"Identification and Estimation of Demand for Bundles:
The Case of the RTE Cereal Industry"
(with Ao Wang)
Abstract: We study the identification and estimation of a mixed logit model of demand for bundles. We propose sufficient conditions for identification and for lack of it, and a constrained MLE that alleviates the challenge of dimensionality inherent in estimation. We use these methods to investigate the welfare implications of mixed bundling pricing in the ready-to-eat cereal industry in the USA. In this context, the constrained MLE reduces the numerical search from approximately 20000 to 150 parameters. The profit gains of mixed bundling pricing with respect to pure components pricing are sharply decreasing in the level of competition: while a monopolist would benefit from mixed bundling, the observed oligopoly would not—even ignoring potential increases in logistics costs. Given any market structure, mixed bundling leads to lower levels of consumer surplus than pure components.