What Is the Sunk Cost Fallacy? How 'But I Already Paid' Makes Losses Grow

2026-07-26

Direct answer: a sunk cost is anything already spent that no future action can recover; the sunk cost fallacy is letting that unrecoverable past steer the choices still open — when rational decisions should weigh only future costs and benefits. Obvious on paper, and yet it may be daily life’s most expensive thinking error, because it specializes in advising you to double down exactly when you are already losing.

The core idea: paid money has no vote on the next step

Thirty minutes into a movie, verdict certain: it’s bad. The $15 ticket is paid and non-refundable. The live options: (A) sit through 90 more minutes, (B) leave and reclaim them. Notice: the $15 is gone under both options — it is sunk, and mathematically irrelevant to the choice. The only live variable is the 90 minutes. Yet most people pick A, “because the ticket’s paid” — spending 90 real minutes to honor 15 unrecoverable dollars. That is the fallacy’s complete anatomy: because we lost, we add more.

Four everyday scenes

“Getting my money’s worth” at the buffet. The $40 entry is paid; every plate from here costs zero. The correct way to “break even” is eating until satisfied, not until stuffed — each forced plate isn’t reclaiming the ticket, it’s paying the sunk $40 a second time in discomfort.
The unused membership. “The annual fee is paid, wasting it would be a shame” — the fee is sunk; the live question is whether each future visit is worth your future time. Regret retains no value; it only retains the charge. The audit method lives in subscription creep.
“If I sell now the loss becomes real.” The paper loss is already real. Whether to hold should be decided by “where is this money best placed today,” independent of your entry price. This is the fallacy’s most expensive arena — and this paragraph explains a bias only; it is not investment advice.
The two-year course, the half-done project. “I’ve put in so much already” — the two years are sunk. Zero-reset test: choosing fresh today, would you start this? Yes → continue (the reason lives in the future). No → the fastest stop-loss is now: year three isn’t sunk yet, and it is the only thing still savable.

The ten-second tool: the zero-reset test

Facing any “but I already…” hesitation, ask: “If I were meeting this choice for the first time — no money paid, no time invested — would I start it today?” Yes → carry on; your reasons point forward. No → the only thing keeping you is the past, and the past issues no refunds. The test’s power is physical: it deletes the sunk term from the equation, leaving only future terms — which is where opportunity cost takes over: every day you stay is a day of the next-best alternative foregone.

Why knowing isn’t immunity: the three pushers

Loss aversion: leaving converts the paper loss into an acknowledged one — acute pain; staying lets you keep pretending it hasn’t happened.
Self-justification: quitting indicts your past self’s judgment; continuing protects the past self — funded by the future one.
Waste aversion: we were raised not to waste. But the real waste isn’t walking out — it’s appending good time to a bad choice. The reversal worth memorizing: a stop-loss isn’t waste completed; it’s waste stopped.

Frequently asked questions

What is the definition of the sunk cost fallacy?

A sunk cost is anything already spent that no choice can recover — money, time, effort. The fallacy is letting those unrecoverable outlays steer what you do next, like finishing a bad movie "because the ticket is paid." Rational decisions weigh only future costs and benefits; the sunk portion is mathematically irrelevant to every remaining option.

Why do people fall for it?

Three drivers: loss aversion (admitting the loss hurts more than continuing), self-justification (quitting means the past you chose wrong), and waste aversion (abandoning something paid-for feels sinful). Knowing the drivers doesn't grant immunity, but it lets you name which one is pushing you in the moment.

How do I quickly test whether I'm stuck in a sunk cost?

The zero-reset test: imagine you are meeting this choice for the first time — no ticket bought, no hours invested. Would you start it today? Yes → continuing is rational. No → the only thing holding you is the past, which offers no refunds. Ten seconds; works on subscriptions, courses, projects, and possessions.

How is this different from healthy persistence?

Check which direction the reasons point. Persistence cites the future ("three more months earns the license" — a concrete coming benefit); sunk cost cites the past ("I've already put in two years"). Cross out every past-tense reason and see whether the remaining future-tense ones still carry the decision.

Are there exceptions where sunk costs should count?

Strictly none — spent money stays spent. Two near-exceptions: reputational commitment (quitting mid-promise damages relationships — but that's a future cost, not a sunk one), and deliberately using prepayment to bind yourself to the gym or the course — weaponizing the fallacy as a commitment device, which is fine when you know that's what you're doing.

This article is general behavioral-economics education, not financial or investment advice; investment passages describe a cognitive bias only — consult licensed professionals for actual decisions. All calculators on this site run locally in your browser; nothing you enter is uploaded to any server.