A Complex Systems Model of Industry Concentration and Broadband Infrastructure Investment

Abstract

As the United States faces challenges in developing and deploying next generation broadband technologies, it has reopened the debates regarding the best way to organize the telecommunication industry to achieve these goals. Yet this requires organizing network service providers, application providers, technology developers, and advertisers, each of whom are competing to further their own self interests which may not necessarily align with the interests of the consumers. Given the challenges of modeling these interactions, we have chosen a complex adaptive system approach to modeling the provision of broadband access and the corresponding applications, content, and consumption. The specific approach is a combination of multi‐agent simulation and genetic programming. Each type of agent in the marketplace co‐evolves its strategy to respond to the actions taken by other agents. Through this process, each agent searches for a good response to the other actions, improving their individual welfare within the constraints of the actions taken by other agents. The processes show that in general competition creates the best results for consumers. However, in cases where the cost of building the infrastructure is high, the consumers may be better served by a monopolist.