PsychologyCounterintuitive resultExplained★ Famous
Anchoring Effect
Also known as Anchoring Bias, Focalism
A random wheel-of-fortune spin changes your estimate of how many African nations are in the UN. The spin was meaningless. The effect was real. Why?
Counterintuitive result. Nothing is wrong with the reasoning. The result is simply correct and human intuition is simply unreliable here. Not a contradiction at all.
Discover
Tversky and Kahneman set up a wheel of fortune in front of subjects. It was rigged — it could only land on 10 or 65. They spun it, let it land, and asked subjects to estimate: what percentage of African nations are members of the United Nations?
The number on the wheel had nothing to do with the question. The subjects knew it had nothing to do with the question. And yet:
Subjects who saw the wheel land on 65 gave estimates averaging 45%. Subjects who saw the wheel land on 10 gave estimates averaging 25%.
A random, meaningless number had moved their estimates by 20 percentage points.
This is the anchoring effect. The first number you encounter, even if it is arbitrary and irrelevant, pulls all subsequent estimates in its direction. The effect is large, robust, and almost immune to awareness.
Understand
The anchoring effect is the tendency for an initial numerical value (the anchor) to disproportionately influence subsequent estimates, even when the anchor is known to be arbitrary or irrelevant.
Tversky and Kahneman's original explanation was anchoring and adjustment: people start from the given number and adjust toward a plausible estimate, but they consistently under-adjust. The adjustment process stops too early, leaving the estimate closer to the anchor than it should be.
More recent research suggests that anchoring also works through priming: the anchor activates information consistent with itself in memory, making anchor-consistent information more accessible and anchor-inconsistent information harder to retrieve.
Both mechanisms contribute, and neither can be fully switched off by awareness.
Why knowing about it doesn't help
The anchoring effect is one of the most robust in psychology partly because its mechanism is pre-conscious. The anchor influences the information you retrieve before you begin conscious deliberation. By the time you are reasoning, the anchor has already done its work. Knowing about anchoring bias produces only modest reductions in the effect — and sometimes none at all.
Anchoring in negotiation
The practical consequence most directly exploited is in negotiation. In any negotiation, the party who makes the first offer sets the anchor. Subsequent bargaining adjusts from that anchor, and the adjustments are systematically insufficient. The first offer predicts the final outcome more strongly than any other single factor. This is why experienced negotiators are taught to make the first offer — and why sellers who ask higher prices receive higher final sale prices, on average, even after negotiation.
The shape of it
Same shape, different costume:
Examples
| Context | Anchor | Effect |
|---|---|---|
| Wheel of fortune (10 vs 65) | 10 → estimate 25%; 65 → estimate 45% | 20-point gap from random wheel |
| Salary negotiation | First offer stated | Final salary closer to first offer regardless of merit |
| Legal sentencing | Prosecutor's sentence request | Judges' sentences correlate with request even when request is random |
| Real estate | Listing price | Final sale price anchors to listing, even above market |
| Retail pricing | "Was $199, now $99" | Perceived value higher than "now $99" alone |
| Self-assessment | Recalled best vs worst performance first | Subsequent overall assessment anchored to initial recall |
In sentencing studies, when mock jurors or actual judges are asked to estimate sentences, their decisions are systematically influenced by the first number they encounter — even a randomly generated number unrelated to the case.
Explore
Extreme anchors and counter-anchoring. Some research finds that sufficiently extreme anchors reduce or eliminate the effect,…
Extreme anchors and counter-anchoring. Some research finds that sufficiently extreme anchors reduce or eliminate the effect, because they trigger a motivated contrast response: the anchor is so obviously wrong that subjects adjust much further than usual. But the threshold for triggering this response is high — anchors need to be implausibly extreme, not just outside the true range.
Expert anchoring. A common assumption is that experts are immune to anchoring in their domain. The evidence says otherwise. Real estate agents shown properties with different listing prices give significantly different valuations. Physicians shown high or low base-rate figures give different diagnoses. Judges give different sentences based on prosecutor requests. Domain expertise reduces but does not eliminate anchoring.
Anchoring in the lab and field. The classic lab finding uses arbitrary, clearly irrelevant anchors (a wheel of fortune). Field replications using natural anchors (list prices, prior sale prices, salary requests) consistently find the same direction of effect, often larger in magnitude. The effect is not a laboratory artefact.
The precision of anchors. Precise anchors (e.g., $9,543) lead to different responses than round anchors (e.g., $10,000). Precise anchors suggest that the anchor-setter has done detailed analysis, increasing its perceived credibility. They also shift final estimates to precise numbers near the anchor. This is why precise first offers in negotiation are more effective than round ones.
Historical context
1974. Tversky and Kahneman publish "Judgment under Uncertainty: Heuristics and Biases" in Science, introducing the…
1974. Tversky and Kahneman publish "Judgment under Uncertainty: Heuristics and Biases" in Science, introducing the anchoring-and-adjustment heuristic alongside availability and representativeness. The wheel-of-fortune experiment is the canonical demonstration.
1992. Gregory Northcraft and Margaret Neale publish a real-estate anchoring study in which actual real estate agents' valuations of houses are significantly influenced by the listing price — demonstrating the effect in expert professionals using ecologically valid stimuli.
2002. Kahneman receives the Nobel Prize in Economics. His work with Tversky, including anchoring, is specifically cited.
2010s. A series of replication and extension studies confirm the basic effect, find limits on expert immunity, and document the role of priming alongside adjustment as a mechanism.
Why it matters
Negotiation. The anchoring effect is the most consistently actionable bias in negotiation research. Making the first offer…
Negotiation. The anchoring effect is the most consistently actionable bias in negotiation research. Making the first offer matters — it sets the anchor. Making a high (or low) first offer pulls the outcome in your direction. Being aware of anchoring does not eliminate its effect on you, but it may prompt more aggressive counter-anchoring.
Legal and medical judgment. Sentencing decisions, damages awards, and diagnostic decisions are all influenced by early numerical information. This creates systematic variation in outcomes based on irrelevant factors — who happens to state the first number, in what order information is presented. Procedural reforms that delay numerical information (presenting facts before numbers, blind scoring) partially mitigate this.
Pricing and retail. The anchoring effect is the scientific foundation of "was/now" pricing, suggested retail price displays, and high-low promotional pricing strategies. Retailers systematically use anchors to make their prices feel like bargains. Understanding the mechanism helps consumers recognize and partially resist these strategies.
Modern relevance
AI price discovery. Algorithmic pricing systems and recommendation engines present anchor prices (similar item prices,…
AI price discovery. Algorithmic pricing systems and recommendation engines present anchor prices (similar item prices, "frequently bought together" prices) that anchor consumer valuations. This is anchoring at scale — millions of anchors personalised to individual consumers.
Salary negotiation in tech. The tech industry's salary negotiation culture is extensively shaped by anchoring: the candidate who gives the first specific number anchors the offer. Online databases of salary data (Glassdoor, Levels.fyi) serve partly as counter-anchors — giving candidates a basis for refusing the employer's opening anchor.
AI bias in estimation. AI systems trained to produce numerical estimates (house valuations, risk scores, sentence recommendations) can function as anchors for human decision-makers who see the AI output before forming their own judgment. The human may nominally remain in control, but the AI's output anchors subsequent human deliberation in ways that are partially invisible.
Sources & further reading
Primary source
(1974)Judgment under Uncertainty: Heuristics and BiasesScience, 185(4157), pp. 1124–1131.
The paper introducing the anchoring-and-adjustment heuristic, including the wheel-of-fortune experiment demonstrating that arbitrary numbers pull subsequent estimates.
How this connects
- Related to
- Halo EffectBoth are cognitive biases in which irrelevant information contaminates judgment. The halo effect is about personal traits; the anchoring effect is about numbers. In both, the first piece of information encountered distorts all subsequent evaluation.
- Dunning-Kruger EffectBoth reveal that human judgment is systematically miscalibrated in predictable ways. Dunning-Kruger is about competence self-assessment; anchoring is about numerical estimation.
- Sunk Cost FallacyBoth concern the disproportionate influence of irrelevant past information on current decisions. Anchoring is about an initial number; sunk cost is about past investment.
- Baader-Meinhof PhenomenonBoth are examples of how prior information, once activated, systematically distorts subsequent perception and judgment. Anchoring is about numbers; the Baader-Meinhof Phenomenon is about concepts and objects.