COSTLY OPTIMIZATION - AN EXPERIMENT

An experimental technique is developed and used to examine how human subjects make choices when optimization is costly. The total cost of making a choice is defined to include two parts: (1) the direct cost of making the decision, and (2) the indirect cost of making a choice that is not best among the available alternatives. The total cost of choice for subjects in the experimental economy ranged from 15 to 42 percent of the economy's resources. The hypothesis that agents are rational on average was strongly rejected, while the hypothesis that agents are rational in the long run fared better. © 1991.