Every state insurer of last resort, named and counted

Four sources count this, and all four give a different answer. The disagreement is definitional rather than factual, so the number is worth nothing without the definition attached.

The Peril Desk uses 35, the GAO figure, because it is the most recent, it is federal and free to read, and it states its own as-of date. Anyone publishing a single confident number here has picked one of these four and hidden the other three.

One state sits outside every count. Arkansas is excluded from the Triple-I table by printed footnote, is absent from the PIPSO member list, and its apparent plan domain resolves to an empty page with a single lander URL, which is what a parked domain looks like. What Arkansas has in statute is a rural-only underwriting association that cannot accept an application until the commissioner finds, after a hearing, that a need exists.

Residential policy counts are fiscal 2024. The change column runs fiscal 2018 to fiscal 2023. Read the coverage limit column with care: several plans set different caps for dwelling, contents and commercial, and a few of these figures are a maximum insurable property value rather than a maximum amount of coverage. Where the two differ, the plan is the authority, not this column.

Two states run two plans each and must never be collapsed into one row. Texas splits the job between the Texas FAIR Plan, which is barred by statute from writing wind and hail in the coastal catastrophe area, and the Texas Windstorm Insurance Association, which writes only that. North Carolina runs a statewide FAIR plan and a separate coastal pool. Mississippi does the same.

The plan is not a replacement for the policy that dropped you, and the gap is the part nobody explains at the point of sale. There is no national rule here. Each of the sentences below is one plan, checked against that plan’s own material.

Theft and liability are the usual casualties. The Oregon FAIR Plan states that personal liability coverage is not available and theft coverage is not available. The Washington plan states that liability, theft and most water-related losses are not available. New York’s regulator states the plan carries no liability, flood or theft coverage. Virginia caps theft at $5,000. Against that, Massachusetts and Rhode Island write full homeowners forms including theft and personal liability, Michigan includes theft and $100,000 to $300,000 of liability, Minnesota writes liability on its homeowners forms, and Connecticut makes liability available. Do not assume a plan is fire and extended coverage only because the plan next door is.

Some plans are narrower still. The California FAIR Plan product is fire, lightning, explosion and smoke. Colorado writes fire, lightning and smoke as its base, with hail and wind as optional add-ons, which in a hail state is the whole question. Alabama’s beach pool and the Texas windstorm association are wind and hail instruments and nothing else.

Settlement terms are frequently worse. The Texas FAIR Plan’s dwelling form settles at actual cash value with replacement cost available only by endorsement, and its TDP-1 form has no replacement cost option at all. Every Wisconsin plan policy is actual cash value. New Mexico is actual cash value only. Actual cash value on a roof after a hailstorm is a different cheque from replacement cost.

No plan checked here writes flood. It is a separate purchase, through the National Flood Insurance Program or a private carrier, and several plans, Florida Citizens among them, require proof of it before they will bind.

The entry rule is the field most often reported wrong, because it looks like it should be uniform and it is not. A neighbouring state is not evidence of anything. The table lists only the jurisdictions where the rule was verified against a plan document, a regulator page or a statute. Where a cell is absent from this table, the rule was not published anywhere that could be read, and the honest answer is to phone the plan.

Two plans bar the door in a way worth knowing before a fire season starts. The Oregon plan will not bind, broaden or increase coverage, or reduce a deductible, while a catastrophe is declared or anticipated. New Mexico runs a moratorium within fifty miles of an active fire. The window to move closes before the smoke arrives, not after.

Two more rules cut against the assumption that the plan is a permanent home. Texas requires the policyholder to go back to the voluntary market every two years and be declined again. Indiana requires the three declinations to be re-proved at every renewal.

This is the half of the map that never gets published, because proving an absence is slower than reporting a presence. Each row below rests on the state’s own regulator or its own code, not on the fact that no plan turned up in a search.

The distinction between the last four rows and the first nine is the whole point. A statute that exists and was never switched on is not the same as no law, and a source that will not load is not the same as a source that says no.

Most of these plans publish no policy count anywhere on their own site. Texas publishes quarterly through its regulator, Florida Citizens publishes monthly, Georgia publishes a quarterly exposure report, and South Carolina’s regulator publishes a coastal report. Almost everybody else publishes a homepage, a phone number and an agent portal. Every count in the big table above therefore comes from one downstream census rather than from thirty eight separate plans, and it is fiscal 2024.

Where a plan’s own regulator publishes something fresher, that figure beats the census row. Five do. The Texas Department of Insurance put the FAIR plan at 124,445 policies and $39.2 billion of exposure on March 31, 2026. Florida Citizens stood at 278,196 policies on July 31, 2026, down from a 2023 peak near 1.42 million. California moves fastest of all: an Assembly committee background paper recorded 668,609 policies and $724 billion of exposure at December 2025, against the 417,310 in the fiscal 2024 census above, which is what a three-year-old number costs you in a state like this one. Georgia reported 9,328 policies at June 30, 2026. South Carolina counted 15,337 wind pool policies and $6.73 billion of insured limits at the end of 2025.

Nine questions on this page need a telephone rather than another search: whether Alabama’s beach pool writes fire as well as wind, whether the West Virginia plan is still writing, whether North Dakota’s facility was ever activated, whether South Dakota has a dormant chapter, whether the Hawaii Hurricane Relief Fund is active, the true Massachusetts policy count, whether Pennsylvania excludes theft or sells it, and everything past the statute in Kentucky and Ohio. The working file behind this page lists each one with the number to call.