In May, Census Bureau survey-takers started asking Americans if they were working at home because of the coronavirus, and 35.4 per cent of employed persons said yes.(2) That percentage dropped in subsequent months, but as data released as part of Friday’s employment report from the Bureau of Labor Statistics indicate, it hasn’t been dropping by all that much lately.

This does not include those who were already working from home before the pandemic. Estimates of the size of this group range from the 5.7 per cent of U.S. workers who said they worked at home when the Census Bureau asked them about commuting patterns in 2019 to the upwards of 20 per cent who have reported working at home at least some of the time in recent years in two BLS surveys.

Yet another survey, conducted in April, found that 14.6 per cent of employed persons had already been working at home and 34.1 per cent had started doing so because of the pandemic, for a 48.7 per cent total — although it’s important to note that huge numbers of people who couldn’t work from home had just been furloughed or laid off, and thus weren’t counted in the denominator.

So one shouldn’t make too much of these exact percentages, which among other things don’t capture what is surely a shifting mix of at-home and office work. I’ve gone from working at home every day this spring and summer to going into the office occasionally, but I would still show up as a work-at-homer in the above chart. But the work-at-home numbers do reveal some important patterns and trends. For example, here are the major occupational categories with the highest percentage of workers reporting that they worked at home in September because of Covid.

The occupations with the lowest percentages were, not too surprisingly, building and grounds cleaning and maintenance; farming, fishing and forestry (both 2.1 per cent); and food preparation and serving (2.2 per cent). The rankings above do not perfectly match the rankings of major occupational categories by annual earnings, but there sure is a lot of overlap in the top 10.

Meanwhile, the ranks of occupations with the highest share of stay-at-homers have actually been reshuffled a bit since May.

Embedded Image

The stubbornly high work-at-home percentage of those in computer and mathematical occupations is at once unsurprising and revealing. Tech firms were among the first to recommend that employees work from home in early March and some, such as Twitter Inc., are preparing for this to be their new normal.

This is proving to be terrible economic news for the cities where their employees once went to the office every day (Twitter is headquartered in San Francisco) — and the same appears to be true for other high-stay-at-home occupations and the places where they’re concentrated.

Harvard University economist Raj Chetty has found that many of the ZIP codes hit hardest by the economic fallout of the pandemic are those where affluent stay-at-home workers have stopped spending money on local services. “The shock is most severe actually in the richest parts of the country and the richest neighborhoods in the country,” he recently told Bloomberg Businessweek. “It’s literally the people you were interacting with who I think are suffering the most.”

The BLS also reports stay-at-home percentages by industry. 

It might be surprising to see mining, quarrying, and oil and gas extraction on the list here, but oil and gas companies in particular are high-technology enterprises with lots of engineers and other skilled office workers on the payroll. In the other direction, information, which is where internet and software companies are classified, has its work-at-home percentage pulled down somewhat by the fact that it also includes telecom, broadcasting, film and television production, and movie theaters. The industry with the smallest share of at-home workers is private household services, at 3.5 per cent, followed by agriculture and related industries (4.2 per cent) and food services (4.8 per cent).

Here the matchup between stay-at-home percentages and earnings is somewhat weaker — and the industry sectors for which the BLS reports earnings monthly are a bit broader than those in the above chart. But the work-at-home big three of finance, professional services and information are all near the top in pay, too.

Note that educational services do not make the top 10 here (as a combined educational and health services, they come in 11th, at US$958.77 a week). And it is of course educators who have seen the biggest declines since May in the percentage of at-home workers, while most of those employed in finance and insurance and nearly half of those in professional and technical services remain at home.

Here’s the breakdown by educational attainment. The more years of formal education you have, the more likely you are to be working at home. Not a big surprise.

Finally, here it is by race and ethnicity. Blacks and Whites actually don’t differ much in their likelihood of working at home; Asians and Hispanics do.

There are, then, at least a few interesting twists to who is still working at home and who is not. But on the whole it is people with professional jobs, many of them concentrated in big cities, who are most likely still not to have returned to the office. Which is great for a lot of those people, but not for their cities.

(1) The Census Bureau conducts the monthly survey of 60,000 households that is then used by the Bureau of Labor Statistics to estimate the unemployment rate and other key labor-market indicators.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Justin Fox is a Bloomberg Opinion columnist covering business. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”