CFDP 573

On the Consistency of Non-Linear FIML

Author(s): 

Publication Date: December 1980

Pages: 19

Abstract: 

Examples are given which show that: (i) normality is not necessary for the consistency of the quasi maximum likelihood estimator in the nonlinear simultaneous equations model (nonlinear FIML) even when there are major departures from linearity; and (ii) the lemma which is used extensively by Amemiya [2] in the theoretical development of the properties of nonlinear FIML under the assumption of normality is, as presently stated, incorrect.

See CFP: 549