Pandora’s Rules in the Laboratory

Abstract

In theory, search conditions are a key determinant of Pandora’s rule (i.e. of the optimal search pattern), and of the eventual payoffs (Weitzman, 1979; Doval, 2018). We compare different search conditions in the laboratory and find strong evidence that they affect subjects’ order of inspection and payoffs greatly. Subjects are more conservative at the beginning of the search, and earn less if they are constrained to choose only among the inspected alternatives, than if they are not. These findings reinforce the empirical pertinence of the formal search literature, and generate novel insights regarding relevant settings of applied interest (e.g. the impact of market digitization on consumer behavior)