Large metropolises
Which leads us to those urban agglomerations. It was a report on the share of working from home by commuting zone by Adam Ozimek and Eric Carlson of the Economic Innovation Group, a Washington think-tank, that inspired my dive into this data. I have chosen to sort by the more familiar designations of metropolitan areas (a county or group of counties with a core urban area of at least 50,000 people) and micropolitan areas (the same but with a core between 10,000 and 50,000).
One micropolitan area, Truckee and Grass Valley in the mountains and foothills north and east of Lake Tahoe, makes the top 15. So do five metropolitan areas of fewer than a million people (Boulder, Durham-Chapel Hill, Ann Arbor, Trenton-Princeton and Corvallis), while the country’s five largest metro areas (New York, Los Angeles, Chicago, Dallas and Houston) are missing. But the rankings are still dominated by large metropolises.
The pandemic exodus from big cities was not a chimera. Urban populations truly did decline, and demand for houses in exurbs and mountain resorts really did skyrocket. Yet most of the white-collar workers who had clustered in big cities and expensive suburbs before the pandemic stayed put even as they stopped going into the office.
One possible interpretation is that there remains a lot of pent-up demand for housing far from traditional job centres — the exodus has barely begun. Another is that cities and close-in suburbs have attractions beyond just proximity to the office.
Bloomberg Opinion. More stories like this are available on bloomberg.com/opinion
The mass shift to remote work during the pandemic allowed people with professional and management jobs to do them effectively from mountaintop aeries, beachfront cottages and exotic foreign locales. Mainly, though, it seems to have enabled residents of big-city neighbourhoods and close-in suburbs to avoid going to offices that in some cases were within walking distance of their homes.
These numbers are from the 2021 edition of the American Community Survey, a sort of minicensus that the US Census Bureau sends out to 3.5-million households annually. They come in response to the multiple-choice question, asked of household members who were already reported to have performed paid work the previous week: “How did this person usually get to work last week?”
“Car, truck or van” remained the box checked most frequently, with an estimated 75.6% of US workers getting to their jobs that way in 2021. But the percentage working from home, which had been rising slowly for years, jumped to 17.9% in 2021 from 5.7% before the pandemic in 2019. In some places, as should already be apparent from the above chart, it jumped much more than that.
The places I have chosen to focus on here are census public use microdata areas (PUMAs), which divide the nation into units big enough (population of more than 100,000, usually) for the Census Bureau to release microdata from which everyone can compile their own custom estimates and tables without compromising the privacy of respondents. These particular numbers are derived not from the microdata (which is not out yet and is released only as five-year averages in any case) but from a table published by the Census Bureau in mid-September that I have seen referenced by one pseudonymous Twitter user but otherwise has not received the ample attention it deserves.
A big advantage of PUMAs is that they divide diverse big cities into more coherent parts while adding together suburbs (and small towns and rural areas) that alone do not have enough people for the Census Bureau to release their data. Sorting by PUMAs makes clear that working from home was most prevalent quite close to office centres last year. Central Washington, DC, in the Census Bureau’s accounting includes the nation’s most famous home office, at the White House, as well as the Capitol Mall and the neighbourhoods north and south of it, from Foggy Bottom to DuPont Circle to Adams Morgan to Shaw to the new residential developments of Southwest and the Navy Yard.
Interesting hotbeds
In San Francisco, the Inner Mission & Castro and North Beach & Chinatown PUMAs are close to downtown and South of Market & Potrero encompasses it, as does the Queen Anne & Magnolia PUMA in Seattle. Even the more suburban areas in the chart above are home to corporate headquarters and other large office complexes, with the nation’s three largest corporations by market capitalisation based in Cupertino (Apple), Redmond (Microsoft) and Mountain View (Google parent Alphabet).
The Census Bureau also released work-from-home estimates for cities and census-defined places of 65,000 residents or more. The top-15 ranking of these makes clear why the listing by PUMAs is more informative, I think, but also highlights some more interesting work from home (WFH) hotbeds in the suburbs of Seattle, Washington, San Francisco and Atlanta.
When I posted a version of this chart on Twitter a few weeks ago (based on a different Census Bureau table that had data only for fewer, bigger cities), many people reacted with surprise that cities and close-in suburbs dominated the rankings rather than exurbs and mountain resorts. That is partly because there are not many exurbs or mountain resorts with more than 65,000 people, though the PUMA statistics covering such places also do not reveal any sky-high WFH numbers.
There surely are parts of scenic, exurban Sonoma and national-park-neighbouring Bozeman and Jackson (which is in Teton County, Wyoming) where most people work from home, but to get near the 100,000-population PUMA threshold, those have to be tallied alongside other neighbourhoods and rural areas where most residents have hands-on jobs. In large urban agglomerations so many people are working from home that they can dominate entire cities of more than 65,000 people or PUMAs of more than 100,000.