Empirical Scholarship and the Future of the Voting Rights Act

Over on Rick Hasen's Election Law Blog I've written a post about empirical legal scholarship and the future of the Voting Rights Act.  I thought I'd cross-post the discussion here, but first I should provide a bit of background:

In 2006, Congress reauthorized Section 5 of the Voting Rights Act -- a core provision of the Act that singles out some jurisdictions' voting laws (mostly in the deep South) for intensive federal oversight.  Section 5, which was initially a enacted as a temporary provision, was designed to provide a powerful institutional mechanism to prevent discrimination against minority voters.  But unsurprisingly, the jurisdictions subject to Section 5 were extremely unhappy to be singled out.  They challenged Section 5 when it was first enacted, and they've done so each time the temporary provision has been reauthorized.  Shortly after Section 5 was reauthorized in 2006, its constitutionality was challenged in Northwest Austin Municipal Utility District Number One v. Mukasey ("NAMUDNO").

Two weeks ago, the special three-judge court hearing NAMUDNO upheld Secton 5's constitutionality.  Judge Tatel authored the decision of the court, and in affirming Congress's authority to pass Section 5 he relied extensively on evidence from lawsuits brought under Section 2 of the Voting Rights Act.  Section 2 is the other core provision of the VRA; it provides a private right of action to attack discrimination against minority voters.  Judge Tatel's reliance in this constitutional case on empirical evidence from Section 2 litigation raises a number of important questions, and it was these questions that I addressed over on the Election Law Blog.  Here's the post:

What can litigation under Section 2 of the Voting Rights Act tell us about NAMUDNO and the constitutionality of Section 5?  Over on Columbia Law Review’s Sidebar, Tom Miles and I have a colloquy with Ellen Katz and Anna Baldwin about our paper Judging the Voting Rights Act. While that paper is concerned principally with the role that a judge’s race and partisanship play in litigation under Section 2 of the VRA, our colloquy raises the question of what we can learn about Section 5 from Section 2 litigation.  With the decision in NAMUDNO less than two weeks old, Rick invited me to comment on how our work bears on the constitutionality of Section 5 of the Voting Rights Act and the prospects for NAMUDNO in the Supreme Court.  Here are a few quick thoughts.

NAMUDNO raises difficult questions about Congress’s power to enforce the guarantees of the Fourteenth and Fifteenth Amendments.  Many others have written elegantly and at length about these doctrinal questions, so I won’t rehash the issues here.  But underlying these doctrinal questions is a basic choice courts must make about the type of evidence they require from Congress.  First, courts can require that Congress produce evidence that intentional discrimination remains a reality in the areas covered by Section 5’s oversight mechanism. (Call this the “things are still bad there” test.)  Second, courts could demand comparative evidence that intentional discrimination is a more significant problem in covered than non-covered jurisdictions.  (Call this the “things are worse there than elsewhere” test.)

These are vastly different evidentiary requirements.  Which one the Supreme Court chooses to emphasize may well be dispositive in NAMUDNO. If it focuses principally on the first, then Section 2 litigation can help build the case for Section 5’s constitutionality by providing anecdotal evidence of intentional discrimination.  But Section 5 is likely in trouble if the Supreme Court demands systematic empirical evidence that the boundaries of the coverage formula are neither over- nor under-inclusive – that is, that intentional discrimination is more of a problem in each jurisdiction covered by Section 5 than in uncovered jurisdictions.  For as Tom and I explain in our Sidebar colloquy, Section 2 litigation cannot provide evidence that discrimination is worse in covered than uncovered jurisdictions.

In the lower court decision in NAMUDNO, Judge Tatel focused almost exclusively on the much easier evidentiary question whether there is evidence of intentional discrimination in some covered jurisdictions. He drew on a wide variety of evidence, including the important evidence gathered by Ellen Katz.  By scouring every published Section 2 decision, she was able to identify each instance in which a court made findings of intentional discrimination in the course of concluding that a government had violated Section 2.  Judge Tatel discussed those findings at length, concluding that “Congress thus knew of a combined total of fourteen judicial findings of intentionally discriminatory or unconstitutional state action across six covered states.”  Tatel’s approach is consistent with prior cases evaluating Congress’s power to enforce the Reconstruction Amendments.  Those cases have never – despite Boerne v. Flores’s doctrinal formulation – required systematic empirical evidence concerning the over- and under-inclusion of Congress’s approach.

That said, near the end of his decision Judge Tatel does include a paragraph in which he  attempts to use Section 2 litigation to show that there are “significant differences between covered and noncovered jurisdictions.”  To support the conclusion that things are worse in covered jurisdictions, he relies on the fact that, “from 1982 to 2005 the success rate of section 2 suits in covered jurisdictions exceeded the success rate of such litigation elsewhere.”  But for the reasons that Tom and I explain in our Sidebar colloquy with Ellen, the Section 2 data do not support the conclusion that things are worse today in the South.

Our full explanation of why the data don’t support this conclusion is here, but there are three basic reasons.  First, while a summary comparison of lawsuits in covered and uncovered jurisdictions does suggest a disparity in success rates across covered and uncovered jurisdictions, this difference disappears when we run regression analysis to control for factors other than coverage itself.

Second, and most fundamentally, it is a mistake to interpret differences in litigation success rates across jurisdictions as evidence of differences in the underlying levels of discrimination. The central methodological difficulty with drawing inferences about the extent of discrimination from litigated Section 2 cases is that the sample of cases is almost surely not representative of the entire class of voting rights claims.  This is a problem that is so pervasive that it has a label among empirical scholars – it is known as the “Priest-Klein” problem.  And fancy monikers aside, it is the intuitive reason why no one thinks that the
success rates in tort lawsuits in Illinois are a solid source of information about the level of tortious conduct in the state, and why no one thinks that conviction rates in criminal cases in New York are a meaningful measure of the state’s crime level.

Third, even were courts to ignore these selection problems, the Section 2 statistics Judge Tatel quotes are problematic because they mask a trend.  He describes a difference between covered and uncovered jurisdictions over the two decade period of Ellen’s initial study.  But an average difference over two decades does not fit the conclusion that a difference justifying Section 5’s reauthorization remains today. And, in fact, the Section 2 data reveal that success rates in covered and uncovered jurisdictions have converged over time.  In the 1980s and early 1990s, courts in covered jurisdictions were indeed more likely to find liability than courts in uncovered areas.  But from 1994 to the end of the study in 2004, plaintiff success rates were nearly identical in covered and uncovered areas.  If one really believed that success rates were a measure of the underlying discrimination, then one would have to conclude from this convergence that discrimination today is no worse in covered than uncovered jurisdictions.

Tom and I disagree with that conclusion because we don’t believe these success rates are a good measure of discrimination.  That said, it suggests that – both as a matter of theory and strategy – Section 5 will be ill-served if we try to aggressively wield data from Section 2 lawsuits in the service of conclusions that the data simply cannot support.

Comments

[FrankMCook]

I claim no knowledge or experience in this area, but if Congress had created an agency with the power to impose additional duties in geographic areas where the agency determined there was need to do so, wouldn't the Courts give great deference to the agency's findings? Are we going to second guess Congress to a greater extent if it chooses to make the determination directly?

[Kimball Corson]

An implicit question here is, in the absence of "wholly sufficient" empirical data, must it be that no action should be taken when the consensus judgment of those informed and willing to address the issue clearly agree there is a problem? Data can always be better, available in larger quantities and come from better sources. Where do we draw the line here, especially if a constitutional issue is involved?