If the Obama Administration’s 2014 “Dear Colleague” letter on school discipline had been just another one of the zillion “guidance” documents sent by federal agencies to school systems every year, it would not deserve much concern. Of course it’s fine to remind schools that, under federal civil rights laws, they cannot treat students of various races or other protected groups differently when meting out punishments for misbehavior. It’s also fine to suggest alternative approaches that educators might try to create healthier school climates and keep more children in class.
But this wasn’t just routine guidance. Instead, by applying a shambolic version of disparate impact theory to school discipline, the letter marked an enormous shift in federal policy and set up the Office for Civil Rights (OCR) to be both judge and jury. Most disconcertingly, it was bound to have a chilling effect on discipline in some of America’s toughest schools, with perverse consequences for classroom order, learning, and student safety.
Confusing Cause and Effect
I contend that school discipline reformers in general, and the Obama team in particular, were cavalier in calling on U.S. schools to cut the use of exclusionary discipline without identifying major resources to help them do so. From the administration’s perspective, though, time was of the essence, both because excessive suspensions and the like were causing material damage to affected students and because racial disparities in discipline represented a violation of students’ civil rights. As Dan Losen of UCLA’s Civil Rights Project told The Atlantic, “I don’t think you can go too fast if you are trying to remedy an injustice.”
These sentiments are undeniably heartfelt, but are they grounded in reality? First, consider the claim that out-of-school suspensions cause negative student outcomes. The 2014 letter declares that “Studies have suggested a correlation between exclusionary discipline policies and practices and an array of serious educational, economic, and social problems, including school avoidance. . . decreased academic achievement. . . increased likelihood of dropping out; substance abuse; and involvement with juvenile justice systems.”
But as any undergraduate knows, correlation is not causation. How can we know whether it’s the discipline that students received or their own behaviors that make them more likely to experience negative long-term outcomes, including prison? Nobody can randomly assign students to break school rules and experiment with different disciplinary responses in order to test their long-term impacts. And there’s good reason to believe that students who get into serious trouble at school are likely to get into trouble with the law as adults, regardless of whether they are suspended or expelled. The argument that a single suspension—especially a brief one—sparks a chain reaction that leads to a life of crime strains credulity. Yet it’s taken as gospel by many discipline reformers.
The civil rights concerns, meanwhile, arise from the stark racial disparities in school discipline rates, with African American students about three times more likely than whites to be suspended. Reformers say this is mostly due to educator bias, if not outright discrimination. As U.S. Secretary of Education Arne Duncan said when the 2014 letter was released, “racial discrimination is a major problem today,” and disparities are “not caused by differences in children.” But again, does the evidence show that to be the case?
Yes, some studies have found evidence of bias among educators—as among everyone—but the impact on discipline rates appears to be quite small. For example, a recent study from Louisiana found that black students who were in fights with white students received longer suspensions—but the differences were tiny. Meanwhile, that same study found that black students were four times as likely to be reported for getting into fights as whites. It is totally appropriate for OCR to address bias where it exists, but it’s extremely unlikely that it can explain the bulk of the suspensions gap.
Solid evidence, meanwhile, indicates that student behavior does in fact vary across different subgroups, often quite significantly—not because of race, but because of the vastly different circumstances that children face. Those growing up in poverty are more likely to experience trauma, to live without their fathers, to go home to more violent neighborhoods, and to otherwise face situations that make it more likely that they may misbehave at school. As African Americans are three times more likely to grow up in poverty, we would expect to see racial differences in student behavior, just as we see racial differences in achievement—which are not driven by race, but by socioeconomic differences. And the most recent evidence, again including the study from Louisiana, indicates that such factors can explain much of the racial discipline gap.
Students themselves acknowledge differences in behavior. In 2015, high school students were asked by the federal Centers for Disease Control if they had been in a fight on school property at any time in the past 12 months. African American students were 2.2 times more likely to say “yes” than white students. Meanwhile, in a 2009 survey of a nationally representative sample of 9th grade students by the National Center for Education Statistics, 17 percent of African American students reported going to class late “sometimes,” versus 10 percent of white students. And 4.1 percent of black students reported going to class late “often,” versus 2.2 percent of white students (See Figure 1).
So according to the kids themselves, compared with white students, African American pupils are more than twice as likely to get into fights at school and almost twice as likely to get to class late.
If these large differences in behavior are mirrored in other types of infractions too—for example, students cursing at their teachers—then we should expect to see large differences in suspension rates, even in schools where educator bias is negligible. It’s hard to see how this equates to violating students’ civil rights.
Disparate Impact Theory Meets School Discipline
The 2014 letter, then, was based on dubious claims about the effects of school discipline and highly questionable assumptions about what is driving racial discipline disparities. In other words, it misdiagnosed the problem. What’s even worse, though, was its supposed cure: disparate impact theory.
This first arose in labor law. The classic example comes from firefighting. For decades, many urban fire departments were the exclusive province of whites, thanks to explicit discrimination against people of color. Once civil rights statutes banned such de jure discrimination, some fire departments came up with ways to maintain the discriminatory status quo, such as requiring applicants to pass standardized tests before being hired, tests that African American and Latino applicants tended to fail. Agencies and courts turned to disparate impact theory to probe practices like those, which may have been “facially neutral” but were intended to have, and did have, discriminatory effects.
The 2014 letter brought a similar approach to bear on school discipline. If districts’ data showed that children of color were being suspended or otherwise sanctioned at higher rates than their peers, that alone would demonstrate that schools’ policies were having an “adverse impact” on protected groups and could trigger an OCR investigation. Schools would then have to justify their policies by proving to investigators that they were “necessary to meet an important educational goal” and that there was no “comparatively effective alternative policies or practices” that they could use instead.
Now imagine that you run a school system with pupil demographics that reflect national averages—call it Springfield Public Schools. In your district, African American students are three times more likely to live in poverty than white students and more than twice as likely to get into fights at school. According to the 2014 letter, if you suspend African American students at twice the rate of white students, you can be subject to an OCR investigation—even if those suspensions are completely justified by student behavior.
Furthermore, when federal investigators come knocking, you must demonstrate to them that (a) your discipline policy serves an educational purpose and (b) there’s no better approach that could have a less adverse impact on students. Otherwise you’re guilty of civil rights violations—even if you had no intent of discriminating against students. (Losen and I agree on this point: The 2014 letter did not require OCR to find evidence of discriminatory intent. That’s at odds with how disparate impact theory works in other domains, and is a major problem in my view.)
This is a breathtaking expansion of federal power into one of the most sensitive aspects of American schooling, and substitutes the judgment of federal bureaucrats for local educators and officials. And it makes a mockery of disparate impact analysis.
From “Suspension Factories” to “School Climate Catastrophes”
There’s another driver of discipline disparities that we must consider, beyond educator bias and differences in student behavior: the school-to-school variation in suspension rates. Evidence from Arkansas and elsewhere indicates that the discipline disparities found at the district level are often driven by sky-high suspension rates in a handful of high-poverty schools. Just as relatively few U.S. high schools are responsible for the majority of our dropouts, it may also be the case that relatively few schools are responsible for our discipline disparities.
It’s hard to know exactly why. It could simply be the challenge of concentrated poverty: high-poverty schools serve more students who have experienced trauma, and thus are more likely to get into trouble at school. It’s surely harder for schools to handle that constructively when such youngsters are more numerous. It may also be that these are simply bad schools, with weak leadership, low-quality teaching, and a broken climate, and their high suspension rates are an indicator of campuses that are barely keeping it together. It could be, in fact, that the school-discipline-to-prison pipeline is really the bad-school-to-prison pipeline. In other words, the correlations that researchers are detecting—including in a brand-new study by Janet Rosenbaum—may stem from attending high-suspension schools, not from suspensions per se.
Either way, fixing or replacing these “suspension factories” is a herculean task, requiring significant political will, expertise, and resources. Alas, with its 2014 policy change, the Obama Administration offered none of that. It just implicitly told these schools to stop suspending so many students. And now, not surprisingly, it appears that this push to reduce exclusionary discipline, and similar district-level reforms that preceded it, may be making some of these schools less orderly and safe.
Up-to-date data about school safety and climate are scarce, but the Manhattan Institute’s Max Eden has been doggedly searching for it, and what he’s found does not present a pretty picture. Take New York City, which adopted discipline-policy reforms during 2014-15 that required principals to ask the central office before suspending students. Just a year later, teachers reported less order and discipline in the schools, and students reported more violence, gang activity, and drug and alcohol abuse. The schools with the highest percentage of students of color experienced the most significant declines in school climate. Or Los Angeles, where a Boston University doctoral candidate found significant declines in achievement in high-poverty schools after a ban on suspensions for low-level infractions was put in place. Or Chicago, where a 2015 study found that changes to its discipline code were associated with a decline in student and teacher perceptions of school safety. Eden is uncovering similar patterns in other large districts, which he’s characterized as a “school climate catastrophe.”
A study by the University of Pennsylvania’s Matthew Steinberg and Mathematica’s Johanna Lacoe, and published by the Thomas B. Fordham Institute, also found a differential response to school discipline reforms in Philadelphia, with high-poverty schools mostly ignoring the district’s new policy and/or suspending even more students for serious infractions. Lower-poverty schools, on the other hand, mostly complied, and did so without any ill effects.
Perhaps it is possible for large, high-poverty schools to reduce suspensions without increasing disorder or violence—but it would take strong leadership, engaging teaching, ample mental health supports, and much else. The problem is that many such schools lack these essential ingredients, which may be why they resort to suspending so many students in the first place. If all we do is tell such schools that they can no longer suspend students, we oughtn’t be surprised when dysfunctional campuses become disorderly or worse. And while discipline reformers acknowledge the need for “more support” for such schools, they haven’t explained where that’s supposed to come from—certainly not from OCR.
To be clear, students and their families must retain the right to file complaints with federal officials if they feel they have been subject to discrimination. Investigators should examine the facts of their cases and come to appropriate judgments. Where bias and discrimination impact school discipline, it’s OCR’s job to stop it. Keeping the parts of the 2014 letter that remind districts that they cannot treat students differently would be appropriate.
It might also be reasonable for OCR to investigate districts that are real outliers in terms of discipline disparities—systems where the suspensions gap is much higher than socio-economics might explain. Such districts should be held to account—but only if OCR can identify compelling evidence that their officials intended to discriminate against certain groups of students.
But the heart of the discipline problem is the few thousand “suspension factories” that face the greatest challenges. Fixing these schools is beyond the capacity of a federal agency. It’s going to take concerted effort by states and districts, and lots of money and political will, to turn around or replace these failing, and flailing, schools. Simply telling them to suspend fewer kids could be making them even worse, and may be putting other children—innocent children—in harm’s way. The 2014 letter must be significantly revised before it does further damage.
Michael J. Petrilli is president of the Thomas B. Fordham Institute and an executive editor at Education Next.