What a kilometre really costs

A monthly figure compares with nothing. It answers whether the car fits the budget, and there it stops. Every further question — train or car, one vehicle or two, a longer commute for cheaper rent — is a question about distance, and distance is counted in kilometres.

The second reason is that the unit makes the alternatives visible. A rail ticket, an hour of car hire, a car-sharing tariff and a mileage allowance are all priced by distance or time. Knowing your own cost per kilometre lets you lay them side by side instead of setting one intuition against another.

And the third: the number behaves differently from how people expect. It is not a property of the car but of the car together with how it is used. The same car costs noticeably less per kilometre at twice the annual distance — not slightly less, but in a different order of magnitude if the starting mileage was low.

The whole calculation hangs on a single distinction: which costs arise because the car exists, and which arise because it moves? The first group does not change if you make one trip more or fewer. The second does.

The allocation is not always obvious, and where it is not, a simple test helps: what would this item cost if the car stood still for a year? Insurance would keep running, so would tax and the parking space. Fuel would not, tyres barely, brakes almost not at all. Servicing sits in between, because it usually falls due by time or by distance, whichever comes first.

The average is the simplest figure: all the costs of a year divided by the kilometres driven in that year. Fixed costs plus variable costs, divided by distance. It answers what driving costs you in total, converted into a unit you can go on calculating with.

Its shape is a hyperbola, and that is the interesting part. Because the fixed costs stay constant in the numerator while the kilometres grow in the denominator, the value falls steeply while the distance is small and then flattens out. At very low annual mileage the fixed costs dominate so heavily that each kilometre becomes absurdly expensive — and that is exactly the calculation a rarely used second car presents.

Which yields a rule of thumb you can calculate rather than take on faith: below a certain annual mileage, any car of your own is more expensive than any form of hire, however cheap it was to buy. Where that threshold lies depends solely on the size of your fixed costs — and you know those.

Alongside the average sits a second figure almost nobody calculates: the cost of the next kilometre. Only the variable items go into it, because the fixed ones arise anyway. That figure is considerably smaller than the average, and the gap widens the less you drive.

The two figures answer different questions, and confusing them leads reliably to bad decisions. The average answers: should I keep this car at all? The marginal figure answers: should I make this trip in the car that is already sitting outside? Costing a weekend trip at the average makes the car look more expensive than the trip makes it.

The same mistake runs the other way too. Anyone weighing up a second car while reasoning in marginal costs badly understates the decision — an extra car brings a complete new set of fixed costs, and those arrive before the first kilometre is driven.

The largest item in both columns is depreciation, and it belongs in both. Part of it comes with the calendar: a car that stands unmoved for a year is worth less afterwards than before. Another part comes with the odometer: two identical cars of the same year with very different mileages fetch different prices.

The split can be read off the market without any theory. Search for the same model of the same year and compare prices across different mileages — that gives you the distance-related part. Then compare the same mileage across different years — that gives you the age-related part. Together the two searches take half an hour.

The result explains something often noticed: on a young, expensive car the age-related loss is so large that it swamps everything else — fuel and servicing barely register beside it. On an older car bought cheaply it has almost vanished, and then consumption, tyres and repairs decide the sum. The same question, two entirely different answers.

Your own figure is worth more than any table, because it contains your car, your routes and your prices. Producing it needs no software, just a folder and two odometer readings: one at the start of the period and one at the end.

If you do not want to wait a year, work backwards. The receipts for the last twelve months already exist: the insurance policy, the tax bill, fuel or charging statements, workshop invoices, a tyre purchase. The odometer reading from a year ago is on the last service record or the inspection report. That gets the calculation done in an hour.

Two corrections make the result honest. First: larger items that last several years — a set of tyres, a clutch, a battery — get spread over their life rather than charged to the year the invoice arrived. Second: depreciation belongs in it, even though it never arrives as a bill. Leave it out and in most cases the largest item is missing.

Once both values exist, questions that were previously matters of taste become answerable. The calculation is always the same: what changes in the fixed costs, what changes in the variable ones, and how many kilometres does it take for the difference to even out?

It is clearest when switching to a more economical car. The benefit sits in the variable costs and arrives per kilometre; the cost sits in the purchase price and arrives once. Dividing one by the other gives the number of kilometres at which the switch pays for itself — and at low annual mileage that number often lies beyond the years you planned to keep it.

Every metric has an edge, and this one has three. The first is deferred maintenance: a year with no workshop bill is not a cheap year but often a postponed one. Leave out the reserve for the next major job and you get a figure that holds exactly until the next invoice.

The second is standstill. Fixed costs keep running in months when the car barely moves — after a move, through an illness, across a long winter. Building the average from an atypical year produces a value that never recurs. Two or three years side by side say more than one.

The third is precision itself. A per-kilometre figure quoted to two decimal places suggests an accuracy the inputs do not support — depreciation is an estimate and so is the future fuel price. What is useful is the order of magnitude, not the digit after it. For every decision this piece describes, that is entirely enough.

The calculation is worth most before the purchase, because at that point every item is still negotiable. For a car you are considering, all the inputs can be obtained: consumption from reported experience, insurance from a non-binding quote, service intervals and prices from a workshop, tyre size from the documents, and expected depreciation from comparing prices of the same model in older years.

That preparation reliably separates two cars that look equally expensive in a listing. Two cars at the same price can differ over five years by an amount in the order of another car — and the difference arises not at the purchase but in the columns this piece describes.

Carvexia shows listings and prices across the European markets; the calculation is yours to make, with your kilometres and your tariffs. Insurance, workshop prices and finance are agreed with the respective providers, and what governs is what their quotes say.