Do Strict School Discipline Policies Increase the Likelihood of Adult Criminal Involvement?

This study investigates whether strict school discipline policies, particularly suspensions, increase the likelihood of adult criminal involvement. Using data from a school boundary change in Charlotte-Mecklenburg Schools and principal turnover across North Carolina, researchers estimate the long-term impact of school suspensions on educational attainment and criminal activity. They find that attending a school with stricter discipline policies increases the likelihood of being arrested by 17% and incarcerated by 20%. It also reduces college attendance by 11%. The negative effects are strongest for male students and students of color, highlighting how strict disciplinary policies exacerbate existing educational and incarceration disparities.

Full Citation and Link to Article

Bacher-Hicks, Andrew, Stephen B. Billings, and David J. Deming. 2024. "The School-to-Prison Pipeline: Long-Run Impacts of School Suspensions on Adult Crime." American Economic Journal: Economic Policy, 16(4): 165–193. https://www.aeaweb.org/articles?id=10.1257/pol.20230052

Extended Summary

Central Research Question

This study examines the long-term consequences of strict school discipline policies, specifically school suspensions, on educational attainment and adult criminal involvement. The research seeks to determine whether attending a school with a higher suspension rate increases the likelihood of students engaging in criminal activity later in life. The authors leverage administrative data from North Carolina, using a natural experiment created by a school boundary change in Charlotte-Mecklenburg Schools (CMS) and principal turnover across the state. The study contributes to the broader conversation on the "school-to-prison pipeline," which suggests that exclusionary school discipline policies may contribute to long-term negative life outcomes, particularly for disadvantaged students.

Previous Literature

Prior research has established connections between school environments, discipline policies, and student outcomes. Several studies have shown that school attendance reduces subsequent criminal activity (Anderson, 2014; Cook & Kang, 2016; Jacob & Lefgren, 2003; Lochner & Moretti, 2004). Other research has demonstrated that attending a higher-quality school or being surrounded by more advantaged peers improves educational and life outcomes (Billings, Deming, & Rockoff, 2014; Cullen, Jacob, & Levitt, 2006; Deming, 2011). However, limited research has explored the mechanisms through which school policies influence long-term engagement with the criminal justice system.

One leading hypothesis is that strict school discipline policies, particularly suspensions and expulsions, contribute to criminal involvement by increasing students’ interactions with at-risk peers and exposing them to law enforcement at an early age (Fabelo et al., 2011; Owens, 2017; Weisburst, 2019). Studies have documented a positive correlation between school suspensions and later-life involvement with the justice system, but causal evidence has been lacking.

Policy changes in the past two decades have shifted away from punitive discipline measures toward restorative justice approaches. For example, in 2012, New York City Public Schools prohibited suspensions for minor infractions and implemented peer mediation programs. Chicago Public Schools followed a similar trajectory, reducing suspensions and increasing behavioral support interventions between 2009 and 2014. The Obama administration also issued guidance in 2014 urging schools to limit exclusionary discipline in favor of restorative practices. These changes highlight a growing concern that traditional discipline policies may do more harm than good, making research on their long-term effects particularly relevant.

Data

The study uses administrative data from CMS and statewide principal records in North Carolina. The primary dataset includes detailed student records from 1998 to 2011, tracking demographics, standardized test scores, school assignments, suspensions, and eventual criminal justice outcomes. The researchers also incorporate criminal justice data from Mecklenburg County, which includes all adult arrests and incarcerations from 1998 to 2013. These records allow the authors to examine whether school suspensions in middle school increase the likelihood of criminal involvement in adulthood.

Additionally, the study integrates college enrollment data from the National Student Clearinghouse (NSC), covering over 93% of students nationwide. This allows the researchers to assess how school suspensions affect college attendance.

To identify causal effects, the study focuses on students impacted by a school boundary change in 2002, which reassigned half of CMS students to new schools. This policy change created a natural experiment, as students who previously attended the same schools and lived in the same neighborhoods were suddenly placed in different educational environments with varying disciplinary policies. The researchers also examine principal turnover data across North Carolina, showing how changes in school leadership affect suspension rates and student outcomes.

Methods

The researchers employ a quasi-experimental design, leveraging the boundary change in CMS as an exogenous source of variation in school assignment. This allows them to estimate the causal impact of attending a school with a higher suspension rate.

To account for student sorting, they compare students who lived in the same neighborhoods and were previously assigned to the same school but were reassigned to different middle schools due to the boundary change. They estimate school-level effects on suspensions by conditioning on student characteristics and prior achievement, ensuring that their findings are not driven by preexisting differences among students.

The primary outcome variables include:

• The number of days suspended in middle school

• High school dropout rates

• College attendance within 12 months of expected graduation

• Adult criminal involvement, including arrests and incarcerations between the ages of 16 and 21

The authors also conduct a secondary analysis using principal movement data across North Carolina to test whether changes in school leadership affect suspension rates and student outcomes. This helps isolate the role of school discipline policies from other school characteristics, such as funding, peer composition, or teacher quality.

Findings and Size Effects

The study finds significant negative long-term effects of attending a strict-discipline school. A one-standard-deviation increase in a school’s suspension effect leads to:

• A 16% increase in the number of days a student is suspended per year

• A 15% increase in high school dropout likelihood

• An 11% decrease in the likelihood of attending a four-year college

• A 17% increase in the likelihood of being arrested between ages 16 and 21

• A 20% increase in the likelihood of being incarcerated between ages 16 and 21

These effects are particularly pronounced for male students and students of color. The likelihood of adult arrest for students of color increases by 4 percentage points, compared to 2.7 percentage points for White students. The impact on incarceration is also greater for students of color (3.1 percentage points) than for White students (1.9 percentage points). Similarly, male students experience significantly larger negative effects than female students.

Interestingly, the researchers find no significant impact on short-term academic achievement, measured by standardized test scores. While suspensions reduce instructional time for disciplined students, any negative effects may be offset by the removal of disruptive peers from the classroom. However, the long-term negative effects on educational attainment and criminal justice outcomes suggest that the costs of exclusionary discipline policies far outweigh any short-term classroom benefits.

Conclusion

This study provides compelling evidence that strict school discipline policies, particularly high suspension rates, have lasting negative effects on students' educational and criminal justice outcomes. The findings suggest that suspensions do not deter misbehavior but rather increase the likelihood of long-term criminal involvement, particularly for marginalized groups.

The research has important policy implications. As school districts continue to shift toward restorative justice approaches, this study reinforces the rationale for limiting exclusionary discipline practices. Rather than removing students from the classroom, schools should prioritize interventions that support behavioral development and keep students engaged in the educational system.

By demonstrating that strict discipline policies disproportionately harm male students and students of color, the study also highlights the role of school policies in exacerbating racial and gender disparities in the criminal justice system. Given the substantial negative effects observed, policymakers should consider further reforms to disciplinary practices to reduce the school-to-prison pipeline and promote more equitable long-term outcomes for all students.