Understanding Peer Effects: On the Nature, Estimation and Channels of Peer Effects

This paper provides evidence on ability peer effects in university education. Identification comes from the random assignment of students to sections. We find that students on average benefit from better-ability peers. Low-ability students, however, are harmed by high-ability peers. We introduce a placebo analysis that provides a simple test to quantify the estimation bias driven by the mechanisms described in Angrist (2013). In our setting, the bias is small and does not drive our results. Analyzing students’ course evaluations suggests that peer effects are driven by improved student interaction rather than adjustments in teachers’ behavior or students’ effort.JEL: I21, I24, J2