The National Review Online decried new federal guidelines that could reduce the number of needless arrests and incarceration of minority students in public schools.

On January 8, the Department of Justice (DOJ) and the Department of Education (DOE) issued new, optional guidelines to help public schools develop non-discriminatory disciplinary policies. Right-wing media were quick to accuse the Obama administration of  playing the "race card" because the guidelines addressed the fact that minority students are far more likely to be disciplined -- often unfairly and excessively -- for nonviolent and minor disruptions in school. Because more and more schools rely on armed police officers known as "school resource officers" to handle behavioral problems, many students of color end up getting arrested and incarcerated.

NRO has previously called the new DOJ guidelines "disturbing." But in a January 16 editorial, the site went further, complaining that the guidelines were an overblown response to “spectral racism” and were based on “arbitrary evidence” (emphasis added):

The Obama administration is no stranger to trying to micromanage complex, intractable problems from Washington. But using the Civil Rights Act to direct schools' disciplinary practices might be its most foolhardy idea yet. Beginning in 2010, the Department of Education, led by the occasionally sensible Arne Duncan, announced that it intended to pursue vigorously civil-rights violations in the American school system. That's led to a number of DOE investigations of various school districts with racially disparate discipline rates.

[...]

The feds contend, as an aside, that discrimination in discipline shows up in studies when controlling for poverty and other factors, but the evidence for this contention is ludicrously weak. Federal civil-rights investigators don't have to publicly disclose the grounds they've used to initiate investigations of racial discrimination, but their work so far leans as heavily as the new guidelines do on evidence of disparate statistical impact, rather than on indications of real bias and disparate treatment. They will not admit that they rely on such arbitrary evidence, since there is little statutory justification in the Civil Rights Act for such a disparate-impact case, but the objection is clear enough: Certain minorities are disciplined at higher rates than whites are, so racism must be at work.

When such a simple heuristic is applied, schools will feel even more pressure than they already do to adopt a simple solution: try to discipline all races, regardless of behavior, at the same rate. This might mean arbitrarily increasing rates of punishment for whites or, much more likely, reducing them for blacks and Hispanics, disadvantaging their classmates of all races who'd like peaceful classrooms.

[...]

No one should be surprised by the Obama administration's zeal for alleging racial discrimination when it isn't there ... But it is still shocking that the federal government is effectively encouraging schools to judge students on the color of their skin rather than the content of their character.

The entire point of the DOJ's guidelines is to encourage schools to stop mistreating students based on the color of their skin, so it's odd that NRO would conclude the exact opposite.

Moreover, NRO's claim that the government's new guidelines aren't in response to “real bias” flies in the face of statistical evidence. As the DOJ pointed out, “over 50% of students who were involved in school-related arrests or referred to law enforcement are Hispanic or African-American.” And according to a recent report from the American Civil Liberties Union, those minority students are disproportionately more likely to be punished with suspension, expulsion, or arrest even if they're exhibiting the exact same disruptive behavior as their white classmates.

This “arbitrary evidence,” as NRO calls it, represents real students who may have been unfairly discriminated against by school administrators on the basis of race. For NRO's editorial board to dismiss these facts as some sort of widespread statistical anomaly is unfortunate.