What is the sunk cost fallacy?

A sunk cost is money, time, or effort already spent and unrecoverable. The sunk cost fallacy is the tendency to keep investing in a losing course of action because of what has already gone in. Rationally, sunk costs are identical across every remaining option, so they should be ignored: only future costs and benefits should count. But abandoning feels like wasting what was spent, so people escalate to avoid that feeling, even though the waste has already happened.

What was the airplane experiment by Arkes and Blumer?

In Arkes and Blumer (1985), people were asked whether to spend the last of a $10 million budget to finish a radar-blank plane that a rival had already beaten to market. About 85% said yes, finish it. When a matched group was offered the same doomed plane with no prior investment mentioned, only roughly one in six chose to spend. Same plane, same future, opposite decision, driven entirely by the money already sunk.

Do animals show the sunk cost fallacy?

For about twenty years the consensus (Arkes and Ayton, 1999) was no: there were no unambiguous cases of the fallacy in animals, and young children seemed immune too, suggesting it was a learned human quirk. Then Sweis et al. (2018) in Science found that mice, rats, and humans all show sunk-cost sensitivity in matched foraging tasks, and crucially the pull only appeared after a decision had been committed. It is not fully settled, but the human-only story broke.

Heuristic ▶ Subscribe

Episode 3 · Cognitive biases · Research notes

Why your brain refuses to cut its losses

You have spent $10 million on a plane. It is 90% done when a rival ships a better, cheaper one. Do you finish yours? In the study that made this famous, 85% said yes, but strip away the money already spent and almost nobody would. That gap is the sunk cost fallacy, and it decides more of your life than you think.

The short answer: sunk costs are gone no matter what you choose next, so a rational mind ignores them. But quitting feels like the wasteful act, so you keep spending to avoid that feeling, even though the waste already happened. And the honest part most videos skip: for 20 years science said only humans do this, then in 2018 mice and rats did it too, and in 2025 the classic experiments were flagged as unreliable. Real, but weirder than the textbook says.

Episode 3 premieres

Wednesday, July 15 · 8:00 AM PT

▶ Subscribe to catch it

The experiment: same plane, opposite choice

The landmark study is Arkes and Blumer (1985), ten experiments on the psychology of sunk cost. Their centerpiece was the airplane: a firm has spent $10 million of a $10 million budget on a radar-blank plane, and at 90% done a rival markets one that is better and cheaper. Finish it? About 85% of people said yes. A matched group offered the very same doomed plane as a fresh decision, with no prior spend, chose to invest only about one in six times.

85%

finish it when $10M is sunk

~17%

spend when it's a fresh choice

They also ran it in the real world. Theater-goers were randomly charged full price ($15) or a surprise discount ($13 or $8) for identical season tickets. Full-price buyers attended more plays early on, not because they liked theater more (the prices were random) but because skipping felt like wasting $15. The revealing detail: by the second half of the season, the difference had faded. The grip of a sunk cost weakens as the memory of paying recedes.

Where you meet it

The dollar auction (Shubik, 1971) is a party game where a $1 bill is auctioned but the runner-up also pays and gets nothing, so it routinely sells for $3 to $5 as neither bidder can absorb the loss of quitting. Barry Staw's "knee deep in the big muddy" (1976) showed managers pour the most money into a failing project when they personally made the original call. Vietnam is the dark version, where each wave of sacrifice became the argument for more. And in everyday life: the dead-end job, the "we've been together five years," the stock held until it "gets back to even," the plate finished only because you paid for it.

The 2018 reversal, and the fight now

For two decades the textbook story was that this trap is uniquely human (Arkes and Ayton, 1999): no clear cases in animals, and children seemed more rational than adults, so it looked like a learned "don't waste" rule that we overapply. Then Sweis et al. (2018), in Science, built matched foraging tasks for mice, rats, and humans, and all three clung harder the longer they had already waited. The striking detail: the sunk-cost pull only appeared after a decision had been committed, not before. The Concorde fallacy was named by biologists in 1976, and biology took it back.

It is not fully settled, and that is the honest part. A pigeon study found the effect only under narrow conditions, critics raised a possible statistical artifact and the original team ran new controls, and in 2025 the classic questionnaire experiments were reported as psychometrically unreliable (the same person's answers to different sunk-cost questions barely correlate). The trap is real, you have felt it, but it looks older than our species, it fades, and its true shape is being argued over right now.

Sources

Arkes, H. R. and Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision Processes, 35(1), 124-140.
Dawkins, R. and Carlisle, T. R. (1976). Parental investment, mate desertion and a fallacy. Nature, 262(5564), 131-133.
Staw, B. M. (1976). Knee-deep in the big muddy. Organizational Behavior and Human Performance, 16(1), 27-44.
Shubik, M. (1971). The dollar auction game. Journal of Conflict Resolution, 15(1), 109-111.
Kahneman, D. and Tversky, A. (1979). Prospect theory. Econometrica, 47(2), 263-292.
Arkes, H. R. and Ayton, P. (1999). The sunk cost and Concorde effects. Psychological Bulletin, 125(5), 591-600.
Sweis, B. M. et al. (2018). Sensitivity to sunk costs in mice, rats, and humans. Science, 361(6398), 178-181.
Bialek, M. and Biesiada, M. (2025). Psychometric limitations of classic sunk cost measures. Brain Sciences, 15(8), 808.