Seminarios académicos y conferencias
"Recovering Social Networks from Panel Data"
9 Mayo 2018 - 15:30 hrs.
Sala de postgrados, Facultad de Ciencias Económicas y Administrativas UC
Abstract: Empirical research on social and economic networks has been constrained by the limited availability of data regarding such networks. This paper develops a method that does not rely on network data to estimate network effects. The proposed method also estimates the probability that pairs of individuals formconnections, which may depend on exogenous factors such as common gender. The method may incorporate imperfect network data, such as with self-reported data, with the dual purpose of refining the estimates and testing whether the reported connections positively affect the probability that a link is formed. To achieve those goals, I derive a maximum likelihood estimator for network effects that is not conditional on network observation. Networks are treated as a source of unobserved heterogeneity and eliminated based on data collected from observing many groups. This is accomplished with recourse to a spatial econometric model with unobserved and stochastic networks. I then apply the model to estimate network effects in the context of a program evaluation. I demonstrate theoretically and empirically that including network effects has important implications for policy assessments.
Keywords:social networks, spillovers, spatial econometrics.
JEL Codes:C21, C49, O12, D85.