Here’s Chris Elmendorf’s second guest post on his forthcoming piece:

Enacted by Congress in 1982, Section 2 of the Voting Rights Act (VRA) bans electoral structures “which result[]” in members of a class of citizens defined by race or color “hav[ing] less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.”  During congressional debates on the bill, critics asked supporters of the proposed results test to identify the “core value” they meant to protect.  Whether for reasons of strategy or genuine confusion, the proponents essentially dodged this question, saying only that they intended to restore the legal status quo prior to the Supreme Court’s decision in City of Mobile v. Bolden.  (The Bolden plurality wrote that the 14th Amendment disallows electoral structures that “dilute” minority political power only if the structure was adopted or maintained for discriminatory reasons.)

In the years since 1982, the Supreme Court has struggled to identify Section 2’s core value.  An antidiscrimination results test necessarily presupposes some benchmark conception of neutrality or fairness against which an allegedly discriminatory result may be measured.  As my paper explains, the Supreme Court’s cases variously hint at three different conceptions of the normative benchmark for vote dilution claims, i.e., the entitlement that Section 2 protects: (1) the opportunity to elect a roughly proportionate number of ideally preferred representatives; (2) the representational opportunity that the minority community would have enjoyed under a typical scheme of single-member districts; and (3) the representational opportunity that the minority community would have had absent race-biased decisionmaking by conventional state actors (or majority-group voters).

The third benchmark is in my view the most defensible.  A number of lower courts agree and the law is moving in this direction.  But there is a serious problem with reading Section 2 such that liability depends on proof of intentional discrimination: the framers of the results test were adamantly opposed to intent tests.  The legislative history voices several specific objections to the Bolden plurality’s approach.  The intent test “asks the wrong question”; it’s “unnecessarily divisive”; and it poses an “inordinately difficult burden for plaintiffs in most cases.”  Yet the legislative history also shows that Congress was very much concerned with impairments of minority political opportunity that result from racial prejudice.

It’s possible to reconcile the legislative history’s objections to intent tests with its tacit endorsement of an intent-sensitive benchmark for what is a discriminatory “result” within the meaning of Section 2.  The trick is to expand the object of the intent test (from the lawmakers who enacted the electoral arrangements at issue, to all public actors and majority-group voters whose biased decisions affect minority electoral opportunities), and to relax the standard of proof that plaintiffs must satisfy (from “more likely than not” to something more lenient, such as “significantly likely”).

Let me be clear: I am not arguing that Section 2’s legislative history compels the conclusion that plaintiffs trace their injury to racially biased decisionmaking shown to a “significant likelihood.”  Section’s 2’s text and legislative history are perfectly compatible with a mushy, unstructured legal standard in which liability turns on the judge’s unexplained weighing of any number of factors related to minority political participation, past discrimination, socio-economic conditions, voting patterns, and the electoral system itself.  But the requirement I propose makes sense of the warring intuitions in the legislative history, and would make Section 2 adjudication more normatively transparent.  And as I’ll explain next, it substantially resolves some lingering doubts about Section 2’s constitutionality.

Congress enacted the Section 2 results test as an exercise of its power to enforce the Fourteenth and Fifteenth Amendments.  The Supreme Court has since held that enforcement legislation under the Fourteenth Amendment is permissible only insofar as it is a “congruent and proportional” response to constitutional violations.  (The same tailoring and proportionality requirement likely applies to enforcement legislation under the Fifteenth Amendment.)

[Continue reading below the fold]

If, as Justice Brennan once argued, Section 2 is violated whenever minority and majority voters tend to prefer different candidates and the minority group is unable to elect a proportionate number of its candidate of choice (but could do so under a suitably designed regime of single-member districts), then Section 2’s “congruence and proportionality” is open to doubt.  Given the ubiquity and long tradition of highly majoritarian electoral systems in American democracy, there is scant basis for suspecting an official intent to discriminate from the mere fact that an electoral system results in a minority community enjoying a less-than-proportionate share of political representation.

If, however, Section 2 plaintiffs were required to prove that their lack of electoral opportunity resulted from de jure racial discrimination, there would be little doubt that Section 2 is constitutional.  Congress has the power to legislate against the downstream effects of de jure discrimination.  If obdurate racism in a state’s education or criminal justice system, for example, impairs minorities’ electoral opportunities given the state’s choice of electoral system, then Congress may require the state to change its electoral system so that it’s less of a conveyer belt for discriminatory state action in non-electoral realms.

The constitutional problem can also be fixed by requiring Section 2 plaintiffs to trace their injury to decisions shown to a significant likelihood to be racially biased, even if plaintiffs do not or cannot prove discrimination in accordance with the conventional “more likely than not” evidentiary standard.  This follows from Justice Kennedy’s opinion for the Supreme Court in City of Boerne v. Flores.

Boerne invalidated the Religious Freedom Restoration Act (RFRA), which subjected burdens on the free exercise of religion to strict scrutiny.  RFRA flunked the congruence and proportionality test because the Act reached too broadly and was “not designed to identify and counteract state laws likely to be unconstitutional because of their treatment of religion.” The Court stated, however, that Congress may ban an entire class of state laws “when there is reason to believe that many of the laws affected by the congressional enactment have a significant likelihood of being unconstitutional” (emphasis added).  My formulation of the intent requirement under Section 2 takes the same idea and applies it at a retail level. Instead of “prohibiting certain types of laws” across the board, my reading of Section 2 subjects particular voting requirements to heightened scrutiny once a case specific significant-likelihood showing has been made.

So far so good, but what if plaintiffs’ “significant likelihood” showing concerns subjective discrimination by majority-group voters, rather than conventional state actors?  The individual citizen is not a state actor, and he may even have a First Amendment right to vote for “whatever reason he pleases,” including reasons that the Constitution denies to the state.  My next post argues that even assuming the individual citizen has this right, the electorate as a whole is a state actor when it puts in office persons who wield the coercive authority of the state.  To the extent that Section 2 ameliorates or compensates for the problem of election outcomes that are unconstitutional because of the racial basis for the electorate’s verdict, Section 2 is readily justified under Boerne–so long as liability depends on a significant likelihood showing of race-biased decisionmaking by conventional state actors or the majority-group electorate.

This is not just a debater’s point.  The circuit courts have spent two decades struggling with the question of whether plaintiffs must prove subjective discrimination to prevail under Section 2.  To date, none has seen my solution.  Meanwhile, the Supreme Court has been chipping away at Section 2, narrowing the very broad statutory text with limiting interpretations that rest on the constitutional avoidance canon.  If Section 2’s constitutionality is not in doubt, these holdings ought to be reconsidered.