A restorative approach to student discipline shows promise in reducing suspensions and arrests

Suspensions are a common form of official discipline meted out in American schools, but does this punishment actually reduce harmful student behavior?

Beginning in the 1970s, schools increasingly turned to exclusionary discipline practices in hopes of curbing disruptions and creating a safe learning environment. By the 2011-2012 school year, nearly 3.5 million students (roughly 7% of all public school students nationwide) were suspended annually, resulting in a staggering loss of 18 million days of instruction. The idea behind exclusionary discipline is simple: Remove disruptive students to preserve order. But the long-term consequences tell a different story. Research shows that suspensions can lead to lower educational attainment and increase the likelihood of incarceration—what some call the “school-to-prison pipeline.” Even more troubling, exclusionary discipline hasn’t affected all students equally. Since 1974, suspension rates for Black and Hispanic students have more than doubled.

Despite the potentially detrimental impacts on life outcomes and racial inequities associated with suspensions, educators are often given few other tools to manage discipline. Educators identify classroom discipline among the most challenging aspects of their role but often face a dilemma: how to address undesirable behaviors, especially in a context where 71% of schools report incidents of violence, without causing further harm. Should students be kept in school, potentially disrupting or even harming their classmates, or be removed, possibly hindering their own academic progress? The challenge is compounded by a scarcity of resources and expertise, which often leads schools to default to suspension as the lowest-cost means of restoring order. In practice, though, the costs of suspensions may be much higher as they are passed on to the students perceived to be disruptive.

Restorative practices as a shift in philosophy

Restorative practices (RP) is a relatively new and alternative approach to school discipline that is gaining traction across U.S. public schools. As opposed to more traditional school discipline practices, which seek to deter misconduct and prevent recurring misbehavior by threatening punishment (e.g., suspension), RP seeks to accomplish the same by helping students understand the harm caused by their misconduct. By focusing on harm reparation and community dialogue instead of punishment after misconduct occurs, RP represents a major philosophical shift relative to the zero-tolerance approaches that proliferated until recently. 

Let’s imagine, for example, a student that talks back to their teacher during class and is unresponsive to the teacher’s initial attempts to curb the misbehavior. Using a traditional approach, the teacher might next send the student to the principal. This could then set off a sequence of escalating punishments possibly including suspension, which can ultimately lead to the student’s disengagement with learning.

Using RP, on the other hand, the teacher might choose a different path. After class, they might sit down with the student to explore what happened. Maybe they involve a counselor or organize a conversation with peers to show the student how their behavior affected others. The goal of such an approach is to build understanding rather than to punish. Restorative practices—and they do take practice—are meant to build conflict-resolution skills on the parts of both teachers and students.

Behavior change from RP in Chicago Public Schools

Can RP reshape student discipline practices in an entire school system? To answer this question, we draw on a unique research-practice partnership with Chicago Public Schools (CPS) to study the implementation of an RP program in CPS. From 2014 to 2019, CPS rolled out annual waves of RP training for school staff with funding from a U.S. Department of Justice grant. Starting in 22 high schools, a total of 73 high schools were eventually chosen to receive the training. A few times each week, school staff in these schools met with professionals who coached them on implementing restorative practices. These 73 schools were selected by CPS’s Office of Social and Emotional Learning (OSEL) to receive RP programming based on a variety of factors including their out-of-school suspension rates, size, and interest in participating.

Our research findings provide compelling evidence that RP positively impacted student behavior. We use a statistical approach called difference-in-differences to isolate the causal impact of RP by comparing changes in student and school outcomes before and after RP was introduced to changes observed in other CPS schools. We find a meaningful reduction in the number of suspension days (0.17 days or an 18% overall decline) served by students in the four years following RP rollout. We cannot discern from the data whether this decline was caused by teachers reducing their use of suspensions or students being less likely to engage in misconduct. However, our findings provide encouraging evidence that RP can reduce school suspension rates over time.

Beyond the classroom, we find that RP led to a 19% decrease in the number of overall student arrests and a 15% decrease in the number of arrests out of school. These findings, enabled by linking CPS students with data from the Chicago Police Department, suggest that the social and emotional learning skills that RP develops for students translate into decreased misconduct outside the classroom.

Black students in CPS particularly benefited from RP

We find that the RP roll out in CPS was especially beneficial for Black students. The disparity between Black students and their peers before RP was stark: In 2013, Black males in CPS were suspended for four times as many days and arrested six times more often than white males. 

After the rollout of RP, Black male and female students saw substantial reductions in out-of-school suspension days, more than double the average reductions observed across all students. The story is similar for arrests: We found that RP reduced the number of arrests faced by Black males by 0.073 (three times the average reduction across all students). Additionally, Black male students were absent for 1.66 fewer days after RP was implemented, suggesting that they became more engaged in school as a result of RP.

These benefits underscore RP’s potential to address long-standing racial disparities in school discipline and contribute to a more equitable educational environment.

RP’s impact on peers’ test scores, school climate

Skeptics may argue: If students prone to misconduct are suspended less often, they may cause more classroom disruption when spending more days in school. In turn, these disruptions might negatively affect the learning and safety of their peers. We address these concerns in two ways, and our findings raise doubts about these concerns.

First, we investigate changes in students’ math and reading standardized test scores using state-of-the-art machine learning methods to classify student peer groups as more or less “disruptive.” Based on these classifications, we find no evidence that RP reduced students’ test scores across peer groups predicted to contain more disruptive students versus fewer disruptive students.

Second, we explore student survey data, and RP-implementing schools report more supportive and participatory school environments after the rollout. On CPS’s annual “My Voice, My School” survey, students reported improvements in their perceptions of student classroom behavior, their sense of belonging in the school environment, and school safety.

Conclusion

Since the COVID-19 pandemic, over 80% of U.S. public schools have reported worsening student behavior and declines in students’ socio-emotional development, leaving educators to grapple with how to manage misconduct while ensuring learning continues.

Suspensions, the default response to disruptive behavior, often push students—especially Black and disadvantaged students—further behind, increasing their risk of academic failure and interaction with the criminal legal system. Restorative practices can offer an alternative approach to discipline. By focusing on RP, our findings suggest that schools can build a more participatory, supportive, and equitable school environment without compromising student learning.

Implementing RP at scale still remains a challenge. Education researchers are still learning about the conditions and implementation supports necessary to make it a success. Though there is still much to learn, CPS’s success with RP offers an important, promising example of how policymakers and practitioners in other school systems might use alternative approaches that are not only less punitive but also more reparative to transform school discipline.