Anchoring Bias

Definition and Basic Phenomenon

Anchoring bias is a systematic tendency for people’s numerical estimates and evaluative judgments to be pulled toward an initial value or anchor, even when that anchor is arbitrary, random, or obviously uninformative. Once an initial number or starting point is introduced—such as a suggested price, an opening offer, or a previously stated guess—subsequent reasoning typically involves insufficient adjustment away from this reference point.

In everyday contexts, anchoring can be seen when a high list price makes a sale price seem more reasonable, when the first salary offer shapes later negotiation, or when a preliminary forecast influences revised predictions. The bias illustrates how human cognition often relies on heuristics rather than recalculating from first principles.

Experimental Origins and Classic Findings

Anchoring bias was systematically described by Amos Tversky and Daniel Kahneman in the 1970s as part of the heuristics and biases program in cognitive psychology and behavioral economics. Their early experiments used random anchors to show how robust the effect is.

A classic paradigm involves asking participants to spin a wheel of fortune rigged to produce a random number (e.g., 10 or 65), then asking whether the percentage of African countries in the United Nations is higher or lower than this number, followed by a numerical estimate. Participants who saw a higher random number gave substantially higher estimates than those who saw a lower one, despite being explicitly told that the wheel outcome was random.

Subsequent studies extended these findings:

• In negotiation, initial offers strongly influence final agreements, even when negotiators are highly skilled and aware of potential bias.
• In legal contexts, recommended sentences (even when labeled as random or coming from non-experts) can influence judges’ sentencing decisions.
• In consumer behavior, list prices, “was/now” discounts, and suggested reference prices powerfully shape perceived value and willingness to pay.

These results suggest that anchoring is not easily eliminated by expertise or by simply warning individuals, indicating that it is a deep and pervasive feature of human judgment.

Theoretical Explanations

Multiple theoretical accounts aim to explain why anchoring occurs and why adjustment tends to be insufficient:

1. Heuristic Adjustment Account
   Early work by Tversky and Kahneman proposed that people start from the anchor and engage in insufficient adjustment. Because adjustment is effortful and often truncated by time, motivation, or cognitive limitations, final judgments remain biased toward the initial value. This explanation highlights the anchoring-and-adjustment heuristic as a general rule of thumb.

2. Selective Accessibility Model
   Later theories emphasize selective accessibility: when an anchor is presented, people test its plausibility by searching their memory for information consistent with it. This confirmatory search makes anchor-consistent evidence more mentally accessible, thereby pulling estimates toward the anchor. On this view, anchoring arises less from mechanical adjustment and more from biased information retrieval and interpretation.

3. Scale Distortion and Representation Accounts
   Other accounts suggest that anchors influence how people represent the scale of possible values. For example, a high anchor may implicitly expand what is considered a “normal” or “plausible” range, whereas a low anchor compresses it. In this sense, anchoring affects the internal frame of reference, not just a starting point for computation.

4. Motivational and Social Explanations
   Some explanations focus on social norms and strategic behavior, especially in negotiation. People might give weight to anchors out of politeness, perceived informational value, or deference to authority. In such cases, the effect reflects not only cognitive bias but also social reasoning about what others “must know” to make such an offer or suggestion.

Philosophically, these theories raise questions about the nature of bounded rationality: whether anchoring reveals flaws in human reasoning, adaptive shortcuts given resource limits, or a mixture of both.

Philosophical and Practical Implications

Anchoring bias has significant implications for epistemology, ethics, and public policy.

In epistemology, anchoring challenges idealized models of rational belief updating. If arbitrary starting points can systematically influence judgments, then people’s credences and estimates diverge from what Bayesian or logical norms would prescribe. Philosophers of mind and cognitive science use anchoring as evidence for heuristic-driven cognition and as a case study in how context-dependence shapes belief formation.

In moral and political philosophy, anchoring raises issues about manipulation and autonomy. Marketers, negotiators, and political actors can exploit anchoring by setting strategic reference points (e.g., framing budget cuts or increases relative to a chosen baseline). Critics argue that such practices may undermine informed consent or democratic deliberation, while defenders sometimes claim that anchoring is ubiquitous and can be used for nudging people toward beneficial choices (e.g., setting default contribution rates in retirement plans).

Practically, awareness of anchoring has influenced:

• Behavioral economics: models of consumer choice, pricing, and market behavior incorporate anchoring to explain anomalies in demand and valuation.
• Law and policy: understanding how initial numbers affect jury awards, sentencing, and damage claims has led to debates about whether to restrict certain forms of numerical suggestion.
• Decision-making and forecasting: experts are advised to generate independent estimates, use structured analytic techniques, or consult multiple anchors to mitigate bias.

While no consensus exists on how fully anchoring can be overcome, many scholars agree that it exemplifies the broader theme that human judgment is reference-dependent and shaped by prior frames, underscoring the gap between normative models of rationality and actual human reasoning.