Before Covid-19, the share of full-time US workers who worked from home either fully or partially on an average day was 24%, according to data from the US Bureau of Labor Statistics.
We know there are many more in their ranks now. But we also know that 16.8 million Americans have filed for unemployment benefits in the past three weeks, many because their jobs (whether at restaurants, hotels, or retail shops, on theater stages, in factories, or in any number of other fields) are venue-specific; their work simply cannot be replicated at home.
So what share of US jobs can be done remotely?
A new working paper from Jonathan Dingel and Brent Neiman, both at the University of Chicago’s Booth School of Business, found that 37% of jobs in the US can potentially be done at home. These roles account for an estimated 46% of all wages, in line with previous findings that remote work is dominated by those with higher levels of education and income.
To classify jobs that can be done remotely, the researchers used responses from the US Department of Labor’s Occupational Information Network (O*NET) surveys covering “work context” (eg, work outdoors) and “generalized work activities” (eg, operating vehicles, mechanized devices, or equipment) to determine whether a job can be done at home or not. The findings were then merged with US occupational employment counts from the BLS.
The researchers identified characteristics of a job that would make working entirely from home not feasible, but did not take into consideration factors that could make working from home difficult. As a result, the researchers point out their estimate is an “upper bound” on what might be feasible. Of course, even if a job can be done remotely, not everyone can work as productively at home.
The share of jobs that can be done remotely varies across cities and industries. For instance, Dingel and Neiman found that 45% or more of all jobs in San Francisco, San Jose, and Washington DC could be performed at home, whereas fewer than 30% of jobs in Fort Myers, Florida; Grand Rapids, Michigan; and Las Vegas could be done at home.
Not surprisingly, the most “remotely-friendly” sectors were white-collar industries such as finance, along with professional, scientific, and technical services. But it turns out that the industry with the greatest share of jobs that can be done from home is educational services, at 83%. Indeed, the pandemic has turned education upside down, as universities and schools have migrated en masse to video conferencing in recent weeks.
The least remote-friendly jobs were found to be in the transportation and warehouse sectors (19%), construction (19%), retail trade (14%), agriculture, forestry, fishing and hunting (8%), and accommodation and food services (4%).
Dingel and Neiman note being able to identify which jobs cannot be performed from home may be useful for policymakers looking to target social insurance payments directly to those who need them most. Their findings also could be useful in predicting the overall health of the US economy during social distancing.