Average expenditure vs average premium vs quoted rate

The averages disagree because they measure three different things. Average expenditure is what insurers collected per insured vehicle across everyone, average premium is the typical price of one coverage or of a full package, and a quoted rate is what an insurer offers one imagined driver on one day. For the 2023 data year, the National Association of Insurance Commissioners (NAIC) reports an average expenditure of $1,281.92 and a combined average premium of $1,438.60, while quote panels for a clean-record driver with full coverage land well above both.

Which number applies to you depends on what you want to know. If you want to see how a state's spending compares, use expenditure. If you want a benchmark for a policy like yours, use the average premium for the coverages you carry. If you want to know what you will pay, only a quote answers that, and even a quote can change before you buy.

The three measures side by side

The first two are historical and cover everyone who bought insurance, including the cheapest and most expensive drivers and the shoppers and non-shoppers. The third is forward-looking and covers only the imagined driver. None is wrong. Each answers a different question, and people often borrow the wrong one when they quote a figure.

How the NAIC builds average expenditure

The NAIC's Auto Insurance Database Report is the standard source for state averages. The edition covering data years 2019 through 2023 defines the measure in its own technical notes.

An exposure, in the report, is a car-year: the risk of insuring one car for one year. A policy covering two cars for six months counts as one car-year. Average expenditure, per the NAIC's formula, is the total written premium for liability, collision and comprehensive combined, divided by liability written exposures. The report states the assumption plainly: every insured vehicle is assumed to carry liability, but not every vehicle carries the physical-damage coverages. See the NAIC 2022/2023 Auto Insurance Database Report.

You can check the arithmetic from the report's own countrywide rows for 2023:

• Liability written premiums: about $172.9 billion
• Collision written premiums: about $83.3 billion
• Comprehensive written premiums: about $44.7 billion
• Liability written exposures: about 234.7 million car-years

The three premium totals add to about $300.9 billion. Divided by 234.7 million car-years, that gives $1,281.92 per insured vehicle, the countrywide average expenditure the NAIC publishes. The Insurance Information Institute repeats the figure in its auto insurance facts and statistics page, rounding it to $1,282 and noting that it rose 14.0 percent from the prior year.

Two details matter for anyone reading the number.

Premium-weighted, not person-weighted. The result is a pooled average. A driver who bought only the state minimum and a driver who bought maximum limits with low deductibles both go into the same pot. The NAIC says its calculation makes no distinction among policyholder classifications, vehicle characteristics, or the limits and deductibles chosen, "all of which significantly impact the cost of coverage."

The report is narrower than "all auto insurance." It covers private passenger auto, and it leaves out vehicle types such as motorcycles, motor homes and golf carts. Residual-market business (the plans for drivers insurers will not otherwise write) is included in the combined figures.

How average premium and combined average premium differ

An average premium is computed one coverage at a time. Liability average premium is liability written premium divided by liability exposures. Collision average premium is collision premium divided by collision exposures, and comprehensive works the same way. The NAIC's countrywide 2023 figures are:

The combined average premium therefore describes a policy that carries all three coverages. The NAIC's own description: it is the average cost of a policy that contains liability, comprehensive and collision. The median state's combined average premium was $1,268 for 2023, and the median state's average expenditure was $1,114, both from the report's narrative.

One wrinkle for the liability figure: the NAIC's liability premium includes more than bodily injury and property damage liability. Its technical notes list uninsured and underinsured motorist coverages, medical payments, and no-fault personal injury protection among the coverages folded into liability premiums. A state that requires no-fault or uninsured motorist coverage will show a larger "liability" average than a state that does not, even if the actual liability cost is the same.

Why expenditure and combined premium diverge

The gap between $1,281.92 and $1,438.60 is not a statistical accident. It comes from the share of vehicles that carry physical-damage coverage.

Expenditure divides every dollar by every vehicle with liability. Combined average premium assumes the buyer has all three coverages. In the NAIC's 2023 countrywide rows, collision exposures were about 179.6 million and comprehensive exposures about 187.6 million, against about 234.7 million liability exposures. In round terms, roughly three of every four insured vehicles carried collision, and about four in five carried comprehensive. Vehicles with only liability pull the expenditure average down while leaving the combined premium untouched.

That makes the two measures move apart fastest in states where fewer people buy collision and comprehensive. The report's own numbers show it.

Reading the NAIC tables for the 50 states and D.C., Texas ranks 13th on average expenditure but 6th on combined average premium. The gap is nearly $300 per vehicle. The reason is not that Texas drivers are cheaper to insure in some mysterious way; it is that, according to the exposure counts, roughly a third of insured vehicles in the state carry no collision and a similar share carry no comprehensive, and those vehicles pull the pooled average down.

Compare Florida, where collision exposures were about 81 percent of liability exposures in the same data and where the two measures differ by only about $129, and where the state ranks first on both. Oklahoma shows the Texas pattern in a smaller form: $1,083.97 in average expenditure against $1,324.15 in combined premium. Neither pair of numbers is a mistake. They answer different questions about the same state.

Quoted rates: what a sample assumes

A quoted rate begins with a driver who does not exist. A quote panel fixes the driver's age, sex, credit tier, ZIP code, vehicle, mileage, record, coverage limits and deductibles, then collects what insurers would charge. Change any item and the answer moves.

LendingTree's February 2026 at-fault-accident study is a typical case. Using Quadrant Information Services rate data pulled in January 2026, it priced a 30-year-old man with good credit and a clean record, driving a 2018 Honda CR-V EX, with $50,000/$100,000 bodily injury, $50,000 property damage, uninsured/underinsured motorist coverage and $500 collision and comprehensive deductibles. For that driver, the national full-coverage average before any accident was $2,277.15. The details are in LendingTree's study.

That is nearly $840 above the NAIC's $1,438.60 combined premium. It would be wrong to say one of them is mistaken. They differ in at least five ways:

• Year. The NAIC figure describes premiums written in 2023. The quote panel is January 2026. Prices moved between those dates.
• Population. The NAIC counts everyone, including older drivers, multi-car households, drivers with discounts and drivers at the state minimum. The panel prices one 30-year-old.
• Limits. The panel uses mid-level limits and UM/UIM coverage, while the NAIC pool includes many minimum-limit policies.
• Price type. The panel prices a new application. The NAIC pool reflects what all policyholders, new and renewing, were actually charged.
• Insurer mix. The NAIC totals weight each insurer by the premium it wrote, while a panel's insurer coverage depends on its data vendor.

Insurance.com's 2026 analysis uses another sample: a 40-year-old with good credit and a clean record driving a Honda Accord LX, with a full-coverage set at 100/300/100 liability and $500 deductibles (see its state-by-state analysis). Different driver, car and limits mean a different average. The AAA "Your Driving Costs" study, as the Insurance Information Institute reports it for 2026, puts the average full-coverage cost across all vehicles at $2,098 for a driver under 65 with more than six years of experience and no accidents, living in a suburban or urban area. The III states that the coverage behind it is $100,000/$300,000 liability, $25,000 medical, $100,000 property, $25,000/$50,000 uninsured/underinsured motorist and $500 deductibles, and that these figures are not comparable with the NAIC expenditure data.

What coverage mix, limits and deductibles do to an average

Coverage choices change the price more than most people expect, which is why a state's average cannot be treated as the cost of "car insurance" in general.

• Liability limits. Higher limits cost more, though not in proportion. See why doubling liability limits does not double the premium for how a doubling of limits does not double the premium. Where state minimums are higher, or buyers choose higher limits, the liability average tends to be higher (the liability limits decoder explains the notation).
• Physical damage. Collision and comprehensive together made up about $702 of the NAIC's $1,438.60 combined premium in 2023, nearly half. A liability-only driver pays none of that.
• Deductibles. Raising a collision deductible lowers that part of the premium. A panel at a $500 deductible and a pool that includes $1,000 deductibles will not match.
• Add-ons. Rental reimbursement, roadside coverage and gap coverage are often absent from published averages but present on real policies.
• Required coverages. No-fault and uninsured motorist requirements go into the NAIC's liability line, as noted above.

Here is a made-up example to show the effect. It is hypothetical and not market data.

Say two neighbors insure similar sedans in the same ZIP code. Neighbor A buys liability only at the state minimum and pays $900 a year. Neighbor B buys 100/300/100 liability, collision and comprehensive at $500 deductibles and pays $2,300. An insurer that writes ten policies like A's and ten like B's collects $32,000 over those twenty cars, or $1,600 per car, and that $1,600 describes neither neighbor. A published state average is the same kind of blend, spread over millions of cars.

Why the same state can rank differently in different sources

A state's position in a ranking depends on the measure, the year and the edition of the data.

Different measure. As the Texas example shows, ranking by expenditure and by combined premium can reorder states.

Different year or edition. Insurers keep revising the data they submit, and the NAIC says so. Its report states that historical averages might not match those published in prior reports, because data for earlier years was adjusted to reflect the most current information. The III's page shows the effect. Its text cites the NAIC's 2023 figure of $1,282, up 14.0 percent from $1,127 in 2022, while the NAIC's current-edition table lists $1,124.57 for 2022. The same page also contains a table labeled 2023 that ranks Louisiana first at $1,494.70, while its text and the NAIC's 2023 table put Florida first at $1,864.63. As I read the page in November 2026, it does not reconcile these figures, and the table appears to carry data from an older edition. This is a reason to cite the original report and its edition.

Different coverage mix. A quote panel run at one set of limits will rank a no-fault state differently from a pooled average that includes lots of minimum-limit policies.

Different geography. State averages hide huge ZIP-code differences. The NAIC itself cautions that direct comparisons between states should be treated with a high degree of caution because of differences in tort laws, rate filing laws, traffic conditions and demographics. See prior approval vs file-and-use: do regulated states pay more or less? for one of these structural differences.

Turning a published average into a personal expectation

An average is a starting point, not a prediction. A reasonable process looks like this.

1. Pick the measure that matches your purchase. Liability only: start from the state's liability average premium. All three coverages: start from the combined average premium. Only "how much do people here spend": expenditure.
2. Check the year. Add the later change in prices if you are comparing a 2023 figure against 2026 quotes. The The CPI for motor vehicle insurance: how price change is measured article explains one official measure of price change.
3. Compare your limits and deductibles to the typical policy. If yours are higher than minimum, expect to pay above the average.
4. List the rating factors that move you. Your age, record, credit-based score where allowed, ZIP code, vehicle, mileage and prior insurance history all matter (see how car insurance rates are calculated).
5. Collect three or more real quotes with identical coverage using the car insurance quote comparison worksheet, and treat the published average as a sanity check on them.
6. Look for a reason when a quote is far off. A quote 40 percent above the relevant average has an explanation: a claim, a young driver, a high-theft car, a coverage mismatch or a mistake in the application.

A worked example, again hypothetical. Say the combined average premium in your state is $1,500 for a policy with all three coverages. You want 100/300/100 liability with $500 deductibles on a six-year-old car, and your three quotes arrive at $1,620, $1,780 and $2,350. The first two sit about 8 percent and 19 percent above the average, which is what higher-than-typical limits might explain. The third is 57 percent above and needs a look: ask what is different in that quote, such as a surcharge, a rating tier, missing discounts or a different coverage set. Do not conclude that $1,620 is a bargain either; the average itself blends many unlike policies.

How to cite an average honestly

A citation that gets the measure, the year and the source right is more useful than a headline number.

• Name the measure: "average expenditure per insured vehicle," not "the average car insurance bill."
• Give the data year and the publication: "NAIC, 2022/2023 Auto Insurance Database Report, 2023 data."
• Say what is excluded or assumed: "pooled across all coverage choices; liability assumed on every vehicle."
• For a quote, state the sample: age, vehicle, limits, deductibles, date and data vendor.
• Don't compare a 2023 NAIC figure with a 2026 quote and call the difference "inflation." Differences in year, population and coverage all contribute.
• Prefer the primary source. When a secondary page repeats a figure, check it against the original, since editions get revised. See Insurance Information Institute statistics and how to use them and fact-checking car insurance headlines about percent increases.

The how much one at-fault accident raises your rate in different states article applies the same discipline to quote-study percentages, and What households spend on car insurance: Consumer Expenditure Survey covers a different kind of "expenditure," what households report spending in a government survey.

Common questions

Which average should I use to estimate my own bill?

Start with the average premium for the coverages you will actually carry, then adjust for your limits, deductibles and rating factors. Average expenditure is better for comparing states than for estimating a policy, because it blends buyers who carry only liability with buyers who carry everything. Then get real quotes. A published average is only a sanity check.

Why is my quote so much higher than the average for my state?

Several ordinary reasons: the average is pooled across older, lower-risk and minimum-limit customers; your limits or deductibles may be richer; you may have a recent accident, ticket or lapse; and the average describes premiums written in an earlier year. Compare quotes with identical coverage before treating the gap as a problem. If it is large with every insurer, why an insurer declines to quote you or write a policy and the list of rating factors are the places to look.

Is a quote-panel average more accurate than the NAIC average?

Neither is more accurate in general. They measure different populations. The NAIC figures reflect what policyholders were actually charged, with lag. Quote panels reflect current prices for one imagined driver and can be fresher. Use the NAIC for statewide comparison and quote panels for a rough read on current prices for a driver like the sample.

Why do two reports on the same year show different state averages?

They may measure different things (expenditure versus combined premium), use different editions of revised data, use different vehicle types or include different coverages. Check the definitions in each report, the data year and the edition before concluding that one is wrong.

Key takeaways

• Average expenditure is total premiums for liability, collision and comprehensive divided by liability car-years; average premium is computed per coverage, and the combined version adds the three. The NAIC's 2023 countrywide values were $1,281.92 and $1,438.60.
• The two diverge where fewer vehicles carry collision and comprehensive. Texas is a clear case: 13th on expenditure, 6th on combined premium.
• A quoted rate is for a defined, imaginary driver on a specific date, so it will not match a pooled average from an earlier year.
• State rankings change with the measure, the year and the data edition; cite the original report and say which edition.
• Use averages to sanity-check, then collect real quotes at identical coverage, since only a quote reflects your own rating factors.