Estimation of a Demand System with Limited Dependent Variables
Chung L. Huang
Department of Agricultural & Applied Economics
The University of Georgia
313-E Conner Hall
Athens, GA 30602-7509
Phone: (706) 542-0747 Fax: (706) 542-0739
E-mail: chuang@agecon.uga.edu
Steven T. Yen
Department of Agricultural Economics
University of Tennessee
302 Morgan Hall
Knoxville, TN 37996-4518
Phone: (865) 974-7231 Fax: (865) 974-7484
E-mail: syen@utk.edu
Abstract
The study employs the full-information maximum-likelihood method to estimate a censored
translog demand system. U.S. household consumption of steak, roast, and ground beef are used
to demonstrate the application of the estimation procedure. The proposed methodology produces
more efficient estimates than the popular two-step procedures found in demand literature.
Key Words: full-information maximum-likelihood method, two-step procedures, censored
translog demand system, demand elasticities.
Selected paper presented at the AAEA annual meeting, July 28-31, 2002, Long Beach, CA.
Copyright 2002 by Chung L. Huang and Steven T. Yen. All rights reserved. Readers may make verbatim
copies of this document for non-commercial purposes by any means, provided that is copyright notice
appears on all such copies.
Estimation of a Demand System with Limited Dependent Variables
Introduction
Red meat consumption in the United States has decreased significantly in the past few decades
following a steady downward trend started in the late 1970s. Per capita consumption of beef,
accounting for approximately 60% of total read meat, reached an all-time high of 88.8 lbs. in
1976. It dropped about 19% to 72.1 lbs. in 1980 and remained relatively flat in the early 1980s,
and then steadily declined from 74.6 lbs. in 1985 to 56.1 lbs. in 1998 (Figure 1). Considerable
interest and concern have focused on the trend of declining red meat consumption with special
attention given to beef consumption. Smallwood, Haidacher, and Blaylock provided a