Car insurance as a share of income by state: NAIC premiums over Census median household income, 2024
Insuring one vehicle with liability, collision and comprehensive coverage cost 1.94% of the national median household income in 2024: the NAIC combined average premium of $1,612.81 divided by the Census median of $83,160 (1.72% on the NAIC average expenditure, which includes vehicles with liability only). By state the share runs from 1.00% in New Hampshire to 3.57% in Louisiana. Of the 51 jurisdictions, 34 sit below 2%, and Louisiana is the only one above 3%.
Bankrate’s 2025 report prints 3.39% nationally and 6.83% for Louisiana, which is 1.75 times the national share here and 1.92 times Louisiana’s. The two sets measure different things. Bankrate divides a quoted full-coverage price for one 40-year-old driver with good credit ($2,638 nationally, $3,978 in Louisiana, January 2025 rates) by 2023 income. The figures here divide NAIC’s 2024 average premium per insured vehicle, built from premium actually written, by income received in 2024. Nationally, a factor of 1.64 comes from the premium and 1.07 from the income, and the two multiply to the 1.75-times gap.
Estimates and published averages, not a quote. Not insurance advice. State rules and insurer rules vary. Confirm requirements with your state insurance regulator or a licensed agent in your state.
All 51 jurisdictions, ranked by share
Each row divides a jurisdiction’s 2024 combined average premium (NAIC Table 5, dollars per insured vehicle per year, liability plus collision plus comprehensive) by its 2024 median household income (Census Table H-8, current dollars). Rank 1 is the highest share. Ranks use unrounded values, so two rows that print the same two decimals can sit one place apart. The range column turns the Census 90% interval for the income, 1.645 standard errors either side of the median, into a range for the share. The last two columns give two further readings of the same premium data: the share when the NAIC average expenditure replaces the combined premium, and the share for a household at 60% of the jurisdiction’s median income. Shares carry two decimals because neighbouring rows differ by less than 0.1 point in places and the comparison source prints two decimals.
| Rank | Jurisdiction | Combined average premium | Median household income (standard error) | Share of income | 90% range of the share | Share using average expenditure | Share at 60% of median |
|---|---|---|---|---|---|---|---|
| 1 | Louisiana | $2,151.12 | $60,320 (1,826) | 3.57% | 3.40% to 3.75% | 3.12% | 5.94% |
| 2 | Florida | $2,174.55 | $75,260 (1,908) | 2.89% | 2.77% to 3.02% | 2.68% | 4.82% |
| 3 | Mississippi | $1,542.36 | $55,620 (1,661) | 2.77% | 2.64% to 2.92% | 2.36% | 4.62% |
| 4 | New York | $2,149.87 | $86,480 (3,352) | 2.49% | 2.34% to 2.66% | 2.29% | 4.14% |
| 5 | Georgia | $1,972.56 | $81,180 (2,077) | 2.43% | 2.33% to 2.54% | 2.17% | 4.05% |
| 6 | Texas | $1,917.05 | $81,280 (1,294) | 2.36% | 2.30% to 2.42% | 1.94% | 3.93% |
| 7 | Nevada | $1,865.83 | $80,020 (4,014) | 2.33% | 2.15% to 2.54% | 2.13% | 3.89% |
| 8 | Michigan | $1,791.22 | $78,860 (2,754) | 2.27% | 2.15% to 2.41% | 2.07% | 3.79% |
| 9 | Arkansas | $1,465.02 | $64,540 (1,913) | 2.27% | 2.16% to 2.39% | 1.82% | 3.78% |
| 10 | New Mexico | $1,421.95 | $62,660 (2,782) | 2.27% | 2.11% to 2.45% | 1.88% | 3.78% |
| 11 | Oklahoma | $1,453.74 | $64,870 (2,612) | 2.24% | 2.10% to 2.40% | 1.82% | 3.74% |
| 12 | South Carolina | $1,664.69 | $76,050 (2,850) | 2.19% | 2.06% to 2.33% | 1.97% | 3.65% |
| 13 | West Virginia | $1,367.18 | $62,800 (1,678) | 2.18% | 2.09% to 2.28% | 1.88% | 3.63% |
| 14 | Alabama | $1,388.11 | $64,840 (5,874) | 2.14% | 1.86% to 2.52% | 1.82% | 3.57% |
| 15 | Kentucky | $1,337.15 | $64,380 (3,067) | 2.08% | 1.93% to 2.25% | 1.79% | 3.46% |
| 16 | Delaware | $1,745.95 | $85,580 (3,376) | 2.04% | 1.92% to 2.18% | 1.89% | 3.40% |
| 17 | Rhode Island | $1,849.80 | $91,360 (3,969) | 2.02% | 1.89% to 2.18% | 1.83% | 3.37% |
| 18 | Arizona | $1,647.25 | $84,350 (3,354) | 1.95% | 1.83% to 2.09% | 1.74% | 3.25% |
| 19 | District of Columbia | $1,986.50 | $104,100 (4,511) | 1.91% | 1.78% to 2.05% | 1.77% | 3.18% |
| 20 | Missouri | $1,462.03 | $77,560 (3,272) | 1.89% | 1.76% to 2.03% | 1.64% | 3.14% |
| 21 | North Carolina | $1,233.32 | $66,850 (2,652) | 1.84% | 1.73% to 1.97% | 1.55% | 3.07% |
| 22 | New Jersey | $1,878.38 | $103,600 (3,074) | 1.81% | 1.73% to 1.91% | 1.69% | 3.02% |
| 23 | Tennessee | $1,353.62 | $75,460 (3,306) | 1.79% | 1.67% to 1.93% | 1.54% | 2.99% |
| 24 | Pennsylvania | $1,395.99 | $79,640 (2,491) | 1.75% | 1.67% to 1.85% | 1.59% | 2.92% |
| 25 | Colorado | $1,835.17 | $105,500 (8,254) | 1.74% | 1.54% to 2.00% | 1.52% | 2.90% |
| 26 | Connecticut | $1,690.07 | $97,790 (4,610) | 1.73% | 1.60% to 1.87% | 1.60% | 2.88% |
| 27 | California | $1,690.23 | $100,700 (1,580) | 1.68% | 1.64% to 1.72% | 1.44% | 2.80% |
| 28 | Illinois | $1,382.89 | $83,090 (4,172) | 1.66% | 1.54% to 1.81% | 1.52% | 2.77% |
| 29 | South Dakota | $1,290.69 | $78,870 (3,300) | 1.64% | 1.53% to 1.76% | 1.31% | 2.73% |
| 30 | Montana | $1,321.59 | $81,610 (4,191) | 1.62% | 1.49% to 1.77% | 1.32% | 2.70% |
| 31 | Maryland | $1,767.57 | $109,300 (4,797) | 1.62% | 1.51% to 1.74% | 1.49% | 2.70% |
| 32 | Wyoming | $1,261.71 | $78,340 (2,515) | 1.61% | 1.53% to 1.70% | 1.29% | 2.68% |
| 33 | Washington | $1,521.63 | $95,980 (5,650) | 1.59% | 1.45% to 1.76% | 1.48% | 2.64% |
| 34 | Nebraska | $1,345.35 | $85,760 (3,537) | 1.57% | 1.47% to 1.68% | 1.29% | 2.61% |
| 35 | Alaska | $1,398.93 | $89,800 (5,723) | 1.56% | 1.41% to 1.74% | 1.37% | 2.60% |
| 36 | Oregon | $1,374.74 | $89,510 (3,867) | 1.54% | 1.43% to 1.65% | 1.42% | 2.56% |
| 37 | Indiana | $1,142.11 | $76,220 (3,571) | 1.50% | 1.39% to 1.62% | 1.34% | 2.50% |
| 38 | Kansas | $1,295.40 | $86,450 (4,341) | 1.50% | 1.38% to 1.63% | 1.23% | 2.50% |
| 39 | Minnesota | $1,372.61 | $91,890 (2,013) | 1.49% | 1.44% to 1.55% | 1.34% | 2.49% |
| 40 | Massachusetts | $1,610.84 | $112,300 (5,737) | 1.43% | 1.32% to 1.57% | 1.34% | 2.39% |
| 41 | Ohio | $1,133.98 | $79,260 (3,973) | 1.43% | 1.32% to 1.56% | 1.31% | 2.38% |
| 42 | Virginia | $1,358.32 | $97,530 (3,951) | 1.39% | 1.31% to 1.49% | 1.25% | 2.32% |
| 43 | Utah | $1,402.65 | $103,400 (5,174) | 1.36% | 1.25% to 1.48% | 1.24% | 2.26% |
| 44 | Wisconsin | $1,112.23 | $82,140 (3,189) | 1.35% | 1.27% to 1.45% | 1.24% | 2.26% |
| 45 | Iowa | $1,132.96 | $84,370 (4,239) | 1.34% | 1.24% to 1.46% | 1.15% | 2.24% |
| 46 | Idaho | $1,086.21 | $81,250 (2,717) | 1.34% | 1.27% to 1.41% | 1.18% | 2.23% |
| 47 | North Dakota | $1,077.51 | $86,840 (4,834) | 1.24% | 1.14% to 1.37% | 1.04% | 2.07% |
| 48 | Vermont | $1,039.63 | $84,480 (4,492) | 1.23% | 1.13% to 1.35% | 1.11% | 2.05% |
| 49 | Maine | $1,000.13 | $90,440 (4,326) | 1.11% | 1.03% to 1.20% | 1.03% | 1.84% |
| 50 | Hawaii | $1,051.00 | $98,300 (4,263) | 1.07% | 1.00% to 1.15% | 0.95% | 1.78% |
| 51 | New Hampshire | $1,103.33 | $110,500 (6,813) | 1.00% | 0.91% to 1.11% | 0.97% | 1.66% |
| United States (countrywide) | $1,612.81 | $83,160 (634) | 1.94% | 1.92% to 1.96% | 1.72% | 3.23% |
Sources: NAIC 2024 Auto Insurance Database Average Premium Supplement (September 2026), Table 5 (combined average premium) and Table 4 (average expenditure), checked 2026-11-03; U.S. Census Bureau, Current Population Survey Annual Social and Economic Supplement, Table H-8 Median Household Income by State: 1984 to 2025, 2024 column in current dollars with standard errors, checked 2026-11-03. Shares, ranges and the 60% column computed by ClearCarRates. The range treats the NAIC premium as exact and carries only the sampling error of the income median. Published averages for groups in a past year, not a quote for any driver.
Louisiana (3.57%), Florida (2.89%) and Mississippi (2.77%) head the list, and New Hampshire (1.00%), Hawaii (1.07%) and Maine (1.11%) close it. The median of the 51 shares is 1.73% and their simple mean is 1.82%. Neither equals the countrywide 1.94%, which divides the national premium by the national median income; the national median is not the median of the state medians, and a vehicle-weighted premium is not a mean of state premiums. Eighteen of the 51 jurisdictions are above the countrywide ratio, among them Florida, Texas, New York and Georgia, while California (1.68%) is below it. Fifteen are below 1.5%.
The dollar ranking and this ranking agree at the extremes and disagree in the middle. Florida holds first place in premium dollars and second in share, and Louisiana holds second in dollars and first in share. Mississippi is 19th in dollars and third in share, a gap taken up in the section on dollar rank. The two rankings correlate at 0.76 (rank correlation across the 51 rows), which is high but leaves room for large individual moves.
The premium side: two NAIC measures
NAIC publishes two state averages built from the same written premium. The combined average premium, Table 5, is the sum of three separate averages: liability written premium over liability written exposures, collision written premium over collision exposures, and comprehensive written premium over comprehensive exposures. The supplement describes the result this way: “The result is the average cost of an auto insurance policy in the state that contains all three coverages (i.e., liability, comprehensive, and collision).” It prices a vehicle that carries everything, and it is the larger of the two in all 51 jurisdictions. The average expenditure, Table 4, is “the total written premium for liability, collision, and comprehensive coverages divided by the liability written car-years (exposures).” The supplement adds: “This assumes that all insured vehicles carry liability coverage but do not necessarily carry the physical damage coverages (i.e., collision and/or comprehensive).” It describes the expenditure as “an estimate of what consumers in the state spent, on average, for auto insurance.”
For a ratio to income, the choice matters. The combined premium answers “what does a vehicle with full protection cost here, as a share of what a household earns?” The expenditure answers “what did the average insured vehicle cost here, as a share of income, given the coverage choices people actually made?” Nationally the first is 1.94% and the second 1.72%. The table carries both. The ranking is nearly the same either way (rank correlation 0.97), but a few jurisdictions move several places, because the gap between the two measures depends on how many vehicles drop collision and comprehensive. Texas is 6th by share on the combined premium and 9th on expenditure; Kansas is 38th and 44th.
Four features of the premium data limit what a ratio built on it can say. First, the figures are written premium for calendar year 2024 in the combined voluntary and residual markets, not what was billed across twelve installments; the supplement states that its figures “do not include adjustments for dividends to policyholders.” Second, the liability figure is a bundle. The supplement lists uninsured and underinsured motorist coverages, medical payments, and, for particular states, Michigan property protection insurance, Virginia medical expenses, and no-fault or personal injury protection, so the liability line in a no-fault state holds coverage that a tort state buys elsewhere or not at all. Third, the supplement says that aggregate premiums and exposures “are used in calculations with no distinction as to policyholder classifications, vehicle characteristics, or the selection of specific limits or deductibles, all of which significantly impact the cost of coverage. Nor do the results consider differences in state auto and tort laws, rate filing laws, traffic conditions, or other demographics.” Fourth, it closes with a warning that applies to every row here: “CAUTION: Because of these differences, direct comparisons between state results should be treated with a high degree of caution.”
The same supplement lists the factors that drive state-to-state differences. Underwriting expense, coverages purchased, weather, traffic density, theft rates, repair costs, medical and legal costs, rate and form filing laws and liability requirements are on the list, and so is “Per capita disposable income.” That entry matters for how to read the table. Income is not only the denominator. NAIC names it as one of the things that can move the numerator, so a high share is not the same as an unexplained one, and a low-income state is not automatically a high-share state. The national level behind the table, the per-month conversion and the reconciliation with other published averages are in the national average per month.
The income side: what Table H-8 counts
Table H-8 comes from the Current Population Survey Annual Social and Economic Supplement, a household survey; the table notes read “Households as of March of the following year.” It reports, for each state and the District of Columbia, the median household income and its standard error. The definitions sit in the technical documentation published with the 2026 supplement. A household “consists of all the persons who occupy a house, an apartment, or other group of rooms, or a room, which constitutes a housing unit,” and the count “excludes persons living in group quarters, such as military barracks and institutions.” Household income is the sum of the money income of every member aged 15 or over, and the survey asks about “money income received in the preceding calendar year.”
Money income has a boundary that matters for an affordability ratio. Census states that its income data “cover money income received (exclusive of certain money receipts such as capital gains) before payments for personal income taxes, Social Security, union dues, Medicare deductions, etc.” So the denominator is income before taxes, and it leaves out non-money support: the same passage names food stamps, health benefits, subsidized housing and energy assistance as non-money transfers that money income does not reflect. A household’s capacity to pay for insurance depends on what remains after taxes and necessary spending, which this denominator cannot show. The ratio is a standard yardstick for comparing places, not a measure of disposable income.
The documentation also says that “there is a tendency in household surveys for respondents to under report their income.” Wages are better reported than public assistance, Social Security and investment income. A median that is too low in some state would make that state’s share too high. The tables give no way to size it, so it is a stated limit and not a correction.
The median is the middle household, not the typical bill-payer. In a state with a wide spread of incomes the median can sit well below the mean, and half of the households have less than the figure in the table. The national table shows how wide the spread is: the upper limit of the lowest fifth of households in 2024 was $34,160 against a median of $83,160, which is 41.1% of the median. The section on households below the median uses that figure.
Matching the premium year to the income year
The NAIC 2024 column counts premium on policies issued during calendar year 2024. Census labels its columns by the year of income, and the 2024 column reports income received in calendar 2024 by households as of March 2025. The table’s source line runs “1985 to 2026 Annual Social and Economic Supplements,” its 2025 column is the 2026 supplement, and its 2024 column carries a footnote for the implementation of Vintage 2025 population controls. The table pairs 2024 premium with 2024 income. That is the only pairing in which both numbers describe the same twelve months.
Other pairings are common and they move the answer. A shopper reading a 2026 quote against the newest Census column would use 2025 income; the national share of 2024 premium over 2025 income is 1.84%, 0.10 point below the same-year figure, and the largest changes among the jurisdictions are Oklahoma and North Carolina, each 0.31 point lower. Bankrate’s release states that its income data are “2023 data, the most recent available,” set against January 2025 rates. A premium from one year over an income from another year mixes price growth with income growth, and the two have not moved together.
The national series shows how far apart they have drifted. The combined premium rose 12.12% in 2024 while the Census median household income rose 3.16%, from $80,610 to $83,160. Over the four years from 2020 to 2024 the combined premium rose 37.05% and the median income 22.28%. The share fell from 1.73% in 2020 to 1.68% in 2021, when the median income grew 4.07% and the combined premium 1.05%, stayed near 1.7% in 2022 (1.69%), then rose to 1.78% in 2023 and 1.94% in 2024. All five rows below use the NAIC 2024 edition. The supplement states that “Data for prior years were also adjusted to reflect the most current information for each year. Therefore, historical averages in this report might not match those published in prior reports,” so earlier years can differ from older editions.
| Year | Combined average premium | Average expenditure | Median household income | Share (combined) | Share (expenditure) |
|---|---|---|---|---|---|
| 2020 | $1,176.79 | $1,047.75 | $68,010 | 1.73% | 1.54% |
| 2021 | $1,189.15 | $1,060.83 | $70,780 | 1.68% | 1.50% |
| 2022 | $1,257.59 | $1,124.61 | $74,580 | 1.69% | 1.51% |
| 2023 | $1,438.51 | $1,281.61 | $80,610 | 1.78% | 1.59% |
| 2024 | $1,612.81 | $1,431.63 | $83,160 | 1.94% | 1.72% |
Sources: NAIC 2024 Auto Insurance Database Average Premium Supplement (September 2026), Tables 5 and 4, countrywide rows; U.S. Census Bureau, CPS ASEC Table H-8, United States row, current dollars; both checked 2026-11-03. Shares and growth rates computed by ClearCarRates.
The countrywide premium is a vehicle-weighted national average and the Census figure is a national median, so the national share is a context number. The state rows are the benchmarks. For the price side of this movement, including the BLS index and why it can fall while the NAIC average rises, see the price trend page.
What the standard errors do to rank
Census publishes a standard error beside every state median, and the standard errors are large relative to the gaps between neighbouring states. The Census technical documentation explains the use: if all possible samples were drawn under the same conditions, “approximately 90 percent of the intervals from 1.645 times the standard error below the estimate to 1.645 times the standard error above the estimate would include the average result of all possible samples.” It adds: “The Census Bureau uses 90-percent confidence intervals and 0.10 levels of significance to determine statistical validity.” The range column in the table applies exactly that rule to the income and converts the two ends into shares.
The size of the error differs a great deal by state. The standard error as a share of the median runs from 1.6% in Texas and California to 9.1% in Alabama, with 4.3% as the middle value. Alabama’s median of $64,840 carries a standard error of $5,874, which makes its share of 2.14% a reading with a range from 1.86% to 2.52%. Colorado’s standard error is $8,254 on a median of $105,500 (7.8%), giving 1.74% with a range from 1.54% to 2.00%. At the other end, Texas’s $1,294 on $81,280 pins its share to 2.30% through 2.42%.
What this does to the ranking is more informative than any single interval. Louisiana’s range, 3.40% to 3.75%, overlaps none of the other 50. Florida’s range overlaps one (Mississippi’s), and Mississippi’s overlaps two. The first three places are therefore firm. In the middle, the picture is different. Alabama’s 14th place is a reading rather than a rank: three jurisdictions lie wholly above its range, 26 wholly below, and the ranges of 21 overlap it. Across all 1,275 pairs of jurisdictions, 899 pairs (70.5%) have ranges that do not overlap and 376 do. Non-overlap is a stricter test than the Census difference test, so overlapping ranges do not prove two states are alike, but they do mean the published ranks cannot separate them with the sampling error taken at face value.
A second check uses Table H-8B, which publishes three-year moving averages of the state medians and so averages down the sampling noise. H-8B is expressed in 2025 dollars, so the 2022-2024 column was converted to 2024 dollars with the national ratio of current to 2025-dollar median in H-8 ($83,160 over $85,210) before dividing the same premium. The ranks barely move: the rank correlation with the single-year ranks is 0.99, the first three places are unchanged, nine of the top ten are shared, and the only jurisdiction to move five or more places is Alaska, from 35th to 40th. Sampling error blurs the order inside the pack without reordering the ends.
Two cautions about the range column itself. It carries only the income side. The NAIC premium comes from insurers’ statistical reports, not from a sample, so it has no sampling error, but it has the coverage limits listed above and the possibility that “some data from other small statistical agencies may not be included,” as the supplement notes. The true uncertainty is therefore at least as wide as the range shown. And the interval is for the median of all households, which is a different object from the income of any one reader.
Why Bankrate prints 3.39 percent
A quote-based version of this ratio exists and gives much higher figures. Bankrate’s 2025 True Cost of Auto Insurance report, announced in a release dated February 10, 2025, states: “The national average cost for full coverage car insurance is $2,638 per year, which is 3.39% of the median annual household income.” It ranks Louisiana highest at $3,978, or 6.83%. Its method is stated in the release: Bankrate “calculated the true cost of car insurance by dividing the average annual full coverage car insurance premium by the median annual income for each state and metro-statistical area (MSA).” The premiums come from Quadrant Information Services, which supplied January 2025 rates for “all ZIP codes and carriers” in the 50 states and Washington, D.C., weighted by population, for “a single, 40-year-old male and female driver with a clean driving record, good credit” and limits of $100,000 and $300,000 for bodily injury, $50,000 for property damage, matching uninsured motorist limits and $500 deductibles, on a 2023 Toyota Camry driven 12,000 miles a year. The income is “2023 data, the most recent available, provided by the U.S. Census Bureau.”
That is a sound way to price one defined buyer, and it answers a different question from the NAIC ratio. Bankrate’s numerator is a quoted price for a clean-record driver at high limits. The NAIC numerator is premium written for every vehicle insured in 2024, at whatever limits, deductibles, ages and records those policies carried, including liability-only vehicles in the expenditure measure. On the national figures the Bankrate premium is 1.64 times the NAIC combined average premium ($2,638 against $1,612.81), and the gap is a mix of a later rate year, higher limits than many buyers choose, and a driver profile that is not the average one. The release gives no way to split that mix.
The income side contributes less. Bankrate’s national 3.39% implies an income of about $77,800 ($2,638 divided by 0.0339), which is 3.5% below the CPS figure of $80,610 for 2023 and 6.4% below the 2024 figure of $83,160 used here. The release says only that the data come from the Census Bureau, without naming the table. The two effects multiply: 1.64 on the premium times 1.07 on the income is 1.75, and 3.39% divided by 1.94% is 1.75. In Louisiana the same decomposition gives 1.85 on the premium, 1.04 on the income and 1.92 overall (6.83% against 3.57%).
Rankings share more than the levels do. The five highest-burden states in Bankrate’s release are Louisiana, Florida, New York, Nevada and Michigan, and all five fall in the top eight here. The five lowest-burden states in its release are Hawaii, Vermont, New Hampshire, Idaho and Washington; four of them are among the six lowest here, while Washington is 33rd of 51. The reason is in the premium column of the table below. Across these ten states the ratio of Bankrate’s premium to the NAIC combined premium is not constant: it runs from 1.21 in Washington to 1.92 in Florida. A quote for one profile marks up differently from written premium in different places. The causes cannot be separated with these two sources; candidates include how far real policies sit below 100/300/50 limits, how old the typical insured vehicle is, and how rate changes between 2024 and January 2025 differed by state. Those are candidates, not findings.
| Jurisdiction | Bankrate average full-coverage premium (Jan 2025 rates) | Bankrate share of income | NAIC combined average premium (2024) | Bankrate premium divided by NAIC premium | NAIC over Census share (2024) | Rank of 51 here (1 = highest share) |
|---|---|---|---|---|---|---|
| Louisiana | $3,978 | 6.83% | $2,151.12 | 1.85 | 3.57% | 1 |
| Florida | $4,171 | 5.69% | $2,174.55 | 1.92 | 2.89% | 2 |
| New York | $3,916 | 4.77% | $2,149.87 | 1.82 | 2.49% | 4 |
| Nevada | $3,562 | 4.66% | $1,865.83 | 1.91 | 2.33% | 7 |
| Michigan | $3,131 | 4.53% | $1,791.22 | 1.75 | 2.27% | 8 |
| Washington | $1,845 | 1.95% | $1,521.63 | 1.21 | 1.59% | 33 |
| Idaho | $1,454 | 1.94% | $1,086.21 | 1.34 | 1.34% | 46 |
| New Hampshire | $1,823 | 1.88% | $1,103.33 | 1.65 | 1.00% | 51 |
| Vermont | $1,470 | 1.81% | $1,039.63 | 1.41 | 1.23% | 48 |
| Hawaii | $1,689 | 1.77% | $1,051.00 | 1.61 | 1.07% | 50 |
| United States | $2,638 | 3.39% | $1,612.81 | 1.64 | 1.94% |
Sources: Bankrate, press release “National Average Cost of Car Insurance in the United States Reaches $2,638, Up 12% from 2024” (February 10, 2025), state and national figures as printed there; the release lists only the five highest and five lowest states, so no other state is shown. NAIC and Census figures as in the main table, checked 2026-11-03. Ratios and ranks computed by ClearCarRates. Bankrate figures are quoted prices for one defined driver and are not comparable with written premium; no ranking of any insurer is implied.
None of this makes the 3.39% wrong for its purpose. A reader pricing a clean-record 40-year-old at 100/300/50 limits has a reason to prefer a quote-based figure, and its Louisiana share of 6.83% describes that buyer’s burden on those rates against 2023 income. A reader asking how heavily car insurance weighs on households in each state, as actually sold in a given year, has a reason to prefer written premium. Which one fits depends on whether the question is about a profile or about the market, and a page that does not say which it uses is not comparable with either.
Dollar rank against income rank
The same premium produces a different standing once income is in the denominator. The jurisdictions where the two rankings diverge most are those with unusually high or low incomes. Mississippi’s combined premium of $1,542.36 ranks 19th in dollars, close to the middle, but its median household income of $55,620 is the lowest in the table, and the share of 2.77% ranks third. Kentucky ($1,337.15, 36th in dollars) and West Virginia ($1,367.18, 32nd) have premiums below the countrywide figure and incomes under $65,000, and rank 15th and 13th by share. North Carolina’s premium, $1,233.32, is 41st in dollars and 21st by share.
The reverse holds where incomes are high. Massachusetts pays $1,610.84, 18th in dollars, on a median of $112,300, and ranks 40th by share. Utah ($1,402.65, 25th in dollars) is 43rd by share on a median of $103,400, Maryland 12th in dollars and 31st by share, and Colorado 10th in dollars and 25th. The table lists the eight with the largest gaps. A high-income jurisdiction can sit among the highest in premium dollars and still show a lighter burden, which is the pattern an affordability measure is built to find.
| Jurisdiction | Combined average premium | Rank by premium dollars (of 51) | Median household income | Share of income | Rank by share (of 51) |
|---|---|---|---|---|---|
| Mississippi | $1,542.36 | 19 | $55,620 | 2.77% | 3 |
| Kentucky | $1,337.15 | 36 | $64,380 | 2.08% | 15 |
| West Virginia | $1,367.18 | 32 | $62,800 | 2.18% | 13 |
| North Carolina | $1,233.32 | 41 | $66,850 | 1.84% | 21 |
| Colorado | $1,835.17 | 10 | $105,500 | 1.74% | 25 |
| Utah | $1,402.65 | 25 | $103,400 | 1.36% | 43 |
| Maryland | $1,767.57 | 12 | $109,300 | 1.62% | 31 |
| Massachusetts | $1,610.84 | 18 | $112,300 | 1.43% | 40 |
Sources: NAIC 2024 supplement, Table 5; Census CPS ASEC Table H-8, 2024; checked 2026-11-03. Ranks computed by ClearCarRates from unrounded values; 1 is the highest premium or share.
A caution before reading cause into any row. The ranking says where a premium weighs most against a middle household’s earnings. It does not say what makes the premium high or low, and it does not say a lower-share jurisdiction is cheap to insure in. Uninsured-driver estimates by state are on the uninsured drivers page; no link between those estimates and the share of income is tested here. A fuller account of what pushes a premium up is in why car insurance is so high, and the statutory floor for coverage in each state is on state minimum limits.
One vehicle is not one household
The numerator is per insured vehicle and the denominator is per household, and that mismatch is the largest limit on the ratio. NAIC’s averages divide premium by car-years, so a household with two insured vehicles at the average pays about twice the figure in the table. Neither of the two files used here counts vehicles per household, so no state-level adjustment can be built from them. What can be done is bracket the national figure with Census medians by household size from Table H-11 and the quintile limit from Table H-1, which are national only.
The scenario table shows the spread. One vehicle at the countrywide combined premium is 1.94% of the all-household median, 3.73% of the median for one-person households ($43,280) and 1.72% of the median for two-person households ($93,840). Two vehicles at the countrywide premium take 3.88% of the all-household median and 3.44% of the two-person median. Each row is arithmetic on stated inputs, not a finding about any household type: it assumes the household’s vehicle or vehicles sit at the national average, which most households’ do not.
The lesson is that the headline share is for a hypothetical household with one average insured vehicle and a median income. A one-person household with a car has the same premium but a median income only 52% of the all-household median. When the share is used for planning, the figure to compute is the reader’s own: the annual premium for every insured vehicle divided by the household’s total money income before tax. The state table then offers a per-vehicle comparison point for the premium.
| Household | Income | Insured vehicles at the national average | Annual combined premium for those vehicles | Share (combined premium) | Share (average expenditure) |
|---|---|---|---|---|---|
| All households, median | $83,160 | 1 | $1,612.81 | 1.94% | 1.72% |
| All households, median, two insured vehicles | $83,160 | 2 | $3,225.62 | 3.88% | 3.44% |
| One-person households, median | $43,280 | 1 | $1,612.81 | 3.73% | 3.31% |
| Two-person households, median | $93,840 | 1 | $1,612.81 | 1.72% | 1.53% |
| Two-person households, median, two insured vehicles | $93,840 | 2 | $3,225.62 | 3.44% | 3.05% |
| Upper limit of the lowest fifth of households | $34,160 | 1 | $1,612.81 | 4.72% | 4.19% |
| 60% of the all-household median ($49,896) | $49,896 | 1 | $1,612.81 | 3.23% | 2.87% |
Sources: NAIC 2024 supplement, countrywide Tables 5 and 4; Census CPS ASEC Table H-11 (median household income of all households by size, 2024), Table H-1 (limits for household income quintiles, 2024, all races) and Table H-8 (United States median, 2024); checked 2026-11-03. Computed by ClearCarRates. The income levels are national, not state; the last row is a stated scenario, not a Census category. Illustrative arithmetic, not a household benchmark.
Households below the median
Half the households in every state earn less than the median, so the median share understates the burden on the lower half. Census publishes quintile limits nationally, not by state. In 2024 the upper limit of the lowest fifth of households was $34,160. The countrywide combined premium is 4.72% of that income and the average expenditure 4.19%. Because the figure is the ceiling for the lowest fifth, every household in that fifth faces a ratio at or above it for the same vehicle.
For states, no sourced low-income threshold exists in these tables, so the last column of the main table uses a stated scenario: a household at 60% of its own jurisdiction’s median income. The 60% is an assumption chosen for illustration; it is not a Census quantile and it is not the poverty line. Under it, Louisiana’s share is 5.94%, Florida’s 4.82%, Mississippi’s 4.62%, New York’s 4.14% and Georgia’s 4.05%. Louisiana is the only jurisdiction at or above 5%. At the other end, a household at 60% of New Hampshire’s median pays 1.66%, which is lower than a household at the median in Louisiana (3.57%) or Mississippi (2.77%).
The comparison shows why a state average hides distribution. A household at $36,192, which is 60% of Louisiana’s median, would put $2,151.12 into one vehicle’s insurance, 5.94% of income, or about $179 a month out of about $3,016 a month of pre-tax income. A household at 60% of New Hampshire’s median, $66,300, would pay about $92 a month for a vehicle at that state’s $1,103.33 average, against about $5,525 of monthly income. The figures illustrate scale only. A given household’s premium may be far from the state average in either direction.
The limits stated earlier apply here with more force. The denominator counts income before tax and excludes non-money support, and under-reporting of income in household surveys is documented. Those make the burden on a low-income household harder to state exactly than the median-household figure, and none is a reason to use the median share as though it described every household.
Rebuilding every row from the sources
Every figure here can be recomputed from two public files and one published release. The NAIC file is the 2024 Auto Insurance Database Average Premium Supplement, dated September 2026. Table 4 (average expenditure) is on printed page 18 and Table 5 (combined average premium) on printed page 19. In each, the first numeric column is 2024 and the last row is Countrywide, which reads $1,431.63 in Table 4 and $1,612.81 in Table 5. The Census file is Table H-8 in the historical income tables, an Excel workbook with the median in one column and its standard error in the next for each year. In the current-dollar block the 2024 pair sits in columns D and E (rows 10 to 61, United States in row 10), and a second block below repeats the series in 2025 dollars. Use the current-dollar block, because the NAIC premium is a 2024 dollar figure.
Take care with the NAIC text. When the PDF is converted to text in a mode that tries to preserve physical layout, several rows of Tables 4 and 5 come out with values attached to the wrong state: in Table 4 the values from Hawaii onward are shifted by one row, and in Table 5 the later year columns shuffle between neighbours. Text converted in content-stream order and text converted in a table-aware mode give identical values for all 104 rows of the two tables (52 rows each including Countrywide, five year columns), and the 25 states already published on this site’s state pages match to the cent. Check any extracted row against the printed page, with Louisiana’s row as a quick test: Table 4 prints $1,880.42 and Table 5 prints $2,151.12.
The state labels need checking too. NAIC lists 50 states and the District of Columbia, and Census lists the same 51 names plus a United States row, spelled identically, so the join needs no mapping. The calculation is then one division per row: premium over median income, multiplied by 100. For the range, divide the premium by the median plus 1.645 standard errors to get the lower end and by the median minus 1.645 standard errors to get the upper end. Louisiana’s check: 2,151.12 divided by 60,320 is 3.566%, rounding to 3.57%. The standard error of 1,826 times 1.645 is 3,004, so the range runs from 2,151.12 divided by 63,324 (3.40%) to 2,151.12 divided by 57,316 (3.75%).
Putting your own premium on the scale
To place a policy against these benchmarks, compute two ratios separately. The first is the premium for one vehicle against your state’s combined average premium or average expenditure in the table, which says whether your price is high or low for the state; the Car Rate Checker does that comparison, and comparing quotes at matched limits comes before it. The second is your annual premium for all insured vehicles divided by household money income before tax, which says how heavily the bill weighs on your household. Only the first says anything about whether a price is out of line. The second can be high because the premium is high, because income is low, or because there are many vehicles.
Here is a labelled illustration with stated inputs. A household in Louisiana with two insured vehicles, each at the state’s 2024 combined average of $2,151.12, and an income equal to the state median of $60,320 would spend $4,302.24 a year, which is 7.13% of income. If the same household held two vehicles at the state’s average expenditure of $1,880.42, the total would be $3,760.84, or 6.23%. Neither is a quote and neither is a prediction for a Louisiana household; they show that a household with two vehicles at the state average is above 6% of the median income in the highest-share jurisdiction, where the single-vehicle table value is 3.57%.
What can be changed is the premium side. Limits, deductibles, driver and vehicle details and discounts are the levers a buyer has, and the guides on lowering a premium and liability-only against full coverage work through them. The state benchmark pages show the premium side for each state, and the state rate map places each NAIC figure beside the national one.
The table cannot answer whether a premium is fair for the household that pays it. A 3% share can be reasonable for a driver with a recent claim in an area with high repair costs, and a 1.5% share can be too much for a clean-record driver whose insurer filed a large increase. A share-of-income ratio describes burden. A price benchmark describes level. Reading one as the other is the most common error in the figures that circulate.
When the next editions arrive
Both inputs are on a schedule that the table will outrun. The NAIC supplement dated September 2026 is the newest edition found on the check date, and carries data through 2024; it says the full 2023/2024 report with loss data will follow. Census Table H-8 runs through the 2025 income year, from the 2026 supplement. A 2025 NAIC edition would pair with the 2025 Census column already published, which puts the national median at $87,460. If premiums keep rising faster than incomes, as they did from 2020 to 2024 (37.05% against 22.28%), the shares will rise; that is a condition, not a forecast. The price trend page tracks the BLS index for the period after the NAIC data end.
Two things would change the table without any price movement. Census revises its population controls, as the footnote on the 2024 column shows, and NAIC restates earlier years in each edition, so a rebuilt table in a later year can differ from this one. And a jurisdiction’s rank would change if its median income moved by more than the sampling error shown, which is between 1.6% and 9.1% of the median depending on the state.
Questions readers ask
What share of income do Americans spend on car insurance?
On the NAIC 2024 combined average premium ($1,612.81) and the Census 2024 median household income ($83,160), one insured vehicle with liability, collision and comprehensive coverage costs 1.94% of income, with a 90% range of 1.92% to 1.96% from the income standard error. On the NAIC average expenditure ($1,431.63) it is 1.72%. A household with two insured vehicles at the average is near double those figures.
Which state has the highest car insurance cost as a share of income?
Louisiana, at 3.57% (range 3.40% to 3.75%) on 2024 data, followed by Florida at 2.89% and Mississippi at 2.77%. Louisiana’s range overlaps none of the other 50 jurisdictions’ ranges. New Hampshire has the lowest share, 1.00%, with Hawaii at 1.07% and Maine at 1.11%.
Why is Bankrate’s percentage so much higher than the one here?
Bankrate divides a quoted full-coverage premium for one 40-year-old with good credit ($2,638 nationally, January 2025 rates) by 2023 income, and gets 3.39%. The table here divides premium written in 2024 on all insured vehicles by 2024 income and gets 1.94%. The quoted premium is 1.64 times the NAIC combined average, and its implied income is about 6% lower than the 2024 Census median; together they give the 1.75-times gap.
Does a higher share of income mean insurance is more expensive in that state?
Not by itself. A share can be high because the premium is high, because incomes are low, or both. Mississippi is 19th in premium dollars and third by share because its median income of $55,620 is the lowest of the 51. Use the state premium benchmark to judge a price, and the share to judge the burden on a typical household.
Why do you divide by median household income and not by personal income?
Insurance is bought by households, usually one policy for several drivers and vehicles, and Census publishes standard errors for the household median in Table H-8. The limit is that NAIC averages are per insured vehicle, so a household with several vehicles pays several times the figure, and the all-household median hides the lower median of one-person households ($43,280 against $83,160 for all households in 2024).
How reliable are the state shares?
The income median is a survey estimate with a standard error from 1.6% of the median in Texas and California to 9.1% in Alabama. Louisiana, Florida and Mississippi are separated from every lower state by that error, while mid-table ranks are not: 899 of the 1,275 pairs of jurisdictions have non-overlapping 90% ranges. A three-year average of the medians (Table H-8B) leaves the order nearly unchanged, with a rank correlation of 0.99.
Can I use these shares to see whether my own premium is fair?
No. A share of income describes the burden of a premium on a household, not whether the premium is in line with its state. For that, compare your premium for one vehicle with the state combined average premium or average expenditure, at matched limits, in the Car Rate Checker. The benchmarks are published averages for groups in 2024, not quotes.
Sources, checked 2026-11-03
- NAIC, 2024 Auto Insurance Database Average Premium Supplement (September 2026): Introduction, Terms and Calculations, Technical Notes, Table 4 (average expenditure) and Table 5 (combined average premium), state and countrywide rows, checked 2026-11-03.
- U.S. Census Bureau, Table H-8 Median Household Income by State: 1984 to 2025 (CPS ASEC; medians and standard errors, current and 2025 dollars), checked 2026-11-03.
- U.S. Census Bureau, Table H-8B Median Household Income by State Using Three-Year Moving Averages (2022-2024 column, 2025 dollars), checked 2026-11-03.
- U.S. Census Bureau, Table H-1 Limits for Household Income Quintiles and Top 5 Percent of All Households: 1967 to 2025 (2024 row), checked 2026-11-03.
- U.S. Census Bureau, Table H-11 Median and Mean Household Income of All Households by Size of Household: 1975 to 2025 (2024 rows), checked 2026-11-03.
- U.S. Census Bureau, Current Population Survey 2026 Annual Social and Economic Supplement technical documentation (glossary: household, income, total money income; standard errors and their use), checked 2026-11-03.
- U.S. Census Bureau, Historical Income Tables footnotes (footnote 42, Vintage 2025 population controls; page last revised August 18, 2026), checked 2026-11-03.
- Bankrate, press release: National Average Cost of Car Insurance in the United States Reaches $2,638, Up 12% from 2024 (February 10, 2025), 2025 True Cost of Auto Insurance report figures and methodology. A later Bankrate edition, if published, is not reflected here, checked 2026-11-03.