Here’ Chris Elmendorf’s latest guest post:
In yesterday’s post, I argued that constitutional doubts about Section 2’s results test would be put to rest if plaintiffs were required to show that their injury resulted from race-biased (prejudiced or stereotyped) decisions by conventional state actors or majority-group voters. This is so even if plaintiffs need to establish only a “significant likelihood” of racial bias, as opposed to proving discrimination “more likely than not.” The constitutional argument is straightforward if the discriminators are ordinary state actors, such as legislators, pollworkers, prison wardens, or public school administrators.
But what if the plaintiff only shows race-biased decisionmaking by voters? Societal discrimination was an overriding concern of the Congress that adopted the results test. Yet private race-discriminatory behavior does not violate the Constitution, so in what sense could a Section 2 that targets such behavior be a congruent and proportional remedy for constitutional violations? The answer lies in the problem of election outcomes that are unconstitutional because they were determined by race-biased voting.
The notion that an election held under constitutionally permissible ground rules can yield an unconstitutional result because of the reasons for voters’ choices will strike many readers as peculiar. Isn’t the reason for voting this way or that an entirely private matter? Pamela Karlan and Daryl Levinson have gone so far as to assert that “[t]he voting decisions of individual . . . citizens are absolutely protected under the First Amendment. This is true whether they decline to support candidates favored by [another racial group] out of ignorance, selective sympathy or indifference, or outright racism.” The Supreme Court’s decision in Anderson v. Martin (1964) provides tacit support for Karlan and Levinson’s position.
Yet the Supreme Court has also written, “It is plain that the electorate as a whole, whether by referendum or otherwise, could not order [public] action violative of the Equal Protection Clause . . . .” City of Cleburne v. Cleburne Living Center, Inc. (1985). This was not a careless dictum. It justified the Court’s invalidation of a popularly adopted restriction on group homes, which “rest[ed] on an irrational prejudice against the mentally retarded . . . .” It made sense of several earlier decisions in which the Court invalidated facially neutral, voter-enacted laws on equal protection grounds. And it has been honored by the Court in unconstitutional motive cases in the years since.
The seemingly incompatible precepts of Anderson and Cleburne can be reconciled. Here is the key: the citizen’s right to vote “for whatever reason he pleases” is a personal right, and does not entitle him to a corresponding measure of influence over election outcomes. The determination of election results—the collective byproduct of these personal choices—is state action. Otherwise private behavior becomes state action when it determines the outcome of an election, assuming that the officials so elected wield the coercive authority of the state.
A court that invalidates an election result because an outcome-determinative share of the votes were cast for discriminatory reasons does not burden anyone’s right to vote. This follows from the personal nature of the right. Though the citizen may vote “for whatever reason he pleases,” he may not exercise political power “for whatever reason he pleases.” It may seem odd to think of the right to vote as distinct from the exercise of power, but this is more or less how the Supreme Court has come to understand it.
The Cleburne principle is well established in constitutional challenges to legislation enacted by plebiscite, but as best I can tell no court has been asked to undo the outcome of an election for public office because it “rested on prejudice.” One might distinguish the two scenarios on the ground that a popularly approved referendum creates binding law, whereas the election of a representative does not bring the power of the state directly to bear on its citizens. Yet the Equal Protection Clause applies to state action generally, not just to state actions that directly coerce or tangibly burden particular, identifiable citizens.
It is state action to appoint a public official with coercive lawmaking or law enforcement authority, just as it is state action to enact rules of law that directly bind the citizenry. A mayor’s decision to appoint David Duke as sheriff, for example, because of the mayor’s stated desire to restore Jim Crow “by any means necessary” would violate the Equal Protection Clause. If the sheriff’s office were elective rather than appointive and Duke ran for sheriff pledging to restore Jim Crow by any means necessary, his victory would be just as unconstitutional as his hypothetical appointment by a racist mayor–as long as his margin of victory was provided by voters who acted on the same motives as the mayor.
The state action in the mayor’s appointment of the sheriff is not simply a byproduct of the mayor’s status as a public official. Rather, the putting in office of an official cloaked with coercive lawmaking or law-enforcement authority should be understood as a “public function” within the meaning of the state action doctrine. As my paper explains, this conclusion follows quite naturally from cases about such seemingly disparate subjects as jury selection (the exercise of peremptory challenges), delegation of regulatory authority to private actors, and exemptions from the constitutional rule of “one person, one vote.”
* * *
That there is no state action difference between the Klansman-for-sheriff hypothetical and the racist ballot initiative does not mean that courts should treat the two cases the same way. The problem of unconstitutional electorate motive (“Cleburne violations”) in elections for representatives should almost certainly be held to present a nonjusticiable political question, even though it is within the judicial competence to adjudicate Cleburne challenges to the outcome of a referendum election. This, I will argue, has important implications for how courts should evaluate the congruence and proportionality of any congressional response to the problem.
I won’t belabor the political question argument here. As developed in my paper, it turns on the lack of a satisfactory remedial counterpart (in the case of elections for representatives) to the enjoining of direct legislation enacted for unconstitutional reasons; on the structure of the Constitution, which contemplates no role for the federal courts in deciding who shall be seated following a disputed election; on the need for a more confining legal standard for adjudicating electorate motive challenges to representative as opposed to plebiscitory elections; on differences in the types of evidence available; and on the potentially disruptive effects of electorate motivate challenges to elections for representatives.
Assuming I’m right that Cleburne claims present nonjusticiable political questions in the context of representative elections, what follows for Section 2? There are two significant implications.
First, courts should apply the congruence-and-proportionality test with a light touch insofar as the Section 2 targets Cleburne violations. Congressional authority under the Fourteenth Amendment’s enforcement clause properly reaches its apogee when Congress undertakes to prevent or remedy a type of constitutional violation that ordinary constitutional litigation cannot reach. This follows from the familiar distinction between Type 1 (false positive) and Type 2 (false negative) errors.
In the context of judicial application of Boerne’s congruence-and-proportionality standard, a false positive is striking down a congressional enactment that amounts to reasonable remedial legislation. A false negative occurs if the court upholds enforcement legislation when Congress made gross errors of judgment in designing the enforcement mechanism. The cost of a Type 1 error depends on the likelihood that the constitutional norm will be violated absent the invalidated legislation. If the courts cannot enforce the norm at all absent suitable prophylactic legislation from Congress, then Type 1 errors are, ceteris paribus, of much greater concern than if the courts can and will provide vigorous enforcement on their own.
The other important implication is an interpretive thumb-on-the-scale in favor of reading Section 2 to authorize heretofore unrecognized “depolarization claims.” As I define it, a depolarization claim challenges electoral arrangements on the ground that they unreasonably induce or sustain race-biased voting, whatever the consequences for minority representation.
Depolarization claims and conventional vote-dilution claims are both ways of responding to race-biased voting by the majority-group electorate. The universal remedy in dilution cases has been to restructure the electoral process (e.g., the design of legislative districts) to enable the election of more minority champions even if race-biased voting continues unabated. The remedy for a meritorious depolarization claim would be to change the electoral arrangements that induce or sustain race-biased voting, thereby reducing the intensity or frequency of such voting in the future. The latter is superior as a means of combating Cleburne violations. The less race-biased voting there is, the less likely it is to determine the winner of any given election.
By contrast, the redesign of legislative districts so as to enable the election of more minority-preferred candidates does not, as such, prevent Cleburne violations. The pattern of Cleburne violations might well continue—with each racial group electing candidates on racial grounds—albeit without so noxious an impact on the distribution of representation across groups.