Federal workers went back to the office. Productivity went down

The U.S. federal government just ran the largest return to office experiment on the planet, on more than a million people, with no pilot phase and no opt out. The results are now sitting in a peer reviewed journal and a 7,463 person survey, and neither one is flattering.

In January 2025, a governmentwide order ended the telework arrangements that had let much of the executive branch work from home, at least part of the week, for years. There was no phased rollout to watch for problems first. Everyone went back more or less at once. Which, if you like natural experiments, is a gift: a single policy, a single start date, applied to a workforce big enough that the numbers mean something.

The number that should worry any manager, federal or not

Federal News Network ran its own survey in May 2026 and got 7,463 responses from people who identified as federal employees. About 76 percent said they were somewhat or much less productive than before the mandate. Twenty one percent said their productivity hadn't changed. Three percent said it had gone up. That is not a close call, and it is not a small sample either.

The written answers are the part that sticks. One respondent put it plainly: "People are much less able to focus, be productive, get the resources they need and perform their job duties versus pre-RTO." Another was more specific about what the actual trade looked like: "I spend time and money commuting longer to do exactly what I do from home, with less productivity, more background noise, more interruptions, worse amenities, worse food and no one I actually work with." That last clause is doing a lot of quiet damage to the whole premise of mandatory proximity: showing up to a building that doesn't contain the people you work with is just a longer, worse commute to the exact same laptop.

The peer reviewed version says the same thing, more carefully

A study published in April 2026 in Public Personnel Management, by Geneviève Morin, Carey Doberstein and Étienne Charbonneau, looked at the same mandate through three waves of survey data collected in April 2024, January 2025 and March 2025. Because that window straddles the mandate itself, the researchers could watch sentiment shift in something close to real time rather than reconstructing it after the fact from memory.

Their finding is less about a number and more about a mechanism: the order disrupted routines people had built over years, took away a degree of control over how and where they worked, and layered new emotional and logistical strain on top of that loss. Losing control of your own schedule, it turns out, costs you something even before you factor in the commute or the parking. Other reporting on the same period backs this up with the mundane details that never make it into a policy memo: parking lots that can't fit the returning headcount, buildings missing basic supplies, offices built for a workforce a third smaller than the one now showing up.

What actually broke, and what didn't

It's tempting to read this as proof that offices are bad, but that's not quite the finding either. What both sources describe is a sudden loss of autonomy landing on top of infrastructure that hadn't been rebuilt to receive everyone back. Nobody planned for the parking. Nobody restocked the building. The policy shipped; the logistics didn't. That's a rollout failure wearing a return to office costume.

We made a version of this argument when we covered the study on remote days and energy variability: what predicts real progress isn't where you sit, it's how much of your day gets eaten by things you didn't choose. A mandate that removes your choice about where you sit, then also fails to provide parking, is removing control on two axes at once. It would be strange if morale held up.

The broader trend we tracked in our look at where U.S. remote work actually settled in 2026 shows the private sector mostly landing on a hybrid compromise rather than a full mandate. The federal government picked the harder version of this policy, all at once, for everyone, and the well being data is the bill for that choice.

The honest counterargument

A self-selected online survey has a real bias problem: people who are frustrated are more likely to click through and answer than people who are quietly fine, so 76 percent is very likely an overstatement of the true population figure. Fair criticism, and Federal News Network says as much in its own reporting. The peer reviewed study is smaller in scope and qualitative rather than a clean before and after productivity measurement, so it can describe the mechanism without proving the magnitude.

And to be fair to the other side of this argument entirely: plenty of federal work genuinely benefits from people being in the same room, particularly for newer employees who need mentorship a video call handles badly. Nobody in either source is arguing that offices are pointless. The argument is narrower and less exciting: a mandate handed down without the logistics to support it, and without any say for the people living under it, produces worse output than the arrangement it replaced. That's a management failure with a return to office label on it, not an argument against buildings.

The part that applies to you even if you don't work for the government

Most of us aren't federal employees, and most workplaces would never attempt a mandate this blunt. But the underlying belief behind it, that putting people in the same physical space automatically produces accountability and visibility, is one a lot of managers still hold quietly, and it's the same belief this data just took apart. Proximity doesn't make ownership visible. It just makes people sit closer to each other while ownership stays exactly as fuzzy as it always was.

We've argued before, in Clear Ownership Without Status Meetings and Team Accountability Without Micromanagement, that what actually surfaces stalled work is a system that makes ownership visible on its own, not a floor plan. A board where every action shows who has the ball does that whether the person holding it is at a desk three feet away or working from a kitchen table across the country. The federal government just spent a great deal of political capital, and apparently a fair amount of morale, proving the floor plan alone doesn't do that job.

Sources

Federal News Network, "After returning to the office, federal employees report low productivity, morale", July 2026. Geneviève Morin, Carey Doberstein and Étienne Charbonneau, "Mandated Return-to-Office Policies and Federal Employee Well-Being: Evidence From Three Waves of U.S. Survey Data", Public Personnel Management, 2026. Additional context via Federal News Network's on the ground reporting, March 2025. Figures accurate as of September 2026.