Estimating Macrofiscal Effects of Climate Shocks from Billions of Geospatial Weather Observations

Abstract

The literature studying the macroeconomics of weather has focused on temperature and precipitation annual averages, while micro studies have focused more on extreme weather measures. We construct hundreds of variables from high-frequency, high-spatial-resolution weather measurements. Using the LASSO, we identify the parsimonious subset of variables that can best explain GDP and key macrofiscal variables. We find that scarcer mild temperatures and an increase in the occurrence of high temperatures and severe droughts reduce GDP. These variables substantially improve the share of GDP variations explained by weather. Additional evidence suggests that fiscal policy mitigates these shocks. (JEL E23, E62, O44, Q54, Q56)