New Econometric Models in Consumer Demand

• Activity Number: 414
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Kai-Sheng Song*
• Companies: University of North Texas
• Keywords: Econometric Models ; Consumer Demand ; Random Utility Maximization

Abstract

Consumer demand theory has been one of the most thoroughly examined, tested, and applied areas in microeconomics. The existing approaches to generate a system of demand equations include the direct random utility maximization subject to the consumer's budget constraint, the indirect utility function through Roy's identity, and the expenditure function via Shephard's lemma. However, there is no standard way of introducing a random component into the utility function. Depending on the particular utility model specified, the random disturbances enter the direct or indirect utility function either additively or multiplicatively in an ad hoc manner implying strong restrictions on the demand equations. In this research, we present a new approach of specifying econometric models for consumer demand and discuss its statistical implications. Our general statistical method is applicable to estimation of demand systems with an arbitrarily large number of goods. Numerical examples are presented to demonstrate the performance of the proposed method.