How premiums change with age: the U-shaped curve from 16 to 85

Premiums follow expected claim costs, and the cost of insuring a driver is highest at 16 to 19, drops steeply through the twenties, bottoms out in the early sixties, then rises again from about the early seventies. Loss data from the Highway Loss Data Institute (HLDI) shows that shape clearly: the youngest drivers file the most claims, the oldest file more than the sixty-somethings, and the older group's claims cost more to pay when people are hurt. The rest is detail about how steep each side is, and that depends on the coverage, the state and the company.

This article uses crash and insurance-loss data, not company quotes, to describe the curve. It then covers the states that restrict age as a rating factor, the difference between age and driving experience, and what drivers in each life stage can do about their own price.

The curve in loss data

The cleanest public look at insurance losses by age comes from the Highway Loss Data Institute, a nonprofit that gathers and publishes data on the losses from owning and operating motor vehicles and works alongside IIHS. Its Loss Facts sheet, Insurance losses by rated driver age (April 2007), compares crash losses under five coverages for 2002-04 model passenger vehicles by the age of the "rated driver," the person assigned to a vehicle for insurance purposes, who is not necessarily the one driving at the time of a crash. Everything is indexed so that rated drivers aged 30 to 59 equal 100.

The data are old, and the vehicles and traffic are different now. They still make a useful illustration, and HLDI's later work agrees with the direction. A 2014 HLDI bulletin on 2000-13 models found a collision claim frequency of 10.7 claims per 100 insured vehicle years for rated drivers 15 to 17, against 5.9 for ages 35 to 50, "nearly twice" the prime-age rate.

_Index values from HLDI's April 2007 Loss Facts, 2002-04 models; 30-59 = 100. Table is a reading of HLDI's published figures, not current premiums._

Three features stand out.

• The low point is in the sixties, not the middle years. The HLDI sheet says claim frequencies under all five coverages were highest for ages 16 to 19 and lowest for people in their 60s.
• The young end is steeper than the old end. The 16-to-19 group's collision, property damage liability and bodily injury losses were about three times those of rated drivers 30 to 59. At the older end, the index passes 100 around the mid-seventies for collision and bodily injury, and reaches 150 to 256 only at 85 and older.
• The sides rise for different reasons. Young drivers drive the loss up mainly through how often they crash (collision frequency 214, against payments per claim of 139). Older drivers do it through a mixture of more crashes and higher payments when injuries are involved, as explained below.

Our guide on claim frequency vs severity: which one actually moves premiums? explains why those two parts are priced separately.

Why risk falls steeply through the twenties

Inexperience and exposure explain most of the early steepness. The IIHS teenagers page says the fatal crash rate per mile for drivers 16 to 19 is about three times the rate for drivers 20 and older (2022 data), and that on police-reported crashes of all severities the rate is more than four times. The Centers for Disease Control and Prevention's teen driver risk page says risk is particularly high during the first months of licensure, that teens are more likely than older drivers to underestimate or fail to recognize dangerous situations, and that the per-mile crash rate is about 1.5 times higher for 16-year-olds than for 18- and 19-year-olds.

Other conditions pile on. IIHS reports that passengers raise crash risk for teen drivers but lower it for drivers 30 and older, and that 35% of teen driving deaths in 2024 occurred between 9 p.m. and 3 a.m. Teen crash deaths have fallen by half since 1996, from 5,819 to 2,899 in 2024, a period in which every state adopted some form of graduated driver licensing. IIHS also cites research linking a licensing age of 17 instead of 16 to a 9% lower collision claim rate among 16-year-old licensed drivers.

So the early part of the curve is a learning curve. Each year of driving removes some of the inexperience, and the next age band's loss index is lower: 214 at 16-19, 164 at 20-24 and 122 at 25-29 in the HLDI collision frequency column. Our guide on when do car insurance rates drop for young drivers? translates this into what to expect year by year, and why teen drivers cost so much to insure covers the first years in detail.

The flat middle and the low point

HLDI groups every rated driver from 30 to 59 into one baseline, so its sheet cannot show differences within those three decades. What sorts drivers within those decades is mostly not age but the other rating factors: your driving record, mileage, vehicle, ZIP code and the coverages you buy. The sheet does not measure those, so it supports only the shape of the curve, not the size of anyone's price.

HLDI's lowest claim frequencies fell in the 60-to-64 band, which fits a picture of long driving experience before age-related risks start to build, though the sheet does not test that explanation. Our guide on how car insurance rates are calculated explains how all those factors combine.

Why the curve turns up again after about 70

Two things push the older end up, and they need to be separated.

More crashes per vehicle. In the HLDI data, collision, property damage liability and bodily injury claim frequencies "increased sharply and doubled by age 85 and older" compared with their low point, while personal injury protection and medical payments frequencies rose about 30 percent. HLDI read the sharp rise in bodily injury and property damage losses beyond age 70 as suggesting drivers' increasing responsibility for their crashes.

More severe outcomes when a crash happens. Payments under the three injury coverages (PIP, bodily injury and medical payments) were about one-third higher among the oldest drivers than among 30-to-59-year-olds, while collision and property damage payments were only slightly different. The IIHS older drivers page says a 2005-08 study found older drivers' elevated per-mile fatal crash rates stem mainly from fragility, meaning a greater risk of dying in a given crash, not from being involved in more crashes, and that similar results hold for 2018-24. Fragility starts rising around middle age and keeps climbing.

Mileage complicates the picture. IIHS notes that per-mile fatal crash rates start rising at about age 70 in part because older drivers generally drive fewer miles. Older drivers' fatal crash involvement rates per licensed driver and per mile have fallen since the 1990s, and the rate per licensed driver for those in their 70s dipped below that of drivers 35 to 54 for the first time in 2015. For drivers 80 and older, IIHS says multiple-vehicle intersection crashes accounted for 43% of fatal crashes in 2024, compared with 24% for drivers ages 16 to 59.

NHTSA's Older Population fact sheet (DOT HS 813 719, May 2025) adds the national picture for 2023. The traffic fatality rate per 100,000 population was 16.50 for ages 80 to 84 and 15.85 for 85 and older, compared with 12.47 for ages 65 to 69 and 11.92 for people under 65. Older drivers made up 22 percent of licensed drivers. These are death rates per resident, not claims, so they speak to severity and exposure rather than premium directly. Our guide on Crash statistics and premiums: reading NHTSA FARS and IIHS data explains how to read this sort of statistic, and the article how car insurance rates change from age 65 to 85 looks at the older end in more depth.

How steep is each side? A worked example with defined numbers

Here is how a loss index turns into a rough price idea. The dollar figures are invented. The only real data are the HLDI index values above. Real premiums are set under each company's filed rating plan and then limited by state law, so treat this as arithmetic, not a forecast.

Say that for a baseline driver in the 30-to-59 band, the expected yearly collision loss is $600 and the expected bodily injury loss is $700, for $1,300 before expenses and profit.

Two lessons from the arithmetic. First, a driver in the early seventies on this basis looks about the same as the baseline, so a large increase at that age would need to rest on something other than age, such as the driver's own record. Second, the older curve gets steep late, and mostly through injury coverages. Which coverages you carry, and in what amounts, decides how much of the older end's cost reaches your bill. Our tool renewal rate increase checker helps compare your own renewal change with what is plausible.

States that limit age as a rating factor

Age is a standard rating factor in most states, but a few limit it.

• Massachusetts. Chapter 175E, section 4 provides that motor vehicle rates for an insured age 65 or older who otherwise qualifies for the lowest rate classification "shall be twenty-five per cent less than the applicable rate for such classification," and that risks may not be grouped by sex, marital status or, except for that reduction, age (M.G.L. c. 175E, section 4).
• Hawaii. Hawaii's insurance commissioner stated in 2006, citing HRS 431:10C-207, that no insurer may base a standard or rating plan "in whole or in part, directly or indirectly" on a person's age, sex, length of driving experience, credit bureau rating or marital status, among other traits (Insurance Journal, 2006). Check the current statute before relying on the wording.
• Louisiana. Act 168 of the 2025 Regular Session (House Bill 258) amends the state insurance code to bar an insurer from increasing an automobile liability premium solely because the insured has attained age 65 or older (Louisiana Legislature bill page; trade summary in Insurance Business, June 2025). Insurers may still use driving history and claims. See Louisiana.
• California. Proposition 103 makes the driver's safety record, annual mileage and years of driving experience the mandatory rating factors, and the California Department of Insurance's auto rating factors study says they have the greatest influence on premiums. Age is not one of the three. See California.

Separately, the NAIC's model law on declination and termination says a policy may not be terminated because of the insured's age, sex or marital status, while explicitly allowing rating differentials based on them (NAIC Model Act 725, 1997). States adopt such rules individually. Our article can an insurer drop you for being old? covers the underwriting side, and the guide on how age, gender and marital status affect car insurance covers the neighboring factors. To see your own state's requirements, use state car insurance requirements lookup and ask your department of insurance about its rating rules.

Age is not the same as experience

The HLDI data are organized by age, but a 16-year-old and a 45-year-old getting a first license are not the same risk, and the data cannot separate them. What the research does show is that inexperience is a large part of the youthful risk: the CDC highlights the first months of licensure and the drop between 16 and 18-to-19.

Insurers' own plans use "years licensed" or "driving experience" alongside age, and some places treat them as separate. California makes experience a mandatory rating factor; Hawaii names length of driving experience among the traits that cannot be used. For an adult who is licensing late, the effect varies by company. Our article on learning to drive as an adult: how insurers price a first policy covers how those first policies are priced. Time insured matters too: continuous coverage and a clean record build the history that lowers the curve whatever your age.

What drivers can control at each stage

Some states mandate mature-driver discounts for completing an approved course. California (Insurance Code 11628.3) and Florida (statute 627.0652) each require a discount for drivers 55 and over who complete a qualifying course, lasting three years; check your state's current rules. For older drivers, our articles retiring and your car insurance: what to tell the insurer and driver assistance features for older drivers: what helps, what pays describe two levers that matter.

Common questions

At what age do car insurance rates start to drop?

In loss data, risk falls year by year from 16 through the twenties and reaches its low in the early sixties. Prices do not follow the data exactly, because insurers also weigh your own record, but a clean history after a few years of driving generally helps. Your actual drop depends on your company, state and record.

Do rates really rise automatically at 65 or 70?

Not in a way that the HLDI data support. Its loss index for ages 65 to 69 was close to the low point, and for 70 to 74 near the baseline. Rates that jump sharply at 65 should come with a reason you can identify, such as a claim, a violation or a change in coverage. Massachusetts requires a 25% reduction at 65 for qualifying drivers, and Louisiana bars a liability increase solely for turning 65.

Why do older drivers sometimes pay more even with clean records?

Because the pricing is by class, and the class's average loss goes up with age, particularly for bodily injury when injuries are more severe. A clean record, low mileage and safety equipment can partly offset it.

Is age discrimination legal in car insurance pricing?

In most states age is a permitted rating factor if it is actuarially supported and filed with the regulator, and a few states restrict it. The details above show how narrowly the exceptions are drawn. A licensed attorney or your state insurance department can address a specific complaint.

What drives teen premiums most?

The age band itself is the biggest piece, but you control more than you might think: the vehicle, the number of teen drivers on the policy, the discounts and the record. Our guide why teen drivers cost so much to insure and the article choosing a first car for a teen using safety and insurance data go into the choices.

Key takeaways

• Premiums follow expected loss, and loss by rated driver age is high for teens, lowest in the early sixties, and rises again past about 70. HLDI's 2002-04 data put collision frequency at 214 for ages 16 to 19 and 160 for 85 and older, against 94 at 60 to 64.
• The teen side is mostly about inexperience and exposure. The older side combines more crashes with more severe injuries; IIHS research points to fragility rather than over-involvement as the main cause of elevated per-mile fatal rates.
• The early seventies sit near the baseline in HLDI's index, so a big jump at that age should have an identifiable reason beyond age.
• Age rules vary: Massachusetts requires a 25% cut at 65 for qualifying drivers, Louisiana bars a liability increase solely for turning 65, and Hawaii's commissioner has cited a ban on age, among other factors. Check your own state.
• Each stage has levers: driving record, continuous coverage, vehicle, mileage and discounts, plus shopping around at each life change.