Essays on learning

This dissertation considers two distinct economic problems where individual and social learning play a central role. The first chapter analyzes the stability of monetary regimes in a decentralized economy where fiat money is endogenously created by the government, information about its value is imperfect, and agents only learn from their private experience. It establishes that in poorly informed economies, monetary stability is only possible with sufficiently patient governments, whereas in economies where agents gather information more easily, monetary stability is possible even with impatient governments. A dynamics of fiat money acceptability is obtained that resembles historical accounts of the rise and eventual demise of overissued paper money. Moreover, the results from this chapter provide an explanation for the fact that, despite its obvious advantages, the widespread use of fiat money is only a very recent development. The second chapter analyzes how learning in society can overcome, in two-armed bandits, the so-called Rostchild effect. It considers an economy populated by a continuum of infinitely lived agents where each one of them faces a two-armed bandit, and the stochastic payoffs of these bandits are the same for each individual. These agents are randomly and anonymously matched in every period, and they observe the current action choice of their partner. Two results are established. First that the resulting game has a symmetric equilibrium in Markovian strategies. Second, and the main result, that in any symmetric sequential equilibrium the fraction of the agents choosing the inferior arm converges to zero