A THEORY OF LEARNING WITH HETEROGENEOUS AGENTS

We study learning in organizations comprised of short-lived agents with heterogeneous beliefs. It is shown that the sequence of actions converges in probability to the set of ex-post optimal actions. Under some identifiability conditions, beliefs also converge in probability to point mass on the truth. Moreover, while actions and posterior beliefs converge in probability, they can simultaneously almost surely fail to converge. Our simulations of Kiefer's 1989 monopoly learning model suggest that if the firm is operated by a sequence of short-lived managers with heterogeneous beliefs then complete learning of the demand curve obtains.