well-supported overall; strength varies by decision type and contextEvidence Which past costs are irrecoverable, and would they change what we choose if we evaluated only the future options?
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.
influential and widely estimated; magnitude and robustness debatedEvidence Is stopping being evaluated as a prospective loss from the current reference point, separately from the sunk costs already paid?
Loss aversion is a central component of prospect theory and many studies estimate losses as receiving greater subjective weight than gains around a reference point. However, the effect should not be summarized with one universal coefficient. A 2024 interdisciplinary meta-analysis of 607 estimates reported a mean coefficient near 1.96, while another 2024 meta-analysis of individual risky-choice datasets estimated about 1.31. A 2025 re-analysis of the larger dataset found little evidence of loss aversion in some symmetric, unordered designs, showing that task structure and analysis can materially change the result.
established, but related constructs should be separatedEvidence Are setbacks leading us to commit more resources without reopening whether this course still deserves the next investment?
Sunk-cost effects and escalation of commitment overlap but are not interchangeable. Sunk-cost research asks whether irrecoverable prior investments influence current choices. Escalation of commitment describes persistence or additional resource allocation to a failing course of action and can also be driven by personal responsibility, self-justification, project structure, and other factors.
well-supported for numerical judgments; strength depends on anchor type and contextEvidence Which original target, budget, valuation, or deadline is still pulling the current judgment toward it?
Anchoring is a well-supported effect in which an initial numerical value can pull a later estimate toward it. A 2026 meta-analysis covering 2,601 effect sizes found a large overall effect, but also substantial variation across studies. The effect should not be treated as a rule that every number changes every judgment: incidental anchors, anchors from a different dimension, clearly random values, incentives, and some debiasing conditions were associated with smaller or null effects.
well-supported for time estimates; size and causes vary by contextEvidence What does comparable completed work say about the cost and time still required from today?
The planning fallacy is a well-documented tendency for people to predict their own task completion times too optimistically. The effect has been observed across different kinds of tasks, but it is not a rule that every plan will run late. Project overruns can also come from changing scope, dependencies, incentives, poor data, deliberate underestimation, or genuinely unusual events, so a late project should not automatically be diagnosed as a planning fallacy.
well established, broad constructEvidence What evidence would make us stop or materially change the project, and have we actively looked for it?
Confirmation bias is an umbrella label for several ways existing beliefs or hypotheses can influence information search and interpretation. It should not be reduced to one behaviour such as reading only agreeable news, and a preference for confirming tests is not irrational in every task or environment.
Would we rate the quality of today’s continue-or-stop process the same way if the eventual outcome went the other direction?
Outcome bias occurs when knowledge of a result changes how people evaluate the quality of a decision even when the information available at the time of the decision is held constant. Outcomes can still be relevant for learning, so the error is not 'never look at results'; it is using luck or hindsight as if it had been available to the original decision-maker.
