Racial Disparities in Education: Bridging Theory and Method

Sarah K. Bruch, University of Wisconsin at Madison
Mara Loveman, University of Wisconsin at Madison

This paper integrates constructivist insights into how we deal with population data, investigating how race is a multidimensional not a unidimensional construct. Race as a social relationship is a product of a constellation of things: self-perception, other ascription, family ancestry, physical appearance, and social context. In this paper, we juxtapose four measures to evaluate how different aspects of race contribute to the apparent “effect of race” on educational outcomes. We articulate the theoretical implications of selecting one measure over another, identifying plausible mechanisms and evaluating alternative measures against existing theories of social processes believed to contribute to educational outcomes. We predict achievement using the four alternative measures, modeling them individually and simultaneously. In preliminary work we find that these measures capture different dimensions of race and operate according to different mechanisms. We find a strong linear association between ascribed skin color and grades in models including both self-identification and skin color measures.