Estimating the Pure Characteristics Demand Model: A Computational Note

17 Pages Posted: 26 Feb 2009

See all articles by Minjae Song

Minjae Song

Bates White Economic Consulting

Date Written: November 26, 2008

Abstract

This paper provides details of the computational algorithm of Berry and Pakes (2005) for estimating the pure characteristics demand model. The main challenge in estimating the model is to find "the mean product quality" that equates the model predicted market shares to real market shares. The algorithm combines three methods: a contraction mapping, a homotopy method with an element-by-element inversion, and Newton-Raphson method. After explaining the algorithm, I evaluate its performance using four simulated data sets with up to 20 products per market.

Keywords: Demand estimation, Discrete Choice Model

Suggested Citation

Song, Minjae, Estimating the Pure Characteristics Demand Model: A Computational Note (November 26, 2008). Simon School Working Paper No. FR 08-28, Available at SSRN: https://ssrn.com/abstract=1349991 or http://dx.doi.org/10.2139/ssrn.1349991

Minjae Song (Contact Author)

Bates White Economic Consulting ( email )

1300 I Street NW
Washington, DC DC 20005
United States
2027471404 (Phone)

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
120
Abstract Views
574
rank
254,164
PlumX Metrics