Do Institutions Promote Rationality? An Experimental Study of the Three-Door Anomaly

The three-door problem is an astounding example of a systematic violation of a key rationality postulate. In this seemingly simple individual decision task, most people initially fail to correctly apply Bayes’ Law, and to make the payoff-maximizing choice. Previous experimental studies have shown that individual learning reduces the incidence of irrational choices somewhat, but is far from eliminating it. We experimentally study the roles of communication and competition as institutions to mitigate the choice anomaly. We show that the three-door anomaly can be entirely eliminated by these institutions.