Files
Abstract
We examine the role of peer effects in childhood obesity outcomes by investigating whether
obesity rates among the highest graders in a public school has an effect on obesity rates among
younger grades. We use a panel dataset with obesity prevalence measured at the grade level.
Our data are from Arkansas public schools. Results provide evidence that changes in the obesity
prevalence at the highest grade are associated with changes in obesity prevalence at younger
grades. The magnitude of the peer effect depends on the type of school, and we find statistically
significant peer effects in both elementary and high schools but not in middle schools. These
effects are also larger in high schools than in elementary schools. We use falsification tests to
provide evidence that these peer effects are more than just a statistical correlation or an
association.