The two questions underneath

How they get tangled

Because an answer to the first appears to settle the second, and does not.

Suppose it were established tomorrow, beyond argument, that co-location produces a substantial benefit. Does it follow that employees should travel three days a week to produce it? No. It follows that there is a benefit and a cost, and the question of who bears the cost is untouched.

Suppose the opposite were established. Does it follow that nobody should come in? Also no — people might still want to, and a company might still value it for reasons that are not productivity.

The two questions are logically independent and conversationally fused.

What separating them does

Try the argument with them apart.

First question. Does being in the same room produce something? Probably yes, for some kinds of work, at some frequencies, and the specific claims are checkable in principle. On this record two of them could be checked today.

Second question. Given that it produces something, who pays for it? The travel time, the fares, the loss of flexibility, the childcare, the house moves — on this record, four of the five costs fall mostly on individuals.

Put that way, the disagreement is not about whether offices work. It is about whether the benefit, whatever it turns out to be, is worth what it costs and to whom.

That is a question about distribution, and it has the useful property of being negotiable. Benefits can be shared, costs can be offset, frequencies can be traded. None of that is available while the argument is about whether co-location does anything.

Why the fusion persists

Because separating them is unattractive to both sides.

Separated, the office side has to say out loud that the benefit is worth somebody else's time — which is a defensible position and an uncomfortable sentence. And the other side has to concede that the benefit may well be real, which feels like losing the argument when the argument has been conducted on whether it exists.

Both prefer to keep arguing about the first question, where each believes it is winning.

The one sentence that separates them

Available to anybody in the room, at any point:

Suppose you are right that it produces something real. What should it cost, and who should pay it?

It concedes nothing — suppose is doing the work — and it moves the discussion to the part that can actually be settled. On this record the one exchange that reached this question produced the only negotiated outcome in eleven weeks.

The third question nobody asks

There is a question behind both of the two, and it almost never surfaces: for whom?

Co-location does not produce one thing for everybody. The evidence that does exist points fairly consistently to the benefit being concentrated — larger for people new to a role, larger for work that requires frequent negotiation, smaller or negative for sustained individual work.

If that is right, then both of the main questions are being asked at the wrong level. The useful version is not "does being in the same room produce something" but "for whom, doing what, and how often".

Why the specific version is avoided

Because it produces an arrangement rather than a rule, and arrangements are harder to write down and harder to administer.

"Three days for everybody" fits in a sentence. "Six weeks alongside the team for anybody new, two days a week for teams doing design work, and nothing for anybody else unless they want it" does not, and it requires somebody to make judgements that a number does not.

That is a real administrative cost and it is the honest reason most organisations prefer the rule. It is worth saying out loud, because when it is unstated the specific version looks like something that was considered and rejected on the merits, and usually it was not considered at all.