Richard Schmalensee, May 17, 2011
Jeffrey Rohlfs’ pioneering 1974 study of demand in the presence of network externalities, which make each actor’s demand for some good or service depend in part on whether others purchase it, laid the foundation for a huge academic literature that has had a major impact on antitrust policy. The government’s case in U.S. v. Microsoft, for instance, relied heavily on network externality arguments.
In most of the post-Rohlfs network-effects literature, buyers are modeled as making long-term product or technology choices because those choices either involve the purchase of significant durable goods or create switching costs. Examples include the choices between VHS and Betamax VCRs, or between Apple and Wintel computers, or the choice to purchase an early fax machine.
In contrast, Rohlfs presents a model that seems better suited to analysis of new Internet-based businesses that rely on network effects, like Facebook and YouTube. These businesses provide services rather than durable goods, and their customers are not required to make long-term commitments. Switching costs are at most moderate, and customers can often participate in multiple competing networks at the same time. (In the terminology of the recent, related literature on two-sided markets, they can “multi-home.”2) Thus I think the Rohlfs paper deserves to be read carefully on its own, apart from the literature it helped to launch.