The woman who never existed sits in an office that did. Lloyd's of London, February 1953, three weeks after the North Sea took 1,836 Dutch lives and left 165,000 hectares underwater.1 Dorothy Catchpole—"Dot" to the handful of colleagues who acknowledge her—is one of perhaps a dozen women calculating risk in a building full of men calculating profit. Her desk faces a window overlooking Lime Street, but she's looking at columns of figures that refuse to cooperate.
The Dutch want flood insurance. Lloyd's syndicates want to sell it. Dot's job is to price it. The mathematics are unforgiving.
You're calculating premiums for Dutch flood insurance three weeks after the worst flood in their modern history. That seems like terrible timing.
Dot: (laughs, but it's not a happy sound) Terrible for whom? The Dutch are still pumping seawater out of farmland, still finding bodies. For Lloyd's? This is when the business comes. Everyone wants flood insurance the week after the flood. No one wants it the week before.
Before the flood, I could have priced this honestly. The Dutch had good dikes, mostly. War damage, yes, but nothing catastrophic. The actuarial tables from the 1930s suggested a major surge once every few centuries. Price accordingly, collect premiums, sleep soundly.
Now I know those tables are fiction.
What changed?
Dot: The water came up 5.6 meters above mean sea level.2 Eighteen feet. The Dutch had built their dikes half a meter above the highest recorded water level. Twenty inches of margin. They thought that was sufficient.
So now I'm sitting here with requests from Rotterdam, from Zeeland, from every river delta town that just watched the sea erase the assumption of safety. And I'm supposed to calculate: what's the premium for insuring against something we just learned can happen, when we have no idea how often it will?
Surely you have historical data.
Dot: (pushing papers across the desk) I have data that just proved itself worthless. Look—the Rijkswaterstaat engineer, Wemelsfelder, published a probability distribution in 1939. Logarithmic. Elegant mathematics. He even warned that dike heights were too low.3 No one listened. Or rather, they listened but decided rebuilding from the war was more urgent than reinforcing dikes that had held for centuries.
"Held for centuries" is not the same as "will hold." And now I'm supposed to price the risk of "will hold" when "held" just failed spectacularly.
What does the honest calculation look like?
Dot: If water can reach 5.6 meters when we thought 5 meters was the ceiling, then our entire understanding of maximum surge is wrong.
(pause, drumming fingers on the desk)
If I assume this was a once-in-500-year event and price accordingly, the premium would be perhaps 2% of insured value annually. A farmer with a 50,000-guilder property pays 1,000 guilders a year. That's a month's income for most. Unaffordable.
But if I'm wrong—if this is actually a once-in-100-year event, or if the sea is rising, or if storms are intensifying, which some of the meteorological chaps are starting to whisper about—then 2% is catastrophically low. The syndicate goes insolvent the next time it happens.
And the optimistic calculation?
Dot: (sharp look) The optimistic calculation treats this as an aberration. Freak storm, freak tide, both at once. Won't happen again for a thousand years. Price it at 0.5%. Makes insurance affordable. Makes the syndicate look generous. Makes the Dutch government happy because their citizens can actually buy coverage.
And in ten years, or fifty, when it happens again?
Well. That's a future underwriter's problem.
You lost family in the 1947 Thames flood.
Dot: (very still) My aunt and two cousins. Canvey Island. The water came at night, just like in Holland. They were asleep.
(long pause)
That flood killed over 300 people in England alone. Everyone said it was extraordinary. Once-in-a-century. Six years later, the North Sea does it again, but worse.
So when I look at these Dutch requests, I don't see actuarial tables. I see my aunt's house with water to the ceiling. I see my cousins who were eight and eleven. And I think: if I price this wrong, I'm not just risking the syndicate's money. I'm risking that some other family trusts the insurance policy I wrote and stays in a house that should be abandoned.
That's not your job, though. Your job is calculating risk, not making evacuation decisions.
Dot: (bitter laugh) If insurance is unaffordable, people stay in flood zones without coverage. If insurance is affordable but underpriced, the system collapses and people are left with worthless policies when the water comes. Either way, I'm making decisions about who drowns.
The men upstairs don't see it that way. They see market share and premium volume and syndicate rankings. They want me to price competitively. Which means optimistically.
(stops, looks out the window)
I keep thinking about something the Dutch engineer wrote—van Veen, the one who kept demanding dike improvements. He filed a report on January 29th, two days before the flood.4 Two days. He was shouting into a void while everyone focused on rebuilding from the war. Then the water came and proved him right, and it was too late.
What did you decide?
Dot: (very quiet) I priced it at 0.75%. Optimistic but not reckless. Affordable but not insulting to the mathematics. The syndicate accepted it. The Dutch are buying it. Everyone's happy.
And I wake up at night doing the calculations again, wondering if I just wrote policies for people who'll drown in twenty years because I made their false security affordable.
The Dutch are building the Delta Works now. Doesn't that change the calculation?
Dot: It should. They're talking about reducing flood risk to once per 4,000 years.5 Extraordinary engineering. But it won't be finished for decades. What do I do in the meantime? Price based on the dikes they have or the dikes they're promising?
And here's what really keeps me up: even if the Delta Works succeed, even if they make Holland safe for 4,000 years—what about everywhere else? The Thames flooded in '47. The North Sea flooded Holland and England in '53. How many other coasts are we underpricing because we're treating each flood as a separate aberration instead of a pattern?
Are you saying this is the beginning of something systemic?
Dot: I'm saying that every underwriter at Lloyd's is facing versions of this choice. Marine insurance after ships sink. Fire insurance after warehouses burn. And now flood insurance after floods that weren't supposed to happen.
And the incentive is always the same: price low enough to win the business, high enough to look responsible, and hope the next disaster happens on someone else's watch.
(bitter smile)
We're very good at optimism in this building. We've been insuring ships since 1686. We've survived wars, fires, earthquakes. The market assumes we'll survive anything because we always have. But what if "always have" is like "held for centuries"? What if it just means we haven't hit the thing that breaks us yet?
Do you think you'll be proven wrong?
Dot: I think I'll be dead before we know. The policies I'm writing now won't mature for decades. Maybe the Dutch build their Delta Works and I'm remembered as the underwriter who helped them afford insurance during the transition. Maybe the sea keeps rising and I'm remembered as the woman who underpriced catastrophe.
Or maybe—and this is what I actually think will happen—no one remembers at all. The premiums get paid, the policies renew, the system keeps running. And the next time the water comes, everyone acts surprised again.
(straightens papers, professional mask returning)
The Dutch requests are due by Friday. I should get back to the calculations.
One last question. If you could price it honestly—truly honestly, regardless of affordability—what would the premium be?
Dot: (long silence) I don't know.
That's the answer that terrifies me most. I genuinely don't know. The mathematics we have aren't adequate for the world we're living in. And I'm pricing insurance anyway, because someone has to, and because the alternative is admitting that some risks are uninsurable.
But if I had to guess? Double what I'm charging. At least. Maybe more.
(looks back at the window, at the London sky)
The Dutch are calling this the Watersnoodramp. Water emergency disaster. As if it's over. As if the emergency was the flood, not what comes after.
I'm writing policies for a world that doesn't exist anymore. We all are. We just don't admit it.