Description

Structural estimation in economics often makes use of models formulated in terms of moment conditions. While these moment conditions are generally well-motivated, it is often unknown whether the moment restrictions hold exactly. We consider a quasi-Bayesian approach for performing inference on structural parameters while relaxing the restriction that moment restrictions hold exactly. Within this context, we prove new Bernstein-von Mises (BvM) type theo- rems for the quasi-posterior distributions, which can be used to obtain tractable approximations in practical settings. We illustrate the approach through simulation and empirical applications. Our applications illustrate that we can obtain informative inference for structural objects, even allowing for substantial relaxations of the requirement that moment conditions hold exactly.