Chapter 11: Anchors

Type: Chapter 11

Category: Psychology

Difficulty: Intermediate

Reading time: 6 min

Core idea

Kahneman and Tversky set up a wheel of fortune rigged to stop on either 10 or 65. After the spin, participants were asked: is the percentage of African nations in the UN higher or lower than that number? What is your estimate? The people who saw 65 guessed 45% on average. The people who saw 10 guessed 25% on average. The wheel had no information about the UN. It was random. Yet it moved their estimates by 20 percentage points.

This is the anchoring effect: an initial value — even an arbitrary, irrelevant, or obviously random one — pulls final estimates toward itself. The first number considered becomes a psychological anchor that subsequent reasoning fails to escape.

Anchoring is one of the most robust and consequential biases in the research literature. The anchoring index — the percentage of the anchor’s influence that survives in the final estimate — runs between 40% and 60% in most experiments. Even professionals with relevant expertise show large anchoring effects on consequential judgments.

Why it matters

Two mechanisms: adjustment and priming

Anchoring operates through two distinct pathways, which map onto the two systems:

Anchoring-as-adjustment (System 2): when people start from an anchor and adjust toward their best estimate, they tend to stop adjusting too early — as soon as they reach a value that seems plausible rather than one that is actually correct. The adjustment is effortful and gets cut short. This is a System 2 process, but System 2’s lazy termination leaves the anchor’s influence intact.

Anchoring-as-priming (System 1): the anchor value selectively activates information in memory that is consistent with it. If asked “Is Gandhi older or younger than 144?” you will find your mind retrieving information consistent with extreme old age — long-lived historical figures, myths about centenarians — even though you consciously know 144 is ridiculous. This is the mental shotgun from Chapter 8: System 1 fills the retrieval environment with anchor-consistent evidence, biasing the judgment without any deliberate adjustment.

The scale of the effect

The anchoring effect is large by the standards of psychological biases. In one study, real estate agents shown a house with a high listing price gave higher appraisals than those shown the same house with a low listing price — and the difference persisted even though the agents claimed to be ignoring the listing price. A study of experienced German judges showed that sentencing recommendations were systematically shifted by a random dice roll before deliberation.

The anchoring index quantifies this: on average, about half the distance between anchor and true value survives in the final estimate. If the anchor is $200 and the true value is $100, final estimates cluster around $150.

Anchors in negotiation and commerce

First offers in negotiations function as anchors. A high first offer tends to produce a higher final settlement — not because the first offer has any logical force, but because it establishes the reference point from which adjustments are made. A low asking price anchors buyers downward.

Retailers exploit this with “original prices” crossed out to show a “sale price.” The crossed-out price is the anchor; the sale price feels like a win relative to it. Whether the original price was ever real is irrelevant to the psychological effect.

Key takeaways

Key takeaways

• Anchoring effect: an initial value pulls final estimates toward itself, even when the anchor is random, arbitrary, or obviously irrelevant to the question.
• Two mechanisms: System 2 anchoring-as-adjustment (stops too early) and System 1 anchoring-as-priming (selectively activates anchor-consistent information in memory).
• The anchoring index runs 40–60% in most experiments — about half the anchor's distance from the true value survives in final estimates.
• Expertise does not protect against anchoring: experienced professionals (real estate agents, judges) show anchoring effects comparable to novices on domain-relevant judgments.
• First offers in negotiation function as anchors: high first offers produce higher final settlements. Opening positions matter far more than subsequent moves.
• Countermeasures are weak: consciously knowing about anchoring, being warned against it, or being motivated to be accurate all reduce but do not eliminate the effect.

Mental model

click to zoom

The two mechanisms of anchoring — how both System 1 and System 2 produce anchor-biased estimates through different pathways.

Read it as: An anchor gets into the final estimate through two pathways simultaneously. System 2 uses it as a starting point and stops adjusting too soon. System 1 primes the mental environment with anchor-consistent information. Both pathways push the estimate toward the anchor. Together, they make anchoring one of the most reliable and hard-to-escape biases in judgment.

Practical application

You cannot simply choose to ignore an anchor

Being aware of anchoring is not enough to eliminate it. Studies show that warning participants about anchoring and instructing them to correct for it reduces the effect but does not neutralize it. The priming mechanism operates below conscious awareness; the adjustment mechanism is stopped by System 2 laziness that awareness does not cure.

Strategies that help at the margin:

• Generate the opposite extreme before estimating: if you are about to make an estimate, first ask what the highest plausible value would be, then what the lowest would be. This activates a broader retrieval environment than the anchor primes on its own.
• In negotiations, set the first offer: the party that names a number first gains the anchor advantage. If the other party anchors first, explicitly name the anchor as irrelevant before proceeding.
• Use reference class data: replace anchor-based adjustment with outside-view base rate data. Instead of adjusting from the anchor, ask “what is the typical value for this type of situation?”

Example

A tech startup is negotiating an acquisition. The buyer’s opening offer is $15 million. The founders, who privately valued the company at $8–12 million, counter at $18 million. They eventually settle at $14.5 million — which both sides feel represents a win relative to their opening positions.

But note what the buyer accomplished: by anchoring at $15M, the entire negotiation occurred above the founders’ original valuation ceiling. The anchor was set deliberately. Had the founders made the first offer — say, $20M — the settlement zone would likely have shifted upward. The anchoring effect explains why negotiation training consistently teaches: make the first offer whenever possible, and make it ambitious.

Related lessons

Related

Chapter 8: How Judgments Happen The mental shotgun — how System 1 generates and biases the retrieval environment, the mechanism underlying anchoring-as-priming. Chapter 12: The Science of Availability The availability heuristic — ease of retrieval as a substitute for frequency judgments, a close cousin to anchoring-as-priming. Anchoring The full concept — how initial values distort final judgments across domains from pricing to sentencing.