Abstract
Research on school discipline disparities often examines the relationship between student behavior, student/teacher demographics, bias, policies, and practices on the likelihood of receiving a discipline outcome. However, often missing from the discourse is the role of repeat infractions and the time of the year that the discipline outcome was given. In this study, we investigate the relationship between race, severity of past disciplinary infractions, and the probability of receiving an exclusionary disciplinary outcome for infractions at the end of the school year. Utilizing data from the School-Wide Information System (SWIS) for the 2017–2018 academic year in 80 Georgian elementary schools, we conducted a binary logistic regression analysis with cluster robust standard errors to accommodate the clustering of students within schools. The final sample comprised 4075 students after applying exclusion criteria. The interaction between the severity of the most recent infraction and race was statistically significant (χ2(4) = 14.93, p = 0.0049). The model’s AUC was 0.82, indicating a strong predictive capacity. The analysis revealed that Black and multi-racial students had a higher probability of receiving exclusionary outcomes compared to White students, even when controlling for previous infractions and the severity of the most recent infraction. The results suggest significant racial disparities in the administration of school discipline. These findings suggest the need for systemic investigation of, and changes in, disciplinary practices to support equity.