A week or so ago, I posted on the final version of my article (with Alison Somin) entitled The Department of Education's Obama-Era Initiative on Racial Disparities in School Discipline: Wrong for Students and Teachers, Wrong on the Law.
As the title suggests, the article makes two arguments: (1) The Obama Administration's aggressive application of disparate impact theory to school discipline has been bad for students and teachers; and (2) The policy also goes beyond the scope of the federal government's authority.
The first post was intended to give the reader a taste of the policy arguments. Giving the reader a taste of the legal arguments may take two or three posts (starting with this one).
First, the Obama Department of Education did not even claim that Title VI itself imposes disparate impact liability. In Alexander v. Sandoval (2001), the Court made it clear it does not. It stated, "[I]t is similarly beyond dispute—and no party disagrees—that [Title VI] prohibits only intentional discrimination."
Instead, the Obama Department of Education's argument was based on a couple of Title VI regulations promulgated in 1966. It argued that these regulations impose a general rule against actions that have a disparate impact. My article responds that (1) If the Obama Department of Education was right about the proper interpretation of those regulations, then, by analogy to City of Boerne v. Flores, they were beyond the scope of the powers of the Executive Branch; and (2) As a matter of textual analysis, the Obama Department of Education is wrong about the proper interpretation of those regulations. They were not intended to impose general liability for disparate impact.
I'll get to all that. But in this post, I just want to make a preliminary point: Everything or nearly everything has a disparate impact on some protected group.
It's easiest to make the point with Title VII, since that provision of covers not just race, color, and sex, but also religion and national origin (and there are a lot religions and national origins):
It is no exaggeration to state that there is always some protected group that will do comparatively poorly with any particular job qualification. As a group, men are stronger than women, while women are generally more capable of fine handiwork. Chinese Americans and Korean Americans score higher on standardized math tests and other measures of mathematical ability than most other ethnic groups. Subcontinent Indian Americans are disproportionately more likely to have experience in motel management than Norwegian Americans, who are more likely have experience growing durum wheat. African Americans are [disproportionately represented] in many professional athletics . . . . Unitarians are more likely to have college degrees than Baptists.
Some of the disparities are surprising. Cambodian Americans are disproportionately likely to own or work for doughnut shops and hence are more likely to have experience in that industry when it is called for by an employer. The reasons behind other disparities may be more obvious: Non-Muslims are more likely than Muslims to have an interest in wine and hence develop qualifications necessary to get a job in the winemaking industry, because Muslims tend to be non-drinkers.
I used to offer a $10,000 check to the favorite charity of anyone who could specify a job qualification that actually had excluded some job candidates that would not have a disparate impact on some group. I have never had to pay a penny. Even right-handedness is substantially more common among ethnic groups that consider left-handedness something that children must be trained out of.
Title VI is a little tougher for my point, since it applies only to race, color, and national origin. Nevertheless, there are a lot of national origins out there, so if disparate impact liability were to apply to Title VI through its regulations, it would make an extraordinary range of decisions by funding recipients presumptively a violation:
For example, in the education context, a university that considers the Math SAT score of an applicant for admission gives Korean-Americans and Chinese-Americans an advantage while disadvantaging many other racial and national origin groups. A college that raises its tuition has a disparate impact on Cajun-Americans, Haitian-Americans, and Burmese-Americans, all groups that have below-average median household incomes.
Similarly, a high school that decides to invest in a basketball team rather than a baseball team has a disparate impact on Latinos, who, on average, are shorter than African-Americans and whites, given that height is an indicator of success for male youth basketball players. And if a "Little Beirut" neighborhood is further from a given high school campus than most neighborhoods, and that school decides to build a tennis court where part of the parking lot used to be, the loss of that parking may have a disparate impact on the Lebanese-American students who have to drive to school, as it would any community far from the school campus.
There is no end to it. A university that gives college credit to students who can pass a foreign language exam has a disparate impact on Irish-Americans, Scottish-Americans, and Anglo-Americans, since they are unlikely to have a language other than English spoken in the home. Even a teacher who decides to seat students in alphabetical order will have a disproportionate effect on Chinese-American students. Chinese surnames are more likely to start with W, X, Y, or Z, which would place such students disproportionately toward the back of the classroom.
Two arguments are often made as to why disparate impact liability isn't that terrible. With Title VII, the EEOC takes the position that it doesn't ban (or at least the agency won't take notice of) job qualifications where the protected group qualifies at a rate of at least 80% of the highest group's rate. Tiny disparate impacts don't count. Such a rule could be applied in the Title VI context too (where it would be just as arbitrary). The problem is that disparate impacts that fail the "80%" rule for some group are by no means rare. They are the rule, not the exception. Even with such a rule in place (and it was made up out of thin air) it will still be so that everything or nearly everything has a disparate impact. Since employers and recipients of federal funds never know who will be applying for a job or for admission, etc., it means they will always be treading on thin ice. Government agencies will thus always have the discretion to rule their actions out of bounds. That's an extraordinary level of discretion.
The second argument is that disparate impact liability only makes actions that have a disparate impact presumptively a violation. It is always possible to rebut the presumption by proving "necessity." But do we really want to live in a society in which the actions of recipients of federal funding under Title VI are always (or almost always) subject to a demand by a government agency that they be "justified" by necessity? That's extraordinary--especially for an underlying statute that prohibits only intentional discrimination.
I will write next time about why it's too extraordinary. Such a regulation would be beyond the scope of the Executive Branch's powers.