Where it sits
The only question that isn't contaminated: would I choose this today, arriving fresh?
Tell it apart
| Aspect | Sunk cost fallacy | Loss aversion |
|---|---|---|
| Is | A behaviour | A feeling |
| Looks | Backward | At the reference point |
| Produces | Continued spending | Reluctance to realise |
| Named | 1985 | 1979 |
Loss aversion is why quitting hurts — you feel the write-off as a loss. The sunk cost fallacy is what you then do about it: keep spending to postpone booking that loss. One is the engine, the other is the behaviour it drives.
What it doesn't mean
It's only the fallacy when the past cost is unrecoverable AND the forward-looking case says stop. Finishing a course because the remaining classes are genuinely worth your time is just a good decision.
Money is only the easiest version to measure. Time, effort and public commitment produce the same pull — and Arkes and Blumer traced it to not wanting to appear wasteful rather than to the arithmetic itself.
What it really means
Arkes and Blumer's 1985 study is elegant because the participants never knew they were in it. People buying a season ticket to a university theatre were handed, at random, either the full $15 price, a $13 price, or an $8 one. Same seats, same plays, same season — the only difference was a discount nobody had asked for and nobody could have earned.
Over the first half of the season, the people who had paid full price showed up more often. Not because they enjoyed theatre more; they'd been assigned their price by chance. Because $15 was a larger amount of money to have thrown away, and going to the play was the only way to stop it being thrown away.
The part worth being careful about is that not every persistence is this fallacy. Past spending sometimes carries genuine information about the future — a car with £2,000 of recent repairs really may be more reliable now. The test isn't whether you spent; it's whether the spending is doing any forward-looking work in your reasoning. When the entire argument for continuing is the size of the bill so far, you're not making a decision. You're paying an old one interest.
Where it comes from
an accountant's term — a cost already gone under
Economics says a rational decision-maker treats sunk costs as if they don't exist. The fallacy is that we can't.
- 1985
Hal Arkes and Catherine Blumer publish 'The Psychology of Sunk Cost'. Buyers of a university theatre season ticket were randomly given the full $15 price, $13, or $8. Over the first half of the season, full-price buyers attended significantly more plays — identical tickets, identical shows, different amounts already spent.
- 2016
Fujimaki and Sakagami report a sunk-cost-like pattern in pigeons, but only in birds with a history of high-cost work paying off — suggesting learned experience, not a uniquely human accounting error.
In real life
- You've paid £60 for a concert ticket, you're ill and it's pouring, and you go anyway because otherwise the money's wasted. The money is gone either way; the only choice left is whether to spend the evening miserable.
- Two hundred pages into a novel you've disliked since page twenty, you keep reading because you've come this far.
- Having spent £3,000 repairing an old car, the owner pays for a fourth big repair rather than throw away the first three, though a replacement would cost less.
- The Concorde fallacy is another name for it: the British and French governments kept funding the supersonic airliner long after it was clear it would not pay its way.
- Not the fallacy: you've paid for a course, it's turning out to be useful, and you finish it. The past cost isn't what's keeping you there.
Questions
Q01What is the sunk cost fallacy?
Continuing something because of what it already cost. You're forty minutes into a bad film at the cinema; the ticket money is gone whether you stay or leave, but you sit through another ninety minutes 'because I paid for it' — and lose the evening on top of the ticket.
Q02Is it always irrational to consider what you've already spent?
No, and this is the part most explanations skip. Past spending can be real information — a car you've already put £2,000 of repairs into may genuinely be mechanically sound now, which is a fact about the future. The fallacy is when the size of the bill is the *only* argument for continuing.
Q03How do you escape a sunk cost trap?
Ask the fresh-start question: if I were arriving today, with none of this spent, would I choose this? For a two-year side project with no users, if the honest answer is no, the two years are gone either way — and only the next two are still yours.
Sources
- The feeling of throwing good money after bad (PLOS ONE, 2019) — PMC.
- Experience that Much Work Produces Many Reinforcers Makes the Sunk Cost Fallacy in Pigeons (2016) — PMC.