Removing Police From Schools: Implications for Racial Equity in Exclusionary Discipline

• In Event: Linkages Between Education and the Criminal Legal System

Abstract

Purpose
This study examined how (a) the removal of school-based law enforcement (SBLE) shapes Black-white racial inequality in exclusionary discipline, and (b) school racial composition conditions this relationship. Research has consistently linked schools’ use of SBLE to higher rates of exclusionary discipline—particularly among Black students—but little is known about what happens when SBLE is removed.

Theoretical framework
Theoretical explanations for racial inequality in school discipline have rested on three hypotheses: differential behavior, differential treatment, and differential sorting. Among these, differential treatment and differential sorting have received the most empirical support. Differential treatment refers to instances where students who exhibit similar behaviors are treated differently based on how they are racialized. Differential sorting refers to how students of different races are sorted into schools that vary in their use of disciplinary actions. SBLE has the potential to contribute to both within-school differential treatment and between-school differential sorting.

Method
This study uses a two-group, two-wave difference-in-differences design. The treatment group consists of schools that removed SBLE between waves. The two comparison groups consist of schools that either (a) had SBLE at neither wave, or (b) had SBLE at both waves.

Data
The data include two-wave panel data from the 2013-14 and 2017-18 waves of the Civil Rights Data Collection, a census of U.S. public schools (N = 81,933 after data cleaning).

Measures
SBLE presence was measured as the presence of sworn law enforcement officers (0 = absent, 1 = present). Exclusionary discipline was measured at each wave as the rate per 100 students of a given race (i.e., Black or white) who were administered (a) at least one out-of-school suspension, or (b) an expulsion. School racial composition was measured as the proportion of white students enrolled in the school.

Analysis
We used a series of ordinary least squares regression models that included time-by-treatment interactions to estimate within-school changes in exclusionary discipline. School racial composition was used as a moderator. Additionally, models included propensity score weights to reduce baseline differences between treatment and comparison schools.

Results
Contrary to expectations, removing SBLE (compared to having SBLE at neither wave) was associated with substantial increases in within-school Black-white racial disparities in out-of-school suspension. This relationship was contingent on school racial composition such that removing SBLE increased racial disparities especially strongly in schools with larger proportions of white students. There were also reductions in expulsion rates for both Black and white students, although the decreased rate among white students was smaller in schools with larger proportions of white students.

Significance
Although removing SBLE has been a commonly proposed strategy for reducing racial disparities in exclusionary discipline, this study did not find evidence that removing SBLE had such an effect. It is important, therefore, to understand what other changes accompany schools when SBLE is removed, and to understand how schools’ disciplinary practices shift in response. The findings point to the utility of both the differential treatment and differential sorting hypotheses, and further interventions for advancing equity should remain targeted at changing systems and structures.

Authors

• Benjamin W. Fisher, University of Wisconsin - Madison

• Deanna N. Devlin, Farmingdale State College - SUNY