The American homeowners’ insurance market has been producing alarming statistics for years now, each one a little worse than the last. Inflation-adjusted premiums rose 28% nationally between 2017 and 2024.
In the five costliest states - Nebraska, Louisiana, Florida, Oklahoma and Kansas - homeowners now pay upwards of $4,400 a year on average, more than $2,000 above the national average. The average premium for a new policy reached approximately $1,950 by December 2025, an 8.5% year-over-year increase even after the market began to stabilise. Deductibles climbed roughly 22% in 2025 alone, as insurers transferred more financial exposure to policyholders.
Insurers have been cancelling policies, exiting markets, and in some cases going insolvent.
California's FAIR Plan - the state's insurer of last resort - received nearly 4,800 claims from the January 2025 Los Angeles wildfires, which produced estimated total losses that could exceed $250 billion, and needed a $1 billion assessment on private insurers operating in the state just to stay solvent. Florida has spent years managing an exodus of private carriers. Louisiana and Texas are not far behind.
Now, three economists with deep expertise in insurance markets are proposing a structural solution: a new federal entity, which they call "US Re," that would sell reinsurance to homeowners’ insurance providers to cover the most extreme climate-driven events.
The proposal, published this week by the Brookings Institution's Hamilton Project, comes from Benjamin L. Collier of the University of Wisconsin-Madison, Benjamin J. Keys of the Wharton School at the University of Pennsylvania, and Philip Mulder, also of the University of Wisconsin-Madison. It is, by design, deliberately limited in scope - and deliberately attentive to the failures of past government insurance experiments.
To understand why the proposal targets reinsurance rather than primary insurance, it helps to understand the peculiar economics of catastrophic risk.
When a hurricane strikes the Gulf Coast or wildfires devastate an entire California county, insurers face what they call correlated losses - vast numbers of claims arriving simultaneously from the same geography. The standard insurance mechanism of pooling risk across many policyholders breaks down when everyone in the pool is affected at once. To manage this, primary insurers purchase reinsurance - insurance for insurance companies - from global markets, effectively transferring some of their peak catastrophe exposure to large reinsurers such as Munich Re, Swiss Re, and Berkshire Hathaway.
But the cost of that reinsurance for US catastrophes has become, in the authors' framing, high and volatile. Reinsurers suffered heavy losses from the California wildfires in early 2025 even as a quiet Atlantic hurricane season provided some financial relief. The fundamental dynamic - more frequent and more severe weather events concentrated in areas where a great deal of insured property sits - is not improving.
When reinsurance costs rise, primary insurers have limited options. They can raise premiums, tighten underwriting, reduce coverage limits, or exit markets entirely. All four of those options have been on display simultaneously in the United States since roughly 2017. The authors argue this dynamic is structural, not cyclical, and will worsen as climate change continues to intensify catastrophic weather.
Under the proposal, US Re would function as a reinsurer of last resort for the most extreme tail events - covering losses so large and so concentrated that the private reinsurance market either prices them prohibitively or refuses to cover them at all.
The key mechanism is the federal government's cost of capital. Because the US Treasury can borrow at rates unavailable to private market participants, the authors argue US Re could price reinsurance for catastrophic risk at lower and more stable rates than private reinsurers can offer, without subsidising risk in the traditional sense. The entity would still price contracts according to expected losses and administrative costs — the point is not to make catastrophic risk cheaper than it actually is, but to remove the premium that private reinsurers must charge to compensate for their own cost of capital and return expectations.
The authors are careful to distinguish this from an implicit government subsidy. US Re would not underwrite at a loss, they argue. It would simply borrow more cheaply and pass that advantage on through contract pricing.
Three design principles govern their recommendation. First, price risk: contracts should reflect expected loss, not political preferences about affordability. Second, target market failures: US Re should intervene only where the private market demonstrably fails to provide coverage at reasonable cost, not crowd out private capacity. Third, maintain credibility: the entity needs clear legal authority to pay claims and structural independence from short-term political pressures - the latter being, as anyone familiar with federal insurance programs will recognise, considerably easier to recommend than to achieve.
No serious proposal for a federal role in property insurance can be written without reckoning with the National Flood Insurance Program, and the authors do not try to avoid it.
The NFIP, created by Congress in 1968 after the private flood insurance market effectively ceased to exist following catastrophic losses, has accumulated approximately $22.5 billion in Treasury debt - a figure that represents decades of premiums set below actuarially sound levels, political pressure not to raise them, and a series of catastrophic loss events that overwhelmed the program's reserves.
Hurricane Katrina, Superstorm Sandy, and Hurricane Harvey each tested the program to or beyond its limits. Congress has cancelled $16 billion of NFIP debt at least once, in 2017, to allow claims from that year's hurricane season to be paid.
The NFIP is the cautionary tale that haunts every federal insurance proposal, and for good reason. The mechanism by which a program intended to price risk accurately becomes instead a vehicle for subsidising development in flood-prone areas - because the political cost of raising premiums on existing policyholders is always higher than the political cost of deferring the reckoning - is well understood. It has happened before and the structural incentives that produced it remain entirely intact.
The Collier-Keys-Mulder proposal tries to inoculate US Re against this pathology through its three design principles, particularly the insistence on pricing risk rather than subsidising it and on political independence in setting prices. Whether such structural provisions can survive sustained political pressure from states with large concentrations of high-risk property - states that have considerable representation in Congress - is a question the proposal acknowledges without fully resolving.
The NFIP has also, in recent years, moved toward purchasing its own reinsurance from private markets, a program that FEMA has expanded steadily since its first placement in 2016. That program transferred some flood risk to private reinsurers and capital markets, reducing the likelihood of Treasury borrowing after catastrophic events. The US Re proposal inverts this model: rather than a federal program buying private reinsurance, it envisions a federal entity selling reinsurance to private primary insurers. The architecture is different; the underlying logic - that the federal government's balance sheet can absorb tail risk that private markets cannot efficiently price - is similar.
The insurance industry is not waiting for Washington. Carriers have been responding to the crisis with the tools available to them: sharper underwriting, property-level risk assessment, higher deductibles, and selective market exits. Several states have pursued legislative and regulatory remedies to stabilise their local markets.
Florida enacted legislation curbing litigation-driven claims inflation, a significant contributor to that state's insurance dysfunction, and has seen private carriers return to its market. California overhauled its rate-setting framework in 2024 to allow insurers to use forward-looking catastrophe models - rather than being limited to backward-looking historical data - and to pass reinsurance costs through to policyholders. Both changes are designed to make the state's market more attractive to carriers who had been pulling back.
AM Best elevated the homeowners’ insurance market segment from negative to stable in December 2025, reflecting some improvement in carrier profitability. Reinsurance prices have eased somewhat following the relatively quiet 2025 Atlantic hurricane season, though they remain at elevated levels after years of increases.
But the improvement is fragile and geographically uneven. The Excess and Surplus lines market - which operates outside standard admitted markets and typically carries higher prices and fewer consumer protections - has become a critical if expensive option for high-risk properties that admitted carriers increasingly decline to write. The Consumer Federation of America found that US homeowners spent $21 billion more on insurance in 2024 than in 2021. Repair and rebuilding costs have jumped nearly 30% over five years, and federal tariffs threaten to add further pressure in 2026 as material inventories decline.
The structural problem the Brookings proposal addresses - that catastrophic weather risk is becoming too concentrated and too correlated for private reinsurance markets to price efficiently - has not been resolved by market or state-level adjustments. Those adjustments have managed symptoms; the underlying disease, as the authors argue, requires a different kind of intervention.
The proposal arrives in a political environment that is, to put it charitably, not predisposed toward new federal programmes. The current administration has been sharply focused on reducing the scope of federal government activity, and a new entity with authority to write hundreds of billions of dollars in reinsurance contracts would represent a significant expansion of that scope, whatever its design principles.
There is, however, a bipartisan case to be made. The housing market distress caused by insurance unavailability and unaffordability crosses state and party lines. Florida and Louisiana are reliably Republican; California and New York are reliably Democratic; Colorado, Georgia, and Texas are contested. When insurance costs rise to the point of threatening home values and mortgage markets - household insurance costs now represent roughly 9% of a typical homeowner's monthly mortgage payment, the highest share on record - the political pressure to act does not respect partisan alignment.
The precedent the authors draw on extends beyond the NFIP. The Terrorism Risk Insurance Act of 2002, created after the September 11 attacks made terrorism coverage essentially unavailable in the private market, established a federal backstop for terrorism-related losses that has been renewed repeatedly with broad bipartisan support. The agriculture sector relies on a public-private partnership for crop insurance in which the federal government subsidises premiums and reinsures losses. Neither of these programs is without critics, but both represent durable federal commitments to underwriting risk that private markets cannot efficiently bear alone.
For insurance professionals, the proposal raises several questions that go beyond its political viability.
The design principle of targeting market failures rather than crowding out private capacity is operationally difficult. Defining the attachment point - the level of loss at which US Re's contracts kick in - determines how much risk remains with private reinsurers. Set it too low and you displace private market capacity; set it too high and the tail risk you leave in the private market remains too expensive for primary carriers to manage. The NFIP's experience suggests that getting that calibration right, and keeping it right under political pressure, is a persistent governance challenge.
The pricing discipline principle is similarly easier stated than maintained. Actuarially sound pricing for catastrophic risk requires forward-looking models, and forward-looking models for climate risk are themselves evolving rapidly. The growing precision of property-level catastrophe models - the same tools that California is now allowing insurers to use for rate-setting - would need to underpin US Re's contract pricing if the program is to avoid the NFIP's pattern of systematic underpricing.
Finally, the proposal's emphasis on political independence deserves scrutiny. The NFIP's pricing failures were not accidents of design; they were the predictable result of a program whose premium rates were subject to congressional override whenever raising them became politically inconvenient. US Re would need structural protections - an independent board, rate-setting authority insulated from the appropriations process, and perhaps a statutory requirement to price at expected loss - that are more robust than anything the NFIP has ever had. Whether such protections can survive sustained political pressure from states whose voters bear the direct cost of actuarially sound premiums is the central governance question the proposal does not fully answer.
That is not a criticism of the proposal so much as an honest acknowledgment of the problem it is trying to solve. The US homeowners’ insurance market is in structural distress driven by structural causes. The market and the states are managing. The question the Brookings authors are raising - whether management is enough, or whether the problem requires federal architecture - is one the industry will be navigating for years to come.
