Cognitive Psychology
The anchoring effect is a cognitive bias in which an initial piece of information—the anchor—disproportionately influences subsequent judgments and estimates. Even when the anchor is arbitrary, irrelevant, or obviously uninformative, people adjust insufficiently away from it, producing systematic errors in decision-making across domains such as negotiation, pricing, legal judgments, and medical diagnosis.1
The anchoring effect was first systematically documented by Amos Tversky and Daniel Kahneman in 1974 using a wheel-of-fortune paradigm: participants who spun a rigged wheel landing on 10 or 65 gave median estimates of the percentage of African nations in the UN of 25% and 45%, respectively. Two main mechanisms explain the bias. The anchoring-and-adjustment model posits that people start from the anchor and make insufficient adjustments, stopping as soon as they reach a plausible range. The selective accessibility model suggests that the anchor primes anchor-consistent information in memory, which is then used to construct the estimate. Both mechanisms operate even when the anchor is clearly random, such as a social security number, as shown in a study where participants' willingness to pay for wine was influenced by the last two digits of their ID number.
Anchoring has profound consequences in high-stakes settings. In legal contexts, mock jurors' damage awards are significantly influenced by the plaintiff's requested amount, even when that amount is absurdly high. In negotiations, the first offer serves as a powerful anchor, often determining the final settlement price. In medical diagnosis, a preliminary (even incorrect) diagnosis can anchor clinicians' subsequent reasoning, leading to diagnostic errors. In consumer pricing, retailers exploit anchoring by displaying a high original price next to a discounted one, making the discount appear larger. The effect is robust across cultures, ages, and levels of expertise; even experienced real-estate agents are influenced by listing prices, although they deny it.
Anchoring is not universal; its strength varies with individual and situational factors. People with higher cognitive ability, measured by need for cognition or intelligence tests, are less susceptible, but not immune. Expertise in a domain reduces but does not eliminate the bias. Debiasing attempts have had mixed success: simply warning people about anchoring does not eliminate it, but instructing them to consider the opposite or to generate counterarguments can reduce its effect. Another effective strategy is to use a 'consider-the-opposite' technique, where decision-makers are asked to list reasons why their estimate might be wrong. In group settings, anchoring can be attenuated by having members independently generate estimates before sharing them, a process known as the Delphi method.
Beyond the classic demonstrations, anchoring appears in surprising contexts. In the courtroom, even irrelevant anchors like the number of years a judge had served on the bench influenced sentencing decisions in a study of German judges. In the field of forecasting, the 'anchoring trap' is a known pitfall in project management, where initial cost estimates become self-fulfilling. The effect also operates in the opposite direction: 'reverse anchoring' occurs when a low anchor leads to higher estimates, as seen in some consumer contexts. A notable edge case is the 'anchoring effect in memory': once an anchor is processed, it can bias the recall of past events, such as remembering one's own past grades as closer to an anchor than they actually were. The effect is also observed in non-human animals, suggesting a deep evolutionary origin; for example, pigeons show anchoring-like behavior in numerical discrimination tasks.
Anchoring is one of the most robust and replicable findings in behavioral economics, with implications for policy, marketing, and law.
Help improve the encyclopedia. Reports go straight to the site manager.