We hope you can join us this Wednesday, April 17, 2013 for the Applied Statistics Workshop. Ariel Pakes, Professor of Economics from the Department of Economics at Harvard University, will give a presentation entitled "Moment Inequalities for Semiparametric Multinomial Choice with Fixed Effects". A light lunch will be served at 12 pm and the talk will begin at 12.15.

"Moment Inequalities for Semiparametric Multinomial Choice with Fixed Effects"

Ariel Pakes

Department of Economics, Harvard University

CGIS K354 (1737 Cambridge St.)

Wednesday, April 17th, 2013 12.00 pm

Abstract:

We propose a new approach to identication for multinomial choice models with a group (or panel) structure. We take a standard random utility model of choice, where the utility for each choice is additively separable in a choice-specic fixed effect, a disturbance, and an index function of covariates and parameters. Observations in the same group are assumed to share the same fixed effects. Examples of this structure include; (i) Chamberlain's (1980) conditional likelihood estimator for panel data problems with choice specic fixed effects and i.i.d. logistic disturbances, and (ii) models of product demand where markets are the grouping device, the within group observations are consumers, and the choice-specic fixed effects represent product level unobservables.

We place no restriction on the variance-covariance of the disturbance vector across choices. The only restriction on the disturbances is a group homogeneity assumption. The main cost of the semiparametric flexibility in our model is that the conditional moment inequalities will, in general, only partially identify the index function parameters. The advantages are that it; (i) is non-parametric in the joint distribution of the disturbance vector across choices, (ii) allows for incidental choice specic effects whose cardinality can grow with sample size, and (iii) can be extended to allow for certain types of endogeneity.

Posted by Konstantin Kashin at April 15, 2013 10:10 AM