Fanning in or out - How Does Schools Affect Inequality and Stratification

In this paper we use data from the Student/Teacher Achievement Ratio (STAR) project – a large longitudinal study of schools in Tennessee, USA to study how student stratification changes across school grades. We ask the question whether schools increase or decrease student ranking and student heterogeneity. If schools add to stratification at large we should expect first that schools exaggerate initial student heterogeneity. That is, variance in student achievement in kindergarten and first grade should be smaller than variance in student at later grades. Because student test scores are comparable across grades in STAR this data set allows at unique opportunity to test the hypothesis of student heterogeneity across grades. Secondly, we can test whether student ranking across grades are stable or change according to social background. We find that for achievement in English, variance in grades increase whereas for achievement in math, student heterogeneity decreases. For ranks, we find them to be stable across grades.