You've paid for the ticket, the film is bad, and you stay anyway. You're forty minutes into a queue that isn't moving and you don't leave, because leaving would mean the forty minutes bought nothing. Economists have a flat, almost insulting name for this: the sunk cost fallacy. Money or time already spent is gone regardless of what you do next, so it shouldn't factor into the next decision at all — only what happens from here forward should. And yet it factors in every time.
The theory that explains the fallacy without calling you irrational
Prospect theory, the framework behind this, doesn't say people are bad at maths. It says people weigh losses and gains on different scales entirely: a loss of a given size hurts noticeably more than an equal-sized gain pleases. Walking away from the film means accepting a loss — the ticket price, final, no upside. Staying keeps that loss unrealised, even though staying can't undo it either. You're not weighing "leave" against "stay" on their actual merits. You're weighing a certain, acknowledged loss against a fuzzier one you haven't had to admit to yet, and the fuzzier one wins because admitting it is the part that hurts.
That's loss aversion, and the sunk cost fallacy is just loss aversion pointed at a cost you can't recover. The money's already spent either way. What changes is whether you have to feel like you spent it for nothing, and continuing is the cheapest way to postpone that feeling — even when it costs more of the thing you have left.
Why the fallacy isn't the whole story
Not every case of "I already started, so I'll finish" is the fallacy in disguise. Finishing a course because you've learned the professor grades the final generously is a rational update, not a sunk cost error — it's a decision about the future, made with new information, that happens to point the same direction as the money already spent. The tell that separates the two is simple to state and hard to apply in the moment: would you make this same choice if you'd paid nothing for it so far? If the answer changes once you imagine the past cost away, the past cost is doing work it has no business doing.
What we're still unsure about
Knowing the name of a bias is a weaker defence against it than it feels like in the reading. Naming loss aversion doesn't switch it off — the queue still feels like it wants finishing even after you've correctly diagnosed why. The honest caveat is that a lot of behavioural economics is better at explaining a decision after the fact than at changing the next one in real time, and this topic is a clean example of that gap.
This sits inside Prospect Theory & Loss Aversion, one of seven topics in Behavioral Economics, one of five domains in Economics, one of seventeen subjects the app can quiz you on.