Accurate projections of macroeconomic climate change damages are crucial for informing policy debates on adaptation, mitigation, and climate justice. Adaptation strategies require an understanding of the magnitude and spatial distribution of future climate impacts, particularly relevant for climate justice considerations. Mitigation policies, such as the Paris Agreement, necessitate balancing implementation costs against the benefits of avoided damages. Long-term projections, however, face challenges due to uncertainties in future emission scenarios and the long-term climate response. This study overcomes these challenges by focusing on near-term damages already committed by historical emissions and socio-economic inertia, minimizing uncertainties associated with long-term projections. This focus simplifies communication and enhances the credibility of projected economic damages.

The paper builds upon recent advances in climate econometrics, which provide evidence of sub-national economic growth impacts from various climate variables. Studies utilizing fixed-effects panel regression models and exploiting within-region variations in temperature and precipitation have identified causal relationships between these variables and economic productivity. These studies have revealed impacts from changes in daily temperature variability, total annual precipitation, the annual number of wet days, and extreme daily rainfall, in addition to the already established effects of changing average temperature. Regional heterogeneity in these effects has also been observed. The selected climate variables align with micro-level evidence on the impacts of average temperatures on labor and agricultural productivity, temperature variability on agricultural productivity and health, and precipitation on agricultural productivity, labor outcomes, and flood damages. The authors utilize these established findings to provide a more comprehensive and spatially detailed description of climate impacts.

A key challenge in estimating future climate damages lies in determining the persistence of climate variable impacts on economic growth. The authors address this by using distributed lag models to test the significance of delayed effects for each climate variable. Unlike previous studies, they use first-differenced climate variables, implying a dependence of growth rate on changes in climate variables. This approach establishes a baseline specification of purely level effects, where a permanent climate change has only an instantaneous effect on the growth rate. By including lags, the study tests for persistent effects. This conservative approach avoids assumptions of infinite persistence, providing a robust lower bound on impact persistence. The methodology includes ten lags of first-differenced climate variables in fixed-effects distributed lag models, identifying substantial effects on economic growth at lags up to 8-10 years for temperature and up to 4 years for precipitation. Model selection involves evaluating information criteria to balance data fit and avoid overfitting. Extensive robustness checks are performed, including Monte Carlo simulations to address autocorrelation, lag selection, multicollinearity, and the necessity of including multiple climate variables. A restricted distributed lag model is also used to limit oscillations in lagged parameter estimates. The uncertainty of lag selection is explicitly accounted for by sampling from models with marginally different numbers of lags in the error-sampling procedure. The study combines empirical economic response functions with an ensemble of 21 CMIP-6 climate models to project macroeconomic damages. Bias-adjusted climate models are used to avoid introducing biases into the projections. A Monte Carlo procedure samples from climate model projections, empirical models with different numbers of lags, and bootstrapped parameter estimates to estimate overall uncertainty. The study also incorporates various robustness checks relating to timescale of moderating variable estimation and the order of accounting for intertemporal and international currency comparisons.

The study projects that global damages are statistically indistinguishable across extreme emission scenarios until 2049, indicating that these damages are already committed. These committed damages represent a permanent income reduction of 19% globally (likely range 11–29%) compared to a baseline without climate change impacts. Even with per capita income increases, most regions experience permanent income reductions, with South Asia and Africa most severely affected. The committed global annual damages in 2049 are estimated at 38 trillion 2005 international dollars (likely range 19–59 trillion). The committed damages outweigh mitigation costs by a factor of six. While near-term damages are committed, damage estimates diverge strongly after 2049, emphasizing the economic benefits of mitigation. Committed damages arise primarily from changes in average temperature. However, accounting for other climatic components (daily temperature variability, total annual precipitation, number of wet days, and extreme daily rainfall) increases damage estimates by approximately 50%. The largest increases in overall damages are caused by daily temperature variability (4.9 percentage points, equivalent to approximately 10 trillion international dollars). The spatial distribution of committed damages reveals a complex interplay between future climate changes and historical economic vulnerability. Damages from increasing annual mean temperature are larger at lower latitudes, where temperatures are already high. Damages from daily temperature variability show strong latitudinal polarization, increasing at lower latitudes and decreasing at higher latitudes. Changes in precipitation bring mainly economic benefits except in drying regions. Changes in extreme daily rainfall produce damages in all regions. The study reveals significant injustice in committed damages, with larger losses in countries with lower historical cumulative emissions and lower current income per capita. Countries with lower income are committed to an income loss 8.9 percentage points (or 61%) greater than the higher-income quartile. Similarly, countries with lower historical emissions face an income loss 6.9 percentage points (or 40%) greater than those with higher cumulative emissions. The projected damages are larger than in previous studies due to the use of sub-national data, accounting for additional climate variables, and a robust lower bound on impact persistence. Despite the seemingly large projected reductions in income (19%) compared to historical variance explained by the empirical models (<5%), this is due to projected changes exceeding historical ones, particularly for average temperature. The methodology mitigates out-of-sample extrapolation by capping moderating variables at the 95th percentile of the historical distribution. The study notes that the projected damages do not account for all climate change impacts (e.g., heat waves, sea-level rise, tropical cyclones, tipping points, and non-market damages), meaning estimates could be conservative. It also doesn't fully account for spatial spillovers of regional impacts, which could potentially amplify damages. A spatial-lag model analysis suggests that including spatial spillovers increases both the magnitude and heterogeneity of impacts.

The findings directly address the research question by quantifying the committed economic damages from climate change. The results highlight the significant and already-present economic costs of climate change, surpassing mitigation costs considerably. This challenges the common perception that mitigation benefits primarily emerge in the latter half of the century, demonstrating that inaction has significant and immediate economic consequences. The study’s significance lies in its focus on committed damages, providing a more immediate and credible argument for urgent climate action. This near-term focus strengthens the case for mitigation, emphasizing that the economic rationale for action extends beyond long-term projections.

This study demonstrates that substantial economic damages from climate change are already locked in due to past emissions and present socio-economic conditions. These committed damages significantly exceed mitigation costs, underscoring the immediate economic imperative for climate action. Future research should focus on a more comprehensive assessment of impacts, incorporating factors such as sea-level rise, extreme weather events, and non-market damages, as well as a more complete accounting of spatial spillovers. Further investigation into regional mitigation cost estimates could also refine the understanding of national incentives for mitigation.

The study acknowledges several limitations. It does not include all potential climate change impacts, such as those from heatwaves, sea-level rise, tropical cyclones, tipping points, and non-market damages, which could further increase the estimated damages. The analysis also does not fully account for spatial spillovers, which are likely to amplify the magnitude and heterogeneity of impacts. The reliance on historical data and extrapolations beyond the range of that data introduces uncertainties into the projections. The simplicity of the comparison of global-level damages and mitigation costs could benefit from a more granular regional analysis.