... like I'm 5 years old
You buy a ticket to a three-hour movie. After an hour, you are bored and want to leave. However, you stay because you paid $20 for the ticket. That is the sunk cost fallacy: allowing something you have already spent to control what you do next.
A sunk cost is money, time, or effort that has already been used and cannot be recovered. Whether you stay or leave, the ticket money is gone. The sensible question is not, “How much have I spent?” It is, “Will the next two hours be worth my time?”
The fallacy appears everywhere. Someone may finish a meal they no longer want because it was expensive, continue reading an unenjoyable book because they have reached page 200, or remain in a failing project because years of work have gone into it. Continuing can feel like protecting the original investment, but it may simply add another cost.
This does not mean people should abandon anything that becomes difficult. Persistence is valuable when the likely future benefits justify the future effort. The mistake is treating the past investment as a reason to continue, even when stopping would now be better.
A useful test is to imagine that you inherited the situation today. If you had invested nothing so far, would you still choose to continue?
The sunk cost fallacy is like missing your highway exit and continuing in the wrong direction because you have already driven so far. The distance behind you cannot be recovered; what matters is the best route from where you are now.
... like I'm in College
Imagine a manager overseeing a product that customers do not want. Early testing is poor, costs are rising, and a better opportunity has appeared. Yet the manager approves another round of funding, saying, “We cannot waste everything we have already invested.”
The problem is that the earlier spending cannot be changed. A rational decision compares the future benefits and costs of continuing with those of stopping or choosing an alternative. Past, unrecoverable expenditure should not determine that comparison.
The sunk cost fallacy is studied within behavioral economics, which examines how psychology shapes real decisions. People often dislike waste, resist admitting mistakes, and become emotionally attached to choices they have defended. Continuing allows them to postpone the painful moment when an investment must be recognized as lost.
This pattern can develop into escalation of commitment. A person first spends money, then adds time to justify the money, and finally spends more money to justify both. Each new investment makes withdrawal feel harder, even as the case for continuing weakens.
A practical defense is to schedule decision points before beginning. At each point, ask:
- What evidence would make me stop?
- What will continuing cost from today onward?
- What alternatives am I giving up?
- Would I recommend continuing if someone else had made the original decision?
Like procrastination, the sunk cost fallacy shows how immediate emotional relief can defeat better long-term judgment. Stopping may feel like failure now, while continuing delays that discomfort—even when it creates a larger failure later.
Imagine that Maya has spent an evening building a Lego spaceship. She has used 600 bricks, but the structure is unstable. The wings keep falling off, the cockpit will not fit, and she has discovered that several essential pieces are missing.
Beside her sits a smaller spaceship design that she could complete with the bricks she still has. Nevertheless, Maya keeps repairing the original model. She tells herself that dismantling it would waste the 600 bricks and three hours already invested.
Those bricks represent sunk costs. Their earlier use cannot be undone, even if Maya takes the model apart. Her three hours will not return either. The question is what she can build from this moment forward.
Suppose repairing the old spaceship requires another 200 bricks and two frustrating hours, with little chance of success. The new design requires 100 bricks and one enjoyable hour. Choosing the smaller spaceship does not recover the evening, but it produces the better future outcome.
Now change the story. Maya discovers that one missing connector will fix the entire model in five minutes. Continuing is then sensible—not because she has already used 600 bricks, but because the future cost is small and the expected benefit is high.
That distinction is the whole lesson. Do not tear down every difficult Lego model, project, career plan, or relationship merely because it has problems. Instead, temporarily ignore the structure already built. Look at the unused bricks, the available designs, and the time remaining. Then choose the next step that offers the strongest future—not the one that best defends the past.
... like I'm an expert
Consider an investor evaluating an underperforming venture. Under a standard forward-looking decision model, an irrecoverable expenditure is excluded from the choice set’s marginal analysis. The decision should depend on expected future payoffs, opportunity costs, risk, and the value of available alternatives—not on an expenditure invariant across the remaining options.
Behaviorally, however, prior investment can alter the decision maker’s reference point. Abandonment converts an ambiguous, potentially recoverable setback into a psychologically realized loss. Within a prospect-theoretic interpretation, remaining in the loss domain may encourage risk-seeking continuation: the agent accepts additional exposure for a chance to avoid acknowledging the loss. Waste aversion, mental accounting, cognitive dissonance, reputational concerns, and self-justification can reinforce the effect.
Hal Arkes and Catherine Blumer’s influential 1985 study of sunk-cost psychology defined the effect as an increased tendency to continue after investing money, effort, or time. In a field experiment, theater subscribers who initially paid more attended more performances during the following six months, consistent with pressure to obtain value from the expenditure.
Not every investment-dependent choice is fallacious. Prior expenditure may produce learning, complementary assets, contractual obligations, switching costs, or information about project quality. Withdrawal can also affect reputation or future cooperation. These are prospective consequences and therefore belong in the analysis.
The diagnostic question is counterfactual: holding all future consequences constant, would the decision change if the earlier investment had never occurred? If yes, sunk-cost sensitivity is probably influencing the choice. Good governance reduces that influence through independent review, predefined termination criteria, staged funding, and separation between project advocates and continuation decisions.