IDENTIFICATION AND KULLBACK INFORMATION IN THE GLSEM

In this paper we employ the Kullback information apparatus in (a) obta ining the strong consistency of the maximum likelihood (ML) estimator in the standard version of the general linear structural econometric m odel (GLSEM); (b) deriving very succinctly the necessary and sufficien t (nas) conditions for identification by the use of exclusion restrict ions. The arguments given in (a), however, are equally applicable to a wide class of nonlinear models and the arguments in (b) are equally a pplicable in the context of more general types of restrictions. (C) 19 98 Elsevier Science S.A.