In August 2026, U.S. insurance regulators published the first comprehensive national picture of the homeowners market — and it confirmed what millions of homeowners already felt in their bills. Between 2018 and 2024, the rate at which insurers refused to renew existing policies rose between 96% and 216%, depending on the region, with the steepest climb in the West [source: NAIC, 2026]. The average U.S. home-insurance premium has climbed about 46% since 2021, roughly three times the pace of general inflation [source: CNBC, 2026]. The headline that usually rides on top of these numbers is simple and alarming: climate change is making homes uninsurable. The numbers are real. But the one-line explanation is where the story goes wrong — because what is measured and what is attributed are two very different things, and the gap between them is the whole point.
What is actually measured
Start with the losses, because they are the least disputed part. In 2025, global insured losses from natural catastrophes reached roughly USD 107 billion — the sixth consecutive year that figure has topped USD 100 billion [source: Swiss Re, 2026]. Wildfires, storms, and floods made up a record 92% of those insured losses [source: Swiss Re, 2026]. The single largest event was the January 2025 Los Angeles wildfires; the Palisades and Eaton fires together caused on the order of USD 40 billion in insured losses by Swiss Re's estimate, with independent tallies ranging from about $25 billion to $45 billion [source: Swiss Re, 2026; source: Verisk, 2025; source: Milliman, 2025]. That wide spread is worth noticing early: even the size of a loss that has already happened is an estimate with a range, not a single certain number.
The market stress those losses produce is also measurable. The California FAIR Plan — the state's insurer of last resort, meant to be a backstop for people who cannot buy coverage on the open market — grew from 242,440 residential policies in September 2021 to 642,010 in September 2025, a 165% increase, while the value it insures rose from $160 billion to $558 billion [source: Stateline, 2025]. Two of the largest carriers pulled back sharply: State Farm stopped accepting new California homeowners applications in May 2023 and then declined to renew about 72,000 policies in March 2024, while Allstate paused new home and condo policies in the state [source: California DOI, 2024]. These are facts, dated and documented. A backstop insurer nearly tripling its book in four years is not a matter of interpretation.
Where attribution begins — and why one cause is the wrong answer
Here is where care is required. That premiums are up, and that a backstop insurer has swelled, tells you the market is under strain. It does not tell you why, and the most popular answer — climate change, full stop — collapses several distinct forces into one. When you pull them apart, at least five drivers show up, and only one of them is the weather.
Rebuilding costs. When a house burns or floods, the insurer pays to rebuild it at today's prices, not the price when the policy was written. Replacement costs for property-and-casualty losses rose about 45% on average between 2020 and 2023, driven by labor shortages, supply-chain disruption, and the cost of materials like lumber and roofing [source: Dallas Fed, 2026]. A large share of premium growth is simply the rising cost of the same physical repair — an inflation story, not a climate story.
Reinsurance repricing. Insurers buy their own insurance, called reinsurance, to cover extreme years. In 2023 the price of property-catastrophe reinsurance jumped in a "hard market," and carriers passed that cost through to homeowners. But that cost is a cycle, not a one-way ratchet: reinsurance prices have since fallen, with global property-catastrophe rates down 6.6% in early 2025 and a further 14.7% at the January 2026 renewals — the sharpest drop since 2014 — as capital flowed back into the market [source: Artemis, 2026; source: Reinsurance News, 2026]. If the underlying driver were purely rising physical risk, reinsurance would not be getting cheaper this fast. Part of what homeowners paid in 2023–2024 was a capital-market squeeze that is now easing.
Litigation and fraud. Nowhere is this clearer than Florida, where for years the cost of claims was inflated less by storms than by litigation — one-way attorney-fee rules and abuse of "assignment of benefits" agreements. Legal reforms in 2022 and 2023 removed those incentives, and the effect was dramatic: homeowners insurance lawsuits fell roughly 30%, lawsuits against the state insurer fell nearly 50%, and 17 new insurers entered the market [source: Florida Realtors, 2026; source: Executive Office of the Governor of Florida, 2026].
Regulatory rate suppression. In California, Proposition 103 (passed in 1988) long barred insurers from using forward-looking catastrophe models or building reinsurance costs into their rates, effectively holding prices below insurers' own estimate of the risk — one reason carriers chose to stop writing rather than sell at a loss. The state's Sustainable Insurance Strategy, launched in September 2023, now lets insurers use catastrophe modeling and reinsurance costs in rate filings, in exchange for a commitment to write at least 85% of new policies in underserved, high-risk areas [source: California DOI, 2025; source: United Policyholders, 2025]. The retreat, in other words, was partly a pricing-rule problem — and a rule change can reverse it.
Exposure growth. Finally, more homes, worth more money, keep being built in the places most exposed to fire and flood. Some of the FAIR Plan's tripling in insured value reflects rising property values and people moving into high-risk zones, not a change in how often fires occur [source: Stateline, 2025]. The bill can rise even if the hazard holds steady, simply because there is more, and more expensive, property in harm's way.
Correlation is not the same as blaming one fire on the climate
None of this means warming is irrelevant — but it has to be stated at the right resolution. The defensible climate statement is about the trend: insured catastrophe losses have grown on a multi-year basis, roughly 5% to 7% a year in real terms, and Swiss Re projects they could reach USD 186 billion by 2030 [source: Swiss Re, 2026]. That rising trend correlates with a warming climate, and it also correlates with the exposure growth described above; Swiss Re itself attributes much of the long-run increase to rising exposure and asset values, with hazard intensification a contributing factor rather than the sole one [source: Swiss Re, 2026].
What the data does not license is the leap from that trend to blaming any single event on climate change. Saying the January 2025 Los Angeles fires were "caused by climate change" is a far stronger and more contested claim than saying decades of rising losses correlate with warming. The $25-to-$45-billion insured-loss figure for those fires is an accounting estimate of damage, not a measurement of how much of that damage the climate contributed [source: Swiss Re, 2026]. Keeping the trend claim and the single-event claim separate is not pedantry — it is the difference between a statement the evidence supports and one it does not.
The protection gap: the number that actually matters
If there is one figure that captures the stakes, it is the protection gap — the share of catastrophe losses that no insurance covers. In 2025 that gap widened to USD 424 billion globally, up from USD 395 billion the year before, and the resilience index (the share of loss need that insurance actually covers) sat at just 27.3% [source: Swiss Re, 2026]. Nearly three-quarters of the world's catastrophe exposure is uninsured. When insurers non-renew, when premiums outrun what families can pay, and when people drop coverage or lean on a bare-bones state backstop, that gap grows — and the uncovered loss does not disappear. It lands on households, lenders, and eventually taxpayers.
Florida shows the market can move — in both directions
The most useful corrective to the "climate makes homes uninsurable" narrative is Florida's own reversal. Its insurer of last resort, Citizens, shed about 541,000 policies in 2025 and was projected to fall to roughly 385,000 by year-end — a 73% drop from its recent peak, and no longer the largest property insurer in the state [source: Florida Realtors, 2026; source: Insurance Business, 2026]. Its depopulation program moved more than 546,000 policies back to private carriers, the largest such return in a decade, and in December 2025 Citizens recommended rate cuts for most policyholders [source: Citizens Property Insurance, 2025]. Florida is not less exposed to hurricanes than it was two years ago; the weather did not improve. What changed were the legal and market rules. A market that can swing this hard on litigation reform is not a market governed by climate alone.
What to watch next
The honest reading is neither "the climate is making insurance impossible" nor "there is no problem here." Premiums and non-renewals are genuinely up, insurers of last resort are carrying more than they were designed to, and the protection gap is widening — those are measured facts. But the causes are a bundle: rebuild-cost inflation, a reinsurance cycle that is already easing, litigation dynamics, rate-setting rules, and exposure growth, sitting alongside a rising loss trend that correlates with warming. Any story that assigns it all to one of those is describing a different, simpler world than the one the data shows.
For the next year, three things are worth watching. First, whether the recent fall in reinsurance costs actually reaches homeowners' bills, or is absorbed as insurer margin — the test of how much of the recent spike was a capital cycle. Second, whether California's modeling reforms and Florida's legal reforms hold, and whether other high-risk states copy them; that is the lever regulators actually control. And third, whether the protection gap keeps widening even in wealthy markets, because that — not any single fire season — is the number that tells you whether insurance is still doing its job.