Measuring the macroeconomic cost of climate change from the ground up

Reading Time: 4 minutes

New OECD analysis combines granular local climate damages with a global spatial model to estimate how heat, floods, shifting crop yields and sea-level rise affect future aggregate economic activity, taking into account key trade, migration and production networks.

By Diogo Baptista, Hélia Costa and Filiz Unsal, OECD Economics Department

Climate change affects economies through local and granular channels. Workers are exposed to heat, firms are hit by floods, farms face changing growing conditions, and coastal areas lose land to rising seas. Yet the economic consequences are not confined to the places directly affected. Through trade, migration and production networks, local climate damages can reshape economic outcomes across regions, countries, and ultimately the global economy.

In a new OECD working paper(Baptista et al., 2026), we develop the Climate and Adaptation Spatial General Equilibrium Model (OECD-CASGEM) to connect these local damages to country specific macroeconomic outcomes. The model aggregates climate impacts across 927 regions in 186 countries while accounting for trade, migration and production linkages within and across countries. This allows us to quantify how local climate shocks propagate through the global economy, shaping both aggregate losses and their distribution across regions.

Quantifying the GDP impacts of climate change from local damages, using detailed micro-level estimates, granular data, and frontier methods, can help policy makers identify where losses are likely to be largest, which channels matter most, and what trade-offs different policy choices may involve.

Why is this approach needed?

Climate damages are often estimated using aggregate relationships between temperature and GDP (e.g., Bilal and Kanzig, 2025). This provides useful evidence on the overall scale of the problem, but it can obscure critical channels through which climate change affects the economy and the way these effects differ across places.

OECD-CASGEM takes a more granular, bottom-up approach. It starts from local climate damages and then aggregates them through the lens of a global spatial model. (Figure 1). This spatial structure matters because climate impacts depend not only on where hazards occur, but also on how regions are connected to one another. A flood that damages firms in one region can raise costs for firms elsewhere that rely on their inputs. Heat stress in one sector can affect other sectors through lower output, delayed deliveries or higher prices. Trade and migration linkages can also transmit local shocks across regions and countries, while allowing households and firms to adjust over time.

The project combines evidence on four major climate hazards: heat stress, flooding, changing agro-climatic conditions and sea-level rise. These hazards affect economic activity through different channels. Heat stress reduces productivity. Floods damage firms’ capital. Changing crop conditions affect agricultural productivity. Sea-level rise reduces the supply of land available for housing and production.

Figure 1. Quantifying the cost of climate risks: a micro-based, bottom-up approach

**Note** (hover to read the text)
Source: Baptista et al. (2026).

From local impacts to country-specific GDP losses

The results point to sizeable GDP losses in the future for most countries. Under a current-policy scenario, global GDP per capita is projected to be around 3% lower by 2050 and 6.3% lower by 2100 than in a world where climate conditions remain fixed at historical levels. Under a high-emissions scenario, losses rise to around 6% by 2050 and 18% by 2100.

OECD economies face smaller average losses, but the impacts remain material. GDP per capita is projected to be around 1.7% lower by 2050 and 3.1% lower by 2100 under a current-policy scenario. Under a high-emissions scenario, the losses rise to about 2.6% by 2050 and 9% by 2100 (Figure 2). Spatial linkages also shape these losses, with for example input-output linkages amplifying local shocks as disruptions in one region or sector affect firms elsewhere through production networks.

These averages hide large differences across and within countries. Warmer and lower-latitude regions tend to face larger losses, especially through heat stress and weaker agricultural productivity. In the OECD, losses are projected to be above average in parts of Southern Europe, the United States, Australia and Chile, while cooler regions tend to face smaller impacts. This region- and country-specific perspective is central for policy, as the scale and composition of climate damages differ markedly across economies.

How will economies adjust?

Climate change will also reshape economic geography. As some places become less productive or more costly, workers and firms have incentives to move towards less affected regions and sectors. Trade patterns can also adjust, as firms and households source more goods from places where climate damages are smaller.

These adjustments to existing patterns in response to climate change have the potential to reduce global losses. Under a high-emissions scenario, allowing trade and labour mobility to reshape relative to current levels lowers global GDP per capita losses in 2100 by around 8%. It also impacts the distribution of losses, reducing the dispersion of climate impacts across regions and acting as a form of loss-sharing.

But adjustment is not costless. Regions receiving new workers face pressure on housing and local prices. Regions losing workers may see weaker local demand and lower wages. Trade shifts also reduce demand for goods produced in the most climate-exposed places.

Figure 3. The ability to adjust to climate change reshapes losses

Additional percentage point loss in GDP per capita in 2100 due to adjustment via labour mobility and trade

**Note** (hover to read the text)
Source: Baptista et al. (2026).

What does this mean for policy?

Mitigation remains the first line of defence. Lower emissions reduce the scale of future damages and limit the risk of extreme outcomes. But adaptation is also essential. Investments in climate-resilient transport, energy, water and flood-protection systems can reduce both direct climate damages and their propagation through supply chains.

Policies that support private adaptation, including better climate-risk information, insurance coverage and access to finance, can strengthen resilience. More broadly, adaptation planning should consider how climate shocks and resilience investments affect interconnected regions, sectors and countries regions through trade, migration and production networks.

References

Baptista, D., H. Costa and F. Unsal (2026), “The climate and adaptation spatial general equilibrium model (OECD-CASGEM): the macroeconomic cost of climate change”, OECD Economics Department Working Papers. https://doi.org/10.1787/bd3ad48b-en.

Bilal, A. and D. Känzig R (2026), “The Macroeconomic Impact of Climate Change: Global Versus Local Temperature”, The Quarterly Journal of Economics, Vol. 141/2, pp. 889–944, https://doi.org/10.1093/qje/qjag011.

Costa, H., Franco, G., Unsal, F., Mudigonda, S., Caldas, M. P. (2025), The heat is on: Heat stress, productivity and adaptation among firmsOECD Economics Department Working Papers, No. 1828, OECD Publishing, Paris, https://doi.org/10.1787/19d94638-en.