Working Paper No. 05-27:
Two Tales of Bundling: Implications for the Application of Antitrust Law to Bundled Discounts
Date Posted: 2005
Abstract (below) | Full text (most recent) on SSRN
The economic literature on bundling has made many theoretical advances. However, several omissions reveal themselves. The advances have largely been on the theoretical side. These models contain restrictive assumptions regarding the existence of monopoly in some markets, and the nature of rivalry in others. The models generally ignore obvious and ubiquitous reasons firms may use bundled discounts. These models have not been subject to robustness checks, nor have their assumptions been tested empirically. As a result, the literature that shows the possibility of anticompetitive harm does not provide a reliable way to gauge whether the potential for harm would outweigh any demonstrable benefits from the practice.
As a result of the underdeveloped nature of the literature, simple rules that result in extreme tradeoffs between type I and type II errors may dominate more complex tests that attempt to differentiate procompetitive from anticompetitive behavior. Such complex tests may work well within the confines of a theoretical model, but not when applied to firms in actual antitrust cases. Improving the reliability of more complex tests for anticompetitive behavior will require economists to expand their understanding of both the anticompetitive and procompetitive reasons firms engage in bundling. This will entail studying the reasons bundling is adopted by firms without market power, relaxing the assumption of monopoly in theoretical models, and generating testable hypothesis and the data to test them.