In this paper we analyze the impact of classroom peers' ability on individual student achievement with a unique longitudinal data set covering all Florida public school students in grades 3-10 over a five-year period. Unlike many data sets used to study peer effects in education, ours identifies each member of a student's classroom peer group in elementary, middle, and high school as well as the classroom teacher responsible for instruction. As a result, we can control for student fixed effects simultaneously with teacher fixed effects, thereby alleviating biases due to endogenous assignment of both peers and teachers, including some dynamic aspects of assignment. Our estimation strategy, which measures the influence on individual test scores of peers' fixed characteristics (including unobserved components), also alleviates potential bias due to measurement error in peer ability. Under linear-in-means specifications, estimated peer effects are small to nonexistent, but we find sizable and significant peer effects in nonlinear models. We find that peer effects depend on an individual student's own ability and on the relative ability level of peers, results suggesting that some degree of tracking by ability may raise aggregate achievement. Estimated peer effects tend to be smaller when teacher fixed effects are included than when they are omitted, a result that emphasizes the importance of controlling for teacher inputs. We also find that classroom peers exert a greater influence on individual achievement than the broader group of grade-level peers at the same school.
