UC Berkeley

In this study, I use multilevel modeling and multivariate matching to estimate the effect of suspension on college attendance. I build upon existing research that has suggested a negative relationship between school suspension and educational outcomes but that has mostly not been able to control for pre-existing differences in academic achievement between suspended and non-suspended students. I also extend my analysis to college attendance, which is currently an under-studied outcome related to high school suspension and thought to be important for many life outcomes. After controlling for baseline academic achievement, socioeconomic status, and other potential confounders, the college attendance rate of students who were suspended once or twice in the first semester of tenth grade was approximately 10 percentage points lower than that of non-suspended students. These results suggest that racial differences in educational attainment are in part due to the disproportionate discipline of African American students.