- Current approaches for estimating discrete choice models of consumer demand are computationally challenging, making them difficult to use for policy analysis.
- This paper presents a simple method for estimating these models by using increasingly common microdata, such as household demographics.
- The method is applied to simulate the effect of tightening fuel economy standards for cars and light trucks on new vehicle sales.
- The simulation results suggest that a 1-percent increase in the stringency of the standards leads to a modest reduction in new vehicle sales.
Discrete choice models remain a key tool for analyzing markets of differentiated products. Recent developments of these models incorporate preference heterogeneity to reflect realistic substitution patterns. Yet the identi fication and estimation of preference heterogeneity remains computationally challenging. I derive a new method for identifying and estimating the parameters defi ning heterogeneity in discrete choice models of product differentiation. I develop a multistage estimation method in which the heterogeneity preference parameters are estimated in initial stages and the average preference parameters are estimated in a final stage. A key advantage of the method is that each stage is straightforward to estimate: the initial stages are estimated with a simple closed-form expression or with a linear regression, and the final stage is estimated with instrumental variables. I apply the method to estimate parameters of new vehicle demand and to simulate the effects of new vehicle fuel economy standards on total new vehicle sales. The estimation results show a weak substitution between new and used vehicles. Therefore, the simulation results suggest that a marginal tightening of the standards has a modest impact on total new vehicle sales.