We examine the contribution of differences in school environments to the gap in education outcomes between Aboriginal and non-Aboriginal students. We find both substantial school-level segregation of Aboriginal and non-Aboriginal students and a substantial gap in test scores. Conventional achievement gap decompositions attribute roughly half of the grade 7 test score gap to between-school differences and half to withinschool differences. The segregation of Aboriginal students suggests that peer effects might explain some of these between-school achievement differences. However, we find that peer effects associated with a greater proportion of Aboriginal peers, if anything, improve value-added exam outcomes of Aboriginal students. JEL classification: I21 Classement, pairs et résultats desétudiants aborigènes en Colombie Britannique. On examine la contribution des différences dans les environnements scolaires pour expliquer l'écart entre les résultatséducationnels desétudiants aborigènes et non-aborigènes. On découvreà la fois un degré de ségrégation substantiel entreétudiants aborigènes et nonaborigènes, et unécart substantiel dans les résultats des tests. La décomposition conventionnelle desécarts de résultats pour lesélèves de 7 e année attribue en gros la moitié desécartsà des différences entreécoles, et l'autre moitiéà des différencesà l'intérieur deś ecoles. La ségrégation desélèves aborigènes suggère que l'effet des pairs pourrait expliquer une portion des différences entreécoles. Cependant, on découvre que les effets de pairs associésà la présence d'une plus grande portion de pairs, s'ils ont un impact quelconque, améliorent les résultats desétudiants aborigènes dans les examens qui mesure les progrès.

No abstract

We use data on students in grades 4–7 in the Canadian province of British Columbia to investigate the effect of having disabled peers on value-added exam outcomes. Longitudinal data for multiple cohorts of students are used together with school-by-grade-level fixed effects to account for endogenous selection into schools. Our estimates suggest that same-grade peers with learning and behavioral disabilities have an adverse effect on the test score gains of nondisabled students in British Columbia. However, these effects are statistically insignificant and are sufficiently small that they are unlikely to raise concerns about the placement of this group of disabled students. The effect of peers with other disabilities is also small and statistically insignificant but varies in sign.

No abstract

No abstract