Sunk Cost Effect – when past costs influence the next choice
Sunk Cost Effect is the tendency for money, time, or effort that cannot be recovered to influence what you choose next, even when the future costs and benefits should be enough to make the decision.
Examples and situations
- Projects – a team wants to continue mainly because it has already spent a large budget or many months on the work.
- Purchases and subscriptions – you keep using something you no longer value because you already paid for it.
- Personal plans – you stay with a course of action because stopping would make earlier effort feel wasted.
Questions to check
- Separate what is already gone from what you still have to spend.
- Ask what you would choose today if the past investment belonged to someone else and could not be recovered either way.
- Compare the future value of continuing with realistic alternatives for the remaining time, money, and attention.
- Do not assume that every decision to continue is a sunk-cost error. New information, future value, switching costs, or being close to useful completion can make continuation reasonable.
Long-form guide
Sunk Cost Effect: why past investment keeps voting on the next decision
Money, time and effort that cannot be recovered can still influence what we choose next. The useful correction is not 'always quit.' It is to price the next step from today forward and keep genuinely future value separate from the emotional weight of what is already gone.
On this page
Past cost is real, but it is not always relevant to the next choice
A team has spent eight months building a product. Adoption is weak. The next three months will cost another €200,000. In the meeting, someone says, 'We cannot stop now after everything we have invested.' The past investment is real and painful, but it cannot be recovered by continuing. The decision in front of the team is what to do with the next €200,000 and three months.
This is the core of the sunk cost effect: irrecoverable prior investments influence a later choice even though the relevant economic comparison should focus on future costs and benefits. The effect is well supported overall, but it does not make every continuation irrational. Switching costs, learning already achieved, contractual obligations and the value of completing nearly finished work can all affect future value.
Completion value and sunk cost can be confused
Suppose a project is 95% complete and the final 5% will unlock a working system. Continuing may be sensible because the remaining investment buys something valuable. That is a future-benefit argument. It is different from saying the project deserves more money because 95% has already been spent.
This distinction is practical because 'we are almost done' can contain both ideas at once. Ask what the next increment buys. If the same future benefit could be reached more cheaply by another route, compare those routes. If completion creates no meaningful value, closeness to the finish line is not enough. The goal is to isolate the forward-looking case from the desire to make history feel worthwhile.
Reprice the decision as if it arrived today
A strong review starts with a reset question: if this opportunity appeared on your desk today, with no emotional ownership of the past, would you invest the remaining resources? You should still include assets created so far if they affect the future. Code, contracts, customer relationships and knowledge may have current value. What you exclude is the need to justify resources that are already irrecoverable.
Then compare the best realistic alternative. Time and attention are also scarce, so a project can be expensive even when the next cash payment is small. What else could the team ship? What risk would disappear if the commitment ended? Which option has the highest expected value from today forward? Opportunity cost makes the next decision concrete.
Design exit criteria before the pain arrives
Sunk costs become harder to ignore when the project is tied to identity, public promises or a visible budget. Precommitment helps because the decision rule is written while everyone is still relatively neutral. Define review dates, success thresholds and conditions that trigger a pause, pivot or stop. The criteria can be changed when new information genuinely changes the model, but changes should be explicit.
Independent review can also help. Someone who did not sponsor the original decision has less need to defend it. Ask that reviewer to evaluate only the next tranche of investment. This turns a large emotional question—'was the whole project a mistake?'—into a smaller operational question: 'is the next unit of resource still the best use of what we have?'
A compact example: the annual software contract
A company pays for an analytics platform that few employees use. Because the annual contract was expensive, leaders fund training, internal champions and extra integration work to 'get value from the investment.' Some of that follow-up may be sensible if adoption can realistically create future benefit. But the original license payment is already gone.
The review should compare the cost and expected benefit of the next adoption effort with alternatives such as a smaller tool, a migration or accepting that the original purchase was a loss. Continuing can win that comparison, but only because of the future case. Past spend can explain why the decision hurts; it should not decide what happens next.
Decision checklist
- Write irrecoverable past costs on a separate line from future costs.
- State what the next increment of investment is expected to buy.
- Compare continuation with the best realistic alternative use of money, time and attention.
- Include current assets or completion value only when they affect future outcomes.
- Ask whether you would fund the next step if the project arrived today.
- Use pre-defined review dates and exit criteria for large commitments.
- Have an independent reviewer assess the next tranche when ownership is emotionally strong.
Evidence review
What the evidence supports
The sunk cost effect is a documented tendency for irrecoverable prior investments of money, time, or effort to influence later choices. A meta-analytic review found clear evidence for the effect overall, while also showing that its size and moderators differ between utilization decisions and progress decisions. The effect should not be used to label every choice to continue: future value, switching costs, uncertainty, and how close a project is to useful completion can all be relevant to a forward-looking decision.
How researchers describe the pattern
Classic experiments linked sunk-cost behavior to reluctance to appear wasteful, while later work has connected it to several decision processes rather than one universal mechanism. In project settings, sunk costs can interact with the perceived need to complete the work. That interaction helps explain why sunk costs are one possible antecedent of escalation of commitment rather than a complete explanation for every escalation case.
Practical interpretation
For a current choice, put irrecoverable past costs in a separate line and evaluate the remaining options using costs and benefits from today forward. Compare continuing with the best realistic alternative use of the remaining money, time, and attention. If completion itself creates future value, include that value explicitly instead of treating 'we are already close' as either automatically rational or automatically biased.
Reviewed sources
- The psychology of sunk cost field study and experiments · 1985 · DOI 10.1016/0749-5978(85)90049-4
- Looking forward and looking back: integrating completion and sunk-cost effects within an escalation-of-commitment progress decision experimental study · 2001 · DOI 10.1037/0021-9010.86.1.104
- On the sunk-cost effect in economic decision-making: a meta-analytic review meta-analysis · 2015 · DOI 10.1007/s40685-014-0014-8
Editorial review: 2026-08-18. The controlled evidence class is an editorial reading aid, not a numeric truth score; the descriptive status above carries the entry-specific qualification.
Continue from here