This paper exploits vast granular data – over 10 million county-industry-month observations – to estimate dynamic panel data models of weather’s short-run employment effects. I estimated the contemporaneous and cumulative effects of temperature, precipitation, snowfall, the frequency of very hot days, the frequency of very cold days, and natural disasters on private nonfarm employment growth. The short-run effects of weather vary considerably across sectors and regions. Favorable weather in one county has positive spillovers to nearby counties but negative spillovers to distant counties. Local climate mediates weather effects: economies are less sensitive to types of weather they are accustomed to.
About the Author
Daniel Wilson is a vice president in the Economic Research Department of the Federal Reserve Bank of San Francisco. Learn more about Daniel Wilson