It's not just one company. The problem is our current driving is something under 4,000 miles/year and the discounts for low driving aren't remotely as big as the difference in mileage.

I don't think my personal situation affects the cost because the puny discounts for low mileage aren't limited to certain groups but here goes: 52 year old married male (my wife doesn't have the vision to drive, she has no bearing on it), no at-fault accidents (one accident where the other side was clearly at fault, it didn't change my rates) or tickets within the window they can look at. The car is garaged in the suburbs. I have a 4 year old car with 17k miles.

The vast majority of their risk comes from driving and thus should be approximately linear with the amount of driving I do. Things rarely happen to cars in garages.

  • 19
    'Driving less' could be argued as giving you less experience/training, so your risk higher. Not sure if they use that argument, or if it's even true, but it could be.
    – Aganju
    Commented Sep 12, 2017 at 3:26
  • 45
    It sounds like your question is "why is there no pay-as-you-go car insurance?" but there is pay-as-you-go car insurance, where insurance costs are scaled per mile driven. Do a web search for pay-as-you-go car insurance and you'll find a number of companies offering this service. Commented Sep 12, 2017 at 12:23
  • 4
    On many cars it's easy to tamper with the odometer. The company would be stupid to trust it. Commented Sep 12, 2017 at 12:59
  • 6
    @sgroves Lying to your insurance company will rarely be a problem until the time comes that you have to make a claim; then they might want to check the service history of your car. A lot of the newer limited mileage deals will require you to have a GPS tracker installed also.
    – richardb
    Commented Sep 12, 2017 at 17:26
  • 2
    @ Eric Lippert: the OP clearly addressed your comment by noting that there are discounts for lower driving, but they aren't commensurate with the decrease in mileage. Metro-Mile, for instance, has a base fee of about $30/month, and then they charge more on top of that for every mile driven. $30/month is commensurate with what someone with a good driving history can get without any restrictions on mileage. Commented Sep 13, 2017 at 14:47

12 Answers 12


There are several aspects to this but at a high level it boils down to

  • To some extent the insurer has fixed costs.
  • Underwriters apparently think there is a marginal decrease in risk for each marginal mile driven.
  • In the case of comprehensive coverage the car is being covered whether or not it's being driven, insurers are about to get pummeled with flood claims even though many of the affected autos were just sitting in garages and on streets.

A lot goes in to insurance rating and risk projecting. You can't adjust a single variable and expect a proportional change in your premium, 7,000 miles per year just won't be 70% of the cost of 10,000 miles per year, because there are a lot of other things in play as well.

To further address premium adjustments. Consider this:

Your fixed cost policy fee is $25
Your liability coverage is $100
Your comprehensive coverage is $75

Your total premium is $200

Even if your liability coverage did scale with perfect correlation to your mileage (using the same 70% from above, 7,000 miles per year versus 10,000 miles per year) then your premium composition is:

Policy fee: $25
Liability coverage: $70
Comprehensive coverage: $75

Total premium: $170

$200 to $170 is 15%. No change will have a direct linear correlation to your total premium because there are different component pieces of the total premium. Fixed costs may be built in to the amounts for other component pieces of the premium, for example maybe no line of coverage ever has a cost below $X.

Obviously these numbers are all made up

Additionally, and also less considered is the fact that your liability also scales because of a lot of factors that have nothing to do with you. It might be the other cars that are on the road, it might be that more densely populated areas have more fender benders. For example if you live in Beverly Hills you have a much higher likelihood of accidentally bumping a $70-$80-$90-$100k+ car than you do in say, rural Wisconsin. If your zip code is gentrifying and everyone starts buying Mercedes, your liability coverage increases.

You can not adjust one single variable and decide that you are lower risk than all insurers think you are. If you shop this coverage and all insurers are within a nominal margin of pricing for the same coverage levels, there isn't much to argue with; you are simply riskier than you think you are and the variable you are focused on is not as meaningful as you think it is.

  • 3
    Great points by bringing up the fixed costs and pointing out that car insurance covers damages even when the vehicle isn't being driven. This also comes into play with hail and other storm related damages (trees or buildings falling on your car).
    – airfishey
    Commented Sep 12, 2017 at 13:21
  • 3
    When it's not being driven that's comprehensive coverage and I would expect that to not scale anything like linearly. Commented Sep 12, 2017 at 18:35

Not all miles carry the same amount of risk. A survey by Progressive indicated that accidents are most likely to occur within 5 miles of home, and 77% of accidents occur within 15 miles of home. Only 1% of accidents occurred 50 or more miles from home. That's from 2002, but it seems unlikely to have changed much. Since the miles closest to your home carry more risk, they cost more, and low-mileage discounts reflect that.

There are per-mile insurance options in a few states which could save you money, but they do constant monitoring via that ODB2 telematics device, and other insurers offer discounts if you accept their monitoring either in perpetuity or for a limited period of time.

Without monitoring, insurers don't know if that 4,000 miles of driving is spread into a few mid-day trips each week, or maybe you're doing all that driving from midnight to 4am on weekends (fatalities far more likely), or from 5-7pm during weekdays (accidents far more likely).

Personally, I save ~10% by being a 'low-mileage' driver, and am currently in the middle of a 90-day monitoring, so might go lower, but given that accidents are far more likely close to home, 10% feels pretty significant and appropriate.

  • 65
    Of course most accidents are near home--that's where you're most likely to be unless you have a long commute to work. That doesn't show that driving near home is the most dangerous. And they have a fair idea that it's not the 5-7pm as my current insurance is in a rate category for driving to work less than 1/week. Commented Sep 12, 2017 at 0:01
  • 4
    @LorenPechtel Right, but half the mileage doesn't mean half the risk on its own, the commuting once a week part is more important than the mileage as independent variables go, so it's great if they actually take that into account for your rate, that's atypical without monitoring. Even if you have a longer commute, the statistics show you're more likely to have an issue close to home, fewer miles independent of other variables doesn't reduce risk that much.
    – Hart CO
    Commented Sep 12, 2017 at 2:38
  • 4
    All that survey shows is that 77% of the time people drive within 15 miles of their homes. That doesn’t make that area more dangerous. If OP spends 10% of his time within 15 and 90% over 15 miles away, his chance of a crash within 15 miles is pretty close to 10%.
    – Tim
    Commented Sep 12, 2017 at 10:34
  • 4
    @LorenPechtel The point about "near to home" isn't about actual distance, it's about the fact that most of us live in towns. From racfoundation.org/motoring-faqs, "In 2015, the majority of injured casualties occurred on built-up roads (72 per cent of total casualties)". If you're spending most of your driving time on roads which are the most likely to result in an accident, that's why your insurance isn't so cheap.
    – Graham
    Commented Sep 12, 2017 at 14:08
  • 8
    This statistic tells us where average drivers have the most accidents. It tells us absolutely nothing about people who drive less than average, who might have completely different accident patterns. As such, it is not a sensible answer to the OP's question. Commented Sep 12, 2017 at 22:46

4000 miles a year is not a few! European average is about 9000... But nevertheless...

But when it comes to risk, then:

1) Nothing stops you from changing circumstances and drive 10 times as much as in previous yers. The insurance remains the same. The only thing the insurance company can do is to charge you more next year (taking the miles you've made this year as a basis for calculations)*

2) Drivers who drive very seldom are a huge risk because of their low experience. I know a few people that drive more than 100 miles only a few times a year, and on average once a year have accident during that drives. It doesn't mean that an average sunday driver have similar risk of accident as daily driver, but it's in no way similar.

*) Germany/Switzerland based, the whole EU is likely to be the same

  • 11
    The point that people who rarely drive don't get enough practice to keep their skills up to scratch is probably quite significant, and isn't mentioned in the other answers so far. Commented Sep 12, 2017 at 13:32
  • 3
    Don't know about OP but 12k miles/year is closer to average for US. Commented Sep 12, 2017 at 15:19
  • 5
    I don't know what it's like elsewhere in the EU but UK insurers can and do void peoples insurance for failing to comply with the terms. Commented Sep 13, 2017 at 0:37
  • 5
    In much of the western US, a drive between neighboring towns can be the same distance as a drive between European countries.
    – jamesqf
    Commented Sep 13, 2017 at 3:28
  • 2
    @jamesqf - I know! Also, when I went to college in my home state my fellow students were all like - what do you mean you've never been out of your state? They were from those dinky states on the east coast, I was from California ...
    – davidbak
    Commented Sep 13, 2017 at 22:24

Many services charge prices that do not scale linearly with usage. This is because the service provider has fixed costs that they must recoup by charging a rate with a fixed component. A 5-mile taxi ride is unlikely to cost half what a 10-mile taxi ride costs. Even a half sandwich at a sandwich place usually costs more than half of what a full sandwich costs. In this respect, insurance is no different from many other items you may purchase.

  • 1
    You are explaining why the pricing structure results in prices not scaling linearly, but utterly failing to explain why the pricing structure is the way it is, which I think is OP's main question. Just saying that there are fixed costs is little more than a restatement of the OP's observation that prices are nonlinear. Why are there fixed costs? Commented Sep 13, 2017 at 15:00
  • 2
    @Acccumulation: Why are there fixed costs? Pardon me, but that is an absurd question. That's just the way the world works. It's not like some policeman comes and in says "Hold it right there, pal, you have to have fixed costs." To run an insurance company you have to decide how much to charge people, set up some way for customers to contact you, and probably pass some regulatory muster, not to mention stockpiling enough cash to pay claims that may arise before you earn much. You can't just go out and sell your first insurance policy to some guy on the street without doing any prep work.
    – BrenBarn
    Commented Sep 13, 2017 at 18:08

Some proportion of the costs of a policy have little to no relationship to miles driven. Think of costs of underwriting, and more especially sales/marketing/client acquisition costs (auto insurance isn't in the same league as non-term life insurance (where the commissions and other selling expenses typically exceed the first year's worth of premiums), but the funny TV ads and/or agent commissions aren't free), as well as general business overhead.

Also, as noted by quid, some proportion of claim risk isn't correlated to distance covered (think theft, flood, fire, etc.).

There are also differences in the miles that are likely to be driven by a non-commercial/vehicle-for-hire driver who puts 25k miles a year vs. one who puts 7k per year. The former is generally going to be doing more driving at higher speeds on less-congested freeways while the latter will be doing more of their driving on crowded urban roads. The former pattern generally has a lower expected value of claims both due to having fewer cars per road-mile, fewer intersections and driveways, and also having any given collision be more likely to result in a fatality (paralysis or other lifetime disability claims are generally going to exceed what the insurer would pay out on a fatality).

  • Things like fire & flood apply only to the comprehensive part of the policy--and to a large degree the risk is related to the number of trips. Commented Sep 12, 2017 at 2:37
  1. Other people lie to the companies about how many miles they drive, so they can't take the mileage figures literally.

  2. You aren't specifying whether you want liability only, or more-comprehensive insurance. Stuff happens when you aren't driving. Cars get stolen. Other drivers hit parked cars and leave. Trees fall on parked cars.

  3. Move to Virginia where insurance is not required. Just pay $500 a year for not having insurance, and be careful.

  • 1
    Point number one is not possible with many insurance companies as they will periodically ask for odometer readings from your car. Mine also takes into account the distance of my home from my job, and uses their own algorithms to determine how much they think I drive. Don't lie to your insurance company.
    – BlackThorn
    Commented Sep 12, 2017 at 15:44
  • 4
    @jamesqf And when they deny your claim because you have way more miles than you told them? Like I said, don't lie to them. Also, disconnecting your odometer is a felony crime in the US. justice.gov/civil/case/federal-odometer-tampering-statutes And purposefully lying to your insurance company is fraud. Underestimating when they ask an approximation is one thing, lying about a reading is another thing entirely.
    – BlackThorn
    Commented Sep 12, 2017 at 17:00
  • 2
    @DavePhD That link does not support your claim about Virginia. It says that someone caught without insurance is fined $500, and then "During [a] three-year period, insurance companies notify DMV if the insurance coverage is canceled."
    – jpaugh
    Commented Sep 12, 2017 at 22:05
  • 1
    @jpaugh you're right that the link isn't the best reference. Better is "Virginia law does not even require a driver to have insurance. Owners can opt to pay a $500 uninsured motor vehicle fee to the Department of Motor Vehicles. " valawyersweekly.com/2011/12/19/… The link in the answer is talking about a fine in a different situation, where you don't pay the fee and don't have insurance.
    – DavePhD
    Commented Sep 13, 2017 at 0:48
  • 1
    @jpaugh and still better is the statutory language itself "In addition to any other fees prescribed by law, every person registering an uninsured motor vehicle, as defined in § 46.2-705, at the time of registering or reregistering the uninsured vehicle, shall pay a fee of $ 500" vacode.org/46.2-706 So it is a fee, not a fine. But if you don't pay the fee, then you are fined.
    – DavePhD
    Commented Sep 13, 2017 at 0:52

because it cost the insurer more, obviously. while this sounds snarky, it's important to realize that actual insurance companies set their insurance rates based on actual historical costs. for some reason people who report low miles have cost the company more dollars per reported mile than people who report high miles. in that sense, insurance is not overpriced. if it were truly overpriced, then an insurer would specialize in such insurance and make a killing on the free market.

the more interesting questions is why do drivers who claim to travel very few miles cost the insurance companies so much per mile? that question has a host of possible answers and it's difficult to say which is the largest cost. here are just a few:

  1. some people lie about how many miles they drive. these liers are also more likely to lie about other things like how many accidents they've been in or whether some damage to their car was caused by a covered accident or some uncovered event. moreover, it's nearly impossible to catch a liar since well-meaning people grossly underestimate their miles driven all the time. in fact, on the rare occasion when an insurer tries not to cover an accident because the driver far exceeded their mileage, the public outcry is so negative they are effectively unable to do so.
  2. some insurance costs are unrelated, or even inversely related to typical miles driven (e.g. hail damage, theft risk, parked car accidents, advertising, billing and customer support overhead, accidents while driving to the hospital, windshields broken by crazy exes, etc.)
  3. people who drive less often tend to be worse drivers. obviously, practice makes perfect. less obviously, some people drive less because they know they're bad at it or simply can't be bothered to pay attention. moreover, people tend to pay less attention to their driving when they are on a familiar road and are therefore more likely to hit a deer, etc. than when on a road they must examine closely to navigate.
  • "because it cost the insurer more, obviously" -> But if it were "obvious", the question and your answer were not written.
    – phresnel
    Commented Sep 18, 2017 at 8:55
  • it's only obvious once someone points it out. en.wikipedia.org/wiki/Egg_of_Columbus Commented Sep 18, 2017 at 21:16
  • Yes, that's what I mean. Likewise Alexander the Great and the Gordian Knot. Obviousity is always relative to the one who uses the term. Personally, I've quit using that word.
    – phresnel
    Commented Sep 19, 2017 at 8:51

First you have to understand that insurance is basically a social system, just with Shareholders. Insurance costs consist of 3 factors:

  1. Estimated mean damage costs.
  2. Overhead for managing the contracts and claims, reinsurance etc.
  3. Profit. (for the shareholders)

Now, to encourage a low-risk behavior a separating factor is search in the vast amount of statistical data. Drivers experience, miles and type of car being the most common, but also other things like oldtimer-status etc. are possible.

If it so happens that the 3-5000 miles driver do only in average have 80% of the damage-costs of a comparable group 5-8000 miles driver, you´ll get the 20% bonus on factor 1.

So the answer is, it is not overpriced, there is just no linear relationship to mileage.

You can´t divide your insureds in too many groups or you´ll miss the mutual aspect of insurance. If everybody just pays his own risk, he can just do so in his bank and save on overhead and profit.


There is plenty of over-rationalisation in the majority of these answers, when the simple answer is that it is simply down to statistics.

Say an insurer had two pieces of information about two separate drivers: annual mileage, and whether they had had an accident in the last 3 years.

Driver A drives 10,000 miles a year and hasn't had an accident in the past 3 years. Driver B drives 500 miles a year and hasn't had an accident in the past 3 years.

Which would the insurer think was the safer bet? The answer is A, and this makes his premiums lower. The reason for this is that the insurer has a lot more data about Driver A than Driver B: they know that Driver A has driven 30,000 miles without having an accident. This could, of course, be luck, or a fluke, but it is likely that Driver A is actually a safe driver. The chance that Driver A hasn't had an accident just through sheer luck and that they are actually a terrible driver is quite slim.

On the other hand, Driver B has only driven 1,500 miles in the past three years. Whilst this seems like prima facie evidence of them being as safe a driver as Driver A, it is much more likely that Driver B could have driven 1,500 miles and avoided an accident through sheer luck, even though they are a terrible driver.

This means drivers who drive low amounts of mileage will be penalised relative to other drivers who have high mileage. It has nothing to do with insurers taking a judgement that 'doing more mileage makes you more experienced' or 'makes you a better driver' as others have suggested here (although, it is probably true - it's not quantifiable from an insurer's perspective).


One reason is because car insurance is mandated. Mandated insurance means the government is forcing people to purchase it, which also means that everyone must have the opportunity to purchase it at a reasonable cost, even if the insurer would normally not choose to insure them. In mandated industries, risk pools are formed which means that as a whole, lower risk members partially subsidize higher risk members. In mandated industries that have a large risk variance, the insurance system would break down if everyone was charged their "fair share" because high risk members would be unable to afford a policy. (This is even more prominent with health insurance than car insurance because the difference in risk is vastly greater.)

On a positive note, perhaps you may get a warm and fuzzy feeling knowing that you are helping out others "in need".

  • 4
    Your statement "the insurance system would break down if everyone was charged their 'fair share' " is incorrect. Insurance systems would work perfectly fine if everyone paid their actual risk share. Imagine for example a scenario where everyone's risk is exactly the same. You can still pool things together. (This transfers the one-off risk of a disastrous event into a steady-payment.) Commented Sep 12, 2017 at 17:55
  • 2
    @user2705196 - You are correct about your scenario where everyone's risk the same (or when the risk variance is low within the pool). I have tweaked that sentence to clarify. However, the statement holds up when there is high risk variance because higher risk members would not be able to afford their insurance, so they can't "pay" their actual risk share.
    – TTT
    Commented Sep 12, 2017 at 18:57
  • @user2705196 Imagine if insurers somehow had perfect models and knew exactly who would get in an accident and how much it would cost. Pricing in such a world would be very interesting.
    – user12515
    Commented Sep 13, 2017 at 16:58
  • 1
    from the perspective of the insurer, no members subsidize any other members. that is precisely why higher risk members are charged more than low risk members. the insurance company expects every single customer to pay in more than they pay out. obviously, the insurer is sometimes wrong because they only have limitted information (e.g. self-reported miles driven), but on average, for a given customer profile, they are nearly always right. Commented Sep 14, 2017 at 20:15
  • @jamesturner - if what you said were true than the cost of health insurance for certain medical conditions would be $1M per year for some people. That is not the case. In auto insurance, they call it "assigned risk".
    – TTT
    Commented Sep 14, 2017 at 20:44

Insurance rates are about assessing risk.

If the insurer has no way to reliably and easily assess usage, they will not reduce the premiums.

Many companies are providing tracking devices that connect to the OBD-II port. This not only tracks actual miles driven, but can typically track aggressive driving, time of day, length of trips, and other information.

Unless you are using this kind of device to give the insurer actionable feedback on your driving habits, do not expect any discounts for mileage or usage.


People who drive long distances tend to do more of their driving on larger, well-built roads (freeways / motorways) that are designed for high-speed driving. Although some people find them intimidating, they are much safer in terms of accidents per kilometre driven for several reasons:

  • No sharp corners or obstructed lines of sight
  • Relatively few junctions
  • No traffic lights
  • Drivers are generally fairly well behaved
  • Relatively few distractions
  • Clear road markings
  • Very smooth, well-maintained road surface
  • Clear signposting
  • And yet all statistics about accident rates always use "accidents per mile" as a metric, not "accidents per trip" or "accidents per year", or "accident per traffic light crossed". Commented Sep 15, 2017 at 13:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .