Do Magnet Schools Reduce Black-White Disparities in Middle-School Suspensions and Achievement?

• In Event: How Can Pre-K Contribute to Reducing the Black-White Achievement Gap?

Abstract

Research has consistently revealed disproportionate rates of suspension across racial and ethnic groups (Krezmien et al, 2006). Black-White differences, specifically, in school suspensions tend to predict racial differences in educational achievement (Morris and Perrin, 2016) and students in schools with higher suspension rates are more likely to have poorer educational outcomes (Bacher-Hicks et al., 2019). Evidence on the role of pre-K education in reducing racial disparities in suspension rates is, however, lacking. With earlier findings indicating that children who attended pre-K settings in the Tulsa School District were more likely to attend a magnet middle school as our departure point (see Figure 1), we examine (1) whether children in magnet schools have lower suspension rates and higher achievement than those in traditional schools, and (2) whether magnet schools have the potential to reduce the suspension and achievement gap, and what the pertinent school features may be.
We use a longitudinal sample of approximately 1,600 students followed from kindergarten through middle school in the Tulsa Public Schools (TPS), initially recruited as part of a study to examine the short and longer-term effects of pre-K education. Our outcomes are middle school in-school and out-of-school suspension rates of students, number of days of the suspensions, and math and reading state test scores. We pair these student-level data (i.e. race of the student, lunch status, attended pre-k) with school-level data (i.e. % of minority teachers, student-teacher ratios, staff attendance rates) to better understand how student characteristics and features of the middle school environment predict suspension rates. We use propensity score matching as our causal identification strategy and use a robust set of covariates including race, gender, age, lunch status, parental education, and early academic achievement measured by Woodcock Johnson test scores.
In exploring suspension rates, we find a similar Black-white suspension gap in TPS as to what has been found nationally, with Black students suspended at much higher rates. We also find that magnet schools have significantly lower suspension rates than non-magnet schools. More importantly, the Black-white suspension gap is smaller in magnet schools than it is in traditional schools. Black students overall are 17.7% more likely to receive in-school suspension compared to white students, but Black students in magnet schools are 13.4% less likely to receive in school suspensions compared to Black students in traditional schools. This reduces the Black-white suspension gap from 17.7% in traditional schools to 2.8% in magnet schools. Similar trends are found when focusing on out-of-school suspensions. With regard to achievement outcomes, both Black and white students in the magnet programs out-performed students in traditional settings, even after controlling for prior academic achievement. However, white students exhibited larger gains from magnet school enrollment on math and reading tests (see Figure 2) and, as a result, we do not find evidence that magnet school enrollment reduces the Black-white achievement gap. Pre-K enrollment may start trajectories of school enrollment that, via Magnet schools, offer one potential avenue for reducing the Black-white suspension gap and improving academic outcomes for all children.

Authors

• Karin Kitchens, Virginia Polytechnic Institute and State University
  Presenting Author

• NaLette Brodnax, Georgetown University
  Non-Presenting Author