Give Us Our Eleven Days

In September 1752 Britain deleted eleven days. Wednesday the second was followed by Thursday the fourteenth.

What everyone knows about this is that people rioted, demanding their eleven days back.

The evidence for that is a satirical journal and a painting. Lord Chesterfield’s The World ran the joke; Hogarth put a placard reading “Give us our Eleven Days” into An Election Entertainment in 1755, three years after the fact, in a picture about electoral corruption. Contemporary riot depositions held at the Public Record Office contain no record of any such disturbance.

The story is a good one. It has a crowd, a grievance, and a line of dialogue. It has been repeated for two hundred and seventy years by people who were not making anything up, each of whom got it from somebody who also was not.

Meanwhile, in the same week, Parliament wrote a clause that is still running.

What the Act actually did #

Britain converted more than a century and a half after the Catholic states did, which created a problem the 1582 changeover never had. In 1582 no contract had been written against the deleted dates. In 1752 rents, wages, annuities and loans had all been signed against a calendar that was about to lose a fortnight.

Section VI of the Calendar Act dealt with it directly: payments falling due in the gap were deferred by eleven days. A payment due on 3 September fell due on 14 September instead — not cancelled, not accelerated, moved. Wage abatement tables circulated to prorate the shortened month, at reductions cited around sevenpence in the pound of monthly wages.

That is the documented cost of the changeover. It is small, real, and per-worker, and it is the part with a paper trail.

There is also a piece that never went away. Lady Day, 25 March, had been the start of the English financial year. Shift it forward by the same eleven days the calendar lost and you get 5 April, which is why the UK tax year still ends there rather than at the calendar year or at Lady Day itself.

Every UK payroll run still executes against a date chosen to accommodate a correction made in 1752. Nobody involved thinks about it. There is no story attached to it at all.

Displacement, not exaggeration #

The interesting property here is not that the riot story is false. Plenty of stories are false.

It is that the false story is not an inflated version of the true one. An exaggeration would be there was some grumbling, which became a riot in the retelling. That is ordinary and easy to correct for.

What happened instead is that a memorable account attached itself to the event and pushed out an unmemorable one that was about something else entirely. The riot is about anger. The Act is about payment schedules. They are not competing descriptions of the same thing; one simply took the slot.

And the selection pressure is obvious once stated. Transmission favours what can be told. A crowd with a slogan is a scene. A deferral clause is a clause. Over enough retellings the scene wins every time, regardless of which one had consequences.

The clause won on consequences and lost on transmission, which is why it is still in force and unremembered — and those two facts are related. Nothing kept it alive except the statute book, and the statute book is not a story.

The same slot, taken #

Financial events have a remembered cause and an operative one, and they come apart in the same direction.

The remembered cause of a crisis is usually a scene: a trading floor, a phone call, a firm failing on a particular morning. The operative change is usually a rule written eighteen months later by a committee, which is still running and which almost nobody can name.

The test is not whether the anecdote is true. It usually is, or true enough. The test is whether the thing you remember about an event is the thing that is still operating.

Take any market episode you can recount. Then ask which convention, threshold, reporting requirement or settlement rule dates from it and is still executing. If you can produce the anecdote and not the mechanism, memory has attached to the wrong artefact — and it will have done so for the same reason it did in 1752, which is that one of them was worth repeating.

The riot is the best-known thing about the British calendar change and did not happen.

The eleven-day deferral is the least-known thing about it and ran on every debt in England.

One of those is still being paid, twice a year, by everyone with a UK payslip, on a date that exists because of a fortnight that did not.