Non-cognitive peer effects in secondary education

Abstract

The peer effects literature has established that peers impact each other in the classroom through academic achievement and cognitive ability, but has not explored many alternative channels. This paper examines how the non-cognitive traits of peers in the classroom impact an individual’s learning outcomes. I estimate a linear-in-means model and alternative models of peer effects with additional peer effects terms accounting for “Big Five” personality traits. Controlling for selection into schools, cognitive and non-cognitive ability, and family background, there is a significant, positive relationship between average peer conscientiousness and individual academic performance of the order of a 0.15 standard deviation increase in math scores and a 0.12 standard deviation increase in language scores. This is the first evidence relating non-cognitive traits to peer effects in secondary school and lends support for programs in schools targeting the development of non-cognitive skills.