Here is the number the headlines keep leaving out: 0.4.
That is how far the American work-from-home share is projected to fall once every return-to-office mandate companies currently plan has actually landed: from 21.2 percent of paid workdays down to 20.8 percent. Less than half a percentage point. Not a wave. A rounding error. The figure comes from Stanford's Survey of Business Uncertainty, a monthly poll of more than a thousand executives across US industries, fielded in February 2025 by Nicholas Bloom and Steven Davis of the Stanford Institute for Economic Policy Research with the Federal Reserve Bank of Atlanta. Their own summary of it is blunt: the headline-grabbing mandates "will barely move the needle on WFH." And then, in case anyone missed the point, "working from home is here to stay."
That number is worth keeping in mind, because the panic it debunks is expensive. Every few weeks a big employer announces that everyone is coming back five days, the story travels, and by lunchtime I have three reader emails asking whether they should quit before their own company does the same. They are making a career decision off a vibe. The vibe and the data have quietly stopped agreeing.
Few employers order it, and attendance barely moves
Start with who is actually issuing these orders. In the Stanford survey, only 12 percent of executives with hybrid or fully remote staff said they plan a return-to-office mandate in the year ahead. Amazon, JPMorgan, AT&T: real, loud, and a small, concentrated group of large employers. They are not the economy. The other 88 percent of companies with flexible staff are holding steady, because they can see what a full mandate costs them. Stanford's Survey of Working Arrangements and Attitudes asked American workers in December 2024 what they would do if ordered fully onsite: only about 44 percent said they would comply. Most of the rest said they would start looking for a job that let them keep working from home, and return only if they could not find one; a smaller share said they would quit outright. Those are stated intentions, not resignation letters, but no manager who can read a spreadsheet wants to test them on their best people.
The gap between what companies say and what workers do is the whole story, and someone has measured it. The Flex Index, run by Scoop, tracks the actual in-office policies of thousands of US firms. By the third quarter of 2025 the average company required corporate staff in the office 2.87 days a week, up from the mid-2024 low of 2.49 days, a rise of about 15 percent. So yes, on paper the screws tightened. The share of firms allowing zero office days fell from 32 to 24 percent; the share demanding three or four days rose from 21 to 30 percent. That is the part that makes headlines. Here is the part that does not: over the same stretch, actual attendance moved between 1 and 3 percent. Policy shifted. Behavior mostly did not. Bloom's phrase for the attendance trend is "flat as a pancake."
This is what a settled equilibrium looks like, not a retreat. Full work-from-home days now account for roughly a quarter of all paid US workdays, up from about 7 percent in 2019, and the retreat from the pandemic peak leveled off years ago. Hybrid is the default: Gallup finds hybrid workers now spend about 2.3 days a week in the office, a little under half the week, and among remote-capable US employees, 52 percent are hybrid, 27 percent fully remote, and only 21 percent fully onsite. When Bloom, Davis and Jose Maria Barrero modeled what it would cost to force everyone back to 2019 norms, they got a drop in aggregate labor productivity of 0.2 to 1.1 percent and a hit to worker welfare of about 2.3 percent. The flexibility is not a perk companies are graciously extending. It is a bargain both sides re-signed because the numbers work.
The mandate is about control, not the office
Why the noise? Because the mandate is doing a job that has nothing to do with where you sit. In a December 2023 working paper, researchers at the University of Pittsburgh, Yuye Ding and Mark Ma, studied return-to-office mandates at S&P 500 firms. What they measured: mandates tend to follow stock declines rather than precede recoveries, and their before-and-after tests found significant drops in employee job satisfaction with no significant change in financial performance or firm value. What they infer from that pattern, and it is their interpretation, not a number in the data, is that the mandates track managers wanting to reassert control and looking for somewhere to pin underperformance, rather than any evidence that offices lift firm value. Read either way, a mandate is often a story a struggling company tells its shareholders. It photographs well. That is different from working.
Averages hide your bill, but your leverage is real
Now the caveat, because I slid past one like this once and regret it. In 2021 I told readers the fix for bad conditions was to move, that the market would reward them, and I watched enough of them land somewhere worse with a nicer title that I stopped saying it. So I will not tell you the aggregate calm means your situation is fine. If you are one of the people at a company that just went to five days, the national average is cold comfort; your commute got real while the country's did not. Averages hide the individual bill. What the data changes is not whether your specific mandate is happening. It is how much power you should assume you have when you respond to it.
And it is more than the panic implies. You are negotiating inside a labor arrangement that most employers have decided they want to keep, against a compliance record that shows attendance rules bend in practice all the time. That is leverage, and leverage is worth using calmly rather than spending on a resignation you send in a bad week.
Don't quit off a headline, negotiate the days
Here is the Monday move. Do not quit off a headline, and do not pre-comply with a mandate nobody has actually given you. Find your real number first: what your role requires in the office now, in writing, versus what you are currently doing. If there is daylight, that daylight is the arrangement most of your peers are quietly living in, and it is defensible. Then, if a change is coming, negotiate the specific days, not the principle. Ask for the two anchor days the team genuinely needs and offer to make those count, which is the trade a manager can actually say yes to. Argue the shared interest, coverage and collaboration, never fairness, because fairness loses and mutual interest wins.
The office is not empty and it is not full. It settled, somewhere around Wednesday. The only people who missed it were the ones writing the headlines.
