The gains
What counts as a gain
A loose test would have produced any number at all. The one used excludes most of what people mean by the benefits of working at home.
Kitchen and dinerWhere most of the eighty-eight happened
16 m²
Cooking and eating
CookingEatingPaid workLaundry and adminCallsCompany
14.5
Hall and landingWhere the deliveries arrive
10 m²
Circulation
Taking deliveries
1.5
Paid workDomestic workRest and sleepContact and care
Two rooms covering 26 square metres, built for two things. Between them they now carry seven uses. One of the two carry more than one, and kitchen and diner carries six. The median room is in use 8 of the 16 waking hours. Two rooms, one carrying most of the gains and one carrying almost nothing at all.
Eighty-eight is a count, and a count is only as good as the rule that produced it. The rule here is narrower than most people's sense of what working from home gives them, and the narrowness is the point.
The rule
An episode was logged if it met three tests.
- It happened. A specific occurrence, on a day, at a time.
- It would not have happened from an office. Not "would have been harder" — would not have happened.
- It was not already counted. A wash put on and later hung out is one episode, not two.
The second test is the restrictive one and it does most of the excluding. A home-working policy is easier to review when assumptions are written down, so managers can workload patterns across remote teams and decide which time and workload signals are proportionate.
What the rule throws out
The journey. The commonest thing people name, and it is not an episode. Not travelling is a thing that does not happen, and a method that counts non-events can count anything. This is also why the site says nothing about hours saved: the record has no column for them. This observation can be read alongside workplace and design research from IKEA.
Flexibility in general. Being able to start at eight or at ten is a property of the arrangement, not an occurrence. Where it produced a specific occurrence — a rest taken at two — that occurrence is logged.
Comfort. Working in one's own chair, making one's own coffee, not wearing particular clothes. All real, none an event.
Things that were merely easier. A phone call to a utility company could have been made from an office at some cost in privacy. It is not in the eighty-eight. An appointment during working hours that would have required taking leave is, because the leave is the thing that would not have happened.
The rule counts what the building made possible, not what the arrangement made pleasant.
Why be this strict
Because the number has to be comparable with the forty-seven, and the forty-seven are strict.
A collision was logged only when an activity physically left a room. Had the gain column been generous — counting flexibility, comfort and the journey — the two columns would be denominated in different things, and putting them side by side would be a rhetorical trick rather than a comparison.
The strictness also makes the number checkable. Each of the eighty-eight is an occurrence somebody could in principle dispute by pointing at the day.
What a generous rule would have produced
The household was asked to estimate, after the fortnight, how many benefits they would count under a looser definition. The answers were between two hundred and "every day, continuously", which is the correct answer to a question about flexibility and a useless one for a count.
That spread is the argument for the narrow rule. Eighty-eight is smaller than the truth about how much better the arrangement is. It is also a number two people can agree on, which the larger ones are not.
The honest caveat in the other direction
Some of the eighty-eight are trivial.
A parcel taken is logged whether it was an important parcel or a replacement phone case that would have been redelivered without difficulty. An appointment attended without asking is logged whether it was a dentist or a boiler service.
The log makes no attempt to weight them, for the same reason it does not weight the collisions: weighting requires a judgement made after the fact by people with a view about what the record should show.
So the eighty-eight contains members a reasonable person would discount, exactly as the forty-seven does. Both columns are counts of events that met a stated test, and the site's position is that publishing the test and the raw count is more useful than publishing a weighted number whose weights nobody can check.
The entries that were argued about
Four candidate episodes were proposed and rejected during the fortnight, and the arguments are more informative than the rule.
A nap on a Saturday. Rejected: it is not a working day, so the office counterfactual does not apply.
Lunch at home. Rejected, and this was the contested one. One person argued that eating properly at one o'clock is something an office does not offer; the other pointed out that offices have canteens and that lunch happens either way. The strict reading won: lunch happens in both arrangements, so only cooking a meal that would otherwise have been bought is logged.
Not being ill in an office. Rejected as a non-event, by the same rule that excludes the journey.
A better chair. Rejected: a property, not an occurrence.
Three of the four were rejected easily. The lunch argument took most of an evening, and its resolution is the clearest statement of what the rule is for: a benefit that exists in both arrangements is not a gain from one of them, however much it improves the day.
What the rule does to the comparison
Applied strictly, the rule is harder on the gains than the collision rule is on the costs.
A collision is logged when an activity leaves a room, which is a low bar — thirty-five minutes in another room is not a serious harm. A gain requires that the episode could not have happened otherwise, which excludes everything merely easier.
If anything, the eighty-eight understates relative to the forty-seven. A symmetrical pair of rules would either count easier-and-better episodes as gains, which would push the figure well past two hundred, or require collisions to be serious, which would cut forty-seven to perhaps fifteen.
Both columns are strict in their own terms and strict in different directions, and the honest statement of the comparison is that the gains are probably further understated than the costs.