You bought a $15 movie ticket. The movie is bad. You sit through it anyway.
You've been at a job for five years. The first three were good. The last two have not been. You stay.
You've been in a relationship for three years. The last six months have made you smaller. You don't leave.
You sank $40,000 into a business that isn't working. You keep going, partly because you've already spent $40,000.
Every economist in the world will tell you these decisions are irrational. The money is gone. The years are gone. Future decisions should be made on future returns, not past investment. That's the rule.
And yet almost no one follows it. Including the economists.
This pattern has a name — the sunk cost fallacy — and over the last fifteen years, neuroscientists have started to map exactly where it lives in the brain. The mapping is humbling. It explains why willpower mostly fails here. It also points at what actually works.
The bias is older than you are
The sunk cost effect was formally described by Hal Arkes and Catherine Blumer in a 1985 paper that's now one of the most-cited in behavioral economics. They showed it across dozens of scenarios, in business and personal decisions, in laboratory experiments and real-world data.
The pattern is consistent. The more you've already invested in a course of action — money, time, effort, identity — the more likely you are to keep going, even when continuing is the worse choice on every dimension that matters going forward.
This isn't a quirk of dumb people. Arkes' early studies showed it in MBA students, executives, and economists. People who knew the rule still broke the rule.
That suggests this isn't a knowledge problem. It's a wiring problem.
What lights up when you can't let go
Recent fMRI work — including studies by Sara Garofalo, Maël Lebreton, and others — has identified the neural circuits involved in sunk cost decisions. The picture is more nuanced than the original viral claim that "two brain regions are hijacking your choices," but a few patterns are real.
When you face a decision where you've already invested, your ventromedial prefrontal cortex — the brain region that calculates the value of options — does something important. It doesn't just weigh future outcomes. It folds in your past investment as part of the value calculation.
Your anterior cingulate cortex — a region involved in tracking conflict and effort — registers the discomfort of "wasting" what you've already put in.
And your insula — which processes the gut-level pain of loss — activates strongly. The insula is part of why losses feel about twice as bad as equivalent gains feel good. That asymmetry is well-documented from Kahneman and Tversky's prospect theory work in the 1970s.
Together, these systems make abandoning a sunk investment feel like an active loss, even when the rational choice is to stop pouring resources into something that won't return them.
The brain isn't broken. It's doing what evolution shaped it to do. In ancestral environments, persistence often paid off — the hunt that takes hours yields a meal; the relationship that takes years builds the alliance that keeps you alive. Quitting was often more dangerous than continuing.
In modern life, with mortgages and careers and capital structures and decades-long trajectories, that same persistence circuit can keep you stuck for a long time.
Why "just be rational" doesn't work
Here's what trips up smart people. You can know about the sunk cost fallacy. You can have read the studies. You can quote Kahneman at parties. And you'll still feel the pull.
That's because the bias operates partly below conscious deliberation. By the time your prefrontal cortex thinks "should I quit this," your insula and ACC have already loaded the question with the felt cost of letting go.
The studies on this are clear. Patients with damage to the ventromedial prefrontal cortex are actually less prone to sunk cost effects in some experiments — they don't fold past investment into the calculation the way intact brains do. That's not a recommendation; those patients have other serious decision-making problems. But it tells us the bias is structural, not weak willpower.
What that means in practice: you don't beat this with self-talk. You beat it with structure.
What actually works
A few moves have evidence behind them.
**Make the decision from the outside.** Pretend a friend came to you with the same situation. They've been at the failing job for five years. They've sunk $40,000 in the business. They're three years into the relationship that's making them smaller. What would you tell them? People are consistently more rational about other people's sunk costs than their own. Use that.
**Pre-commit to exit conditions.** Before you start a project, a job, an investment, write down what would have to be true for you to walk away. "If revenue isn't at $X by month Y." "If I'm still feeling Z by date W." Decisions made before you've invested are cleaner than decisions made after.
**Ask the future-only question.** Not "should I have done this?" but "starting from where I am now, with no obligation to my past, what's the best move from here?" This is the question economists prescribe and almost no one actually asks. It works.
**Notice the felt cost.** When you feel the pull to keep going specifically because you've already invested — not because the future case is strong — that's the bias. Name it. Then make the decision again.
Where this shows up that you might not expect
The classic examples are money and careers. But the sunk cost fallacy runs through almost every long-running commitment in your life.
Health goals you set five years ago that no longer fit. Friendships that drained out a decade ago but you keep tending. Beliefs about who you are that were true at twenty-two and aren't true now. Investments — financial and emotional — in a version of your life you've outgrown.
The cost isn't usually the loss of the original investment. The cost is the years you spend after, defending it.
The reframe
You're not throwing away the past. You can't. The past already happened. The money is already spent. The years are already gone.
The only question that's actually live is: what do you want to do with the rest of it?
Some things are worth continuing. The investment was real, the trajectory is good, the future case is strong. Stay.
Other things are worth ending. The investment was real, but it doesn't make the future case stronger. Leave.
Your brain will not make this distinction cleanly on its own. It's not built to. You have to build the structure that makes the distinction for you.
That's not irrational. That's the most rational thing you can do.
---
*Pairs well with: "Your Brain Has Three Different Fear Circuits for Money" and "What Neuroeconomics Says About Why You Bought That."*