A report suggests a relationship between school suspension rates and students’ perceptions of safety in Milwaukee schools. Specifically, it finds that following an agreement with the U.S. Department of Education Office for Civil Rights, lower suspension rates for African American students predicted higher numbers of students feeling unsafe in schools. The report asserts that “reduced suspension rates for African American students resulted in lower reports of safety.” Thus, the report erroneously communicates to readers that decreases in suspensions of African American students are causing decreased feelings of student safety, and that the Department of Education agreement harms, rather than helps, African American students. This review of the report finds numerous concerns, including unsupported claims, misleading interpretations, conflation of correlation with causation, and the use of racially criminalizing stereotypes of African American students. The report is therefore not useful to policymakers as a basis for policy decisions about school discipline.

A report from the Fordham Institute investigated the impact of a reform in the School District of Philadelphia that eliminated suspensions for certain low-level misbehaviors. The report considered whether the policy change was associated with any of the following: (a) district-wide out-of-school suspension rates, (b) academic and behavioral outcomes for students (looking separately at students who had a record of prior suspensions and those with no prior suspensions), and (c) racial disparities in suspensions. While the report concluded that the reform was a failure, the actual results were mixed, with the positive trends for students who were earlier suspended being much stronger in magnitude than evidence of negative outcomes for students who were not. A strength of the report is the use of advanced statistical methods and a longitudinal dataset to answer the questions of interest. However, the report is plagued by logical fallacies, overly simplified interpretations of findings, and inflammatory language. Moreover, the report uses misleading causal (“consequences”) language in the title and to describe study results, even though the study design is limited by unmeasured confounding factors and inappropriate comparison groups. Thus, while the analyses upon which the report is based have some technical merits, the narrative seems more of an attempt to advance a political agenda opposed to the reform studied than to improve understanding of complex policy issues.