Updated on September 22 at 12:30 p.m. ET
About six months ago, a colleague asked me to guess what percentage of Americans were still working from home. I was still spending eight hours a day making calls just a few feet from my fridge. So were most of my friends. Maybe 40 percent? I guessed. I was off by half. Twenty-one percent of employed Americans were still teleworking as of March 2021; the other 79 percent were leaving their home like the old days.
This was a case of Beltway bias on my part, and I should have known better. My parents, in Iowa, had gone back to in-person work at that point, and many of my friends in the state had teleworked for only a few weeks at the beginning of the pandemic—if at all. Last week, The Atlantic commissioned a poll from Leger, asking Americans to estimate how many people had worked from home during the pandemic. The results weren’t entirely surprising: Those working remotely tended to overestimate how many other people were doing the same.
Seventy-three percent of survey respondents who had teleworked because of the pandemic guessed that at least half of American employees had done the same. But the actual number of people who worked remotely because of COVID-19 was, at its highest point, roughly 35 percent, way back in May 2020. Let’s skip ahead to last month: About 90 percent of surveyed respondents who worked from home in August because of the pandemic guessed that at least 40 percent of American workers did too. In reality, only 13.4 percent worked from home in the final month of summer.
Part of the reason for the discrepancy comes down to basic psychology. Human beings tend to believe that other people are like us, that our thoughts and opinions are more common than they actually are, the sociologists I interviewed for this story told me. But when I answered that question so wrongly back in March, I felt a pang of embarrassment; I was out of touch. “People don’t have a great sense anymore of what the lives of others across the economic divide look like,” Jonathan J. B. Mijs, a sociology professor at Boston University, told me.
So much of the media coverage throughout the COVID-19 pandemic has focused on the phenomenon of remote work and how it’s here to stay. For months, we were drowning in articles about ergonomic desk chairs, Zoom fatigue, and the comfiest teleworking sweatpants. (I wrote a story last spring about how much it sucks to work from home in a group house.) All of those articles were relevant for a certain segment of the population—one that is quite likely to read this publication. But the prevalence of these stories also helped obscure the reality that most Americans were still going to work in person, risking their life and the health of their families. Nurses, line cooks, delivery drivers, assembly-line workers, grocery-store clerks, and plumbers simply don’t have the option of getting an extra hour of sleep and hopping on Zoom in their pajamas.
Americans with the option to telework are much more likely to be highly educated; they’re also more likely to be upper-income workers, rather than low- or middle-income. And the professionals who telework tend to be concentrated in urban centers. That all might seem intuitive. But it’s difficult for most people to perceive that disparity, given the increasing socioeconomic segregation in the United States: Americans are more likely to live and work and pal around with others of similar educational backgrounds and job types than they used to be, the sociologists I spoke with told me. Online social networks, which in theory should help us burst out of our social bubbles, have instead tended to reinforce them.
Exclusively socializing with people who have similar lifestyles has consequences. Highly educated white-collar workers who spend a lot of time with like-minded people are more likely to underestimate income disparities and the structural barriers many people face, Mijs said. Being trapped in a class bubble can also affect Americans’ political and policy preferences. In her 2016 book, Strangers in Their Own Land, the sociologist Arlie Hochschild described how feeling misunderstood breeds resentment and political division.
Americans were not so divided by income 40 or 50 years ago, when neighborhoods, churches, and social organizations, while less racially diverse, were more socioeconomically diverse. Cross-class mixing happened often, but much of that mixing disappeared when Americans began to group themselves based on identity. That has happened in other wealthy countries too. “Economic clustering became lifestyle clustering became political clustering,” Bill Bishop, the author of the 2004 book The Big Sort, told me. One of the outcomes of that clustering is that most people don’t know how other Americans are living.
Hochschild believes mutual understanding is still possible between Americans. She offered some advice for people—like me—who assumed incorrectly that more Americans were living like them. “Ask yourself,” she said, “Why didn’t I know this?”
* This article has been updated to clarify that the U.S. Bureau of Labor Statistics figures refer to employed Americans.