The Evolving US Homeowners Insurance Market: Four Major Trends and What We Know About Them

The Evolving US Homeowners Insurance Market: Four Major Trends and What We Know About Them

1. Introduction

Homeowners insurance markets in the United States have exhibited issues in recent years. Premiums are on the rise, obtaining and keeping coverage has become difficult, and the coverage provided by insurance has not kept up with the costs of repair and rebuilding. News stories of problems in insurance markets abound, especially in the wake of major disaster events such as the January 2025 Los Angeles wildfires (Darmiento 2025).

In this issue brief, I describe what we know and don't know about four major trends in homeowners insurance markets: (1) rising premiums, (2) increasing numbers of policy cancellations and nonrenewals, (3) growth in residual market plans (often referred to as "insurance of last resort"), and (4) coverage gaps. I synthesize the trends and findings from a rapidly growing body of research in finance and economics.

Insurance acts as an important risk transfer mechanism, shifting the financial burden of infrequent but high-cost weather events from individual households to a diversified pool of risk-bearers. This function not only protects valuable household assets but also enables insurance to serve as a key pillar of US housing and mortgage markets. The four problems examined in this issue brief represent distinct ways in which this financial arrangement is coming under strain.

2. Insurance Premiums

Average homeowners insurance premiums rose almost 63 percent from 2017 to 2024, according to a study by Keys and Mulder (2025), who backed out individual insurance costs from mortgage escrow accounts. A 2025 report from the US Treasury's Federal Insurance Office, using data collected by the National Association of Insurance Commissioners, shows an inflation-adjusted increase in average premiums of 8.7 percent over 2018-22 period (US Treasury 2025).

These increases in average premiums mask significant geographic variability. Both studies find that average premiums and premium increases over time are significantly higher in areas where disaster risks are high. Grouping by disaster risk quintile at the zip code level, Keys and Mulder (2025) show that the top quintile paid significantly higher average premiums than the four lower quintiles in every year from 2014 through 2024 and that the gap increased over time. The US Treasury (2025) study reports that consumers in the top risk quintile paid average inflation-adjusted premiums over 2018-22 that were 82 percent higher than those in the bottom quintile. Both studies use expected annual building loss data from the Federal Emergency Management Agency's National Risk Index to create the risk quintiles. Keys and Mulder (2025) combine these data with data from First Street, a private data provider. See https://www.fema.gov/flood-maps/products-tools/national-risk-index for more information.

2.1. What Factors Explain the Trends?

In explaining premium increases, the insurance industry emphasizes the role of industry underwriting losses, which have been growing because claim payouts have outpaced premiums earned. To make up for the losses, insurers have increased the premiums they charge. Claims have risen as a result of a combination of rising costs of repairs and rebuilding and increases in property damage from weather extremes and natural disasters. Litigation costs are often held up as a contributor to rising premiums as well. They are a significant cost in Florida, but some reforms enacted in 2019 have reduced the number of lawsuits (Yaworsky 2026).

• Construction costs. According to data from Verisk, an insurance industry data analytics and catastrophe modeling firm, construction costs in the United States increased 2 to 9 percent per year every year from 2015 through 2025. Since 2020, the federal government's Producer Price Index for the construction sector has outpaced that for all commodities by a wide margin. Construction costs are available in quarterly reports from Verisk. See https://www.verisk.com/resources/campaigns/360value-quarterly-cost-updates/reconstruction-cost-analysis/?__FormGuid=6317ea9c-9d4b-41d4-beae-e368524905e5&__FormLanguage=en-US&__FormSubmissionId=288de012-0edd-4c0a-96f6-8c1f31dc471d. PPIs for each year, compiled by the US Bureau of Labor Statistics, are available at https://fred.stlouisfed.org/categories/31. Costs have risen as a result of the rising costs of materials, labor shortages, and productivity declines. Economists have written about stagnating productivity growth as a longer secular trend in the construction sector (Goolsbee and Syverson 2023). The reasons for the trend are unclear, but without productivity growth, wages and other input costs translate more directly into higher output prices.
  
  
• Damages from weather extremes. Losses from extreme weather events, which account for more than 90 percent of all homeowners insurance claims (Sastry et al. 2026a), have been on the rise because of increasing frequency and severity of events and greater risk exposure (i.e., more people and assets in harm's way). Increasing costs from weather extremes have also added to the cost of reinsurance, which is another contributor to rising premiums (Keys and Mulder 2025). Studies by the insurance industry emphasize the role of exposure. According to Swiss Re Institute (2026), three-quarters of the rise in global underwriting losses since 1970 are due to a combination of increases in population and per capita economic growth in risky areas. Wildfire losses are the exception. Only about one-third of the growth in losses from wildfires in North America comes from increasing exposure; much of the rest is a result of more frequent and intense wildfires.
  
  Several studies link losses to premiums, whether these losses are coming from escalation in hazard or exposure. Keys and Mulder (2025) find a statistically significant positive relationship between disaster risk and average premiums at the zip code level, and the relationship has become more pronounced over time: a one standard deviation increase in risk was associated with a $220 increase in premiums in 2017 and a $615 increase in 2024. Blonz et al. (2026), using individual property level data, estimate that every $1 increase in average annual disaster losses results in an increase of $0.94 in premiums. Walls et al. (2026) estimate that large losses from extreme events in a single year lead to premiums that are 6.2 percent higher, on average, the following year.

3. Insurance Policy Nonrenewals

Homeowners insurance is typically offered in a one-year contract, and most policies are renewed year to year. The US Treasury (2025) study shows that average nonrenewal rates changed very little over 2018-22 but rose by 120 percent in the highest risk quintile. According to an analysis of data from the National Association of Insurance Commissioners by Weiss Ratings, an independent insurance rating agency, the average nonrenewal rate across the 15 states with the highest nonrenewal rates in the country was three times higher in 2024 than in 2018 (Weiss Ratings 2025). In 2023, the average nonrenewal rate for the United States as a whole was 1.06 percent, but several Florida counties had 9-10 percent nonrenewal rates, California counties with high wildfire risks had rates of 7-8 percent, and the nonrenewal rates in Louisiana coastal parishes (county equivalents) were 4-5 percent (US Senate Budget Committee 2024).

3.1. What Factors Explain the Trends?

Although insurers may drop homeowners for property-specific problems such as an aging roof or other deficiency, In California, several roof types (e.g., flat roofs, wood roofs, roofs with double layers of shingles) can cause problems for homeowners looking for insurance coverage (Munce 2026). more often the phenomenon is a result of weather-related risks increasing to a level at which premiums earned cannot cover expected losses. Why haven't insurers simply increased premiums to compensate for the rising risk? In some places, they have; as described in Section 2, premiums have been rising in recent years in the highest-risk locales. However, 27 states require prior approval of rate increases from a state regulator. This process can take time, and increases might be denied (Oh et al. 2026). Most of the remaining states are considered "file and use"-that is, if a company files a rate increase, it can immediately raise the rates (NAIC 2026). There are some nuances in these categories; for example, some states allow "file and use" for a small increase and prior approval for larger ones. A few states are "use and file," meaning insurers do not have to file until after they begin using the new rate (LegalClarity 2026). Some economic research shows that regulations are a factor in nonrenewals.

• Regulations. California has long been recognized as having some of the most restrictive regulations of any state. Any proposed annual rate increase above 7 percent triggers a public hearing, in which consumer advocacy and other groups may participate. Insurers tend to want to avoid these hearings, and studies have found that most rate filings in California propose increases just below the 7 percent threshold (Boomhower et al. 2025).
  
  After major wildfires, California issues a one-year moratorium on nonrenewals in areas affected by the fires. Examining this policy with a difference-in-differences regression approach across regulatory boundaries, Taylor et al. (2025) find that the moratoriums backfired, causing larger-than-expected increases in nonrenewals and enrollments in residual market insurance after the moratorium periods ended. Several industry experts have highlighted these problems and recommended changes to California's rate regulations (Powell et al. 2024).
  
  Oh et al. (2026) exploit the regulatory differences across all states in a regression analysis and find that insurers adjust premiums less to underlying risks in highly regulated states and cross-subsidize with higher rates in less-regulated states.
  
  
• Not just regulations. Nonrenewals are more than just a California problem and are not only a result of regulations. Foster (2026) shows that nonrenewal rates have increased in recent years in 13 states, many in the Midwest, where losses from severe convective storms and hail have risen sharply. Even Oklahoma, which is a "use and file" state and has the highest average premiums in the country, has seen growth in nonrenewals.
  
  There is a paucity of research that fully examines these trends. In one study, however, underwriting losses show up as important. Using zip code-level data, Walls et al. (2026) find that a large loss year-that is, a year in which claims paid are greater than premiums earned-increases the average nonrenewal rate by 1.7 percentage points in the following year, an approximate doubling of the baseline rate.
  
  

Some insurers have opted to stop writing new policies altogether in some areas. In 2022, Allstate announced it would pause writing new policies in California, and in 2024, the company with the largest market share, State Farm, followed suit (Jacobson 2024). In the worst cases, insurers are completely leaving some states; nearly 20 companies pulled out of Louisiana's homeowners insurance market in 2022 and 2023 (US Senate Budget Committee 2024). Sastry et al. (2026b) show that exits left the Florida market served by financially fragile insurers-in other words, those that are less diversified, possess less capital, and have less reliable reinsurance.

4. Residual Markets

When homeowners lose their insurance and cannot find another policy in the state-regulated market (known as the admitted market), they have three options: go without insurance, purchase a policy from the surplus lines (unregulated) market, or purchase a policy from the residual market, also known as "insurance of last resort."

There is some evidence that an increasing number of homeowners are going without insurance (Dagher 2023; Cooley 2024), including a relatively high share of low- and moderate-income households (Zhu et al. 2026b). However, homeowners with a mortgage are required to have insurance. Surplus line insurance policies are expensive and historically have been used sparingly for residential properties. Although their use has grown in recent years, surplus lines still make up a small part of the residential market. A study by AM Best and the Wholesale and Specialty Insurance Association reports that surplus lines made up 12.3 percent of written premiums nationwide at the end of 2024, but their share of policies-in-force is undoubtedly much lower. See https://www.wsia.org/wcm/wcm/Foundation/Summary.aspx. According to the California Department of Insurance (CDI 2025), there were almost three times as many residential surplus line policies-in-force in California in 2023 as in 2015, but they still made up less than 1 percent of all policies in the state.

Most homeowners who cannot obtain insurance turn to the residual market, state-created insurance programs designed specifically for providing coverage to people who cannot find it elsewhere. Residual market insurance is provided by Fair Access to Insurance Requirements (FAIR) Plans, Beach and Windstorm Plans, and in two states (Florida and Louisiana), a state-run insurance company, Citizens Property Insurance Corporation. According to the Insurance Information Institute, there were an estimated 2.7 million residual plan policies-in-force in the United States in 2024, half of which were in Florida, with fully 86 percent in Florida, California, Louisiana, North Carolina, and Texas combined (III, n.d.). Total exposure in residual market plans in 2024 was $1.1 trillion.

Many of the residual market plans provide single-peril coverage, such as wind or fire damage; homeowners must have a separate homeowners insurance policy to insure against other losses. Beginning in 2024, the California FAIR Plan introduced a comprehensive homeowners policy in response to the rapid growth in FAIR Plan enrollment (albeit without liability coverage). Colorado, which introduced a FAIR Plan in 2025, has gone the opposite direction, with a policy that is fire only, covers only physical damage to dwellings and contents, has a coverage limit of $750,000 for residential properties (compared with California's limit of $3 million), and provides coverage for actual cash value, not replacement. See https://www.coloradofairplan.com/consumercoverageresources.

4.1. What Factors Explain the Trends?

Most of the factors that explain the trends in nonrenewals also explain the growth in residual market plans, as a nonrenewal is what often causes a homeowner to turn to the residual market (Taylor et al. 2025; Powell et al. 2024).

The California Department of Insurance reports that FAIR Plan policies increased from 1.6 percent of the state's residential market in 2015 to 3.7 percent by 2023, with most of the growth in high wildfire risk areas. By 2023, the FAIR Plan accounted for 32.6 percent of policies in the 10 counties with the highest wildfire risk (CDI 2025). As of March 2026, there were 684,388 California FAIR Plan policies-in-force covering $750 billion in asset value ($700 billion of that in the residential market) (California FAIR Plan Association, n.d.).

Citizens Property Insurance Corporation historically had a very large share of the homeowners insurance market in Florida because it provides full homeowners insurance, covering all perils and liability, and because its rates were below market for many years. Legislation passed in 2009 limited Citizens' rate increases to a maximum of 10 percent per year, which resulted in its rates falling below the market (Born et al. 2021). In addition, market instability in Florida driven by insurer insolvencies and third-party litigation costs led many homeowners to turn to Citizens (Born et al. 2021). At its peak in 2011, Citizens accounted for 23 percent of all Florida residential policies-in-force, but that figure had dropped to 3 percent by the end of 2025 (Citizens Property Insurance Corporation 2025). The state has actively sought to reduce homeowners' reliance on Citizens by requiring property owners who are offered a policy in the private market to take it and providing financial incentives to private insurers to take on property owners from Citizens.

4.2. Why Is a Large Residual Market a Problem?

A large and growing residual market can be a problem for several reasons.

• From a homeowner's perspective, most of the plans do not provide full coverage of all perils, necessitating a second policy, and the plans often have limits on the coverage they do provide. Thus homeowners often end up paying more for less coverage than with traditional private market plans.
  
  
• From a broader societal perspective, residual plans create an adverse selection problem-property owners in the riskiest areas and thus with the highest probability of loss end up in the pool. This leads to concentrated risk that is more correlated and volatile than a more diversified private insurer's risk would be. A study of the Florida market found an additional adverse selection problem: residual policyholders tend to choose lower deductibles, meaning that the pool isn't just riskier on average; it self-sorts to the riskiest coverage choices within the pool (Dumm et al. 2013).
  
  
• Losses in residual plans are typically covered by assessments on all insurance policyholders in a state (Watkins et al. 2023). This raises insurance costs for everyone and leads to a cross-subsidization from less risky to more risky areas and property owners. The cross-subsidization is exacerbated by residual plans, like Florida's, that are priced below market. Low prices draw in more customers than insurance of last resort is meant to cover, swelling exposure, a phenomenon seen in Florida and now in California (Nguyen et al. 2026). Furthermore, it can lead private insurers to retreat from the market.
  
  

5. Coverage Gaps

A growing body of research shows that many homeowners are underinsured. Klein (2026) examines claims for fire damage from more than 74,000 homeowners in California over the 2018-23 period and concludes that 71 percent of homeowners are underinsured. Sastry et al. (2026a), using nationwide individual household data, find that half of US households have less than 70 percent, and one-fifth less than 50 percent, of their rebuilding costs covered by insurance. Cookson et al. (2026) analyze data from the 2021 Marshall Fire in Colorado and find that of the 5,000 policyholders who filed claims, 74 percent were underinsured. Biswas et al. (2023) estimate that 40 percent of households in California that suffered damage in wildfires received insurance payments below replacement costs.

5.1. What Factors Explain the Trends?

• Coverage neglect. Cookson et al. (2026) conclude that an important reason for underinsurance is that homeowners tend to shop mainly on price-that is, they receive quoted premiums from different insurers and assume that each offers the same coverage, even though they do not, a phenomenon the authors refer to as "coverage neglect."
  
  
• Industry practices. Some studies lay the blame for inadequate coverage at the feet of insurers-specifically, the data and software algorithms insurers use, usually from third-party providers, which rely on faulty or out-of-date information to set replacement costs (Neilson and Munce 2025; Klein 2023, 2026).
  
  
• Mortgage lenders and government-sponsored enterprises. Mortgage lenders and government-sponsored enterprises set rules for minimum insurance coverage that are based on a home's outstanding loan balance. For borrowers who have equity in their homes, Sastry et al. (2026a) show that this leads to minimum insurance requirements that are less than replacement costs (with home equity essentially making up the difference). For homeowners who look for coverage above the regulatory minimum, the authors find that coverage is price-elastic, so as premiums rise, homeowners tend to drop coverage back to the mandated threshold, leaving themselves exposed. The extreme end of this phenomenon is homes without a mortgage altogether, and some evidence suggests that these homes are less likely to have insurance, leaving the household fully covering the costs if a disaster strikes (Zhu et al. 2026a; Bass and Fogerty 2026).
  
  
• Construction costs. Compounding these factors that researchers have identified is the fact that rebuilding costs often escalate after a disaster, when the need to repair and rebuild many homes in a given geographic area strains local construction markets and drives up prices and wages. This factor is not accounted for in typical replacement cost estimates.
  
  

Some evidence suggests that insurance deductibles create an additional coverage gap problem. Setting deductibles as a percentage of a home's value, for example, is becoming more common, which means that the higher a home's value, the larger the amount a homeowner must pay out of pocket when a claim is filed. Nineteen states have hurricane deductibles, which usually run 2 to 10 percent of a home's insured value (III 2025; UP 2026). Some California homeowners were surprised to learn when they filed claims that their policy had a separate wildfire deductible (Sacks 2026). This is a legally questionable practice in California, but most instances have been cases of carriers in the surplus lines market, which are not subject to the same regulations as carriers in the admitted market. These deductibles can also be another form of coverage neglect, something that homeowners fail to fully realize and account for when choosing a policy.

6. Conclusions

Topics

• Climate Risks and Resilience
• Insurance and Disaster Aid
• Climate Finance and Financial Risk

Topics

• Climate Risks and Resilience
• Insurance and Disaster Aid
• Climate Finance and Financial Risk