A sunk cost is an expenditure already incurred that cannot be recovered through a current or future decision. In economics, it is distinguished from costs that can still be avoided or changed. Because an irrecoverable expenditure remains the same whichever alternative is chosen, it does not affect the relative value of those alternatives. The concept also applies to investments of time and effort, and is important in microeconomics, managerial accounting, and behavioral economics. (openstax.org)
Definition and related costs
The defining feature of a sunk cost is irrecoverability, not simply that a payment occurred in the past. An asset purchased earlier may still be sold or put to another use. Its original purchase price is historical, but its recoverable value remains relevant to a present decision. A nonrefundable payment for an already completed service, by contrast, cannot be restored by changing course. Managerial accounting therefore separates historical expenditure from future cash flows that differ between alternatives. (openstax.org)
A sunk cost is not synonymous with a fixed cost. Fixed costs do not vary with the level of output over the period considered; sunk costs cannot be recovered. Many short-run fixed costs are sunk, but the two classifications answer different questions. A payment may be fixed relative to output yet recoverable through resale or cancellation. Conversely, expenditure on inputs that originally varied with production becomes sunk once those inputs have been consumed without recoverable value. (openstax.org)
An opportunity cost is the benefit forgone by choosing an action instead of its next-best alternative. Unlike sunk expenditure, it concerns a choice still available. Continuing to use a saleable machine, for example, forgoes the proceeds from selling it. Likewise, time already spent cannot be recovered, whereas additional time devoted to a project displaces other possible activities. (books.core-econ.org)
Role in economic decisions
The sunk-cost principle evaluates alternatives by their incremental consequences. In decision theory, the relevant comparison is between benefits and costs that change with the decision, rather than an attempt to recover expenditure that cannot be altered. This logic is consistent with constrained choice: available actions are compared according to their consequences and the opportunities they displace. (openstax.org)
For illustration, suppose a project has already consumed $100,000, all irrecoverable. Completing it would cost another $20,000 and generate $30,000, while abandonment would generate no additional receipts or costs. Completion adds $10,000 relative to abandonment, despite leaving the project with an overall loss. If completion instead generated only $15,000, abandonment would yield the better outcome under those assumptions. The past $100,000 appears equally in both comparisons and cancels out. This is an application of the relevant-cost principle, not a claim about a particular project. (openstax.org)
Ignoring sunk expenditure in this comparison does not imply ignoring all history. Earlier spending may have created assets or information that changes present alternatives. The distinction is between the expenditure itself and consequences that remain relevant. A financial loss on the project as a whole also differs from the incremental gain obtainable by continuing it. (openstax.org)
Production and shutdown
The concept explains why a firm can rationally continue production while reporting an economic loss. In the standard short-run model of perfect competition, unavoidable fixed costs remain payable even if output falls to zero. Production can reduce the total loss when total revenue exceeds avoidable operating costs, because the surplus contributes toward those fixed costs. (openstax.org)
In that model, the shutdown threshold is the minimum average variable cost. A profit-maximizing firm produces where price equals marginal cost, provided operating is preferable to shutting down. A price below average total cost therefore does not automatically imply immediate shutdown. At the minimum average variable cost, the firm is indifferent between operating and shutting down under the model’s assumptions. Long-run exit differs because commitments that are unavoidable today may become avoidable when contracts expire or assets can be adjusted. (openstax.org)
The sunk-cost effect
The sunk-cost fallacy occurs when irrecoverable investment itself is used to justify continuation, rather than the prospective value of continuing. It is studied in psychology as the sunk-cost effect: a greater tendency to persist after investing money, effort, or time. Hal Arkes and Catherine Blumer’s 1985 paper, “The Psychology of Sunk Cost,” presented experimental evidence and identified a desire not to appear wasteful as a psychological explanation. (sciencedirect.com)
Their field study found that theater subscribers who initially paid more attended more performances during the following six months. Questionnaire studies also found that prior investment could influence continuation decisions and estimates of project success. These findings concern the effect of past investment on subsequent choice, not merely the existence of an unsuccessful outcome. (cognition.aau.at)
The phenomenon overlaps with escalation of commitment, in which decision-makers continue supporting an unsuccessful course of action. Nevertheless, persistence alone does not establish a fallacy: additional investment may still improve the outcome relative to stopping, as the incremental project example illustrates. (openstax.org)
Market entry and irreversibility
Before expenditure occurs, potentially sunk investment remains relevant. Prospective entrants must consider the risk that setup expenditure or specialized assets will not be recoverable on exit. High sunk costs can therefore create barriers to entry, even though an incumbent should not treat already irrecoverable spending as a reason to continue operating. Recoverability depends partly on asset specificity and the availability of resale or rental markets. These entry risks are distinct from economies of scale, although both can influence the feasibility of entering an industry. (accc.gov.au)