Claim Denial Rates by State: NAIC's Real 2026 Numbers Behind the Viral Ranking

On April 16, 2026, Weiss Ratings published a list of 15 large home insurers that closed at least half of their 2025 claims without paying anything.³ Within a week, MoneyGeek, AOL, Newsweek and a run of adjuster and attorney blogs had converted that list into a ranking of the worst states for claim denials. The insurers on it are domiciled in places like Texas and Florida. Whether any of them wrote a single Texas or Florida policy is a question the ranking never asks. MoneyGeek's own methodology note concedes that its state column shows insurers headquartered in each state rather than where claims were filed, and the state-labeled table runs anyway.⁵

State of risk versus state of domicile

A state-level version of this measurement already exists, published every year, sitting in a spreadsheet almost nobody in the reposting chain opened.

The Market Conduct Annual Statement collects claim-handling data from licensed homeowners writers, and the NAIC's State Ratio Distribution Reports aggregate it by the state where the insured property sits rather than by the state on the carrier's charter.¹ Ratio 1 in the homeowners blank is claims closed without payment divided by total claims closed, computed separately for each reporting jurisdiction. It is, in other words, the thing the viral maps present themselves as being.

Domicile aggregation fails at a mechanical level. A carrier chartered in Texas that writes wind business across four Gulf states reports one company-wide figure, and that single number is what a domicile-keyed table drops into the Texas row. A national writer chartered in Illinois carries claims from all fifty states into the Illinois row. Several of the largest personal-lines groups in the country are chartered in states where they write a modest share of their book, so the row carrying that state's name is mostly describing losses that happened somewhere else entirely. The state column in those tables is recording where corporate counsel filed the articles of incorporation. MCAS assigns each claim to the jurisdiction whose insurance code governed the adjustment, whose adjuster licensing rules applied to the person who inspected the roof, and whose department would field the complaint.

For data year 2025 the report covers 52 jurisdictions: the fifty states, the District of Columbia, and Puerto Rico. The median jurisdiction, excluding Puerto Rico, closed 31.23% of homeowner claims without a payment. The range runs from Wyoming at 18.87% to Puerto Rico at 47.12%,¹ the same kind of state-by-state spread that shows up whenever a rule that sounds national turns out to be fifty different rules wearing one label, the way public adjuster fee caps vary by statute rather than by the flat "10 percent" figure the trade repeats.

Puerto Rico posts the highest figure on the board and the least comparable one, a small, storm-exposed market whose ratio moves on claim volumes that would barely register in a mainland state; it belongs in the reading as an outlier rather than a rank. The District of Columbia at 38.04% is one city. Among full states, New Mexico's 35.29% leads, and the distance from there down to Wyoming is roughly sixteen points across the entire country.

The top of the state-of-risk list looks nothing like the top of a domicile-keyed one. Maine, Vermont, Massachusetts and New Hampshire sit inside the top ten alongside Florida, Hawaii and California, a distribution shaped at least in part by deductible structure and claim mix rather than by any regional story about carrier conduct. New York and North Dakota began reporting to MCAS with data year 2025, so neither carries a comparable prior-year series and neither should be read for a trend.²

Five data years, no spike

The framing repeated all year has been that denials are climbing. The medians do not show it.

Across data years 2021 through 2025 the median state ratio moved 31.11%, 29.58%, 29.16%, 29.77%, and 31.23%.² The 2025 median sits about a tenth of a point above where 2021 started, having dipped in between, noise around a stable center produced over five years of heavy catastrophe activity and steep homeowners rate increases across most of the coastal South. A dataset generated during that stretch could plausibly have moved several points in either direction. It moved a fraction of one.

Texas is the clearest case of a number traveling further than its source. Texas Ratio 1 ran 34.26% in 2021, 32.89% in 2022, 30.80% in 2023, 34.20% in 2024, and 30.45% in 2025, placing it 33rd of the 52 reporting jurisdictions last year and below the national median², also the state where a 2026 appraisal ruling is already reshaping how carriers there account for what they owe on a disputed claim. The 47% figure appearing in Texas attorney marketing and in a run of syndicated 2026 explainers comes out of insurer-level aggregation in the Weiss release, not from the Texas Department of Insurance or any state-of-risk dataset, where a state label describes a company's home office rather than a policyholder's address.

Florida's series swings harder: 30.30%, 35.37%, 29.44%, 40.26%, 34.16%.² One footnote governs all five of those figures. NAIC's MCAS homeowners data for Florida excludes Citizens Property Insurance Corporation, the state's insurer of last resort and one of its largest property writers by policy count. Every Florida ratio quoted from this dataset describes the private market only, which means the published all-carrier picture for the state is understated rather than inflated.

MCAS Ratio 7: suits opened per claim closed without payment

The homeowners blank carries a second ratio that gets almost no attention outside regulator offices and matters considerably more to anyone working files. Ratio 7 divides suits opened during the reporting period by claims closed without payment, by state.

Florida's 14.28% is roughly 4.2 times New York's 3.40%, and the remaining jurisdictions in the top five sit in a tight band just above 3%.¹ Most of the country reports well under that. A claim closed without payment in Florida is, at the state level, several times more likely to become a lawsuit than the identical outcome in a median state, and the gap between first place and second is wider than the gap between second place and the bottom of the table.

One reading of that number is that Florida carriers are closing files without payment that a neutral reviewer would have paid, and litigation is functioning as the correction mechanism, the same dynamic behind the bad-faith hail litigation building against carriers in Oklahoma. The competing reading is that Florida holds the densest concentration of public adjusters and first-party property attorneys in the country, so a non-payment there meets organized resistance it would never meet in a thinner market. Both are consistent with a 14.28% suit ratio.

Ratio 7 cannot separate them. It counts suits opened, not merits, and it reports nothing about which side prevailed or what the file looked like at the point of closure. Florida also leads Ratio 3, with 41.65% of its paid homeowner claims paid more than 60 days after report, against 35.37% in second-place Hawaii.¹ A slower payment clock is equally consistent with either explanation.

What "closed without payment" does not mean

On June 1, 2026, the Insurance Information Institute pushed back publicly on how the closed-without-payment figures were being characterized, after a Weiss report framed the data as a sharp rise in homeowner claim denials, itself following a Trump social-media post calling for more transparency around denials.⁴ Triple-I's language was direct: "Claims closed without payment are not synonymous with denied claims, and neither should be automatically interpreted as evidence of improper claims handling or an unwillingness to pay covered losses."⁴

The mechanics behind that objection are real, and any adjuster who has pulled a carrier's claim log recognizes them. A loss that comes in under the deductible closes without payment. A flood claim ultimately covered by the NFIP frequently requires the private carrier to close its own file administratively first. Duplicate filings on a single loss close without payment on all but one. And in the weeks after a named storm, contractors and adjusters canvassing a neighborhood generate a wave of filings on damage that turns out to fall below the deductible or outside the perils the policy covers. Every one of those lands in the numerator of Ratio 1 without anyone having mishandled anything.

What the data does not support is the inference the rankings are built on. A jurisdiction's ratio sitting three points above the median is not evidence that its carriers are handling claims improperly, and nothing in the MCAS scorecard distinguishes a correctly denied claim from an incorrectly denied one. What the data does support is narrower: the spread across states is modest, the five-year trend is flat, and one jurisdiction's litigation ratio sits four times above the next.

None of that settles what a carrier did on any particular loss, though it does settle which benchmark belongs in a demand letter: the state-of-risk column in a NAIC spreadsheet, not a headline assembled from company charters.¹