Erik Brynjolfsson, J.R. Horton, Christos Makridis, Alex Mas, Adam Ozimek, Daniel L. Rock, Hong‐Yi TuYe
ABSTRACT Remote work surged during the COVID‐19 pandemic, but estimates vary widely. To address this, we field the Remote Life Survey (RLS), a nationally representative survey. In October 2020, we find that 31.6% of continuously employed workers always worked from home (WFH), and 21.9% did so sometimes or rarely, totaling 53.5%. We compare our results with government surveys and assess four factors contributing to measurement differences: (a) web versus mail‐based respondents, (b) inclusion of self‐employed workers, (c) occupation mix, and (d) exclusion of pre‐pandemic remote workers. We find that (d) explains most of the discrepancy between the Current Population Survey (CPS) and other measures. Policymakers and researchers relying on CPS data should note that it may underestimate remote work prevalence by up to 25 percentage points. Our preferred estimates suggest that about half of the U.S. workforce worked remotely at least one day per week as of December 2020.