Skip to content
mentua.org

What Is the Sunk Cost Fallacy?

Line drawing of a path that forks at a point: one solid branch climbs to a filled circle, the other dotted branch bends down to a line under which three dotted coins have sunkLine drawing of a path that forks at a point: one solid branch climbs to a filled circle, the other dotted branch bends down to a line under which three dotted coins have sunk

6 min read≈ 1,230 words3 questions · up to 3 points

What this is about

You have paid for two ski trips on the same weekend and can only take one: the cheaper trip you would enjoy more, or the expensive one. You learn why most people in a classic study chose the expensive trip, what a theater and a supersonic airliner show about the same error, and why some animals seem to make it while young children often do not. You also learn when looking at the past is sensible.

In a nutshell· skip ahead

The sunk cost fallacy is the mistake of continuing something because of what you have already spent on it, in money, time or effort, although that spending cannot be recovered. A rational choice compares only what each option will cost and bring from now on, because the sunk cost is lost whichever option you choose.

Imagine that you have paid $100 for a weekend ski trip to Michigan, and a few weeks later you buy a $50 trip to Wisconsin, which you expect to enjoy more. Then you notice that both trips are on the same weekend, and neither ticket can be sold or returned. In 1985 the psychologists Hal Arkes and Catherine Blumer gave this problem to 61 college students and asked which trip they would take.

What is the sunk cost fallacy?

The sunk cost fallacy is the mistake of continuing something because of what you have already spent on it, although that spending cannot be recovered. A sunk cost is any money, time or effort that is gone whatever you decide next. A good choice compares only the future costs and benefits of each option, and the ski trip shows why in four steps.

  1. What is already spent? Both tickets cost $150 in total, and that money is gone whichever trip you take.
  2. What still differs between the options? Only the weekend itself, and you expect to enjoy Wisconsin more.
  3. What follows? You should go to Wisconsin, because the $100 ticket is lost either way and cannot make Michigan any better.
  4. What did people choose? 33 of the 61 students, or 54 percent, chose the Michigan trip they expected to enjoy less.

To test real spending, Arkes and Blumer also sold season tickets for the Ohio University Theater and, at random, gave some buyers a normal $15 ticket and others a discount of $2 or $7. Over the first five plays, those who had paid full price used 4.11 tickets on average, while the two discount groups used about 3.3. Six to nine months after the purchase, the difference had disappeared.

Arkes and Blumer’s explanation was that people do not want to appear wasteful, and an unused ticket feels like waste. In 1976 the organizational psychologist Barry Staw had described a related pattern, escalation of commitment, which means putting more resources into a failing course of action. In his role-playing experiment, 240 business students invested the most when they had chosen a course of action themselves and it had gone badly, which fits the self-justification described in What Is Cognitive Dissonance?.

Do animals fall for it too?

In biology the fallacy is named after a supersonic airliner. In 1962 Britain and France signed a treaty to build Concorde. By the end of 1975 they had spent a little more than £1 billion on it, although nearly all the airlines that had once wanted the plane had dropped their options. In 1976 the biologists Richard Dawkins and T. R. Carlisle found the same error in theories about animals, and it became known as the Concorde fallacy.

Dawkins and the biologist Jane Brockmann then seemed to find it in nature. Female great golden digger wasps sometimes end up sharing a burrow and fight over it, and the burrow is worth more the more paralysed katydids, a kind of insect, it holds as food for the young. In 23 observed fights, the winner tended to be the wasp that had brought more katydids herself, as if she were fighting for her own investment rather than for the value of the burrow.

It amounts to investing further in a project simply because one has invested in it heavily in the past

Richard Dawkins and H. Jane Brockmann, “Do Digger Wasps Commit the Concorde Fallacy?”, Animal Behaviour (1980)

A wasp probably cannot count the katydids in a burrow, but her own contribution is a fair guide to their number, so Dawkins and Brockmann concluded that fighting for it may be a sensible rule of thumb. In 1999 Arkes and the psychologist Peter Ayton reviewed the animal studies and found no clear case of the fallacy in any animal. They also reported that young children, in a similar choice, decide more in line with economic theory than adults do.

What the sunk cost fallacy teaches you

When you hesitate to stop something, take away what you have already spent and ask: “If I had not paid anything yet, would I choose this today?” If the answer is no, the money or time behind you is not a reason to continue.

In 1995 Staw and Ha Hoang found that NBA basketball players picked early in the draft got more playing time than their performance justified. When the economists Colin Camerer and Roberto Weber re-examined the data in 1999, the effect was still there but weaker.

When is looking back not a fallacy?

The most common misunderstanding is that the sunk cost fallacy means “quit when things get hard”. The rule only says that costs you cannot recover should not decide, so continuing is right whenever the future benefits are larger than the future costs.

The strongest objection is that past spending often comes with future effects, and those do count. In 2010 the economists Preston McAfee, Hugo Mialon and Sue Mialon gave three kinds of reasons. Past spending can carry information, for example about how good a project is. Quitting can harm your reputation as someone who finishes what they start, and limits on money and time can make it sensible to stay. Penalties for quitting are future costs as well, and Concorde had one, since the 1962 treaty had no clause that allowed either country to leave it.

The evidence is mixed. In Many Labs 1, a 2014 project that repeated 13 findings in 36 samples, a question about a paid or free ticket to a game in freezing weather showed the effect again, although it was small. Effects with real stakes are less certain, because the theater effect faded after six months and a 2021 replication of Staw’s experiment with 365 participants found no escalation.

Learn more

The sunk cost fallacy has no single originator, because economists had long taught that only future costs and benefits should count. In 1980 the economist Richard Thaler described a family who drive through a snowstorm to a basketball game because they paid $40 for the tickets, although they would have stayed home with free ones. Arkes and Blumer named the sunk cost effect and tested it in 1985, after Staw had studied escalation of commitment and Dawkins and Carlisle had carried the idea into biology.

To judge a past decision by what you knew at the time, not by its result, read A Good Decision Can End Badly: How to Judge Choices When Luck Is Involved.

Quiz

3 questions · up to 3 points

Open a question and pick one answer. Only your first answer counts: +1 if it is right, −1 if it is wrong.

  1. In Arkes and Blumer's ski-trip question, both tickets are already paid and neither can be sold. Why should everyone choose the $50 trip to Wisconsin?
    Show answer

    The $150 is spent whichever trip you take, so the only difference left is that Wisconsin is more enjoyable. Both payments are sunk costs. Nothing you do on the weekend brings them back, so they cannot favour either trip.

  2. Lena has paid €400 for an online course that cannot be refunded. After six weeks she finds it useless, and a free course would teach her more. Which question helps her avoid the sunk cost fallacy?
    Show answer

    If I had not paid anything yet, which course would I choose today? The question removes the €400, which is gone in both cases, and leaves only what each course will cost and bring from now on.

  3. Which statement about the sunk cost fallacy matches the article?
    Show answer

    It is not a rule to quit, and past spending can matter when it carries information or makes quitting costly. The rule only says that unrecoverable costs should not decide. Information, reputation and penalties for quitting are future effects, so they belong in the decision.

The takeaway

The sunk cost fallacy is continuing a plan because of money, time or effort already spent that cannot be recovered, although only future costs and benefits should decide. In a 1985 study, 54 percent of students chose a $100 ski trip over a $50 trip they would enjoy more, and theatergoers who paid full price went to more plays. Past spending can matter when it carries information, protects a reputation or comes with costs of quitting. Before you continue, ask whether you would start this today.

Key quotes

“To repeat: sunk costs are irrelevant to current decisions.”
Hal R. Arkes and Catherine Blumer, "The Psychology of Sunk Cost", Organizational Behavior and Human Decision Processes 35 (1985), p. 126

References

  1. The Psychology of Sunk Cost. Hal R. Arkes and Catherine Blumer, Organizational Behavior and Human Decision Processes 35, 124–140, 1985

    The original paper with the ski trip, the theater season tickets and the radar-blank plane, and the explanation that people do not want to appear wasteful.

  2. Knee-Deep in the Big Muddy: A Study of Escalating Commitment to a Chosen Course of Action. Barry M. Staw, Organizational Behavior and Human Performance 16, 27–44, 1976

    The role-playing experiment with 240 business students that started the research on escalation of commitment.

  3. Do Digger Wasps Commit the Concorde Fallacy?. Richard Dawkins and H. Jane Brockmann, Animal Behaviour 28, 892–896, 1980

    A short and readable paper on 23 fights between wasps over shared burrows, and why the apparent error may be a sensible rule of thumb.

  4. Investigating Variation in Replicability: A "Many Labs" Replication Project. Richard A. Klein, Kate A. Ratliff, Michelangelo Vianello and others, Social Psychology 45 (3), 142–152, 2014

    The replication of 13 effects in 36 samples, including the question about a paid or free ticket to a game in freezing weather.

  5. Do Sunk Costs Matter?. R. Preston McAfee, Hugo M. Mialon and Sue H. Mialon, Economic Inquiry 48 (2), 323–336, 2010

    Three economists argue that reacting to sunk costs is often rational because of information, reputation or limits on money and time.

sunk-cost-fallacysunk-cost-effectsunk-costsconcorde-fallacyescalation-of-commitmentbehavioral-economicsdecision-makingcognitive-biaseshal-arkesrichard-thalerbarry-stawrichard-dawkinsanimal-behaviorreplicationpsychology-experiments