Narrative-first, evidence-driven

Back to case studies [pdf]

Receipts: tracked in Methods and Sources by type: Official data | Independent analysis

Three things that happened to real people

A homeowner’s escrow payment jumped by hundreds of dollars a month. Nothing about their property changed. No claim, no renovation, no new risk. Their insurer had repriced the ZIP code.

A driver with a clean record, no accidents, no tickets, saw their full-coverage premium rise for the third consecutive year. The increase had nothing to do with them. Repair costs and litigation in their state had risen across the board.

A buyer made an offer on a house, got financing, and then couldn’t close. The property was no longer insurable at a price that fit the loan terms. The standard carriers had exited that market the year before.

These are not separate problems. They are insurance acting as a cost amplifier.

How the pressure built

The US property and casualty insurance market ran without major reform through decades of rising climate exposure, construction cost inflation, and expanding litigation - then absorbed all of it in a compressed repricing window between 2022 and 2024. Some states responded with targeted reforms. Others tried to hold prices down. The responses targeted different layers of the problem, and the outcomes differed accordingly.

How the mechanism works

Insurance is where a lot of separate pressures show up as one bill. Climate risk, rebuild costs, repair costs, litigation, and reinsurance do not arrive separately at the household level. They get bundled into a premium. When several of those pressures rise at once, the bill can jump all at once too. The basic mechanism:

1. Risk rises. More severe weather events, higher-value assets, longer rebuilds.
2. System costs rise. Reconstruction labor and materials spiked 16.6% in a single year (Q2 2020 to Q2 2021, CoreLogic). Auto repair costs rose approximately 33% over four years. Litigation costs rose in high-liability states.
3. Reinsurance tightens. Property catastrophe reinsurance rates rose 45% to 100% in the US in 2023 alone, forcing primary insurers to absorb more risk per event. This is invisible to consumers but directly drives primary premiums.
4. Premiums rise. US home insurance premiums rose 40.4% cumulatively from 2019 to 2024. Auto full-coverage premiums rose 46% from 2022 to 2024. CPI for motor vehicle insurance rose 20.3% in a single year (Dec 2022 to Dec 2023), then an additional 11.3% the year after - fast enough to dominate changes in total transportation affordability even when vehicle prices were flat.
5. Where pricing cannot adjust cleanly, markets thin or exit. In California, where Proposition 103 (1988) required prior regulatory approval for rate increases - averaging 138 days without intervention, 343 days with - insurers had a harder time pricing forward-looking climate risk. In a market already under pressure, that helped push some carriers from repricing into withdrawal.

This is the core mechanism: holding down the signal does not remove the pressure underneath it. It just delays the moment when the bill catches up.

The amplifier has two visible outputs: higher prices for the insured, and silent exposure for the underinsured or uninsured.

Two additional channels that hit people who never buy a policy:

• Renters. Federal Reserve analysis finds multifamily property insurance costs rose from $39/unit/month (2019) to $68/unit/month (2024) in real terms - up 75% (Federal Reserve FEDS Notes on rental pass-through; specific paper not individually identified; label: plausible). Rents tend to rise alongside those costs. A renter who never touches an insurance application still absorbs the increase through their monthly payment.
• Force-placed insurance. When a homeowner’s coverage lapses or is canceled, the loan servicer can impose lender-placed insurance - which can be twice as much as regular coverage (CFPB). This hits financially stressed households at the worst possible moment: the premium spikes just as they can least afford it, and the coverage primarily protects the lender, not the homeowner.

Mechanism evidence

Key metrics:

• Home insurance premiums: +40.4% nationally, 2019-2024 (Insurify / insurance.com)
• Auto full-coverage premiums: +46% from 2022 to 2024 (The Zebra / Insurify); CPI for motor vehicle insurance: +20.3% (Dec 2022 to Dec 2023), +11.3% (Dec 2023 to Dec 2024) - these annual figures show how concentrated the spike was
• AAA 2025: total annual cost to own and operate a vehicle is $11,577; insurance accounts for $1,694 of that, roughly 15% (AAA Your Driving Costs 2025; label: confirmed directionally, specific figures from AAA annual report). When insurance inflation runs at 20%, it moves the total cost of car ownership more than almost any other line item
• Home insurance now accounts for 9% of a typical homeowner’s monthly payment - highest share on record
• NBER Working Paper 32579: premium increases have reduced home price growth by over $40,000 in the most-exposed zip codes
• 2025: national auto average dropped 6%; 39 states saw reductions - the first evidence of a reversible cycle
• Industry combined ratio (overall U.S. P&C; NAIC/industry sources): 2022 approximately 100.7 (unprofitable), 2023 = 101.8 (unprofitable), 2024 = 96.6 (profitable) - see profit tension discussion below

Time horizon: 2019-2025 for premium trends; reform outcomes are 2019-2024 for Michigan; 2022-2024 for Florida.

This transmission mechanism works best in competitive markets with forward-looking pricing. When pricing is administratively constrained or markets collapse into an insurer of last resort, the pattern changes - not into accurate prices, but into exit and coverage gaps.

Premium trends are confirmed. What remains unclear is how much of any given increase comes from risk, system cost, or margin expansion. No study has cleanly separated those pieces at the consumer level.

Full premium trend data, profitability figures, reinsurance detail, and catastrophe loss data are in the Research receipts appendix below.

Three state comparisons

Where reform matched the mechanism

Michigan auto reform (2019)

Michigan was the only state requiring unlimited personal injury protection (PIP) - uncapped medical coverage - under a no-fault system. This drove the highest auto insurance premiums in the country. Governor Whitmer’s bipartisan reform legislation (signed May 2019, effective July 2020) created PIP coverage tiers, giving consumers options from unlimited down to $50,000.

Result: average auto insurance premium per vehicle fell $357 (18.8%) from 2019 to 2024. Wayne County (Detroit) average reduction: $539 per vehicle. Drivers over 64 - already Medicare-covered - saved 22% on average. The mechanism was direct: remove a mandated cost that exceeded what most consumers needed or wanted.

Florida property insurance reform (SB 2-A, 2022)

Florida’s property insurance crisis had a specific mechanism: assignment of benefits (AOB) fraud and litigation abuse. Between 2008 and 2018, property insurance AOB lawsuits increased 900%. Florida accounted for 7% of US homeowners insurance claims but 76% of US homeowners insurance lawsuits. Litigation filings ran at approximately 8,000 per month in early 2023.

SB 2-A eliminated AOB for property insurance and restricted attorney fee arrangements. Post-reform: litigation filings dropped to approximately 4,000 per month by November 2024. Property claims lawsuits returned to 2019 levels. Florida domestic P&C insurers moved from a net loss of $751 million (2022) to net income of $944 million (2024). Fourteen new insurance companies entered the market.

Michigan’s reform addressed the mandate itself rather than trying to regulate prices around it. Florida targeted the litigation abuse mechanism directly. Both point to the same lesson: reforms work better when they reach the actual cost driver.

Where policy missed the mechanism

California Proposition 103 rate suppression

Prop 103 required prior approval for all rate changes. Insurers could not use forward-looking catastrophe models or include reinsurance costs in rates. The result: prices that did not reflect current or projected risk - and consumers still lacked a simple, visible fairness test for what they were being asked to pay. The gap between regulatory timing and market conditions widened until exit became the only rational insurer response.

The January 2025 Los Angeles wildfires tested this directly. The Palisades and Eaton fires burned 37,000+ acres and destroyed 16,000+ structures. Estimated insured losses: $25-45 billion. Total economic losses: $250-275 billion (AccuWeather). The insured fraction was approximately 10-18% of total economic loss. Many affected homes had been dropped by standard carriers in 2023-2024 and were on the FAIR Plan - the state’s insurer of last resort - or were uninsured entirely.

California’s Sustainable Insurance Strategy (2023-2024) is attempting to correct this: it permits forward-looking catastrophe models, allows reinsurance costs in rate calculations for the first time, and requires insurers accepting these tools to write a minimum 85% of new policies in historically underserved areas. Whether this has reversed the exit trend is not yet confirmed - implementation is too recent for outcome data.

Where the story gets messy

This is the part where the clean story breaks down a little.

The profit tension

The 2024 P&C profitability surge complicates any simple “pass-through” story. The industry made $169 billion in net profit in 2024, up 90% year over year. That figure appears across multiple sources, though AAJ foregrounds it as evidence of extraction. The ex-catastrophe combined ratio was 88%, the best in at least 20 years.

Two interpretations fit the data:

1. Catch-up lag: The industry ran unprofitable combined ratios in 2022 and 2023 after years of underpriced risk. The 2024 recovery reflects a repricing cycle completing - not steady overearning.
2. Overshoot: The repricing cycle went further than underlying risk warranted. Once claims normalized, rates stayed elevated. Insurers captured excess margin during normalization.

A third structural factor shaped the 2024 result directly. Record profitability in a high-catastrophe year seems counterintuitive until you account for market exits. When State Farm, Allstate, and other carriers stopped writing in California and other high-exposure states, those policies migrated to FAIR Plans - which sit outside the private market’s combined ratio calculation. The carriers who remained were writing a less-exposed book; the worst risks had been selected out. Higher premiums on a smaller, lower-risk portfolio is the mechanical setup for a record combined ratio improvement even in a hard cat year. That structural shift does not resolve the catch-up vs. overshoot debate - it complicates both.

Both narratives have evidence. The American Association for Justice and Consumer Watchdog cite the profit surge as evidence of extraction. The industry response is that two years of losses came first, and that Fortune 500 average net margin is 13% versus a P&C operating ratio of approximately 7% in 2024. The 2025 national auto premium drop of 6% across 39 states - without any major underlying cost reversal - is at least some evidence for the overshoot view: rates may have stayed high longer than the risk justified before starting to come back down. The answer probably differs by line, carrier, and state. The honest version is that this dispute does not resolve cleanly.

The social inflation dispute

Nuclear verdicts (jury awards exceeding $10 million) more than quadrupled between 2020 and 2024. In 2024: 135 nuclear verdicts (+52% year-over-year), $31.3 billion in awards (+116% year-over-year), median award $51 million (Marathon Strategies Nuclear Verdicts Report 2025 - a defense-aligned source).

The Consumer Federation of America disputes the “social inflation” narrative directly, arguing it is an industry-constructed concept to justify price increases and that actual claims payouts were not consistently rising in periods when insurers cited litigation as a driver. The increase in nuclear verdicts is clear. The causal link to overall premium levels remains contested.

The distributional angle

The amplifier does not hit evenly. The FTC found that credit-based insurance scores are predictive of risk - but also that African-Americans and Hispanics, as groups, tend to have lower scores than non-Hispanic whites and Asians, which means the price burden can land unevenly across groups. Neighborhood-level pricing compounds this: places already carrying more economic stress can also get hit with higher insurance premiums through the same pricing system.

The uninsured feedback loop makes this worse. When premiums rise, some households drop coverage rather than pay. The Insurance Research Council puts uninsured motorists at 15.4% of drivers in 2023 and underinsured at 18.0%. More uninsured drivers in the pool means more uncompensated losses, which pushes premiums higher for everyone who is still paying - a regressive spiral. The households priced out of insurance are not just absorbing risk themselves; they are raising costs for the households that remain.

The uninsured and underinsured gap

High prices are one visible output of the amplifier. The other is unavailability. A third channel runs through existing policyholders who are structurally underinsured: the 2021 reconstruction cost spike left many properties insured at pre-spike replacement values. A policy priced in 2019 may now cover less than 85 cents on the dollar of a current rebuild - the gap opened silently, with no new bill and no warning. The premium stayed manageable; the coverage gap did not.

Approximately 7-14% of owner-occupied homes are uninsured nationally (CBO 2024, JEC 2024 - the range reflects methodological differences). Flood insurance gap: $24.4 billion in estimated future annual flood losses on single-family homes; more than $17 billion currently uninsured. In low-income neighborhoods, over 90% of households are underinsured; average uninsured flood losses exceed 20% of annual income [label: plausible - secondary citation in Philadelphia Fed paper; not independently verified].

The amplifier works in both directions: too expensive for some markets, unavailable for others, often because of the same underlying cause.

These conditions affect tens of millions of households and shape mortgage qualification, regional housing markets, and transportation costs across the country. The Florida and Michigan reforms matter, but neither addresses climate exposure, construction cost inflation, or the broader uninsured gap at scale. They are tools within a larger system problem.

What the market showed

After the California exits, policy volume collapsed into the FAIR Plan, which held 2.5% of the statewide market but 20.4% of policies in wildfire-exposed ZIP codes. That is not a healthy competitive market. It is a sign of market failure under stress.

In Florida, market recovery post-reform is measurable: 14 new carrier entrants, return to collective profitability, litigation filings halved. That recovery is confirmed for 2022-2024. It has not been stress-tested against a major catastrophic hurricane event, and the Governor himself flagged Citizens as potentially insolvent under that scenario.

In Michigan, reform passed in 2019 and produced confirmed aggregate savings by 2024. Whether those benefits flowed fairly to the most burdened Detroit drivers remains unresolved. The open question is distributional: whether low-income Detroit drivers saw their expected share of the savings. The current data does not settle that.

Auto insurance nationally: the 6% price drop in 2025 across 39 states is early evidence that the 2022-2024 spike was partly a repricing overshoot, not a permanent step change.

What to do

Short - do this now

• Check your homeowner policy’s replacement cost estimate against current construction costs. The 2021 spike left many policies covering less than 85 cents on the dollar of a current rebuild - silently, with no new bill. Call your insurer and ask for a current estimate.
• If you’re in a state where major carriers have exited, ask your agent to compare FAIR Plan terms against surplus lines options before defaulting to the state backstop. FAIR Plans are designed for availability, not value.
• Housing operators: verify your Statement of Values is current. Stale or undervalued data is one of the most common reasons for pricing penalties at renewal.

Medium - requires planning or investment

• Home hardening (resilient roofing, fire-resistant materials, stormwater barriers) reduces insurer-assessed risk and can qualify for premium discounts. Several states fund part of this: Alabama’s Strengthen Alabama Homes Program, Louisiana’s Fortify Homes Program, Colorado HB25-1182. Check what your state offers before paying out of pocket.
• Housing operators: document every capital improvement and share it proactively with your underwriter. Bid the full program to market at each renewal rather than auto-renewing.
• Housing operators with larger portfolios: investigate captive or risk retention group structures. The Housing Partnership Network model covers 85,000+ units across 23 organizations and has demonstrated measurable cost stabilization (Enterprise Community Partners, 2026).

Long - legislative and structural

• Ask where your state stands on forward-looking catastrophe modeling in rate filings. States still on backward-looking approval are building the next coverage gap.
• Support home-hardening grant programs that tie state funding to mandatory insurer premium discounts - the mechanism that makes the investment durable, not one-time.
• Track Fannie Mae / Freddie Mac deductible reform. Current GSE guidelines cap property deductibles at $50,000 for properties under $10M in value, well below commercial norms - reform here would unlock risk-sharing options for affordable housing operators.

Bottom line

Insurance costs are not some separate category of squeeze. They are where a lot of other pressures land all at once, making housing and transportation harder to afford even when nothing changed about the household itself. And the amplifier runs in both directions: too expensive in some places, unavailable in others, often for the same underlying reasons.

That makes this both a cost problem and a trust problem. Real risk and real system costs do show up in premiums, but most people cannot see the logic well enough to tell the difference between warranted pricing and a ripoff. When that happens, anger goes looking for villains instead of mechanisms.

North Star verdict

Put plainly, the three state cases show three things:

Reform works best when it reaches an actual cost driver. Michigan eliminated a mandated cost that exceeded consumer needs. Florida reduced a litigation mechanism that was inflating claims costs inside a broader crisis. Both produced measurable results within their scope - though Michigan’s aggregate savings may not have reached low-income Detroit drivers proportionately, and Florida’s recovery has not been stress-tested by a major hurricane event. Neither reform addressed climate exposure, construction cost inflation, or the broader uninsured gap. Those remain live system problems.

Suppressing the price signal does not reduce the underlying pressure. California’s Prop 103 framework held prices down for decades, but when the gap between regulatory timing and actual risk became too large, that delay became part of what pushed some insurers out instead of letting them reprice gradually. The FAIR Plan absorbed the overflow. The January 2025 wildfires exposed what that meant for coverage gaps. California’s Sustainable Insurance Strategy is an attempt to maintain oversight while allowing pricing to track real costs - whether it succeeds is not yet known. The takeaway is not to remove oversight. It is to make oversight responsive to present conditions rather than anchored to old ones.

Residual markets are a system signal, not a solution. When the FAIR Plan holds 20% of policies in wildfire-exposed ZIP codes, that is market failure visible in aggregate data. The Florida and Michigan reforms moved markets back toward private coverage. That is the direction reform should push - but it requires reaching the mechanism, not just adjusting the price.

All three cases rest on competition doing some of the disciplining - once a major cost driver was addressed, private markets re-entered and prices moved. But that only goes so far. Insurance is often a captive market: mandatory for access to housing and mobility, with exit options that collapse under stress. Where market discipline cannot do enough on its own, oversight is not the enemy of accurate measurement. It is part of what makes the measurement trustworthy.

Insurance is a measurement device - when it breaks, fix what it is measuring, not just the reading. But in a captive market, someone still has to check the scale.

Research receipts

Receipts

Supporting data for the main narrative. Claims carried into the main flow are drawn from this evidence base.

Homeowners premium trend:

• Cumulative increase 2019-2024: +40.4% nationally (Insurify / insurance.com)
• Year-by-year: +3.0% (2021), +5.4% (2022), +11.0% (2023), +11.4% (2024)
• Highest cumulative increases 2019-2024: Colorado +76.6%, Nebraska +72.3%, Utah +70.6% - not California or Florida (confirmed counterintuitive finding; driven by severe convective storms and construction cost inflation, not coastal climate disasters)
• Smallest 2024 increases: Florida (+1.7%), Texas (+3.4%), New York (+3.8%) - interpreted as markets that had already repriced earlier

Auto premium trend:

• Full-coverage average rose 46% from 2022 to 2024 (The Zebra / Insurify)
• 2025: national average dropped 6% to $2,144; 39 states saw reductions

Industry profitability:

• Combined ratio (overall U.S. P&C; NAIC/industry sources): 2022 approximately 100.7 (unprofitable), 2023 = 101.8 (unprofitable), 2024 = 96.6 (profitable, best in a decade). AM Best (via Enterprise Community Partners, Feb 2026) reports the same period at 103.1 / 103.7E / 100.7P - both series show the same directional arc; the gap likely reflects different line-of-business cuts or publication vintage
• Ex-catastrophe combined ratio 2024: 88% - best in at least 20 years
• Net profit 2024: $169 billion (+90% year-over-year, +333% from 2022) - see profit tension discussion
• Return on equity: dropped to 4.8% in 2022, recovered to approximately 10% in H1 2024

Reinsurance:

• Property catastrophe reinsurance rates rose 45% to 100% in the US in 2023 alone (Beinsure / industry data)
• Reinsurers compelled primary insurers to accept higher retentions - primary insurers absorbed more loss per event before reinsurance paid
• Reinsurance market moderated in January 2024; hard market of 2022-2023 eased

California market detail:

• State Farm halted all new homeowners applications May 2023; began nonrenewing 30,000+ residential homes starting July 2024 and 42,000 commercial apartment properties starting August 2024
• Allstate paused new home insurance sales 2022
• Five additional carriers stopped writing new policies between summer 2023 and early 2024
• California FAIR Plan policy count increased 276% from 2018 to 2024; premiums jumped from $87.2 million (2018) to $1.4 billion (2024) - approximately 16x
• FAIR Plan exposure reached $649.4 billion by June 2025, up 42% since September 2024; policies in force reached 610,179, up 31% over the same period (NAIC/CIPR)
• Average insurance quotes available per person fell to 1.07 in June 2024, a 27% decrease from June 2023

Housing affordability spillover:

• 68% of homeowners with escrow accounts saw mortgage payments rise over the past two years due to higher taxes and insurance (annual escrow survey - source organization not identified in search results; confirmed directionally across multiple sources)
• 55% said they were surprised by the change - source organization not identified; label: plausible, source unknown
• In Pensacola, FL, property taxes and homeowners insurance together account for 44% of the average monthly mortgage payment
• Cotality Housing Affordability Index: insurance premiums, property taxes, and PMI now exceed 40% of buyer monthly housing obligation in many markets

Florida litigation detail:

• AOB lawsuits: 900% increase from 2008 to 2018; approximately 135,000 lawsuits filed by end of 2018
• Florida: 7% of US homeowners insurance claims, 76% of US homeowners insurance lawsuits (III / Florida OIR)
• Post-reform litigation: dropped from approximately 8,000/month (early 2023) to approximately 4,000/month (November 2024)
• Frivolous lawsuits fell 25% H1 2025 vs H1 2024

Louisiana (stress case): Louisiana ranks 50th in auto insurance affordability at approximately $236/month for full coverage. Drivers pay approximately $1,334/year above the national average. The primary structural driver was the “direct action statute,” allowing plaintiffs to sue insurers directly - unique among states, creating settlement leverage. Legal settlements average 23% higher than in comparable tort states (label: plausible - primary source not independently reviewed). The 2024 reform revised the direct action statute and extended the tort prescription period to two years. Premium trajectory is too recent to evaluate reform impact. The Louisiana pattern confirms the litigation-mechanism hypothesis from Florida but is an earlier-stage test.

Flood insurance gap:

• $24.4 billion in estimated future annual flood losses on single-family homes; more than $17 billion currently uninsured (Philadelphia Fed Working Paper 24-23)
• 80% of households at risk of flood damage lack adequate coverage
• NFIP owes approximately $22-25 billion to the US Treasury; since 2005, NFIP losses have exceeded premiums by $36 billion
• NFIP expired September 25, 2025; pending reauthorization as of early 2026
• GAO: median NFIP premium was $689 but would need to rise to $1,288 to reach full-risk pricing under Risk Rating 2.0 - a direct parallel to the California pattern: suppressing the price signal does not reduce the pressure, it just defers the reckoning

Research gaps

• [RESEARCH GAP: Clean decomposition of any specific premium increase into risk vs. system cost vs. margin components. No study has produced this at the consumer level for property insurance. This gap limits how precisely policy can target the right layer.]
• [RESEARCH GAP: Exact household count of uninsured US homes. Current confirmed range: 7-14%.]
• [RESEARCH GAP: National HHI data for homeowners insurance concentration by state over time. NAIC Competition Database Report (2022) exists but aggregate figures were not surfaced.]
• [RESEARCH GAP: Whether Michigan’s aggregate premium savings distributed equitably to low-income Detroit drivers specifically.]
• [RESEARCH GAP: Whether California’s Sustainable Insurance Strategy (2024) has materially reversed insurer exits. Implementation is too recent for confirmed outcome data as of early 2026.]
• [RESEARCH GAP: Long-run premium stability in Florida post-reform. Market recovery is confirmed for 2022-2024; durability through a major hurricane event is structurally unquantified.]

Back to case studies