Trends in Work from Home in the U.S.: Insights from Six Datasets

The rise of work from home (WFH) is one of the most significant shifts in the U.S. labor market in recent decades. When COVID-19 was declared a pandemic in March 2020, millions of U.S. workers abruptly transitioned to full-time WFH as an emergency response. While this transition initially seemed temporary, it prompted a reevaluation of traditional work arrangements, raising questions about the lasting role of WFH in the postpandemic landscape. Today, researchers and policymakers alike are still exploring whether WFH is here to stay or whether many workers will eventually return to in-person settings.

In this blog post, we highlight key findings from our recent research paper, which documents the evolution of WFH in the U.S. from the eve of the pandemic to mid-2024.For details, see “Measuring Trends in Work from Home: Evidence from Six U.S. Datasets.” Working Paper 2024-023B, Federal Reserve Bank of St. Louis, September 2024. Our analysis is distinguished from other work by the breadth of datasets considered: We base our findings on six nationally representative U.S. datasets. By creating comparable WFH measures across these datasets, we provide the broadest picture of WFH in the U.S. to date.Other papers have compared WFH estimates in multiple surveys, including Alexander Bick, Adam Blandin and Karel Mertens’ article, “Work from Home before and after the COVID-19 Outbreak,” in the American Economic Journal: Macroeconomics; Erik Brynjolfsson, John J. Horton, Christos Makridis, Alexandre Mas, Adam Ozimek, Daniel Rock and Hong-Yi TuYe’s 2023 working paper, “How Many Americans Work Remotely? A Survey of Surveys and Their Measurement Issues”; José María Barrero, Nicholas Bloom and Steven J. Davis’ article, “The Evolution of Work from Home,” in the Journal of Economic Perspectives; and José María Barrero, Nicholas Bloom, Shelby Buckman and Steven J. Davis’ 2024 slide presentation, “How Much Work from Home Is There in the United States?” (PDF). However, our paper involves more datasets and a longer time horizon.

How Do We Measure WFH?

We draw on four surveys conducted by the U.S. Census Bureau: the American Community Survey (ACS), the American Time Use Survey (ATUS), the Current Population Survey (CPS) and the Survey of Income and Program Participation (SIPP). We also include two independent online surveys: the Real-Time Population Survey (RPS) and the Survey of Working Arrangements and Attitudes (SWAA).The RPS is a nationwide survey that asks the same core questions and follows the same timing and structure as the Current Population Survey (CPS), the monthly labor force survey conducted by the U.S. Census Bureau for the Bureau of Labor Statistics; for a general overview, see Alexander Bick and Adam Blandin’s 2023 article, “Employer Reallocation during the COVID-19 Pandemic: Validation and Application of a Do-It-Yourself CPS,” in the Review of Economic Dynamics. The SWAA was first introduced by José María Barrero, Nicholas Bloom and Steven J. Davis in their 2021 working paper, “Why Working from Home Will Stick”; this survey has been used for multiple papers by these authors.

Our key conceptual notion of WFH is a “WFH workday,” defined as a day in which some work was done for pay but the worker did not commute for that work. Based on this definition, we consistently construct two measures of WFH across these datasets:

1. The WFH Only Rate: The share of workers who work from home every workday in a given week; that is, the share of workers who are fully remote during that week.In the ACS, individuals are classified as WFH only if they usually work from home in the previous week, which captures individuals who work from home every workday or on the majority of workdays in the previous week.
2. The WFH Share: The share of all workdays that are WFH among all workdays in a given week. This includes WFH workdays by those who are hybrid workers during that week.

We can construct the WFH only rate for all datasets except the ATUS, and the WFH share for all datasets except the ACS and CPS. Given the broader coverage across datasets, this blog post focuses primarily on the WFH only rate.

The Persistent Rise in WFH

The first figure shows that the WFH only rate increased sharply after the COVID-19 outbreak in early 2020. While it has declined slowly since then, it remains well above prepandemic levels.

WFH Only Rate for All U.S. Workers

SOURCES: U.S. Census Bureau, IPUMS, WFH Research, Real-Time Population Survey and authors’ calculations.

NOTES: ACS=American Community Survey, CPS=Current Population Survey, RPS=Real-Time Population Survey, SIPP=Survey of Income and Program Participation, and SWAA=Survey of Working Arrangements and Attitudes. See Endnote 5 for more information on the data.

In February 2020, just before the pandemic, the WFH only rate was 5% in the ACS, 7% in the RPS, and 11% in the SIPP.We compute moments on the annual level in the ACS and SIPP and plot these in July of the corresponding year. We also plot data from the 2019 releases of the ACS and SIPP in February 2020. In addition, we omit the 2020 ACS release because we cannot distinguish prepandemic months from months after the pandemic started. The 2020 SIPP data uses responses from April through December 2020. In May 2020, this rate rose to 32% in the RPS, and although it gradually declined to 19% by December 2020, it has remained at elevated levels. For example, in June 2024, the RPS was 12%, 1.6 times higher than its prepandemic value. At the same time, the ACS and SIPP rates were also higher than their prepandemic level as of the last observations available, which were in 2023 and in 2022, respectively.

While the availability of WFH only rates for CPS and SWAA began in 2022, their movement is insightful. The SWAA rate was 20% in June 2024, down just four percentage points from 24% in January 2022, while the CPS rate bumped up from 10% in October 2022 to 11% in June 2024. The elevated WFH rates derived from these five surveys suggest that present work arrangements reflect a lasting shift to WFH rather than a temporary response to emergency conditions.

Why Do Datasets Disagree About Levels of WFH?

While the trends in WFH are similar across datasets, the actual levels of WFH differ considerably, as seen in the first figure. Estimates are consistently highest in the SIPP and SWAA. The next figures show the primary source of this disagreement: WFH rates among the self-employed (second figure) vary much more than those of employees (third figure).

WFH Only Rate by Employment Status: Self-Employed

WFH Only Rate by Employment Status: Employees

SOURCES FOR BOTH FIGURES: U.S. Census Bureau, IPUMS, WFH Research, Real-Time Population Survey and authors’ calculations.

NOTES FOR BOTH FIGURES: ACS=American Community Survey, CPS=Current Population Survey, RPS=Real-Time Population Survey, SIPP=Survey of Income and Program Participation, and SWAA=Survey of Working Arrangements and Attitudes. See Endnote 5 for more information on the data.

As shown in the second figure, in 2022, the WFH only rate among the self-employed is most similar in the ACS, CPS and RPS, with their averages ranging from 14% to 23%, while the SIPP and SWAA estimates averaged above 35% and 51%, respectively. In contrast, WFH estimates for employees are more closely aligned, with averages ranging from 10% to 15% across datasets in 2022.

Differences in WFH estimates for the self-employed likely reflect some combination of (1) different definitions of employment or self-employment, and (2) unobserved differences in the sample of self-employed workers between surveys. Notably, in a companion paper,See “Heterogeneity in Work from Home: Evidence from Six U.S. Datasets.” Working Paper 2024-038A, Federal Reserve Bank of St. Louis, December 2024. we investigate how WFH levels and trends vary by demographic groups (such as sex, age and education), firm characteristics, and geography since the pandemic, and do not find comparably dramatic differences across datasets along any of these dimensions. An upcoming blog post will examine the potential capacity for WFH among different industries and whether these industries took advantage of WFH before and after the pandemic.

The Role of Hybrid Work

Unlike the WFH only rate, our second WFH measure—the WFH share—includes hybrid work arrangements. In our research paper, we show that during the first year of the pandemic, the increase in the WFH share was almost entirely driven by the rise in the WFH only rate. However, hybrid arrangements have since contributed significantly to the overall increase. By June 2024, the WFH share in the RPS was 23%, about 1.6 times higher than its prepandemic level. Approximately half of that increase can be attributed to the higher WFH only rate, while around one-quarter is due to the increase in WFH days among hybrid workers, which rose from about 2 in 5 workdays to 2.5 in 5 workdays.

Conclusion

Our findings show broad agreement on WFH patterns across datasets, with rates consistently elevated above prepandemic levels. This persistence suggests WFH has become an enduring feature of the U.S. workforce.

Higher WFH rates bring notable implications. For employers, hybrid models may require rethinking office space, potentially reducing demand for commercial rentals. This effect extends to local economies beyond commercial rentals by reshaping housing and transit demands, especially in commuter-dependent cities. As a concrete example, previous research by Alexander Bick, Adam Blandin, Karel Mertens and Hannah Rubinton shows that more than half of the increase in interstate migration since the pandemic can be accounted for by the increase in the WFH only rate.

For employees, remote work offers flexibility and cost savings, though it can limit in-person training, mentorship and networking opportunities, especially for new workers. These challenges, along with concerns about collaboration and company culture, are often cited as reasons to encourage a return to the office.

In summary, although the U.S. public health emergency for COVID-19 ended in May 2023, WFH remains far more prevalent than before the pandemic. This enduring trend suggests that while WFH may not persist at current levels, it is likely to remain well above prepandemic rates, with both advantages and challenges shaping how we work and the communities in which we live.

Notes

1. For details, see “Measuring Trends in Work from Home: Evidence from Six U.S. Datasets.” Working Paper 2024-023B, Federal Reserve Bank of St. Louis, September 2024.
2. Other papers have compared WFH estimates in multiple surveys, including Alexander Bick, Adam Blandin and Karel Mertens’ article, “Work from Home before and after the COVID-19 Outbreak,” in the American Economic Journal: Macroeconomics; Erik Brynjolfsson, John J. Horton, Christos Makridis, Alexandre Mas, Adam Ozimek, Daniel Rock and Hong-Yi TuYe’s 2023 working paper, “How Many Americans Work Remotely? A Survey of Surveys and Their Measurement Issues”; José María Barrero, Nicholas Bloom and Steven J. Davis’ article, “The Evolution of Work from Home,” in the Journal of Economic Perspectives; and José María Barrero, Nicholas Bloom, Shelby Buckman and Steven J. Davis’ 2024 slide presentation, “How Much Work from Home Is There in the United States?” (PDF). However, our paper involves more datasets and a longer time horizon.
3. The RPS is a nationwide survey that asks the same core questions and follows the same timing and structure as the Current Population Survey (CPS), the monthly labor force survey conducted by the U.S. Census Bureau for the Bureau of Labor Statistics; for a general overview, see Alexander Bick and Adam Blandin’s 2023 article, “Employer Reallocation during the COVID-19 Pandemic: Validation and Application of a Do-It-Yourself CPS,” in the Review of Economic Dynamics. The SWAA was first introduced by José María Barrero, Nicholas Bloom and Steven J. Davis in their 2021 working paper, “Why Working from Home Will Stick”; this survey has been used for multiple papers by these authors.
4. In the ACS, individuals are classified as WFH only if they usually work from home in the previous week, which captures individuals who work from home every workday or on the majority of workdays in the previous week.
5. We compute moments on the annual level in the ACS and SIPP and plot these in July of the corresponding year. We also plot data from the 2019 releases of the ACS and SIPP in February 2020. In addition, we omit the 2020 ACS release because we cannot distinguish prepandemic months from months after the pandemic started. The 2020 SIPP data uses responses from April through December 2020.
6. See “Heterogeneity in Work from Home: Evidence from Six U.S. Datasets.” Working Paper 2024-038A, Federal Reserve Bank of St. Louis, December 2024.

Fed Working Papers Delivered Right to Your Inbox

Want an easy way to keep up with scholarly economic research from the Federal Reserve? Sign up today for Latest Working Papers from the Federal Reserve Banks, a monthly digest showcasing the academic work of economists from all 12 Federal Reserve banks.