Tom Hoffer, Shobha Shagle

Schools that enroll disproportionately high percentages of pupils from low-income families are widely believed to have negative consequences for student performance. Prior research has investigated the relationship of school poverty and outcomes in numerous ways, but the basic proposition is that school composition affects student learning, such that otherwise similar students realize different levels of achievement in schools with different proportions of low-income. This paper updates and extends the research on compositional effects in several respects. First, we extend the research to the early elementary level of schooling. Second, the data needed to assess the relative importance of individual and school factors have not been available and the mechanisms that mediate the school-level effect independent of student background factors are thus not clear. This paper draws upon nationally representative data on kindergarten pupils and the schools they attend to estimate both the overall impact of school poverty on mathematics achievement and its impact on a variety of other school and schooling-experience variables that may in turn affect student learning. Finally, this paper analyzes the effects of high poverty schools on several schooling variables hypothesized to affect student achievement.