Measuring Discrimination in Education: Are Methodologies from Labor and Markets Useful?

This paper reviews the methodologies most frequently used by social scientists when measuring discrimination in housing and labor markets and assesses their potential usefulness for analyzing discrimination in education. We begin by briefly reviewing what economists typically mean by racial discrimination in housing and labor markets and note that the standard definition is somewhat more complicated in the case of education. We then review four methodologies—standard statistical methods (e.g., multiple regression), methods using more complete data (e.g., productivity vs. wages or applications for jobs/housing vs. acceptances), experimental/audit methods, and natural experiments based on actual policy changes. The strengths and weaknesses of the four methods when used on housing and labor markets are assessed. We then consider the extent to which each of the methods can be applied to education and the potential usefulness and limitations of trying to do so. We conclude that each of the four methods has some potential uses in analyzing discrimination in education.