Merger Simulation in Second-Score Auctions: A Nested Logit Model

Abstract In several recent merger investigations involving companies competing through bidding, economists have used second-score auction models. The standard application of the model makes a “flat logit” assumption, which has the strong implication that business diversion from one bidder to the other rival bidders is proportional to the rivals’ shares (or win probabilities). We show how to simulate merger effects under a more general “nested logit” assumption, which allows for diversion to be more than proportional to shares for bidders in the same nest, and less than proportional to shares for bidders in different nests. We show how to apply the framework to the 2012 acquisition of Power Reviews by Bazaarvoice, which was challenged by the U.S Department of Justice.