Varadarajan V. Chari, Principal Investigator

Solving and Estimating Dynamic General Equilibrium Models with Heterogeneity

This research group applied Bayesian statistics tools that they had developed in their previous work at the Supercomputing Institute to estimate and compare a set of sticky price models widely used in the literature. Their results were presented in a paper entitled “How Important Are Real Wage Rigidities? A Bayesian Approach.”

The main results of the research were:

• The researchers’ model delivers reasonable point estimates for the duration of prices and wages, of three and five quarters respectively. The fraction of wage setters that follow a real wage indexing rule is about one half.
• As real wage rigidities are removed, the estimated duration of prices increases to high levels. The researchers view this result as a natural consequence to substitute for the lack of endogenous persistence associated with real wage rigidities.
• The group’s estimates point at little evidence suggesting backward-looking behavior in the inflation equation.
• Using the marginal likelihood of the data implied by each model, the model presented in this paper performs best.

A second project involved working on a method to perform nonlinear likelihood-based inferences in Dynamic Equilibrium Economies. Although this type of model is a standard tool in quantitative economics, existing literature has been forced so far to use moment procedures or linearization techniques to estimate these models. The researchers used Monte Carlo filtering to evaluate the likelihood function implied by the nonlinear solution of the model. As a consequence, they could perform likelihood-based inference, either searching for a maximum (Quasi-Maximum Likelihood Estimation) or simulating the posterior using a Markov Chain Monte Carlo algorithm (Bayesian Estimation).

Research Group

Jesus Fernandez-Villaverde, Graduate Student Researcher
Juan F. Rubio, Graduate Student Researcher