Walk into a store that has marked a jacket down from four hundred dollars to two hundred and fifty, and something interesting happens in your brain before you have consciously evaluated anything. The four hundred dollar figure has already established itself as the reference point against which two hundred and fifty will be measured. Never mind that two hundred and fifty might be more than the jacket is worth. Never mind that you would never have considered paying four hundred in the first place. The original number has done its work, and two hundred and fifty now feels like a bargain rather than a price to be evaluated on its own terms.
This is anchoring bias operating in one of its most visible forms. The mechanism is the same whether the anchor is a price tag, a project estimate, a salary figure, a real estate listing, or a seemingly random number introduced before a judgment. The first number heard, seen, or considered sets a gravitational field around subsequent thinking that is extremely difficult to escape, even when people know about the bias and are actively trying to compensate for it.
Contents
The Research That Established the Effect
Anchoring bias was identified and named by Amos Tversky and Daniel Kahneman in their landmark 1974 paper on heuristics and biases, the same work that introduced availability bias to the scientific literature. Their original demonstration was striking in its simplicity. Participants watched a wheel of fortune spin to a number between zero and one hundred, then estimated the percentage of African countries in the United Nations. The wheel was rigged to land on either ten or sixty-five. Participants who saw sixty-five gave estimates averaging around forty-five percent. Participants who saw ten gave estimates averaging around twenty-five percent. A completely arbitrary, visibly random number, one with no logical connection to the question, produced a systematic twenty-point gap in the answers.
Subsequent research has replicated and extended the finding across every domain studied. Judges given higher recommended sentences hand down longer sentences. Real estate agents shown higher listing prices estimate higher property values. Negotiators who open with higher initial offers consistently achieve better final outcomes than those who open modestly, even when the counterpart knows this and consciously tries to resist it. The anchor does not merely suggest a starting point. It restructures the entire evaluative landscape around itself.
Why the Brain Cannot Simply Ignore an Anchor
The persistence of anchoring even when people know about it and try to correct for it points to a mechanism that operates below the level of deliberate control. Psychologist Nicholas Epley and colleagues proposed that anchoring works through an adjustment process: people start from the anchor and adjust until they reach a value that seems plausible, then stop. The problem is that adjustment consistently stops too early, leaving the final estimate closer to the anchor than an objective analysis would warrant. Knowing that you should adjust more does not reliably produce larger adjustments. The stopping signal that says “this seems plausible now” fires before sufficient distance from the anchor has been achieved.
A complementary mechanism involves selective retrieval: exposure to an anchor activates information in memory that is consistent with that anchor, making anchor-consistent evidence feel more available and salient. This connects anchoring to availability bias in a way that compounds both effects. The anchor not only sets the starting point for adjustment. It also shapes which evidence feels relevant during the evaluation.
Where Anchoring Costs You Most
The financial consequences of anchoring are the easiest to quantify and often the largest in absolute terms. In salary negotiations, the first number offered, whether by employer or candidate, significantly shapes the final outcome. Candidates who anchor high, when they have the market knowledge to justify it, consistently achieve better results than those who wait for an offer and respond to an employer-set anchor. The asymmetry is substantial: the person who sets the anchor has a structural advantage that the person who responds to it must work actively to overcome.
Real estate transactions are similarly saturated with anchoring effects. List prices anchor buyer assessments of value even when buyers are explicitly told the list price is arbitrary. First offers in negotiation anchor the midpoint around which subsequent haggling occurs. The agent who sets the first number in the room is, in a meaningful sense, shaping the entire transaction.
Anchoring in Professional and Medical Judgment
The stakes are higher still in professional domains where anchoring distorts judgments that carry serious consequences. The medical literature on diagnostic anchoring documents a well-established pattern: initial diagnoses, made quickly and often on limited information, disproportionately influence subsequent clinical reasoning even as new, potentially contradictory information arrives. The first diagnosis functions as an anchor that makes confirmatory evidence feel more salient and disconfirmatory evidence harder to integrate. This is a version of confirmation bias operating through an anchoring mechanism, and it contributes meaningfully to diagnostic error rates.
Legal sentencing research reveals an equally troubling pattern. Judges, who are specifically trained to be impartial and evidence-based, show measurable anchoring effects in response to prosecution sentencing recommendations, defense recommendations, and even, in some studies, numbers generated entirely at random. The professional training, the deliberate commitment to neutrality, and the explicit awareness of the bias do not eliminate the effect. They reduce it modestly at best.
Practical Defenses Against Anchoring
The most direct protection against anchoring in any evaluative context is generating your own independent estimate before encountering any external number. When buying a car, a house, or a professional service, researching comparable prices and forming a view of reasonable value before entering any negotiation or seeing any list price fundamentally changes the anchor’s power. The independently formed estimate becomes your anchor, and the external number must contend with it rather than filling an empty field.
In negotiations, being the first to set an anchor is almost always strategically preferable to responding to one, provided the anchor is grounded in defensible reasoning rather than arbitrary aspiration. An anchor with a clear rationale is harder to dismiss than one that simply names a large number. The rationale does not need to be airtight. It needs to provide enough justification to establish legitimacy, after which the anchor’s gravitational pull does the rest of the work.
In your own deliberations, naming the anchor explicitly and asking what you would estimate if you had never seen that number is a useful discipline. It rarely produces perfect deanchoring, but it produces better deanchoring than simply knowing that the bias exists and hoping the awareness is sufficient. Awareness reduces the effect without eliminating it. The discipline of actively constructing an independent estimate is what closes the remaining gap.