The Sunk Cost Fallacy: Why Smart People Keep Paying for Bad Decisions

Learning to spot this fallacy in yourself will save you more money than any other item in this series.

What is the sunk cost fallacy?

A sunk cost is any resource you have already spent and cannot get back. Money, hours, reputation, years.

The fallacy is letting that spent resource influence what you do next. The correct question is always what the next dollar or the next hour buys. The question people ask instead is what the last hundred dollars or the last five years would have been for.

Economists have been clear about this for a long time and it changes nothing, because the fallacy doesn’t live in the reasoning. It lives in the part of a person that cannot stand waste.

Consider a project eighteen months in and clearly failing. The rational analysis looks only forward. What does finishing cost from here, and what’s finishing worth. The eighteen months appear nowhere in that calculation, and in the meeting they’re the only thing anyone talks about.

Why do intelligent people fall for sunk cost?

Because intelligence isn’t the faculty being tested.

Loss aversion is doing the work. People feel a loss roughly twice as strongly as they feel an equivalent gain, so abandoning something converts a vague ongoing situation into a definite loss with a number attached. Continuing keeps the loss theoretical.

Public commitment adds to it. Somebody argued for this project, staffed it, defended the budget. Stopping it means standing in front of the people who watched all that and saying the money is gone.

Identity comes in on top of those. A person who has spent nine years becoming something doesn’t experience quitting as a decision about the next nine years. They experience it as a statement about who they’ve been.

None of that is stupidity. It’s a set of ordinary human responses producing an expensive answer.

How does sunk cost show up in business decisions?

Four places, and every one of them is expensive.

Failing projects get finished because stopping would waste what was already spent. The waste already happened. Finishing adds to it.

Bad hires stay because of what recruiting and training cost. The training is gone whether the person stays or leaves, and every additional month is a fresh decision to keep paying.

Software gets maintained past the point of sense because the build was expensive. The build is done. The only live question is what maintaining it costs against what replacing it costs.

Vendor relationships continue because of the integration work already completed. That work is spent. The switching cost is real and belongs in the analysis. The original integration doesn’t.

The pattern in all four is the same. A number from the past is being entered into a calculation about the future, where it doesn’t belong.

What is the difference between sunk cost and persistence?

Separating the two is the hard part, and anyone who says the distinction is obvious hasn’t run anything.

Persistence and the sunk cost fallacy look identical from the outside. Both involve continuing when things are difficult. Both feel like commitment. Plenty of successful ventures survived because somebody refused to quit during a period that looked exactly like failure.

The distinction is in the reason, not in the behavior.

Persistence says the plan is still sound, the conditions still favor it, and the evidence hasn’t changed. Sunk cost says look how much we have put in.

A useful test: if this project landed on your desk today, with the current state and no history, would you fund it. If yes, continuing is persistence. If no, the only thing holding it up is the money already spent.

Most people cannot run that test on their own work. Somebody else has to ask.

How do writers use the sunk cost fallacy in fiction?

The sunk cost fallacy is one of the most useful engines available for a character who needs to make a bad decision the reader still believes.

Readers reject characters who behave stupidly for no reason. They accept characters trapped by reasoning that feels correct from inside, and sunk cost feels extremely correct from inside.

A detective who has built a case against the wrong suspect cannot easily abandon it, because eight months of work argue against every piece of contrary evidence. That’s not a stupid detective. That’s a human one, and the reader can see the trap closing while the character cannot.

The same engine drives a marriage nobody wants to leave after twenty years, a business that swallows a family’s savings one decision at a time, and a soldier who cannot accept that the position is lost because of what taking it cost.

What makes it work in fiction is that the audience does the math the character won’t. That gap is tension, and it costs nothing to build.

How do you write a character escaping a sunk cost trap?

The escape has to cost something visible, or the reader won’t believe it.

A character who simply realizes the truth and walks away hasn’t escaped anything. Realization is cheap. The trap demands that the character accept the loss publicly, in front of people who watched them commit.

Write the moment where they say the number out loud. Eight months. Nine years. Every dollar we had. The saying is the price, and it’s what makes the exit land.

Then let the loss stay lost. A story that rewards the character with immediate compensation has quietly argued that leaving was free, and leaving is never free.

What question breaks the sunk cost fallacy?

One question, and it works on projects, relationships, hires, and drafts.

If I were starting from here today, knowing nothing about what it took to get here, what would I do next?

The question strips the past out of the calculation, which is where the past belongs. Everything already spent stays spent regardless of the answer.

Two things make it work in practice. Ask it out loud to somebody who didn’t make the original decision, because the person who made it cannot hear their own answer. And ask it on a schedule rather than in a crisis, since a project reviewed every quarter never accumulates the emotional weight that makes stopping unthinkable.

The fallacy grows with time and silence. Regular review starves it.

Frequently Asked Questions