The Sunk Cost Trap: Why We Throw Good Time After Bad

Imagine you have paid 100 dollars for a weekend ski trip. A few weeks later, you find a much better ski trip for only 50 dollars, and you buy that ticket too. Then you discover the two trips are on the same weekend, and both tickets are non-refundable. Which one do you go on?

Economically, the answer is obvious: go on the better trip. The 100 dollars is gone whichever ticket you use - it is what economists call a sunk cost, money already spent that no future decision can recover. Only the trip ahead of you should matter.

In 1985, Hal Arkes and Catherine Blumer put a version of this scenario to real participants, and a sizeable share of them chose the more expensive, worse trip anyway. Losing 100 dollars felt worse than losing 50, even though by the time the choice was made, both amounts were already gone regardless of what happened next.

The trap, defined

The sunk cost fallacy is the tendency to keep investing time, money or effort into something because of what you have already put into it, rather than because of what it is likely to produce going forward. It shows up far beyond ski trips: staying in a job you dislike because you spent years building seniority there, finishing a bad book because you are halfway through, or a company continuing to fund a failing project because of how much it has already cost.

Arkes and Blumer ran several versions of this experiment and consistently found the same pattern: people treated money, time and effort already spent as a reason to continue, even when told explicitly that it should not matter. In one version, subjects who had personally paid for something were more reluctant to abandon it than subjects who were simply told the same amount had been spent by someone else - suggesting the effect is at least partly about not wanting to feel that your own past choice was wasted, not just about the abstract quantity involved.

Where it comes from: escalation of commitment

A related and older line of research looked at the same pattern in organisations, under the name escalation of commitment. In a widely cited 1976 study, Barry Staw had participants act as financial managers deciding whether to invest further in one of two company divisions. Some were told they had personally chosen the division that later performed poorly. Compared to participants who had not made the original choice, those who had were significantly more likely to sink additional money into the very division they had picked - trying to justify their earlier decision rather than evaluate the current one on its own terms.

This gives the sunk cost fallacy a second engine beyond simple loss aversion: it is not just that abandoning a project feels like admitting a loss, it is that abandoning it can feel like admitting you were wrong to start it. Protecting your own track record and protecting your money get tangled together, and the tangle can be worse than either factor alone.

How solid is the effect, really

This is one of the better-replicated findings in behavioural science, but it is not simple or uniform. A large 2012 meta-analysis by Dustin Sleesman and colleagues, pooling decades of escalation-of-commitment studies, confirmed the effect is real on average but also found that it depends heavily on context: it is stronger when the decision-maker feels personally responsible for the original choice, when there is social pressure to appear consistent, and when the project's failure is ambiguous rather than obvious. It weakens, and can disappear, when clear negative feedback is available and organisational incentives reward changing course rather than sticking to it. In other words, this is not a fixed bias that always fires at full strength - it is a tendency that context can amplify or shrink.

What this does not mean

Persisting with something is not always a sunk cost error. Genuinely useful information you gained along the way, momentum that would be expensive to rebuild, or a project that only needed slightly more time to succeed are all legitimate reasons to continue, and they have nothing to do with what you already spent. The fallacy specifically describes cases where the amount already sunk is the actual reason for continuing, separate from any honest reassessment of what lies ahead. Distinguishing the two in the moment is genuinely hard, which is part of why the bias persists even among people who know it exists.

One practical tactic follows directly from Staw's finding about personal responsibility: the effect is weaker when someone other than the original decision-maker evaluates whether to continue. Asking a friend, a colleague or even a future version of yourself who was not attached to the original choice to look at the decision fresh removes exactly the ingredient - protecting your own past judgment - that the escalation-of-commitment research identifies as doing much of the damage. It will not eliminate the pull of sunk costs entirely, but it targets the part of the bias that pure willpower struggles to touch.

What this means for you

The next time you are deciding whether to keep going with something - a subscription, a relationship, a half-finished course, a struggling side project - try a version of the question Arkes and Blumer's experiments point to: if you were deciding today, with no memory of what you had already invested, would you still choose this path? If the honest answer is no, the money, time or effort already spent is not a reason to keep going. It is only a reason you feel like you should.