Anchoring Bias: Examples, Studies, and Your Money

Spin a wheel, watch it stop on a random number, and that number will quietly bend your next guess about something completely unrelated. In 1974, Daniel Kahneman and Amos Tversky did exactly that. They rigged a wheel of fortune to land on 10 for half their subjects and 65 for the other half, then asked everyone the same question: what percentage of African countries are in the United Nations? The people who saw 10 answered a median of 25%. The people who saw 65 answered 45%. A meaningless spin had moved the answer by 20 points.

That is anchoring bias, and it does far more expensive work than guessing at UN membership. The first price on a tag, the first number in a salary talk, the price you once paid for a stock: each becomes a reference point that pulls every later judgment toward it. This guide covers what anchoring is, the five experiments that pin it down (with the actual numbers), why it happens, and, because psychology sites stop at the wheel, exactly how it reaches into your rupees and dollars through pricing, salary, and investing.

What is anchoring bias?

Anchoring bias is the tendency to rely too heavily on the first number you encounter, the "anchor", when making an estimate or decision, then adjust away from it too little, even when that number is random or irrelevant. The name comes from Tversky and Kahneman's 1974 paper in Science, the study that launched the whole "heuristics and biases" research programme and eventually earned Kahneman the 2002 Nobel Prize in economics.

The wheel experiment is the clean version, but the same paper carried a second demonstration that is just as sharp. Ask people to estimate 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1 in five seconds and the median guess is about 2,250. Ask a different group for 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8, the identical product, and the median drops to 512. The first few numbers anchor the estimate. Both groups are badly wrong, by the way: the real answer is 40,320.

The five experiments that prove anchoring is real

Anchoring shows up across randomised experiments spanning UN estimates, product auctions, house appraisals, and criminal sentencing, which is why it is one of the most replicated findings in behavioural science. The table below is the canon, with the numbers most articles leave vague.

Study	The anchor	The result	What it proves
Tversky and Kahneman, 1974 (wheel of fortune)	Rigged spin to 10 or 65	Median UN estimate 25% vs 45%	A random number moves an unrelated judgment
Strack and Mussweiler, 1997 (Gandhi's age)	"Older or younger than 9? vs 140?"	Mean estimate 50 vs 67 years	Even an impossible anchor still works
Ariely, Loewenstein and Prelec, 2003 (SSN auction)	Last 2 digits of your Social Security number	Top-quintile bidders paid roughly 200% to 350% more	An arbitrary number changes what you pay with real money
Northcraft and Neale, 1987 (real estate)	Same house, list price $119,900 to $149,900	Agent appraisals moved ~$114,204 to ~$128,754	Experts anchor too, and deny it
Englich, Mussweiler and Strack, 2006 (judges and dice)	Dice loaded to sum 3 or 9	Sentence recommendations 5.28 vs 7.81 months	A self-rolled random number sways expert judgment

The Ariely study is the one to sit with, because it involves real money. MBA students wrote the last two digits of their Social Security number, treated it as a price, then placed genuine bids on wine, chocolates, and gadgets. Students whose digits fell in the top fifth bid two to three-and-a-half times more than those in the bottom fifth for the same items. A rare bottle of wine drew average bids near $27.90 from high-number students and about $8.64 from low-number ones. Their own Social Security digits, which have nothing to do with wine, set what they were willing to pay.

Why does anchoring bias happen?

Anchoring has two competing explanations, and the money examples make more sense once you can tell them apart. Both are backed by decades of research; the second one is the one most articles miss.

The first is anchoring-and-adjustment (Tversky and Kahneman, 1974). You take the anchor as a starting point and adjust toward your real answer, but you stop as soon as the number feels plausible, so the estimate stays parked near the anchor. This explains anchors that sit in a believable range, like a list price or an opening offer.

The second is selective accessibility (Strack and Mussweiler, 1997). When you test "is the answer higher or lower than the anchor?", your mind pulls up facts consistent with the anchor, and those primed facts stay accessible when you produce your actual estimate. This is the deeper mechanism, and it explains the thing adjustment cannot: why an absurd anchor still moves you. Nobody thinks Gandhi lived to 140, so there is nothing to "adjust down" from, yet the 140 group still guessed higher. Anchor-primed memory did that, and current research treats selective accessibility as the dominant driver of everyday anchoring.

Does anchoring bias affect experts?

Yes, and the experts usually insist it does not. This is the counterintuitive part, and it is well documented in the two professions whose whole job is putting a fair number on something.

Northcraft and Neale (1987) walked real-estate agents through an actual Tucson house and gave them a full information packet. The only thing that differed between groups was the printed list price, spread from $119,900 to $149,900. The agents' appraisals climbed right along with it, from a mean near $114,204 at the low anchor to about $128,754 at the high one, roughly a $14,000 swing driven by a number they were free to ignore. Then the kicker: only about 19% of the agents said the list price had affected them at all. The bias ran below their awareness.

Englich, Mussweiler and Strack (2006) went further and used judges with more than fifteen years on the bench. Each read an identical case, then rolled a pair of dice, secretly loaded to sum to either 3 or 9, and treated the roll as a prosecutor's sentencing demand. Judges who rolled low recommended an average of 5.28 months; judges who rolled high recommended 7.81 months. A number they generated themselves, from dice, changed a prison sentence by roughly half. Expertise did not protect them, and it will not protect you either.

How anchoring bias hooks your money

In money decisions the anchor is almost always a number someone chose for you: a printed MRP, a strikethrough price, an opening salary offer, an analyst's target, or the price you once paid. This is where the rupees and dollars actually leak.

Retail pricing and the MRP trick

The struck-through "was" price is anchoring in its purest commercial form. In India, the printed Maximum Retail Price is a legal ceiling (selling above it breaches the Legal Metrology Act), which is precisely what makes it such a clean anchor: brands print a high MRP, then sell well below it, so a shopper paying a healthy margin still feels like they scored 40% off. Online, Flipkart and Amazon India strike through a higher figure for the same effect, and research on Indian shoppers finds that "Save 40%" and "Save ₹400" land differently even when the final price is identical.

The tactic is old and, past a point, illegal. In the US, JCPenney settled a false-reference-pricing class action for $50 million, with court filings alleging only about 0.2% of its products ever sold at the "regular" price the discounts were measured against; Kohl's settled a similar suit for $6.5 million. The classic decoy version is Williams-Sonoma adding a $429 bread machine beside its $275 model: the $429 unit barely sold, but it reportedly nearly doubled sales of the $275 one by making it look like the sensible middle choice.

Salary negotiation: the first number usually wins

The opening offer in a pay negotiation is one of the strongest predictors of where it lands. A 2006 meta-analysis by Orr and Guthrie put the correlation between the first offer and the final outcome near 0.5, and a 2025 synthesis of 90 studies pushed it to about 0.62, a very large effect. In a worked illustration: if an employer opens at $95,000, a candidate who counters at $103,000 tends to settle near $99,000, while a candidate who opens first at $115,000 against a $102,000 counter tends to settle near $107,000. Same job, roughly an $8,000 first-year gap, decided mostly by who dropped the anchor. It compounds, because raises are a percentage of base and the next employer often anchors to your current pay.

That last point is why the "current CTC" and "expected CTC" boxes on Indian job applications matter so much: they anchor your next offer to your last one, so past underpayment follows you. In the US, more than twenty states have banned employers from asking salary history for the same reason, and research links those bans to a gender pay gap about two percentage points narrower. Our guide on how to negotiate your salary covers the mechanics.

Investing: the price you paid is not fair value

The most expensive anchor in investing is your own purchase price. Buy a stock at ₹800, watch it slide to ₹550, and the pull is to hold until it "gets back to ₹800", as though the market owes you your cost back. It does not; the anchor lives only in your head, and this pattern overlaps heavily with the disposition effect covered in our piece on loss aversion. Other investing anchors include a stock's 52-week high (a share down from ₹2,000 to ₹1,400 looks "cheap" against its own past, not against earnings), the IPO issue price, brokerage target prices, and round index levels like Sensex 80,000 or Nifty 25,000 that get treated as meaningful lines rather than arbitrary points on a continuous scale. This section describes a bias, not a strategy; it is general education, and a SEBI-registered adviser can speak to your own portfolio.

Two everyday Indian anchors worth naming

Two more show up constantly in Indian spending. Gold jewellery bills anchor you on the gold rate, which feels fixed and fair, so the making charges (typically 6% to 20% of the gold value, and the most negotiable line on the bill) get waved through as a small add-on. And "just ₹4,999 a month" on a phone, a scooter, or a gold scheme anchors you on the monthly number so the ₹60,000-plus total, plus interest and processing fees, never becomes the reference point. The frame is chosen so the big number stays out of view.

Anchor in the wild	The number	What it really is
India MRP	Inflated ceiling; "40% off" measured against it	Seller-chosen anchor
JCPenney "regular" price	$50M settlement; ~0.2% ever sold at it	Fictitious anchor
Salary first offer	Correlates ~0.5 to 0.62 with final pay	Negotiation anchor
Gold making charges	6% to 20% of gold value, waved through	Add-on hidden behind a "fixed" rate
EMI framing	"₹4,999/month" vs the ₹60,000+ total	Monthly-amount anchor
Round index levels	Sensex 80,000, Nifty 25,000	Psychological round-number anchor

Anchoring bias vs framing effect vs confirmation bias

Anchoring is about a number, framing is about wording, and confirmation bias is about a belief you already hold, which is why the three get muddled. The distinctions are worth one clean table, because people routinely call all three "anchoring".

Bias	The trigger	Money example
Anchoring bias	The first number you see pulls your estimate toward it	A ₹2,499 strikethrough makes ₹999 feel cheap
Framing effect	The same fact worded differently changes your choice	"Save 40%" versus "Save ₹400" on an identical deal
Confirmation bias	You favour evidence that fits a belief you already hold	Reading only the bullish takes on a stock you own

How do you reduce anchoring bias?

The reliable fixes are process, not willpower, because awareness alone barely moves the effect. Research on debiasing points to a few moves that genuinely help.

Generate your own number first. Form an independent estimate (the market salary, the fair price, the comparable home value) before you ever see the seller's or the recruiter's figure, so your adjustment starts somewhere honest. Consider the opposite, the best-supported technique in the debiasing literature: deliberately ask why the anchor might be too high or too low before you commit. Get a second independent data point instead of deferring to the first or loudest one; a second salary benchmark or a second valuation blunts the pull. And for everyday spending, widen the frame back out: convert a monthly EMI into its full total, turn a "40% off" into the actual rupee price, and price an item on its own merits with the strikethrough covered up. The goal is not to out-think the anchor in the moment, which even judges fail to do, but to arrange things so you meet your own number before you meet theirs.

What this guide does not cover

This guide explains anchoring bias and how it operates in money decisions; it is not personalised financial, legal, or negotiation advice. It does not tell you what price to accept, what salary to demand, or what to do with any specific investment, and the investing section describes a documented pattern and is not a recommendation to buy, sell, or hold. For decisions tied to your own portfolio or taxes, a SEBI-registered adviser or a qualified professional is the right call. For the wider set of money biases this one travels with, see our behavioural finance overview.

Frequently asked questions

What is anchoring bias in simple terms? Anchoring bias is the tendency to rely too heavily on the first number you encounter (the "anchor") when making an estimate or decision, then adjust away from it insufficiently, even when the anchor is random or irrelevant. In Tversky and Kahneman's 1974 experiment, people watched a wheel rigged to stop on 10 or 65, then estimated what percentage of African countries are in the UN. Those who saw 10 answered a median 25%; those who saw 65 answered 45%. A meaningless spin moved the answer 20 points. In money, the first price, the first salary offer, or the price you paid becomes the reference everything else is measured against.

What is a simple example of the anchoring effect? A shop tag that reads "was ₹2,499, now ₹999" is anchoring in one line. The ₹2,499 sets your reference point, so ₹999 feels like a steal, even if the item never sold at ₹2,499. The same trick runs a whole restaurant menu (an ₹2,500 dish near the top makes the ₹900 mains look moderate) and every three-tier pricing page (a pricey top plan makes the middle plan you were meant to buy look sensible). The first number does the work; the "discount" is measured against a figure the seller chose.

Why does anchoring bias happen? Two mechanisms explain it. Anchoring-and-adjustment (Tversky and Kahneman, 1974) says you start from the anchor and adjust toward your real estimate but stop too early, so the answer stays pulled toward the anchor. Selective accessibility (Strack and Mussweiler, 1997) says that testing "is it higher or lower than the anchor?" makes anchor-consistent facts more mentally available, which then biases your estimate. Selective accessibility is why even an absurd anchor works: adjustment cannot explain why "older than 140?" moves anyone, but anchor-primed memory can.

Does anchoring bias affect experts too? Yes, and they usually deny it. When Northcraft and Neale (1987) showed real-estate agents the same house with only the printed list price changed, professional appraisals swung about $14,000 across a $30,000 anchor range, yet only about 19% of agents admitted the list price had influenced them. Englich, Mussweiler and Strack (2006) had experienced judges roll dice loaded to sum 3 or 9, then recommend a sentence for an identical case; the low roll averaged 5.28 months and the high roll 7.81 months. Expertise does not switch anchoring off.

How does anchoring bias affect investors? Investors anchor to a reference price and judge value against it instead of against current fundamentals. The most common anchor is the price you paid: buy at ₹800, watch it fall to ₹550, and the urge is to wait for it to "come back to ₹800" as if the market remembers your cost. Other anchors are a stock's 52-week high, the IPO issue price, an analyst's target, and round index levels like Sensex 80,000 or Nifty 25,000, which get treated as meaningful thresholds rather than arbitrary points. This is general education, not investment advice; a SEBI-registered adviser can address your specific situation.

What is the difference between anchoring bias and the framing effect? Anchoring is about a number; framing is about wording. Anchoring bias pulls your estimate toward the first figure you see, such as a strikethrough price or an opening offer. The framing effect changes your choice when the same fact is worded differently, such as "90% survival" versus "10% mortality", or "Save 40%" versus "Save ₹400" on an identical deal. Anchoring is a reference-point problem; framing is a presentation problem. Confirmation bias is a third thing: favouring evidence that supports a belief you already hold.

How do retailers use anchoring bias? They show a high number first so the price they actually want is judged against it. In India, an inflated printed MRP acts as a legal ceiling and a psychological anchor, so a product sold well below MRP feels like a win even at a healthy margin. Online, the struck-through "original" price does the same job. US regulators have treated fake reference prices as illegal: JCPenney settled a false-pricing class action for $50 million, with filings alleging only about 0.2% of products ever sold at the "regular" price. Decoy items and expensive menu dishes are the same mechanism.

How can you reduce anchoring bias with money? Decide your own number before you see theirs. Research supports three moves: generate an independent estimate first (research the market salary, the fair price, or the comparable home before anyone quotes you), consider the opposite by asking why the anchor might be too high or too low, and get a second independent data point rather than deferring to the first or loudest one. For everyday spending, convert the frame back: turn a "₹4,999 per month" EMI into the full ₹60,000-plus total, and a "40% off" into the actual rupee price. Awareness alone barely helps, so the fix is process, not willpower.

In summary

Anchoring bias is the first number you see quietly setting the terms for every number that follows, and fifty years of experiments show it barely cares whether that first number makes any sense. A rigged wheel moved UN estimates 20 points. Loaded dice moved real prison sentences by half. A house's list price moved trained appraisers by $14,000 while they insisted it did not. The effect is not a novice's mistake; it is how the mind handles numbers under uncertainty.

Where it costs you is in the anchors someone else picks: the inflated MRP, the strikethrough, the low opening offer, the price you paid for a stock that has since moved on. You cannot out-stare an anchor in the moment, because the experts who try still fail. What you can do is meet your own number first, question why theirs might be off, and convert every clever frame back into the plain total. The next read in this cluster is the sunk cost fallacy, the bias that keeps you paying because of what you have already spent.

Sources

• Tversky, A. and Kahneman, D. (1974), Judgment under Uncertainty: Heuristics and Biases, Science 185(4157), science.org
• Ariely, D., Loewenstein, G. and Prelec, D. (2003), Coherent Arbitrariness, Quarterly Journal of Economics 118(1), academic.oup.com
• Northcraft, G. and Neale, M. (1987), Experts, Amateurs, and Real Estate, Organizational Behavior and Human Decision Processes 39(1), sciencedirect.com
• Englich, B., Mussweiler, T. and Strack, F. (2006), Playing Dice With Criminal Sentences, Personality and Social Psychology Bulletin 32(2), journals.sagepub.com
• US Federal Trade Commission, Guides Against Deceptive Pricing, ftc.gov