Association for Public Policy Analysis & Management

Based on recent research (see Perry & Morris, 2014), the rationale that the use of exclusionary discipline policies in schools mitigate disruptions and by doing so increases the learning of non-offending students appears to be unfounded.

However, many policy relevant questions have been left unanswered by previous research. For example, since much of the previous research uses localized samples we do not know under what circumstances the effects of suspensions might apply more broadly to schools throughout the nation. Also, given the focus of these studies on out-of-school suspension (OSS), existing research cannot inform the consequences of the current shift from the use of OSS to the use of in-school suspensions (ISS). Additionally, a focus on outcomes at a single point in time within existing work has not revealed the duration of educational consequences that extend from exposure to these social control extremes. We also do not know whether these effects persist when methodologies are used that limit selection bias by addressing the issue of non-random selection of students into schools.

In extending the previous literature, we (a) rely on students from a nationally representative sample; (b) establish the impacts of less-severe exclusionary policies through measures of ISS; (c) explore both the short-term (math achievement) and long-term (college attendance) impacts associated with high and low-social control contexts and demonstrate how these impacts are related; and (d) limit bias associated with selection into schools that vary in ISS rates by using a counterfactual model based on propensity scores. We also explore the difference between the direct and indirect effects of ISS, as well as the interactions of gender, race, and class variables with suspensions, high-suspension schools, and academic predictors of the outcomes.

Through our results, we demonstrate that even the least severe forms of exclusion, such as ISS, are associated with detrimental short and long-term effects on students that do not directly receive them, but—by no fault of their own—merely attend schools that overuse them. While a low-suspension school may be prone to some collateral consequences of their own, our study demonstrates that it is far worse to attend a high-suspension school. Attending a high-suspension school can lower students’ math achievement scores and, ultimately, decrease their ability to access post-secondary educational opportunities. Moreover, we found that these collateral effects were similar—and in many cases—worse than the direct effects related to receiving a suspension when not attending a high-suspension school.

Furthermore, when considering interactions with demographic variables we found that even if school suspension rates were equalized across high and low suspension schools, Black students would likely face other significant obstacles in their pursuit of post-secondary educational opportunities, such as early math achievement.

In total, a greater reliance on ISS provides an additional mechanism of educational stratification, and by doing so, further exacerbates inequities between schools. Thus, while ISS has been thought of as a policy alternative to out-of-school suspensions, this study has demonstrated that ISS might need an alternative as well.