Indirect inference under stochastic restrictions

This paper investigates inference methods to introduce prior information in econometric modelling. The set up includes the nonlinear least squares and indirect inference estimators. The goal is to show that stochastic restrictions method estimates can be asymptotically more efficient than estimates ignoring prior information and can achieve efficiency of the restricted estimate if prior information grows faster than the sample information in the asymptotics. Finally, the proposed approach is applied to a macroeconomics model where high efficiency gains are shown.