(Behavioural Science) #2 Anchoring Effect

Principle #2 — Framing

The anchoring effect

The first number or piece of information a person encounters in a decision context exerts a disproportionate and lasting pull on all subsequent judgments — even when that first number is arbitrary, irrelevant, or obviously wrong. People do not evaluate numbers in isolation. They evaluate them relative to whatever was seen first. The anchor does not determine the final answer, but it dramatically constrains the range of answers that feel plausible.

1974

Tversky & Kahneman identify the effect

~50%

Adjustment from anchor is typically insufficient

Robust

Persists even when people know they're being anchored

Universal

Affects experts, novices, and judges equally

1. What it is — science and research

Amos Tversky and Daniel Kahneman introduced anchoring in their landmark 1974 paper on heuristics and biases. In their original demonstration, participants watched a wheel of fortune spin to land on either 10 or 65 — a number determined entirely by chance. They were then asked to estimate the percentage of African countries in the United Nations. Those who saw 65 guessed significantly higher than those who saw 10, even though the wheel had nothing to do with the question. An arbitrary, irrelevant number had colonized their estimates.

What makes anchoring particularly powerful as a nudge is that it is remarkably resistant to awareness and expertise. Knowing that an anchor exists does not free you from its influence. In subsequent research, judges given sentencing recommendations anchored their decisions to those recommendations even when the recommendations were generated randomly — a finding with serious implications far beyond commercial nudging.

The mechanism operates through two distinct cognitive pathways. The first is confirmatory hypothesis testing: when we see an anchor, we search our memory for information consistent with it, which inflates our estimate of its plausibility. The second is insufficient adjustment: even when we consciously try to move away from an anchor, we tend to stop adjusting too soon — we reach a number that feels plausible but has not traveled far enough from the starting point.

"Anchoring is not a sign of stupidity or laziness. It reflects the same cognitive economy that makes human judgment efficient in most real-world settings. The problem arises because the same mechanism that helps us use context to interpret numbers also makes us vulnerable to manufactured context." — Kahneman, Thinking, Fast and Slow

Key research

Tversky & Kahneman — wheel of fortune (1974)

Foundational

The original anchoring demonstration. A rigged wheel landing on 10 or 65 significantly shifted participants' estimates of the percentage of African UN member states — despite no logical connection between the wheel and the question. Participants who saw 65 estimated 45% on average; those who saw 10 estimated 25%. A 55-point difference in the anchor produced a 20-point difference in estimates. This established the core finding: even transparently arbitrary numbers anchor judgment.

20-point estimate gap from a purely random anchor

Ariely, Loewenstein & Prelec — arbitrary coherence (2003)

Field experiment

Participants were shown consumer products (wine, chocolate, computer equipment) and first asked whether they would pay more or less than the last two digits of their social security number. They were then asked their actual willingness to pay. Those with high SSN endings (80–99) bid 216–346% more than those with low endings (01–20). The last two digits of a completely personal identifier — a number no one treats as relevant to pricing — created massive, systematic differences in perceived value. The authors coined the term "arbitrary coherence" to describe how anchors create internally consistent preference structures from arbitrary starting points.

High SSN group bid up to 346% more than low SSN group

Northcraft & Neale — real estate appraisals (1987)

Expert study

Real estate agents and undergraduate students were given identical property information packets, including a listing price that was manipulated to be either high or low relative to actual value. Both groups anchored their independent valuations to the listing price — and the effect was equally strong for experienced agents as for students. When asked what factors influenced their estimate, agents listed structural features, neighborhood comparables, and market conditions — they were entirely unaware that the listing price had driven their judgment. Expertise did not confer immunity.

Agents' estimates shifted $11,000+ per $12,000 of anchor manipulation

Englich, Mussweiler & Strack — legal sentencing (2006)

High-stakes field

Experienced German judges read a case file for shoplifting and were then given a sentencing "recommendation" generated either by a roll of dice or by a prosecutor. The dice-generated recommendations — obviously random — still anchored sentencing decisions. Judges who received a high dice roll gave sentences averaging 8 months; those receiving a low roll averaged 5 months. The finding is disturbing: professional judges making consequential legal decisions were influenced by numbers they knew to be random. Awareness of the anchor was not sufficient to neutralize it.

3-month sentencing gap from a random dice anchor

Four forms anchoring takes in the wild

Price anchoring

The most common commercial application. A high original price makes a discounted price feel like a bargain. A premium tier makes mid-tier pricing feel reasonable. The anchor is not what you pay — it is what shapes what feels like fair value.

Quantity anchoring

Suggested purchase quantities anchor how much people buy. "Limit 12 per customer" raises average purchases far above "no limit." The anchor communicates what a normal amount looks like — and people adjust toward it.

Negotiation anchoring

The first offer in any negotiation sets the range within which all subsequent counter-offers are evaluated. Opening high in salary negotiation or asset pricing consistently produces better outcomes than waiting for the other party to open.

Default / suggestion anchoring

Pre-filled form fields, suggested donation amounts, and recommended serving sizes all anchor behavior toward a suggested quantity. People adjust away from suggestions less than they would from a blank field.

How price anchoring shifts perceived value

High anchor shown first

$999

Original price

→

Sale price

$399

Feels like a great deal

No anchor / low anchor

—

No prior reference

→

Same price shown

$399

Feels expensive or uncertain

Same $399 price. Different anchor. Different perceived value. The product has not changed — only the reference point has.

---

2. Real application examples

Business

Retail pricing — the "original price" anchor

The most pervasive commercial use of anchoring is the struck-through original price displayed alongside a sale price. The original price is not informational about current market value — it is an anchor. Research by Urbany, Bearden & Weilbaker (1988) found that even implausibly high "original prices" that consumers doubted increased their perception of deal quality. The anchor does not need to be credible to be effective — it only needs to be present. Retailers who display original prices see higher conversion rates and higher willingness to pay than those showing price reductions as absolute discounts ("save $200") because anchoring the full value is more powerful than stating the saving amount.

Struck-through pricing increases purchase likelihood even when anchor is doubted

SaaS pricing tiers — the decoy and anchor combined

Most SaaS pricing pages display three tiers: a low-feature entry tier, a mid-tier "most popular" option, and a high-feature enterprise tier. The enterprise tier is often not intended to convert at high volume — its primary function is to anchor the mid-tier price as reasonable by contrast. A $299/month plan feels very different when displayed next to a $999/month enterprise plan than when displayed alone. The "most popular" badge on the mid-tier reinforces this with social proof, but the anchor is what makes the price feel justified. This structure, now nearly universal in SaaS, was popularized partly by Dan Ariely's research on the decoy effect and is a textbook anchoring + decoy combination.

Supermarket quantity limits — "limit 12 per customer"

Brian Wansink's research on quantity anchors found that "limit 12 per customer" signs dramatically increased average purchase quantities versus no-limit conditions — even for products people had no prior intention of stocking up on. The number 12 functioned as a quantity anchor, communicating what a normal purchase amount was. People who would have bought 1–2 cans adjusted toward the suggested "normal" of 12. This is anchoring operating through quantity perception rather than price perception, and it works even when the limit is transparently promotional in intent.

Quantity anchors increase average purchase volume by 30–50%

Salary negotiation — the power of opening first

Negotiation research by Galinsky & Mussweiler (2001) established that the party who makes the first offer in salary or price negotiations consistently achieves better outcomes, because the first number becomes the anchor around which all subsequent discussion orbits. The study also found a partial counter-strategy for recipients of anchors: rather than adjusting from the anchor (which leads to insufficient adjustment), deliberately consider information inconsistent with the anchor — what the lowest plausible salary is, what comparable roles pay — to counteract the pull. Knowing about anchoring is not enough; active counter-anchoring strategies are required.

Public policy

Charitable giving — suggested donation amounts

Donation solicitations that display suggested amounts systematically anchor giving behavior. A landmark study by Desmet & Feinberg (2003) found that donors given high suggested amounts gave significantly more than those given low suggestions — even when the amounts were acknowledged as arbitrary. Critically, the anchor's effect extended to the distribution of gifts: high-anchor conditions saw more donors give large amounts and fewer give token amounts. The anchor did not just shift the average; it reshaped the entire distribution. Charity: Water and other high-performing nonprofits use anchored ask strings ("$50 / $100 / $250 / other") rather than open fields, which produce consistently higher average gifts than blank input boxes.

High anchors increase average donations by 20–100% vs. low or no anchor

Pension contribution defaults — anchoring savings rates

When employees are auto-enrolled in pension plans, the default contribution rate functions as a powerful anchor on savings behavior. Research by Choi, Laibson & Madrian (2004) found that employees who were auto-enrolled at 3% tended to remain at 3% even years later, treating the default as both the normal and the appropriate amount. Auto-escalation schemes that gradually raise contributions attempt to counter this anchor by making the anchor itself move. The policy implication is significant: setting a low default contribution rate does not just reflect a policy choice — it creates a psychological anchor that keeps contributions low for years, with compounding effects on retirement adequacy.

Default rate becomes modal contribution rate even after years of enrollment

Legal judgments — the disturbing case of anchor immunity

The Englich et al. sentencing research described above has direct policy implications. If experienced judges are anchored by randomly generated sentencing recommendations, then the sequence and framing of information presented in legal proceedings matters independently of its content. Prosecutors who open with a high sentencing ask anchor the judge's range of consideration. Defense attorneys who open with a low counter-anchor may partially neutralize this. Some jurisdictions have explored structured sentencing guidelines partly as a response — providing a normatively derived anchor that replaces the influence of procedurally arbitrary first numbers.

Personal habit change

Exercise goals — anchoring the minimum, not the target

Most fitness advice anchors people to aspirational targets: run 5km, exercise 5 times a week. Behavioral research suggests this is often counterproductive for habit formation — high anchors feel distant and prompt avoidance rather than action. BJ Fogg's Tiny Habits framework deliberately anchors at the minimum: "do two push-ups," "floss one tooth." This low anchor makes starting trivially easy, and once started, people consistently do more than the anchor — but the anchor ensures they start at all. The framing insight is that for habit initiation, the anchor should be so low it is embarrassing not to exceed it.

Budgeting — the first number you write down

In personal budgeting, the first category allocation a person sets anchors all subsequent allocations. If someone begins by allocating $500 to dining out, this number shapes how reasonable $400 or $600 feels for other lifestyle categories. Zero-based budgeting — starting from $0 for every category each period — is partly a de-anchoring technique: it forces fresh evaluation rather than adjustment from last month's anchor. Apps like YNAB that require explicit re-allocation each month are structurally counteracting the anchoring effect that makes people simply repeat prior spending patterns.

Negotiating your own salary — the anchoring advantage

For individuals navigating salary negotiations, anchoring research provides clear practical guidance. State your number first and make it specific rather than round — "I'm looking for $127,500" anchors more effectively than "$125,000" because specific numbers imply careful research and calculation, making them harder to dismiss as arbitrary. Research by Mason, Lee & Wiley (2013) found that precise anchors generate smaller counter-offer adjustments than round ones, because round numbers signal that the asker hasn't done detailed analysis and is prepared to move significantly. Precision is a credibility signal that stiffens the anchor.

Precise anchors ($127,500) generate smaller adjustments than round anchors ($125,000)

---

3. Design guidance — when and how to use it

When it works — use anchoring if these conditions hold

• The decision involves a number, quantity, or amount that the person has no strong prior value for
• The person will be making a judgment under uncertainty — anchoring is strongest when people don't have confident prior beliefs
• You control the order in which information is presented — anchoring requires first-mover advantage
• The anchor is plausible enough not to trigger immediate skepticism (implausible anchors still work, but less effectively)
• You want to shift a distribution of responses, not just an average — anchoring reshapes the entire range of what feels reasonable
• The context allows for a high anchor that makes the target option appear as a contrast or value comparison

When it won't work or may backfire

• The audience is highly expert and has strong prior knowledge — expertise reduces but does not eliminate anchoring, so effects will be smaller
• The anchor is so implausible it triggers disbelief and reactance — audiences may then adjust further away than they would without any anchor
• The person has already formed a clear prior — anchoring is weakest when people already have confident values for what something should cost
• The anchor is transparently manipulative in a context where trust matters — e.g., inflated "original prices" in categories where consumers track real prices
• The goal is to encourage people to think carefully and independently — anchoring undermines deliberation, which is sometimes the wrong outcome

How to design the nudge — six steps

1

Identify the number or quantity that needs to be shaped

Map out every numerical decision in the journey: price perception, quantity purchased, contribution rate, suggested amount, negotiation range. Each is an anchoring opportunity. Prioritize those where the person has no strong prior belief — that is where anchoring has the largest effect.

2

Set the anchor before the target number is presented

Order is everything. The anchor must appear first. In pricing, show the high-tier or original price before the sale price. In donation asks, show the suggested amounts before the input field. In pension enrollment, communicate the contribution rate before asking for confirmation. Reversing the order neutralizes the effect.

3

Make the anchor specific rather than round

Precise numbers ($127,500 rather than $125,000; "87% of members" rather than "most members") signal that the number was calculated, not chosen arbitrarily. This increases credibility and reduces the psychological permission to adjust away. Round numbers invite round counter-offers; precise numbers invite precise ones.

4

Calibrate anchor height to context

For pricing, the anchor should be high enough to make the target feel like a deal, but not so high as to break credibility entirely — research shows implausibly high anchors still work but generate more consumer skepticism and lower trust. For habit formation and goals, low anchors that ensure action beats aspirational anchors that cause avoidance. Match anchor direction to the behavioral goal.

5

Use visual and spatial contrast to reinforce the anchor

Struck-through original prices, color contrast between anchor and target, size hierarchy that makes the anchor visually prominent — these all amplify the cognitive contrast effect that makes the anchor feel more distant from the target, increasing perceived value of the gap. Pure text anchoring is less effective than visually emphasized anchoring.

6

Test whether your anchor is too obvious

In contexts where consumers are sophisticated and track prices (electronics, travel, financial products), transparent anchoring with inflated "original prices" can undermine trust when exposed. Test whether consumers believe the anchor. If skepticism is high, real comparison anchors — competitor prices, market averages, prior period rates — are more credible and nearly as effective.

What good vs. bad anchor design looks like

Charitable donation ask

Weak — open field, no anchor

"Please enter the amount you'd like to donate: $____"

Strong — anchored ask string

"$35 / $75 / $150 / $300 / Other $____" — with $75 highlighted as "most common"

Pension auto-enrollment

Low anchor — shapes long-term behavior badly

Auto-enrolled at 3% with no escalation — employees anchor to 3% for years

Higher anchor + escalation

Auto-enrolled at 6% with annual 1% escalation — anchors to a more adequate rate and drifts upward

Salary negotiation

Weak — round, invites adjustment

"I'm looking for around $120,000."

Strong — precise, implies calculation

"Based on my research and experience, I'm targeting $127,500."

The ethical boundary with anchoring

Anchoring sits closer to the ethical edge than social proof does, because it can be used to manipulate rather than inform. Fabricated "original prices" on products that were never sold at those prices is a common form of consumer deception — and one that regulators in the UK, EU, and US are increasingly scrutinizing. The test is whether the anchor reflects something real (a genuine prior price, a market rate, a meaningful comparison) or is invented purely to distort perception. Real anchors are legitimate nudges. Invented anchors are manipulation. The practical distinction matters: consumers who discover they were deceived by a fake anchor do not just avoid that product — they lose trust in the brand entirely.