Throwing good money after bad — the sunk cost trap
🕳️

The sunk cost fallacy: why you can't quit

"I've come too far to stop now" is one of the most expensive sentences in the language. Here's the bias behind it — and why your past spending should have a vote of exactly zero.
Written & fact-checked by the StupidGames editorial team Last updated: June 2026 About the team
advertisement · 728×90

You're forty minutes into a film that is, objectively, terrible. You know it. You could turn it off right now and spend the rest of the evening doing anything else. But you paid £14 for the ticket, or you've already committed forty minutes, and so you stay — grimly, joylessly, determined to get your money's worth. This is the sunk cost fallacy in its most harmless form. It scales up considerably from there.

What is a sunk cost?

In economics, a sunk cost is any cost that has already been incurred and cannot be recovered. The ticket price is gone whether you walk out at the interval or sit through the credits. The years you've put into a career, a relationship, or a failing business venture are spent whether you continue or not. Past investment does not change the future — it only changes how you feel about the future, which is where the trouble starts.

The sunk cost fallacy is the tendency to factor these irrecoverable costs into decisions about what to do next — to keep throwing resources at something not because the future looks promising, but because of what you've already put in. Rational decision-making says: judge every choice purely on its future costs versus future benefits. The past is a sunk cost. It does not get a vote.

Where it shows up

The fallacy is everywhere once you start looking:

The Concorde fallacy: when governments do it

The most cited large-scale example is the Concorde supersonic passenger jet, which gives the phenomenon an alternative name: the Concorde fallacy. Britain and France began developing Concorde jointly in the 1960s. It became clear fairly early that the aircraft would never be commercially profitable: the market was too small, the operating costs too high, and the economics fundamentally difficult. Yet both governments continued funding it for years, in part because they had already committed so much money and national prestige that stopping felt worse than continuing. The plane flew commercially from 1976 until 2003 and never turned a profit for its operators. The prior investment did not make continuing rational — it just made stopping feel unbearable.

The classic experiment

Psychologists Hal Arkes and Catherine Blumer demonstrated the sunk cost effect cleanly in a series of experiments published in 1985. In one, participants who had paid more for a ski-resort season ticket were more likely to push through bad weather to use it than those who had paid a discounted price — even though the future experience was identical. The only thing that differed was the past cost. That past cost had no logical bearing on whether the trip was worth taking today. But it changed behaviour anyway.

Why the brain falls for it

The sunk cost fallacy is not stupidity. It is the predictable output of two well-documented cognitive tendencies:

Both instincts are useful in many contexts — perseverance and commitment are often genuinely valuable. The problem is applying them to situations where the underlying facts have changed and the honest assessment is that the future no longer justifies the cost.

How to actually escape it

Knowing about the sunk cost fallacy does not automatically make you immune to it — but it gives you something to work with. A few practical approaches:

None of this is a guarantee. The sunk cost pull is strong precisely because it mimics virtues — loyalty, perseverance, not being a quitter. The difference is that genuine perseverance is driven by a clear-eyed assessment of the future. The sunk cost fallacy is driven by the past. The past is a receipt, not a strategy.

Sources & further reading

Related

Is it you, or the money?

Six factors that separate a habits problem from an income problem. Get the verdict.

💰 Take the quiz