Instrumental variables are often used to identify peer effects. This paper shows that instrumenting the ‘peer average outcome’ with ‘peer average characteristics’ requires the researcher to include the instrument at the individual level as an explanatory variable.We highlight the bias that occurs when failing to do this.

Cross-Sectional Models; Spatial Models; Treatment Effect Models; Quantile Regressions (jel C31), Instrumental Variables (IV) Estimation (jel C36), Microeconomic Behavior: Underlying Principles (jel D01), Health Production: Nutrition, Mortality, Morbidity, Substance Abuse and Addiction, Disability, and Economic Behavior (jel I12), Education and Research Institutions: General (jel I20)
dx.doi.org/10.1111/obes.12299, hdl.handle.net/1765/131509
Oxford Bulletin of Economics and Statistics
Department of Applied Economics

von Hinke Kessler Scholder, S.M.L, Leckie, G, & Nicoletti, C. (2019). The use of instrumental variables in peer effects models. Oxford Bulletin of Economics and Statistics, 81(5), 305–9049. doi:10.1111/obes.12299