A report from the Department of Education Reform at the University of Arkansas examines the association between out-of-school suspensions and student test scores. Using dynamic and multilevel regression modeling of six years of school records from all K-12 public schools in Arkansas, the paper purports to estimate a causal relationship between exclusionary discipline and academic performance. It concludes, in contrast to prior work, that the number of days of suspension a student receives has a positive relationship to math and language arts test scores. However, the outcomes (end-of-year test scores) are measured at least a full year after the hypothesized causal factor (days suspended), and there is no control for days suspended in the year the test was taken. Consequently, the results do not reflect missed instructional time for the tested material or other associated mechanisms through which suspension might adversely affect test scores. The analyses also control for a large number of infractions that are strongly related to days of out-of-school suspension, which may produce biased results. The findings also have weak face validity, in light of the weight of evidence suggesting that exclusionary discipline and school absences have adverse effects on test scores, GPA, grade retention, and dropping out. For these reasons, the reviewers caution that this paper is not useful for policymakers.

A response from the authors is below the review.

The reviewers of “Understanding A Vicious Cycle: Do Out-Of-School Suspensions Impact Student Test Scores?” respond to criticisms of their review in two parts, Part I by Daniel Losen and Part II by Brea Louise Perry.