Reconsidering the macroeconomic damages of severe warming

Projections of macroeconomic damage from future climate change tend to suggest mild to moderate impacts. This leads to welfare-optimal climate policies in Integrated Assessment Models (IAMs) that recommend very slow emissions reductions over the coming decades, in sharp contrast with the ambitions of the Paris Agreement. These econometric models assume that weather impacting a single country is all that affects the economy of that country. We examine whether the addition of global weather conditions in the empirical modelling of economic growth affects the projections of the impact of climate change on global GDP. In effect, we explore whether the interconnectedness of the global economy makes individual countries vulnerable to weather changes that impact other countries. Using three influential econometric models we add global weather to the regressions. We find that this leads to significant worsening of the projections of macroeconomic damage for given future emissions scenarios. Damage to world GDP in 2100 under SSP5-8.5, averaged across both econometric models and climate models increases from ~11%  under models without global weather to ~40% if global weather is included. Further, we demonstrate that when the damage function used in IAMs is estimated from empirical models augmented with global weather conditions, they reduce the welfare-optimal amount of climate change from ~2.7C to ~1.7C which is consistent with the Paris Agreement targets. Our results highlight the need for econometric modelling and climate science’s understanding of extreme events to be integrated much more consistently to ensure the costs of climate change are not underestimated.