Linear Regression Estimation of Discrete Choice Models with Nonparametric Distributions of Random Coefficients

摘要

In a discrete choice demand model, consumer i chooses product j out of the set J if u_(i,j) > u_(i,k) any k = 1,...,J, k ≠ j. A standard setup has a vector of D characteristics x_(i,j) for product j, and utility is parametrized as u_(i,j) = x'_(i,j)β + ε_(i,j), where β is a vector of parameters to estimate that reflect the marginal utility of the product characteristics, and ε_(i,j) is a product-specific error term. Typically ε_(i,j) is assumed to have a logit or normal marginal distribution.