The evidence
The measurement nobody will make
Two claims on this record could have been settled with data already held. Neither was. The reason is not difficulty.
Claims this part is about, and what each was answered with
New starters take longer to become productive remotely
Empirical→EvidenceThe one claim answered with evidence that measured it
Productivity fell when we went remote
Empirical→DataRemote workers put in longer hours than office workers
Empirical·Nothing; it was never answeredthree claims · three empirical · none of the answered ones was answered in the wrong kind · one was never answered at all
Three empirical claims with three different fates.
Two of the nine empirical claims could have been answered in an afternoon with data the organisation already had. One of them was. The other was not, and has never been.
Understanding why is more useful than any amount of discussion about evidence in general.
The claim that was not measured
Remote workers put in longer hours than office workers.
Made by somebody on the flexibility side of the argument. The measurement is logged hours by working pattern; the company has logged hours for other reasons; the comparison is a spreadsheet.
Nobody produced it in eleven weeks.
Why not
Not difficulty. Three reasons, and none of them is anybody's fault exactly.
The person who made the claim did not need it. They believe it because they live it, and they experience the claim as reporting rather than asserting. Producing evidence for something you know from the inside feels redundant.
The other side preferred the doubt. An unchecked claim can be discounted. A checked one has to be accepted or disputed on the data, and disputing data is harder than disputing an assertion. There is no incentive to call for a measurement you might lose.
Asking for it would have looked like disbelief. "Can we check that?" directed at somebody's account of their own working hours reads as an accusation about their honesty, not a request for data.
The asymmetry with the one that was measured
The claim that did get measured was about new starters — about a third group, not about the person making the claim or the person doubting it.
That is the whole difference. An empirical claim about a third party can be checked without anybody's account of themselves being doubted. An empirical claim about the speaker's own conduct cannot, which is why those never get checked and why they accumulate.
| Claim is about | Can it be checked without insult? | Was it checked here? |
|---|---|---|
| A third group | Yes | Yes, once |
| The speaker's own behaviour | Not easily | Never |
| The organisation's own costs | Yes | Never asked |
The third row is the odd one: the organisation's costs could have been stated at any time by the organisation, without anybody asking, and were not.
How to get the measurement made anyway
One move works, and it is to offer your own first.
I will pull the logged hours by working pattern — I think they support what I am saying but let us look. Could you pull whatever you have on the other side?
It converts a request into an exchange, removes the implication of disbelief, and commits you in advance to a result you have not seen. It is also the only version on this record that anybody responded to.
What it is worth
Modest and real. Two claims out of twenty-four were settleable with existing data, and settling both would have left twenty-two claims standing.
The value is not in winning the argument; it is in shrinking the part of it that is pretending to be empirical. Every claim that gets checked is one fewer thing to exchange studies about, and on a record like this one that is two conversations that can stop.
The measurement that was deliberately not requested
There is a sharper version of this on the record and it is worth stating.
Partway through, somebody on the office side worked out that logged hours by working pattern existed and could be produced. They did not ask for it.
Asked afterwards why, the answer was honest: they thought it would probably support the other side's claim, and they did not want to be the person who produced it.
That is not bad faith in the sense used elsewhere on this site. It is an ordinary human calculation and nearly everybody makes it. It is also the precise mechanism by which an argument stays empirical in form and unexamined in fact.
The rule that would prevent it
One rule, agreed at the start, costs nothing and removes the calculation entirely:
Any measurement either side asks for, we produce, whatever it shows.
With that in place, asking is free. Without it, asking is a bet, and people do not make bets they expect to lose.
On this record no such rule existed, two checkable claims went unchecked, and one of them was unchecked by somebody who knew exactly how to check it.