The White House has proposed that a new school safety commission review whether to rescind current federal guidance that clarifies schools’ legal obligations to avoid race-based discrimination in school discipline. The announcement follows recent debates over federal enforcement of civil rights laws, and whether suspensions and expulsions are needed to maintain order and safety. It also comes at a time when state legislators are forging ahead with legislative proposals to reduce suspensions for nonviolent behaviors, and to encourage the use of preventative strategies such as positive behavioral interventions and supports and restorative practices.

Federal data have long shown that children of color experience disproportionately high rates of suspension, and various studies have shown a relationship between suspensions and harmful outcomes, including an increased likelihood of juvenile justice interaction. These findings have proven convincing to state lawmakers: In 2017 alone, there were at least 34 bills proposed on the topic of suspension and expulsion and at least 25 proposed on alternatives to exclusionary discipline.

So why, at the federal level, are officials reconsidering the school discipline guidance? The guidance clarifies that a seemingly neutral policy that disparately impacts students of color is discriminatory if it is not necessary to meet an important educational goal. A key question for federal officials, then, is whether disciplinary removals—which we know disparately impact students of color—are necessary to meet an important educational goal.

While the determination of whether a particular discipline policy is discriminatory is situational, research has long suggested that disciplinary removals are unlikely to serve education goals. Suspensions have not been shown to improve school safety or student behavior. Indeed, one study of U.S. and Australian students showed that suspension increased the likelihood that students would engage in problem behaviors. Further, dozens of studies have shown an association between suspension and poor educational outcomes, including decreased graduation rates and increased grade retention.

Decades of school discipline research informing state and federal initiatives have been largely consistent on these points; however, as is often the case in social science, this research is made up of mostly correlational studies. This has left room for questions—particularly in federal policy debates—as to whether suspensions could be in students’ best interest, even if children of color are disproportionately targeted.

Researchers generally reserve conclusions about causal impacts for studies where the topic of interest is randomly assigned, but there are statistical methods that allow researchers to investigate causal relationships in instances where random assignment would be ethically dubious (as would be the case with suspension). These techniques ground a new peer-reviewed study by Dr. Janet Rosenbaum, which provides new evidence that school suspensions directly lead to detrimental outcomes, particularly for children of color. Its findings are familiar but troubling, given the rigor of the study.

In this case, the author compared students who were the same on a host of characteristics—including demographic, health, and family characteristics—that could place the youth at risk for negative outcomes. After 12 years, youth that were suspended between the eighth and twelfth grades were six percent less likely to have graduated from college and 24 percent less likely to have obtained a four-year degree when compared to youth who had not been suspended. Such youth were also 30 percent more likely to have been arrested once, 51 percent more likely to have been arrested two or more times, and 23 percent more likely to have been in prison. The results were worse for children of color: Black youth who had been suspended were 58 percent more likely to have been arrested compared to black youth who had not been suspended.

The Rosenbaum study provides the latest evidence that suspensions do have severe and negative consequences for youth, and particularly for youth of color. And this cost comes without benefit to schools. State and federal officials should provide tools to help educators manage classrooms without disciplinary removals, and support initiatives to reduce discipline disparities. At a time when communities are looking for answers on how to improve school safety, education leaders should look for ways to increase investment in developing supportive approaches that keep children safe and in school.