Seminar Vol. 182

Title: Aggregation Bias in Discrete Choice Models

Speaker: David Brownstone, University of California, Irvine

Time: October 10th, 2019      13:30-15:00

Venue: Conference Room 106B, IESR, Zhonghui Building (College of Economics)

About the speaker:

David Brownstone is a Professor of Economics at the School of Social Sciences, University of California, Irvine since 1999. After receiving his Ph.D. in Economics from University of California, Berkeley in 1980, he taught at Princeton University and later spent several years in Sweden (Stockholm School of Economics, and The Industrial Institute for Economic and Social Research). David Brownstone’s main research interests lie in the areas of panel study methodology, discrete choice models, bootstrap methodology, and housing demand. Over the years, he has published widely in such journals as Journal of Choice Modelling, Journal of Urban Economics, Journal of Econometrics, Transportation Research, and Marketing Letters.

Abstract:

This talk will start with an overview of missing data problems in applied econometrics, and I will argue that we do not pay enough attention to these problems. The main part of the talk examines the common practice of aggregating choice alternatives within discrete choice models. We carry out a Monte Carlo study based on realistic vehicle choice data for sample sizes ranging from 500 – 10,000 individuals. We consider methods for aggregation proposed by McFadden (1978) and Brownstone and Li (2014) as well as the more commonly used methods of choosing a representative disaggregate alternative or averaging the attributes across disaggregate alternatives. The results show that only the “broad choice” aggregation method proposed by Brownstone and Li provides unbiased parameter estimates and confidence bands. Finally, we apply these aggregation methods to study households’ choices of new 2008 model vehicles from the National Household Travel Survey (NHTS) where 1120 unique vehicles are aggregated into 235 make/model classes. Consistent with our Monte Carlo results we find large differences between the resulting estimates across different aggregation methods.