Sunk Cost Fallacy

In plain terms

A sunk cost is money, time, or effort that's already spent and can't be recovered. The sunk cost fallacy is letting those past investments drive future decisions, even though, by definition, nothing you do now changes what you've already put in.

The rational rule is almost insultingly simple: the only thing that should affect a decision is what happens from this point forward. The past is a constant. Yet nearly everyone, in nearly every domain, struggles to stop factoring it in.

Why it matters

Past investment feels like it should earn continued commitment, because walking away seems to admit the original investment was wasted. The feeling isn't imagined. Abandoning a project does mean the spent resources won't produce returns. But that was already true the moment they were spent. Quitting doesn't make it more true, and continuing doesn't make it less true. The only question, at any decision point, is whether the next investment is worth the next expected return.

Knowing this intellectually doesn't make it easier to act on. The fallacy is so persistent that economists had to give it a name to get people to notice it.

Canonical example

"I've already spent two years on this PhD, so I have to finish even though I hate it and the job market for my field has collapsed."

The two years are gone. Finishing the PhD will take additional years and additional effort. The decision is not "should I have started" — that's answered. It's "given where I am, is continuing worth what continuing will cost?" The previous two years enter the analysis only as context for where you are now, not as a reason to keep going. Many people, when they work through the math cleanly, realize the sunk cost was doing most of the persuading.

Counter-example (not the fallacy)

"I've put in eighteen months on this company. The first year was rough, but the last six months show a clear trajectory: revenue is up, the team is stable, and I've learned enough about the industry to see specific near-term opportunities. Continuing makes sense based on what's in front of me."

This isn't the sunk cost fallacy. The continuation is justified by forward-looking evidence (trajectory, opportunities), not by the prior investment itself. The past eighteen months show up as a source of information about the future, not as an argument for the future. That's a legitimate role for past effort. The fallacy only happens when the past is treated as a debt the future is obligated to pay.

How to apply it to yourself

A simple test: if you had just arrived at today's situation with no history, would you make the investment required to continue? Strip out all the effort already in, imagine you're starting fresh with only the current state and the expected returns, and decide from there. If the answer is "no, I wouldn't start this now," you've just identified the sunk cost at work, and the honest next question is why you're still paying it.