Alternative Semi-parametric Likelihood Approaches to Generalised Method of Moments Estimation.

Since L. P. Hansen's (1982) seminal paper, the generalized method of moments (GMM) has become an increasingly important method for estimation and inference in econometrics. This paper examines semiparametric quasi-likelihood approaches. Essentially, these methods embed sample versions of the moment conditions used in GMM in a nonparametric quasi-likelihood function by use of additional parameters associated with these moment conditions. Specification and misspecification tests may be defined which are similar in nature to the classical tests and are first-order equivalent to the corresponding GMM statistics. The structure of the semiparametric quasi-maximum likelihood estimator is explored for models estimated by instrumental variables. Copyright 1997 by Royal Economic Society.