Racial discipline discrepancies in Louisiana schools

A recent study from the Education Research Alliance at Tulane University uses thirteen years of student-level data from Louisiana to examine differences in suspension rates for black and white students, as well as poor and non-poor students. Overall, it finds that black students are about twice as likely as white students to be suspended and low-income students are about 1.75 times as likely to be suspended as non-low-income students. However, as with previous studies of this topic, it is difficult to know whether (or to what extent) these gaps reflect educator bias, as opposed to differences in behavior or school culture.

According to the authors, suspensions for black students in Louisiana last an average of 0.40 days longer than suspensions for white students who commit the same type of infraction. This difference could be interpreted as evidence of bias (or at least systemic inequity). However, when comparisons are restricted to students in the same school, grade, and year, the difference between black and white students is just 0.10 days. So, at the very least, the first estimate overstates the bias exhibited by individual educators.

Further complicating matters, since suspensions for low-income students are 0.18 days longer than suspensions for their high-income peers, even treating these smaller within-school differences as evidence of bias implies that educators harbor a particular animus against the poor (which seems strange, since poverty is less visible than race). Then again, unlike other studies (which have typically found that differences between schools account for most socioeconomic and racial disparities), this one finds that “within-school differences account for at least 50 percent of the black/white and poor/non-poor gap in kindergarten and grades 5 through 12.” It’s not clear what explains this difference. But assuming it’s real, the implication is that discipline gaps are (at the very least) more “visible” to students and teachers in Louisiana than in other places.

Uncomfortable yet? Well buckle up because we’re just getting started.

So far, most of what I’ve written about this study could probably have been written about any number of discipline studies over the years. Consequently, in an effort to advance the conversation, the authors go on to examine fights between one black student and one white student, which in their view represent “a very particular setting” where disciplinary “disparities most likely would reflect discriminatory school discipline practices” (emphasis added). As the authors acknowledge, even this assumption is contestable. After all, it’s possible that black and white students behaved differently in these fights—and thus warranted different punishments. (Which kid started it, for example?) But, for the sake of argument, let’s assume that the authors have succeeded in putting a lower bound on the racial bias of Louisiana educators.

After controlling for student demographics, the school the students in question attend, and the number of prior fights each was involved in during the year in question (to address the possibility that students who are involved in multiple fights are disciplined differently), the authors estimate that black students receive suspensions that are an average of 0.05 days (or just 1.6 percent) longer than those handed out to the white students. In other words, they receive one additional day of suspension for every twenty interracial fights.

Frankly, if that number accurately quantifies the amount of racial bias in Louisiana, then we should all start talking about something else (like our criminal justice system). After all, only a quarter of black students in Louisiana are suspended in a given year, and less than a third of these suspensions are for violent offenses, so the “average” black student in the state is probably suspended for about one violent offense per lifetime. In other words, if we assume that the bias detected in interracial fights applies to violent offenses in general, then by the end of grade twelve the average black student in Louisiana is suspended for about 0.05 days longer than she should be on account of violent offenses.

Is a student who gets in one fight going to miss out on college because she is suspended for an additional twenty-five minutes? Will the life outcomes of a student who gets in twenty fights be significantly altered by one additional day of missed classes? And what about the life outcomes of his classmates?

Of course, it’s possible that this estimate really is a “lower bound” that understates the racial bias of Louisiana educators. But even if that’s the case, how is the “true” extent of that bias to be disentangled from the other factors that determine suspension rates? And, more to the point, what is the appropriate remedy—other than a less racist society?

SOURCE: Nathan Barrett, Andrew McEachin, Jonathan N. Mills, and Jon Valant, “What are the Sources of School Discipline Disparities by Student Race and Family Income?,” Education Research Alliance at Tulane University (November 2017).

David Griffith
David Griffith is a Senior Research and Policy Associate at the Thomas B. Fordham Institute.