Estimating Social Network Models with Link Misclassification

Abstract: We propose an adjusted 2SLS estimator for social network models when the network links reported in samples are subject to two-sided misclassification errors (due, e.g., to recall errors by survey respondents, or lapses in data input). Misclassified links make all covariates endogenous, and add a new source of correlation between the structural errors and peer outcomes (in addition to simultaneity), thus invalidating conventional estimators used in the literature. We resolve these issues by adjusting endogenous peer outcomes with estimates of the misclassification rates and constructing new instruments that exploit properties of the noisy network measures. Simulation results confirm our adjusted 2SLS estimator corrects the bias from a naive, unadjusted 2SLS estimator which ignores misclassification and uses conventional instruments. We apply our method to study peer effects in household decisions to participate in a microfinance program in Indian villages.