We are grateful for the opportunity to publish this article in a volume in honor of Daniel McFadden. The article illustrates just one way in which Dan's path-breaking work on the economics and econometrics of choice theory has impacted Industrial Organization. We would also like to thank two referees and Charles Manski for helpful comments. Please address correspondence to Steven Berry, Department of Economics, Yale and NBER. E-mail: email@example.com.
THE PURE CHARACTERISTICS DEMAND MODEL*
Article first published online: 11 DEC 2007
International Economic Review
Volume 48, Issue 4, pages 1193–1225, November 2007
How to Cite
Berry, S. and Pakes, A. (2007), THE PURE CHARACTERISTICS DEMAND MODEL. International Economic Review, 48: 1193–1225. doi: 10.1111/j.1468-2354.2007.00459.x
Manuscript received October 2005; revised June 2007.
- Issue published online: 11 DEC 2007
- Article first published online: 11 DEC 2007
In this article, we consider a class of discrete choice models in which consumers care about a finite set of product characteristics. These models have been used extensively in the theoretical literature on product differentiation and the goal of this article is to translate them into a form that is useful for empirical work. Most recent econometric applications of discrete choice models implicitly let the dimension of the characteristic space increase with the number of products (they have “tastes for products”). The two models have different theoretical properties, and these, in turn, can have quite pronounced implications for both substitution patterns and for the welfare impacts of changes in the number and characteristics of the goods marketed. After developing those properties, we provide alternative algorithms for estimating the parameters of the pure characteristic model and compare their properties to those of the algorithm for estimating the model with tastes for products. We conclude with a series of Monte Carlo results. These are designed to illustrate: (i) the computational properties of the alternative algorithms for computing the pure characteristic model, and (ii) the differences in the implications of the pure characteristic model from the models with tastes for products.