Local Externalities in Real-World Networks: Does Topology Matter?

While the widespread realization of the importance of the embededdness of economic activity in social settings has been fundamental to sociologists for some time, it was largely ignored by economists until twenty years ago. Since then, the study of social interactions in economics-decision making has become an important area of research. The motivation lies in the common observation in the social sciences of large differences in outcomes in the absence of corresponding differences in fundamentals. Models of social interactions present frameworks where an agent’s behavior depends, among other things, on choices of other agents in a reference group. This property of models of social interactions has resulted in a large literature that uses these models to explain economic and social behavior in a variety of contexts. This literature, however, assumes very special interaction structures, implying that agents care only about some average behavior in the chosen (typically aggregate) group. Such a structure presents huge difficulties for empirical tests. The problems are clear since the work of Mansky (1995), and virtually impossible to overcome in absence of a clear external source of variation or experimental data. This is why the actual existence, extent and functioning of social interaction effects is still a matter of debate in many policy context. Data on social networks provide a natural solution to the Mansky’s reflection problem in the identification of social interaction effects. In this paper we relate a particular form of social interactions, i.e. peer group influences, to the structure of the network from which they originate. Our results suggests that the individual heterogeneity stemming from network topology, i.e. the individual position in the network of her/his social contacts, might be an important actor in economics-decision making