Sections

Research

What are we learning from school suspension bans?

Catherine Mata
catherine mata
Catherine Mata Postdoctoral Research Associate - Annenberg Institute, Brown University, Policy Fellow - University of Maryland, Baltimore County, Affiliate - Stanford Pathways Network, Research Affiliate - Maryland Longitudinal Data System Center

October 21, 2024


  • Maryland’s ban on suspensions in early grades significantly reduced the number, probability, and duration of suspensions but did not eliminate them entirely.
  • Despite the overall reduction in suspensions, racial, economic, and gender disparities in suspension rates persisted, and gaps widened for students with disabilities.
  • The limited impact of the suspension ban suggests that broader, more targeted strategies addressing the root causes of disciplinary disparities are necessary to fully eliminate inequities.
A frustrated student is shown sitting in her room after being suspended.
A frustrated student is shown sitting in her room after being suspended. Antonio Guillem/Shutterstock

Exclusionary school discipline—i.e., suspension and expulsion—remains a controversial practice in U.S. schools. Researchers have documented that students of color and other historically marginalized groups are disproportionately subjected to exclusionary discipline, and these practices are associated with negative impacts such as reduced instructional time, lower achievement, and increased school dropout. However, there’s ambiguity about what the best alternative might be, and discipline advocates see more of a role for exclusionary discipline.

While some states and school districts have advocated for developing and implementing new discipline strategies that keep students in the classroom and emphasize behavior improvement, others have pushed for limiting or outright banning the use of suspensions. In some cases, both approaches have been adopted, combining suspension bans with alternative disciplinary methods. No state has fully banned suspensions, but a growing number of states have taken steps to limit allowable uses of exclusionary discipline. Partial bans restrict suspensions to students in specific grade levels or for specific infractions.

We are starting to see evidence of the effects of these restrictions. States that have restricted the use of exclusionary discipline for specific infractions, such as truancy (Arkansas) and absenteeism (Rhode Island), have seen modest reductions in suspensions. This limited impact may result from the limited scope of the policies or from schools exploiting loopholes to evade compliance. We see similar patterns at the local level, where districts like New York City, Philadelphia, and the Los Angeles Unified School District have banned suspensions for certain infractions and (in some cases) adopted alternative practices— all aimed at reducing the use of exclusionary discipline. In each case, reforms led to reductions in suspensions for the prohibited infractions but did not effectively reduce the disproportionate use of exclusionary discipline against historically marginalized students.

In a new study, co-authored with Jane Lincove and Kalena Cortes, I look at the first statewide ban that is focused on suspensions in early primary grades, enacted by the state of Maryland.

Examining Maryland’s ban on suspensions in early elementary grades.

Beginning in the fall of 2017, Maryland severely restricted the use of out-of-school suspensions in grades PK-2. The policy presents a valuable opportunity to examine the effects of state-level discipline policies. First, the Maryland ban is very broad. It applies to all public schools in the state, and compared to other state bans, it bans the use of suspensions with only a few exceptions in cases of an imminent threat of violence. Second, while policymakers intended for the ban to reduce disproportionality by race, gender, and disability status, the ban is neutral regarding student characteristics, so it offers an opportunity to test whether states can eliminate inequities in suspensions simply by restricting suspensions for all. Finally, the ban focuses on early primary grades, which provides an opportunity to assess whether grade-specific suspension restrictions can have schoolwide impacts, figuring that schools might adopt alternative practices or voluntarily apply the suspension ban at the school level in ways that may positively impact students in older grades.

Descriptive evidence of the effects of the ban is shown in Figure 1. A student’s probability of suspension in K-2 grades increased from 1.4% to 1.9% from 2015 to 2017, and then declined to less than one percent after the ban in 2018. In grades 3-5, which were not subject to the ban, the rate remained steady at 3.1% from 2017 into the post-ban years.

Figure 1

The state ban substantially reduced, but did not eliminate, suspensions in treated grades.

We show that the probability, number, and duration of suspensions declined substantially and significantly in all three grades subject to the ban (K-2). In second grade, for example, the number of suspension events decreased by 60% after the ban, while suspension days were reduced by 64%.

While we observed statistically significant reductions in all suspension outcomes across the treated grades (K-2), suspensions were not fully eliminated. This raises an important question: Why did students in grades K-2 continue to get suspended after the ban? Were these really cases of an imminent threat of violence? Or did schools change how they recorded disciplinary infractions in order to continue suspending students?

We find no evidence that schools engaged in strategic behavior to continue exclusionary discipline after the ban.

We expected to see an increase in the use of violent codes to justify out-of-school suspensions during the ban (since those were still allowable), or possibly a shift toward increased use of in-school suspensions. However, we did not see this in our data. In fact, we observed a significant decrease in the use of violent codes in grades K-2. We also found that in-school suspension was rare before the ban and no more likely to occur after the ban.

While all student subgroups experienced significant declines in frequencies and suspension rates, disparities by race, income, and gender persisted, and gaps related to disability status widened.    

We find that the suspension ban significantly reduced the probability, number, and duration of suspensions in K-2 grades for Black, male, economically disadvantaged, and students with disabilities (SWDs), with no increases in the use of violent codes or in-school suspensions. However, these reductions were not large enough to eliminate the disproportionate use of suspensions against certain groups.

The probability of suspension before and after the ban for different subgroups of students is illustrated in Figure 2. Despite the reduction in suspensions, Black students continue to be suspended more frequently than white students, male students more than female students, and economically disadvantaged students more than their non-economically disadvantaged peers. While SWDs saw a decline in their chances of being suspended, suspension rates of students without disabilities declined as well. As a result, disparities between SWDs and their non-disabled peers widened after the ban. For example, before the ban, SWDs were suspended five times more often than their peers. After the ban, this rate increased to six times more often.

FIGURE 2

Probability of getting suspended before and after K-2 suspension ban, by student subgroup

What can we learn from Maryland’s ban?

The policy implemented in Maryland provided limited flexibility for schools to suspend students in early grades while allowing schools to continue suspending students in upper grades. Such exceptions were necessary to gain the support of teachers who are hesitant to relinquish their existing disciplinary tools and worry that other students might be negatively affected if a truly disruptive student cannot be removed from the classroom.

The Maryland policy also lets us investigate whether restricting suspensions at lower grade levels sparks schoolwide changes in suspension practices. Our results suggest that policymakers aiming to reduce suspensions across all grades will need to extend their focus beyond just the early grades. We observe progressively smaller reductions in suspension rates in grades not covered by the ban, with no significant effects by fifth grade. This suggests limited schoolwide impacts if a policy only changes school practices in some grades while maintaining business-as-usual practices in other grades.

Finally, it is unsurprising that the ban failed to eliminate disproportionalities in suspensions and even exacerbated some inequities, particularly for students with disabilities. The ban is a broad policy that does not take aim at any of the underlying causes of discipline disparities. To effectively eliminate entrenched disparities, schools will likely need to attend to the multitude of reasons why some student groups are disciplined at higher rates than others. For example, schools might need to address factors that contribute to educator biases in referrals and suspensions, rather than simply implementing a race-neutral strategy and hoping it will reduce racial inequalities.

Author

The Brookings Institution is committed to quality, independence, and impact.
We are supported by a diverse array of funders. In line with our values and policies, each Brookings publication represents the sole views of its author(s).