Resilience

Adequate to the Last Emergency

The warning that had nowhere to go

Take any decision that matters and follow it along its chain:

1. Governancewho decides
2. Riskwhat could go wrong, and what it would cost
3. Compliancewhat must be true
4. Resiliencewhat all three amount to when the thing actually happens

Each link can be checked on its own, and each can be done well. So work a real one through. The example below is a single night on the North Sea, in the winter of 1953 — a good one precisely because nothing about it was hidden, and nobody involved was incompetent.

The Netherlands had bodies whose duty in law was to hold back the sea, with people in them who could be named and asked — governance. It understood what a spring tide driven by a northerly gale does to a coast lying below the water, and what such a coast is worth — risk. Its sea defences were built to the standard of the day, and inspected against it — compliance.

Three links, all of them in place. The exercise is the fourth.

On the evening of Saturday 31 January 1953, the Royal Netherlands Meteorological Institute got it right. A spring tide and a northerly gale were driving the North Sea into the narrow end of a funnel, and the institute said so, in time.

The warning went out by telegram to everyone who had subscribed to the service. The service had about thirty subscribers. In the region that would be worst hit, one water board out of roughly two hundred received it.

The radio might have carried it further. Dutch radio closed down at midnight and did not return until eight in the morning. The sea came in the dark, in the small hours of the Sunday, through a country that was asleep and had not been told.

By the time it went out again, at least eighteen hundred people were dead in the Netherlands and about three hundred more in eastern England, where twenty-four thousand homes were damaged.

Everything worked except the last yard

It is worth being precise about what failed, because it was not the science and it was not the forecast. The institute was competent. The prediction was correct. The message was composed and sent. Institutions existed, with names and duties and a place in law.

What did not exist was anyone on the far end of the chain at three o'clock on a Sunday morning.

The defences had a version of the same fault, arrived at from a different direction. They were built to the standard of the day, and the standard of the day was the highest water anyone had written down. The sea went higher than anyone had written down.

A standard is a statement about the past. It is the worst thing that has happened, written down and turned into a line on a drawing. It is useful, and it is honest, and it is the only thing anyone can build to. But it describes what has already been survived, and an event has no interest in the drawing.

What the two countries changed, and why it was not the same thing

The Netherlands convened a Delta Commission, which did something that sounds like engineering and is closer to philosophy. It stopped setting the height of a defence from the highest recorded flood, and started setting it from a stated tolerance for failure. Central Holland was to be defended against a flood with a one-in-ten-thousand chance of arriving in any given year.

That number is not a prediction. Nobody believes a sea can be forecast ten thousand years out. It is a decision about how much risk a society agrees to carry, taken in the open, by people who could be named and asked about it. The Delta Works are what the decision cost.

Britain built the missing yard instead of the wall. The Storm Tide Warning Service was established inside the Met Office: a standing route by which a forecast reaches the people who have to act on it, at whatever hour it happens to be ready.

Neither country had been negligent by the standards of its day, and that is the uncomfortable part. In the language we would use now, both were compliant. Their arrangements were adequate to the last emergency, and they were tested by the next one.

The keep

• A standard tells you what has already been survived. It does not tell you what you will meet.
• A control with nobody on the end of it is a document, not a control.
• The hour of the test is not yours to choose, and it will not be a convenient one.
• Ask who is named, not which team is responsible. Teams do not answer telephones.
• A tolerance you have stated out loud can be argued with. A tolerance you have merely inherited cannot.

First move: take one control you genuinely rely on, and find out who is on the end of it at three o'clock on a Sunday morning — not which team, which person, and by what route you would actually reach them. If that takes more than a minute to establish, you have found something worth the afternoon.

Sources

The telegram subscription figure, the single water board, and the radio's midnight-to-eight closedown: Environment & Society Portal, The North Sea Flood of 1953 and the Zeeuws Archief. Casualty and damage figures: Britannica, North Sea flood. The Delta Commission's one-in-ten-thousand standard: Watersnoodmuseum, The history of the Delta Works. The Storm Tide Warning Service: Royal Meteorological Society, The UK east coast floods of 1953. Checked 20 September 2026.

‹ The Second Opinion