Information Matrix Tests for the Composed Error Frontier Model

University of Illinois, Bureau of Economics & Business Research Working Paper No. 99-0101

25 Pages Posted: 15 Apr 1999

See all articles by Anil K. Bera

Anil K. Bera

University of Illinois at Urbana-Champaign - Department of Economics

Naresh C. Mallick

Georgia State University - Department of Economics

Date Written: March 1999

Abstract

Estimation of composed error frontier models is generally conducted under certain strict assumptions. In practice however, these assumptions are not tested thoroughly. This is probably because simple workable tests are not yet available for these models. This paper develops easily computable specification tests for half-normal composed error frontier models. The tests are based on the information matrix (IM) and moment test principles. These tests are applied to the well - known Cowing (1970) steam-electric data set. Our tests reveal no serious misspecification of the cost model, while for the output model the null hypothesis of correct specification of rejected strongly.

JEL Classification: C19

Suggested Citation

Bera, Anil K. and Mallick, Naresh Chandra, Information Matrix Tests for the Composed Error Frontier Model (March 1999). University of Illinois, Bureau of Economics & Business Research Working Paper No. 99-0101. Available at SSRN: https://ssrn.com/abstract=159085 or http://dx.doi.org/10.2139/ssrn.159085

Anil K. Bera (Contact Author)

University of Illinois at Urbana-Champaign - Department of Economics ( email )

410 David Kinley Hall
1407 W. Gregory
Urbana, IL 61801
United States
217-333-4596 (Phone)
217-244-6678 (Fax)

Naresh Chandra Mallick

Georgia State University - Department of Economics ( email )

P.O. Box 3992
Atlanta, GA 30302-3992
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
18
Abstract Views
763
PlumX