The Bureau of Labor Statistics published the American Time Use Survey June 25. One number is getting most of the attention: in 2025, more Americans worked from home than the year before. Up from 33.4% to 34.9% of full-time workers doing at least part of their job remotely on an average day.
The headline writes itself. RTO mandates hit a wall. Remote work is back.
I think that’s the wrong read.
The more interesting number is buried a little deeper. Required office days increased 12% from 2024 to 2025. Actual attendance increased 1 to 3%. That is not a story about remote work winning. That is a story about a decision that didn’t travel.
Every organization has a transmission system. Some move executive decisions through the organization with remarkable fidelity. Others lose clarity, consistency, and accountability at every layer.
This data is a picture of the second kind.
There’s a distinction I’ve come to think matters quite a bit in how organizations function. There are mandates. And there are decisions. They are not the same thing.
A mandate is a statement. A decision is a statement plus the infrastructure to implement it, the feedback loops to know whether it’s working, and the organizational capacity to surface what it’s actually costing. Most of what the last two years produced were mandates.
What the Gap Is Telling You
When required time goes up 12% and actual attendance goes up 3%, something happened between the executive conference room and day-to-day behavior. The policy was announced. Somewhere between that announcement and execution, the organization’s operating system produced a different outcome than leadership expected.
That gap could mean many things. The follow-through infrastructure wasn’t built. The people responsible for implementing the policy made different calculations than leadership anticipated. The stated rationale didn’t survive contact with competing priorities at the operational level. Probably some combination of all three.
A Baylor University study examining more than three million workers at 54 large technology and financial firms found that RTO mandates produced a 13 to 14% increase in employee turnover. Senior and highly skilled employees departed at significantly higher rates than junior employees. Female employees left at nearly three times the rate of male colleagues.
This research was available before many of these mandates were announced. Whether leaders concluded the tradeoff was acceptable or simply reached a different conclusion, it illustrates why organizations need mechanisms to evaluate the expected costs of major decisions before implementation.
What Success Looked Like
Before concluding the policy failed, leaders should ask a different question: “What did success look like?”
If reducing voluntary turnover, increasing in-person collaboration, lowering real estate costs, or reshaping the workforce were among the objectives, the evaluation has to be made against those goals, not attendance alone. A policy that produced significant senior departures may have failed on one dimension and succeeded on another, or it may have produced an outcome no one was willing to name out loud. Organizations that cannot answer the success criteria question clearly cannot distinguish between those possibilities.
That’s not a remote work problem. That’s a decision quality problem. And it’s one that exists before the first mandate is ever announced.
The Population That Stays
There’s a subtler outcome in this data that deserves more attention than it’s getting.
Work-from-home flexibility has remained heavily concentrated among highly educated, senior professionals. That’s not new. What’s new is what happens when you layer a compliance mandate over that reality.
Employees with the strongest external market alternatives often have greater ability to negotiate exceptions, move to firms with different postures, or leave altogether. The employees who comply fully are often the ones with fewer options.
The result is that the mandate shapes the workforce even if it never produces the culture the organization said it wanted. You end up with offices full of people who couldn’t leave, and an absence of the experienced employees whose judgment was the point. The policy succeeded as a filter when it was intended as a cultural tool.
That is a governance outcome. Not a management one.
Every Policy Becomes a Middle-Manager Decision
There is a layer of this problem that doesn’t get named often enough.
RTO policies are not administered by the executives who announce them. By the time a mandate reaches the people it’s meant to govern, it has passed through every layer of the organization, and at each layer, someone made an interpretation. Middle managers translate strategy into behavior. That translation is where most organizational policies either hold or dissolve.
A manager who doesn’t believe in the policy, or who would themselves prefer flexibility, or who has ongoing relationships with the employees they are being asked to hold accountable, faces a straightforward calculation. The cost of friction is visible and immediate. The cost of non-compliance is diffuse and distant. The rational move is to look the other way, absorb the gap quietly, and avoid generating the kind of information that creates problems for everyone.
That’s not a management failure in the conventional sense. That’s an information flow failure. When the people responsible for implementing a policy have competing incentives, the information leadership needs to assess whether the policy is working doesn’t get produced. Not because the systems failed. Because the humans in the middle made a reasonable choice not to generate it.
A policy administered by people who don’t believe in it isn’t a policy. It’s an announcement with a longer half-life.
How Faithfully Decisions Travel
Stanford economist Nick Bloom describes overall work-from-home levels as flat, and points to what he calls a composition effect: older, shrinking firms are cutting flexibility while younger, growing ones are quietly expanding it. The averages wash out. Only the mandates make the news.
Which means the organizations winning on talent right now are probably not the ones generating headlines. They’re the ones that made a decision, built the infrastructure around it, and have some way of knowing whether it’s working.
The gap between a 12% increase in required office days and a 1 to 3% increase in actual attendance isn’t just about remote work.
That’s the question worth asking. Not whether the mandate was right, but whether the organization was designed to know if it was right.
Policies reveal priorities.
The gap between policy and practice reveals the operating system.
The Inside Advantage 
Leave a comment