Foundationsbehavioral
Anchoring: First Numbers Stick
Overview
In a famous experiment, spinning a rigged wheel (landing on 10 or 65) shifted people's estimates of African UN membership by 20 percentage points. The first number encountered — even a random one — anchors judgment, with adjustments falling short. Anchoring affects negotiations, pricing, legal damages, and valuations. This lesson explains the mechanism, the evidence across domains, why experts aren't immune, and practical defenses.
Learning objectives
- Define anchoring and the adjustment mechanism
- Reproduce the classic wheel experiment
- Identify anchoring in negotiations, pricing, and law
- Explain why anchors work: priming vs. adjustment
- Apply defenses: counter-anchors, ranges, and structured processes
The wheel experiment
Tversky and Kahneman (1974): participants spun a rigged wheel landing on 10 or 65, then estimated the percentage of African nations in the UN. Median estimates: 25% (anchored at 10) vs. 45% (anchored at 65) — a random number moved judgments 20 points. The mechanism: insufficient adjustment — people start at the anchor and adjust toward the truth, but stop too early. Stunningly, anchors work even when people know they're random, even when they're implausible, even with financial incentives for accuracy. The mind treats the first number as a suggestion it can't fully shake.
The core mechanism: judgment = anchor + (partial) adjustment toward the truth. Adjustments are systematically too small — the anchor's gravitational pull exceeds its information.
Anchoring in negotiations
First offers anchor negotiations powerfully: the final price correlates strongly with the opening bid, even controlling for value. Experiments: listing prices anchor home sales; initial salary offers anchor lifetime earnings paths; 'suggested donations' anchor charitable giving. Strategy follows: open ambitiously (the anchor pulls toward you), but credibly (absurd anchors break trust or end talks). Counter-strategy: prepare your own anchor (market data), explicitly discount theirs ('that number reflects X, not this deal'), and anchor the range, not just the point. Negotiation is substantially a contest of anchors.
Meta-analyses confirm: the opening offer explains a large share of final settlement variance. Making the first offer — well-prepared — is usually advantageous.
Prices, law, and valuations
Marketing runs on anchors: 'was 120' (the $200 anchors value perception), 'limit 12 per customer' (12 anchors quantity), prestige pricing. Law: damage awards anchor on the plaintiff's demand — identical cases get different awards based on the ask; sentencing anchors on prosecutors' requests. Finance: analysts anchor on past prices and management guidance; IPO valuations anchor on comparable multiples; 52-week highs anchor investor expectations (a documented trading signal). Wherever numbers are uncertain, the first number rules — which is everywhere.
Anchors need not be real: inflated 'original' prices anchor discounts deceptively. Regulators police fake reference prices precisely because anchors work even when fabricated.
Why anchors work
Two mechanisms, both supported. Insufficient adjustment: we start at the anchor and adjust toward our true estimate, stopping early (effortful, System-2 process). Selective accessibility (priming): the anchor makes anchor-consistent information come to mind — hearing '65' activates thoughts of high African representation, biasing the subsequent estimate (automatic, System-1). Which dominates depends on the task: numeric estimation favors adjustment; preference judgments favor priming. The practical upshot is the same: the anchor contaminates the information brought to bear, not just the final number.
Anchors work both by lazy adjustment from a starting point and by priming consistent thoughts. Defenses must address both — effortful re-derivation plus considering the opposite.
Defenses
You can't simply 'ignore' anchors — warnings barely help. What works: generate your own anchor first (do independent analysis before seeing others' numbers — the 'anchor hygiene' of good procurement). Consider the opposite: deliberately argue why the anchor is too high/low (activates inconsistent information, countering priming). Use ranges and structured processes: independent estimates averaged (wisdom of crowds dilutes individual anchors), decision checklists, and 'premortems.' In negotiations: never accept their anchor as the frame — re-anchor with your own data, or negotiate the criteria before the numbers. The defense is procedural, not willpower.
Key points
- Anchoring: the first number encountered pulls judgments toward itself via insufficient adjustment (and priming).
- Classic proof: a rigged wheel (10 vs. 65) shifted UN-membership estimates by 20 points.
- Anchors drive negotiations (first offers), pricing ('was $200'), legal awards, and financial valuations.
- Experts, incentives, and awareness barely attenuate it — the bias is deep.
- Defenses are procedural: independent first estimates, consider-the-opposite, ranges, and re-anchoring — not willpower.
Common misconceptions
✓ Truth: Judges, analysts, and experts anchor robustly. Domain knowledge shifts the adjustment, not the mechanism — the pull persists with incentives and expertise.
✓ Truth: Warnings and awareness barely reduce anchoring. The bias operates partly automatically (priming); only procedural defenses — independent estimates, structured processes — reliably help.
✓ Truth: Reasonable-but-ambitious opens anchor best. Absurd opens break trust; timid opens anchor against you. The evidence favors confident, well-justified first offers.
Practice questions
Q1. In Tversky and Kahneman's wheel experiment, participants' UN-membership estimates differed by ~20 points depending on a rigged wheel landing because of:
- Their geography knowledge
- Anchoring — starting from the random number and adjusting insufficiently toward the truth
- Peer pressure
- Faulty wheels
A random 10 vs. 65 moved median estimates from 25% to 45%. The anchor provides a starting point; adjustment toward the true answer falls systematically short.
Q2. Which negotiation finding about anchoring is best supported?
- First offers don't matter
- The opening offer strongly predicts the final settlement — making a well-prepared first offer is usually advantageous
- Only the final offer matters
- Anchors only affect amateurs
Meta-analyses confirm the first-offer advantage: final prices correlate strongly with opening bids. Preparation + ambitious-but-credible opening is the evidence-based strategy.
Q3. The most effective defense against anchoring is:
- Trying harder to ignore the anchor
- Procedural: generate independent estimates before exposure, consider the opposite, use ranges and structured processes
- Avoiding all numbers
- Trusting your gut
Awareness and willpower barely help (the bias is partly automatic). Procedural defenses — independent first analysis, deliberate counter-arguing, aggregation — actually work.
Click a question to reveal the answer and explanation.
Frequently asked questions
What's the difference between anchoring and priming?
Anchoring is the general phenomenon (first numbers pull judgments); priming (selective accessibility) is one mechanism — the anchor activates consistent thoughts. Insufficient adjustment is the other mechanism.
Do absurd anchors work?
Weakly — implausible anchors get discounted, though they still pull slightly. The most effective anchors are ambitious but defensible: high enough to pull, credible enough to stick.
How do stores use anchoring ethically vs. deceptively?
Genuine reference prices ('our regular price') inform; fabricated 'was' prices deceive. Regulators require the reference price to have been real — because even fake anchors work.
Can you anchor yourself beneficially?
Yes — set ambitious personal goals (they anchor effort upward), or consider a high savings target before budgeting. Self-anchoring is the bias turned into a commitment device.