What a used Fit costs

What each model year costs

Three numbers per year: the cheap end, the middle and the dear end. The gap between them is the useful part — it tells you whether the year has a clear market price or whether condition decides everything.

A spread near 1.5× means the market agrees what the car is worth. Past 3× it usually means the sample mixes tidy examples with tired ones — and that a careful buyer can do well.

Dealer or private seller

Both figures are measured against the median for the same model year, so age is taken out of the comparison — this is the forecourt markup on its own.

The gap comes to 18.7 percentage points. Some of that buys something real — reconditioning, a warranty, paperwork handled. The rest is the cost of not looking further.

What a branded title costs

A salvage or rebuilt title follows a car for life and shows up in every future sale. Here is what the market actually knocks off for it — again against the median of the same year.

⚠ A discount is not automatically a bargain: branded titles are harder to insure, harder to finance and much harder to sell on. Check the specific car before assuming the saving is real — run its VIN against the records.

How fast price falls with mileage

Taken from a single model year — 2013, the one with the most listings — so the curve shows mileage alone rather than age creeping in alongside it.

Between the lowest and highest mileage bands the asking price falls 63%. Most of that drop happens early: the first hundred thousand miles cost far more than the second.

Which options hold value

Same normalisation: each figure is the average premium or discount against the median of its own model year.

Where it's cheaper

States where this car asks least and most, measured against the national median for each model year. Worth knowing before a long drive: a fifteen-percent gap on a twenty-thousand-dollar car is three thousand.

Only states with enough listings to mean something are shown. Rust-belt winters, local demand and how many were sold there originally all pull these numbers around.

How we count

Every figure comes from live classified listings — asking prices, not sale prices. Sellers ask more than they get, so treat these as the top of the range, not the bottom.

Why the percentages, and not plain averages. This is the part worth explaining, because getting it wrong produces confident nonsense. Take a naive average price per state and you learn that this car costs four times more in Alabama than Indiana — which is false. What actually happened is that Indiana's listings were older cars. Do the same by title and you get the absurd result that rebuilt cars sell for more than clean ones, because the rebuilt ones in the sample happened to be newer.

So we never average prices across a slice. Each listing is first converted into its ratio against the median for its own model year, and those ratios are averaged instead. «Fifteen percent below median» then means what it says, and holds up whichever years the slice happens to contain.

Slices with too few listings are dropped rather than shown with a wide error bar — a percentage built on eight cars is decoration, not data.