Tuesday, January 15, 2013

Race, Politics and the Voting Rights Act

by Mike Dorf

I have sometimes criticized Jeff Toobin's New Yorker articles as dumbing down the law too much in order to reach a broad audience (e.g., here), so I feel some obligation to say that his latest New Yorker piece--on the Voting Rights Act case of Shelby County v. Holder, which will be argued in the SCOTUS next month--is excellent. Toobin argues that in light of the GOP's efforts to disenfranchise likely Democratic voters in November's election, it is ironic bordering on perverse that the Court is choosing this historical moment to consider whether the VRA (or, to be more precise, the 2006 reauthorization of the Section 5 preclearance mechanism of the VRA) is unconstitutional as beyond the powers of Congress.

As Toobin concedes--and as I would concede as well--contemporary disenfranchisement is rooted more in politics than in the racism of the Jim Crow period.  Republicans don't want African Americans (or non-Cuban Latinos) to vote because they generally vote for Democrats, not because Republicans (necessarily) harbor racist views about African Americans (or other minority voters).  Republicans are happy trying to disenfranchise Democratic-leaning affluent white college students too.

But that shouldn't matter because the touchstone of forbidden race discrimination is the use of racial categories, not the purpose for which those categories are being used. The Court's affirmative action cases make that clear. A state university or employer that expressly employs race as a means of advancing a benign objective--like redressing past discrimination or increasing racial diversity--must satisfy strict scrutiny, even though the use of race is not rooted in animus towards any racial group. Likewise here. Even assuming that suppression of Democratic votes is a legitimate government objective (about which more, momentarily), pursuing it through race-conscious means would violate the Fourteenth and Fifteenth Amendments--and thus Congress acts within its authority to enforce those amendments when it requires special procedures for scrutinizing laws that may be restricting the vote based on race.

To be sure, none of the laws being used these days is expressly race-based, but that was true under Jim Crow as well. Just as back then, a literacy test or a poll tax would have been imposed in Alabama for the obvious purpose of disenfranchising African Americans--and was therefore a sufficient predicate for the original VRA--so too today, the adoption of Voter ID laws for the obvious purpose of exploiting their disparate impact on poor minority voters (who tend to vote for Democrats) should count as a sufficient predicate for the continued validity of the VRA.

Moreover, it's not even clear that the argument for the VRA's continuing validity rests on the conclusion that today's voter suppression efforts are tacitly race-based. In Vieth v. Jubelirer the Supreme Court held that complaints about political gerrymandering are not justiciable, although a concurrence in the judgment by Justice Kennedy, providing the necessary fifth vote, left open the possibility that in some future case the Court might be amenable to invalidating gerrymandering if sufficiently determinate judicial criteria could be developed. As I read the case, it does not stand for the proposition that complaints about voting procedures are non-justiciable (a position that might be attributable to Justice Frankfurter a couple of generations ago) but rather that the nature of the districting process is so inherently political that the courts cannot establish judicially manageable standards. Vieth thus leaves open the possibility that other efforts to shut out or diminish the political impact of particular voting blocs--even if for political rather than ultimately racial reasons--could give rise to a justiciable claim.

Moreover, even if Vieth were read broadly to express judicial reluctance to enter the fray of regulating elections, it would only reflect the Court's skepticism of its own ability to develop manageable standards. In the Voting Rights Act, Congress has provided the standards.

The foregoing argument responds to justiciability objections rooted in the separation of powers but what about federalism? Where does Congress get the affirmative authority to regulate the voting process? The answer--quite apart from any interest in tamping down race discrimination via the Fifteenth Amendment--is Section 5 of the Fourteenth Amendment. Although the Supreme Court's Section 5 cases are arguably somewhat contradictory, the closest thing to a synthetic principle that emerges from them is this: Where Congress seeks to remedy or prevent violations of rights that are protected by heightened scrutiny under the judicially developed doctrines, the Court will give substantial deference to Congress. That principle explains the results in the Hibbs case (where the semi-suspectness of sex discrimination gives rise to heightened scrutiny) and the Lane case (where the fundamental right to court access gives rise to heightened scrutiny).

Thus, because the right to vote is "fundamental" within the Court's cases, Congress has fairly wide latitude to enact preventive and remedial measures. If the Court applies its existing Section 5 jurisprudence, it therefore ought to uphold the VRA. But don't count on it. In the NAMUDNO case, the conservative Justices pretty clearly signaled that they're itching to invalidate the VRA.


Joe said...

Given Prof. Dorf's column on the case, off topic, it would be interesting to get his response to today's USSC ruling. I appreciate btw that he reads his columns over at Verdict and wish others would as well. Seems like a good use of research assistants or students or the like.

Hashim said...


Two problems with your analysis:

First, your casual attribution of implicit racism to facially race-neutral laws is unwarranted here (or at least unsupported). Even assuming, as you do, that voter-id laws are not legitimately motivated by a desire to eliminate any potential for fraud, what basis do you have for further claiming that the GOP's motive is to disenfranchise poor *minorities* specifically, as opposed to poor people *more generally*? After all, in the States where voter-ID has been passed, I suspect that the latter group is *also* disproportionately Democratic. Unlike with ancient literacy tests, whose application was specifically manipulated to ensure that even illterate whites still passed, nothing in the design or application of modern voter-ID laws uniquely targets poor people who specifically happen to be minorities, by somehow exempting poor whites. Indeed, I bet this is empirically provable, given that these laws have been passed by numerous states across the Nation, which likely have a huge variance in the size of the affected minority population. In short, at worst, the GOP's intent is to discriminate against poor people in the relevant states because they're disproportionately Democrats, and the fact that that partisan intent may have a disparate impact on minorities does not violate minorities' constitutional rights under the 14A or 15A. See Washington v. Davis; Bossier Parish.

Second, even assuming that Congress has power under Section 5 of the 14A to prevent partisan manipulation of the franchise wholly apart from any racial intent (though I disagree, given the implications of Vieth under Boerne), the preclearance regime's *coverage formula* can't possibly be justified w/r/t partisan discrimination. After all, there's no evidence whatsoever that the covered jurisdictions are in any way worse than the non-covered jurisdictions when it comes to partisan discrimination, rather than racial discrimination, which is unsurprising, since Section 5 has always been about racial discrimination. There's simply no justification for allowing states like Indiana, Pennsylvania, and Wisconsin to engage in partisan discrimination free from preclearance, but to subject the covered jurisdictions to intrusive federal oversight--especially since no one has ever successfully brought a federal lawsuit against voter-id laws in the non-covered jurisdictions. In short, Section 5's coverage formula is demonstrably irrational with respect to partisan discrimination. (It's also irrational with respect to the preclearance standard, which focuses on the rights of racial minorities, rather than the rights of partisan minorities--e.g., Section 5 wouldn't stop Democrats from discriminating against white Republicans.)


Paul Scott said...

This is nice and all, but what does it have to do with big coins?

Michael C. Dorf said...

Here a couple of thoughts for Joe and Paul, respectively, then a separate comment for Hash.

1) Joe refers to the Lozman case, in which the Court today ruled that the object at issue was not a "vessel." I think that's a fair conclusion and one that might have been anticipated in light of the difficulty that the respondent had in offering a limiting principle. However, I'm disappointed that the Court did not address the argument offered by my colleague Kevin Clermont (in a brief I was on but for which Kevin did all of the work) that when the jurisdictional test is the same as the merits test, jurisdiction should be decided by a prima facie test.

2) Paul: Good one! Yes, this was at risk of becoming "Dorf and Buchanan on the Debt Ceiling and Nothing Else." Neil will have another debt ceiling post tomorrow and then I have another non-debt-ceiling post coming up on Thursday. Plus we may have a bonus post on the fact that Justice Thomas said something from the bench yesterday (even though no one knows quite what he said). Stay tuned!

Joe said...

Scotusblog has a two posts on the Garbo, I mean Thomas, talks matter, one actually also talking about the argument overall and a bonus comment on the lack of law school diversity on the USSC bench. But, hey, welcome Dorf on Law's .02, platinum or otherwise.

Michael C. Dorf said...


1) I thought I was careful to distinguish between "racism"--an ideological view that says some races are inferior and others superior--from race-consciousness. Perhaps you mean "implicit racism" to connote only the latter but I think it's better to reserve all uses of the terms "racist" and "racism" for conscious animus, given their emotional salience. My basis for assuming that the GOP aimed at minority voters rather than poor voters is that I assume that politicians of both parties are very good at counting. According to exit polls, as reported, e.g., at http://tiny.cc/mcuyqw , low-income voters split 60-38 for Obama, which, on its face, would make it seem like a good strategy for the GOP would be to suppress all poor votes. But you have to cross-index that split with the fact that Romney won the white vote 59-39, whereas Obama won African Americans 93-6, Latinos 71-27 and Asians 73-26. I don't have a graph that breaks down voters by both race/ethnicity and income, but it's pretty obvious that suppressing the poor white vote does little good for Republicans these days and probably harms them. These laws suppress poor white votes along with minority votes, but the political goal is to suppress the minority vote. This seems just patently obvious.

Accordingly, I read you to be asking whether a law that is race-neutral on its face and motivated ultimately by political goals that merely use race as a way of getting at those political goals should be treated as race-based. I agree that's technically an open question, but I think the answer should be yes, especially given the ugly history of using race-neutral tools like literacy tests and poll taxes to suppress minority votes. The absence of a racist motive does not save these laws, in my view.

2) If you're right that the coverage formula of VRA Sec. 5 is an unconstitutional fit, then applying equal protection precedents, the remedy could be to expand the VRA, rather than to invalidate it. (See, e.g., Orr v. Orr.) Be careful what you wish for.

But are you right? I don't think that's at all obvious. The VRA coverage formula is based on the history of race discrimination in voting. Surely it's rational for Congress to think that all disenfranchisement efforts are suspect but that it is going to concentrate federal enforcement efforts on those jurisdictions that have a history of engaging in what our history tells us are the most pernicious sort of disenfranchisement efforts--those aimed at disadvantaged minorities. I think your argument is better if you say that while this satisfies the rational basis test, it fails the congruence and proportionality test. I think that's a much closer question (although I don't like the test).

Hashim said...

A few thoughts on your #1:

1. You're right that I intended "implicit racism" only to connote facially-neutral laws that nonetheless are *intended to harm racial minorities* (regardless of whether that is the ideological end goal or merely a means to some other end), in contradistinction to facially-neutral laws that are intended to harm some non-racial group and merely have a disparate impact on racial minorities. I think that all such intentionally racially harmful laws are invidious, whether motivated by ideologicial views about racial inferiority or "merely" a desire to harm racial groups for some non-racial reason, and so I felt comfortable using the term implicit racism to describe them all. That said, it probably was a broader-than-conventional use of the term, and I certainly didn't mean to suggest that you were attributing the idelogical-inferiority component of racism to proponents of voter-id laws, so I apologize for the semantic confusion.

2. More substantively, I totally agree that it's unconstitutional -- indeed, clearly so, in my opinion -- to adopt facially-neutral laws that are designed to achieve non-racial goals through the intentional effect of harm to minorities. This is just a sub-set of the types of laws where the intent to achieve the disparate impact is unconstitutional.

3. That said, I still question the confidence of your empirical prediction about the GOP's motive here. First, you're equating the poor white electorate nationwide with the poor white electorate in the specific states that adopted these laws. Second, you're equating the poor white electorate generally with the specific types of poor whites who don't even have an ID in our modern society. I strongly suspect that, in the States where voter-ID laws were enacted, even the poor whites adversely affected by such laws vote disproportionately Democratic, or at least that the GOP could reasonably have been operating on that assumption. It'd be interesting to see if anyone has tried to measure that empirically -- e.g., I assume that, if I were wrong, the minority challengers to these laws would have tried to prove it, in order to rebut any potential defense based on partisanship.

4. In all events, even your own empirical numbers refute your "motive" argument. It's quite clear from your 60-38 number that the target is in fact poor Democrats generally, not poor minority Democrats specifically. The fact that minorities happen to make up a large percentage of the 60% poor Democrats is utterly irrelevant to the GOP. Do you really think they'd be acting differently if poor whites *also* overwhelmingly supported Democrats, or if the *only* poor Democratic voters were white? I don't, and I don't see any reason why you'd blithely assume otherwise. Indeed, given that even 39% of whites supported Obama, and that there are likely more poor whites than poor minorities in absolute numbers, it seems pretty likely that the *majority* of the 60% poor Democrats is in fact poor white Democrats, which makes it hard to conclude that suppressing the minority vote is the goal here. Instead, at worst, the GOP was willing to throw overboard a few poor white Republicans, knowing full well that there were more poor Democrats of all races.

Hashim said...

And a few more thoughts on your #2:

5. Although it's perfectly rational for Congress to concentrate on race discrimination in voting because that's the most pernicious form of discrimination, it's not rational to use a substantive standard and coverage formula that's focused on race if the goal is to prevent partisan discrimination instead. There's really no reason at all to assume that the jurisdictions with the worst history of racial discrimination are the jurisdictions with the most recent partisan discrimination, and, in any event, the preclearance standard doesn't prevent partisan discrimination. So I don't think any difference between Katzenbach and Boerne matters here, though I certainly agree that it's an easier case for me under Boerne.

6. I don't think the equal-protection remedial precedents apply under Boerne's fit analysis, and, even under those precedents, the SCt would never expand Section 5 because it would have *to re-write* the coverage formula, which is always barred in such remedial situations. Leveling up only works when you can strike an exception and revert to the general statutory provision.

7. That said, I'd love it if the SCt judicially expanded Section 5's coverage formula as a constitutional remedy. That would be the surest way to ensure prompt *legislative* repeal, since no non-covered legislator would allow that to happen, which is why they didn't get the coverage formula right in the first place in 2006. And then the liberal cries of judicial activism and the importance of Section 5 would be neutered.

Michael C. Dorf said...


1) On your first follow-up (re my point earlier labeled 1): I think we've now narrowed our differences to what I regard as a not-very-interesting factual question, as opposed to a theoretical question. (I'm not saying factual questions are inherently uninteresting, just that I'm much more interested in the purely legal questions here). I have been assuming that there's no political advantage to Republicans in diminished voting by poor whites. Where that assumption is false, I agree that it makes little sense to describe general vote-suppression of the poor vote as racial in any regard. But what about where the GOP sees a racially mixed group of poor voters and decides to suppress their votes because poor people are on the whole Democrats (even though the minority poor are overwhelmingly Democratic and the poor whites are, let's say, just a bit Republican)? I agree with you that if the thinking is "let's just suppress the poor vote because it cashes out for us on net" then that's not race-based. But my impression from anecdotal evidence is that this is not in fact what was going on--that at least in particular places (perhaps rogue elements of) the GOP specifically targeted minority neighborhoods (not with voter ID laws but with, e.g., threats of enforcement of such laws) and that more generally the tactic of suppressing poor votes was chosen not in spite of its disparate impact, but because of it. Perhaps I'm wrong about the facts, but if I'm right, then I take it we agree about how that should be treated constitutionally.

2) Just a quick reaction w/r/t remedy: Severing the coverage formula would be analytically clean. Therefore, I don't think it would fall into the "rewriting exception". (A really fun point of comparison is the NY case of State v. Liberta, in which the court "severed" the marital rape exception, thereby expanding its coverage. True, it's a state case, but NY law is not appreciably different from federal law on this point.) But the cases are somewhat inconsistent about when expansion-by-severance is an available remedy.

Cicy said...

Paul: Good one! Yes, this was at risk of becoming "Dorf and Buchanan on the Debt Ceiling and Nothing Else." Neil will have another debt ceiling post tomorrow and then I have another non-debt-ceiling post coming up on Thursday. Plus we may have a bonus post on the fact that Justice Thomas said something from the bench yesterday (even though no one knows quite what he said). Stay tuned!

Fifa 14 Coins -http://fifa14.mmo18.com/PC-Coins.html
League of Legends boost