The sunk cost fallacy is the habit of pouring more time, money, or effort into something because you already spent so much, even when every signal says stop. It costs you twice: once on the original loss, and again on the new resources you waste trying to justify it. The fix isn't willpower. It's a math rule you can apply in under ten seconds.
A sunk cost is money, time, or effort you've already spent and cannot get back. It's gone. The fallacy kicks in when you let that gone money influence a decision about the future, as if spending more now could somehow retroactively make the past spending worth it. It can't. The $40,000 MBA you hate doesn't get less wasted because you finish it out of spite.
Economists have a clean rule for this: only future costs and future benefits belong in a decision. Past costs are irrelevant, full stop. Every dollar or hour already spent should carry zero weight in what you do next. Most people nod at that rule and then ignore it completely the moment it applies to their own life.
Loss aversion is the real engine here. Research from Kahneman and Tversky showed that losses feel roughly twice as painful as equivalent gains feel good. Walking away from a sunk cost means admitting a loss right now, today, with your name on it. Continuing the project lets you defer that admission, even if it guarantees a bigger loss later.
There's also a social layer. Quitting a failing startup, a bad marriage, or a stalled degree program invites judgment. Continuing invites none, at least not yet. So people choose the path that delays blame over the path that minimizes damage. That's not irrational in a reputational sense. It's just expensive.
Identity gets tangled in here too. If you've spent five years calling yourself "the person building this company," shutting it down feels like erasing a version of yourself, not just closing a business. The fallacy survives because it protects ego, not because it protects money.
The pattern repeats across every domain where effort precedes outcome. Once you can name it, you'll see it everywhere, including in your own last six months.
| Domain | The Trap | The Real Question |
|---|---|---|
| Career | Staying in a job because of years invested in climbing the ladder | Would you take this job today, knowing what you know now? |
| Investing | Holding a losing stock to "get back to even" before selling | Would you buy this stock today at this price, with this information? |
| Relationships | Staying together because of years, a wedding, or shared history | Does this relationship add value going forward, independent of the past? |
| Projects | Finishing a product nobody wants because the team already built 80% of it | If you started today with zero built, would you still build this? |
| Education | Finishing a degree you no longer want to avoid "wasting" tuition already paid | Does finishing add more value than the time cost of finishing? |
Notice the pattern in the right column. Every real question strips out the past entirely and asks only about the future. That reframe is the entire fix, and it works in about three seconds once you train yourself to ask it.
Here's the calculation that matters, and it has nothing to do with what you've already spent. Compare two numbers: the expected future benefit of continuing, and the expected future benefit of your best alternative. Whichever number is bigger wins. The amount you've already sunk doesn't enter the equation at all, because it's identical under both options. You can't get it back either way.
Take a trading example. You bought a stock at $100. It's now at $60. The question isn't "how do I get back to $100." The question is "if I had $60 in cash right now, would I use it to buy this stock, or would I put it somewhere else." If the answer is somewhere else, sell. The $40 you already lost is lost regardless of your next move. Holding on doesn't un-lose it. It just risks losing more.
The same math applies to a four-year relationship that's been dead for the last eighteen months. The four years are spent. They don't come back whether you stay or leave. The only real question is whether the next four years are better spent in this relationship or outside it. Stated that plainly, most people already know the answer. They just haven't let themselves say it out loud.
You need a trigger phrase, because in the moment, emotion moves faster than math. Use this one: "If I were starting today, with zero already spent, would I choose this?" Ask it out loud, not just in your head. Saying it changes how seriously you take the answer.
Set a pre-commitment before you start anything big. Before launching a project, name the exact condition that would make you kill it: a revenue number by a date, a user count, a specific red flag. Write it down somewhere you'll see it later. Future-you, deep in sunk cost territory, will try to move the goalposts. A written pre-commitment makes that harder.
Get an outside read from someone with zero stake in your past decisions. A friend, a mentor, a financial advisor who wasn't in the room when you made the original investment will see the current numbers clearly, because they have no ego buried in the history. Ask them one question: knowing only where things stand today, what would you do? Their answer is usually the one you already know but can't say to yourself.
Separate the decision from the emotion by writing both down on paper. List the future costs and future benefits of continuing. List the future costs and future benefits of stopping. Leave out everything already spent. Looking at the two lists side by side, on paper, strips out the narrative your brain has built to justify the sunk cost. Numbers on a page don't care about your feelings, and that's exactly the point.
Not every instance of "keep going despite sunk costs" is a mistake. If new information genuinely improves the future outlook, independent of the past spending, continuing can be the correct call. The error isn't persistence itself. The error is persistence driven by the size of the past investment rather than the quality of the future opportunity.
A founder who's burned through two years and a product pivot might have a genuinely better business today than the one they started with. If the current trajectory stands on its own merits, keep going. But test that claim honestly. Ask whether you'd invest fresh capital into this exact business today, from zero, with no history attached. If the honest answer is yes, that's conviction. If the honest answer is "I don't know, but I can't quit now," that's the sunk cost fallacy wearing a business suit.
The discipline isn't about quitting more. It's about deciding on the right variable. Past spending feels urgent and present. Future value is the only thing that actually determines outcomes. Train yourself to weigh the second one, and you'll stop losing money, time, and years to decisions your past self already made for you.
It's the tendency to keep investing time, money, or effort into something because of what you've already spent, even when continuing no longer makes sense. The money or time already spent is gone regardless of what you choose next, so letting it influence your decision is a logical error, not a smart strategy.
Loss aversion makes the pain of admitting a loss today feel much worse than the risk of a bigger loss later, so people delay the admission by continuing. Ego and identity also play a role, since quitting something you've invested years into can feel like erasing part of who you are.
Ask one question before any big call: if I were starting today with nothing already spent, would I still choose this path? Compare only future costs and future benefits between your options, and leave the amount already spent out of the comparison entirely.
No, but new information can be. If fresh evidence shows the future outlook has genuinely improved, continuing can be correct, as long as the decision is based on that future evidence and not on how much has already been spent. The test is whether you'd start fresh today knowing only what you know now.
Holding a losing stock just to get back to your original purchase price is a classic case. The right question isn't what you paid for it, but whether you'd buy that same stock today at its current price with the cash you'd get from selling it.