SMVaddadi/propensity-data-dot-airfare

dot_airfare — mirror

A convenience mirror of a public dataset, pushed by propensity.data.hf_mirror so that benchmark runs do not depend on a flaky upstream origin.

This is a mirror, not a grant of rights

Copyright and licence remain with the original source named below. Mirroring moves bytes; it does not and cannot change the terms they arrived under. If you intend to redistribute this further, or to use it commercially, read the original terms — not this card, and not the license: field above, which is our best reading and may be wrong.

Provenance

• Source: Consumer Airfare Report, U.S. Department of Transportation, Office of the Assistant Secretary for Aviation and International Affairs, via datahub.transportation.gov. Public domain as a matter of US law: 17 U.S.C. section 105 denies copyright protection to works of the United States Government. Declared as license: other + license_name: us-public-domain rather than cc0-1.0, because CC0 is a rights waiver only a rights holder can execute and this repository is not the rights holder. NOT INSURANCE DATA, and that matters: this collection is otherwise insurance regulatory returns, and nothing in this file is an insurance outcome. It is here as the PRICING comparator -- an administratively-collected quarterly price panel with a long time axis, a clean entity axis, and a quantity (passengers) beside every price, which is what a price-sensitivity control needs. Treating it as an insurance outcome would be a category error. WHAT THIS IS: three arms. table_1a 262,087 rows x 23 published columns, city-market pair x quarter, 126 quarters 1993-2026, 1,332 city-market pairs and 4,214 airport pairs. table_5 15,481 x 21, carrier x market, 121 quarters 1996-2026. table_7 24,126 x 23, airport x quarter, 121 quarters 1996-2026. THREE FARE ESTIMANDS ON ONE ROW in table_1a, which is the subtlety: fare is the market average, fare_lg the average of the LARGEST carrier (share large_ms), fare_low that of the LOWEST-FARE carrier (share lf_ms). label_columns names all three; 'the fare' is ambiguous here. TWO MEASURED CAVEATS. (1) The published tbl label changes spelling mid-series: 'Table 1a' covers 1993-2018 (200,897 rows) and 'Table1a' with no space covers 2018-2026 (61,190 rows). A groupby on the raw column reports two tables where there is one and splits the panel at 2018; this mirror normalises it and keeps the published string in source_table_label. (2) table_7's Total* columns are the exact arithmetic SUM of their short-haul and long-haul components -- TotalMkts == SHMkts + LHMkts and TotalFaredPax == SHPax + LHPax on 23,485 of 24,126 rows, max absolute difference 0.0000, with all 641 exceptions being rows where one haul type is absent. They are derived and leak if fed beside their components; they are named in DERIVED_TOTAL_COLUMNS and droppable in one argument. Checked and CLEAN: no aggregate pseudo-entities -- no city or carrier value matches total/all/average/national, unlike the sibling OSFI and OECD mirrors where aggregates sit in the entity column. WHAT THIS IS NOT: not microdata (every row is already a quarterly market average, so there is no personal-data question at all), not insurance, and not a grant of rights by this repository.
• Origin URL: https://www.transportation.gov/policy/aviation-policy/competition-data-analysis/research-reports
• Loader: propensity.data.public_loaders.load_dot_airfare
• Extract: full pull from origin
• Reported source: dot_airfare
• Rows pushed: 262,087
• Columns: 26

Licence and redistribution

• Licence as found: other (us-public-domain)
• Redistribution verdict: open
• Basis: US Department of Transportation, Office of the Assistant Secretary for Aviation and International Affairs. PUBLIC DOMAIN BY STATUTE: 17 U.S.C. section 105 denies copyright protection to any work of the United States Government, and the Consumer Airfare Report is such a work. That is the load-bearing basis and it does not depend on any page being reachable. CORROBORATION, read 2026-11-09 from each Socrata view's own /api/views/.json on datahub.transportation.gov (views tfrh-tu9e, bkh6-tj42, d6nc-s8v6): licenseId USGOV_WORKS, license {"name": "Public Domain U.S. Government", "termsLink": "https://www.usa.gov/government-works"}, attribution "Department of Transportation Office of the Assistant Secretary for Aviation and International Affairs", provenance official. RECORDED HONESTLY: that termsLink target is NOT READABLE -- https://www.usa.gov/government-works returned HTTP 200 with 521 bytes and 115 characters of readable text, i.e. a stub, not terms. So the link is corroboration that could not be checked. It does not matter here because the statute is the basis; it WOULD matter for a US state body, which has no equivalent statutory position -- see the ct_doi_complaints reasoning. Card uses other + us-public-domain, never cc0-1.0: CC0 is a waiver only a rights holder can execute and this repo is not one. No personal data: every row is already a quarterly average over a market; there is no ticket, passenger or itinerary record.

This is US-government material and therefore in the public domain as a matter of US law. The frontmatter says other rather than cc0-1.0 on purpose: CC0 is a waiver that only a rights holder can execute, and we are not the rights holder. Claiming it would be the same error that makes much of this Hub's licence metadata untrustworthy.

Citation

U.S. Department of Transportation, Office of the Assistant Secretary for Aviation and International Affairs. Consumer Airfare Report. Work of the United States Government, public domain under 17 U.S.C. § 105.

Loading

from propensity.data.hf_mirror import load_mirrored

result = load_mirrored("dot_airfare", org="SMVaddadi")
result.frame.head()

load_mirrored falls back to the original loader (origin, then its own offline replica) if this mirror is unreachable, so it is safe to depend on.