Program in Applied Mathematics Colloquium

I formally relate the consequences of climate change to the time series variation in weather extensively explored by recent empirical literature. I delineate the classes of payoff functions for which reduced-form fixed effects estimators can exactly recover the effects of climate. The conditions become more restrictive when regressions do not control for forecasts or for lagged weather, as has been common in the empirical literature to date.  I also show how to recover structural estimates from reduced-form weather regressions.  My median estimates indicate that following the RCP 4.5 climate change trajectory will reduce U.S. agricultural profits by nearly 50% over this century.