Effects of Reduced Workplace Presence on Covid-19 Deaths: An Instrumental-Variables Approach

Abstract

Numerous government policies have attempted to keep workers out of the workplace, on the assumption that this will lower transmission of COVID-19. We test that assumption, measuring the effect of aggregate workplace absence on US COVID deaths at the county level through August. Instrumenting with an index of how many local workers pre-pandemic can work from home, based on differences in county occupational mix, we find no effect of workplace absence until mid-May, then a sharply rising effect. By August, moving 10 percent of a county's workers from the workplace would lower deaths there by three quarters one month later.