anchoring bias

You're buying a car. The salesperson names the sticker price. You don't accept it - you're a careful negotiator. You push back, ask for a lower number, get a counter. The final price is lower than the sticker price.

But here's the question: was your final number closer to the market value of the car, or closer to the sticker price?

Almost always, it's closer to the sticker price. The negotiation began there; the process moved away from it, but not far enough. The first number set the frame, and every subsequent judgment happened within that frame.

That's anchoring bias - the tendency to rely too heavily on the first piece of information encountered when making judgments or estimates, even when that information is irrelevant or arbitrary.

What It Is

Anchoring bias (also called focalism) is the tendency to rely too heavily on the first piece of information encountered - the "anchor" - when making subsequent judgments, even if that information has no rational bearing on the estimate.

The effect is systematic and robust. When people need to estimate an unknown quantity or make a decision, they start with an initial value and adjust away from it - but these adjustments are typically insufficient. The final judgment remains biased toward the anchor, pulled by a reference point that was set before the reasoning began.

How the Mechanism Works

Anchoring arises from a heuristic process: when an anchor is presented, it activates related knowledge and establishes a reference point. Subsequent adjustments away from that point are limited by cognitive effort and the natural stopping point of the adjustment process - people stop adjusting when they reach a value that seems plausible, not when they reach the actual best estimate.

Two processes reinforce this:

Selective accessibility. The anchor activates information that is consistent with it - evidence, associations, and estimates that are adjacent to the anchor value. The adjustment starts from a biased information set.

Insufficient adjustment. Even when people know they should move away from an anchor, they tend not to move far enough. The anchor functions as a gravitational field: the estimate is pulled toward it before the reasoning process can complete.

Arbitrary numbers presented as anchors systematically shift people's estimates of unknown quantities. The anchor doesn't need to be plausible or even relevant. Research has shown that irrelevant anchors - such as the last two digits of a social security number - influence willingness to pay for goods in auction settings. People whose social security numbers ended in high digits were willing to pay more for unrelated items than people whose numbers ended in low digits. The number was random. The anchor effect was real.

A Hypothetical Scenario

Priya is negotiating the purchase of a second-hand piece of audio equipment. The seller lists it at GBP1,800 - significantly above its typical resale value of around GBP900. Priya researches the market before the conversation, knows comparable units sell for GBP800-1,000, and plans to offer GBP850. In the negotiation, the seller opens at GBP1,800 and Priya counters at GBP1,100, which feels like a significant move down from the ask. They settle at GBP1,250 - still GBP350 above the typical market rate.

The initial list price of GBP1,800 was the anchor. Priya's preparation gave her genuine information about market value, and she moved toward it - but not all the way. The anchor compressed the negotiating range. Even a well-prepared, research-backed buyer adjusted less than the information warranted.

Where It Shows Up

In negotiation, the anchor is set by whoever names the first number. That number becomes the reference point for the entire discussion, regardless of its legitimacy. In salary negotiations, job offers, property purchases, and commercial contracts, the party that frames the first number typically achieves a better outcome - not because the number is accepted, but because the final result gravitates toward it.

In consumer pricing, retail anchoring is deliberately engineered: the crossed-out original price beside the sale price sets an anchor that makes the sale price feel like a good deal, regardless of what the product's actual value is. The comparison is between the current price and the anchor, not between the current price and what the item is worth.

In estimation and forecasting, anchoring distorts professional judgments. When experts are given a starting estimate before producing their own, their outputs cluster around the starting value - even when they're instructed to ignore it and generate an independent estimate.

The Common Misunderstanding

The most common response to anchoring is: I know about this, so I can correct for it. Even when people are informed about the anchoring effect, their judgments remain biased toward the anchor. Awareness reduces the magnitude of the effect somewhat, but doesn't remove it. The mechanism operates partly below conscious deliberation - the anchor activates information and sets the reference point before the explicit reasoning begins.

A second misunderstanding: anchoring only applies to numbers. Any salient initial information can function as an anchor - an initial description of a person, a first impression of a plan, an opening characterisation of a situation. The numerical anchoring experiments are the most cleanly documented, but the principle extends to non-numerical framing.