Extreme Weather in a Changing Climate: Implications of Learning and Adaptation for Predicting Economic Damages

Joint with Jiangmin Xu at Peking University

Abstract

The adverse consequences of global warming for extreme weather are uncertain. Theory predicts time-varying damages as society learns about these consequences from extreme-weather arrivals and adapts. Using tropical cyclones and temperature data, we show that learning results in a misspecification of a widely-used panel regression for making causal inferences regarding damage to economic growth from extreme-weather events. To address this misspecification, we estimate a varying-coefficient model where damage from a given event declines as society expects more frequent recurrence of extreme weather. Our projections for damages over the coming century are substantially lower than current ones that ignore adaptation.

Harrison Hong Profile