ISER Discussion Paper, Institute of Social and Economic Research, Osaka University 567
The widespread use of illicit substances by American teenagers has attracted the interest of both the public and academic researchers. Among the various factors that people believe influence youth substance use, peer effects are identified as a critical determinant; substance use is considered a highly social behavior. Identifying peer effects, however, is not an easy task. Common teenage behaviors can be due to similar unobserved characteristics of the group members or peer effects. Moreover, it is difficult to pinpoint whether a subject is affecting the group members' behaviors or vice versa. In an attempt to overcome these difficulties, I estimate peer effects on substance usage among American teenagers using perceived peer behavior in National Longitudinal Survey Youth 97. School and household fixed effect estimation are also employed to ensure the robustness of the results. The data indicate robust peer effects. Moreover, the results do not change substantially in school and household fixed effect estimations.