Abstract
The theory of consumer demand is the oldest and one of the most fruitful research areas in economics. Theoretically, it is concerned with the foundation of microeconomics, that is, allocation problems. Empirically, demand analysis provides policy-makers with important references for their decisions on welfare issues such as poverty and inequality measurement. The tax reform in Europe and the US is the drive for the new round of great interest in this area because demand analysis can roughly depict welfare consequences by its indices.
The understanding of consumer demand analysis contributes to knowledge of the methodology of economic studies. In the preceding study, the present author reviews the theory of consumer demand and discusses the properties of demand functions. The literature review in that study also discusses some relevant topics in this area such as the choice of functional forms, the aggregation problem, and the comparison of different systems. It provides some examples of demand specifications and describes their merits and weakness. However, it is a great step to make from theoretical models to empirical application of the models; how to deal with the data, how to make the models econometrically workable, especially how to interpret results are the most important issues in empirical work.
This paper estimates three demand system specifications, namely, the AIDS, the Rotterdam, and the LES using British household expenditure data for the years 1974-1993. The purpose of this study is to provide an empirical comparison of these models. The focus is on demonstrating the differences in the results they yield.
Section 2 describes the data used in this empirical work. Section 3 presents a brief overview of the empirical models and the estimation method. Section 4 reports the empirical results in reasonable detail, especially through a complete interpretation of the calculated elasticities. A sequential test of the homogeneity and symmetry restrictions for the AIDS and Rotterdam models are also presented in this section. Section 5 concludes this paper.