Adapting to climate change – finding ways to prevent the harshest effects or deal with crises when they occur – is a first-order issue. But research on the economics of adaptation has offered little guidance for policy. This column reviews the existing literature on adaptation and considers how it can more directly inform policy design and implementation. The authors identify two distinct streams of adaptation literature, outline how each can be used to evaluate policy, highlight limitations and opportunities for public intervention in private adaptation markets, and provide guidance for future work.
With each subsequent year breaking new temperature records, climate change is not a problem of the future: the problem of adapting to climate change – finding ways to prevent the most severe impacts, or dealing with crises when they occur – is a first-order issue. Policymakers have taken notice, and adaptation has become more of a priority. However, the academic literature on the economics of adaptation offers little direct guidance for adaptation policy. This column – based on our chapter in the Handbook of the Economics of Climate Change (Carleton et al. 2024) – reviews the growing body of research studying the impacts of and barriers to adaptation through the lens of a stylised framework. We highlight key methodological challenges and insights that can help inform adaptation policy.
Climate change increasingly shapes our everyday lives, leaving individuals, communities, governments, and companies across the globe with the task of improving their resilience to harmful weather, natural disasters, and other climate shocks (Kala et al. 2023). These adaptive decisions – ranging from installing irrigation to investing in indoor cooling systems to shifting outdoor working hours – are not made in isolation, but shaped by the economic, political, and cultural conditions that influence decisions and may themselves respond to an evolving climate.
Mounting evidence suggests that a host of financial barriers, market frictions, and behavioural biases prevent or constrain adaptation that can substantially reduce climate-related losses. These barriers often disproportionately affect low-income individuals and countries, where extreme heat and weather variability are already more pronounced and where governments and individuals lack the resources to invest in resilience (Premand et al. 2022, Macours et al. 2022). Without policy intervention, adaptation may widen inequality, as wealthier individuals and countries protect themselves while vulnerable populations bear the brunt of climate damages (Abajian et al. 2025, Mendelsohn et al. 2006).
Understanding where adaptation is working, where it is insufficient, and how it is shaped by policy is critical to realising the potential for adaptation to reduce losses – particularly among the most vulnerable.
Defining adaptation: Two channels of adaptive decision-making
Adaptation takes two main forms. First, there are reactive (ex-post) responses to realised weather, such as increased air-conditioning use during heat waves or moving to non-agricultural work during a poor harvest season. In contrast, proactive (ex-ante) investments are undertaken before weather is experienced: examples include planting drought-resilient crops or building climate-resilient infrastructure such as bridges and sea walls (Lemoine 2021). Unlike ex-post adaptation, these ex-ante investments are made based on expectations about weather realisations that will occur in the future, so they depend on information and beliefs about future climate.
Whether and to what extent each of these adaptation channels are pursued in practice depends on a set of external factors unrelated to climate itself and regularly shaped by policy interventions. These factors include prices of adaptation technologies, income, market structure, and any market frictions. As an example, if flood tolerant crop seeds get cheaper, farmers will be more likely to adopt them. Another example is income – increased wealth in India will fuel expansive growth in air-conditioning installations, regardless of future climate change. Numerous other individual, market, and policy factors determine adaptation decisions and impede optimal adaptation.
A growing body of literature exploits variation in historical weather and long-run climate as well as changes in the external factors shaping adaptation decisions to measure the impact of and barriers to ex-post and/or ex-ante adaptation. In our chapter, we synthesise this research and point out opportunities for improved policy relevance.
Why study adaptation? The evolution of the adaptation literature and its policy motivations
The adaptation literature has evolved through two main streams with very different aims. The first focuses on measuring climate change damages, with a goal of informing mitigation policy (e.g. deriving an optimal carbon tax). The second treats adaptation as a policy objective itself, evaluating specific interventions to enhance climate resilience through infrastructure, finance, or regulation. The two streams have not spoken much to each other, which has been a limiting factor for them both.
The study of adaptation to inform mitigation policy
Estimates of the effect of climate change on welfare-relevant outcomes (agricultural profits, economic output, health, etc.) help governments decide which emission mitigation policies to pursue. However, to accurately estimate such climate damages, we must account for how people and firms will adapt to a warming world. Thus, a stream of literature has developed a set of methods to measure the extent to which adaptation lowers future climate change damages (Hogan and Schlenker 2024, Kolstad and Moore 2020).
Some of these studies examine empirical relationships without specifying how adaptation is occurring. They combine historical observations of short-run weather events (e.g. a summer heat wave or drought), long-run climate records (e.g. a 30-year moving average of annual temperature), and outcome data (e.g. crop yields, mortality rates) to estimate reduced form effects of weather variations, which account for how people endogenously adapt to their long-run climate. While methods are diverse, the basic intuition is to observe how populations in an already hot climate (such as Phoenix, AZ, in the US) respond to heat, and compare their response to populations that rarely experience such heat (e.g. Seattle, WA). All else being equal, the increased resilience in Phoenix can be attributed to improved ex-ante adaptive investments (like home air conditioning) and ex-post responses (like choosing to stay indoors). Instead of enumerating specific adaptive margins, these methods implicitly account for unobservable changes in adaptive choices. Key findings in this literature show that ex-ante adaptation significantly reduces climate damages across outcomes including mortality, energy consumption, and labour supply (Auffhammer 2022, Carleton et al. 2022, Heutel et al. 2021, Rode et al. 2021, Rode et al. 2022). Another key finding of this literature is that adaptation is constrained by income levels. Even places with a long history of hot weather experience more damage when they are poor (Carleton et al. 2022). On the other hand, adaptation has been less effective in protecting agricultural output – particularly in the US (Burke and Emerick 2016, Schlenker and Roberts 2009, Schlenker et al. 2013) – and aggregate economic outcomes such as GDP (Burke et al. 2015, Dell et al. 2012).
Other methods explicitly model specific adaptive margins – such as international trade, migration, and infrastructure investments – via quantitative general equilibrium models (Bilal and Rossi-Hansberg 2023, Desmet and Rossi-Hansberg 2024) or Integrated Assessment Models and process-based models (Diaz and Moore 2017, Martinich and Crimmins 2019). By construction, these studies omit many plausible adaptive margins, but they enumerate the welfare effects of specific barriers and frictions impeding the deployment of adaptation. In part because the studies rely heavily on functional form assumptions and parameter calibrations, the existing literature has yet to converge. For instance, two of the most studied adaptive margins, trade and migration, show substantial variation in the magnitude of their benefits (Docquier et al. 2019, Laborde and Gouel 2019, Rossi-Hansberg et al. 2021).
Together, these approaches offer innovative methodological solutions to account for adaptation in the quantification of climate damages. However, they tell us relatively little about how to design context-specific interventions that could improve or hinder ex-post or ex-ante adaptation.
The study of adaptation to inform adaptation policy
A second distinct motivation for studying adaptation is to inform policies and programmes that influence adaptation. To do so, a separate body of literature focuses on directly recovering estimates of the impact of certain adaptive actions on welfare outcomes. These studies have relied on quasi-experimental or experimental variation in 1) prices of ex-ante or ex-post adaptation technologies (e.g. Aker and Jack 2023, Björkman-Nyqvist et al. 2023); 2) income (e.g. Macours et al. 2022, Premand and Stoeffler 2022); or 3) frictions and constraints in the adaptation decision environment to assess their impact on climate damages and on specific adaptive actions, behaviours, or investments (e.g. Barwick et al. 2022, Burlig et al. 2023, Lane 2024, Shrader et al. 2023). Figure 1 summarises how the existing literature is spread across categories, revealing an emphasis on income transfer interventions and the manipulation of climate information and market frictions.
Figure 1 Source of exogenous variation exploited and sectors studied in papers leveraging variation in adaptation interventions
Several studies show that introducing climate-adapted agricultural technologies such as new seed varieties can improve household welfare (e.g. Emerick et al. 2016, Aker and Jack 2023). Improvements in healthcare services and access to them have also been shown to moderate the mortality-heat relationship in the US (Mullins and White 2020), Mexico (Cohen and Dechezleprêtre 2022), and Uganda (Björkman Nyqvist et al. 2023). Many papers have studied how conditional and unconditional cash transfers moderate the effects of weather shocks on outcomes – including educational and labour market outcomes in Mexico (Adhvaryu et al. 2024), consumption and food security in Zambia (Asfaw et al. 2017), savings, asset accumulation, and income smoothing in rural Niger (Premand and Stoeffler 2022) and Nicaragua (Macours et al. 2022), and violence in Mexico (Garg et al. 2020) and Indonesia (Christian et al. 2019). In the study of frictions in the broader adaptation decision environment, the literature has focused mostly on access to information in the form of forecasts, which has been shown to benefit both high-income (e.g. Molina and Rudik 2023, Shrader et al. 2023) and low-income countries (e.g. Burlig et al. 2024, Rosenzweig and Udry 2013). Other studies have examined the role of frictions in input and output markets, including access to financial services like mobile money (Burgess et al. 2017, Rajan and Ramcharan 2023, Riley 2018, Jack and Suri 2014). Fewer studies have focused on the role of market power and market structure, which, for example, restricts adaptation in agricultural markets in India (Kochhar and Song 2024).
These studies face two major empirical challenges. First, they typically rely on limited spatial and temporal variation in weather conditions, decreasing their relevance for adaptation to long-run climate change. The choice of the functional form used to model weather also plays a crucial role in analyses of the adaptation interventions, with implications for the interpretation of results (see Carleton et al. 2024, Figure 2). Second, studies focused directly on adaptation rarely consider explicit projections to future climate states. As a result, they focus on weather shocks identified in historical data that may not represent the types of shocks relevant to future climate change. For example, interventions that have been effective in addressing historical droughts may be less helpful for future increases in precipitation (the situation that many regions in Africa will experience). In addition, as climate changes, the relevance of different adaptive margins may also change: irrigation may be an effective response to a single hot growing season, but not for decades of heat in a row that see ground and surface water stores dwindle.
How should public policy address adaptation?
While private adaptation is crucial to reducing climate damages, its effectiveness depends on financial resources, institutional capacity, and the functioning of associated markets, which introduce potential inefficiencies and frictions. This raises several questions regarding how, when, and for whom the public sector should intervene to guide adaptation. From a traditional economic perspective, adaptation policy interventions are justified when market imperfections distort private adaptation decisions. Such interventions can improve information access (e.g. Burlig et al. 2024, Shrader et al. 2023); provide public goods (e.g. Benetton et al. 2022, Bradt and Aldy 2022, Kelly and Molina 2023); correct distortions generated by externalities associated with adaptation investments (e.g. Abajian et al. 2023, Colelli et al. 2022); or address incomplete markets, notably insurance and credit (e.g. Collier et al. 2024, Lane 2024).
Without policy intervention, adaptation may exacerbate inequality. If wealthier individuals and countries are better able to protect themselves, vulnerable populations will bear the majority of climate damage. For this reason, even in the absence of traditional motives for policy intervention driven by market failures, equity may serve as another welfare-based motivation for policy intervention. Public spending to support adaptation can help reduce inequality by making adaptation more affordable for everyone or by directing more benefits to poorer communities. This principle has shaped international discussions of funding mechanisms like the Loss and Damage fund, which could be used to justify public spending on adaptation with transfers both across and within countries. In a world where unrestricted income transfers are politically feasible, they may theoretically be preferable to adaptation as they leave full discretion to recipients; if income transfers are constrained, adaptation policies could be the best alternative.
Policymakers also need frameworks to evaluate or prioritise among many possible adaptation policies (Carleton et al. 2024). We see two general approaches. In the first, policy evaluation accounts for all ancillary costs and benefits of a policy that might or might not explicitly target adaptation. For example, public adaptation policies like sea walls might also create recreational opportunities that count toward the costs and benefits of the policy. Likewise, policies that do not explicitly target adaptation but create adaptation co-benefits may end up ranking highly. For example, restrictive gun laws mitigate the effect of temperature on homicides (Colmer and Doleac 2023). In the second approach, a narrow evaluation and ranking considers only policy impacts on adaptation itself, thus accounting only for the cost-effectiveness of reducing the impacts of climate shocks. This approach is gaining traction in recent policy debates, through the Green Climate Fund and the Adaptation Fund supported by the UN Framework Convention on Climate Change. While the latter approach might offer a transparent way to compare policies based on their adaptation benefits and guide spending toward the most cost-effective investments from the narrow perspective of climate change adaptation, it might also end up prioritising projects with negative consequences in other sectors or neglecting policies with large welfare gains but smaller adaptation benefits.
Future directions for research on adaptation
Future research on climate adaptation should focus both on improving methodologies and on informing adaptation policy. Priorities include a more theoretically and empirically grounded justification for functional form assumptions that model climate and weather supported by robustness tests when short-run weather variation is limited (Carleton et al. 2024). To improve the relevance of findings for future climate change, studies should focus on understanding how society adapts to moderate changes in weather conditions in addition to often-used extreme conditions. Importantly, they should also focus on climatic characteristics and shocks that are geographically relevant and projected to increase in frequency and/or intensity.
To better align adaptation research with policy demand, researchers should focus on unpacking the mechanisms through which policy interventions have been found to reduce climate damages. In doing so, research can be effective at informing targeted adaptation interventions – or assessing the holistic costs and benefits of interventions that influence adaptation – as well as other socially relevant outcomes.
While research needs remain urgent, the adaptation literature is growing quickly, drawing on a diverse set of tools, data, settings, and questions. Combining insights across fields and learning from each other’s methods will continue to advance the frontier. For example, interrogating model-driven approaches using well-identified variation from adaptation interventions can leverage the strengths of both, generating future climate scenarios based on causal adaptation evidence. This expanding body of work can more directly inform adaptation policy by explicitly considering policy motivations (i.e. the welfare function), barriers to private adaptation (e.g. market failures), and opportunities (e.g. adaptation policy co-benefits) in the study design, interpretation, and conclusions. Together, these insights have the potential to shape the future of adaptation policy design and human well-being in a warming world.
Source : VOXeu
