SSRI/DuPRI Seminar Series~Linear Social Networks Models

This paper provides a systematic analysis of identification in linear social networks models. This is both a theoretical and an econometric exercise in that it links identification analysis to a rigorously delineated model of interdependent decisions. We develop a Bayes-Nash equilibrium analysis for interdependent decisions under incomplete information in networks that produces linear strategy profiles of the type conventionally used in empirical work and which nests linear social interactions models as a special case. We consider identification of both contextual and endogenous social effects under alternative assumptions on the a priori information on network structure available to an analyst and contrast the informational content of individual-level and aggregated data.