Heterogeneity and the Dynamics of Technology Adoption

This paper analyzes the role of heterogeneity and forward-looking expectations in the diffusion of network technologies. Using a detailed dataset on the adoption of a new videoconferencing technology within a firm, we estimate a structural model of technology adoption and communications choice. We allow for heterogeneity in network benefits and adoption costs across agents. We find that ignoring heterogeneity in the interplay between adoption costs and network effects will underpredict the size of the steady-state network size by almost 50 percent. We develop a new 'simulated sequence estimator' to measure the extent to which agents seek diversity in their calling behavior, and characterize the patterns of communication as a function of geography, job function, and rank within the firm. We find that agents have significant welfare gains from having access to a diverse network, and that a policy of strategically targeting the right subtype for initial adoption can lead to a faster-growing and larger network than a policy of uncoordinated or diffuse adoption