Disparities By Race and Disability across Discipline Outcomes: Multilevel Correlates of in- and out-of-School Suspension

Background:

A growing body of research considers the overrepresentation of youth of color and students with disabilities among those who receive out-of-school suspensions (OSS), yet patterns related to in-school suspensions (ISS) have gone relatively unexamined. Many consider ISS to less severe than OSS, but it still involves removing students from their classrooms and lost instructional time. Implementing ISS might help some schools reduce their reliance on OSS, but if it perpetuates the same educational and civil rights concerns as OSS, it will not address the goals of recent discipline reforms. To further our understanding of disparities by race and disability in school discipline, this study compares the correlates of ISS and OSS.

Methods:

We constructed our dataset by matching administrative data on district-wide student and school characteristics with discipline records. The sample included more than 100,000 K-12 students enrolled over 200 schools. The sample was predominantly youth of color (75%) and 10% had one or more disabilities. District-wide discipline rates were 3.3% for ISS and 2.7% for OSS.

Independent variables were student race and disability. Dependent variables were dichotomous indicators of whether or not a student received one or more ISS or OSS. At the student-level, all analysis controlled for gender, gifted and talented program eligibility, identification as homeless, English language learner status, and grade-level. At the school-level, covariates included grade configuration (high, middle, elementary, or alternative e.g. K-8), school size, governance model (charter or traditional), the proportion of student body that was eligible for free and reduced lunch, and the degree of racial segregation.

We used Stata13 to analyze a multilevel logistic regression model estimating the likelihood of students of color and youth with disabilities experiencing one or more ISS or OSS compared to their more privileged peers.

Results:

Results indicate disparities in both OSS and ISS by race and disability. Black (OR 2.8, p < .001), Latinx (1.6, p<.001), Multiracial (OR 1.8, p < .05) and youth with disabilities (OR 1.5, p < .001) were significantly more likely to experience an ISS than their White or mainstreamed peers. Similarly, Black (2.7, p<.001), Latinx (1.5, p<.001), Multiracial (1.8, p <.001), and youth in with disabilities (OR 1.4, p<.001) were also significantly more likely to experience an OSS.

Conclusions:

ISS accomplishes the goal of keeping students in school, but this practice still limits students’ access to high-quality learning opportunities. Like most research on exclusionary discipline, it is difficult to prove the causal impact of these practices. However, correlational and longitudinal research suggests that ISS is negatively associated with standardized test scores, GPA, school persistence, and student reports of school connectedness. In light of those findings, our results suggest that ISS may raise many of the same concerns as OSS, as disparities across the two practices were remarkably similar. Future research in school social work would be strengthened by including a range of discipline outcomes create a more complete picture of exclusionary ad punitive practices in schools.