La modélisation macro-économétrique dynamique

This paper presents the recent developments of macro-econometric modeling and discusses their advantages and limits. We first present the Sims critique and the Lucas critique. These two critiques have opened two new ways of macro-modeling. On the one hand, the Structural VAR approach allows to simply represent aggregate data with a small number of restrictions and to easily conduct various quantitative exercises (forecasting, dynamic multipliers of economic policy). On the other hand, the DSGE approach considers structural dynamic models wherein reduced forms are deduced from dynamic optimization problems (households, firms,...) with respect to a set of constraints. We thus present various quantitative methods and we discuss their advantages and limits.