The Sunk Cost Fallacy

Why past spending that can't be recovered shouldn't factor into a decision going forward - and why it so reliably does anyway.

A sunk cost is money, time or effort already spent that cannot be recovered no matter what choice gets made from this point forward. Economic theory is genuinely clear on how a sunk cost should factor into a future decision: it shouldn’t, at all, because no choice available going forward can change what’s already been spent. In practice, people factor sunk costs into decisions constantly, a pattern with real, well-documented consequences.

The fallacy, defined precisely

The sunk cost fallacy is the tendency to continue investing in a decision - a project, a relationship, a purchase - specifically because of what’s already been invested in it, rather than based on the actual expected value of continuing from this point forward. The rational question in any decision is always “given where things stand right now, what’s the best choice going forward?” The sunk cost fallacy substitutes a different, backward-looking question instead: “how much have I already put into this?” - a question whose honest answer has no actual bearing on what should happen next.

The movie ticket that's already been paid for

Imagine buying a $15 movie ticket, then discovering thirty minutes in that the movie is genuinely terrible. The $15 is gone either way - leaving now versus staying to the end doesn't bring that money back under any circumstance. The only real decision left is whether the next ninety minutes are better spent finishing a movie you're not enjoying or doing literally anything else instead. A person who stays "to get their money's worth" is, without realizing it, treating a sunk cost as though it were still somehow recoverable - when the truthful comparison is between ninety more minutes of a bad movie and ninety more minutes of almost anything else.

Why this bias runs so deep

Part of what makes the sunk cost fallacy so persistent is that abandoning something already invested in can feel like admitting the original decision was a mistake, and most people find that admission genuinely uncomfortable. Continuing to invest, by contrast, can feel like the original decision is still being vindicated, even when the honest facts on the ground no longer support that conclusion at all. This emotional pull toward consistency, not any careful calculation, is usually what’s actually driving the decision to keep going.

Escalation of commitment: the fallacy at a larger scale

When the sunk cost fallacy plays out repeatedly, growing larger each time, economists call it escalation of commitment - a pattern of continuing to invest more and more into a failing course of action, specifically because of how much has already been invested in it. This shows up in business constantly: a company that has already spent millions on a struggling product line may keep funding it well past the point a fresh, unbiased evaluation would recommend, purely because walking away would mean acknowledging those millions produced no return.

"Recognizing sunk costs means never finishing anything hard"

Avoiding the sunk cost fallacy doesn't mean quitting the moment something gets difficult - a genuinely good future opportunity is still worth pursuing even if it's hard. The fallacy specifically describes continuing something *because of* past investment rather than because of its actual future value. The test is whether the decision to continue would still make sense to someone with no memory of what had already been spent - if it would, continuing is a reasonable choice on its own merits, not a fallacy at all.

A practical way to catch it in yourself

A genuinely useful habit is to mentally ask, before continuing any costly commitment, “if I were starting completely fresh right now, with none of what I’ve already spent, would I still choose to begin this?” If the honest answer is no, continuing purely because of past investment is very likely the sunk cost fallacy at work.

Key takeaways

A sunk cost is money, time or effort already spent that can't be recovered no matter what happens next.
The sunk cost fallacy means continuing a decision because of past investment rather than its actual future value.
Rational decisions should only weigh the best choice going forward, ignoring what's already been spent.
Escalation of commitment describes the fallacy growing larger over repeated rounds of continued investment.
A useful test is asking whether you'd still choose to start fresh today, with none of the past investment already made.