This study uses nonparametric techniques to analyze the stability of demand for nineteen major food categories among various demographic groups in the United States. Households are divided into population groups by income, the head's age, and the spouse's education level. The data used are from the 1980-90 "Diary" portion of the Bureau of Labor Statistics's annual Consumer Expenditure Survey. The programming model developed by Sakong and Hayes, with the modifications suggested by Chalfant and Zhang, is used to test for and measure taste changes. Substantial differences in preference trends between population groups are found for many of the food commodities.
Bibliographical noteFunding Information:
Financial support for this research was provided by the Agricultural Experiment Station at the University of Minnesota. The authors wish to thank James Chalfant, who provided his computing language regarding the test for preference stability, and Alexander Meeraus, who provided the GAMS 386 version 2.25 software on a free trial basis. Chalfant and the two anonymous Journal referees provided very useful comments which improved the final version of this paper.
- Food consumption
- Nonparametric analysis
- Taste changes