Why your insurer no longer wants your house

The letter is polite, almost dull. It informs you that your home insurance policy will not be renewed when it lapses. You read it again, sure you have misunderstood. You have never filed a claim, you pay on time, your roof is new. Nothing in your conduct explains this notice. And yet the decision is firm, and it targets the place where you live rather than you. An identical house a few kilometers away stays insurable without trouble. Yours is on the wrong side of a line you cannot see, a line insurers see more and more clearly, and that creeps a little further inland every year.

No underwriter's whim is behind this, and no excess of caution. Insurance has reached its own limit. Insurance is a bet on chance, and it promises nothing beyond that. It works only on three conditions, that misfortunes be rare, that they strike at random with no one knowing in advance who will be hit, and that one person's misfortune tell you nothing about another's. Climate change breaks all three at once. When a peril becomes frequent, foreseeable and collective, it stops being a risk and becomes a fate, and no one insures a fate. Your insurer walks away because the very object it needs, a poolable risk, has dissolved under your roof. Greed and indifference have nothing to do with it.

The village and the common pot

To see why, we have to return to what insurance really is, stripped of its jargon. Picture a village where everyone drops a coin into a common pot at the start of the year. If one person's barn burns, it is rebuilt from the pot. The system rests on plain facts: barns rarely burn, no one knows which will burn, and one fire tells you nothing about the next. The many who are spared pay for the few who are struck, and the following year it begins again. That is pooling, the mainspring of all insurance: an unpredictable misfortune turned into a small, predictable cost shared by all.

Now move that village into a floodplain where the river rises a little higher every decade. Three things break down, and each one attacks a pillar of the system. The first is frequency. The barn is no longer flooded once a generation, but every few years. The pot empties faster than it fills, so the coin each person drops in must rise, until it costs nearly as much as the barn itself. At that point insurance loses its meaning, because paying, every year, a sum close to your own loss, with a fee on top, amounts to funding your own damage, only more expensively. Nobody cancels the policy out of malice; arithmetic does it.

The second thing that breaks down runs deeper: correlation. When the river overflows, it does not take one barn, it takes them all at the same time. But the pot could rebuild one unlucky member out of the contributions of the many spared, it cannot rebuild everyone out of no one. The very principle that held the structure up, the small number of victims funded by the large number spared, inverts. There are no spared members left, only simultaneous victims. You endure a catastrophe together, and a collective catastrophe cannot be pooled, only borne.

The third is the most counterintuitive: predictability. As climate models sharpen, the insurer sees more and more clearly which barns will flood and which will be spared. But a risk you can foresee with confidence becomes a cost to be assigned. The owner of the threatened barn becomes uninsurable precisely because the danger is known. Here is the most unsettling reversal, the very knowledge that made insurance possible ends up making it impossible. As long as the die looked fair, everyone agreed to wager. Once everyone can see the die is loaded, and which way it lands, there is no bet left to make, only a bill to hand to the loser named in advance.

On top of these three cracks comes a vicious circle that hastens the collapse: adverse selection. When the premium climbs to catch up with the risk, it is the least exposed who leave first, because they judge, often rightly, that they are paying for others. But their departure drains the pot of the good risks that were funding it, which forces the premium up again, which drives the next ones out, and so on. The insured group then shrinks around the most threatened, until it holds only those almost certain to be struck. At that point there is no pooling left at all, only a club of future victims splitting a bill they cannot pay. The spiral does not stop on its own; it ends when the market simply disappears, the exact point at which your letter was mailed.

So the three pillars, rarity, independence, uncertainty, crack together, and pooling, that silent engine, seizes up. Your insurer's retreat comes out of the pooling calculation itself, followed through to the end. It leaves you because it knows how to do one thing, turn a shared chance into peace of mind, and where you live, chance has stopped existing.

The fair price or fair access

This is where the subject stops being a matter of actuarial plumbing and becomes a real debate, with no obvious side. One reading vindicates the insurer, and even finds it virtuous. A price that climbs to the true level of the risk is an honest signal. It says what politicians dare not, that people should stop building there, and stop rebuilding there. The retreat of insurance would then be the market whispering a truth, the price system gently steering people away from danger. On this view, the insurer that walks away does a service, it refuses to subsidize settlement in places doomed to burn or to drown, and its signal, if heeded, saves lives and money.

The opposite reading is just as solid, and far more troubling. This retreat abandons exactly those least able to leave. It turns insurability into a privilege of safe places and well-off owners, and manufactures a kind of exile by risk map, whole neighborhoods where no coverage exists, where house values collapse, where the poorest are trapped in an asset they can neither insure nor sell. The price signal, however clear-eyed, does not only say where the danger is great, it also says who can afford to flee it and who will have to stay. Actuarial truth and social justice then stare each other down, because what is mathematically honest can be humanly unbearable.

Between the two, almost always, the same actor appears, the state. When private insurance retreats, public power steps forward, through guarantee funds, subsidized pools, an insurer of last resort. But this intervention resolves nothing, it moves the problem. The taxpayer now pools what the market refused to pool, and that solidarity can be noble, yet it can also become a subsidy for building in harm's way, delaying the very retreat the price signal was trying to force. We end up paying, through taxes, to keep people where the risk has already delivered its verdict. There is no clean solution to this dilemma, only a permanent trade-off between the honesty of the price and the fairness of access, two just values pointing in opposite directions.

None of this is theoretical. In California, large insurers such as State Farm and Allstate stopped accepting new home policies in 2023, citing soaring wildfire risk and the surging cost of rebuilding and of reinsurance. Homeowners then fall back on the FAIR Plan, the state's insurer of last resort, dearer and thinner in coverage, whose rolls swell year after year. In Florida, the public company Citizens has grown into one of the state's very largest insurers, as private players retreated from the hurricane-exposed coasts. The retreat is already under way, redrawing the map of what is habitable, insurable and financeable.

The retreat faster than the water

Worse than the slowness of the physical danger is the speed of the financial retreat. A house does not become unlivable the day the water rises, it becomes unsellable long before, the day no one will insure it any more. Insurance is the link that holds the whole chain of ownership together. A bank does not lend without requiring home insurance, so an uninsurable property instantly becomes unfinanceable, and therefore unsellable, since the next buyer cannot borrow to acquire it. The property's value then collapses because the financial system anticipates the flood, long before the water.

The consequence is dizzying, economic destruction precedes physical destruction, sometimes by decades. The market does not wait for the catastrophe to arrive, it prices it today and inflicts it today, turning a future threat into a present ruin. This is the cruelest paradox of the mechanism, the financial loss strikes people whose house is still perfectly standing, and might stay standing for twenty years. The insurer that withdraws does not merely announce that your risk has changed, it triggers part of the very loss it claims only to measure, because in leaving it takes with it the value, the credit and the liquidity of your property. The retreat is a prophecy that comes true in the speaking of it.

What the letter really announces

Your insurer's letter is not really about your house. It is the first place where a global, abstract fact, the climate is changing, turns into a personal, local, priced verdict. Insurance is the canary in the mine, because it is the one industry paid to judge the future without flattery. When it leaves a place, it is telling you that place's future has been repriced, before the official maps, before the political debates, before you feel it in any other way. The price of insurance is a prophecy no one wants to hear, and that is exactly why it is reliable.

The letter raises a harder question than how to get covered again: who bears a risk that can no longer be shared. And that question is going to spread, from the coasts and the fire lines inward, from the shore toward the interior, from the extreme case toward the ordinary one, until it reaches almost everyone. Where the market stops, a political choice begins: deciding together who will pay for a world that has become, in places, too certain to be insured.