“I’ve already put $5,000 into this — I can’t stop now.” Sound familiar? That’s the sunk cost: money already spent and unrecoverable. The trap is that it feels like an argument for continuing — when it’s not an argument for anything. It’s gone either way, whatever you choose.
How it works
One rule: decide looking only forward. Continuing is worth it if what you get by finishing beats what it still costs to get there — compared against what you recover by stopping now.
keep going if: value when done − cost to finish > what quitting returns
The money already spent doesn’t appear in the formula. Not out of coldness — out of arithmetic. The calculator shows it anyway, at the bottom, on both ledgers: so you can watch it shift the two totals by the exact same amount while the gap between the paths doesn’t move a dollar.
A worked example
You’ve put $5,000 into a project — a small shop, a van you’re fixing up, a site you’re building. Getting to the finish takes another $3,000, and once done it will return $2,500. Quit now, sell the gear and materials, and you recover $500.
From here on: continuing is worth 2,500 − 3,000 = −$500; quitting is worth +$500. Quitting leaves you $1,000 ahead, full stop. And the $5,000 already spent? Crank it to $50,000 in the calculator: both final tallies get uglier, but the gap stays $1,000. The more you’ve spent, the more it hurts — and the less it has to do with the choice.
Why your brain fights this
Kahneman’s explanation is loss aversion: as long as you keep going, the loss is only “on paper”; quitting makes it final, and a sure loss weighs about twice as much psychologically as an equal gain. The SEC sees the same wiring in investors — its bulletin on behavioral patterns describes the disposition effect, “the tendency of an investor to hold on to losing investments too long” — which is the sunk-cost fallacy wearing a brokerage account. The same reflex finishes appetizers nobody likes, watches bad shows to the final season, and — with more serious money — keeps funding projects the numbers have already voted against. Economists also call it the Concorde fallacy, after the supersonic jet funded for years past the point where it could ever pay for itself.
The honest limits
- “Value when done” is an estimate, usually the shakiest number on the page. Run a cautious value and an optimistic one: if the verdict holds for both, you’re done; if it flips, the real decision is about the value, not about quitting.
- Not everything is money. A promise made to other people, what you learn by finishing, your reputation: if they’re worth something to you, fold them into “what it returns, once finished” — as long as they’re future value, not regret in disguise.
- It cuts both ways. Sunk-cost thinking sometimes makes people quit too early (“I’ve lost so much, I’m done”) when the forward numbers say stay. The formula doesn’t root for either side: it just looks ahead.
Once you’ve decided to quit, the next question is what those freed-up future dollars could do instead — that’s exactly what opportunity cost computes. And in the cost of a habit, the same forward-only logic applies to spending that repeats. Meanwhile, try it above: raise “already spent” as high as you like and watch the gap hold still. It’s the most convincing proof that reading about the fallacy isn’t the same as seeing it.