Lower suspension rates can hide unequal discipline, working paper finds
Using detailed records from one large Southern district, researchers found that schools usually responded to incidents with a single punitive action — and non-suspension consequences were uneven too.
A drop in school suspensions can look like progress on discipline equity. A new July 2026 [working paper] challenges that assumption, arguing that schools may be redistributing punishment rather than reducing it. In Discipline Beyond Suspensions, researchers Youngsun Lee and Jing Liu of the University of Maryland and Emily K. Penner of the University of California, Irvine use unusually detailed administrative records from a large Southern district to examine a wider range of consequences after student incidents. Their central finding: schools most often respond with a single punitive action, and the “alternatives” to suspension are not necessarily applied more evenly. (edworkingpapers.com)
That matters because the paper is asking a more practical question than much of the recent discipline literature. Earlier studies helped show that racial disparities arise before a suspension is ever recorded: in who gets referred, how often, and what happens after a referral. This new paper pushes further down the chain of consequences, asking whether a district can lower suspension counts while leaving unequal treatment intact in other forms. (edworkingpapers.com)
The headline example is blunt. Among students involved in the same fight, the paper finds that White students were more likely to have loss of privileges treated as their harshest consequence, while Black students were more likely to receive suspensions or be reported to a school resource officer. Because that comparison is within the same type of incident, the finding is harder to dismiss as a simple difference in behavior categories alone. It suggests that looking only at whether suspensions fell can hide who is still getting the most severe response when trouble happens. (edworkingpapers.com)
The study arrives as suspension numbers remain one of the most visible discipline metrics in public policy and school improvement debates. That visibility is understandable: exclusionary discipline is linked to serious long-run harms. An NBER paper on school suspensions found that students assigned to schools with higher suspension rates were 15 to 20 percent more likely to be arrested and incarcerated as adults, with negative effects on educational attainment as well. Separate work led by Penner linked school suspensions and expulsions to later disparities in criminal justice involvement, education, and income measures. (nber.org)
Looking past the headline metric
The new paper does not argue that suspensions are harmless or that schools should stop trying to reduce them. Its point is narrower, and more uncomfortable: a district can improve on the metric everyone watches while still preserving an unequal punishment system underneath it. That concern fits with federal data practice. The U.S. Department of Education says the Civil Rights Data Collection has consistently shown discipline disparities for students of color, especially Black students, and students with disabilities. The same federal data system also tracks referrals to law enforcement and school-related arrests, not just suspensions and expulsions. (ed.gov)
In other words, the accountability problem is not that schools lack categories to examine. It is that public conversations and local dashboards often flatten discipline into a small set of headline numbers, especially out-of-school suspension. Lee, Liu, and Penner’s result suggests that once a school is under pressure to cut that one number, the real equity question becomes: what replaced it, for whom, and with what consequences? If the common response to an incident is a single punitive action, then changing the preferred sanction may simply reroute students into a different lane of punishment. That is an inference from the paper’s design, but it is a strong one. (edworkingpapers.com)
That is also why this paper feels useful for principals and discipline teams rather than only for researchers. A school can honestly report fewer suspensions and still leave students of different racial groups with different chances of losing privileges, being excluded from school activities, or encountering police or school-resource-officer involvement. A narrower metric can miss that redistribution entirely. (edworkingpapers.com)
What the paper adds to earlier discipline research
The authors are building on a line of work that has steadily moved upstream in the discipline process. A 2022 paper by Liu and co-authors showed that racial disparities appear both in referral rates and in how often referrals become suspensions, and that Black and Hispanic students received harsher discipline than White students involved in the same incident and with similar histories. A later study co-authored by Liu and Penner found that just 5 percent of teachers in a California district generated nearly 35 percent of all office discipline referrals, often for subjective offenses. Those studies made it harder to treat suspension gaps as a neutral readout of student behavior. The new paper extends that logic: even after schools try to move away from suspensions, inequity can persist in the substitute responses. (edworkingpapers.com)
That is an important clarification at a moment when many schools are trying to soften zero-tolerance approaches without appearing soft on safety. There is credible evidence that some alternatives can improve outcomes. In a 2025 American Economic Review study of Chicago Public Schools, researchers found that restorative practices reduced suspensions and arrests, with effects driven by Black students, and found no meaningful average decline in test-score value added. So the lesson from the new working paper is not that alternatives fail. It is that “alternative” is not a synonym for “equitable,” and that implementation details matter enough to change who bears the burden. (aeaweb.org)
What school leaders should and should not change
Readers should be careful not to overread one district study into a national rule. This is a July 2026 EdWorkingPaper, and the series says its papers are circulated for comment and discussion before rigorous peer review. The district in the study is unnamed and located in the South, which protects confidentiality but limits what outsiders can compare directly. That means the paper is best read as a warning about a plausible mechanism in discipline reform, not as a final national estimate of how often every alternative sanction is used. (edworkingpapers.com)
Still, the practical takeaway is strong. School leaders who want a serious discipline-equity audit should not stop at asking whether suspensions fell. They should review the full chain of responses after incidents by race, ethnicity, disability status, grade span, and school: who lost privileges, who was referred onward, who encountered law enforcement or a school resource officer, and which students were still getting the harshest available response after similar incidents. The question is not simply whether schools are suspending less. It is whether they are punishing differently in ways that are actually less exclusionary, less escalatory, and more evenhanded. (edworkingpapers.com)
That is the accountability blind spot this paper exposes. A cleaner suspension chart may reassure a board or calm a dashboard, but it does not by itself show that discipline is becoming fairer. The next thing worth watching is whether districts start publishing and internally reviewing a broader set of consequence data — and whether future peer-reviewed work in other districts finds the same pattern of inequity shifting rather than shrinking. (edworkingpapers.com)