The Center for Equal Opportunity on “Disparate Impact” at Cato 

Roger CleggDisparate Impact

The libertarian Cato Institute was kind enough to ask me to contribute an article to its annual Cato Supreme Court Review, inviting me to write on the Supreme Court’s recent (and unfortunate) decision to allow “disparate impact” causes of action under the Fair Housing Act. 

So I thought I would excerpt some of that article for this week’s email, and have done so below.  You can read the full article here

I should also note that last week I spoke at Cato about the case on a civil-rights panel.  You can watch the event here (“Panel II:  Civil Rights”); I’m introduced starting at the 0:22:10 mark, and my discussion runs from 0:23:40 through 0:40:25. 

*          *          *

Introduction

In Texas Department of Housing and Community Affairs v. Inclusive Communities Project, the Supreme Court at last resolved the issue of whether “disparate impact” causes of action may be brought under the Fair Housing Act, which was first passed in 1968 and then substantially amended and expanded in 1988. In brief, disparate impact cases are based on an inference of illegal discrimination if a defendant’s actions have a disproportionate adverse effect on individuals in a protected class, such as race. By contrast, disparate treatment cases, which are indisputably covered by the Act, are triggered when a defendant’s actions arise out of a discriminatory motive or intent.
… Justice Anthony Kennedy wrote the 5-4 majority opinion in Inclusive Communities, in which he was joined by Justices Ruth Bader Ginsburg, Stephen Breyer, Sonia Sotomayor, and Elena Kagan. Justice Alito’s dissent was joined by the remaining justices; Justice Clarence Thomas also wrote a separate dissenting opinion …

The Court’s decision is disappointing. It fails to follow the clear language of the statute and will not only result in unfair liability for many defendants, but will encourage race-based decisionmaking in the housing area—exactly what the Fair Housing Act was meant to prohibit. The only silver linings are that Justice Kennedy’s opinion itself recognizes these problems, and some of the language toward the end might be useful in stemming the worst abuses.

To elaborate: The question presented in this case was, “Are disparate-impact claims cognizable under the Fair Housing Act?” Under a disparate-impact claim, discriminatory motive is irrelevant: It need not be alleged or proved, and it doesn’t even matter if the defendant proves that there was no actual disparate treatment. If a policy or procedure results in a disproportion of some sort—on the basis not only of race, color, or national origin but also (under the FHA) of religion, sex, or familial status (that is, having children)—then that’s enough, even if the policy is nondiscriminatory by its terms, in its intent, and in its application. The defendant can prevail only by showing—to the satisfaction of a judge or jury who may know or care nothing of the defendant’s needs—some degree of “necessity” for the policy. This numbers-driven, we-don’t-much-care-about-your-reasons approach inevitably results in pushing potential defendants away from perfectly legitimate and race-neutral policies and toward race-based decisionmaking: again, just the opposite of what civil-rights laws are supposed to do.

This article will begin by summarizing the various opinions in the case, and will then explain some of the problems with the disparate-impact approach, both generally and with respect to housing discrimination in particular. It will then discuss what might be done to address these problems in the future, through litigation and through legislation.

Problems with the Disparate-Impact Approach

 In General
As noted, under a disparate-impact claim of discrimination, discriminatory motive is irrelevant: It need not be alleged or proved, and it doesn’t even matter if the defendant proves that there was no discriminatory motive. If a policy or procedure results in a disproportion of some sort, then that’s enough, even if the policy is nondiscriminatory by its terms, in its intent, and in its application. The defendant can prevail only by showing—to the satisfaction of a judge or jury who may know or care nothing of the defendant’s needs—some degree of “necessity” for the policy.

Now, suppose that you are a potential defendant and that you have some non-discriminatory selection criterion that has helped you run your business well, but the criterion has a disparate impact on some group. You know you are vulnerable to a lawsuit, which you may or may not win, depending on the judge or jury you draw, and you know that lawsuits are expensive, win or lose. If you don’t want to get sued—and who does?—the potential of a disparate-impact lawsuit is going to push you to do one of several things, none of which is good. You might keep the criterion but apply it in a way that gets your numbers right: In other words, you will adopt surreptitious quotas. Or you might get rid of the criterion altogether, and just accept the fact that your business will not be run quite as well as it could be. Or you might decide to replace the old criterion with a new one, which you will choose and/or apply in a race-conscious way. You might, that is, now choose a criterion because of the racial outcomes that will result, or choose some criterion that can be applied in a biased way so that the resulting racial double standard will ensure that the numbers come out right. No matter what, you are no longer using the criterion you freely chose because you thought it to be the best, but are instead weighing race—directly or indirectly—in what you do. 

In other words, we’re supposed to stop judging people by the content of their character, and start judging them by the color of their skin. In addition to this moral dilemma, there is this overwhelming practical one: There is probably no selection or sorting criterion that does not have a disparate impact on some group or subgroup.

And here’s the most fundamental point of all: If a business, agency, or school has standards for hiring, promoting, admissions, or offering a mortgage that aren’t being met by individuals in some racial or ethnic groups, there are three things that can be done. First, the standards can be relaxed for those groups. That’s what racial preferences do. Second, the government or aggrieved private party can attack the standards themselves. That’s what the disparate-impact approach to enforcement does. Third, one can examine the underlying reason why a disproportionate number of individuals in some groups aren’t meeting the standards—such as failing public schools or being born out of wedlock—and do something about that. But this option holds little interest on the political left.

Speaking of which, the Obama administration has made no secret of its love for disparate-impact civil-rights enforcement, and has been aggressive in applying it to every imaginable situation. In employment, for example, the government complains if fire or police departments administer physical or written tests that have politically incorrect results, or if companies use criminal background checks; in voting, it objects if voter ID is required; in education, it is hostile to school discipline policies if they have a disproportionate racial or ethnic result; it has even insisted on drawing distinctions between acceptable and unacceptable pollution, depending on the skin color and national origin of those affected by the pollution. The disparate-impact approach is also employed to require the use of a foreign language—on driver’s license exams, for example—on the theory that using only English might have a disproportionate effect on the basis of national origin. And it has been used to pressure banks with regard to their lending requirements, even though many believe this to have been a contributing cause of the mortgage meltdown and the following recession.

III. Going Forward

Litigation

While the Supreme Court’s ruling here is misguided, potential litigants should not lose sight of this counterintuitive fact: The law is actually better now than it was before Justice Kennedy wrote the opinion.

This is true partly because the bar was so low: All the courts of appeals to entertain this issue had adopted this approach, too, and the Obama administration and its allies in the civil rights establishment were already interpreting the law this way. So things could not have gotten a lot worse, no matter what the Court had done.

It is also true, however, that the law is now better because Justice Kennedy’s opinion recognizes that the disparate-impact approach can lead to very bad results. … [T]he Court has now set some limits on the law that will be useful. For example, Kennedy warns the lower courts against “second-guess[ing]” the nondiscriminatory reasons for challenged policies, requires a “robust causality requirement” rather than relying simply on racial disproportions, recognizes that “racial quotas” and “racial considerations” and “abusive . . . claims” can result from threatened and actual lawsuits, cautions that any “remedial orders must be consistent with the Constitution.” He all but says that he expects the plaintiffs to lose in this case. He even calls Justice Alito’s dissent, which of course makes similar points, “well-stated.”

Given that the Court was unanimous, then, in recognizing the constitutional problems and bad policy results that can arise from the disparate-impact approach, litigators should continue to press courts to reject or at least limit the approach.

For example, the door is still open for courts to reject disparate impact use under the Equal Credit Opportunity Act, to limit it under Section 2 of the Voting Rights Act, and to strike down disparate-impact regulations that have been promulgated under Title VI of the 1964 Civil Rights Act. (Those regulations have been used, for example, to challenge school discipline, policing policies, and English-language requirements where they have a disproportionate effect on this-or-that racial or ethnic group.)

Legislation

While much can be accomplished through litigation in stemming the abuses of the disparate-impact approach to civil rights enforcement, ultimately there is no substitution for action by Congress, which ought now to amend the Fair Housing Act. And, while at it, Congress should clarify that, in other contexts as well, the disparate-impact approach is invalid.  Most civil rights laws have no “disparate impact” provisions—rather, they prohibit actual disparate treatment—but they have been expanded to include disparate impact through agency interpretation and unwarranted court rulings. The FHA is, of course, a case in point. Thus, Congress should make clear that laws prohibiting discrimination do not extend to mere disparate impact. Legislation has been drafted to do just that, and includes the FHA and a number of other statutes.

Conclusion

The disparate-impact approach to civil-rights enforcement is untenable as a matter of law and policy. It second-guesses nondiscriminatory selection criteria and encourages race-based decisionmaking. 
Those are disturbing abuses of federal power at the expense of liberty and limited federal government. As a general matter, the presumption should be that the decisions of private, state, and local actors are no business of the federal government; an exception can be made in extraordinary circumstances of racial discrimination, but the disparate-impact approach is used precisely when discrimination has not been shown. And the problem is compounded here since it will be the federal government that is encouraging discrimination.

While Justice Kennedy’s opinion for the Court in Texas Department of Housing and Community Affairs v. Inclusive Communities Project unfortunately now allows this approach under the Fair Housing Act, it recognizes the problems with it, leaving the door open to future litigation that limits this approach under that statute, as well as to litigation that challenges or limits the approach under other statutes. Instead of leaving this matter to the courts and the uncertain course of future litigation, however, Congress should act to preclude or at least limit the disparate-impact approach.