Measuring Trends in Work from Home: Evidence from Six U.S. Datasets

Abstract

This article documents the prevalence of work from home (WFH) using six nationally representative U.S. surveys. These surveys measure WFH using different questions, reference periods, samples, and survey collection methods. After constructing comparable samples and WFH measures across surveys, we find that the surveys show broadly similar trends in the trajectory of aggregate WFH since the COVID-19 outbreak. The most important source of disagreement in WFH levels across surveys is in WFH by self-employed workers; by contrast, WFH rates for employees are closely aligned across surveys. All surveys show that, in 2024, WFH remains substantially above pre-pandemic levels. We also highlight that while full-time WFH drove most of the increase in aggregate WFH during and after the pandemic, part-time WFH has become a more significant contributor since 2022. Finally, we validate the findings from the survey data by comparing self-reported commuting behavior to cell phone geolocation data from GoogleWorkplace Visits.

Introduction

Following the COVID-19 outbreak in March 2020, many U.S. workers stopped commuting and started to work from home (WFH). Almost immediately, a large literature emerged to document this trend and study its implications for the economy and society more broadly. This literature relies on several surveys with information on commuting and WFH. These surveys measure WFH in different ways and use different reference periods, samples, and survey collection methods. As a result, they often exhibit very different WFH rates (Brynjolfsson, Horton, et al., 2023). This variation makes it difficult to understand exactly how much WFH increased and whether differences in WFH rates across surveys reflect concepts of WFH, different samples, or true disagreement between surveys.