Merger simulations are commonly used to simulate the effects of potential mergers. Despite the large resources devoted to merger review, little evidence exists on the accuracy of these methods. This paper uses the acquisition of Tambrands by Proctor and Gamble to provide evidence on the efficacy of merger simulation. Two simple demand systems are estimated under several identification assumptions and combined with a static model of price competition. Simulations predict small price effects of about 1 percent for the merging firms' brands, while direct estimates indicate the merger raised prices by 5–8 percent.

Abstract This paper evaluates the efficacy of a structural model of oligopoly used for merger review. Using premerger data, we estimate several demand systems and use a static Bertrand model to simulate the price effects of two mergers. Using pre- and postmerger data, we directly estimate the price effects. The direct estimates imply that one merger resulted in moderate price increases, while the second left prices essentially unchanged. While some simulations are similar to the directly estimated price effects, overall simulations overstate the price effects in one case and understate them in the other. Explanations for the discrepancies are explored.

We document abrupt increases in retail beer prices just after the consummation of the MillerCoors joint venture, both for MillerCoors and its major competitor, Anheuser‐Busch. Within the context of a differentiated‐products pricing model, we test and reject the hypothesis that the price increases can be explained by movement from one Nash–Bertrand equilibrium to another. Counterfactual simulations imply that prices after the joint venture are 6%–8% higher than they would have been with Nash–Bertrand competition, and that markups are 17%–18% higher. We relate the results to documentary evidence that the joint venture may have facilitated price coordination.

We study a repeated game of price leadership in which a firm proposes supermarkups over Bertrand prices to a coalition of rivals. Supermarkups and marginal costs are recoverable from data on prices and quantities using the model’s structure. In an application to the beer industry, we find that price leadership increases profit relative to Bertrand competition by 17 percent in fiscal years 2006 and 2007, and by 22 percent in 2010 and 2011, with the change mostly due to consolidation. We simulate two mergers, which relax binding incentive compatibility constraints and increase supermarkups. These coordinated effects arise even with efficiencies that offset price increases under Bertrand competition. (JEL G34, K21, L13, L14, L41, L66)