California and Florida Are Running the Same Insurance Experiment Backwards

Two state-run property insurers, both created as a last resort, spent 2026 moving in opposite directions.

California's FAIR Plan received approval for a 29.1% rate increase taking effect for new and renewal policies on October 15, 2026. Its exposure has reached roughly $768 billion, up about 250% since September 2022, spread across something like 680,000 to 696,000 policies. Florida's Citizens Property Insurance Corporation, which peaked near 1.42 million policies in October 2023 and was briefly the largest property insurer in the state, has fallen to an all-time low, somewhere under 400,000 depending on which week's disclosure you read.

The tempting conclusion writes itself: Florida fixed its market, California broke its own. Both halves of that sentence are incomplete, and the more useful story is that these are not two crises. They are one process, running through two different sets of state machinery.

The California number that is going the wrong way for the narrative

Almost every account of the FAIR Plan describes relentless growth. Raw policy counts support that. But Stanford's Climate and Energy Policy Program has been tracking something more revealing: how often a FAIR Plan policy sits behind a new mortgage. That share peaked at 8.1% of owner-occupied originations in the first quarter of 2025, immediately after the January Los Angeles fires, and had fallen to 5.6% by the first quarter of 2026.

That is not a collapse in slow motion. It is a residual market that absorbed a shock and then began, at the margin of new lending, to hand some business back. California's regulatory overhaul, which allows catastrophe modeling and reinsurance costs into rate filings in exchange for commitments to write in distressed areas, is the most plausible reason. The FAIR Plan's own 29.1% increase belongs in the same frame: an insurer of last resort is supposed to be expensive. When it is cheap, it competes with the private market it is meant to backstop.

The more counterintuitive finding is buried in the same Stanford work and appears in essentially none of the secondary coverage. FAIR Plan penetration is disproportionately high in ZIP codes classified as low and moderate wildfire risk, not only in fire country. That inverts the standard picture. If the insurer of last resort were purely a wildfire phenomenon, its footprint would trace the hazard maps. Instead it is tracking something closer to carrier withdrawal decisions, which are made at the portfolio level for reasons including reinsurance cost, statewide accumulation and rate adequacy, not just the fire risk of an individual street.

The implication is uncomfortable for anyone treating this as a story about people who chose to build in canyons. It is becoming a housing finance story, because a mortgage requires insurance, and a FAIR Plan policy is a narrow, expensive, dwelling-fire-only product that many buyers must supplement.

Florida's decline is engineered, not spontaneous

Citizens shed roughly 541,000 to 546,000 policies during 2025 alone, with more than 20 new or returning carriers entering the state. Coverage almost universally describes private insurers "returning," which is true and which has a real cause: the December 2022 and 2023 tort reforms that ended one-way attorney fees and curtailed assignment-of-benefits litigation. Florida had a litigation problem wearing a hurricane costume, and legislators addressed the litigation.

But the mechanism that actually moves policies off Citizens' books gets far less attention. Under the depopulation program, if a participating private carrier makes a takeout offer within 20% of the Citizens premium, the policyholder loses eligibility to remain with Citizens. That is not a market clearing through consumer preference. It is an eligibility rule doing the clearing, with the price band set by statute.

This matters for how the achievement should be read. A shrinking Citizens genuinely reduces the state's assessment exposure, which was the policy goal. It does not necessarily mean hundreds of thousands of Floridians shopped their way into a better deal. Some were, quite literally, made ineligible for the alternative.

Who actually holds the risk

Here the two states converge, and here the coverage is thinnest.

When the January 2025 Los Angeles fires produced roughly $4 billion in FAIR Plan losses, the plan issued a $1 billion assessment on California's admitted insurers. Those are the same companies writing ordinary policies in Sacramento and San Diego, and assessment costs do not stay where they land. Florida's structure is more explicit still: Citizens deficits are recovered first from its own policyholders and then, if needed, through surcharges on policyholders across the state who never bought a Citizens policy.

This is the part worth internalizing. A residual market does not reduce catastrophe risk. It converts a private, priced, capital-backed liability into a public, partially unpriced, post-loss one. The bill arrives after the fire or the landfall, spread across everyone with a policy, which is a tax with worse timing and no vote.

What can and cannot be concluded

Florida's reforms worked in the sense they were designed to work: litigation costs fell, capital returned, Citizens shrank. That is a real result and cynicism about it is unearned. But the depopulated market has not yet been tested by a major Florida landfall since the reforms fully matured, and the depopulation rate has been assisted by a rule that limits the choice being celebrated.

California's FAIR Plan is enormous and getting more expensive, but the mortgage-origination data suggests the peak stress may have been in early 2025 rather than now, and the geography of its growth suggests the problem is broader and less hazard-specific than the fire-country framing implies.

The honest synthesis is that neither state has removed risk from the system. They have chosen different places to park it: California in a growing public pool being slowly repriced, Florida in a private market re-entered on the strength of a legal reform, with a statutory nudge closing the exit back to the state. One of those bets will be graded by a wildfire. The other by a hurricane. Neither grade has been issued.