A note on the asymptotic efficiency of the restricted estimation

Abstract

This paper provides a unified framework for the analysis of the stochastic and deterministicconstrained estimation. In a general framework it is show that stochastic restrictions methodestimates can be asymptotically more efficient than estimates ignoring prior information, andcan achieve efficiency of the restricted estimate if priorinformation grows faster than the sampleinformation in the asymptotics. As an example of the applicability of the previous result, themaximum likelihood stochastically restricted criterion is provided.