Two-sided networks: Implications for Internet intermediaries and technology platforms

Many Internet intermediaries and technology platforms are two-sided networks, i.e. they provide the infrastructure to bring two sides (e.g. buyers and sellers) together. This dissertation develops a two-sided networks framework to analyze (a) the optimal pricing strategy and the relationship between the ownership and the design of Internet intermediaries, (b) the competition between intermediated and peer-to-peer electronic commerce infrastructures, and (c) the structure of competition between a proprietary and an open source technology platform.; The first chapter develops a model that characterizes the value created by two-sided networks, and the allocation of that value across the two sides. When the asymmetry of the network effects is large, then the side with the low network effect participates for free. The analysis departs from existing networks literature by endogenizing the network effects and focusing on the network design resulting from investments in network effects. It shows that typically the design of a two-sided network is characterized by maximally asymmetric allocation of surplus independent of the ownership regime.; The second chapter characterizes the future structure of e-commerce infrastructure. Will the intermediaries have a dominant role, or firms will link directly to their partners, or these two types of infrastructure will co-exist? It shows that the asymmetry of the network effects between buyers and sellers, along with the costs of infrastructure, are crucial determinants of the infrastructure equilibrium. It predicts different scenarios for the evolution of an industry's e-commerce infrastructure and the profitability of intermediaries.; The third chapter develops a framework to characterize the optimal two-sided pricing strategy of a platform firm. It compares industry structures based on a proprietary platform with those based on an open-source platform and analyzes the structure of competition and industry implications. It finds that, when the platform is proprietary, the equilibrium prices can be below marginal cost. The social welfare in the software industry may be higher when the platform is open source. The open-source platform industry can also be more profitable than the proprietary platform industry. Under competition, the open source system is typically dominant in terms of market share.