The cost nobody sends you a bill for
Ask someone what their car costs to run and they will quote the fuel. It is the number they see, at the moment they pay it, several times a month.
The full sum has three parts:
what it loses in value (depreciation)
what it costs to keep (insurance, tax, servicing, tyres)
what it costs to move (fuel)
For most drivers of most cars, the first is the largest — often larger than the other two combined. And it is the only one that never arrives as an invoice. Nobody charges you for depreciation; you simply have less than you had, discovered years later when you sell.
A car bought at 30,000 and sold five years on for 15,000 has cost 3,000 a year in value alone. At 10,000 miles a year that is 30 cents a mile before a drop of fuel, before insurance, before a service. Fuel at 32 mpg and 3.40 a gallon adds about 11 cents. The invisible cost is nearly three times the visible one.
This is why the official reimbursement rates look so high. Both the IRS standard mileage rate and the HMRC approved rates are set to cover the whole cost of running a vehicle — depreciation included — not the fuel. People who assume the rate is a fuel figure conclude they are being paid generously, and they are being paid roughly correctly.
Average against marginal: two different questions
The tool reports two per-mile numbers, and confusing them leads to bad decisions in both directions.
Average cost per mile is everything divided by the distance driven. It answers: what does having this car cost me, per mile of use?
Marginal cost per mile is what one more mile adds. Mostly fuel, plus a share of wear that is real but small and hard to pin down. It answers: what does this journey cost?
The distinction matters because the insurance is paid whether the car moves or not, and the depreciation is largely a function of age and market rather than of the specific trip. So:
Deciding whether to drive somewhere — use the marginal figure. Comparing a train fare against the full average cost per mile overstates the driving cost, because most of that average is being spent regardless.
Deciding whether to own the car at all — use the average. Here the standing costs are exactly the thing in question.
Deciding between two cars — use the average, on each of them, over the same distance. The comparison lives almost entirely in depreciation and insurance rather than in fuel economy, which is where most buyers concentrate.
There is a common error in each direction. People compare a train ticket against the total cost of ownership and conclude driving is expensive; and people compare a new car's fuel savings against an old car's fuel bill and conclude replacing it pays, having quietly ignored that the new car will shed several thousand a year in value.
Depreciation is not a straight line
The tool divides the loss evenly across the years you keep the car, which is fine for planning and is not what actually happens.
The real curve is steepest at the start. A large share of the total loss happens in the first year, and a substantial part of that in the first weeks, when the car stops being new. From there the annual loss shrinks in absolute terms, and by year eight or ten the value moves slowly.
Three consequences:
Buying at two or three years old transfers the steepest part of the curve to someone else. The car is mechanically almost identical and has already lost the money.
Keeping a car longer spreads a loss you have already taken. Depreciation per year falls every year you hold on, which is why a well-maintained old car is usually cheaper to run than its repair bills suggest.
Mileage moves the curve. Value falls with distance as well as age, and the used market has thresholds — crossing a round number moves a price more than the mile justifies. Drive a lot and a car that has already flattened out costs you less to use.
If precision matters, look up what your specific model actually sold for at the age you plan to sell, rather than trusting a rule of thumb. It is the largest term in the calculation, so the estimate deserves the most attention and usually gets the least.
What the standing costs really include
Insurance. Annual, largely independent of distance, and the second biggest fixed item for many drivers. Worth re-shopping every renewal — auto-renewal pricing rarely favours the customer.
Road tax or registration. Fixed and usually easy to look up exactly.
Servicing, tyres and repairs. The line that lies. Entering an average is right for planning and wrong for any particular year: an older car does not cost a smooth 600 a year, it costs 150 twice and then 1,400 once. Tyres are the same shape — nothing for two years, then a set.
If a car is older than about eight years, treat the servicing figure as a savings target rather than a forecast, and revise it upward when a big item lands.
Missing from most calculations, including this one unless you add them: parking and permits, tolls, breakdown cover, cleaning, and finance interest. If the car was bought on credit, the interest is a real cost of that car. If it was bought outright, the capital is no longer earning anything, which is a smaller and equally real cost.
Where electric changes the sums
Electric cars redistribute the three terms rather than removing any.
Energy per mile is usually much cheaper, and the size of the saving depends almost entirely on where you charge. A home overnight tariff and a motorway rapid charger can differ by a factor of three or four per unit, which means an owner who mostly charges at home and one who mostly does not have completely different running costs on the same car.
Servicing is lighter. No oil changes, far fewer moving parts, and brakes that last longer thanks to regenerative braking. Tyres often wear faster, from the extra weight and torque.
Purchase prices are higher and depreciation has been less predictable. This is the term that dominates the total, and it has moved sharply as battery prices, subsidies and second-hand demand have changed. It is the number to research rather than assume.
The method does not change. Put your own charging mix and your own realistic resale figure in, and the answer will be specific to you in a way no published average can be.
A sanity check on the result
Two quick tests once you have a figure.
Compare it against the official rate. Your own average cost per mile, set beside the current IRS or HMRC rate, tells you something immediately useful about business mileage — well below it and reimbursement is generous, well above and it is not. Look up the current rate rather than a remembered one; they are revised.
Compare it against your actual spending. Add up a year of bank statements for anything car-related, add the value the car has lost, and see whether it lands near the total the calculator gives. Where the two disagree, the calculator is usually missing a cost rather than the statements being wrong — and finding out which one is the entire value of doing this.