Who goes uninsured and why: cost, lapses and enforcement gaps

Roughly one driver in seven was uninsured in 2023, and no single cause explains it. Price against income is the best-documented driver, but the states with the least affordable insurance are not always the states with the most uninsured drivers, which points to the other forces: short gaps between policies, vehicles that are barely driven, and how hard a state works to find and penalize people without coverage. The evidence on remedies is thinner than the headlines suggest, because most of it comes from the agencies that run the programs.

This article takes the forces one at a time, starting with how the number itself is built, and ends with what a driver can do when cost is the barrier. For the state-by-state counts, see our guide to uninsured drivers by state.

How the uninsured rate is measured, and why it is an estimate

No agency counts every uninsured driver. The Insurance Research Council (IRC) estimates the rate from claims: it compares how often people injured by uninsured drivers file claims against how often people injured by insured drivers do. The Insurance Information Institute (Triple-I) states the method in its summary of the IRC's 2025 study: the rate is "measured by the ratio of uninsured motorists (UM) claims to bodily injury (BI) claim frequencies" (Triple-I, Uninsured motorists).

Three consequences follow.

• It measures drivers who cause injury crashes, not every registered car. A state that counts uninsured registrations will report a different figure.
• State rules change the meaning. Triple-I notes that in Florida, compulsory laws apply to personal injury protection and property damage but not to third-party bodily injury, so Florida's 20.6% in 2023 is not the same kind of number as a state that requires liability coverage.
• State and IRC figures often disagree. Arkansas's Department of Finance and Administration reported its own rate falling from 16% in early 2020 to 8.7% by October 2023 (Route Fifty, republishing the Arkansas Advocate), while the IRC's 2023 estimate for Arkansas was 12.1%. They count different things, so neither is wrong, and you should never compare one source's number for one state with another source's number for a second.

The national trend

The IRC's series, as reported by Triple-I, ran 12.4% in 2017, 12.1% in 2018, 11.6% in 2019, then jumped to 14.3% in 2020 and reached 15.4% in 2023. In 2023, 15 states and the District of Columbia were above the countrywide rate and 36 were below. The spread is wide: Mississippi at 28.2%, New Mexico at 24.1% and Michigan at 22.3%, against Maine at 5.7%, Utah at 6.2% and Idaho at 6.4%.

The IRC's own explanation, in its February 2025 release, is economic. UM rates rose in nearly every state when the pandemic hit household incomes in 2020, and the council said it had expected them to ease as employment recovered but that they kept rising in most states, which it attributed to deteriorating insurance affordability outweighing the economic recovery (IRC release, February 2025). That is an interpretation by an insurer-funded research group, not a proven cause, but it matches the direction of the data.

Cost: price measured against income

The IRC measures affordability as average auto insurance spending divided by median household income. For 2018, the latest year in its public summary, the national figure was 1.67% of income ($1,057 average expenditure against a $63,179 median income), down from a peak of 2.03% in 1993. Iowa was the most affordable state at 1.02% and Louisiana the least at 3.09% (IRC, Auto Insurance Affordability).

An average hides the households that matter here. Say a household earns $32,000 and pays $2,400 a year to insure two older cars. That is 7.5% of its income, nearly five times the national ratio, and the cost arrives as a lump sum or a monthly bill that competes with rent and food. The IRC says plainly that its index looks at the overall population and does not address lower-income or underserved households, which need more granular data.

Two details from the IRC page show why price alone is a poor predictor:

• Income can matter more than price. Mississippi's average spending was 8% below the countrywide average in 2018, yet its median income is the lowest in the nation, so it ranked fourth least affordable. Mississippi also had the highest 2023 uninsured rate in the IRC series.
• The least affordable state is not the most uninsured. Louisiana's affordability ratio was by far the worst, but its 2023 uninsured rate was 11.7%, below the national 15.4%. (The affordability data are from 2018 and the uninsured data from 2023, so treat the comparison as a hint, not a measurement.) Whatever Louisiana is doing, price is not the whole story, which is why the sections below on gaps and enforcement matter.

Gaps, lapses and part-time vehicles

A driver can be uninsured without ever deciding to be. Several routes lead there.

Short gaps between policies. A policy ends for non-payment or is cancelled, and a new one starts days later. Regulators treat even a short gap as a lapse. New York's DMV defines an insurance lapse as any period a registered vehicle has no liability coverage, including the time between the date insurance is cancelled and the date new insurance begins (NY DMV, Insurance Lapses). Nevada says it has no grace period and that a one-day lapse may bring suspension and reinstatement fines (Nevada DMV). Pennsylvania is more forgiving for a short gap, with an exception to the registration suspension when the lapse was 30 days or fewer and the owner proves the vehicle was not operated (PennDOT). Our guide to what happens if your car insurance lapses explains what a gap does to prices later.

Non-payment and cancellation. A missed payment can turn a policy into a lapse, and a cancelled driver may be offered only more expensive coverage. See cancelled for nonpayment: reinstate the policy or start over?.

Vehicles that are rarely driven. A 1998 California Department of Insurance study, which matched insurer records against the DMV vehicle file, estimated 22.6% of vehicles uninsured in mid-1997 and found two distinct groups. "Pure" uninsured owners had one or two vehicles, none insured, and tended to be younger, lower income and drive daily to work or school; they tended to give cost as the reason. "Hybrid" uninsured owners had some insured vehicles and some not, tended to be older and better off, and tended to cite a vehicle-related reason such as the car not running or being used occasionally (California Department of Insurance, 1998). The study is old, from one state, and its authors said their sample probably under-represented the pure uninsured, so use it as a map of the types of non-compliance rather than as current numbers. The report also found that close to 90% of the pure uninsured had one or no tickets or accidents, and noted that with three years of driving experience they would qualify for California's good driver discount.

Cars that do not count. The same study noted that no law required liability coverage on a vehicle that does not run or is operated only off-road, so some "uninsured" vehicles are legally uninsured. Florida's DMV tells owners to surrender the plate before cancelling insurance for a car that is not being driven (FLHSMV), which is how an owner stays on the right side of the line.

Enforcement: the gap between the law and the likelihood of being caught

Almost every state requires coverage. What differs is the chance a driver without it is noticed, and what happens when they are.

How states find them. The traditional checks happen at registration, during a traffic stop or after a crash. Many states now run electronic verification in which insurers report policies to the motor vehicle agency and the agency matches them to registrations. Triple-I says more than half the states have passed laws and begun to develop online verification systems. Florida's statute requires insurers to report cancellations and nonrenewals to the state within 10 days, and the state can then suspend registration and license (Fla. Stat. 324.0221). Arkansas moved in 2020 from checking only at registration, stops and crashes to an instant verification system; the state's finance department says a letter and a $100 reinstatement fee follow a detected lapse, and the American Property Casualty Insurance Association counted Arkansas as the 17th state with an instant system. For how matching works and its flaws, see state insurance verification systems. A driver who gets a notice should read after an insurance lapse: notices, suspension and reinstatement.

What the penalties look like. They vary a great deal, and it is not obvious which design works.

The table does not prove that harsher penalties cut uninsured driving. The states differ in income, cost, population density and how coverage is verified, and Florida's figure reflects its different mandate. What it does show is that the money penalty for a first lapse ranges from a few hundred dollars to over a thousand, and that in these states suspension of registration comes with it.

What a penalty does to a low-income driver. A fee that is small to one household can be unaffordable to another, and a suspended registration on a car needed for work can tempt a driver to keep driving, which adds the risk of further penalties. We found no study measuring how often that happens.

Rates inside a state

Averages hide big differences within a state. The IRC affordability page cites an analysis of NAIC data showing that higher premiums in lower-income ZIP codes were in line with higher claim costs in those areas, and says that efforts to improve affordability there must address those costs. That is the industry's explanation: pricing follows local claim frequency and severity. States regulate rating differently. Our overview of how car insurance rates are calculated explains territory as a rating factor, and Uninsured drivers and your premium: how UM costs are priced into rates explains the feedback loop in which uninsured drivers raise the price everyone else pays for uninsured motorist coverage.

What states have tried, and what the evidence shows

A few notes on the evidence.

Verification. The reported drops are plausible, since a driver who knows a lapse will be detected has a reason to avoid one. But the figures come from the same agencies that run the systems, and agency counts of registered vehicles are not comparable across states or with IRC estimates. The Arkansas Advocate reported that the state could not say how many fines had been rescinded since the system began.

California's low-cost program. The CLCA requires a valid California license, a vehicle worth $25,000 or less, a good driving record and household income under a limit that, in the department's 2026 report, was $39,900 for one person and $54,100 for two. Annual liability rates effective May 1, 2025 vary by county; for example, $199 in Butte County and $418 in Los Angeles County. The department's report also said that California's January 2025 increase in minimum limits to 30/60/15 was raising premiums statewide and sending more drivers to look for low-cost options. A program of about 67,000 policies is small next to a state whose IRC uninsured rate was 20.4% in 2023, and whether a larger low-cost market would cut uninsured driving is a policy question the data cannot answer.

Limited-benefit policies. New Jersey's product is medical-only, so it does not provide the third-party liability coverage that protects other drivers. It targets a narrow group.

No pay, no play. These laws punish being uninsured after a crash; they do not make coverage cheaper. They change who can recover damages, not what coverage costs. See no pay, no play laws.

If cost is the barrier: what a driver can do

1. Check what your state requires. Use state car insurance requirements lookup. Some states require only liability; the minimum differs widely.
2. Ask about state programs. California runs CLCA through the state; other states have assigned-risk plans for drivers insurers decline (assigned risk plan). Your state insurance department can say what exists.
3. Compare several quotes at the minimum limits and price annual payment against monthly installments.
4. Surrender plates before dropping coverage on a car you are not driving, where your state allows it, to avoid a lapse penalty.
5. Keep proof of coverage in the car and update the policy when you move or change vehicles.
6. If a notice arrives, respond by its deadline even if you think it is an error.

Common questions

Is the 15.4% figure the share of cars with no insurance?

No. It is an IRC estimate based on injury-claim ratios for 2023, not a count of registrations, and state agency figures use other methods.

Does a short gap in coverage count as being uninsured?

For legal purposes, often yes. New York treats any period without liability coverage on a registered vehicle as a lapse, and Nevada has no grace period. A few states, such as Pennsylvania, provide an exception for brief lapses when the vehicle was not driven. A gap can also raise your next quote, which our guide on what happens if your car insurance lapses covers.

Are uninsured drivers mostly drivers with bad records?

The California study found otherwise for the lowest-income group: nearly 90% of the pure uninsured had one or no tickets or accidents. That is from 1997 data, so do not treat it as current.

Do higher fines reduce uninsured driving?

The evidence we could verify does not say. In the five states in the table above, the 2023 rates do not line up neatly with penalty size, and the states also differ in income, prices and how coverage is verified.

What happens to me if an uninsured driver hits me?

You rely on your own uninsured motorist coverage if you have it, or on collision and medical coverage. See uninsured motorist coverage.

Key takeaways

• The 15.4% uninsured rate for 2023 is an IRC estimate built from injury-claim ratios, not a count of cars, and it rose from 11.6% in 2019.
• Price against income is the best-documented factor, but the least affordable state (Louisiana in 2018) did not have the highest uninsured rate, so cost is not the whole story.
• Lapses, cancellations and rarely driven vehicles account for coverage gaps that look like deliberate non-compliance; the old California study found that the cost-driven and the part-time-vehicle groups differ.
• Enforcement matters, but the public evidence on verification, penalties and low-cost programs is mostly self-reported by the agencies that run them.
• If cost is the barrier, check state programs and assigned-risk plans, and avoid a lapse penalty by surrendering plates on cars you are not driving.