Recovering the Anchoring of Economic Valuations

We revisit the interpretation of the anchoring effect as evidence for nonexisting or arbitrary preferences. A theory of endogenous information acquisition (i.e., deliberation) is developed to rationalize the causal dependence of economic valuations on arbitrary anchor numbers. We identify theory-driven moderators to reconcile seemingly discrepant findings among original and follow-up anchoring experiments. We demonstrate in a meta-analysis that the anchoring effect may be systematically moderated by unintended experimental/individual differences. Moreover, apparent replication failure of the anchoring effect may be a false negative for self-countervailing treatment effects, because an anchoring treatment may cast opposite effects on different subjects within an experiment.