Estimation of Theoretically Plausible Demand Functions from U.S. Consumer Expenditure Survey Data

Abstract

The purpose of this paper is the application of four popular theoretically plausible consumer demand systems to a common cross-sectional data set that combines expenditure data from the quarterly BLS consumer expenditure surveys with price data that are collected quarterly in cost-of-living surveys conducted by ACCRA. Six broad categories of expenditure that exhaust total expenditure are analyzed: food consumed at home, housing, utilities, transportation, health care, and miscellaneous. The four demand systems investigated are the Almost-Ideal-Demand-System, the Linear Expenditure System, and the Indirect and Direct Addilog models. Despite absolute differences in magnitudes that in some instances are rather large, there is substantial agreement in the rank-orderings of elasticities. In general, the largest elasticities (for both own-price and total expenditure) are for transportation, miscellaneous, and housing expenditures, while the smallest elasticities (again for both own-price and total expenditure) are for food and utility expenditures. Engel's Law for food is confirmed in all instances.