What is the Future of State-Created Insurance Programmes?

Carolyn Kousky, Environmental Defense Fund, and Hannah K. Friedrich, University of Arizona and Cornell University, explain the various state-created disaster insurance programmes in the United States and discuss their future role in an era of increasing risk.

As the risk of climate-related disasters continues to escalate, property insurance markets in many locations around the United States have been destabilized (Kousky, 2024). In communities from California to Florida, it is getting increasingly difficult for residents in risky areas to find homeowners insurance from the private market – or find it at a price they can afford. When this occurs, many residents turn to so-called residual insurance markets or state-established insurance programmes, often referred to as “markets of last resort,” to provide coverage. In recent years, several of these programmes have been expanding, with increasing numbers of policyholders and more exposure. And at least one state recently felt it had to create a programme for the first time: over concerns about insurance availability, Colorado took steps in 2024 to establish a residual programme in response to growing wildfire and hail risk.

Here, we provide an overview of these programmes, current dynamics, and discuss the urgent policy questions as to how the costs of disasters should be distributed and whether these programmes still are, or can remain, true markets of last resort.

What is a residual insurance market?

There is a long history of the U.S. government providing disaster insurance when the private sector cannot do so. For flooding, this has been done federally through the National Flood Insurance Program (NFIP). The NFIP was established over 55 years ago to offer flood insurance in participating communities nationwide because such coverage was, and still is, hard to find on the private market (Kousky, 2018). In the NFIP, private insurers write policies and process claims, but the prices are set by the Federal Emergency Management Agency and the federal government holds all the risk, although they transfer some of this through the purchase of reinsurance and insurance-linked securities. Challenges in insuring other natural hazards have been addressed at a state level and those are our focus here.

Federal legislation in 1968 created state FAIR (Fair Access to Insurance Requirements ) plans, which were designed to address a lack of insurance in urban areas due to a combination of civil unrest and exclusionary financing resulting from redlining and other discriminatory housing practices (Nwokolo, 2023). The federal legislation provided federally subsidised reinsurance for losses from riots or civil disorder. Many of these programmes have since expanded to cover other risks, such as the California FAIR Plan, which is largely now a programme for wildfire coverage. Other state-created programmes include beach plans or wind pools that offer wind-only dwelling policies or full homeowners coverage along hurricane-prone coasts and the California Earthquake Authority for earthquake coverage. Two states, Florida and Louisiana, have a hybrid FAIR and beach plan in one combined programme. In total, 39 unique programmes operate in 34 states, plus the District of Columbia.

Most of these programmes were created decades ago, often when a large disaster led to turmoil in the private market (Kousky, 2011). For example, the Louisiana and Alabama programmes were created in response to Hurricane Camille in 1969, while the California Earthquake Authority was created after the Northridge earthquake in 1994. Some of the longer-standing programmes have seen changes or reorganisation, such as reforms to Florida’s state programmes following Hurricane Andrew in 1992.

Residual insurance programmes vary in structure. They are typically created by state action, but operate independently and are non-profits. FAIR plans, for example, are not state agencies but involuntary associations of all admitted insurers in the state. Residual programmes are not directly backed by taxpayers. They typically have assessment authority to cover large losses, either on admitted insurers in their state or directly on other policyholders statewide. This creates a cross-subsidy from low-risk and future policyholders to current policyholders, which we discuss further below.

Table 1 provides an overview of select residual market programmes that have a non-trivial number of policyholders and cover climate-related disasters, our focus in this comment. Data comes from reports from the Property Insurance Plans Services Office (PIPSO, 2024a; 2024b).

How has participation in these programmes been changing?

The absolute size of these programmes varies considerably across states. Figure 1 shows total exposure in the state programmes over the time period for which we have consistent data: 2013 – 2023. Florida, as seen in Figure 1, is by far the largest programme. More recent public information indicates that as of January 2025, Florida Citizens had approximately 983,000 residential policyholders and $405 billion in coverage. In contrast, programmes in Alabama, Mississippi, and South Carolina, for example, are much smaller, with approximately 18,800, 13,700, and 16,300 policyholders in each programme, respectively, as of 2023. These differences in size are driven by many factors, including the total population in areas eligible for coverage within the residual programme and the availability of private insurance offerings, which itself is a reflection of both risk levels as well as state regulatory dynamics.

Figure 1: Total exposure (billions USD) by year for select residual insurance programmes; 2013-2023
Data source: PIPSO (2024a).

These programmes also have different relative footprints within their states. Table 2 provides information on the percent of eligible households that have an insurance policy with the state residual programme, as well as the average claim payout in 2023. The programmes vary in geographic scope; those that offer coverage statewide are in bold. While a sizable share of households are in Florida Citizens, the statewide share for California is fairly small — but if we were to limit examination to only areas of very high wildfire risk, it would be larger. For example, almost 15% of the CA FAIR Plan market share is in the zip code where Pacific Palisades is located (Araullo, 2025). The coastal programmes also vary dramatically across states in the share of coastal homes they cover, from a low of 1% in Georgia to a high of over 60% in North Carolina. Again, this is driven by many factors. There is concern in North Carolina, for example, about suppression of rates driving more residents into the residual programme (APCIA 2025), but another key factor is the extent of the geography. The North Carolina programme, for example, only offers policies in a very narrow high-risk area. If its footprint of eligible homes were larger, including lower-risk areas, the share of households needing the programme would be smaller.

As shown in Figure 1, total exposure in some residual programmes has been growing more recently. This is most pronounced for Florida Citizens and the California FAIR plan. Louisiana Citizens, although smaller in absolute numbers, has been increasing since 2021. North Carolina and Texas have seen smaller increases. The share of policyholders in North Carolina has been fairly stable; their policyholder growth reflects an increasing population in the areas they serve.

The change in the size of these programmes reflects both availability and pricing in the private market, as well as the policy terms and requirements of the residual programme. For example, Figure 2 shows the number of policyholders in Florida Citizens back to 2009. The variations are both a reflection of the extent of pullback from the private sector, which, for example, drove increases in Citizens policyholders over the past 5 years, but also state policy toward the residual programme, which has varied substantially over the decades (Kousky and Medders, 2024). For example, after the 2004 and 2005 hurricane seasons, changes were adopted that made Citizens less expensive and more attractive, leading to a peak in policyholders in 2011. This was then reversed, and measures were adopted to shrink Citizens and keep people in the private market, leading to declining policy counts for many years. Similarly, recent legislative changes have reversed some of the private market pullback and refocused Citizens on moving policyholders into the private market, leading to a drop again in 2024.

Figure 2: Florida Citizens Policy by Year, 2009-2024
Data source: Florida Citizens website, www.citizensfla.com/policies-in-force

As shown in Figure 3, the California FAIR programme has also been growing recently, beginning with the 2017 and 2018 severe wildfire seasons that triggered exits of several insurers (Kousky, 2025). Since 2019, the number of policies in force has grown by more than 200%. The California FAIR plan had $458 billion in exposure this past fall, with a reported $5.9 billion in exposure in Pacific Palisades alone. This is up from only $50 billion in 2018 (Roach, 2024).

Figure 3: California FAIR Plan Policies by Year, 2009 - 2024
Data source: PIPSO (2024a) and CalMatters.

After a decade of steady decline, policies and exposure have also risen in Louisiana as shown in Figure 4 (note y-axis scale and years represented varies from Figures 2 and 3). Since the catastrophic 2020-2021 hurricane seasons, the number of policies underwritten by Louisiana Citizens has more than tripled (Louisiana Citizens, 2023), from about 48,000 in 2021 to 184,000 in 2023. In addition to the significant increase in policies, the insured value of properties increased from $6.7 billion in 2021 to $33.3 billion in 2022, a 397% increase in total insured value (Louisiana Legislative Auditor, 2022).

Figure 4: Louisiana Citizens Policies by Year, 2009 - 2023
Source: PIPSO (2024a).

Figure 5 shows several other residual programmes that are much smaller in size. Of these, Texas and North Carolina are the largest and both programmes have also seen increases in recent years. The other coastal programmes are substantially smaller. The Massachusetts FAIR plan covers many coastal properties, and has been declining.

Figure 5: Policies in Force for Other State programmes Covering Climate-related Perils, 2009-2023
Data source: PIPSO (2024a).

What makes a “market of last resort”?

Most residual market programmes were created to provide coverage in periods of hard insurance markets, when supply is scarcer and prices high, or to serve high-risk but geographically localised areas with insufficient private market supply, such as along a hurricane-prone coast. Now, these programmes face questions about their role in states where more people are struggling to find coverage – or affordable coverage – in the private market. The risks of climate-related disasters are growing as the planet continues to warm. Absent widespread and large investments in risk reduction and climate adaptation, this increasing risk will continue to make private market insurance more expensive and difficult to obtain, driving up demand in residual programmes.

Initially envisioned as markets of last resort, state residual programmes adopted policies to ensure residents only purchased coverage from them when they had truly exhausted all private market options. Once in the programme, the goal was to get consumers back to the private market as soon as possible. To this end, some programmes put in place surcharges, limited the coverage available via restrictions in policy forms or the type of coverage, capped the maximum limit available, restricted the geography in which programmes could write, or required proof that a policyholder had been denied coverage in the private market. Others have taken active steps to “depopulate” their programmes and move policyholders back to private market carriers, although sometimes with concern from consumer groups about placing policyholders with riskier insurers or higher-priced policies. All these various measures make these programmes a true last resort option for consumers and not competitive with the private market.

Amidst increasing strain on the availability and affordability of coverage in the private market in some locations, however, pressure has been building from consumer groups and others that residual insurance programmes relax these restrictions and shift to become programmes that offer affordable coverage to any resident — that is, transition from being markets of last resort to markets of first resort. For instance, Louisiana recently paused a surcharge that Louisiana Citizens historically charged policyholders, and the CA FAIR plan increased the coverage it would offer homeowners to $3 million and greatly expanded coverage for commercial policyholders. Such changes make the residual programme more attractive and thus also more competitive with the private market.

The cost of insurance in residual programmes: who pays for disasters?

The price of insurance is fundamentally related to the probability that the insurer has to pay claims and the potential magnitude of those claims. The higher the risk, the higher the price of insurance. The price of insurance, however, also includes other components that cover various firm costs, including marketing and distribution, underwriting, commissions for agents and brokers, and claims handling. It will also include some amount for profit and taxes. As such, insurance costs more than just the expected damages. While residual markets also have operational expenses, they generally do not have to pay taxes (except for programmes in Connecticut, Kentucky, Massachusetts, Rhode Island, and South Carolina) and many function as non-profits, limiting the profit load. As such, they may be able to offer coverage somewhat cheaper than the private sector.

That said, residual programmes face the same fundamental economics of disaster risk as the private market (Kousky, 2022). Since disasters can impact everyone simultaneously and losses can be quite high, insurers need access to substantial amounts of capital to prevent insolvency. For private firms, this financing must be in place pre-disaster, and includes the surplus they hold, reinsurance purchases, and use of any financial instruments. All of that will be reflected in the price.

Many residual insurance programmes, however, purchase less reinsurance or other pre-event financing mechanisms and instead rely on post-event financing in the event of large losses, which spreads the cost of disaster claims beyond current policyholders. This is typically done using post-disaster assessments either on admitted carriers in the state (some of which will then pass the cost through to their policyholders) or directly on policyholders, perhaps coupled to issuance of bonds. Typically policyholders across many lines of coverage and across the entire state are subject to assessments and pass-throughs, not just those in the programme. The structure of assessments varies somewhat across programmes but all result in cross-subsidies, with the cost of a disaster being partially covered by future policyholders and by lower-risk policyholders elsewhere in the state (Watkins et al., 2023).

For instance, after Hurricane Katrina in 2005, Louisiana Citizens issued $978 million in bonds to cover deficits, repaid through annual 1.36% assessments on all residential and commercial policies in the state. These assessments to pay Katrina losses started in 2007 and were planned to end in 2026 but ended a year early. Similarly, the 2004 and 2005 hurricane years also required Florida Citizens to go to the bond market to pay claims and then repay the debt with assessments on policyholders throughout the state. In Florida, Citizens first assesses its own policyholders up to 15% of the premium. If this surcharge is insufficient, Citizens can then levy an emergency assessment of up to 10% per year on all lines of insurance written in the state, with the exception of workers' compensation, medical malpractice, and flood. This means that hurricane losses for Citizens policyholders are partially paid for by households not impacted by the storm and on policy lines not related to homeowner damage, such as business and auto policies (Kousky and Medders, 2024). In California, following losses from the L.A. fires, the FAIR plan faced a funding gap for claims, necessitating a $1 billion assessment on insurance companies operating in the state, half of which they will be able to pass on to future policyholders everywhere in California.

These approaches to paying disaster claims create hidden cross-subsidies from lower-risk and future policyholders to those in higher-risk areas that suffer disasters. Private insurers may also have cross-subsidies when regulators suppress rates within their states, but they do not utilise post-event financing. This raises many questions about who should pay the costs of disaster insurance. Keeping costs lower than the amount needed to cover the risk for some people means others will pay higher costs. Keeping costs lower today means that the costs will need to be paid in the future. This raises questions of intergenerational, geographic, and financial fairness and equity, as well as economic efficiency. Should residents in low-risk areas pay more so the costs in high-risk areas are lower? Should higher-income residents pay more to keep costs lower for disadvantaged residents? Does suppressing pricing lead to lower incentives to reduce risk or focus development in safer areas?

What is the goal of residual markets?

Decisions about pricing and financing of residual insurance programmes should be driven by the overall objectives of the programme and who they are designed to serve. Most residual insurance programmes have written in their mission that their task is to provide insurance only when private firms do not, and they view success as moving policyholders back to the private market. For example, the California FAIR plan website notes, we “reach success when we are no longer needed,” and Florida Citizens highlights on their website that they are “focused on promoting a healthy property insurance market.” (Note, this is quite different than the federal NFIP, which is a programme of first resort, as it provides over 90% of residential flood insurance in the country and has no programme goals to keep policyholders out of the programme.)

While most state residual programmes have these goals to encourage private market policies first, some have adopted programmatic approaches at different points that make them more competitive with the private market, such as earlier choices in Florida to keep rates lower (Kousky and Medders, 2024). Indeed, Florida Statutes note an affordability goal for Citizens, beyond just availability, stating “The Legislature intends, therefore, that affordable property insurance be provided and that it continue to be provided, as long as necessary, through Citizens Property Insurance Corporation,” (Section 627.351(6)(a)). Affordability seems to have also motivated the recent decision to remove the Louisiana Citizens surcharge through 2027. Decisions around pricing, however, could be better made following an explicit discussion of which populations should receive lower rates, who pays those costs, what it means for public sector budgets, and how to mute perverse incentives. For instance, instead of suppressing insurance prices across the board, these programmes could instead have means-tested discounts for lower-income policyholders. This could either be paid for as a state or federal benefit programme or through a programme that provides rebates to lower-income residents on a sliding scale that becomes a fee for higher-income policyholders. This would assist those who need the financial protection of insurance the most, without jeopardising the financial soundness of the programme and without distorting risk signals for others.

Indeed, there has long been hope from economists that risk-based insurance prices would act as a signal about risk levels, as well as a financial incentive to lower risks. For well-documented loss reduction measures, such as installing a FORTIFIED roof, many insurers will reward homeowners who make such retrofits with lower premiums (or are required to do so by insurance regulators). This can be an important benefit, but there are many other confounding factors that can make it difficult for homeowners to invest in loss reduction measures, even if they get a small premium reduction, such as budget constraints, long payback periods, lack of information on risks and the necessary investments, and high transaction costs (such as finding a reputable builder who is knowledgeable about resilient construction). There are other explicit steps residual programmes can take to lower losses and overcome these challenges for policyholders. Possible support could include endorsements, which pay for stronger buildings at the time of a claim, or offering direct grants for critical retrofits. The residual programme in North Carolina, for example, has implemented a grant programme to partially cover the costs of installing Fortified homes on roofs for qualified policyholders at the time of rebuilding. They have found that these investments generate cost savings for the programme in lower reinsurance costs and future claims potential (Turner, 2024).

As we have failed to decarbonise at the pace and scale needed, the risks of weather-related extremes will continue escalating in the coming years and decades. This will continue to stress the private insurance market and increase demand for public insurance programmes. It will also necessitate higher premiums — in the private sector or residual programmes. Focusing on lowering losses through climate adaptation can stabilise insurance and keep people safer. In addition, a broader conversation about when and how to socialise insurance costs is needed in the United States to guide reform of our residual insurance programmes.

References

Acknowledgments

Thanks to Stephen Jablonski and Jason Foote for their assistance with the PIPSO data, including clarification of coverage details and financing structures. We also thank Talley Burley, Jesse Gourevitch, and John Aloysius Zinda for comments on earlier versions of this commentary.

Image Credit

Shutterstock