Sunk cost

"Sunk Costs" redirects here. For the TV series episode, see Sunk Costs (Better Call Saul).

In the fields of economics and business decision-making, a sunk cost—sometimes called a retrospective cost—refers to an expense that has been paid and cannot be recovered. This contrasts with prospective costs, which represent future expenses that can be avoided through certain actions. Simply put, sunk costs are past expenditures that should not influence future decision-making. Despite this economic principle, individuals frequently factor in previous spending, such as on vehicle or home repairs, when making ongoing decisions about those assets.

Bygones principle

Within classical economics and mainstream microeconomic frameworks, rational choices are guided solely by prospective (future) costs. Decisions at any point rely exclusively on the current options available. What truly matters are the future outcomes. Past errors hold no bearing on present decisions. Costs already incurred before deciding are unavoidable regardless of the chosen path. These can be likened to "water under the bridge", and basing decisions on them is often described as "crying over spilt milk". Thus, rational decision-making dictates ignoring sunk costs. For instance, if a factory initially estimated to cost $100 million and produce $120 million in value has already consumed $30 million but its projected value drops to $65 million, abandoning the project is advisable rather than investing an additional $70 million. Conversely, if the value forecast only declines to $75 million, continuing the project aligns with rational behavior. This concept is known as the bygones principle or alternatively the marginal principle.

The bygones principle finds its foundation in normative decision theory, specifically within rational choice theory and the expected utility hypothesis. Expected utility theory depends on a property termed cancellation, which asserts that rational decision-making involves disregarding any scenario yielding identical outcomes regardless of choice. Since past choices—including sunk costs—fit this criterion, they should be excluded from consideration.

Another formalization of the bygones principle is "separability". This requires decision-makers to evaluate options based solely on possible future events unaffected by prior history. In terms of decision trees, it means that choices at a given node must not depend on unreachable branches of the tree. This concept is pivotal in economic theory, underpinning methods such as the folding-back algorithm for sequential decisions and game-theoretic ideas like sub-game perfection.

Until resources are irrevocably allocated, prospective costs remain relevant and avoidable, thus warranting inclusion in decision analysis. For example, contemplating pre-purchasing movie tickets but not yet buying them means the cost is still avoidable.

Both retrospective and prospective expenses may be classified as either fixed costs—constant regardless of output—or variable costs—dependent on production volume. However, many economistsTemplate:Who argue against labeling sunk costs as either fixed or variable. For instance, a $400 million one-time payment for enterprise software installation is sunk because it cannot be recovered once spent. In contrast, fixed costs would include ongoing monthly fees for service contracts or licenses related to the software. The initial irretrievable payment should not be categorized as fixed costs amortized over time; instead, sunk costs warrant separate treatment. Variable costs might include expenses such as data center electricity usage linked to the project.

There are scenarios where factoring sunk costs into decisions—thereby contravening the bygones principle—is justifiable. For example, a manager aiming to demonstrate persistence or avoid blame for prior errors may rationally continue a project despite it being unprofitable for their company. Similarly, if possessing private information about why abandoning a project is undesirable, persisting with it can be rational even if outsiders perceive it as falling prey to sunk cost fallacy.

Fallacy effect

The principle of disregarding past losses does not always align with actual human conduct. Often, sunk costs exert a significant influence on decision-making, as individuals tend to believe that previous investments justify further commitments. This phenomenon, known as the sunk cost fallacy, describes the tendency to persist with an endeavor after money, effort, or time has already been spent. Such behavior is colloquially referred to as "throwing good money after bad," reflecting a reluctance to "cut one's losses." For instance, people may stay in failing relationships due to the amount they have already invested, or continue supporting wars because of lives already lost. Similarly, victims of manipulative schemes frequently continue dedicating resources despite growing doubts. These patterns deviate from rational choice theory and are often categorized as behavioral mistakes.

Research by Rego, Arantes, and Magalhães highlights the sunk cost effect within committed relationships. Their studies revealed that individuals who have invested money and effort into a relationship are more inclined to maintain it rather than end it. Additionally, those who have devoted considerable time tend to invest even more time in the relationship. This demonstrates how the sunk cost fallacy can influence personal decisions, although ideally, future choices should be made rationally without regard for irrecoverable past investments.

Evidence from De Bondt and Makhija (1988) indicates that many U.S. utility managers hesitated to halt economically unviable nuclear power projects. During the 1960s, nuclear energy was heralded as "energy too cheap to meter," but public support waned in the following decades amid regulatory prudency reviews. These reviews often denied utilities recovery of construction costs due to mismanagement consistent with throwing good money after bad.

The sunk cost fallacy is also known as the "Concorde fallacy": British and French governments continued funding the expensive supersonic jet project despite clear economic impracticality instead of abandoning it.

Government officials have similarly demonstrated difficulty ignoring sunk costs. The term "Concorde fallacy" originates from the British and French governments' persistence in financing the costly Concorde supersonic aircraft project even after its commercial viability was lost. Internally, the British government regarded the project as a financial failure that should have been terminated earlier; however, political and legal constraints prevented withdrawal.

In business contexts, sunk costs frequently arise in decisions such as brand promotion expenses—investments that cannot typically be recovered or reversed—and research and development (R&D) expenditures. Once incurred, these costs should not influence pricing or continuation decisions. For example, a pharmaceutical company justifying high prices solely based on R&D recoupment falls prey to this fallacy; pricing would remain high regardless of actual R&D spending. Instead, R&D expenses should factor into initial investment decisions rather than future pricing.

Dijkstra and Hong's research suggests that emotional states significantly affect susceptibility to the sunk cost fallacy. Their experiments found that negative emotions like anxiety increase stress related to sunk costs, leading individuals under pressure to persist with failing projects rather than explore alternatives. Thus, psychological condition and external factors play crucial roles in influencing this behavior.

The sunk cost effect can contribute to cost overrun situations. For example, if $20 million has been spent on an incomplete power plant now valued at zero with no resale options, completing it might require an additional $10 million while constructing an alternative facility would cost $5 million. Rationally, abandoning the original project and building the new one is preferable despite total loss of prior investment—the initial expenditure represents a sunk cost. However, decision-makers may irrationally choose completion due to incentives such as avoiding visible losses or political repercussions. Such scenarios illustrate an incentive problem, distinct from sunk cost issues themselves. Moreover, some studies observe cases where individuals prematurely abandon prior investments in favor of new ventures, exhibiting a reversed sunk cost effect.

Plan continuation bias

Plan continuation bias is a closely related cognitive distortion that compels individuals to persist with an original plan or course of action despite evolving circumstances. In aerospace, this bias has been identified as a major contributor to accidents; a 2004 NASA investigation revealed that in nearly half of the 19 accidents examined, flight crews demonstrated this tendency.

This risk is particularly relevant for ship captains and aircraft pilots who may adhere rigidly to their initial plans even when such adherence leads toward catastrophic outcomes, when instead they should consider aborting the operation. A notable incident illustrating this bias is the Torrey Canyon oil spill, where the tanker ran aground because its captain chose to continue on a hazardous route rather than accept delays. The bias has also played a role in numerous aviation disasters; an analysis of 279 approach and landing accidents (ALAs) ranked it as the fourth most frequent cause, responsible for 11% of cases. Furthermore, another study covering 76 accidents found plan continuation bias to be a contributing factor in 42% of them.

Two primary elements characterize this bias: first, an excessively optimistic belief in the likelihood of success, which may serve to alleviate cognitive dissonance associated with decision-making; second, the influence of personal accountability, making it challenging for individuals to acknowledge errors when they bear responsibility.

Project management frequently experiences cost overruns and schedule delays due to the planning fallacy and associated biases. These include undue optimism, reluctance to concede failure, groupthink dynamics, and loss aversion related to sunk costs.

Psychological factors

The psychology behind cognitive bias includes a group of "need to act quickly" biases (light green) that influence economic rationality. [click twice to browse].

Research in behavioral economics identifies at least four key psychological elements that contribute to the sunk cost phenomenon:

• Framing effects, a cognitive distortion where individuals make decisions based on whether options are framed with positive or negative connotations, such as gains versus losses. People generally exhibit risk aversion when the scenario is positively framed but tend to pursue risks under negative framing.
• An overly optimistic bias regarding probabilities, leading individuals to inflate the expected returns of an investment after committing resources.
• A sense of personal accountability, where the sunk cost effect predominantly influences those who feel personally responsible for prior investments.
• The reluctance to appear wasteful—individuals may continue investing to avoid admitting that previous expenditures were squandered.

Collectively, these findings imply that the sunk cost effect may stem from non-traditional interpretations of utility, which are inherently subjective and vary by person.

Framing effect

The framing effect linked to the sunk cost phenomenon builds on the principle of extensionality, which states that outcomes should remain constant regardless of how information is presented. This contrasts with intentionality, which concerns whether the way information is conveyed alters the context or decision.

Consider two mathematical expressions:

1. Template:Math
2. Template:Math

Although these formulas are framed differently, they yield identical results for any input "x". Therefore, a rational decision-maker should be indifferent between them. Nevertheless, framing effects introduce biases that cause preferences to diverge despite equivalent outcomes.

The most studied form of framing was proposed by Kahneman & Tversky in 1979 as valence framing effects. This type distinguishes between positive and negative frames: a "sure thing" option framed positively emphasizes gains, whereas the same option framed negatively emphasizes losses, even though both are logically equivalent. For instance, rescuing 200 out of 600 people from a sinking ship is mathematically identical to losing 400 people; the former represents a positive frame, the latter a negative one.

Ellingsen, Johannesson, Möllerström, and Munkhammar have classified framing effects within social and economic contexts into three main theoretical categories. First, the variable sociality hypothesis suggests that how options are framed can influence internalized social norms or preferences. Second, the social image hypothesis posits that framing affects how decision-makers believe others perceive them, thereby shaping their behavior. Finally, framing can alter expectations about others’ actions, which subsequently impacts one’s own decisions.

Overoptimistic probability bias

In 1968, Knox and Inkster conducted a study involving 141 horse race bettors. Among them, 72 individuals had just placed a $2.00 wager within the preceding 30 seconds, while 69 were about to make a $2.00 bet in the following 30 seconds. The researchers proposed that those who had recently committed to their bet would alleviate post-decision dissonance by increasingly believing that their chosen horse was likely to win. Participants were asked to evaluate their horse's winning probability on a seven-point scale. Results showed that bettors yet to place their bet rated the likelihood at an average of 3.48, indicating a "fair chance of winning," whereas those who had just wagered gave an average rating of 4.81, corresponding to a "good chance of winning." This supported the hypothesis that confidence in the outcome rises immediately after making a financial commitment. Knox and Inkster further validated their findings by applying a similar analysis to the horses' owners, obtaining nearly identical results after adjustments. Additional research has also documented tendencies toward overestimating probabilities.

Sense of personal responsibility

Staw and Fox examined the influence of personal accountability in a study involving 96 business students. Participants were asked to decide whether to allocate funds to research and development (R&D) in either an underperforming division or other parts of a hypothetical company. The subjects were split into two groups based on responsibility levels: high and low. Those in the high responsibility group were informed that they, acting as managers, had previously made an unsuccessful R&D investment in the struggling division. Conversely, participants in the low responsibility group were told that a prior manager had made the earlier investment, with both groups receiving identical profit information. Each participant was then tasked with deciding on a new $20 million investment. Findings revealed a significant interaction between perceived responsibility and investment amount; individuals in the high responsibility condition committed an average of $12.97 million, whereas those in the low responsibility condition invested an average of $9.43 million. Comparable outcomes have been observed in related studies.

Reluctance to Seem Wasteful

Purchasing a ticket ahead of an event and later finding it unenjoyable creates a partial public commitment to attending. Exiting prematurely reveals this misjudgment to others, an exposure the individual might prefer to avoid. Additionally, having already paid, the attendee may hesitate to leave early to avoid feeling that their money has been squandered. Conversely, some may derive a sense of pride from recognizing the opportunity cost associated with how else their time could have been spent.

Insights from Neuroeconomics and Neuroscience

Recent years have seen renewed interest in exploring how the brain processes sunk cost information. Laboratory assessments of sunk cost sensitivity are often complicated by the challenge of separating sunk cost effects from anticipated future returns. In a comparative study involving humans, rats, and mice, Sweis et al identified an evolutionarily conserved sensitivity to sunk costs across these species.

This discovery prompts further investigation into the evolutionary reasons behind the brain's capacity for such processing, the potential benefits of sunk cost sensitivity, and the neural pathways involved. Research is examining how specific brain circuits generate these valuation patterns depending on question framing, environmental context, or individual condition. Current studies focus on how neurons encode sunk cost sensitivity, why such sensitivity emerges only after certain decisions, and its possible influence on emotional states.

See also

• Disposition effect
• Endowment effect
• Escalation of commitment
• Region-beta paradox
• Foot-in-the-door technique
• Prospect theory
• Psychology of previous investment
• Stop-loss order
• Thinking, Fast and Slow

Further reading

• Amankwah-Amoah, J. (2014). "A unified framework of explanations for strategic persistence in the wake of others' failures." Journal of Strategy and Management, 7(4), 422–444. doi:10.1108/JSMA-01-2014-0009
• Arkes, H.R., & Ayton, P. (1999). "The Sunk Cost and Concorde effects: are humans less rational than lower animals?" Psychological Bulletin, 125(5), 591–600. doi:10.1037/0033-2909.125.5.591
• Bade, Robin; Parkin, Michael. Foundations of Microeconomics. Addison Wesley Paperback, 1st Edition, 2001.
• Bernheim, D., & Whinston, M. Microeconomics. McGraw-Hill Irwin, New York, NY, 2008. ISBN 978-0-07-290027-9.
• Doody, Ryan (2020). "The Sunk Cost 'Fallacy' Is Not a Fallacy." Ergo: An Open Access Journal of Philosophy, 6(40), 1153–1190. Available at: http://rdoody.com/SunkCostFallacyIsNotaFallacy.pdf
• Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. ISBN 978-0374275631.
• Klein, G., & Bauman, Y. The Cartoon Introduction to Economics Volume One: Microeconomics. Hill and Wang, 2010. ISBN 978-0-8090-9481-3.
• Samuelson, Paul; Nordhaus, William. Economics. McGraw-Hill International Editions, 1989.
• Sutton, J. Sunk Costs and Market Structure. MIT Press, Cambridge, MA, 1991. ISBN 0-262-19305-1.
• Varian, Hal R. Intermediate Microeconomics: A Modern Approach, Fifth Edition. New York, 1999. ISBN 0-393-97830-3.