Sunk Cost

A sunk cost is money, time or effort that has already been spent and cannot be recovered whatever happens next. Examples include non-refundable deposits, research already carried out, advertising already aired and specialized equipment with no resale value. Economists, managers and investors use the concept to separate past spending from the costs and benefits that a decision can still change.

Standard economic reasoning says sunk costs should be ignored when choosing between options, because no choice can bring them back. In practice people often let past spending sway them, a tendency known as the sunk cost fallacy or sunk cost effect, which is studied in behavioral economics. A sunk cost differs from a fixed cost: a fixed cost such as rent may still be avoided in future by ending the lease, while a sunk cost is gone for good.

Suppose you paid for a concert ticket weeks ago, and on the night you feel ill and it is pouring with rain. The money is spent either way; you cannot get it back by going. The only real question is whether the evening at the concert will be better than the evening at home. Yet many people drag themselves out "so the ticket isn't wasted".

The same pull shows up in bigger choices: finishing a degree that no longer suits you, repairing an old car again and again, or staying in a failing project because so much has gone into it. Looking forward, not backward, is the core lesson of sunk costs.

A sound decision compares only future, avoidable costs with future benefits. In capital budgeting, a project's net present value is calculated from cash flows from today onward; money spent on earlier studies is excluded. The relevant costs are the marginal cost of continuing and the opportunity cost of the resources involved, meaning what they could earn in their next best use.

Take a company that has spent $4 million developing a product. Finishing it will cost another $2 million, and the finished product is expected to bring in $1.5 million in today's money. The $4 million is sunk and irrelevant. Spending $2 million to earn $1.5 million destroys $0.5 million, so the rational choice is to stop. If the expected return were $3 million, the right answer would be to continue, again regardless of the $4 million already spent. A full cost-benefit analysis follows the same rule.

Psychologists Hal Arkes and Catherine Blumer reported in 1985 that theater-goers at Ohio University who had paid full price for a season subscription attended more plays in the first half of the season than those randomly given discounts, even though the money was equally spent in both cases. Richard Thaler, awarded the 2017 Nobel Memorial Prize in Economic Sciences, discussed the effect in his work on mental accounting. Biologists Richard Dawkins and Tamsin Carlisle named it the "Concorde fallacy" in 1976, after British and French governments kept funding the supersonic airliner long after its commercial case had weakened.

Researchers link the fallacy to other biases: loss aversion, because abandoning a project turns a paper loss into a real one; the framing effect, because quitting is framed as "wasting" money; the endowment effect; and a wish to appear consistent. Within organizations it overlaps with the principal–agent problem: a manager who launched a project may keep it alive to protect a reputation, even when shareholders would be better off if it were stopped.

Sunk costs also shape how markets work. Industries that require large unrecoverable investments, such as building a factory, a brand or a network, are harder to enter. William Baumol, John Panzar and Robert Willig argued in 1982 that markets with low sunk costs are "contestable": even a monopoly must keep prices close to cost if rivals can enter and leave cheaply. Where sunk costs are high, incumbents are better protected.

Public projects are vulnerable to sunk-cost reasoning. Large infrastructure schemes are often defended on the grounds of money already spent. Investors face the same temptation; venture capital funds and research and development teams try to judge each new round of funding on its own forward-looking merits. Some economists also connect sunk investments to hysteresis, where temporary shocks leave lasting effects because entering or leaving a market cannot be reversed cheaply.

Ignoring sunk costs is sound logic, but honoring past commitments is not always irrational:

• Information. Past spending may reflect earlier research suggesting the project is worthwhile, and that information is still relevant.
• Reputation. Abandoning projects too readily can damage credibility with partners, staff and lenders, a consideration explored in game theory.
• Bounded attention. With bounded rationality, a simple rule of finishing what you start may beat constant re-evaluation.

The lesson is not that past spending tells us nothing, but that it should influence decisions only through what it reveals about the future.

• Arkes and Blumer (1985), The Psychology of Sunk Cost, Organizational Behavior and Human Decision Processes
• Nobel Prize: Richard H. Thaler, Facts