Adaptive economizing in nonlinear environments: Implications for economic modelling and policy analysis

That trajectories of variables in economic models can become chaotic when natural non-linearities are incorporated, and do so generically, is a fact well established in a great variety of contexts and, presumably, one that is well known. Its implications, however, for empirical modelling and policy analysis seem hardly to have been considered at all. In these remarks I want to suggest what some of these implications are and what they imply for our research agenda. To illustrate the central points at issue three alternative business-cycle models and their implications for monetary policy will be briefly described.