Description:
Base rate neglect typically involves the underweighting of priors in a world with objective probability distributions (Kahneman and Tversky, 1973). In reality, however, individuals often face scenarios with ambiguous priors, in which they have no full confidence. We show, theoretically, that base rate neglect, which is typically regarded as irrational behavior under risk, can be a preference maximizing strategy of ambiguity-averse individuals, who have an incentive to hedge the ambiguity in the base rate by assigning less weight to it. The evidence from a laboratory experiment supports our theoretical analysis: 1) without accounting for base rate ambiguity, subjects seem to underweight the base rate, and more so for more ambiguous base rates; 2) when accounting for base rate ambiguity, subjects are surprisingly sophisticated in weighting the base rate; 3) more ambiguity-averse subjects underweight the base rate more