Social Interaction with Endogeneity

Abstract

We identify and estimate peer and contextual effects in social interactions models with endogenous covariates (e.g., self-selected treatments). Our method uses individual instruments for endogenous covariates, but does not require additional instruments for simultaneity in outcomes, which are often hard to find in models with contextual effects. Nor does it require exclusion restrictions that some observed covariates in the structural form have no contextual effects. The method can be applied to relax the Stable Unit Treatment Value Assumption (SUTVA) in program evaluation, allowing individual treatments to influence the outcomes of others through peer and contextual effects. We apply our method to estimate peer effects in Grade 3 math scores of elementary school students in the State of Tennessee. Using lagged class sizes and teacher qualification as instruments for Grade 2 scores, we find significant evidence for positive peer effects and path dependence on G2 scores.

Keywords: Social Interactions, Endogeneity, Reflection Problem, Control Functions, Peer Effects, Academic Achievement.

JEL Classification: C31, C36, I21