3 Conceptual framework
This chapter presents the conceptual framework this thesis will use to understand the National Adaptation Plans. As we saw in Chapter 2, climate adaptation is being mainstreamed into development discourse. This makes it natural to look into the literature on development. This chapter will contrast the institutional and critical approaches in the three themes of development studies, and thus the NAPs: implementation, financing, and knowledge production.
The first section presents the debates on how development projects interact with realities on the ground. The second section discusses how these projects are, and should be, financed. The third section examines the competing knowledge production paradigms. The last section presents some of the recent research on climate adaptation, focused on Paprocki (2018) and Dewan (2022)’s work in Bangladesh, and synthesizes the concepts into the adaptation nexus, the institutional approach, and the adaptation regime, the critical approach.
The institutional approach to development begins with a genuine response to real problems. Climate change is set to bring suffering to people across the Global South and are not invented concerns, but rather daily realities for billions of people. The development apparatus come to this work with sincere commitment to improving lives. They bring technical expertise, financial resources, and accumulated knowledge from other contexts. Still, scholars argue that climate adaptation might actually create more vulnerability than it removes, and simply shift it between groups.
3.1 Implementation
It is in the implementation development theories meet local realities and local peoples. This section reviews the difficulties development projects have with reaching the goals they have set for themselves.
Early development efforts focused on large-scale infrastructure and modernization. When these top-down approaches produced disappointing results, the sector underwent significant self-reflection (Escobar 1995). By identifying the most efficient interventions, often through statistical testing (Banerjee and Duflo 2019), the projects could reach their goals with as few funds as possible. This would again free up funds for new projects. This discussion of aid effectiveness is very prominent, and shapes how institutional development projects are implemented (Banks, Hulme, and Edwards 2015). Through standardized frameworks, such as the logical framework approach (LFA), the projects could be developed, monitored and evaluated, to find what worked and could be used in other contexts (Banks, Hulme, and Edwards 2015).
Participatory development emerged in the 1980s and 1990s as a critique of the top-down development at the time, promising to revolutionize practice by incorporating local knowledge, empowering communities, and ensuring interventions respond to actual needs rather than imposed blueprints (Chambers 1994, 1438). Communities would gather information about their own situation, identify their priorities, and take action themselves.
Chambers identifies three basic components of Participatory Rural Appraisal (PRA): methods, behavior/attitudes, and sharing (Chambers 1994, 1438). The behavioral shift proved especially significant. Outsiders had to step off their pedestals, sit down, hand over the stick, and listen and learn. Through field learning experiences, workshops, and villager-to-villager training, PRA methods spread across continents (Chambers 1994, 1438). Local people demonstrated they could map, model, rank, score, estimate, diagram and analyze, better than outsiders.
What began as radical critique of expert-driven development became a standardized toolkit applied everywhere. Chambers (1994) warned about formalism as a long-term danger. The urge to standardize and codify, often in the name of quality, leads to manuals and fixed procedures (1994, 1441). Participatory exercises, originally designed to empower communities, became efficient methods for gathering information required by project frameworks. The radical promise of participation, redistributing power and resources, gave way to participation as technique: stakeholder consultations, focus groups, participatory mapping (Banks, Hulme, and Edwards 2015).
The critical approach has focused on how development interventions interact with local realities. Ferguson (1994)’s analysis of development in Lesotho argues that development operates as an anti-politics machine. Development agencies create knowledge that about places and peoples that might have little in common with the realities on the ground, but matches the projects that the development agency is able to offer (1994, 20–21). In Lesotho, he argued, the fact that the economy was tightly integrated in the South African mining industry, was strategically overlooked, as it was out of reach of the project. Rather, the focus became overgrazing and establishing markets for the sale of cattle. As with a flick of a button, like an anti-gravity machine removes gravity, development removes politics (Ferguson 1994, 256).
By removing politics, questions of land, resources, jobs, and wages become technical problems, that require technical interventions, rather than political issues requiring political solutions. This depoliticization is not just a side-effect, but central to the projects. While appearing to fail in their stated objectives, development projects succeed in expanding bureaucratic state power and restructuring social relations. Ferguson observes that the most important political effects of planned interventions may occur unconsciously, behind the backs or against the wills of planners (Ferguson 1994, 47). Establishing markets, building roads and outposts are all expansions of state power, despite the development targets not being met. Plans are always important, but never in quite the way planners imagine (Ferguson 1994, 48).
Li (2007) extends this analysis through her examination of development in Indonesia. She identifies two key practices that translate the will to improve into programs. First is problematization: identifying deficiencies needing rectification. Second is rendering technical, representing the domain to be governed as an intelligible field with specific boundaries and characteristics (2007, 7–8). These practices are inseparable. Problem identification links tightly to available solutions. They emerge together where certain diagnoses, prescriptions, and techniques are available to properly trained experts.
Questions rendered technical are rendered nonpolitical. Experts focus on the capacities of the poor rather than practices through which one group impoverishes another (Li 2007, 10). The local thus becomes a site of intervention, but not a source of political analysis. Development projects identify local problems, local needs, local capacities, but rarely examine how local vulnerability connects to broader political and economic structures. This localization of both problems and solutions allows development to proceed without confronting uncomfortable questions about global inequality or structural violence.
Implementation in practice involves complex processes of translation and brokerage that official accounts might miss (Bierschenk, Chauveau, and de Sardan 2002). Development professionals must navigate between different worlds of meaning. Mosse (2011) show how development brokers actively build social, political, and economic roles rather than simply following scripts (2006, 14). They are not passive recipients, but active agents reconciling competing worldviews and interests.
Dewan (2022)’s ethnography in Bangladesh provides insight into this brokerage work. Development brokers engage in strategic code-switching between official narratives and contextual realities, maintaining donor-facing scripts in English while preserving contextual knowledge in Bangla (2022, 547). Mosse (2011) argues that these contradictions development professionals face are standard. They must secure their place within complex institutional contexts, managing relationships, negotiating positions, building networks, dealing with contingency and compromise, producing viable data, meeting targets and budgets. Simultaneously, they must present themselves as bearers of transferable knowledge and universal expertise. Professional status requires transforming messy particulars into clean universals, extracting technical knowledge from the relationships that produce it (Dewan 2022; Bierschenk, Chauveau, and de Sardan 2002).
Participatory development creates an additional paradox. Development workers must hide their expertise and agency to maintain their identity as mere facilitators of community action (Mosse 2011). One Indian community worker insisted his contribution was nothing because he only facilitated and mobilized the community who held the real power (Mosse 2011). When being an expert means pretending not to be one, when action requires apparent inaction, professional identity becomes increasingly difficult to maintain.
Mosse (2011), Bierschenk, Chauveau, and de Sardan (2002) and Dewan (2022) thus all argue that development professionals, being its own profession, are skilled at donor management. By learning each agency’s preferences, buzzwords, and cycles, they develop projects that can attract funding from multiple sources, carefully packaging the same activities in different ways for different audiences. The time and expertise required for this financial diplomacy diverts resources from actual development work. As one project manager explained, success means “maintaining relationships with donors and the government” rather than achieving development outcomes (Mosse 2011).
Scott (1998) argues that the simplification, and thus rendering technical, is not a byproduct of interventions, but central to government itself. The result is radically simplified designs for social organization that inevitably fail because they cannot accommodate the complexity of actual social life (Scott 1998, 7). Crucially, Scott highlights the role of practical knowledge, often also referred to as vernacular knowledge (Escobar 2018), that comes only from experience, which development schemes systematically ignore (Scott 1998, 6). These high-modernist development schemes fail precisely because they attempt to replace practical, local knowledge with abstract, standardized, technical knowledge. This is a practice that Slaughter (2024) argues still is central to thinking today, especially with the belief in data and computer science, and artificial intelligence, to solve global issues.
3.2 Financing
This section explores how development financing is based on a system of conditions, debt, and how financial mechanisms that shape what kinds of development gets funded and thus become possible.
While the section above dealt with the implementation of projects, the financing for these projects exists for itself, in the global aid system. All aid is accounted in the Organisation for Economic Co-operation and Development (OECD) dataset for Official Development Assistance (ODA). The OECD Development Assistance Committee (DAC) maintains the taxonomies for the data, and makes it available through the Creditor Reporting System (CRS). This system is changed regularly, to better reflect the knowledge and policies of the OECD (WP-STAT 2022, 2020b, 2020a). While the system is centralized, aid donors are responsible for reporting their own aid.
The data is then used for either research or for donor countries reporting. It is through these taxonomies that the money pledged at COP (as discussed in Chapter 2) is accounted. Here, the dollar amount could be written as loans, grants, or grant equivalent loans (loans given at a lower-than-market rate) and through NGOs, states or global funds (UN 2024a). Yet, over 70% of flows are from a donor state to a donor country, as grants, although around half is implemented by actors other than the government (UN 2024a).
The financing gap is the starting point for development aid (Bernards 2024, 88). Poor countries need finance they do not have: money for roads, schools, hospitals, agricultural programs. As we will explore more below, this financing gap could be calculated separate from implementation, creating a structure where implementation, financing, and knowledge production are handled by different actors. The institutional approach sees this financing architecture as neutral plumbing for moving money where it’s needed and most effective. They’ve shifted from advocacy to service delivery, becoming contractors implementing donor projects rather than representatives of grassroots movements (Banks, Hulme, and Edwards 2015, 710).
Development finance now increasingly happens through complex financial instruments: blended finance, green bonds, public-private partnerships, that promise to mobilize private capital for development goals (UN 2024a; CPI 2025). This financialization risk deepening, rather than resolving, dependencies (Bernards 2024, 94). Developing countries must offer increasingly favorable terms to attract private investment, including guarantees that transfer risk from private investors to public institutions.
Public-private partnerships (PPPs) exemplify these new financing mechanisms. The PPPs are often presented as as win-win arrangements that combine public purpose with private efficiency (Asian Development Bank 2008, 9). More recently though, they have been reimagined as an easier way of getting private capital involved in development (Bernards 2024, 98). Governments provide guarantees, tax breaks, and regulatory concessions to attract private investment, effectively subsidizing private profit with public resources. When projects fail, as they might, public institutions bear the costs while private partners are protected through complex contractual arrangements (Bernards 2024, 98). A third of the 100 billion dollars pledged to climate adaptation was to come from private funds, and only half of it actually materialized (Bernards 2024, 92).
A central point of discussion is thus the role of the state, or the developing state specifically. With the post-World War II international order established the sovereign nation-state as the fundamental unit of international relations and development, decolonization created new independent states that were expected to pursue their own developmental paths (Mamdani 1996). The institutional approach to development took this sovereignty as given, treating states as autonomous actors capable of making policy choices that would determine their futures. International cooperation would facilitate technology transfer and capacity building between sovereign equals, enabling newly independent states to modernize and eventually join the ranks of developed countries (Escobar 1995, 4).
This formal sovereignty did not address the history of colonialism, and scholars argue that the new states were a continuation of colonial relationships in new forms. As Rodney (1989) argues, the very process of creating independent African states occurred within a framework that ensured continued exploitation (Rodney 1989, 13). The wealth created by labor and resources in the Global South continued to be grabbed by capitalist countries, while restrictions were placed upon capacity to make maximum use of economic potential. Rodney (1989) named the newly independent nation leaders “African sell-outs” (1989, 36). They became part of the definition of underdevelopment, local elites whose interests aligned with metropolitan capital rather than national development. Political independence without economic power meant that real decision-making remained centered outside national boundaries (Rodney 1989, 34).
Mamdani (1996)’s analysis of the bifurcated state reveals how structural dependency operates through inherited colonial institutions (Mamdani 1996, 18). Post-independence African states maintained fundamental structures of colonial governance. The state is both present, in its ability to enforce laws and regulations through its bureaucratic and coercive apparatus, and too absent, in its failure to provide development or represent popular interests. This institutional inheritance means that states, even with local leadership, might function more as mechanisms for managing populations rather than representing them (Mamdani 1996).
As the sovereign power, the state is also supposed to provide security (Figueroa-Helland and Borg 2014). The process in what is deemed a security issue is described by Wæver (1995) as securitization. In this process, an issue is transformed into a security issue through language (Wæver 1995, 48). And when that happens, the issue is transferred from society and into the scope of Government (Wæver 1995, 52). Securitization is thus not simply that an issue is important or produces vulnerability, but that it has been transferred to the state to deal with it. Insecurity is a securitized issue, without a security solution. Security is when the issue has a security solution (Wæver 1995, 52).
3.3 Knowledge Production
This section presents the competing knowledge production paradigms. As the future is uncertain, it contrasts the institutional view of expert knowledge to navigate this, with critical perspectives on how development knowledge override other ways of knowing.
The future is uncertain. These questions, a central issue in development, involve interconnected systems, long time horizons, and radical uncertainty (De Jong et al. 2025). The institutional approach argues that expert knowledge, properly gathered and applied, can help societies navigate these uncertainties and make better decisions about their futures. As the World Bank argues, technical expertise can help manage this uncertainty through systematic analysis, risk assessment, and evidence from other contexts (WB 2021). By using tools for thinking about the future: scenario planning, probabilistic forecasting, sensitivity analysis. Without these tools, societies might be paralyzed by uncertainty or make choices they later regret.
International institutions serve as repositories of global experience. The World Bank positions itself as a “knowledge bank” that collects lessons from development interventions worldwide (WB 2021; De Francesco and Guaschino 2020; Mizuno and Okano 2024). This accumulated knowledge, properly synthesized and transferred, promises to help countries avoid repeating mistakes and adopt proven solutions. The alternative, learning only through trial and error, seems wasteful and unnecessary when knowledge already exists. Technical expertise also promises to overcome the limitations of local knowledge. Communities understand their immediate contexts deeply, but may lack information about broader systems.
The institutional approach emphasizes evidence-based policymaking (De Francesco and Guaschino 2020). Rather than relying on ideology or intuition, development decisions should follow rigorous analysis. Randomized controlled trials can identify what interventions actually work (Banerjee and Duflo 2009). Cost-benefit analysis can determine optimal resource allocation. Monitoring and evaluation systems can track progress and enable course correction (Örtengren 2004).
The production of development knowledge occurs through what Haas (1992) calls “epistemic communities,” networks of professionals with recognized expertise and authoritative claims to policy-relevant knowledge (Haas 1992, 3). These communities share beliefs, validity tests, and policy projects. They define problems in ways their expertise can address (Haas 1992). The development professional might be understood as their own epistemic community (De Francesco and Guaschino 2020). Policy documents result from complex negotiations between team members with different disciplinary backgrounds, institutional pressures, and career incentives (Mosse 2011). What appears as technical analysis often reflects political compromises, like the outcome documents from the COPs discussed earlier Chapter 2. Statistics that seem objective emerge from hurried fieldwork, strategic assumptions, and creative accounting. The “will to believe” in development, the need to maintain institutional legitimacy and career prospects, shapes what knowledge gets produced and validated (Mosse 2011).
Chambers’ work on Participatory Rural Appraisal revealed both the promise and limits of changing who produces knowledge. PRA methods showed that rural people could conduct sophisticated analyses of their situations, often revealing insights that outside experts missed (Chambers 1994). Communities created detailed maps, seasonal calendars, wealth rankings, and problem trees. Yet as Chambers (1994) recognized, PRA could become another standardized toolkit, a way of extracting local knowledge for external use rather than genuinely shifting power over knowledge production (1994, 1442).
Yet as discussed above, development practice and expert knowledge operates through specific mechanisms that shape the results. The post-structuralist critique, beginning with Ferguson’s analysis of development in Lesotho, demonstrates how development discourse transforms political questions into technical problems requiring expert solutions (Ferguson 1994; Li 2007; Scott 1998). Through the anti-politics machine, complex histories of dispossession might become “lack of human capital.”, while political conflicts over resources become “weak property rights.” Structural inequality becomes “market failure.” This rendering technical, as Li (2007) describes it, extracts problems from their political contexts and inserts them into frameworks where expert knowledge can be applied (Li 2007, 7).
The future orientation of climate adaptation and development makes imagining the future a central part of development (De Jong et al. 2025). The interest in more holistic future planning has also reached the global stage, with the recent Futures summit and Pact for the future (UN 2024b). Inayatullah (1990) distinguishes between predictive epistemologies that assume the future can be known through scientific methods, interpretive approaches that see multiple possible futures, and critical perspectives that examine how power shapes which futures can be imagined (Inayatullah 1990, 116).
Development planning relies on predictive epistemologies, such as economic projections, demographic forecasts and climate models. These tools don’t simply predict the future but actively construct it by making certain pathways appear inevitable and others unthinkable (Inayatullah 2013; Godhe and Goode 2018). This inability to imagine other futures is what Goode and Godhe (2017) argues is “capitalist realism”. These technical tools embed particular assumptions about what matters, whose knowledge counts, and what changes are possible.
Santos (2016) extends the idea of capitalist realism to advance the concept of “epistemicide” to describe how certain ways of knowing are not simply marginalized but actively destroyed (Santos 2016). Western modernity operates through what he calls “abyssal thinking” that renders certain forms of knowledge as non-existent (2016, 177). On one side lies science, expertise, and rational planning; on the other, traditional knowledge, local practices, and alternative cosmologies dismissed as belief or superstition. This is not merely exclusion but active destruction, as communities internalize the superiority of Western knowledge and abandon their own ways of knowing (Santos 2016).
By constructing countries as “underdeveloped”, Escobar (1995) argues that development creates the very reality it claims to describe. Countries learn to see themselves through categories that development institutions recognize: GDP per capita, human development indices, governance indicators (1995, 5). These categories don’t simply describe reality but actively construct it. A country becomes “least developed” not through some natural process but through classification systems that privilege certain measures over others. Once classified, countries must present themselves through these categories to access resources, reinforcing the reality that development discourse creates. Their futures are locked in, as they end goal is for them to become like the industrialized countries in the Global North (Escobar 1995).
Escobar (2018)’s concept of the “pluriverse”, offers a framework for understanding an opposition to the abyssal thinking (Escobar 2018). The key idea is that different communities inhabit different worlds constituted by distinct relationships between humans, non-humans, and territories. The “One-World World,” a singular reality defined by Western modernity that eliminates other ways of being and knowing is the most powerful. Development, even when it claims to incorporate local knowledge, operates within this singular world, unable to escaping the abyssal thinking (Escobar 2018). So to counter an epistemicide by the One-World World, to create a “World where many worlds fit” (Escobar 2018), is to construct a pluriverse.
3.4 The Adaptation Nexus and the Adaptation Regime
This section synthesizes recent climate adaptation research to present two frameworks. The adaptation nexus sees integration of climate considerations into development as necessary and beneficial, while the adaptation regime critique examines how adaptation may reproduce colonial patterns through technical governance.
Climate adaptation has emerged as the primary framework through which developing countries must now articulate their development needs Chapter 2. The patterns discussed above are visible in the implementation, financing, and knowledge production of responses to climate change. Two frameworks emerge from the literature: the adaptation nexus and Paprocki (2018)’s adaptation regime. These represent not simply different perspectives but fundamentally different ways of understanding what adaptation is, how it operates, and what it accomplishes.
The adaptation nexus presents climate adaptation as an urgent technical challenge requiring coordinated response. Climate change poses real threats to development gains (Calvin et al. 2023). Rising temperatures, changing precipitation patterns, and extreme weather events affect agriculture, infrastructure, health, and livelihoods. The institutional approach sees adaptation as the necessary integration of climate considerations into development planning.
From this perspective, the convergence observed in adaptation planning reflects successful coordination and learning. Countries face similar challenges and can benefit from shared approaches. The World Bank’s Climate Change Action Plan positions adaptation at the intersection of climate science, development planning, and disaster risk reduction (WB 2021). This nexus requires breaking down silos between sectors, integrating across scales from local to global, and coordinating among multiple actors. The complexity demands common frameworks and standardized approaches to enable comparison and aggregation.
Within the nexus framework, implementation challenges are primarily matters of capacity and coordination. Countries need technical assistance to conduct vulnerability assessments, as the IPCC guidelines recommend specific methodologies for evaluating exposure, sensitivity, and adaptive capacity (Calvin et al. 2023). They require institutional strengthening to manage complex planning processes that span multiple ministries and levels of government. The logical frameworks and results-based management systems that structure adaptation planning provide necessary tools for managing this complexity and ensuring accountability to both citizens and donors (WB 2021).
The financing architecture represents a pragmatic response to adaptation’s enormous costs. Stern, Songwe, and Bhattacharya (2022) reviewed adaptation costs in developing countries and put it at tens of billions annually (Stern, Songwe, and Bhattacharya 2022; Buchner et al. 2023). Here, the traditional aid insufficient, requiring innovative mechanisms to mobilize private capital (Bernards 2024; UN 2024a). This capital is to be raised through blended finance, green bonds, and climate funds (Buchner et al. 2023).
Knowledge production within the nexus operates through scientific assessment and technical innovation. The IPCC synthesizes thousands of studies to provide authoritative climate projections (Calvin et al. 2023). Development banks collect and disseminate adaptation experiences across countries. This accumulated expertise helps countries avoid mistakes and adopt proven approaches. The nexus acknowledges problems but sees them as challenges to be overcome through better practice. The gap between planning and implementation can be narrowed through improved project management (UNEP 2024). The mismatch between global frameworks and local realities can be addressed. The persistence of vulnerability despite adaptation investments indicates the need for more resources and better targeting, not fundamental flaws in the approach.
The adaptation regime presents climate adaptation as a mechanism of control that extends existing patterns of domination. Paprocki (2018)’s analysis of Bangladesh reveals how adaptation operates as a “socially and historically specific configuration of power” that governs the landscape of possible intervention (Paprocki 2018, 957). The regime doesn’t simply respond to climate change but actively constructs it as a particular kind of problem requiring particular solutions that serve specific interests.
Agrawal (2005)’s concept of “environmentality” extends Ferguson’s analysis into environmental governance, showing how subjects are transformed through technologies of government. Through community-based forest management in India, villagers become environmental subjects who internalize conservation logics not through coercion but through participation in regulatory practices. This transformation, from forest users to forest guardians, occurs through new institutions, knowledge practices, and the formation of environmental subjects (Agrawal 2005, 8).
From this perspective, convergence in adaptation planning might just be the operation of power. The rendering technical of climate vulnerability obscures climate change’s political roots. Eriksen et al. (2021) argue that vulnerability emerges from other causes than just climate change itself (2021, 3). Adaptation planning might focus on technical adjustments rather than addressing structural causes. The anti-politics machine that Ferguson (1994) identified in development could operate through climate discourse, transforming the political question of adaptation into technical questions of climate change.
To access climate finance, countries have to present themselves through categories that donors recognize. As Dewan (2022) argues, development professionals describe climate change as “the most amazing spice” to attract funding, adding it to any project regardless of relevance (2022, 539). The elaborate apparatus of plans, assessments, and frameworks disciplines both states and populations into particular ways of understanding and responding to climate change that maintain rather than challenge existing arrangements. Climate finance could just be relabeled aid, as Scoville-Simonds, Jamali, and Hufty (2020) argue, with donors retrofitting adaptation into development assistance, rather than providing anything new (2020, 8).
Because of the future orientation of climate adaptation, this could lead to a form of anticipatory ruination (Paprocki 2018, 969). By casting the future as a climate dystopia, where damages from climate change are inevitable, all other issues are seen as connected. This is climate reductionism, collapsing all political issues into a single, inevitable process of climate change (Hulme 2011). These different interventions are justified both by present knowledge, and in the uncertainty of the future (Paprocki 2018, 958).
The experiments are meant to test out and evaluate the different possible approaches to adaptation, much in line with the institutional approach to development approaches like the LFA (Örtengren 2004), and to find the most effective interventions through constant evaluation (Paprocki 2018, 963). All approaches become “no-regrets” approaches, since anything would still be better than the dystopian climate future. Paprocki (2018) found that this process led to dispossession, the loss of land by peasants, to shrimp agriculture and migration to the big cities (2018, 966). As rural livelihoods, the most vulnerable to climate change, also is seen as obsolete and backwards, climate adaptation is seen as an “opportunity” to shift to something else (Paprocki 2018, 969).