Other shapes
Annualised hours, and the deferral problem
An arrangement that moves hours rather than removing them produces no list of removals, and on the evidence of this record that is a disadvantage rather than.
The Friday clean-down of the workshopThe removal a compressed week cannot make
week 9 · crept back after 22 weeks
A named person on the phone until sixThe removal staggered cover is designed to avoid
week 7 · crept back after 5 weeks
Visits to the two furthest sitesThe removal a nine-day fortnight makes instead
week 7 · asked back after 12 weeks
Crept back 2Asked back 1
Three removals were written down with the week each of them stopped. Every one of them produced an event inside the 78 weeks watched. Of these, one was asked back and two crept back. Three removals that a different shape of week would have handled differently.
Annualised hours sets a yearly total instead of a weekly one. Busy periods are long, quiet periods are short, and the shorter week is taken where the work allows it.
Nothing is removed. By the logic of this record that ought to make it the safest arrangement of the four, and on one reading it is. On another it is the most dangerous, and the reason is worth setting out because it applies more widely than to annualised hours.
The case for it
No subtraction means no list of removals, no slow failures of the kind documented here, and no need for an eighteen-month read-back.
It also matches the work rather than fighting it. An organisation whose demand has a shape — seasonal, cyclical, tied to somebody else's calendar — is already working long weeks and short ones. Annualised hours makes that explicit and lets people plan around it, which is most of what a shorter week is for.
For a genuinely cyclical operation this is the arrangement that imposes the least and delivers something real.
The case against it
Procedures do stop under annualised hours. They just do not get written down, because nothing has been decided.
During a long period, the things that would be dropped in a four-day week get dropped anyway — the reconciliation that slips, the induction folder that does not get updated, the contract review that waits until after the season. Nobody calls these removals. They are deferrals, and a deferral is a removal with an intention attached.
The intention is the problem. A removal goes on a list. A deferral goes on nobody's list, because it is understood to be temporary, and it is temporary until it is not.
A removal has an entry, an owner and an expectation that somebody will look. A deferral has a vague plan to catch up, held in the head of a person who is busy.
What this record suggests about deferrals
Every mechanism that worked on this record depended on the removal being recorded.
The finding about quiet removals required a column recording whether anything was said. The read-back required a list. The two useful questions — who here has done this? and what did you used to do on the day that went? — both attach to a moment at which something stopped.
A deferral has none of those. It has no moment, because it happened gradually. It has no record, because nobody thought it needed one. And it has an explicit expectation that it will be reversed, which suppresses exactly the attention that a removal attracts.
Of the thirty-eight entries on this record, the two that were never proposed were the two slowest failures. A deferral is structurally the same thing: a stop that nobody announced.
The practice that would help
The same practice, applied to a different trigger.
Instead of writing down what was removed, write down what is being deferred, at the moment of deferral, with the date it is expected to resume. Three columns and the third one is a date.
| deferred | since | due back |
|---|---|---|
| quarterly reconciliation | March | after the season |
| induction folder | March | after the season |
| contract review | February | before renewals |
A list like that is no harder to keep than the one this site is built on. Its advantage over the removal list is that the third column is usually fillable, because a deferral comes with an intention and the intention has a rough date in it.
Its disadvantage is that nobody writes it, for the same reason nobody writes the removal list: nothing goes wrong if it is not written, and the moment at which it becomes interesting is after everybody has stopped thinking about it.
Where annualised hours is clearly right
Where the demand genuinely has a shape, where the shape is outside anybody's control, and where the alternative is people working long weeks anyway without acknowledgement.
In that situation the arrangement is not a way of getting a shorter week. It is a way of making an existing pattern legible and compensable, and the deferral problem it carries was already there before the arrangement existed.
What it should not be chosen for is the avoidance of removals. An arrangement that produces no list because nothing is decided is not avoiding the risk this site describes. It is taking the same risk without the record.
Why a deferral is worse than a removal
The claim deserves spelling out, because it is counter-intuitive: a thing that is coming back ought to be less risky than a thing that is gone.
Three reasons, and they compound.
It attracts no attention. A removal is a change and changes get watched, however briefly. A deferral is understood as temporary, so nobody watches it at all.
It has no date. "After the season" is not a date. On this record's evidence, the thing that makes a read-back happen is a calendar entry with a name against it, and a deferral generates neither.
Its consequences accumulate during the period nobody is looking. The reconciliation failed by accumulation — four quarters at once rather than one quarter badly — and accumulation is exactly what a deferral does. A deferral of six months in a cyclical business is six months of accumulation with an explicit expectation that it is fine.
The remedy is the same as everywhere else and it is the third column: a date, written down at the moment of deferral, by somebody who will still be there. Deferrals are easier than removals in this one respect, because the intention to resume usually comes with a rough date attached. It simply never gets recorded.