Using generalized impulse response functions to estimate nonlinear dynamic models

This article proposes the use of generalized impulse response functions as a natural solution to the issues that arise when estimating nonlinear dynamic general equilibrium models by impulse response matching. Using a small-scale New Keynesian model with downward nominal wage rigidity as a testing ground, Monte Carlo analysis shows that the proposed estimation strategy delivers sharper parameter estimates than the matching of traditional impulse responses currently employed in the literature. An empirical application using U.S. data illustrates the use of the proposed estimation strategy. Results support the view that nominal wages are downwardly rigid, predict a frequency of nominal wage cuts in line with the microdata, and quantify frictional adjustment costs as a proportion of total output.