Anchoring

In plain terms

The first number you hear becomes the place your thinking starts. Every estimate after that is an adjustment away from it, and the adjustment is almost always too small.

Amos Tversky and Daniel Kahneman demonstrated this in 1974 with a rigged wheel of fortune. Participants watched the wheel land on a number, then estimated what share of United Nations member states were African. People who saw a high number gave high estimates. The wheel was visibly random and had nothing to do with the question. The estimates moved anyway.

Anchors are built into most things you buy or negotiate. The list price, the opening offer, the crossed-out "was" price, the default tip percentages on the screen. None of them are neutral, and all of them are first.

Why it matters

An anchor hands you a starting point that may carry no information about the answer, and starting points are sticky. Adjustment stops when the number feels defensible, not when the evidence runs out.

This matters most where the stakes are highest. In salary talks, house sales, contract disputes, and budget negotiations, whoever names a figure first often defines the range everyone argues inside, even when that figure was invented on the spot.

Knowing about anchoring helps less than you'd hope. The effect survives when people are warned in advance, when they're told the anchor is random, and when they're offered money to be accurate. Awareness is not a defense. Process is.

Canonical example

Seller's listing: $640,000.

Buyer, to their agent: "That's too high. Let's come in at $600,000 and see what they say."

The buyer never looked up what comparable homes on the street actually sold for. The $600,000 offer feels like a firm move because it's forty thousand below the ask, and "forty thousand below the ask" is a relationship to the seller's number, not to the house.

Run the same buyer past a $560,000 listing for the same property and the offer comes in near $540,000. The house didn't change. The anchor did.

Counter-example (not anchoring)

Buyer: "Comparable sales on this block are running $520,000 to $555,000. They're asking $640,000. I'll offer $535,000 and show them the comps."

The list price was still the first number the buyer saw. It just isn't doing the work. The buyer built an independent estimate and treated the asking price as information about the seller's expectations rather than as a starting point for arithmetic.

Not every anchor is arbitrary, either. An appraiser's valuation, a published fee schedule, or last year's actual spend are all reference points, and they deserve weight because of where they came from. The bias is in treating an uninformative number as informative.

The line: did the number inform your estimate, or did it stand in for the work of making one?

How to fix it

If you've been linked here, someone thinks a figure you're working from arrived before your reasoning did. The fix is order. Write down your own number, with a range, before you look at theirs. If you've already seen theirs, build the estimate from scratch anyway: what are the comparables, what did this cost last time, what would you have said if nobody had spoken first? Then compare. When you catch yourself adjusting up or down from a figure, stop and ask where that figure came from. If the answer is "they said it," it isn't evidence.

If you're on the receiving end, don't argue with the anchor, replace it. "Before we go back and forth on that number, what's it based on?" followed by your own sourced figure moves the conversation onto ground where the evidence decides. Haggling downward from someone else's opening bid concedes the range before you start.