The Official Story
The stated case for return-to-office mandates is almost always some version of the same claim: spontaneous hallway conversations, mentoring that "just happens" near a desk, a culture that erodes without shared physical space. It's a comfortable argument for one specific reason: it can't be falsified with a spreadsheet. Nobody can produce a number for "how much culture was lost" the way they can produce a number for revenue, defect rate, or ticket resolution time. That's not proof the argument is wrong. It is proof the argument was never designed to be checked.
Compare that to what companies can measure and mostly choose not to publish: whether the teams that went back to the office actually shipped more, resolved incidents faster, or retained customers better than the teams that stayed distributed. When that comparison does surface, it rarely favors the office as cleanly as the announcements imply. The team-spirit argument survives less because it's demonstrated and more because it's unfalsifiable.
A Cheaper Story Nobody Tells
Here is the part that gets left out of nearly every recall announcement: home office was, for the company, structurally cheaper. Not marginally, structurally. Every seat that isn't occupied five days a week is a seat that doesn't need to be lit, heated, cooled, cleaned, or fed.
| Cost Category | Effect of Distributed Work |
|---|---|
| Office space / rent per headcount | Lower occupancy needed, or unrealized downsizing potential |
| Electricity and heating | Reduced building-wide consumption on partially empty floors |
| Canteen / catering subsidies | Fewer subsidized meals served per day |
| Cleaning and facility services | Reduced frequency and scope of service contracts |
| Desks, printers, on-site IT support | Partially shifted onto the employee's own home setup |
None of this is speculative; it's why "hot-desking" and office footprint reduction became standard finance-department vocabulary the moment distributed work proved viable. A company that genuinely optimized for cost would have kept as much of this as it could get away with, and quietly resized the real estate around it. Instead, a lot of the same companies signed new leases, rebuilt open-plan floors, and then mandated attendance to justify the expense after the fact. That's not cost discipline. That's a sunk-cost argument dressed up as a culture argument: we already pay for the building, so use it, restated as it's about the team.
The Availability Nobody Measured
Set the cost side against the productivity side, and the case gets harder still to make honestly. Since switching to home office, working hours for a lot of people in technical roles didn't go down. They went up, and the shape of the workday changed in a way that should matter to anyone arguing the office produces more output.
The old call had a built-in boundary: "we have a problem, can you help?" "Yes, maybe in 45 minutes, I'm still stuck in traffic outside Munich." The commute wasn't just dead time. It was the last reliable buffer between being reachable and not being reachable. Home office removed that buffer without anyone deciding to.
An idea at 23:00 now gets opened in a laptop instead of waiting for the next morning's drive in. A message that used to sit until someone physically arrived at a desk now gets answered from the kitchen table. None of this shows up as an explicit policy anywhere; it's just what happens when the boundary between "at work" and "not at work" stops being a building and a commute. If the honest complaint were "people aren't working enough at home," the data most technical managers actually have access to, response times, commit timestamps, incident-response latency, tends to say the opposite.
What Presence Actually Proves
Here's the more honest mechanism, rarely stated out loud: a person visible at a desk is legible to a manager in a way a person answering messages from home is not, regardless of what either person actually produced that day. Presenteeism, being seen working, has always been an easier signal for a manager to read than output, especially for managers who were themselves promoted for being visibly present rather than for what they shipped. Remote work didn't reduce anyone's actual output for most knowledge-work roles. It removed the visual proxy that a certain style of management has always substituted for measuring output directly.
That's the actual driver behind a lot of these mandates: not a demonstrated culture deficit, not a cost calculation, but a discomfort with not being able to see whether people are working, paired with a lack of better tools (or will) to measure whether they are. The office doesn't produce more collaboration. It restores a manager's ability to confirm, by looking, that someone is at their desk, which is a much older and much less reliable proxy for "working" than most executives would say out loud.
| Stated Reason | What It Actually Optimizes For |
|---|---|
| "Team spirit / culture" | Unfalsifiable, so it can't be contested with data |
| "Collaboration needs the office" | Visibility as a stand-in for trust |
| "We already pay for the building" | Sunk-cost framing that ignores achievable downsizing |
| (rarely said directly) "We can't tell if you're working" | Presenteeism as a substitute for measuring output |
The Honest Version Would Be Better
None of this is an argument that offices are worthless or that every remote arrangement is fine as-is; some roles and some team dynamics genuinely benefit from shared physical space, and that's a legitimate case to make on its own terms. The argument is narrower: the reasons usually given for blanket return-to-office mandates don't survive contact with the company's own cost structure or with the lived experience of the people being recalled. If leadership said plainly, "we don't have a reliable way to measure your output remotely, so we want you where we can see you," that would at least be a specific, debatable claim. Instead, the justification stays soft and unmeasurable on purpose, because a claim about "team spirit" can't be checked, and a claim about control usually doesn't sound good enough to put in the announcement.