New damage curves and multimodel analysis suggest lower optimal temperature

Economic analyses of global climate change have been criticized for their poor representation of climate change damages. Here we develop and apply aggregate damage functions in three economic Integrated Assessment Models (IAMs) with different degrees of complexity. The damage functions encompass a wide but still incomplete set of climate change impacts based on physical impact models. We show that with medium estimates for damage functions, global damages are in the range of 10% to 12% of GDP by 2100 in a baseline scenario with 3 °C temperature change, and about 2% in a well-below 2 °C scenario. These damages are much higher than previous estimates in benefit-cost studies, resulting in optimal temperatures below 2 °C with central estimates of damages and discount rates. Moreover, we find a benefit-cost ratio of 1.5 to 3.9, even without considering damages that could not be accounted for, such as biodiversity losses, health and tipping points.