Anchoring Bias

Anchoring and Awareness

Anchoring describes our tendency to rely heavily on the first piece of information we encounter when making a decision or estimate. That initial value—whether a price, a number from a spinning wheel, or a sales projection—serves as a reference point from which we adjust, often insufficiently. Classic demonstrations include Kahneman and Tversky’s wheel-of-fortune experiment: participants who saw a low random number gave lower estimates for the percentage of African nations in the UN, while those who saw a high number gave higher estimates, despite knowing the anchor was arbitrary. A similar effect emerges when people are asked to consider the last four digits of their Social Security number before guessing the number of dentists in Manhattan—the guess correlates with that unrelated number. In everyday life, a shopper sees a $50 shirt marked 25% off and judges $37.50 as a good deal, anchored to the original price, even if the same shirt is available elsewhere for $30.

People may not recognize how much an anchor influences them. Wilson et al. (1996) found only a moderate correlation between self-reported awareness and the size of the anchoring effect, and the vast majority of participants who denied being influenced still showed a clear bias. A person can show anchoring without reporting awareness of it. Explicit warnings may also fall short: both Wilson et al. and Quattrone et al. (1981) told participants not to be influenced by the anchor, yet anchoring persisted.

Anchoring can occur even without an overt comparative judgment that draws attention to the anchor, as long as the anchor is processed sufficiently. Wilson et al. demonstrated this by having participants perform five pages of computations on a large number; later estimates of cancer incidence were higher than in a control condition. One page of computations did not produce the effect, suggesting that the amount of processing mattered in that experiment. Processing the number did not necessarily mean recognizing its influence.

Further complicating the awareness picture is the perception of informativeness. Chapman and Johnson (1999) found that participants who believed a randomly generated anchor was informative sometimes showed a larger effect and sometimes a smaller one, but in both cases a significant anchoring effect remained among those who explicitly judged the anchor uninformative. Recognizing that the anchor is uninformative may therefore be insufficient.

Forecasts can acquire an unwarranted sense of solidity when a spreadsheet turns uncertain projections into neat rows and columns. Dragging a cell to extend a trend ad infinitum creates a tangible object that then serves as an anchor. Planners tunnel on that number, building beliefs around it because, as Kahneman noted, it is mentally easier to compare an idea to a reference point than to evaluate it in absolute terms. A figure we generated ourselves can become the reference point, even when its basis is weak.

A first offer can also anchor a salary negotiation, influencing counteroffers and the outcome. That can favor the first mover, but it does not establish that opening first is always advantageous. The same dynamic plays out in consumer purchases, where an original asking price can make a discount seem attractive, and even in predicting our own future emotions, where our current emotional state serves as the reference point, making it hard to imagine feeling differently.

Because we are so often oblivious, simply knowing about anchoring does not reliably reduce it. Some practical strategies involve deliberately breaking the single-reference-point mindset: comparing multiple options rather than evaluating one in isolation, reframing the problem to consider absolute value instead of relative adjustment, and slowing down decisions to think them through. While mere awareness may fall short, actively seeking alternative anchors and checking independent evidence can help challenge an initial estimate.