Does This Really Work? Measuring Restorative Practices, Racial Disparities, and School Relationships in a Data-Driven Era

• In Event: A Cross-Institutional Examination of Discipline Policy Implementation Processes in K–12 Schools

Abstract

Purpose
Striking racial disparities exist in school discipline, as Black and Latinx students are overrepresented in office referrals, suspension and expulsion practices (U.S. Department of Education, 2014). While the sources of school discipline disparities are multiplicative and debated, schools and policymakers have embraced a variety of interventions to reduce the racial discipline gap. Among the most popular of these are restorative justice practices (RP), which seek to build and repair relationships to form supportive and inclusive school communities that are less reliant on exclusionary discipline. Despite the promise of RP, research finds a nuanced impact on student discipline that may reduce inequity in some contexts (Augustine et al., 2018; Anyon et al., 2016) but unintentionally exacerbate it in others (Davison, Penner and Penner, 2019; Hashim et al., 2018).
This mixed-method study adds to this literature by investigating and describing the adoption of RP in Saint Paul Public Schools (SPPS) as a teacher-focused and teacher-led school climate improvement initiative. Framed by sociological understandings of organizational relationships, the goal of this study is to understand the extent to which restorative practices impacted racial disparities in SPPS, and to explore the organizational factors that impacted the trends we observe.

Data and Methods
We draw on multiple data sources for this project. First, we utilize longitudinal student-level administrative records (including discipline and attendance records) and school survey data to calculate interrupted time series estimates of the impact of restorative practices on student outcomes. In addition, we pay particular attention to racial disproportionality within these schools, adjusting reference groups and matching comparison schools along similar demographical lines. We supplement this administrative data with qualitative data collected at each RP school to capture elements of RP implementation that are unaccounted for in quantitative analysis.

Preliminary Findings
We find variation in RP effects across school sites, where some schools experienced pronounced improvements in teacher-student relationships and other aspects of school climate. There were also reductions in chronic student absenteeism and relatively slight reductions in use of exclusionary discipline practices (i.e. suspension or expulsion). The strength of these findings, however, is seemingly influenced by multiple implementation factors, including staff turnover, the reach of training and coaching within the school, and differential focus on the extent to which practices that build community school wide were successfully installed.

Significance
Our findings underscore both the promise and struggles of implementing school-wide RP and similar relational discipline practices. While the theoretical framework of RP has strong alignment with other research-based models (i.e., the social-ecological model), the success of its implementation hinges on the malleability of existing institutional relationships and dynamics of power and privilege that impact educators and students on a daily basis. Although policymakers often seek quick fixes for racial achievement and discipline gaps, these issues are multi-layered and context-dependent. Further, we note the limits of pre-existing data systems in measuring the effectiveness of relational discipline programs such as restorative practices.

Authors

• Miles Davison, University of California - Irvine

• Barbara McMorris, University of Minnesota - Twin Cities

• Kara Beckman, University of Minnesota