Sunk Cost Fallacy: What It Is, How to Avoid It

The sunk cost fallacy is the habit of sticking with a plan because of the money, time or effort you’ve already put into it, even when walking away would now serve you better. Psychologists Hal Arkes and Catherine Blumer called it the sunk cost effect and defined it as “a greater tendency to continue an endeavor once an investment in money, effort, or time has been made” (Arkes & Blumer, 1985). Those past costs are gone whatever you decide next, so the only thing that should matter is what each option will cost and bring you from here on (Roth, Robbert & Straus, 2014).

Why economists say sunk costs should be ignored

A sunk cost is money or effort that has already been spent and can’t be recovered. According to microeconomic theory, people should base their decisions only on current and future benefits and costs (Roth, Robbert & Straus, 2014). Arkes and Peter Ayton put the same rule this way: a prior investment shouldn’t influence how you weigh your current options, and only the extra costs and benefits of each option should count (Arkes & Ayton, 1999).

Say you paid for a concert ticket and wake up on the day with a bad cold. The ticket price is spent whether you go or stay home, so the real question is whether a night out feeling miserable is worth more to you than a night in bed. Plenty of people still drag themselves out “so the ticket doesn’t go to waste.” That’s an illustration. Researchers have found the same pull in controlled studies.

The classic studies

Arkes and Blumer’s 1985 paper is the standard starting point. In one scenario, people imagined they had already paid for two ski trips, one more expensive than the other, and then found out both fell on the same weekend. Even though the scenario set up a preference for the cheaper trip, a higher sunk cost made people more likely to pick the pricier one (Roth, Robbert & Straus, 2014).

They also ran a field study. People who came to the Ohio University Theater ticket booth at the start of the season were sold season tickets at one of three prices: roughly a third paid the full $15, a third paid $13 and a third paid $8. Compared with full-price buyers, those who got either discount went to fewer plays over the next six months (Arkes & Ayton, 1999). Every buyer had the same right to attend any play, so the price already paid shouldn’t have changed how often they went (same source).

Later field research found similar results in other settings. In one study at an all-you-can-eat pizza restaurant, diners who paid full price for the buffet ate more slices than those who paid half price (Roth, Robbert & Straus, 2014).

Why people do it

Researchers have proposed several reasons, and they probably work together.

The first is a wish to avoid waste. Arkes and Blumer pointed to the motive “not to appear wasteful” as one likely cause (Roth, Robbert & Straus, 2014). Arkes and Ayton later argued that adults overgeneralize a “Don’t waste” rule, applying it even in cases where it leads them astray (Arkes & Ayton, 1999).

The second is the way people feel losses. Building on Kahneman and Tversky’s prospect theory, economist Richard Thaler argued that people keep a kind of mental account for each purchase. Skipping something you’ve paid for means closing that account “in the red,” which feels like taking a loss, so people use what they’ve paid for to avoid that feeling (Roth, Robbert & Straus, 2014). In prospect theory, losses loom larger than gains of the same size, which helps explain why a failing project’s past spending weighs so heavily (same source).

The third is self-justification. In a 1976 study, Barry Staw had 240 business school students make a simulated investment decision. They committed the most money to a previously chosen course of action when they were personally responsible for its bad results (Staw, 1976). Staw linked this to a self-justification process, in which people try to rationalize their earlier behavior (same source). Management researchers call this pattern escalation of commitment, and they count both justifying past decisions and not wanting to appear wasteful among its drivers (Roth, Robbert & Straus, 2014). If you’ve noticed how people defend a belief once they’ve committed to it, our guide to confirmation bias covers a related habit.

The Concorde fallacy

Biologists have their own name for this error. In a 1976 letter in Nature, Richard Dawkins and T. R. Carlisle argued that one idea in the theory of parental investment was a fallacy: the idea that an animal should keep investing in offspring because of how much it had already put in. They compared it to a government that has invested heavily in a supersonic airliner and is reluctant to abandon it, even when a sober look at its prospects says it should. What ought to matter, they wrote, is how much investment is still needed in the future (Dawkins & Carlisle, 1976).

The error became known as the Concorde fallacy, after the Anglo-French supersonic jet. Economists Sandeep Baliga and Jeffrey Ely note that France and Britain kept investing in the Concorde after it was known it would be unprofitable (Baliga & Ely, 2011).

Arkes and Ayton explain that “Concorde fallacy” has been used for animals and “sunk cost effect” for humans. After reviewing the animal research, they concluded there were no unambiguous cases of the Concorde fallacy in animals. They also presented evidence that young children in a situation like a sunk cost one behaved more in line with economic rules than adults did (Arkes & Ayton, 1999).

Is it always irrational?

Not always, and the evidence for the effect is more mixed than its fame suggests.

Baliga and Ely argued that paying attention to sunk costs can be a sensible response to limited memory. If you can’t recall every reason you started a project, the amount you were willing to spend tells you something about how promising it looked at the time (Baliga & Ely, 2011). In their lab experiment, people in the baseline version leaned the other way, becoming less eager to finish projects with high sunk costs. When the researchers added memory limits, that reversed and the Concorde bias appeared (same source). They also describe other economic models in which a person loses reputation by reversing an earlier investment, which gives a strategic reason to stay the course (same source).

Paying up front can also act as a commitment device. One study of health club members found attendance was highest right after people paid (Roth, Robbert & Straus, 2014).

The effect doesn’t show up everywhere. In a large field experiment in Zambia, households paid different prices for a water purification product, and researchers found no sign that the price paid affected how much they used it (Roth, Robbert & Straus, 2014). Baliga and Ely note that a survey of the economics and psychology literature found convincing evidence for sunk cost bias elusive (Baliga & Ely, 2011). Even so, a meta-analysis of 98 effect sizes concluded that the sunk cost effect does emerge, though its size depends on the type of decision (Roth, Robbert & Straus, 2014).

How to avoid the sunk cost fallacy

No one has tested these habits as a training program. They follow from the definition and the studies above.

1. Ask the fresh-start question. If you were offered this plan today with nothing already spent, would you take it? That’s the economists’ rule turned into a question: only future costs and benefits count.
2. Name the sunk cost out loud. In one lab study, when people could clearly see that a cost was sunk, the sunk cost effect almost disappeared (Baliga & Ely, 2011).
3. Separate wasting from losing. The money is already gone. Spending more time or money on something you don’t want doesn’t bring it back.
4. Watch for decisions you’re responsible for. Staw’s participants escalated most when the earlier choice was their own (Staw, 1976), so be most careful with projects you championed.
5. Remember that sometimes the past is informative. If you’ve forgotten why you started, the effort you once put in may tell you something, so look up your original reasons before you quit.

For more ways to slow down and check your reasoning, see how to improve your critical thinking and our list of logical fallacies.