Are Policy Variables Exogenous? : The Econometric Implications of Learning while Maximizing

This book explores the econometric implications of a policy-maker's learning. In a simple discrete model the policymaker maximizes the discounted sum of tax revenues subject to a constraint which involves an unknown parameter specifying the occurrence of Bayesian learning. An optimal balance needs to be found between maximizing current payoff and generating information which enhances the efficiency of maximization in subsequent periods. Learning is demonstrated to affect the exogeneity status of policy variables in small samples and to have implications analogous to the phenomenon in the focus of the Lucas critique. Active learning is proven to be a distinct cause of time inconsistency of optimal plane. It is argued that learning cannot be dismissed as a merely transitory source of these phenomena. A simulation exercise supplies quantitative evidence. Finally, the data generated are used to perform empirical exogeneity tests. Results show that the effect of Bayesian learning is empirically detectable. Learning is generally accepted as the driving force towards a rational expectations equilibrium with perfect information. In this context, the result on the loss of exogeneity of policy variables is an out-of-equilibrium econometric implication of the rational expectations hypothesis. This book is the first to explore the effect of an informational feedback on exogeneity