Moment-Based Estimation of Latent Class Models of Event Counts

University of California at San Diego, Department of Economics, Discussion Paper No. 98-12

Posted: 14 Aug 1998

See all articles by Partha Deb

Partha Deb

City University of New York, CUNY Hunter College - Department of Economics

Xing Ming

Indiana University Purdue University Indianapolis (IUPUI)

Pravin K. Trivedi

Indiana University Purdue University Indianapolis (IUPUI) - Department of Economics

Abstract

This paper develops and implements a GMM estimator for latent class models suitable for count data. The estimator uses conditional moment restrictions derived from standard count models. Both the efficient and consistent variants are considered. The implementation of optimal GMM based on semiparametric estimates of the weighting matrix appears to be problematic as the matrix is not guaranteed to be positive definite. A suboptimal variant which ensures positive definiteness is found to work well in computer simulations. The paper compares maximum likelihood and GMM estimators for Poisson based mixtures in two applications to U.S. health utilization data for the elderly from the National Medical Expenditure Survey.

JEL Classification: C21, C25, D12, I11

Suggested Citation

Deb, Partha and Ming, Xing and Trivedi, Pravin K., Moment-Based Estimation of Latent Class Models of Event Counts. University of California at San Diego, Department of Economics, Discussion Paper No. 98-12. Available at SSRN: https://ssrn.com/abstract=106548

Partha Deb

City University of New York, CUNY Hunter College - Department of Economics ( email )

695 Park Avenue
New York, NY 10021
United States

Xing Ming

Indiana University Purdue University Indianapolis (IUPUI) ( email )

1309 E. 10th St.
Bloomington, IN 47405
United States

Pravin K. Trivedi (Contact Author)

Indiana University Purdue University Indianapolis (IUPUI) - Department of Economics ( email )

Wylie Hall
Bloomington, IN 47405-2100
United States

Register to save articles to
your library

Register

Paper statistics

Abstract Views
439
PlumX Metrics