Shelby County v. Holder: Why Section 2 matters

The following contribution to our Shelby County v. Holder symposium comes from Ellen D. Katz, Professor of Law at The University of Michigan Law School.  Her work focuses on voting rights and anti-discrimination law and includes a detailed empirical study of the Voting Rights Act.

Four years ago, when the Supreme Court last considered the constitutionality of Section 5 of the Voting Rights Act (VRA), Justice Kennedy questioned why “[t]he sovereignty of Alabama is less than the sovereign dignity of Michigan,” and why the government of one is “to be trusted less” than the government of the other. Should the Justices now strike down the statute, as many think they are poised to do, the reason why will likely be their belief that places like Alabama are no longer any different from places like Michigan –  or, better yet, Ohio, where Section 5 is wholly inapplicable. Voters may confront difficulties in Alabama, the Justices would posit, but these difficulties appear no worse than those faced by voters in those states left unregulated by Section 5. Therefore, Section 5 must be invalid.  Q.E.D.

Sounds plausible perhaps, but take a closer look. As an initial matter, it is not at all clear that the Court needs to compare covered and non-covered jurisdictions in order to assess the constitutionality of the VRA. The issue presented in Shelby County v. Holder is not whether the Justices think Alabama is worse than Ohio, or even whether Congress might permissibly conclude that it is. Instead, Shelby County presents a different question: whether Congress has the power to extend a remedial regime that everyone agrees it lawfully adopted based on its conclusion that the regime continues to do critical work in the places where it operates. That conclusion should not be suspect, much less invalid, simply because problems have since developed in other jurisdictions that Congress might also appropriately regulate.

The Justices are nevertheless likely to view a comparative inquiry as relevant to the question presented in Shelby County, and with good cause. Congress’s decision to reauthorize the regional provisions of the VRA rests on its belief that the statute remains necessary, and a comparison of covered and non-covered jurisdictions provides one lens through which to assess that decision. Thus, although Shelby County does not require a comparative inquiry, it invites one.

Such an inquiry, however, must take seriously Section 5’s status as an operational statute. The Court, to be sure, has made clear that Congress now needs evidence of rampant unconstitutional conduct in order to adopt new civil rights legislation, but Section 5 is not new.  If the problems that prompted Congress to enact the VRA in the first instance persisted wholly unchanged today, Section 5 should be discarded as ineffective. To require such evidence as a prerequisite to reauthorization (as opposed to initial enactment), therefore makes little sense.  It would allow Section 5 to continue only if the statute had been a failure.

No one thinks the statute has been a failure. What is disputed is the scope of its success. Critics of the VRA claim that conditions have improved in places like Alabama because the problems Section 5 targets have been solved, and decidedly not because Section 5 actively shapes public conduct in covered jurisdictions in significant and productive ways.  In other words, those who insist that similarities between Alabama and Ohio render Section 5 invalid discount and often disregard Section 5’s blocking and deterrent effects.

There is, however, extensive evidence showing that Section 5 significantly shapes governance decisions in covered jurisdictions. Hundreds of proposed changes to election laws have been blocked by the Department of Justice; hundreds more have been withdrawn or altered in the course of Section 5 review; and many more changes were never proposed because local officials knew they would not be Section 5 compliant.

Given these blocking and deterrent effects, we should expect to find fewer instances of discriminatory practices in places subject to Section 5 if, as critics of the VRA contend, places like Alabama are truly no different from places like Ohio. Covered jurisdictions should look markedly better than jurisdictions not subject to Section 5 insofar as public officials nationwide have comparable inclinations to engage in discriminatory practices and confront comparable opportunities calling for such judgments.

And yet, places like Alabama do not look better than places left unregulated by the statute. A study I did with students here at the University of Michigan Law School suggests that voting problems remain more prevalent in places covered by the Act than elsewhere. Our examination of claims brought under the core permanent provision of the VRA – known as “Section 2” – shows that plaintiffs have been more likely to succeed and succeeded more often in covered jurisdictions than in non-covered ones, and that this disparity is even more pronounced when Section 2 challenges were brought against local voting requirements and procedures. We found, moreover, that courts hearing Section 2 claims in covered jurisdictions were more likely to find certain conditions linked to voting discrimination, including things like intentional discrimination, extreme racial polarization in voting, and a lack of success by minority candidates.

The court of appeals in Shelby County called the regional disparity our study identified “particularly dramatic” in light of Section 5’s blocking and deterrent effects. Section 2 and Section 5, of course, are not coextensive, but a large number of electoral practices run afoul of both provisions. Where they do, Section 5’s preclearance requirement typically blocks implementation of the offending practice and eliminates the need for plaintiffs to challenge it under Section 2. Thus, although the precise effect of Section 5 cannot be quantified, the court of appeals was surely correct that it “reduc[es] the need for section 2 litigation in covered jurisdictions.”

But even with this “reduc[ed]” need, Section 2 plaintiffs have been more likely to succeed and, in fact, have succeeded more often in covered jurisdictions than in non-covered ones. Making the disparity all the more dramatic is the fact that covered jurisdictions are home to less than one quarter of the population, and, by a lopsided margin, contain far few local governmental units.

Our Section 2 results also suggest that intentional discrimination may be more pervasive in covered jurisdictions than a cursory comparison of covered and non-covered jurisdictions suggests. As the court of appeals observed, Section 2’s results tests “requires consideration of factors very similar to those used to establish discriminatory intent based circumstantial evidence.” Because courts need not find intent to find a Section 2 violation and have an obligation to avoid resolving constitutional questions when they are able, some adjudicated Section 2 violations capture conduct that is also unconstitutional even though the decisions do not explicitly say so.

Our Section 2 study did not examine claims that were settled or decided without a published decision and accordingly addressed only a portion of the Section 2 claims filed or decided since 1982.  And yet, as the court of appeals found, available data suggests that a fuller accounting of Section 2 litigation would reveal an even greater disparity in successful plaintiff outcomes between covered and non-covered jurisdictions.  Where, moreover, the disparity is less pronounced (as, for instance, it has become over time), the Section 2 data still attests to Section 5’s continued importance.  Even a rough equivalence in outcomes is significant, given Section 5’s blocking and deterrent effects and the disparities in population and relative numbers of political subdivisions.

The question whether places like Alabama are really any different from places like Ohio is sure to occupy the Court’s attention in Shelby County. The answer should be pursued with a clear understanding of what a comparison of covered and non-covered jurisdictions can be expected to yield. Section 5 is an operational statute with significant blocking and deterrent effects. Conditions in covered jurisdictions cannot be examined meaningfully or compared responsibly to those in non-covered regions unless these operational effects are considered. It is willful ignorance to do otherwise.