File size: 39,303 Bytes
51b3964
 
 
 
 
 
 
31a06e1
 
 
51b3964
0c2e8f6
 
b6c1096
 
 
0c2e8f6
 
51b3964
 
 
 
0c2e8f6
51b3964
 
 
31a06e1
 
 
51b3964
0c2e8f6
 
b6c1096
 
 
0c2e8f6
 
51b3964
 
 
 
0c2e8f6
51b3964
 
 
31a06e1
 
 
51b3964
0c2e8f6
 
b6c1096
 
 
0c2e8f6
 
51b3964
 
 
 
0c2e8f6
51b3964
 
 
31a06e1
 
 
51b3964
0c2e8f6
 
b6c1096
 
 
0c2e8f6
 
51b3964
 
 
 
0c2e8f6
51b3964
 
 
31a06e1
 
 
51b3964
0c2e8f6
 
b6c1096
 
 
0c2e8f6
 
51b3964
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
[
  {
    "case_id": "CS31",
    "chunk_index": 0,
    "section": "overview",
    "location": "Kazungula district, Zambezi River",
    "country": "Zambia",
    "country_normalized": "Zambia",
    "continent": "Africa",
    "admin_area": "Zambezi River",
    "source": "Pipeline Ver2 β€” Kazungula district",
    "source_links": [
      {
        "title": "Ahmed, I., & McEvoy, D. (2014). Post-tsunami resettlement in Sri Lanka and India: site planning, infrastructure and services. International Journal of Disaster Resilience in the Built Environment, 5(1), 53-65. https://doi.org/10.1108/ijdrbe-08-2012-0028",
        "doi": "10.1108/ijdrbe-08-2012-0028",
        "url": "https://doi.org/10.1108/ijdrbe-08-2012-0028"
      }
    ],
    "text": "Case study: Kazungula district (Kazungula district, Zambezi River, Zambia)\nSection: CASE STUDY OVERVIEW\n\n**Case Name and Location:** Flood Resettlement Scheme, Kazungula District, Southern Province, Zambia (along the Zambezi River)\n\n**Hazard(s):** Riverine flooding from the Zambezi River and its tributaries. The district also experiences changing rainfall patterns, including shorter rainy seasons and drought stress affecting agriculture, but flooding was the direct driver of the resettlement scheme.\n\n**Type of Retreat:** State-led flood resettlement scheme β€” a planned relocation initiative for flood-affected rural communities from the Zambezi floodplain to higher ground, driven by the national Disaster Mitigation and Management Unit (DMMU).\n\n**Scale:** The precise number of households or individuals involved in the resettlement scheme is not specified in the reviewed sources. The scheme operated at the district level in a highly rural context and is described as a \"model flood resettlement scheme.\" No household or population figures are available.\n\n**Key Actors:** The primary implementing actor was the DMMU, operating under Zambia's Vice President's Office. The Kazungula District Council (local government) participated initially but became increasingly wary of the scheme. Multilateral donors, including the World Bank-managed Pilot Programme for Climate Resilience (PPCR), provided district-level adaptation funding from approximately 2015. NGOs including CARITAS, Red Cross, Save The Children, WWF, and the African Wildlife Foundation were active in broader adaptation activities in the district. Central government line ministries β€” including the Ministry of Agriculture and the Department of Water Affairs β€” also had adaptation roles.\n\n**Timeline:** The resettlement scheme was a response to a series of major Zambezi River floods throughout the 2000s; the 2005–2006 rainy season is cited as particularly extreme. The primary source (Funder, Mweemba, and Nyambe, based on research conducted 2012–2016) describes the scheme as ongoing at the time of writing. No post-scheme evaluation or completion date is recorded.\n\n**Key Outcomes:** The scheme was publicly framed by the DMMU as a climate change adaptation measure. However, it generated local conflicts and was regarded with growing wariness by the Kazungula District Council due to its top-down nature. PPCR funding that had initially been intended for the DMMU-led flood-resettlement scheme in Kazungula was diverted to other areas after the District Council argued this to a World Bank planning mission. The case illustrates a broader tension between central government control over resettlement and meaningful local government participation in adaptation planning.\n\n---\n\n### Narrative Summary\n\nThe Kazungula district in Zambia's Southern Province sits along the Zambezi River in one of the country's most climate-vulnerable agro-ecological zones. The main ethnic groups β€” Lozi, Tonga, and Toka Leya β€” sustain livelihoods through rain-fed agriculture (maize as the dominant crop), livestock (especially cattle), and artisanal riverine fisheries. Poverty rates are high, with more than 60% of the population in some areas categorized as food-insecure. During the 2000s, a series of extreme Zambezi floods, attributed by the source to climate change, severely disrupted local livelihoods. In the aftermath of particularly extreme flooding in the 2005–2006 rainy season, the national Disaster Mitigation and Management Unit (DMMU) β€” originally charged with coordinating and supporting immediate emergency relief efforts β€” expanded its mandate and began driving the development of a flood resettlement scheme in Kazungula, publicly dubbed a \"climate change adaptation\" scheme. The scheme aimed to relocate flood-affected rural communities from the Zambezi floodplain to higher ground.\n\nThe scheme was characterized by strong centralized control. The Kazungula District Council, the elected local government body, initially participated in and supported activities associated with the scheme β€” including establishing an emergency camp after the 2005–2006 floods β€” but over time became wary of its top-down nature and the local conflicts it generated. The broader institutional environment in Zambia is one in which district councils are under-resourced and in practice subordinate to central government line agencies, despite a formal decentralization policy dating from 2002 (revised 2013) that nominally assigns land-use planning and disaster management functions to district councils. The source suggests that the DMMU's scheme had a longer-term aim of opening territory to government control and subjecting land and production systems to technical interventions by state agencies.\n\nDistrict councils in Kazungula and Sesheke responded to their marginalization through several strategies: actively \"piggy-backing\" on central government adaptation activities to assert legitimacy; seeking alternative alliances with donors and NGOs; and, in one notable instance, successfully lobbying a World Bank planning mission to redirect PPCR funding away from the DMMU-led resettlement scheme toward other district priorities. This redirection of funding represents a rare instance of local government successfully challenging central dominance in adaptation governance. For the communities affected by the scheme, the degree of voluntariness in the relocation is not documented, but the top-down nature of the process and the absence of any documented community consultation suggest that meaningful household-level choice was limited.\n\n---"
  },
  {
    "case_id": "CS31",
    "chunk_index": 1,
    "section": "coding_questions",
    "location": "Kazungula district, Zambezi River",
    "country": "Zambia",
    "country_normalized": "Zambia",
    "continent": "Africa",
    "admin_area": "Zambezi River",
    "source": "Pipeline Ver2 β€” Kazungula district",
    "source_links": [
      {
        "title": "Ahmed, I., & McEvoy, D. (2014). Post-tsunami resettlement in Sri Lanka and India: site planning, infrastructure and services. International Journal of Disaster Resilience in the Built Environment, 5(1), 53-65. https://doi.org/10.1108/ijdrbe-08-2012-0028",
        "doi": "10.1108/ijdrbe-08-2012-0028",
        "url": "https://doi.org/10.1108/ijdrbe-08-2012-0028"
      }
    ],
    "text": "Case study: Kazungula district (Kazungula district, Zambezi River, Zambia)\nSection: CODING QUESTIONS\n\n---\n\n### Is there any evidence of a proactive or reactive trigger in the MR decision process? (Reactive/Proactive)\n\n**Answer:** Primarily **Reactive**, with subsequent proactive framing applied by the DMMU.\n\n**Narrative:** The resettlement scheme was triggered by a series of major Zambezi River floods throughout the 2000s, with the 2005–2006 rainy season cited as particularly extreme. The primary source (Funder et al.) describes the DMMU as originally charged with coordinating and supporting immediate emergency relief and subsequently expanding its mandate toward longer-term resettlement activities. The scheme was therefore fundamentally reactive β€” a response to recurring flood disasters rather than an anticipatory strategy developed in advance of hazard impacts.\n\nHowever, the DMMU subsequently framed the scheme as a \"climate change adaptation\" measure, lending it a proactive narrative presentation. The PPCR programme, which began funding district-level activities including in Kazungula and Sesheke from approximately 2015, represents a more genuinely forward-looking institutional effort, though it did not originate the resettlement scheme.\n\nThe Arnall (2019) comparative article, while describing Mozambique and other African/Asian cases rather than Kazungula specifically, documents a broader pattern in which state-led relocation schemes triggered by floods are subsequently rebranded as proactive climate adaptation β€” a dynamic consistent with the Kazungula case as described in the primary source. This is cited as comparative context, not as direct evidence about Kazungula.\n\n**Coding selected:** Reactive (primary); proactive framing applied post-hoc by the DMMU, but the underlying trigger was reactive.\n\n---\n\n### Date, Number of Households Affected, Number of Population Affected, Total Funding\n\n**Date:** Floods triggering the scheme occurred throughout the 2000s; the 2005–2006 rainy season is cited as particularly extreme. The primary source is based on research conducted 2012–2016; the scheme was described as ongoing at time of writing.\n\n**Number of households affected:** Not specified in the reviewed sources.\n\n**Number of population affected:** Not specified in the reviewed sources.\n\n**Total funding:** Not specified for the resettlement scheme specifically. The PPCR (World Bank-managed Climate Investment Funds) began funding sub-national adaptation activities including in Kazungula and Sesheke from approximately 2015, but no funding amount for the resettlement scheme is provided. Most adaptation activities by government agencies in the district were described as financed through programmes and projects funded by multilateral or bilateral donors such as UNDP, FAO, and the World Bank, but no figures are given.\n\n**Note:** These are significant data gaps. The primary source is an institutional and governance analysis, not a program evaluation, and does not report implementation figures.\n\n---\n\n### What is the occurrence of the hazard(s) considered in the MR decision process? (Single Hazard/Multiple Hazards)\n\n**Answer:** **Single Hazard**\n\n**Narrative:** Riverine flooding from the Zambezi River and its tributaries is the hazard explicitly identified as driving the resettlement scheme. The primary source describes \"a series of extreme floods of the Zambezi and its tributaries throughout the 2000s\" as the direct trigger. The source also notes changing rainfall patterns, including shorter rainy seasons, and drought stress affecting maize production β€” but these are described as broader climate stresses on livelihoods in the district, not as hazards explicitly considered in the retreat decision itself. The DMMU's scheme is consistently referred to as a \"flood resettlement scheme,\" not a multi-hazard resettlement scheme. Drought and rainfall variability are better understood as contextual stressors on an already vulnerable population rather than as co-equal drivers of the retreat decision.\n\n**Coding selected:** Single Hazard (riverine flooding). Note: drought/rainfall variability is a significant broader contextual stressor in the district but is not documented as part of the retreat trigger or decision rationale.\n\n**Alternative interpretation:** One could argue that the compound flood-drought climate stress pattern qualifies as multiple hazards under the codebook's definition of \"two or more hazards impacting the same area.\" However, given that the scheme is specifically called a \"flood resettlement scheme\" and no evidence in the source links the retreat decision to a drought or rainfall-variability assessment, Single Hazard is the more defensible coding.\n\n---\n\n### What is the origin of the hazards considered in the MR decision process? (Technological Hazard/Natural Hazard)\n\n**Answer:** **Natural Hazard**\n\n**Narrative:** The floods are described as natural hazard events. The primary source states they \"have been attributed to climate change,\" meaning climate change is identified as a driver or amplifier of the hazard pattern, but the underlying origin is natural flooding of a river system. There is no evidence of technological or industrial hazard origin (no dam collapse, industrial contamination, or similar). The codebook notes that \"climate change\" as a standalone origin code was discarded from the main analysis; the appropriate coding is Natural Hazard, with the note that the source attributes the flood pattern to climate change.\n\n**Coding selected:** Natural Hazard. Note: source attributes the extreme flood pattern to climate change.\n\n---\n\n### Was this hazard considered in the MR decision process?\n\n**Answer:** **Riverine/Pluvial Flooding**\n\n**Narrative:** The Zambezi River and its tributaries are explicitly identified as the source of flooding driving the resettlement scheme. The primary source refers to \"a series of extreme floods of the Zambezi and its tributaries.\" This is consistent with the codebook definition of Riverine/Pluvial Flooding: \"flooding connected to a river delta, river system, river floodplain, or riparian habitat.\"\n\n**Coding selected:** Riverine/Pluvial Flooding.\n\n**Not selected:** Coastal Flooding/Erosion, Lake Flooding/Erosion, Wildfire, Volcano, Land Subsidence, Landslide/Debris Flow, Earthquake, Tsunami, Hurricane β€” none of these are relevant to this case. \"Other\" (drought/rainfall variability) is a broader contextual stressor, not the retreat hazard.\n\n---\n\n### What was the geographic scale of the MR project? (Individual homes / A few streets / Neighbourhood / Entire community / Sub-watershed / Watershed region)\n\n**Answer:** **Entire community**\n\n**Narrative:** The resettlement scheme operated at the district level, involving flood-affected rural communities across Kazungula district along the Zambezi floodplain. The scheme is not a household-level, street-level, or neighbourhood-level intervention; it involves communities across a rural district. The primary source does not describe the scheme in sub-watershed or watershed-region terms. \"Entire community\" best captures the collective, district-scale character of the scheme.\n\n**Coding selected:** Entire community. Note: the scheme is district-level, encompassing multiple rural communities across Kazungula District, but the source does not frame it as a watershed or sub-watershed intervention.\n\n**Not selected:** Sub-watershed β€” this coding is not supported by the source text, which does not describe the scheme in hydrological planning terms.\n\n---\n\n### Who or what was retreated? (Individuals / Groups/Neighbourhoods / Communities / Critical infrastructure / Hazard protection structures / Industrial assets)\n\n**Answer:** **Communities** (rural farming communities)\n\n**Narrative:** The scheme involved the planned collective relocation of flood-affected rural households β€” smallholder farmers, livestock keepers, and artisanal fisherfolk β€” from the Zambezi floodplain to higher ground. The DMMU drove the development of a flood resettlement scheme described as a \"model\" district-level initiative. This is a community-level retreat, not individual household buyouts or infrastructure-only relocation.\n\n**Coding selected:** Communities.\n\n---\n\n### Other notes (General Information)\n\n- The case is located in the Global South (Sub-Saharan Africa, Zambia).\n- The primary source (Funder, Mweemba, and Nyambe) is a governance and institutional analysis based on research conducted 2012–2016; it is not a program evaluation and does not report implementation metrics (households relocated, funding amounts, or outcomes).\n- The Arnall (2019) article provides comparative context from Mozambique and other African/Asian state-led relocation cases. It is used in this memo as comparative and conceptual context, not as evidence about Kazungula specifically.\n- The Ahmed and McEvoy (2014) article describes post-tsunami resettlement in Sri Lanka and India. It is a different geographic and hazard context; it is not used to infer facts about Kazungula.\n- The term used in the primary source for the scheme is \"flood resettlement scheme\"; the DMMU dubbed it a \"climate change adaptation\" scheme.\n\n---"
  },
  {
    "case_id": "CS31",
    "chunk_index": 2,
    "section": "institutional_environment",
    "location": "Kazungula district, Zambezi River",
    "country": "Zambia",
    "country_normalized": "Zambia",
    "continent": "Africa",
    "admin_area": "Zambezi River",
    "source": "Pipeline Ver2 β€” Kazungula district",
    "source_links": [
      {
        "title": "Ahmed, I., & McEvoy, D. (2014). Post-tsunami resettlement in Sri Lanka and India: site planning, infrastructure and services. International Journal of Disaster Resilience in the Built Environment, 5(1), 53-65. https://doi.org/10.1108/ijdrbe-08-2012-0028",
        "doi": "10.1108/ijdrbe-08-2012-0028",
        "url": "https://doi.org/10.1108/ijdrbe-08-2012-0028"
      }
    ],
    "text": "Case study: Kazungula district (Kazungula district, Zambezi River, Zambia)\nSection: INSTITUTIONAL ENVIRONMENT\n\n---\n\n### What is the focus or scope of the MR project? (Individual buyout/retreat / Community buyout/retreat / Leasebacks, easements and buffers / Other acquisition tools / Managed realignment of infrastructure)\n\n**Answer:** **Community buyout/retreat**\n\n**Narrative:** The scheme involved the collective relocation of rural communities from the Zambezi floodplain to higher-ground resettlement sites. There is no evidence of individual property buyouts, financial compensation mechanisms, easements, or infrastructure realignment. The relocation was state-organized and community-scale. This is consistent with the codebook definition of \"community buyout/retreat\" as a collective relocation scheme, although the degree of community engagement in this case was limited (the scheme was top-down, as discussed throughout this memo). The \"community engagement\" element of the codebook definition was not fulfilled in practice.\n\n**Coding selected:** Community buyout/retreat. Note: collective in organizational form, but with limited community engagement or participation in planning.\n\n**Not selected:** Individual buyout/retreat β€” no evidence of household-level property acquisition. Leasebacks, easements and buffers β€” no evidence. Managed realignment of infrastructure β€” no evidence; the retreat involved people, not infrastructure alone.\n\n---\n\n### What legal constraints or requirements were identified during the MR decision process?\n\n**Answer:** **Legal authority over retreat; Relevant regulations/acts**\n\n**Narrative:** The primary source highlights significant institutional complexity around legal authority over retreat activities. The DMMU was originally charged with coordinating and supporting immediate emergency relief efforts but expanded its mandate to encompass long-term resettlement β€” raising questions about the legal basis for this expanded role. Zambia's decentralization policy (2002, revised 2013) nominally assigns land-use planning and disaster management functions to district councils, but in practice central government line agencies retain control. Throughout the 2015–2017 period covered in the source, devolution remained incomplete and subject to central approval.\n\nThe sources suggest possible legal ambiguity because district councils have formal planning roles under the decentralization policy while land in rural Zambia is largely controlled by the central state (through State Forest Reserves and other mechanisms) or by chiefly structures (communal land). However, the sources do not document the specific legal process used to acquire or designate the resettlement sites, and no litigation, formal land title transfers, rezoning of retreat lands, or compliance and enforcement mechanisms are described.\n\n**Coding selected:** Legal authority over retreat (the DMMU's expansion of its mandate is directly discussed); Relevant regulations/acts (decentralization policy and its tensions with central control are central to the case).\n\n**Not selected:** Litigation, Home acquisition, Land acquisition, Title(s) transfer, Land rezoning, Retreat land development, Compliance & enforcement β€” none of these are documented in the reviewed sources for the Kazungula resettlement scheme specifically.\n\n**Uncertainty:** The reviewed sources provide no detailed legal documentation of the resettlement scheme's legal framework. The legal analysis in the primary source is framed at the institutional and governance level, not as a legal case study.\n\n---\n\n### What were the funding sources for the MR project?\n\n**Answer:** **Multiple funders: Federal (national government/DMMU); International aid (World Bank PPCR and broader multilateral/bilateral donors)**\n\n**Narrative:** The DMMU, as a national government agency under the Vice President's Office, led the resettlement scheme. The primary source states that most adaptation activities by government agencies were \"financed through programmes and projects funded by multilateral or bilateral donors such as the UNDP, FAO and the World Bank.\" The PPCR (World Bank-managed Climate Investment Funds) began funding sub-national adaptation activities including in Kazungula and Sesheke from approximately 2015, though the PPCR funding originally intended for the DMMU-led resettlement scheme was subsequently diverted to other areas.\n\nDistrict councils had a very limited revenue base and were critically dependent on central government budget allocations; they are not identified as funders of the scheme. NGOs were active in the broader district adaptation landscape with internationally sourced funding, but the primary source does not establish that NGO funding supported the resettlement scheme specifically.\n\n**Coding selected:** Multiple funders: Federal (DMMU/national government); International aid (World Bank PPCR, and broader multilateral/bilateral donor financing of government adaptation activities).\n\n**Note:** NGOs are not coded as funders of the resettlement scheme itself; they were present in the broader district adaptation context. Local/Municipal funding was negligible.\n\n---\n\n### What was considered to maintain (or improve) the tax base in the MR decision process?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The sources do not discuss tax base considerations, local revenue implications, or economic development objectives for receiving areas. In rural Zambia, district councils have very limited revenue bases. The primary source focuses on governance and authority dynamics, not fiscal planning.\n\n**Coding selected:** N/A β€” no evidence.\n\n---\n\n### What level/type of compensation mechanisms were considered in the MR decision process?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The primary source (Funder et al.) does not describe any compensation payments, property valuation processes, financial incentives, moving assistance, or other compensation mechanisms provided to relocated households in Kazungula. The scheme appears to have been state-directed without documented compensation mechanisms. The Arnall (2019) article describes, in the context of Mozambique, how the \"promise of new lifestyles in the high area, as exemplified by the provision of modern houses constructed from concrete and corrugated iron rather than traditional materials, undoubtedly acted as a draw for some groups\" β€” but this describes Mozambique, not Kazungula, and cannot be used to infer compensation arrangements in the Zambia case.\n\n**Coding selected:** No evidence. This is a significant data gap.\n\n---"
  },
  {
    "case_id": "CS31",
    "chunk_index": 3,
    "section": "planning_processes",
    "location": "Kazungula district, Zambezi River",
    "country": "Zambia",
    "country_normalized": "Zambia",
    "continent": "Africa",
    "admin_area": "Zambezi River",
    "source": "Pipeline Ver2 β€” Kazungula district",
    "source_links": [
      {
        "title": "Ahmed, I., & McEvoy, D. (2014). Post-tsunami resettlement in Sri Lanka and India: site planning, infrastructure and services. International Journal of Disaster Resilience in the Built Environment, 5(1), 53-65. https://doi.org/10.1108/ijdrbe-08-2012-0028",
        "doi": "10.1108/ijdrbe-08-2012-0028",
        "url": "https://doi.org/10.1108/ijdrbe-08-2012-0028"
      }
    ],
    "text": "Case study: Kazungula district (Kazungula district, Zambezi River, Zambia)\nSection: PLANNING PROCESSES\n\n---\n\n### What is the time horizon considered in the MR decision process?\n\n**Answer:** **Immediate risks (1–5 years)**\n\n**Narrative:** The resettlement scheme was driven primarily by immediate flood disaster response. The DMMU responded to the 2005–2006 floods and subsequent flood events and expanded into longer-term resettlement, but the planning horizon for the scheme itself appears to have been short-term and reactive. There is no evidence in the reviewed sources of long-range climate projection modelling, multi-decadal planning, or consideration of anticipated future risks (10–100 years) in the design of the Kazungula resettlement scheme.\n\nThe PPCR programme introduced some forward-looking climate-screening of district development plans from approximately 2015, but this was a supplementary and later-arriving initiative, not the basis of the resettlement scheme. No post-retreat follow-up activities or evaluation frameworks are described.\n\n**Coding selected:** Immediate risks (1–5 years).\n\n**Not selected:** Pre-retreat planning β€” no evidence of formal pre-retreat planning processes (risk mapping, vulnerability analysis, site evaluation) for the scheme. Emerging risks (5–10 years), Anticipated future risks (10–100 years), Post-retreat follow-up β€” no evidence for any of these time horizons in relation to the scheme.\n\n**Note:** The sources do not specify whether relocation was implemented in a single stage or phased manner.\n\n---\n\n### What planning elements were considered in the MR decision process?\n\n**Answer:** **Top-down; Multi-sectoral (limited, state agencies only)**\n\n**Narrative:** The planning process was explicitly top-down, driven by the DMMU with minimal meaningful local government or community input. The primary source states directly that the Kazungula District Council \"eventually became wary of their top-down nature and the local conflicts they have generated.\" There is no evidence of bottom-up or participatory planning, multi-criteria decision analysis (MCDA), adaptive management, cost-benefit analysis (CBA), degree of transparency, public-private partnerships, or bureaucracy reduction efforts.\n\nSome multi-sectoral coordination occurred β€” the DMMU worked alongside the Ministry of Agriculture and the Department of Water Affairs on \"associated land and agricultural development activities\" β€” but this was coordination among central government agencies rather than genuine multi-sectoral planning involving diverse governance levels or community actors. The PPCR example cited in the source illustrates that even relatively progressive donor programmes were locked into conventional planning and decision-making structures requiring multiple approval layers, suggesting bureaucracy expansion rather than reduction.\n\n**Coding selected:** Top-down; Multi-sectoral (limited β€” state agency coordination only).\n\n**Not selected:** Bottom-up, Participatory/collaborative, Adaptive management, MCDA, CBA, Degree of transparency, Public private partnerships, Bureaucracy reduction, Single stage retreat, Multi-stage retreat β€” no evidence for any of these. The sources do not specify whether the scheme was single-stage or multi-stage.\n\n---\n\n### What decision support tools were used in the MR decision process?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The primary source does not describe any decision support tools used in planning the Kazungula resettlement scheme. No hazard mapping, climate projection modelling, social vulnerability assessments, cost-benefit analysis, demographic projections, housing assessments, or new site evaluation processes are mentioned. The PPCR programme introduced climate-screening tools for district development plans from approximately 2015 β€” but this was a later and supplementary initiative, not the decision support basis for the resettlement scheme.\n\nThe Ahmed and McEvoy (2014) article, describing post-tsunami resettlement in Sri Lanka and India, notes that \"very little site planning guidelines were available specifically for resettlement programmes\" in those contexts β€” but this observation pertains to Sri Lanka and India and cannot be used to characterize the Kazungula case.\n\n**Coding selected:** No evidence.\n\n---\n\n### What planning measures were taken for receiving areas in the MR decision process?\n\n**Answer:** No evidence found in reviewed sources, beyond the general selection of higher ground.\n\n**Narrative:** The primary source describes the scheme as relocating communities from the Zambezi floodplain to \"higher ground,\" indicating that some site selection was involved (higher-ground sites were identified as less flood-prone). However, no detailed receiving-area planning process is described. There is no evidence of assessments of existing housing availability, social carrying capacity analysis, construction of new housing, infrastructure planning for receiving sites, or evaluation of whether receiving sites were appropriate for agricultural livelihoods. No site planning or receiving-area development documentation is cited in the primary source.\n\n**Coding selected:** At most, weak implicit selection of priority areas (higher ground chosen over floodplain); beyond this, no evidence.\n\n**Not selected:** Availability of existing housing, Construction of new housing, Social carrying capacity β€” no evidence for any of these in the reviewed sources.\n\n---\n\n### What uses were planned for retreat lands?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The primary source does not describe post-retreat land use planning for the Zambezi floodplain areas from which households were relocated. The source states that the DMMU's scheme had \"the longer-term aim of opening up territory to government control and making land and production systems subject to technical interventions by state agencies\" β€” but this describes a political aim regarding territory and agricultural governance, not a specific land use plan for vacated floodplain lands. No rezoning, natural use designation, demolition, infrastructure removal, conservation trust, or ownership transfer arrangements are documented.\n\n**Coding selected:** No evidence.\n\n---"
  },
  {
    "case_id": "CS31",
    "chunk_index": 4,
    "section": "community_engagement",
    "location": "Kazungula district, Zambezi River",
    "country": "Zambia",
    "country_normalized": "Zambia",
    "continent": "Africa",
    "admin_area": "Zambezi River",
    "source": "Pipeline Ver2 β€” Kazungula district",
    "source_links": [
      {
        "title": "Ahmed, I., & McEvoy, D. (2014). Post-tsunami resettlement in Sri Lanka and India: site planning, infrastructure and services. International Journal of Disaster Resilience in the Built Environment, 5(1), 53-65. https://doi.org/10.1108/ijdrbe-08-2012-0028",
        "doi": "10.1108/ijdrbe-08-2012-0028",
        "url": "https://doi.org/10.1108/ijdrbe-08-2012-0028"
      }
    ],
    "text": "Case study: Kazungula district (Kazungula district, Zambezi River, Zambia)\nSection: COMMUNITY ENGAGEMENT\n\n---\n\n### What degree of choice was provided to affected people in the MR decision process?\n\n**Answer:** **Involuntary with some degree of opt out** (most defensible given available evidence) β€” though this remains uncertain.\n\n**Narrative:** The primary source does not directly document the degree of voluntariness for relocated households. However, the scheme is characterized throughout as top-down and as generating local conflicts, with no mention of voluntary participation mechanisms, community consultation processes, or household-level choice procedures. This strongly implies that the retreat was not fully voluntary.\n\nThe Arnall (2019) comparative article, drawing on Mozambique and other African/Asian cases, documents that many state-led relocation schemes \"operate within a 'grey area' between voluntariness and involuntariness\" and that programmes claiming to be voluntary often involve \"'induced' or 'compulsory' volunteerism β€” where officials promote resettlement as beneficial while also making clear that not resettling was not an option.\" This is cited as comparative conceptual context, not as direct evidence about Kazungula. The article also notes that \"the voluntary-involuntary framework fails to adequately capture the complexity of the processes involved in state relocation.\"\n\n**Coding selected:** Involuntary with some degree of opt out β€” selected as the most defensible coding given the top-down, state-directed character of the scheme, while acknowledging that direct evidence of compulsion or opt-out mechanisms is not documented in the reviewed sources.\n\n**Uncertainty:** The reviewed sources do not directly document the household-level experience of choice or coercion in the Kazungula scheme. This remains a significant information gap.\n\n---\n\n### What land user groups were involved or considered in the MR decision process?\n\n**Answer:** **Community/NGO; Government**\n\n**Narrative:** The affected population are described as small-scale farmers, livestock keepers, and artisanal fisherfolk in Kazungula district. In rural Zambia, land is largely controlled either by the central state (through State Forest Reserves and similar mechanisms) or by chiefly structures on communal terms; formal individual property ownership is not described. Chieftaincy structures are noted as significant actors who \"directly contest the legitimacy of the central state.\" NGOs were active in the area. Government agencies (DMMU, line ministries, district councils) were central actors.\n\n**Coding selected:** Community/NGO (affected rural communities and NGOs active in the area); Government (DMMU, line ministries, district councils).\n\n**Note on \"Homeowners\":** In rural Zambia's communal land tenure context, \"homeowners\" in the formal property-title sense is not applicable. Residents hold customary/communal land rights rather than formal title. The codebook \"Homeowners\" category is not applied to avoid misleading implications about formal property ownership.\n\n**Not selected:** Homeowners (not applicable given communal land tenure), Leased land owners, Renters, Commercial β€” no evidence for any of these categories in the Kazungula context. First Nations/Indigenous β€” see socio-cultural section below.\n\n---\n\n### What actors participated in the governance structure of the MR decision process? Is there evidence that they shaped the MR outcomes?\n\n**Answer:** **Funders** (World Bank/PPCR, broader multilateral/bilateral donors); **Government agencies** (DMMU, Ministry of Agriculture, Department of Water Affairs, Kazungula District Council); **NGOs/CBOs** (active in parallel adaptation activities); **Political champions** (district councillors who challenged DMMU authority and lobbied World Bank)\n\n**Narrative:** The DMMU led the scheme. The District Council participated but in an increasingly resistant role. Donors (World Bank PPCR) provided funding and convened planning missions. NGOs implemented parallel adaptation projects. The affected community is not documented as a meaningful actor in the governance structure of the scheme.\n\n**Evidence of shaping outcomes:** The Kazungula District Council successfully argued to a World Bank planning mission that PPCR funding was better spent elsewhere, resulting in that funding being diverted away from the DMMU resettlement scheme β€” a clear instance of a local governance actor shaping MR outcomes. Conversely, the DMMU shaped outcomes by expanding its mandate beyond emergency relief to drive the resettlement scheme. These are directly supported by the primary source.\n\n**Coding selected:** Funders; Government agencies; NGOs/CBOs; Political champions (district councillors).\n\n**Not selected:** Affected community (not meaningfully included in governance), Receiving communities (not documented), Private sector (no evidence), First Nations/Indigenous communities (not applicable as a formal governance category), Academics/researchers (the research authors conducted fieldwork but are not described as governance actors in the program β€” their role was research and analysis, not program governance).\n\n---\n\n### Who ultimately made decisions in the MR decision process?\n\n**Answer:** **Government** (national, through DMMU)\n\n**Narrative:** Decision-making authority in the resettlement scheme resided with the DMMU as the central government agency. While the District Council succeeded in redirecting some PPCR funding away from the scheme, the scheme itself was planned and driven by the DMMU. Property owners/affected households, community representatives, and financial institutions are not documented as having decision-making roles in the scheme.\n\n**Coding selected:** Government (national/central β€” DMMU).\n\n---\n\n### Was this engagement approach explicitly incorporated into the MR decision process?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The primary source does not describe any structured engagement approaches incorporated into the Kazungula resettlement scheme β€” no visualization tools, community consultation processes, transparency mechanisms, ongoing public updates, or participatory mapping. Meaningful affected-community engagement is not documented; the scheme is characterized as top-down. The PPCR programme involved some community-level activities and district council training from approximately 2015, but these were supplementary and later-arriving, not part of the resettlement scheme's design.\n\n**Coding selected:** No evidence for any engagement approach categories. This is a notable gap.\n\n---\n\n### What community resources supported the transition during the MR process?\n\n**Answer:** No evidence found in reviewed sources.\n\n**Narrative:** The primary source does not describe caseworkers, temporary housing advisors, relocation assistance programs, community capacity building, or livelihood restoration support for relocated households in Kazungula. This is a significant information gap. Arnall (2019) describes income diversification projects provided post-relocation in the Mozambican case, but this cannot be applied to Kazungula.\n\n**Coding selected:** N/A β€” no evidence.\n\n---\n\n### What supports were provided"
  }
]