Introduction

Vulnerability to climate change serves as a fundamental guiding principle for the allocation of international climate assistance. This principle is enshrined in the UN Framework Convention on Climate Change (UNFCCC), which prioritises assistance to “particularly vulnerable” developing countries. Commitments have evolved from the 2009 Copenhagen Accord target of USD 100 billion annually to the New Collective Quantified Goal adopted at COP29 in 2024, which increased the target to at least USD 300 billion annually by 2035. Despite these commitments, the relationship between climate vulnerability and the observed distribution of adaptation funding remains contested, with cross-national studies offering mixed evidence regarding whether funding reaches those most in need.

Methods

This study addresses whether the subnational distribution of international adaptation funding aligns with patterns of climate vulnerability within recipient countries. The research tracks the distribution of finance from twelve international climate funds across 1,358 municipalities in seven countries: Panama, Costa Rica, Nicaragua, Guatemala, Honduras, El Salvador, and the Dominican Republic. The analysis covers the period between 2007 and 2022, developing the first systematic dataset tracking subnational adaptation funding across multiple funds, countries, and years.

Summary of variables

The study focuses on two outcome variables at the municipal level: climate adaptation projects by total current USD and total project counts. The independent variables are the three disaggregated components of climate change vulnerability: adaptive capacity, sensitivity, and exposure.

Selecting climate finance funds and geocoding projects

Funds were selected from the OECD Climate Fund Inventory. Out of 35 identified funds operating in Latin America and the Caribbean, twelve remained after filtering for location-specific adaptation projects. The research geolocated 87 unique projects to the municipal level out of 191 evaluated projects. A three-step methodology involving geoparsing, georeferencing, and geocoding was employed to identify specific administrative locations. For projects impacting multiple municipalities, the investment was divided equally, and project counts were also aggregated to ensure validity.

Climate change vulnerability

Vulnerability was separated into three core components: exposure, adaptive capacity, and sensitivity. Indicators were chosen based on established methodologies and were weighted equally. Higher index scores indicate greater vulnerability to climate-related challenges.

Climate change exposure

Exposure was measured using climate projections from 23 General Circulation Models (GCMs) for the period 2021–2040 under the energy-intensive SSP5 scenario. Indicators included maximum temperature of the warmest month, minimum temperature of the coldest month, and annual precipitation, compared against historical baselines from 1970–2000.

Adaptive capacity vulnerability index

This index comprised four indicators drawn from census data: percentage of the population with access to water, electricity, and sanitary toilets, and the percentage of adults with only primary education or less.

Sensitivity vulnerability index

Sensitivity focused on three indicators: the dependency ratio, the percentage of indigenous populations, and sensitivity to disasters based on records of deaths, injured persons, and damaged homes between 2005 and 2020.

Analysis

Regression models with fixed effects at the country and year levels were used to assess the relationship between vulnerability indices and adaptation investments. Interactive models were also applied to capture overlapping or intersectional forms of vulnerability across municipalities.

Results

The analysis tracked 191 unique adaptation projects valued cumulatively at over USD 2 billion. Of these, projects geolocated to the municipal level amounted to USD 660,424,579.08 in funding, impacting 817 municipalities. The Green Climate Fund was the largest contributor to geolocated funding at USD 270,145,114.33, followed by the World Bank at USD 124,789,000.00.

Mapping subnational climate aid allocations

Maps revealed wide variation in aid distribution. Approximately 50% of the original list of 191 projects could not be traced to the municipal level due to insufficient documentation, representing a significant challenge in evaluating the efficacy of subnational targeting.

Mapping subnational vulnerability

Vulnerability levels also showed extensive variation. While socioeconomic vulnerability (adaptive capacity and sensitivity) patterns were clear, they did not always overlap with areas of high projected climate exposure.

Low adaptive capacity and high sensitivity are associated with higher adaptation funding

The findings indicate that adaptive capacity is consistently associated with adaptation allocations. A one-unit increase in the adaptive capacity vulnerability index was associated with a 212.64% increase in adaptation allocations and a 0.14 increase in project counts across the cross-sectional models. Sensitivity also showed a significant association when examined independently or in robustness analyses.

Exposure to climate change is not associated with higher adaptation funding

Crucially, projected future exposure to climate change showed no consistent association with funding patterns. This suggests a “blind spot” where international funds prioritised socioeconomic vulnerability alone, potentially overlooking communities at the intersection of high exposure and low adaptive capacity.

Discussion and conclusion

The research concludes that while donors target poorer areas, they may fail to prioritise based on comprehensive climate risk modelling. Socioeconomic indicators are institutionalised in aid frameworks, but complex future climate data are less utilised in allocation decisions. The author advocates for a re-evaluation of vulnerability considerations to address intersectional risks and calls for climate finance donors to enhance transparency and standardized geocoding of fund recipients at the subnational level.