What is the sunk cost fallacy? Why past costs shouldn't decide

The sunk cost fallacy is the mistake of letting money, time or effort you have already spent, and can't get back, decide what you do next. A sunk cost is gone whichever way you choose, so it can't make one choice better than another. Sitting through a bad film because you paid for the ticket, or pouring more money into a failing project because so much has gone in already, are the classic cases.

In short

• A sunk cost is a cost already paid that no future choice can recover. Because every option leaves it the same, it shouldn't tip the decision.
• The test: compare only what each option costs and brings from now on. Ask whether you would start this today, knowing what you know.
• People fall for it because stopping feels like admitting a loss. Yet carrying on is still right when what's left to pay is worth what's left to gain.

What is a sunk cost?

A sunk cost is a cost that was paid in the past and can't be recovered, whatever you decide now. A non-refundable ticket, the years spent on a degree you no longer want, the money already spent building half a bridge: each is sunk. OpenStax's Principles of Economics states the economist's rule plainly: "forget about the money and time that is irretrievably gone and instead to focus on the marginal costs and benefits of current and future options."

The reason is simple once it's laid out. A decision is a choice between futures. A cost that is the same in every future, because it has already happened, can't make any of them better or worse than the others. The only costs that count are the ones a choice can still change. That is the forward-looking idea economists call thinking at the margin.

Not every past cost is sunk. If a concert ticket can be resold for $60, then going to the concert costs you that $60, because you give up the chance to sell it. That is an opportunity cost, and it does count. Only the part you can't get back is sunk.

How do you decide without the sunk cost?

Take a concert. You paid $80 for a ticket that can't be refunded or resold. On the night, a storm sets in, and the venue is a 40-minute drive away.

1. List the options from now on. Go, or stay home.
2. Price each one from now on. Suppose the concert would be worth $100 of enjoyment to you, and the drive in the storm $60 of misery and risk. Going is worth $100 − $60 = $40 more than staying home. Go.
3. Change the storm. Suppose instead the concert is worth $50 and the drive $70. Going is now worth $50 − $70 = −$20 compared with staying home. Stay home.
4. Notice what never appeared. The $80 is missing from every line. It is spent whether you go or not. The answer would be the same if the ticket had cost $300, or been a gift.

A useful check: if a friend offered you this ticket free tonight, would you go? If not, the price you paid is the only thing pulling you out into the storm. The economist Richard Thaler used almost this example in 1980: a family that paid for basketball tickets is more likely to drive through a snowstorm to the game than one that was given them, though the money is gone either way.

The same arithmetic works for a project. A company budgeted $10 million for new software and has spent $18 million, 80% over budget. Finishing needs another $5 million.

• If the finished software will earn $4 million: pay $5 million to get $4 million, and the company is $1 million worse off than stopping now. Stop.
• If it will earn $7 million: pay $5 million to get $7 million, and it's $2 million better off than stopping. Finish.

In the second case the project as a whole still loses money: $23 million in, $7 million back, a $16 million loss. But stopping now would lose the full $18 million already spent, so finishing makes the overall loss smaller. The $18 million is gone in either case and decides nothing; only the $5 million still to spend and what it buys do.

Why do people fall for it?

The sunk cost effect is one of the best-documented cognitive biases. In 1985 the psychologists Hal Arkes and Catherine Blumer published a set of experiments, The Psychology of Sunk Cost. In one, people imagined they had paid $100 for a ski weekend in Michigan and $50 for one in Wisconsin that they would enjoy more, then found the trips fell on the same weekend and neither could be refunded. More than half chose the $100 trip they expected to enjoy less. In another, theatergoers buying season tickets were sold them at random at full price or at a discount; those who paid full price went to more plays in the first half of the season.

Several pulls work together:

• Not wanting to waste. Stopping seems to turn the money already spent into money thrown away, though it was spent either way. Arkes and Blumer called this a wish not to appear wasteful.
• Loss aversion. People feel a loss more than an equal gain. Walking away makes the loss feel final; carrying on keeps alive the hope of making it back.
• Responsibility and pride. In experiments by the organizational psychologist Barry Staw, from 1976 on, people who had made the first decision themselves put more new money into a failing course than people who inherited it, which he called escalation of commitment. Admitting the first choice was wrong is uncomfortable, the tension cognitive dissonance describes.

The supersonic airliner Concorde gave the error a second name. The program's first cost estimate, in 1962, was £70 million; by 1976 it had cost between £1.5 and £2.1 billion, according to the aircraft's history. The British government privately came to see it as a commercial disaster that should never have been started, yet Britain and France kept paying, and only 20 were built. Biologists studying animal behavior later called the mistake the "Concorde fallacy". Whether finishing Concorde was wrong at each stage depends on what was left to pay and gain at the time: the 1962 treaty between the two countries imposed heavy penalties for cancelling, a cost that lay in the future, and national prestige and jobs counted too, not only money.

When is carrying on the right call?

Ignoring sunk costs doesn't mean always quitting. It means deciding on what's left. Carrying on is rational when:

• What's left is worth it. As in the software example, finishing can beat stopping even when the whole project loses money.
• Your reputation is at stake. Someone known to abandon commitments may lose future deals, a real cost of quitting that lies in the future, not the past.
• You know something others don't. A manager with private evidence that a project is about to turn may be right to persist where outsiders see only the fallacy.
• The past is information. Having already learned to play the first three chords, you're closer to playing the guitar. The effort is sunk, but the skill it bought is real and makes the next step cheaper.

The opposite mistake is real too: quitting things too early because they're hard, without asking what finishing would bring.

Still argued

Researchers argue over how irrational the sunk cost effect really is. Some point out that honoring past investments can be a sensible rule of thumb in a world where switching often costs more than it seems, and where people who persist build reputations for reliability. Others stress the experiments: people let sunk costs tip choices even when nothing else is at stake. Studies of animals, including a 2018 study of mice, rats and humans, suggest the tendency is old and widely shared, which leaves open whether it once served a purpose.

Where does this lead next?

Sunk costs are the mirror image of opportunity costs. Opportunity cost says to count what you give up that isn't on the bill; sunk cost says to ignore what's on the bill but can't be undone. Both follow from scarcity: since resources are limited, the question is always where the next dollar or hour does the most good, never where the last one went.

Check yourself

• Without looking back, explain why the $80 ticket price shouldn't affect whether you drive through the storm.
• In the software example, why can finishing be right even though the whole project loses money?
• Give one case where taking account of a past commitment is rational, and say what future cost or benefit makes it so.

Sources

• 2.1 How Individuals Make Choices Based on Their Budget Constraint — OpenStax, Principles of Economics 3e, 2022
• The Psychology of Sunk Cost — Hal R. Arkes and Catherine Blumer, Organizational Behavior and Human Decision Processes, 1985
• Sunk cost — Wikipedia
• Escalation of commitment — Wikipedia
• Concorde — Wikipedia