In this paper we develop a structural estimation approach of a large-scale combinatorial auction (CA); the Chilean auction for school meals in which the government procures half a billion dollars worth of meal services every year. We use bidding data to estimate the firms' cost structure and margins, and use this information to evaluate the performance of the auction in terms of its efficiency and payments to the bidders. To overcome the curse of dimensionality introduced by the huge number of possible package bids allowed, we propose a novel “restricted” model, where each mark-up is set as a function of a small number of package characteristics. Our estimates suggest that the bidders' cost synergies are economically significant in this application, and that the current CA mechanism achieves high allocative efficiency. We also compare the efficiency and revenues of the actual pricing rule (first-price) with alternative mechanisms such as VCG. These counterfactuals inform the auction design. Finally, the proposed structural estimation framework is quite general and can be applied to other large-scale CAs applications.
(joint work with Sang Won Kim and Marcelo Olivares)