The anchoring effect

A reader's summary of the tendency to lean on the first number we see: the experiments that named it, the explanations psychologists have offered, where it shows up in courtrooms, salary talks and checkout screens, and why it matters most in the week every shop crosses out a price.
The term at a glance
The anchoring effect is the tendency for a judgment about a number (a price, a probability, a sentence, a salary) to be pulled toward a reference value encountered just before, even when that value is arbitrary. It is one of the best replicated findings in behavioral science. Online it is mostly discussed around shopping: the strikethrough “was” price, the “compare at” label, the premium plan placed first on a pricing page so the middle plan looks reasonable, and the tipping screen whose first suggestion is 25 percent, which is a large part of tipflation. It is also the standard piece of advice in salary threads: whoever names a number first sets the terms of the conversation.
Origin
The term comes from Amos Tversky and Daniel Kahneman's 1974 paper in Science, “Judgment under Uncertainty: Heuristics and Biases”, which described three mental shortcuts: representativeness, availability, and “adjustment and anchoring”. In the best-known demonstration, participants watched a wheel of fortune that appeared random but was rigged to stop at either 10 or 65. They were asked whether the percentage of African countries in the United Nations was higher or lower than that number, and then to give their own estimate. The median answer was 25 percent among those who had seen 10, and 45 percent among those who had seen 65.
A second demonstration needed no wheel. Students were given five seconds to estimate a product written either as 8 × 7 × 6 × 5 × 4 × 3 × 2 × 1 or as 1 × 2 × 3 × 4 × 5 × 6 × 7 × 8. The median estimate was 2,250 for the descending sequence and 512 for the ascending one. The correct answer, 40,320, was far above both: the first few steps of the calculation acted as an anchor, and neither group adjusted enough.
How it works
Tversky and Kahneman's explanation was insufficient adjustment: people take the anchor as a starting point and move away from it until they reach a value that seems acceptable, then stop. Later work by Nicholas Epley and Thomas Gilovich in the early 2000s found that this describes anchors people generate themselves quite well (knowing that water boils at 100 °C at sea level, and adjusting from there to guess its boiling point on Everest), and that people under time pressure or distraction adjust even less.
For anchors supplied by someone else, Fritz Strack and Thomas Mussweiler proposed in 1997 a selective accessibility model. Asking whether a quantity could be as high as the anchor makes people search memory for reasons that it might be, and those reasons are then at hand when they give their own estimate. That explains why the effect is strongest when the anchor is at least loosely plausible, and why it links naturally to confirmation bias: the anchor works by deciding which evidence comes to mind first.
Main ideas
The first number sticks
When people estimate an uncertain quantity, a number they have just seen pulls the estimate toward it, even when they know the number is irrelevant.
Adjustment stops too early
People start from the anchor and adjust away from it, but stop once they reach a value that seems plausible, which leaves them closer to the anchor than they should be.
Anchors make evidence feel relevant
Considering whether the answer could be the anchor brings to mind the facts that fit it. The anchor does not just set a starting point; it changes what the mind retrieves.
Expertise does not protect
Estate agents, judges and car mechanics have all been shown to drift toward an anchor in their own field, and they generally deny that it influenced them.
The experts
In 1987 Gregory Northcraft and Margaret Neale gave estate agents in Tucson, Arizona, a tour of a real house and a ten-page information packet that was identical except for the listing price. Agents shown a higher listing price gave higher appraisals, and most of them said that the listing price had not been among the factors they considered. Students given the same task showed a similar pattern, but were more willing to admit the influence.
The most striking result came from courtrooms. In a 2006 study, Birte Englich, Thomas Mussweiler and Fritz Strack asked experienced German judges to read a case file about a shoplifter and then roll a pair of dice to determine the prosecutor's sentencing demand. The dice were loaded to show either 3 or 9. Judges who rolled 9 handed down sentences averaging about 8 months; those who rolled 3, about 5 months. They knew the number had come from dice.
Prices and negotiation
Retail is anchoring's natural habitat. A reference price, real or not, tells shoppers what the item is “worth” before they see what it costs. Marketing courses often repeat the story, told by the consumer researcher Itamar Simonson, of a Williams-Sonoma bread maker that sold poorly at $275 until the shop added a larger $429 model next to it, after which the cheaper one began to sell. The expensive option did not need to sell; it needed to be seen. Pricing pages with an “enterprise” tier at the top, wine lists with one very expensive bottle and the first price a salesperson mentions all work the same way. A related trick, the decoy, adds an option that is clearly worse than one of the others to make that one look better.
The clearest evidence that shoppers rely on anchors came from a shop that removed them. In 2012 the American department store J.C. Penney, under its new chief executive Ron Johnson, replaced constant promotions with “fair and square” everyday prices. The prices were often no higher than the old sale prices, but there was nothing to compare them with. Sales fell by roughly a quarter that year, Johnson was replaced in 2013 and the coupons came back.
In negotiation, research by Adam Galinsky and Thomas Mussweiler in 2001 found that the side making the first offer tended to end with a better deal, because the opening number anchored the bargaining range, unless the other side deliberately focused on information that contradicted it. That is the basis of the advice in salary threads to let the employer name a figure first, or, if you must go first, to open ambitiously with a reason attached.
Black Friday and the law
Inflated “was” prices became common enough that regulators stepped in. Since 28 May 2022, under the European Union's Omnibus Directive (2019/2161), a shop in the EU that announces a price reduction must show as the previous price the lowest price it charged in the preceding 30 days. A retailer that raises prices in October to cut them in November can no longer use the October price as the anchor. Consumer groups and price-tracking sites still find many Black Friday deals that are no lower than the price a few weeks earlier, and outside the EU rules on reference prices are generally looser. The strikethrough remains the single most efficient piece of typography in retail, which is why it sits alongside dark patterns and shrinkflation in discussions of how prices are presented.
Critique
- Robust, but not everywhere. The classic anchoring tasks replicated strongly in the large Many Labs project of 2014, among the most reliable effects tested. Anchors that are merely incidental are another matter: the famous result in which the last two digits of a person's social security number shifted what they would pay for wine and chocolate has proved much weaker when others tried to reproduce it.
- Sometimes the anchor is information. A listing price or a prosecutor's demand often does carry real information about value or seriousness. Leaning on it is not always a bias; the problem is the inability to discount it when it is known to be meaningless.
- Knowing does not cure it. Warning people about anchoring, or offering money for accurate answers, reduces the effect only modestly. The more reliable defense is to generate one's own reference point first, or to argue deliberately against the anchor, the “consider the opposite” strategy.
- It can be overused as an explanation. Online, any price someone dislikes gets called anchoring. Many high prices are simply high prices, and many discounts are real.
Impact
Anchoring helped launch the heuristics-and-biases programme that earned Kahneman the 2002 Nobel prize in economics and became one of the founding results of behavioral economics. It shaped how prices are displayed, how negotiators are trained and how lawyers think about damages claims, and it has been found in new places, including in large language models, whose numerical answers can shift with a number planted in the prompt. For a reader in the last week of November, the practical lesson fits on a sticky note: before looking at the crossed-out price, decide what the thing is worth to you.