The Evolution of Remote Work across Industries: From Potential to Practice

As we discussed previously, work from home (WFH) rose rapidly in the U.S. following the onset of the COVID-19 pandemic. Although WFH rates are down from their pandemic peak, they have stabilized well above prepandemic levels. For example, the share of those working from home all workdays rose from 7% just before the pandemic to 32% in May 2020 and has remained at about 12% since 2022.

This raises the questions: Why are some workers continuing to work from home when they did not before the pandemic? And why have others resumed commuting even though they worked from home during the pandemic? In this blog post, we focus on one key factor in understanding WFH variation across workers: the industry in which they are employed.This blog post is based on our December 2024 working paper, “Heterogeneity in Work from Home: Evidence from Six U.S. Datasets” (Federal Reserve Bank of St. Louis Working Paper 2024-038A).

Industries matter because job tasks vary widely across them, and some tasks are much easier to perform remotely than others. This variation in WFH feasibility, or potential, plays a crucial role in determining how much industries were able to pivot to remote work both during and after the height of the pandemic.

Measuring Potential and Actual WFH

Data in this analysis are from the Real-Time Population Survey (RPS), a nationwide survey that follows the same core questions, 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. The RPS closely aligns with labor market outcomes in the CPS, including hours worked, earnings distributions and industry composition of workers.For a general overview of the RPS, see the July 2023 Review of Economic Dynamics article “Employer Reallocation during the COVID-19 Pandemic: Validation and Application of a Do-It-Yourself CPS” by Alexander Bick and Adam Blandin. Additionally, in separate research, we found that WFH prevalence in the RPS closely mirrors findings from other surveys, such as the American Community Survey and the CPS.For details, see Federal Reserve Bank of St. Louis Working Paper 2024-023B, “Measuring Trends in Work from Home: Evidence from Six U.S. Datasets,” published September 2024.

A key element of our analysis is the estimate of WFH potential described by Jonathan Dingel and Brent Neiman,See Jonathan I. Dingel and Brent Neiman’s September 2020 Journal of Public Economics article, “How Many Jobs Can Be Done at Home?” which identifies the share of jobs in an industry that can plausibly be done entirely from home.These industries, which are identified by NAICS code, are agriculture; mining; utilities; construction; manufacturing; wholesale trade; retail trade; transportation and warehousing; information; finance and insurance; real estate and rental and leasing; professional and business services; education; health care; arts, entertainment and recreation; accommodation and food services; other services; and public administration. This measure aligns with the WFH only rate in the RPS, which captures the share of workers working from home every workday during a given reference week.

The (Almost) Irrelevance of WFH Potential before the Pandemic

In February 2020, just before the COVID-19 outbreak was declared a global pandemic, most industries were operating far below their theoretical WFH potential. This can be seen in the following figure from the significant gap between the 45-degree line (the dashed line) and the line of best fit among industries (the solid line). The 45-degree line represents full realization of industries’ WFH potential. Although industries with high WFH potential, such as finance and information services, showed somewhat higher rates of actual WFH, the correlation between potential and actual WFH was still weak at 0.35. Even more striking, each percentage point increase in WFH potential translated to a meager 0.05 percentage point increase in actual WFH.

WFH Potential and Actual WFH by Industry in February 2020

SOURCES: Dingel and Neiman (2020), Real-Time Population Survey (RPS) and authors’ calculations.

NOTES: The figure plots the measure of WFH potential from Dingel and Neiman (2020) by industry on the horizontal axis and the actual WFH only rate in the RPS by industry on the vertical axis. Industries are categorized according to NAICS classifications and are plotted as bubbles with areas directly proportional to their share of employment in the 2019 American Community Survey release. The solid line is the line of best fit using these industry weightings, while the dashed line is simply the 45-degree line. The slope of the line of best fit and correlation coefficient are reported in the upper left-hand corner.

The Pandemic: A Shift toward Full Utilization of WFH Potential

By May 2020, shortly after the COVID-19 outbreak in the U.S., this dynamic had completely reversed. (See the next figure.) WFH potential and actual WFH were almost perfectly correlated (0.94). Moreover, a 1 percentage point increase in WFH potential translated to a considerable (0.7 percentage point) increase in actual WFH. Most industries were close to realizing their full WFH potential, with five industries, including accommodation and food services and retail trade, even exceeding it. This could reflect a situation, especially early in the pandemic, in which many workers in these industries who couldn’t work from home were laid off, while roles that could adapt to remote work did so quickly.

WFH Potential and Actual WFH by Industry in May 2020

SOURCES: Dingel and Neiman (2020), Real-Time Population Survey (RPS) and authors’ calculations.

NOTES: The figure plots the measure of WFH potential from Dingel and Neiman (2020) by industry on the horizontal axis and the actual WFH only rate in the RPS by industry on the vertical axis. Industries are categorized according to NAICS classifications and are plotted as bubbles with areas directly proportional to their share of employment in the 2019 American Community Survey release. The solid line is the line of best fit using these industry weightings, while the dashed line is simply the 45-degree line. The slope of the line of best fit and correlation coefficient are reported in the upper left-hand corner.

These data highlight the adaptability of industries under crisis conditions. WFH became a crucial strategy for business continuity and worker safety, but at the time it remained unclear whether such high levels of WFH would continue once the immediate crisis had passed.

Postpandemic WFH Trends: A Partial Reversion

By 2022, the more immediate impact of the pandemic on WFH had subsided, but even as late as mid-2024,The 2024 data reflect the pooled results of the April and June RPS. the relationship between WFH potential and actual WFH had not returned to its prepandemic state. Although the correlation between WFH potential and actual WFH dropped to 0.54 from 0.94, it remained well above the prepandemic value of 0.35. Each percentage point increase in WFH potential was associated with a 0.12 percentage point increase in actual WFH—more than double the prepandemic rate (February 2020) but far below the pandemic peak (May 2020).

WFH Potential and Actual WFH by Industry in 2024

SOURCES: Dingel and Neiman (2020), Real-Time Population Survey (RPS) and authors’ calculations.

NOTES: The figure plots the measure of WFH potential from Dingel and Neiman (2020) by industry on the horizontal axis and the actual WFH only rate in the RPS by industry on the vertical axis. Industries are categorized according to NAICS classifications and are plotted as bubbles with areas directly proportional to their share of employment in the 2019 American Community Survey release. The solid line is the line of best fit using these industry weightings, while the dashed line is simply the 45-degree line. The slope of the line of best fit and correlation coefficient are reported in the upper left-hand corner. We use 2024, with data pooled from the April and June RPS, as a postpandemic baseline to distinguish the short-run impact of the pandemic on the WFH only rate from longer-run trends.

High-potential industries like finance and information services continue to embrace WFH, but many others have scaled back remote work arrangements. Education is a particularly striking example: Despite a WFH potential above 80%, this industry has almost entirely returned to in-person work. This shift likely reflects productivity challenges associated with remote learning; in-person environments are generally considered more effective for students, especially younger ones.

Lasting Impact and Implications of the Pandemic on WFH

These findings suggest that WFH potential became a strong predictor of actual WFH during the emergency conditions triggered by the pandemic, but it is only weakly predictive of actual WFH under normal conditions. While high-potential industries can pivot to remote work when necessary, many reverted to in-person operations as conditions stabilized. This pattern may suggest that WFH potential acts as an upper bound for actual WFH during crises rather than a standard benchmark for typical operations.

Nonetheless, WFH potential is now a stronger predictor of actual WFH than it was before the pandemic. Many employers have invested in technology and processes to support long-term remote work, perhaps recognizing that WFH can be just as effective as in-office work for certain roles. Consequently, while not all industries will sustain the high levels of remote work seen during the pandemic, the experience has permanently altered workplace dynamics, making flexible work arrangements a viable option in many sectors.

Notes

1. This blog post is based on our December 2024 working paper, “Heterogeneity in Work from Home: Evidence from Six U.S. Datasets” (Federal Reserve Bank of St. Louis Working Paper 2024-038A).
2. For a general overview of the RPS, see the July 2023 Review of Economic Dynamics article “Employer Reallocation during the COVID-19 Pandemic: Validation and Application of a Do-It-Yourself CPS” by Alexander Bick and Adam Blandin.
3. For details, see Federal Reserve Bank of St. Louis Working Paper 2024-023B, “Measuring Trends in Work from Home: Evidence from Six U.S. Datasets,” published September 2024.
4. See Jonathan I. Dingel and Brent Neiman’s September 2020 Journal of Public Economics article, “How Many Jobs Can Be Done at Home?”
5. These industries, which are identified by NAICS code, are agriculture; mining; utilities; construction; manufacturing; wholesale trade; retail trade; transportation and warehousing; information; finance and insurance; real estate and rental and leasing; professional and business services; education; health care; arts, entertainment and recreation; accommodation and food services; other services; and public administration.
6. The 2024 data reflect the pooled results of the April and June RPS.