The effects of being racially, ethnically, & socioeconomically different from peers

Classroom diversity in observable characteristics has been shown in prior literature to have different effects on the educational and behavioral outcomes of students with observable characteristics that differ from their peers. Given that similarity in observable characteristics plays a critical role in social interactions (homophily), students may experience social costs to being racially, ethnically, and socioeconomically very different from their peers. We test these implications using reduced form regressions on Add Health data, where identification comes from exogenous variation in the demographic composition of grades within a school. Empirical findings suggest that students who are more socioeconomically different from same-gender peers are less happy, feel less socially accepted, and do worse academically, particularly in low diversity environments. However, racially different students put forth more effort by participating in extracurricular activities when diversity is low. High diversity environments generate more nonconformity (e.g., altercations with other students) among racially different students, although academic outcomes improve. The results highlight the complexity and importance of considering group composition when weighing policy initiatives that alter the distribution of student types across high schools.