Anchoring Persistence Error

Anchoring persistence error happens when an initial piece of information keeps affecting later decisions. People stick to the first idea even when new facts show it is wrong.

Anchoring persistence error denotes the cognitive bias in which initial information disproportionately influences subsequent judgments and estimates. This bias persists despite additional evidence, causing under-adjustment from the initial anchor in decision-making contexts.

A car dealer tells you the sticker price is $40,000. Even after negotiating a "big discount" down to $36,000, you feel satisfied — anchored to that opening figure. If you had started from a market-value search showing comparable cars sell for $31,000, your final offer would have been far lower.

In a merger valuation, the acquiring firm's analyst opens the discounted-cash-flow model with a precedent transaction multiple of 12× EBITDA drawn from a peak-market deal. Despite subsequent comparables analysis yielding a central estimate of 8.5×, the final fairness opinion anchors near 10.8× — a classic under-adjustment from the initial reference node. The endowment-induced anchor introduced by the precedent transaction propagates through the evidence accumulator, suppressing posterior variance collapse around the unanchored central estimate and biasing the commitment node toward the inflated multiple.

A first clue sets a starting point for thinking. Later information is ignored or changed too little because the start feels right.

An initial anchor imposes a reference constraint on the decision representation, with higher weight assigned to early evidence versus subsequent signals; the asymmetry in weighting limits corrective adjustments. Structural elements like the initial reference node and downstream update pathways mediate biased propagation of estimates.

Compare the first idea to other numbers or opinions before deciding. Ask yourself how you would judge without the first suggestion.

Explicitly elicit and record an unanchored estimate before presenting any reference values, then apply calibrated debiasing adjustments informed by external benchmarks. Introduce procedural checks that reweight incoming evidence to counteract initial reference dominance.

Insufficient adjustment; Overreliance on initial source; Delayed correction

An adversarial actor can deliberately introduce an extreme or fabricated initial figure — a salary offer, a negotiation price, a risk estimate, or a casualty projection — knowing that even after partial correction the final judgment will remain biased toward that anchor. This technique is routinely weaponized in high-stakes negotiations, auction design, and disinformation campaigns where the first number to reach a target audience sets a durable reference-point constraint that subsequent rebuttals struggle to fully displace.

Before any reference value is introduced, practitioners should elicit and commit to an unanchored prior estimate in writing, making the baseline explicit and harder to silently revise. Structured adversarial review — assigning a team member to argue from an alternative anchor — forces explicit reweighting of downstream evidence against the initial reference node. Calibration training with historical base rates and external benchmarks builds the habit of reference-class calibration, reducing default weighting assigned to the first datum encountered.