Work From Home Is Here to Stay at 4 Times Pre-Pandemic Levels

The 28% That Refuses to Budge

Here’s a number that still surprises me: 28 percent. That’s the share of full paid workdays Americans now do from home, as of mid 2023. (Barrero et al., 2023). Before the pandemic, that figure was 7 percent. In the mid 1990s, it was roughly 3 percent. So we are not just above pre pandemic levels. We are at four times the 2019 rate and nearly ten times the rate of a generation ago.

The pandemic forced a massive experiment. But the conventional wisdom, repeated endlessly in 2020 and 2021, was that it would snap back. People would miss the office. Managers would demand returns. Productivity would suffer. None of that happened. Instead, the experiment became a permanent shift. The question is why.

José Barrero, Nicholas Bloom, and Steven J. Davis, three economists who have tracked remote work since the early days of COVID, published a comprehensive analysis in the Journal of Economic Perspectives that answers that question. Their paper, "The Evolution of Work from Home," draws on multiple large scale surveys of American workers and executives, collected continuously from 2020 through mid 2023. They did not just ask people how they felt. They asked what they actually did, day by day. The result is the clearest picture we have of why working from home is not a trend but a structural change.

Why Didn't It Snap Back?

The most common guess is that people simply prefer working from home. That is true, but it is not the whole story. Barrero, Bloom, and Davis found that the average worker values the ability to work from home two or three days a week as much as an 8 percent pay raise. That is a huge number. It means that for many workers, a hybrid schedule is not a nice perk. It is a deal breaker.

But preferences alone do not explain persistence. Companies also adapted. Early in the pandemic, managers had no playbook. They learned. By 2022, many firms had invested in new software, revised performance metrics, and changed meeting norms. The authors note that "the ongoing adaptation of managerial practices" is one of the factors that portends an enduring shift. Once a company has restructured its workflows around remote work, reverting to the old model is costly and disruptive.

There is also a selection effect. The workers who can work from home are not a random slice of the population. They are disproportionately college educated, high income, and employed in industries like tech, finance, and professional services. These workers have leverage. When they demand flexibility, companies often give in. And once one company in an industry offers hybrid schedules, competitors must follow to attract talent.

Who Actually Works From Home?

The numbers are not uniform. Barrero, Bloom, and Davis break down the data by age, sex, education, industry, and geography. Here is what they found:

• ▸Education is the strongest predictor. Among workers with a bachelor's degree or higher, about 40 percent of paid workdays are done from home. For those with a high school diploma or less, the figure is below 10 percent.
• ▸Age matters, but not in the obvious way. Younger workers (under 25) are less likely to work from home, partly because they hold jobs that require physical presence. Workers in their 30s and 40s are the most likely to work remotely, often because they have professional roles and the leverage to negotiate flexibility.
• ▸Parents, especially mothers, work from home more. The authors found that mothers with children under 18 are significantly more likely to work from home than women without children. This suggests that remote work is serving as a tool for managing the persistent asymmetry in childcare responsibilities.
• ▸Industry concentration is extreme. In information technology and finance, over 50 percent of workdays are remote. In retail, hospitality, and healthcare, the rate is below 10 percent. This is not a universal shift. It is a class divide expressed through geography.

Why the United States Leads the World

The United States has a higher work from home rate than almost any other country. The authors note this and offer an explanation that is both simple and uncomfortable: American workers have less legal protection.

In Europe, strong employment protections make it harder for companies to force workers back to the office, but they also make it harder for workers to demand flexibility. In the U.S., the labor market is more fluid. Workers can quit and take a remote job at a competitor. That threat forces employers to offer remote options. The result is a paradox: a country with weaker worker protections ends up with more worker friendly flexibility, because the market enforces it.

What About Productivity and Innovation?

This is where the debate gets heated. Many executives claim that remote work hurts productivity. The authors find a more nuanced picture.

On average, studies of remote work during the pandemic showed no productivity decline for most workers. Some even showed gains, especially in tasks that require deep focus. But the authors also point to evidence that remote work reduces collaboration and slows the pace of innovation. Patents, for example, are less likely to be filed by teams that are fully remote. New hires learn slower. Serendipitous interactions vanish.

The key insight is that the trade offs are not evenly distributed. For routine tasks that require concentration, remote work is a net positive. For tasks that require spontaneous idea exchange or mentorship, it is a net negative. The optimal arrangement, the authors suggest, is hybrid: two or three days in the office for collaborative work, the rest at home for focused work.

What the Research Does Not Prove

The Barrero, Bloom, and Davis paper is rigorous, but it leaves some big questions open. First, the data comes from surveys, not controlled experiments. People may overstate how much they work from home. They may also understate how often they are distracted or unproductive. The authors acknowledge this and use multiple checks to correct for bias, but it is not perfect.

Second, the paper does not measure long term career effects. If remote workers get promoted less often, or if they miss out on informal networks, the 8 percent wage equivalent they value might be an illusion. The authors note that early evidence suggests remote workers are promoted less, but the data is still thin.

Third, the paper focuses on the United States. The global picture is different. In countries with stronger labor protections, remote work rates are lower, but the quality of remote work may be higher because workers have more bargaining power to set terms. We do not know which model produces better outcomes over a decade.

What This Actually Means

The Barrero, Bloom, and Davis paper is not just a description of where we are. It is a map of where we are going. Here is what the findings imply for workers, managers, and policy makers.

• ▸If you are a college educated professional, you have enormous leverage. The ability to work from home is worth 8 percent of your salary. If your employer demands full time return to the office, you can negotiate for a raise or leave. The data shows that companies that refuse flexibility lose talent.
• ▸If you are a manager, stop trying to replicate the office at home. The worst model is a 9 to 5 schedule on Zoom. The best model is asynchronous communication for focused work and scheduled in person days for collaboration. The firms that adapt their management practices will outperform those that try to force the old model.
• ▸If you are a policy maker, recognize that remote work is a class issue. The benefits of work from home go overwhelmingly to the highly educated and high income. Lower income workers in retail, hospitality, and manufacturing get none of the flexibility and often bear the costs of increased traffic from those who do work remotely. Any policy response must account for this divide.
• ▸If you are a young worker, be careful. Remote work is great for productivity but bad for learning. New hires who work fully remote miss out on mentorship, informal feedback, and the social cues that teach you how an organization really works. The data suggests that hybrid is safer than fully remote for the first few years of a career.
• ▸If you are an executive, plan for 28 percent, not 7 percent. The pandemic did not create a temporary spike. It created a new equilibrium. Companies that design their real estate, IT, and management structures around a 28 percent remote rate will be more resilient than those that bet on a return to 2019. The authors found that business executives themselves anticipate modest increases in remote work over the next five years. The direction is clear.

The 28 percent figure is not a snapshot. It is a signal. The way we work has changed, not because of a virus, but because millions of people discovered something they value more than a commute. And they are not giving it back.