Remote work has settled at about a quarter of US workdays

After six years of headlines predicting that remote work would either take over or collapse, the data has done the least dramatic thing available. It stopped moving.

The Survey of Working Arrangements and Attitudes, run monthly by the WFH Research group and covering somewhere between 2,500 and 10,000 US residents aged 20 to 64, put working from home at roughly 25 percent of paid days in the US as of January 2026. That works out to about 1.3 days a week for the average worker. The figure has been broadly flat since 2025.

For scale: before 2020, the same measure sat somewhere around 5 to 7 percent. So remote work is down from its peak and still roughly four times its pre-pandemic level, which is a boring conclusion that satisfies nobody who has staked a public position on this.

The mandates were loud, the aggregate barely moved

That flat line is strange given the past eighteen months of return-to-office announcements, several of them from very large employers. The explanation is mostly composition. A handful of enormous companies calling people back five days a week generates enormous coverage and moves the national average by very little, because the national average includes a very long tail of employers who never announced anything and quietly kept doing what they were doing.

Where you do see movement is in the shape of hybrid rather than the amount of it. Flex Index data indicates that among US firms running a structured hybrid policy, the share requiring three days a week on site has risen to around two thirds, up from about half in mid-2024. The three-in, two-out week has become the default. Estimates of what fraction of all US firms are structured hybrid vary noticeably between summaries, somewhere in the high thirties to low forties percent depending on the cut, so treat any single number there as approximate.

Two different weeks, stapled together

Here is the part that we think gets underdiscussed. A three-two split does not give you a week. It gives you two weeks, running on different rules, and a Tuesday evening handover between them.

On office days your work is interrupt-driven and social. Things get decided in corridors and you leave with four verbal commitments that exist nowhere in writing. On home days your work is long-form and quiet, and the main risk is that you spend the morning reconstructing what you agreed to on Tuesday.

That reconstruction cost is real and it recurs twice a week, forever. It shows up in the focus numbers too. The Hubstaff benchmarking report we looked at earlier this month found hybrid teams reporting the lowest share of deep-focus hours of any working style, below both fully remote and fully in-office. One dataset does not settle anything, and the measurement has real limits, but the direction is consistent with what anyone working a split week will tell you.

This is a personal systems problem, not an HR one

You are not going to win the argument about your company's office policy. Very few people do. What you can control is whether your own work survives the transition between the two modes, and that is a smaller problem than it feels.

The requirement is unglamorous: one place, outside your head, where every open commitment lives, that you can open on either kind of day and immediately see what is yours and what is with someone else. Not a calendar, which tells you where you're expected to be. Not an inbox, which tells you what other people want. A board.

The specific habit that pays for itself is capturing on the office day rather than after it. Corridor commitments have a half-life of about four hours. If a thing gets promised in a hallway and it isn't written down before you sit back down, it exists only as a vague sense of unease on Thursday. We have made this case at length in The Art of the Brain Dump, and hybrid weeks are where it matters most.

The second habit is a five-minute scan at the end of each office day: what did I agree to, what did I hand to someone else, what am I now waiting on. That last category is the one that hybrid weeks destroy, because the person you handed it to was standing right there and now they're at home and neither of you remembers. Keeping it visible is exactly what the ownership-first approach is for.

The honest counterpoint

We should concede something. The in-office advantage in those focus figures is not nothing, and the case for co-location has never really been about focus anyway. It's about the cheap, unplanned conversations that surface problems nobody had scheduled a meeting about. That is a genuine benefit and it is genuinely hard to replicate on a call.

The trouble is that those conversations produce commitments, and commitments made in the highest-bandwidth setting available are the ones least likely to get written down. The office is good at generating work and bad at recording it. That's the gap your system has to cover, and no policy from HR is going to cover it for you.

Sources

WFH Research, Survey of Working Arrangements and Attitudes (SWAA), monthly updates through 2026. Hybrid policy composition per Flex Index reports. Focus-time comparison from Hubstaff's 2026 Global Trends and Work Benchmarks Report. Figures accurate as of August 2026.