Estimating the Price Effects of Mergers by Exploiting a Natural Experiment in Cross-Section Data

Posted by D. Daniel Sokol

Michael J. Doane, Competition Economics LLC, Luke Froeb, Vanderbilt University - Strategy and Business Economics, Gregory J. Werden, U.S. Department of Justice - Antitrust Division, David M. Zimmer, Western Kentucky University - Department of Economics describe Estimating the Price Effects of Mergers by Exploiting a Natural Experiment in Cross-Section Data.

ABSTRACT: Past mergers in the car rental industry have resulted in markets with the same number of brands but different numbers of owners of those brands. Consequently, data at a single point in time exhibit between-market variation much like the variation over time associated with mergers. We construct an estimator that isolates this between-market variation, and the predicted price effects of mergers are roughly what simple theoretical models predict. For markets with few incumbent brands, our natural experiment predicts significantly higher price effects than a prior study using a structural model.