Recent Advances in Solving and Estimating Dynamic, Stochastic Macroeconomic Models

Two of the major objectives of macroeconomic research are to explain the behavior of aggregate economic data and to predict the effects of policy interventions. Within the macroeconomics literature, there are two identifiable approaches to these issues. The reduced-form method involves specifying a statistical model for the variables of interest, estimating the parameters of the model, and answering the underlying question by analyzing the estimated values of the parameters or some function of the parameters. The coherence between the model and the data is of primary concern; theory, in general, plays a subordinate role. The structural approach, on the other hand, entails describing a theoretical model for the relevant macroeconomic variables and analyzing the relationships implied by the model to answer the questions of interest. An important feature of the theoretical model is that the parameters of the model be policy invariant; the parameters are structural, remaining fixed under hypothetical interventions. The magnitude of the roles that measurement and observation play in the structural approach have varied greatly over time, being central in the work of the Cowles Commission and, more recently, subsidiary in the real business-cycle (RBC) literature. The point of this essay is to discuss the second approach—the structural program in macroeconomics.