Black, Indigenous, and Latine students, students with disabilities, and students from low-income backgrounds are disciplined at much higher rates than their peers, and this gap is not explained by differences in behavior.

• An analysis of nearly 16,000 U.S. public middle and high schools using Civil Rights Data Collection data found that Black students are at significantly higher risk of suspension or expulsion compared to White students in both middle and high school (see Figure 1).
• A review of 183 studies found that even when you compare students who broke the same rule, had the same discipline history, and came from the same economic background, race still predicted who got punished more harshly.
• Black economically disadvantaged students are suspended at higher rates than White economically disadvantaged students. Middle- and upper-income Black students are suspended at higher rates than White students at the same economic level.

Figure 1: Black boys are nearly twice as likely as White boys to receive an out-of-school suspension in K-12 public schools.

Source: Civil Rights Data Collection.   *AIAN = American Indian/Alaska Native

Three distinct disparity patterns operate inside any district. Each has different drivers and, thus, requires different solutions, and most districts face some mix of all three.

1. Same behavior, different response, within a school. Two students commit the same infraction in the same building, and one is referred or suspended while the other is warned. The drivers are interpersonal: implicit bias, vague offense categories that hand judgment to adults, classroom relationships, and how repeated behavior is read as a “pattern” for some students and not others.
2. Same behavior, different response, across schools. A student who shoves a peer in one school is suspended; in another school across town, the same shove gets a restorative conversation. This pattern is driven by school- and district-level conditions: discipline codes that make exclusion the default, school climate, leadership stance, staffing ratios for counselors and psychologists, and whether alternatives to suspension exist and are supported.
3. Different histories and resources. Because of historical policies and contemporary funding formulas, some schools are more likely to have resources to sustain restorative approaches, while others might be highly policed, with fewer wrap-around services. This pattern reflects structural conditions in which students arrive at and then experience school (e.g., concentrated poverty, exposure to violence, under-resourced schools with larger classes and fewer adults), and it is the pattern most shaped by structural context.

How has history shaped today’s discipline landscape?

Discipline disparities are, in large part, produced by decades of policy choices that still shape the conditions in which schools work.

• Historical housing and segregation policies: Redlining and related policies concentrated students of color in under-resourced schools with larger class sizes, fewer counselors and social workers, and stronger police presence. Recent research links these historical policies to current discipline disparities.
• Federal and state policy in the 1980s and 1990s: Mandated zero-tolerance responses brought police officers into schools as a routine presence and required schools to refer certain offenses to law enforcement. Under these policies, suspension rates climbed sharply, with Black students two to four times more likely to be suspended than White students. More generally, punitive school discipline is shaped by broader beliefs in society that prioritize control and punishment, especially for marginalized groups. For example, the United States has one of the highest incarceration rates in the world, and Black Americans are imprisoned at nearly five times the rate of White Americans, a pattern that closely mirrors racial disparities in school discipline.
• Resource inequities: Schools in low-income communities often lack counselors, social workers, and mental health supports, making exclusionary discipline a default response. More resourced schools are better positioned to implement preventative and restorative approaches.

This is the structural reality. It helps explain why two districts running the same intervention often get different results, and why the same fight in two schools can produce two completely different responses. The next section is about what leaders can change inside this reality.

Where do schools and districts have leverage to reduce disparities?

Many of the decisions that produce day-to-day disparities are made within schools and districts, and most of them can be changed. But each lever’s reach is shaped by the structural conditions in the section above.

1. Punitive discipline codes and vague offense categories. Most district codes are punitive by design: they specify exclusion as a primary response across a wide range of behaviors, including minor ones. A national review of district policies found they included fewer than half of the recommended elements for proactive, prevention-oriented discipline. When a school’s written policy lists suspension as the primary response, administrators and teachers who are under time pressure, or who lack confidence in other approaches, default to it. Vague categories such as “defiance,” “insubordination,” and “disrespect” are significantly more likely to be applied to Black and Latine students, where adults exercise judgment rather than observe a clear rule violation. 
2. Bias at the moment of referral. The discipline process starts when a teacher decides to send a student to the office. That decision is shaped by implicit racial bias. Some teachers may watch Black students more closely when monitoring for problem behavior, perceive their expressions as angrier, and are more likely to treat an incident as part of a pattern when the student is Black. In one experimental study, teachers read about hypothetical students who misbehaved twice. After the second infraction, teachers were more likely to see a Black student’s behavior as part of a pattern and to recommend harsher discipline than for a White student committing the same behaviors.
3. Anti-Black racism in discipline. Anti-Black racism is endemic in schools and filters through every level of discipline. At the policy level, even race-conscious leaders can fall back on deficit views of students that justify punitive measures, especially under pressure, shaping which behaviors get routinely punished and which get coached. At the classroom level, Black students are treated more harshly for similar behavior, and research has found that Black students admitted to being defiant specifically when teachers showed low expectations and little care. Teachers then perceived those same students as more defiant and referred them more often, creating a self-reinforcing cycle. Teachers with stronger student-teacher relationships tend to have lower rates of referrals.
4. Discipline as the default response to mental health needs. Children with higher rates of mental health symptoms are at greater risk of receiving exclusionary discipline, while being suspended or expelled in turn produces adverse mental health outcomes, creating a reinforcing cycle. Black and Latine students face higher rates of exposure to adversity and lower access to school-based mental health services. Without counselors and psychologists in genuinely therapeutic roles, discipline becomes the default response to mental health needs.
5. Data use and accountability. When schools don’t disaggregate discipline data by race and offense type, individually justifiable decisions accumulate into a racially inequitable pattern that no one is positioned to see.

What are the consequences of exclusionary discipline?

Exclusionary discipline does not deter future misbehavior, and it produces lasting academic, mental health, and life-course harms.

• It doesn’t deter future misbehavior and often makes things worse. A large study analyzing New York City public schools found that exclusionary discipline did not reduce future misbehavior for disciplined students or their peers, nor did it improve academic achievement or school climate. The more severe the exclusionary discipline, the greater its negative effects on a student’s future academic performance, attendance, and behavior.
• Academic harm: Students who are repeatedly removed from class lose valuable instructional time, which contributes to lower test scores, lower academic grades, and higher dropout rates.
• Mental health harm: A University of Minnesota study found an increase in the likelihood of depression and anxiety symptoms among students who had experienced exclusionary discipline. School suspensions are associated with reduced social connections and reduced feelings of belonging within schools.
• Juvenile justice involvement: Longitudinal studies show that suspended students are more likely to become involved with the juvenile and criminal justice systems. 

Why do well-intentioned reforms so often fail to close the racial discipline gap?

Efforts to reduce discipline disparities often fall short because they are short-lived, layered onto punitive systems, or overlook the central role of relationships and school climate in addressing inequity.

• Short-lived initiatives: Reforms such as restorative practices require time and consistency, but are often not sustained or implemented symbolically.
• Layered but conflicting approaches: Schools frequently layer new strategies onto existing punitive systems, limiting their effectiveness.
• Single-level fixes. Policies that ban suspensions for low-level offenses, on their own, often reduce overall numbers without reducing disparities because they address the school/district level but not the interpersonal-bias level or the structural conditions that concentrate behavioral incidents in particular schools.
• Insufficient focus on relationships and climate: Without addressing the underlying culture, beliefs, and relationships that shape discipline decisions, disparities are likely to persist.

A FOCUS ON SCHOOL AND DISTRICT POLICY

1. Monthly discipline data reviews, disaggregated by race, gender, grade, and infraction type, paired with coaching, provide a first step in helping teachers and administrators translate data patterns into changes in practice.

• Research on discipline reform in Maryland found that when schools used disaggregated data to identify specific patterns of disproportionality, they were better positioned to direct technical support toward the schools and classrooms driving the disparities and to measure progress over time. 
• A recent study found that when principals were given data on their own discipline referral patterns and coached on their use, they became less likely to escalate referrals to exclusionary consequences, suggesting that data feedback loops can shift individual decision-making, not just school-level policy.
• Schools receiving the ReACT intervention, which combined disaggregated discipline data analysis, professional development on implicit bias, and ongoing coaching, showed significant reductions in racial discipline disparities for Black students, while control schools showed minimal change. Notably, a prior RCT by the same research team found that sharing discipline disparity data with administrators alone produced no meaningful change, suggesting that data is necessary but not sufficient, and that coaching and professional development are what translate awareness into practice change.

2. Revising discipline policies to limit subjective offense categories and reserving exclusionary discipline for serious infractions provides a second step in reducing the role of implicit bias in disciplinary decisions. These revisions only work if they are paired with the development of alternatives to exclusionary discipline.

• Several large urban districts, including Denver and Los Angeles, have revised their discipline codes to end schools’ ability to suspend students for vague things like “willful defiance.” Since implementing the policy, LAUSD has seen a 75% drop in suspensions across all categories and a narrowing of racial disparities among students who are suspended.
• The Denver Public Schools (DPS) discipline reform defined six levels of intervention in a discipline ladder, with a core principle that disciplinary action should begin and be resolved at the lowest level possible, consistent with the nature of the violation. Rather than giving administrators an open menu where suspension was the easiest option, the policy defined a sequence that had to be worked through first. This was one of many parts of the DPS reform, but the district lowered suspension and expulsion rates by nearly 40%.
• Racial gaps in school discipline are widest where adult judgment plays the biggest role. This suggests that requiring a second-level review before a suspension is issued, especially for minor or subjective infractions, introduces an accountability check and has a strong theoretical justification for reducing bias-driven suspensions.
• When administrators have less-exclusionary options built into their approach, they use them, and disparities narrow as a result.

3. Restorative justice (RJ) has demonstrated the ability to reduce suspensions. It is most effective when implemented as a schoolwide framework to replace punitive discipline, but only when implemented with ongoing staff development.

• Restorative justice is a schoolwide approach that focuses on building relationships and repairing harm after conflicts, rather than relying on punishment. It was developed as an alternative to zero-tolerance policies and exclusionary discipline. Instead of asking which rules were broken and how to punish students, RJ asks who was harmed and how to repair relationships.
• There are different RJ models, but consensus across models is that RJ must be embedded in school culture to be effective. Common practices include restorative circles, restorative conferences, and mediation between those involved in a conflict.
• Studies in California and Pittsburgh show that well-implemented RJ can reduce schoolwide suspension rates, and a study in Chicago showed improved school climate and reduced out-of-school arrests. Student surveys also suggest RJ can strengthen relationships and reduce discipline referrals, especially for racialized and minoritized students.
• Race-neutral restorative practices risk widening disparities, even as they reduce overall rates. A study tracking restorative justice implementation across one large urban school district over nearly a decade found that by the fifth year of implementation, the Black-White suspension ratio had grown substantially. This occurred not because Black students’ absolute suspension rates increased, but because White students’ rates declined faster, suggesting White students benefited most from the practices implemented. 
• Restorative approaches that extend beyond responding to incidents, to include proactive regular check-ins, class community meetings, and relationship-building routines, show stronger effects on both suspension rates and student engagement. 
• The most promising evidence on RJ comes from programs that combine relationship-building structures with explicit anti-bias and culturally responsive staff development. A systematic review of interventions designed to reduce the race-discipline gap found that those addressing racial identity, cultural relevance, and implicit bias produced stronger results than those that did not.
• Implementing restorative approaches as an add-on to, rather than a replacement for, zero-tolerance practices weakens their effectiveness and may reinforce the inequities they are meant to address. When RJ is layered on top of existing punitive systems [2] rather than replacing them, the same biased discretion that drives racial disparities in the first place now governs who gets the restorative option and who gets suspended, meaning RJ may disproportionately benefit White students.

4. Schoolwide Positive Behavioral Interventions and Supports (SWPBIS) reduces overall rates of exclusionary discipline, but whether it narrows racial disparities depends heavily on whether it is implemented with an explicit, race-conscious equity focus.

• SWPBIS is a tiered framework for proactive behavior management that establishes consistent schoolwide expectations, explicitly teaches them to students, and reinforces positive behavior through recognition and support. Studies show that when implemented with fidelity, SWPBIS reduces overall suspensions, including for those from racialized and minoritized groups. Examples include districts in California, Hawaii and Illinois, and elementary schools from across the country.
• The critical caveat, documented consistently, is that standard SWPBIS implementation does not reliably reduce racial disparities. The framework was originally designed as a race-neutral approach. Multiple studies find that overall suspension rates decline in SWPBIS schools, but Black and Latine students’ relative risk of suspension compared to White students remains largely unchanged.
• Equity-focused SWPBIS, added to the standard framework, can reduce racial disparities. In one study, schools were randomly assigned to receive a whole-school equity intervention, the ReACT approach, layered onto existing SWPBIS. ReACT involved identifying race-specific “vulnerable decision points” where implicit bias is most likely to influence referral decisions, providing professional development on culturally responsive behavior strategies, and explicitly teaching staff about implicit bias. Intervention schools showed significant decreases in racial disparities in office discipline referrals for Black students, while control schools showed minimal change.
• Emerging models like Culturally Responsive SWPBIS and Learning Labs, which bring together educators, families, and students to redesign discipline practices, are promising practices. Studies have shown that Learning Labs can create structured space for collaboration, helping schools strengthen relationships and develop shared, culturally responsive discipline solutions. These approaches may also enhance the equity impact of SWPBIS.
• A decade-long synthesis of discipline reform research found that coordinating SWPBIS with policy changes, restorative practices, and culturally responsive professional development produces better outcomes than any single approach implemented in isolation.

5. Schools with more mental health professionals and counselors have fewer disciplinary incidents and lower suspension rates.

• A national study that analyzed data from about 60,000 schools found that schools with more school psychologists and social workers had lower suspension rates for students with disabilities. The presence of school psychologists, specifically, was associated with a reduced risk of suspension for Black and Latine students with disabilities. A separate large-scale analysis of schools using the Civil Rights Data Collection found that schools with higher-quality mental health teams had lower suspension and expulsion rates, especially for minority students.
• School mental health professionals serve as an alternative pathway to exclusionary discipline: rather than escalating a behavioral incident to an administrator, counselors, social workers, and psychologists can conduct needs assessments, provide brief interventions, connect students to services, and coach teachers on de-escalation and relationship-building strategies.
• Staffing alone is insufficient. The impact of mental health professionals depends on clear roles, goals, and integration into discipline reform. Studies consistently find that when counselors are assigned non-school counseling activities such as clerical work and testing duties, therapeutic effectiveness declines and burnout increases.

A FOCUS ON CLASSROOM IMPLEMENTATION

1. Practices to build positive teacher-student relationships with a race-conscious lens are associated with lower suspension rates for students of color. The most effective interventions avoid explicitly naming bias as the target.

• A brief “empathic discipline” intervention delivered to teachers cut racial disparities in suspension rates by up to 50%. The intervention, delivered in two 45-minute online sessions, helps teachers adopt a growth mindset about student behavior, build relationships proactively, and practice perspective-taking by asking students directly what they are experiencing. Notably, the intervention’s materials do not focus explicitly on bias or racial disparities, an approach that can put teachers on the defensive and backfire, but instead reorient teachers toward their core purpose of helping students learn and grow.
• The My Teaching Partner-Secondary (MTP-S) program, a video-based professional development coaching model, produced specific gains for Black students in reducing exclusionary discipline referrals in a randomized controlled trial.
• Teachers who intentionally build relationships can reduce exclusionary discipline. In a longitudinal study following middle schools in the Pacific Northwest, teachers used the WOW strategy (Welcome students, Own your classroom environment, and Wrap up class with intention) to foster positive relationships within a larger schoolwide intervention. Middle schools that employed this strategy observed reduced rates of in-school suspensions and out-of-school suspensions, as well as reduced risk of out-of-school suspensions for students of color.
• Structured opportunities for teachers to learn about students as whole people, through advisory periods, home visits, community walks, and family engagement practices, are associated with more supportive responses to misbehavior and fewer referrals. Teachers who build this kind of relationship knowledge are more likely to respond to misbehavior with curiosity and support rather than escalating to a disciplinary referral.

2. Culturally responsive classroom practices have been shown to reduce discipline disparities by strengthening teacher-student relationships and decreasing cultural mismatches that generate conflict.

• Teachers who use more culturally responsive practices, such as integrating cultural connections into the curriculum, understanding cultural differences, and holding high expectations of all students, have higher levels of academic engagement and lower disruptive behavior than peers who use fewer. 
• Learning students’ backgrounds and adjusting expectations accordingly reduces discipline referrals. In one study, middle school students completed a worksheet describing how expectations looked at school, at home, and in their neighborhood. Teachers who used this information to adjust their classroom expectations observed more equitable treatment of Black students in their classrooms. The key mechanism was that discipline disparities often arise not from intentional defiance, but from mismatches in expectations that neither party recognizes as cultural rather than behavioral.
• Teachers’ implementation of restorative circles must be culturally responsive to be effective. A year-long ethnographic case study revealed how emphasizing the needs of students and the greater school community, laying ground rules and goals of circles, and allowing students to co-construct understandings facilitated a meaningful restorative circle that motivated change over time.

3. Individualized coaching that helps teachers improve the quality of their interactions with students can reduce racial disparities in discipline referrals.

• The MTP-S coaching program is one of the only coaching interventions with rigorous experimental evidence showing it can close the racial discipline gap specifically. With the support of a coach for an entire school year, teachers regularly reflect on video recordings of their classroom instruction to improve the quality of their interactions with students. The study concluded that the racial discipline gap reduces when teachers’ instruction improves: the degree to which teachers facilitated higher-level thinking, problem-solving, and metacognition was significantly linked to their equitable and infrequent use of discipline referrals.
• Double Check is a professional development and coaching framework built around five domains of culturally responsive practice: connection to curriculum, authentic relationships, reflective thinking, effective communication, and sensitivity to students’ culture. Multiple RCTs of Double Check have demonstrated positive impacts: one trial showed it reduced educators’ disproportionate use of office referrals for Black students and student disruptive behavior, while also increasing student cooperation and teachers’ use of proactive behavior management strategies.

A FOCUS ON STUDENTS

1. Interventions targeting students’ own sense of belonging can help students approach new teacher relationships with more security and interrupt the negative cycles that drive discipline disparities.

• Brief “identity safety” exercises delivered early in the school year, targeting belonging, growth mindset, and values affirmation, produced dramatic, lasting reductions in discipline citations for Black and Latino boys. A study combined social-belonging, values-affirmation, and growth-mindset interventions delivered in several class sessions in two middle schools with large Latine populations. The combined treatment reduced discipline citations among negatively stereotyped boys in 7th and 8th grade by 57% compared to a control condition. 
• Middle schoolers who participated in a self- and values-affirmation journaling activity built positive perceptions of their identities. This treatment reduced Black-White suspension and discipline referral gaps by approximately two-thirds, with the strongest effects for Black students with prior discipline experiences.
• Short classroom activities designed to help students build a sense of belonging have also been found to reduce discipline citations for Black and Latino boys. In one study, students read testimonials from older students describing how they initially felt uncertain or out of place but came to feel more connected over time. They then reflected on their own experiences and wrote about how belonging can grow with effort and relationships. The goal was to normalize uncertainty, reduce fears of being stereotyped or excluded, and shift how students interpreted everyday setbacks, which the researchers hypothesized would interrupt negative cycles of mistrust between students and teachers.
• These interventions are most effective when delivered at the start of the school year, before negative cycles take hold.

Changes to inequitable discipline systems must not simply seek to teach students to cope with injustice. For this reason, these student-level interventions may complement, but do not replace, system-level reform.