Blog

Rate caps, rollbacks and freezes: how states limit increases

What a state can actually order on car insurance prices, from California's 1988 rollback to Michigan's mandated cuts, and what it means for the bill you pay.

States do limit car insurance prices, but almost never with a flat ceiling on what you pay. The usual tools are a rollback (cut rates to a past level), a mandated reduction tied to a law change, an approval requirement that slows increases, a refund rule that claws back profit above a set line, and a hard backstop: a rate so low that it threatens an insurer's survival is not allowed. Whether any of these lowers your own bill depends on the rating factors in your policy, which stay legal even when the average is capped.

Because the cases below come from statutes and court opinions that this article opened, they are a sample, not a complete history of every freeze ever ordered. Where the evidence on effects is thin, the article says so.

Every cap runs into the same constitutional wall. A state can say that an insurer's prices must come down, but it cannot force the company to sell below what is needed to pay claims and costs and earn a fair return, because that would take its property without compensation.

California's Supreme Court worked through exactly this in 1989. In Calfarm Insurance Co. v. Deukmejian, 48 Cal.3d 805, it reviewed Proposition 103, the voter initiative passed in 1988. The opinion describes the initiative's rollback provision: for the rollback year, rates could be no higher than 80 percent of the rate in effect on November 8, 1987. The court held that the rollback was not invalid on its face, but only because insurers had a way to show that a particular rate was, as applied to them, confiscatory. It also held that a narrow exception for insurers "substantially threatened with insolvency" could not stand by itself.

The court came back to the same ideas in 20th Century Insurance Co. v. Garamendi, 8 Cal.4th 216 (1994). Quoting its own earlier reasoning, it said a confiscatory rate is necessarily an inadequate rate, so the general standard for rates (excessive, inadequate or unfairly discriminatory) already forbids confiscatory ones. It reversed the lower court and directed judgment for the Insurance Commissioner on the rollback order against the insurers who had challenged it.

Two lessons carry to every state. First, a cap with no pressure valve is likely to fail in court. Second, the valve itself becomes the battleground: who may use it, how much cost evidence it takes, and how fast the regulator decides. The companion article on the legal rate test covers the three prongs.

Rollbacks: California, 1988

A rollback pushes rates back to an earlier level. Proposition 103 is the best-documented American example. According to the two court opinions, it set a maximum rate for the rollback year at 80 percent of the 1987 rate and then moved the state into permanent prior approval starting November 8, 1989, under which rates must be approved by the commissioner before use. The opinions also describe the permanent standard: no rate may be approved or remain in effect that is excessive, inadequate or unfairly discriminatory, with no consideration given to the degree of competition, and the commissioner must consider whether the rate mathematically reflects the insurer's investment income.

What a driver saw depended on the rollback process, which the opinions say ran through application to the commissioner and hearings, with an insurer able to ask for relief if the rollback would be confiscatory as applied to it. The more durable change was the permanent regime. The Proposition 103 guide and the article on California's good driver discount: the Prop 103 rule and who qualifies explain what stayed.

Mandated reductions: Michigan, 2019

Michigan's 2019 no-fault reform tied lower prices to a law change instead of a flat rollback. Michigan Compiled Laws 500.2111f requires insurers to file personal protection insurance (PIP) premium rates for policies effective after July 1, 2020, that produce, as nearly as practicable, an average reduction per vehicle from the rates in effect on May 1, 2019, and the requirement continues for policies effective before July 2, 2028. The size of the required average reduction depends on the PIP limit the buyer chooses:

PIP choice under the statute Required average reduction
Limit under section 3107c(1)(a) 45% or greater
Limit under section 3107c(1)(b) 35% or greater
Limit under section 3107c(1)(c) 20% or greater
No limit under section 3107c(1)(d) 10% or greater

Three features show how real caps are built.

  • The regulator checks the filing. The statute directs the director to review each filing and disapprove one that does not produce the required reductions; an insurer then has 15 days to refile.
  • There is a pressure valve. An insurer can apply for a lower reduction or an exemption, and the director must approve if complying would push it to the company action level of risk-based capital, a regulatory solvency measure. Two constitutional-violation exceptions are in the statute but, according to its text, stop applying after July 1, 2023. The idea behind the solvency test is covered in risk-based capital: how regulators spot a failing insurer early.
  • Individual increases stay legal. The statute says it does not prohibit an increase for any individual policy if the increase results from applying rating factors approved under the chapter. The cut applies to the average, not to every household.

The statute also requires insurers to pass on savings from the medical fee schedule created by the same reform, and calls for compliance filings for policies issued or renewed in the years starting July 1, 2024 and July 1, 2026. The wider story is in michigan's no-fault reform explained.

Approval requirements as a brake: New York, 2026

A softer cap is a rule that makes increases slower or harder to get. New York's fiscal 2027 budget moved in that direction. According to Insurance Journal's September 2026 report, the Department of Financial Services proposed a regulation requiring auto insurers to obtain express prior approval before implementing any upward rate change. Under the existing flex-rating system, the report says, an insurer could implement no more than two overall average private passenger auto increases without department approval if the cumulative effect stayed within 5 percent. The proposal also requires insurers to explain rate changes to policyholders, including any decrease resulting from the reforms.

The Governor's announcement describes the budget as setting a legal threshold that prevents excess profits and returns savings to consumers, adding approval safeguards, and barring rates based on homeownership status, occupation, education level or ZIP code. As of November 2026 these are reports of a proposed regulation and a press release, not final text, and the effect on prices is unproven. Read the final rule on the department's site before relying on any detail.

How do flex bands work in a state that has them? The flex-band article explains, and a 2008 academic note by Jason Strauss describes the design: a threshold lets an insurer move rates within a band without prior approval and requires approval outside it. His note lists example thresholds, such as 5 percent in North Dakota, and discusses a design flaw in which a one-time rate cut can permanently lower the level an insurer can reach by simplified filing. That is an academic argument about incentives, not proof of how any state's market behaved.

Excess-profits rules: Florida

Another cap works backward. Instead of limiting the price going in, it limits the profit coming out. Florida Statute 627.066, titled "Excessive profits for motor vehicle insurance prohibited," tests each private passenger auto insurer group's three most recent calendar-accident years combined. Excessive profit has been realized if the underwriting gain is greater than the anticipated underwriting profit plus 5 percent of earned premiums for those years. The anticipated profit is built from the profit and contingencies factors in the insurer's own filed rates.

If the test is failed, the state's regulator must order a return of the excess after a hearing, unless the insurer shows the refund would make a group member financially impaired or insolvent. The statute says the money goes to voluntary private passenger auto policyholders of record on December 31 of the final year, as a cash refund or a credit toward future insurance, pro rata to earned premium. A cash refund has to be completed within 60 days of a final order; credits go on renewals sent more than 60 days after it.

Florida has a second claw-back. Under Statute 627.062, if an insurer implements a "use and file" rate and the regulator finds it excessive, the regulator must order the portion above what was actuarially justified returned as a credit or refund. The same statute says a rate given final approval, or deemed approved, cannot be disapproved as excessive for one year after the effective date unless the insurer made a material misrepresentation or error.

New York's governor's release cites a Florida regulator analysis and says Florida's largest carrier returned nearly $1 billion in excess profits to 2.7 million policyholders in 2025. That is the governor's statement of another state's result; we did not open Florida's analysis, so treat it as an unverified claim.

Emergency and refund orders: the pandemic example

A freeze can also be temporary and event-driven. In April 2020 the California Insurance Commissioner ordered insurers to return premiums for private passenger auto and other lines. According to the department's press release, the order covered premiums for at least March and April, and May if shelter-in-place restrictions continued, and required a credit, reduction or return of premium as soon as possible and no later than August 2020. A refund using the methods in the commissioner's bulletin did not need prior approval. Companies reported back within 60 days.

In October 2021, Insurance Journal reported that the commissioner directed three insurers (Allstate Northbrook Indemnity, Mercury and CSAA Insurance Exchange) to reimburse excess premiums or face legal action. The department's analysis, it said, found insurer groups returned on average 9 percent of auto premiums from March to September 2020, against nearly double that, 17 percent, that it said they should have refunded. The longer story is in premium refunds: when regulators order money back to drivers.

Texas has a standing mechanism for crises. Under Insurance Code 2251.151, the commissioner may by order require an insurer to file rates for approval when its financial condition or rating practices call for supervision, or on the finding of a statewide insurance emergency, and the insurer has a right to a hearing on request.

Do caps work? What the evidence supports

The honest answer is: it depends on what you measure, and this article opened little of the empirical literature.

  • Mechanism. A cap works as a lower price only while it binds. Rates below cost are "inadequate" in the statutes, so regulators cannot hold them there indefinitely. Strauss's note says regulators who suppress rates generally recognize that rates must increase with costs for insurers to stay in the market.
  • Availability. A well-known worry is that insurers write fewer policies where the cap bites. The court opinions treat that risk as the reason a pressure valve is required, and economists' work on regulated markets, such as Suponcic and Tennyson's chapter on rate regulation and the industrial organization of auto insurance, analyzes it. We could read only that chapter's abstract line, which says it analyzes the impact of rate regulation on the structure of private passenger insurance markets; we cannot report its findings.
  • Comparisons across states. Prior approval and file-and-use states differ in price effects in studies surveyed in prior approval vs file-and-use: do regulated states pay more or less?. Treat any single before-and-after claim, including press-release claims, with caution, because loss costs, tort law and the repair market move at the same time as regulation.

What drivers see on bills

  • An average cut is not a personal cut. Michigan's statute allows individual increases from approved rating factors. A filing can lower the average while your rate rises.
  • A refund can arrive as a credit. Florida's excess-profit refund can be a renewal credit, and the California pandemic order allowed credits and reductions. The bill might show a lower renewal premium instead of a check.
  • An approval delay is not a price cut. A longer review slows the timing; it does not change what a regulator ultimately finds justified.
  • Insurers may stop selling. When a regulator will not approve a rate the company thinks it needs, the company can narrow or stop writing, which pushes drivers toward the residual market and assigned risk plan in the worst cases.

How to shop when rates are capped

  1. Assume prices still differ. A cap limits the average increase at each company, not the gap between companies. Get at least three quotes on identical coverage.
  2. Look up the filing. The filing lookup routine shows whether your insurer's recent change was capped, withdrawn or approved.
  3. Check your own factors. If your increase beats the filed average, ask which factor moved: record, mileage, vehicle, territory or a lapsed discount.
  4. Ask about refunds and credits. In states with excess-profit rules, ask whether a refund or renewal credit is pending, and whether it follows you if you cancel (in Florida's statute, a cancelled policy gets the cash refund).
  5. Watch availability. If an insurer is not accepting new business, a lower filed average will not help you; call your state department. See what your state insurance department does.
  6. Date your notes. Caps change with legislatures. Write the date next to every rule you rely on.

Common questions

Can a state freeze my premium?

Rarely in the sense of a flat freeze on your own bill. States more often limit average rate changes or require approval, and your individual premium can still change when your driving record, vehicle or other rating factors change.

Why not just cap every increase at a fixed percent?

A fixed cap below the actual cost trend produces rates that are inadequate under the statutes and that courts may treat as confiscatory. California's opinions explain why a relief mechanism is required.

Do caps lower rates for everyone?

No evidence in the sources opened here shows a uniform effect. Michigan's statute sets average targets by coverage choice, and New York's changes are too recent to judge.

Where do I check if my state is doing this?

Start with your state's insurance department and the filing pages described in the rate approval guide.

Key takeaways

  • States limit prices through rollbacks, mandated reductions, approval requirements and excess-profit refunds, not through a flat ceiling on your bill.
  • Every cap needs a valve for rates that would be confiscatory; California's court opinions explain why, and Michigan's statute includes one.
  • Averages are not individual outcomes: approved rating factors can still raise your own premium.
  • Evidence on effects is mixed or thin, and press-release claims should be checked against filings and statutes.
  • When rates are capped, shop anyway: price gaps between insurers remain, and your own factors matter most.

Sources

Terms used in this guide

Related states

Related guides

Related tools

Latest on this topic