The Sunk Cost Fallacy: Why We Throw Good Money After Bad
Money already spent should not influence future decisions. Almost everyone violates this principle. The reasons we do, and the costs we pay for it, are worth understanding.
The sunk cost fallacy is one of the most studied cognitive biases in behavioral economics, and one of the easiest to demonstrate. A sunk cost is a cost that has already been incurred and cannot be recovered. According to standard decision theory, sunk costs should not factor into future decisions. The only thing that should matter is what happens next.
In practice, almost no one actually thinks this way. We treat sunk costs as if they were relevant to current decisions, even though we know, on reflection, that they aren’t.
The classic example
You buy a non-refundable ticket to a concert. The day of the show, you feel sick and would rather stay home. The standard reasoning would be: the cost of the ticket is gone either way. The only question is whether you’d enjoy the concert more than staying in. If staying in wins, stay in.
Most people don’t reason that way. They reason: “I paid for this ticket; I should use it.” They go to the concert, often regret going, and have demonstrated the fallacy.
The error is treating the past expenditure as if it were a future cost. The money is gone whether you go or not. Going does not bring it back. Staying home does not waste it any further.
Why it’s persistent
Sunk cost reasoning persists because it satisfies an intuitive emotional need: we want our past efforts to count for something. Acknowledging that a sunk cost is gone means acknowledging that the original decision didn’t pay off. We resist this acknowledgment, even at the cost of further bad decisions.
The bias compounds. Once we’ve invested in a project, we resist abandoning it. The longer we stay invested, the more we’ve sunk, the harder it becomes to leave. Investment escalates not because the project is going well, but because we’ve already invested.
Gambling as the paradigm case
The clearest real-world examples of the sunk cost fallacy involve gambling. A player loses several bets in a row and feels compelled to continue playing to “win back” the losses. The reasoning is fallacious: every new bet is statistically independent of the previous ones. The losses do not change the odds of the next bet. They have already happened and cannot be undone. Industry consumer-education resources document this pattern in detail, precisely because it is the single most common pattern in problem gambling behavior.
The gambling case is illustrative because the math is stark. Each bet has its own expected value, independent of the previous bets. A player operating on sunk-cost logic is essentially treating the previous bets as if they were a cost the next bet could recover. They cannot.
How to counter it
The standard advice is to reframe the decision: imagine you had not yet incurred the sunk cost. Would you, today, make this investment? Would you, today, attend this concert? Would you, today, place this bet?
If the answer is no, you should not continue, regardless of what has already been spent. The past expenditure is not an argument for continuing; it is an argument against the original decision, which is too late to revise.
This is a simple test that takes seconds to apply. It will not eliminate the bias entirely — the intuitive pull of sunk costs is too strong — but it will catch the worst cases, and it will make you a more rational decision-maker in real time.
