Sunk Cost Fallacy

• Bias: Continuing to invest resources (time, money, effort) in a project or decision solely because a substantial amount of those resources has already been committed, even when current costs exceed benefits.
• What it breaks: Rational decision‑making, assessment of current alternatives, ability to timely discontinue unprofitable projects, efficient allocation of resources.
• Evidence level: L1 — multiple laboratory experiments, interdisciplinary research in psychology and economics, documented mechanisms through loss aversion and emotional reactions.
• How to spot in 30 seconds: You justify continuing the action with phrases like “I’ve already invested so much,” “It would be a waste to quit after all this effort,” “We have to see it through since we started” — instead of analyzing future prospects.

Why do past investments control our future?

The sunk‑cost fallacy is a cognitive bias in which people make irrational decisions by considering factors other than current alternatives and future prospects (S001). It is a systematic deviation from rational economic behavior, where ideally only future costs and benefits should be taken into account, and past investments—being sunk—should not influence the current choice. The phenomenon manifests as individuals continuing to invest in a venture with a low probability of payoff solely because investments have already been made (S004).

This bias is most common in contexts of financial investing, project management, personal relationships, and career decisions (S005). People keep putting money into an unprofitable business, stay in unsatisfying relationships, see a hopeless project through to the end, or continue watching a boring movie—all because they have already invested time, money, or emotional energy. The sunk‑cost fallacy affects individuals across all levels of cognitive ability and expertise, making it a universal systematic bias (S002).

There is a substantial interdisciplinary gap in understanding this phenomenon. Psychologists widely acknowledge the sunk‑cost effect as a robust phenomenon supported by numerous studies, whereas economists find limited support for this effect in controlled experiments (S003). This divergence suggests methodological differences: psychologists often examine real‑world decision contexts where emotional and social factors play a significant role, while economists create tightly controlled laboratory settings with clear monetary incentives.

A key distinction should be made between the “sunk‑cost effect” and the “sunk‑cost fallacy.” The former term describes a broader behavioral pattern in which past investments influence current decisions, which can be separated from the “fallacy” aspect that implies irrationality. Some research suggests that the sunk‑cost effect may represent an optimal response to memory constraints in sequential investment models rather than pure irrationality (S008).

The practical importance of understanding this bias is hard to overstate. The sunk‑cost fallacy leads to inefficient allocation of resources at both personal and organizational levels (S007). In a business context this means continuing to fund failing projects; in personal life, maintaining toxic relationships; in education, persisting in a field of study that no longer matches one’s interests. Recognizing the mechanisms of this bias is critically important for improving decision‑making processes in business management, personal finance, and project management.

The connection with other cognitive biases complicates the decision‑making picture. Illusion of control often amplifies the sunk‑cost effect, leading people to believe they can “save” a project if they keep investing. Hindsight bias can cause an overestimation of the original decision, making it harder to recognize the error and abandon the project. Outcome bias leads people to judge the quality of a decision by its result rather than by the information available at the time of the decision.

How Past Investments Capture Future Decisions

Loss Aversion: An Emotional Anchor

The core of the sunk‑cost fallacy is loss aversion—a psychological mechanism whereby the pain of a loss feels stronger than the pleasure of an equivalent gain (S004). Once we have already invested resources, terminating a project is perceived as crystallizing the loss, generating strong emotional discomfort. Continuing to invest allows us to preserve hope of recouping past expenditures, postponing the moment of admitting failure.

Emotional investment—time, effort, and psychological energy—acts just as powerfully as financial investment (S005). When a person feels negative emotions linked to the prospect of “wasting” already‑spent resources, this creates additional pressure to continue an ineffective course of action.

Cognitive Patterns That Lead Us Astray

The intuitive appeal of the sunk‑cost fallacy is rooted in social norms of consistency: we are taught to “see things through” and not be “quitters.” Abandoning a project after substantial investment is perceived as admitting failure, threatening self‑esteem and social standing.

The illusion of control also operates—belief that additional effort can change the situation even when the data say otherwise. The escalation of commitment effect creates a self‑reinforcing cycle: the more invested, the stronger the psychological pressure to continue.

Cognitive Factor	Mechanism	Outcome
Loss Aversion	Loss pain > gain pleasure	Continuation of losing projects
Social Norm	Pressure to “see it through”	Avoidance of admitting failure
Illusion of Control	Belief that effort can change the situation	Overestimation of success likelihood
Escalation of Commitment	Increasing psychological pressure	Self‑reinforcing investment cycle

An Alternative View: Rationality Under Constraints

Limited‑memory theory offers an unexpected perspective: the sunk‑cost fallacy may be an optimal response to cognitive constraints (S008). When an agent cannot perfectly recall the entire history of decisions and outcomes, using past investments as a heuristic can be a rational strategy. What appears as a bias may actually be an adaptive mechanism that evolved in response to the real limitations of the human brain.

In a sequential investment model, reliance on past expenditures serves as a proxy for assessing opportunity quality when full information is unavailable. As new information arrives, this mechanism helps navigate uncertainty, albeit at the cost of occasional errors.

The Universality of the Effect: When Intelligence Doesn’t Save You

Laboratory experiments have shown that even individuals with high cognitive abilities are susceptible to the sunk‑cost fallacy (S002). This underscores that the bias is not a product of lacking intelligence or education, but a fundamental feature of human cognition.

The bias operates at a level that often bypasses rational analysis, activating emotional and intuitive decision‑making systems before analytical thinking kicks in. This explains why confirmation bias frequently amplifies the effect: we seek information that validates the desirability of continuing investment, ignoring contradictory data.

Decision-making, economic behavior, project management

Examples of the Sunk Cost Fallacy in Real Life

Scenario 1: Personal Finance and Leisure

Anna bought an annual membership at a fitness club for $300 in January, motivated by New Year's resolutions. She attended regularly for the first two months, but then her workout intensity began to drop. By April she realized the club was too far from her new workplace, the workouts took too much time given the commute, and she found a free sports area near her home that suited her better.

However, Anna forces herself to keep traveling to the expensive club, experiencing stress and losing time because she feels “bad about the money spent” and “needs to get her money’s worth” (S007). A rational analysis shows that the $300 has already been spent and cannot be recovered regardless of whether Anna continues to attend the club — it is a sunk cost. The current decision should be based on a comparison: what is better for her health, time, and enjoyment now — continuing the inconvenient trips to the club or switching to the more convenient option near her home.

By continuing to go to the club solely because of the money already spent, Anna adds to her financial loss the loss of time, energy, and a decline in quality of life. This is a classic example of how the sunk cost fallacy leads people to persist with ventures even when current costs outweigh benefits (S007). A similar situation occurs with entertainment: a person keeps watching a boring movie in a theater because they “already paid for the ticket,” even though they could spend the remaining time on something more enjoyable.

The ticket is already purchased, the money spent — that’s a sunk cost. The decision to stay or leave should be based solely on whether watching the rest of the film will bring more pleasure than alternative ways of spending the time. Emotional reactions and negative influences amplify this error (S006), creating a sense of guilt over “wasted money” that outweighs a rational assessment of the current situation.

Scenario 2: Business and Project Management

TechNova invested $500,000 and two years of work by a team of 20 developers in creating a new software product. During beta testing it became clear that the market had shifted: competitors released a more advanced solution, the target audience lost interest in this type of product, and the technology platform on which the project was built was becoming obsolete. Market research shows that even with a successful launch the product is unlikely to recoup the additional investments needed to bring it to market readiness and promote it (S005).

However, the company’s leadership decides to continue the project, arguing that “so much has already been invested that we can’t stop,” “we can’t let two years of work go to waste,” and “we’re almost there, just a little more to go.” The company allocates an additional $200,000 for final development and marketing. The product launches, receives a weak market response, and is shut down after six months, incurring losses.

In the end the company lost not $500,000 but $700,000, plus missed opportunities — the team’s resources could have been redirected to more promising projects (S005). This scenario illustrates how the sunk cost fallacy manifests in a corporate context where decisions are made collectively and should be rational. Psychologists widely acknowledge the robustness of this effect in real business situations (S003), although economic models assume rational agents should not succumb to this bias.

An additional factor in the organizational context is the fear of admitting error: managers who launched the project worry that its termination would be seen as a personal failure, creating extra pressure to continue the losing course (S004). This is linked to self‑serving attribution, where people tend to credit successes to their abilities and blame failures on external circumstances.

Scenario 3: Politics and Public Projects

The regional government began construction of a large sports complex for international competitions, investing $50 million from the budget. Three years into construction it became clear that the international federation had changed venue requirements, the original estimate was three times too low, and an economic downturn had sharply reduced fiscal capacity. Independent experts calculated that completing the project would require an additional $100 million, and that once finished the facility would generate annual losses due to high operating costs and low demand for such a large sports venue in the region.

Despite this data, the political leadership insists on continuing construction, publicly stating that “we cannot let the billions already spent disappear” and “we must see the project through for the region’s prestige.” The media actively back this stance, appealing to voters’ emotions: “Will we really let our tax dollars go to waste?” This is a classic example of how the sunk cost fallacy operates at the level of public discourse, where emotional reactions benefit from the bias (S006), and rational analysis of future costs and benefits is replaced by a focus on past investments.

A rational approach would compare two scenarios: halt the project now, losing $50 million but preserving $100 million for other regional needs, or spend another $100 million to obtain a facility that will generate ongoing losses. The initial $50 million is a sunk cost—it is lost either way and should not influence the decision. Yet political logic, fear of being accused of “wasting funds,” and the escalation of commitment effect lead to continued financing of the failing project (S001).

As a result the region loses not $50 million but $150 million, ending up with a loss‑making facility that requires continual budget subsidies. This inflicts even greater damage on the regional budget and citizens’ welfare, funds that could have been used for education, healthcare, or infrastructure. Such cases demonstrate how the anchoring effect on initial investments can lead to massive economic losses at the level of public policy (S004).