Presentation of paper

Bayesian Estimation of Policy Effects in Dynamic Discrete Choice Models

Speaker

Neelam Jain - City University  London

Abstract

The standard procedure of policy evaluation in dynamic structural models does not capture the parameter uncertainty. The policy evaluation there is based on the parameter estimate. We argue that the appropriate policy effect should reflect the inaccuracy of the estimate, or parameter uncertainty. Calculating the standard error in the standard estimation methods is computationally demanding. We provide a Bayesian method of policy evaluation in structural estimation that is computationally superior and captures parameter uncertainty. Our algorithm solves the dynamic model, estimates the parameters and conducts policy experiments simultaneously in a single algorithm.

For more information: crenos@unica.it

Tel. +39 070 6756414    Fax +39 070 6756402