Sunk Cost Fallacy: Meaning, Examples and the Research Behind It

By the IQ Revealed Team · Updated November 3, 2026

The sunk cost fallacy is the tendency to keep investing in something because of what you have already put into it, rather than because of what it will bring from now on. Money, time and effort that are already spent cannot be recovered whatever you decide next, so on standard economic reasoning they should not count. People count them anyway: they finish films they are not enjoying, stay with failing projects and keep paying into plans they no longer want. Psychologists call it the sunk cost effect; the everyday phrase is throwing good money after bad.

The sunk cost fallacy at a glance

What happens: a past, unrecoverable investment pushes a person to continue, even when continuing is the worse option from here on.
Classic study: Hal Arkes and Catherine Blumer, Ohio University, 1985: the ski-trip question and a real theatre-ticket experiment.
Other names: sunk cost effect, sunk cost trap, and in animal research the Concorde fallacy.
How strong: real across 98 measured effects in a 2015 meta-analysis, but smaller and less reliable in some real-money and field studies than the classic vignettes suggest.
Family: one of the classic cognitive biases, closely linked to cognitive dissonance.

The ski-trip question

Arkes and Blumer's most quoted demonstration, adapted from an example by the economist Richard Thaler, asked college students to imagine they had paid $100 for a weekend ski trip to Michigan and then $50 for a trip to Wisconsin on the same weekend, which they expected to enjoy more. Only one ticket could be used, and neither could be sold or refunded.

If only future enjoyment counted, everyone would go to Wisconsin. In fact, of 61 students, 33 chose the $100 trip and 28 the $50 one: only 46% picked the trip they expected to enjoy more. Both tickets were equally paid for; the bigger loss simply weighed more. The same paper reported that people who had taken economics courses showed the effect just as much, and that people who had incurred a sunk cost rated a project's chances of success higher than people who had not.

The theatre-ticket experiment

Questionnaires involve no real money, so Arkes and Blumer also ran a field test. The first 60 people buying a season ticket at the Ohio University Theater for the 1982–83 season were randomly sold either a normal $15 ticket, a ticket with a $2 discount, or one with a $7 discount. Couples were removed, leaving 18, 19 and 17 buyers. Colour-coded stubs showed who attended which play.

Over the first five plays, people who had paid full price used 4.11 tickets on average, against 3.32 and 3.29 for the two discount groups. Every buyer had the same right to attend, so the difference reflected only what they had paid. Over the last five plays the difference disappeared. The higher sunk cost kept people going for about six months, and then faded.

Why people honour sunk costs

“Don't waste.” Arkes and Blumer argued that the main driver is the wish not to appear wasteful. In their 1999 review, Arkes and Peter Ayton described it as an over-generalised “don't waste” rule, useful in many situations and misapplied in this one.
Losses loom larger. The 1985 paper found the basic pattern well described by prospect theory: abandoning a project turns a paper loss into a certain one, and people tend to take risks to avoid certain losses.
Optimism after investing. People who have already invested rate the venture's chances more highly, which makes continuing feel justified.
Trouble letting go. In 2010 Marijke van Putten, Marcel Zeelenberg and Eric van Dijk found that people who tend to ruminate about past events were especially prone to the effect, while people who let go more easily showed none.
Consistency and reputation. Abandoning a plan can look like admitting a mistake. In the first of three preregistered experiments (1,462 participants in all) published in 2026, Marco Mattei and John Michael found that people rated political leaders more highly for continuing a rail project than for abandoning it, whether or not money had already been sunk into it.

Everyday examples

Films, meals and tickets: watching to the end, eating past full, or going out in bad weather because it is already paid for.
Studies and careers: finishing a course or staying in a job mainly because of the years already spent on it. Our guide to changing careers looks at that decision on its future merits.
Relationships: staying mainly because of the time already invested, rather than because of where things are going.
Business and investment: continuing to fund a project because of past spending. In 2024, Fabian Hogrebe and Eva Lutz analysed 30,602 venture capital decisions and found that the more capital and monitoring effort an investor had already put into a company, the more likely it was to invest again.
Public projects: Arkes and Blumer opened their paper with two United States senators in 1981 arguing that stopping a waterway project would waste the money already spent.

The Concorde fallacy: do animals do it too?

Biologists named the same error after the Concorde airliner. In 1999, Arkes and Ayton argued that there were no unambiguous cases in lower animals, and that young children placed in a similar situation behaved more rationally than adults. That made the effect look like a distinctly human over-generalisation.

A 2018 study in Science complicated the picture. Brian Sweis and colleagues designed matching foraging tasks for mice, rats and humans, and found that all three became harder to stop the longer they had already waited for a reward, but only after they had decided to pursue it. In 2022, Torben Ott and colleagues argued that this pattern can arise from a statistical artefact, a form of attrition bias, and can be reproduced by a fully rational decision maker. The animal question is still open.

How strong is the effect really?

The classic results come mostly from questionnaires. Studies with real stakes give a more mixed picture:

A meta-analysis. In 2015 Stefan Roth, Thomas Robbert and Lennart Straus at the University of Kaiserslautern pooled 98 effect sizes. The effect was clearly real, but its size depended on the type of decision, and in decisions about using something already paid for it weakened with time, as in the theatre experiment.
A treasure-hunt game. In 2007 Daniel Friedman and colleagues varied eleven features of a computer game with real payments and found a surprisingly small sunk cost effect.
Tuition fees. In an Amsterdam field experiment published in 2016, students signing up for extra tutorials were randomly given fee discounts. Paying more did not increase attendance or performance for the sample as a whole, only for students identified in advance as sunk-cost prone.
A reverse effect. In the 2022 investment experiment by Negrini, Riedl and Wibral, larger initial investments made people less likely to continue with real money, while the same people showed the standard bias in hypothetical scenarios.

The honest summary: people do let sunk costs sway them, especially in imagined choices and shortly after paying, but the effect is smaller and less universal in real decisions than its fame suggests.

Does a high IQ protect you?

Apparently not. In seven studies published in 2008, Keith Stanovich and Richard West found that the sunk-cost effect was among the classic thinking biases largely unrelated to cognitive ability, along with framing, anchoring and myside bias. Formal training did not help either: Arkes and Blumer found the effect was not reduced in students who had taken economics. Age seems to matter more than ability. JoNell Strough and colleagues asked in 2008 whether older adults are less subject to the fallacy, and the 2015 meta-analysis concluded that they are less likely to fall for it than younger adults. Our pages on critical thinking and deductive and inductive reasoning explain which reasoning errors intelligence does and does not predict.

Sunk costs, dissonance and confirmation

Cognitive dissonance explains part of the pull: admitting that a big investment was a mistake clashes with seeing yourself as a good decision maker.
Confirmation bias keeps the project alive: after investing, people look for evidence that it will still work out.
Escalation of commitment is the organisational version: decision makers keep adding resources to a failing course of action they chose themselves.

How to avoid the sunk cost fallacy

Ask the fresh-start question. If I were starting today, with what I know now, would I choose this?
Set exit criteria in advance. Decide before you start what result, cost or date would make you stop, and write it down.
Name the opportunity cost. What else could this time or money achieve from here?
Separate the people. Let someone who did not make the original decision review whether to continue.
Do not overcorrect. Quitting simply because a lot has been spent is the reverse mistake. Judge the future on its merits in both directions.

Trust and scope notes

This page summarises published research for general information, with figures taken from the studies listed below. It is not financial, career or relationship advice. The IQ Revealed test is an online reasoning test of 40 original items for adults aged 18 and over; it does not measure the sunk cost fallacy or any other thinking bias. Our guide to how IQ tests work explains what an online test can and cannot tell you.

IQ Revealed is not affiliated with, endorsed by or connected to the researchers, universities, journals or publishers named on this page. Names are used only to identify the research being described.

Frequently asked questions

What is the sunk cost fallacy?＋

The sunk cost fallacy is the tendency to keep going with something because of the money, time or effort you have already put into it, rather than because of what continuing will bring from now on. The past investment is gone whatever you decide, so on standard economic reasoning it should not count. Psychologists Hal Arkes and Catherine Blumer described it in 1985 as a greater tendency to continue an endeavour once an investment has been made.

What is an example of the sunk cost fallacy?＋

In Arkes and Blumer's best-known question, people imagined holding tickets for two ski trips on the same weekend: a $100 trip to Michigan and a $50 trip to Wisconsin they expected to enjoy more. Only one could be used and neither could be sold. Of 61 students, 33 chose the more expensive Michigan trip, so only 46% picked the trip they expected to enjoy more. Everyday versions include sitting through a bad film because the ticket was paid for, or finishing a degree you no longer want because of the years already spent on it.

What is the difference between a sunk cost and an opportunity cost?＋

A sunk cost has already been paid and cannot be recovered, so it is the same whichever option you choose. An opportunity cost is the value of the best alternative you give up by choosing one option, so it lies in the future and should count. The sunk cost fallacy often means overweighting the first and ignoring the second: staying in a failing project because of what it has cost so far, and not asking what the same time or money could achieve elsewhere.

What is the Concorde fallacy?＋

The Concorde fallacy is another name for the sunk cost effect, taken from the Anglo-French supersonic airliner. Biologists used the term for animals, psychologists the term sunk cost effect for people. In a 1999 review, Arkes and Peter Ayton argued that there were no unambiguous cases in lower animals and that young children behaved more rationally than adults in similar situations, although a 2018 study in Science reported a comparable sensitivity to time already invested in mice, rats and humans.

What is the sunk cost fallacy in relationships?＋

It is staying in a relationship mainly because of the years, effort or shared plans already invested in it, rather than because of how it is likely to go from here. The reasoning is the same as with money: the past years are spent either way, and the question that matters is whether the future of the relationship is worth having. Past investment can be part of an honest reason to keep working at something, but on its own it is not a reason to stay.

Do smarter people fall for the sunk cost fallacy less?＋

Not measurably. In seven studies published in 2008, Keith Stanovich and Richard West found that the sunk-cost effect was among a group of classic thinking biases largely unrelated to cognitive ability. Arkes and Blumer had already found in 1985 that people who had taken economics courses showed the effect just as much. Age seems to matter more than ability: a 2015 meta-analysis concluded that older adults are less prone to it than younger adults.

What is the opposite of the sunk cost fallacy?＋

Researchers sometimes find a reverse sunk cost effect, in which a larger past investment makes people less likely to continue. In a 2022 experiment with real money, Marcello Negrini, Arno Riedl and Matthias Wibral found exactly that in an investment task, even though the same participants showed the standard bias in hypothetical scenarios. Giving up too early to avoid any further loss is the mirror-image mistake; the goal is to judge the future on its merits in both directions.

How do you avoid the sunk cost fallacy?＋

Ask the forward-looking question: if I were starting from today, with what I now know, would I choose this? Set exit criteria in advance, before money or effort pile up, and write them down. Let someone who did not make the original decision review whether to continue. And name the opportunity cost out loud: what else could this time or money achieve? Knowing about the bias is a start but not a cure; procedures that force the forward-looking comparison work better than good intentions.