The Yale Law Journal


Section 5 as Simulacrum

08 Jun 2013
Justin Levitt

The Voting Rights Act is widely hailed as the most significant civil rights statute in American history.1 It represents a majority commitment to minority political representation—one of the only national legal provisions to do so, and the only one that does so with respect to racial or ethnic minorities. Section 5 is the innovative provision that was, in some ways, at the heart of the Act; it delivered the strongest medicine to the jurisdictions with the worst problems. And though abuses persist, the prospects for representation of racial minorities in those jurisdictions promptly and dramatically improved. Among its other attributes, section 5 is therefore a provision of enormous expressive and historical importance. In the recent Shelby County v. Holder case reviewing a constitutional challenge to the 2006 reauthorization of section 5,2 the origin story and symbolic significance of the statute have been front and center. Conventional wisdom suggests that emphasizing section 5’s evocative aura raises the psychic stakes of striking the provision down.

The symbolic significance of section 5, however, has some substantial and underexplored downsides. A section 5 that serves as the representation of more than just a statute turns the fight over its constitutionality into a fight over a preconceived image that is linked to, but also departs from, the section 5 enacted by Congress. That is, there is a distinct possibility that Shelby County will turn more on a simulacrum of preclearance than on the actual statutory provision at issue in the case. That should be cause for concern.

The privileging of imprecision begins with the very reference to “section 5,” which is more metonymy than description. The challenged federal requirement is not truly a single statutory section, but a sophisticated regulatory regime spanning three different sections of the law. Section 5 provides the “what”: federal authorities (the Department of Justice or a Washington, DC, federal court) must “preclear” any change to election-related procedures in a covered jurisdiction, ensuring that the new procedures neither decrease the existing pragmatic political power of racial or ethnic minorities, nor are intended to discriminate against those minorities.3 Section 4 provides a part of “where” (and perhaps “when”): a formula for determining which jurisdictions were initially covered, a provision for individual jurisdictions to “bail out” from coverage when coverage is no longer warranted, a provision for Congress to reconsider after fifteen years whether the entire coverage regime is no longer necessary, and a sunset provision to terminate coverage after twenty-five years.4 Section 3 provides another part of “where”: a “bail-in” provision, allowing courts to impose a slightly modified preclearance process upon a finding of intentional discrimination.5

The case currently before the Court ostensibly challenges this package of procedures as beyond the proper exercise of congressional power in 2006. But the limited nuance of the preceding description—much less the actual details of how these provisions operate in concert and in practice—is largely absent from the prevailing narrative. The vision of section 5 that emerged from much of the Shelby County oral argument, which mirrored the broad understanding evinced in the popular press, is a construct meaningfully different from the statute itself.

The “likeness” of the preclearance regime rooted in the popular imagination may perhaps be sketched as follows.6 This section 5 simulacrum is a crown jewel of a civil rights regime forged in a different American environment. It was designed to respond to the urgency of the Southern racial suppression of the 1960s, achieved enormous success in that social environment, and continues to excite profound passion in communities that benefited from that success. It requires conscious consideration of race and ethnicity in electoral policy—indeed, in some circumstances, consideration of race and ethnicity will be the but-for cause for choosing some policies over others. Policies with a disparate impact on minorities may be blocked, regardless of the other merits of those policies. In the redistricting context, in particular, decisionmakers operating under the section 5 simulacrum seek pockets of minority voters to lump together into districts casually identified as “black districts,” “Latino districts,” “Asian districts,” or “Native American districts,” which may also become vehicles for legislators seeking partisan advantage. These districts become safe seats for minority legislators with positions distinct from the political mainstream. They also become anchors for future cycles, establishing floors for levels of effective minority participation that limit local discretion.

To accomplish these ends, on top of other, more familiar, legal provisions confronting racial discrimination, the section 5 simulacrum delays all proposed election-related changes by state and local governments within the regime, including changes that are not even plausibly discriminatory; the vast majority of decisions for which jurisdictions must seek federal approval are uncontroversial. And only certain jurisdictions are subject to preclearance. Those that are included were placed on the list in the 1960s, but no longer release police dogs and fire hoses on black protesters; indeed, on some measures of racial electoral performance, they perform as well as or better than other jurisdictions. Moreover, the list of jurisdictions subject to preclearance does not include portions of the country with serious racial problems of their own. Under the section 5 simulacrum, all of these elements—features or flaws, necessary or gratuitous, depending on your perspective—will persist indefinitely unless the Court intercedes. And though the Court launched a shot across the bow in 2009, articulating some areas of concern with the statute, Congress has not modified the preclearance regime since.7

This description of the section 5 simulacrum is much like an editorial cartoonist’s rendering of a political figure, in which particular features take on exaggerated salience. Most elements of the section 5 simulacrum—particularly when they are seen as isolated elements—have at least the ring of truth, which helps to explain the staying power of the image. But though the cartoon version of the preclearance regime resembles the original, the exaggerated features distort rather than clarify our understanding of the legality of the portions of the Voting Rights Act at issue. To detractors, the overall image that emerges is that of a petulant Congress needlessly and clumsily forcing unwilling states to engage in racial essentialism, at the behest of legislators and interest groups who cannot recognize that we no longer live in the past. It is a powerful construct, and one worthy of scorn. But this simulacrum does not describe the statute that Congress actually passed, and therefore should not drive discussion of the actual statute’s constitutionality.

This short Essay investigates some of the more striking elements of the section 5 simulacrum, contrasting the cartoon vision of section 5 with the more fully contextualized operation of the actual statute. Part I discusses the perception of the coverage formula determining the geographic scope of preclearance. Part II reviews the perception of preclearance’s substantive scope: the types of election-related changes properly governed by section 5. Part III engages the perception of the comparative merits of alternative remedial regimes. Part IV reviews the perception of section 5’s impact on race relations more generally, including charges of racial essentialism and now-infamous “racial entitlements.”8 Part V takes an even longer view, confronting the perception of the preclearance regime’s overall place in the constitutional jurisprudence of race-consciousness and federalism. In each case, the real-world statute stands on firmer constitutional ground than its flattened and distorted simulacrum.

I. The Coverage Formula

Consider first the coverage formula. Certain jurisdictions must preclear; others need not. There is no question that preclearance involves an unusual exercise of federal power, and that there must be some reason to subject certain jurisdictions—and presumably only those jurisdictions9—to the prophylactic procedure. Much of the current constitutional challenge concerns the degree to which the 2006 Congress adequately distinguished covered jurisdictions from those that are not covered. And much has been made of the fact that Congress kept essentially the “same” coverage formula that it first put in place in 1965.

Coverage attaches, as an initial matter, if a jurisdiction had used a test or device as a prerequisite to voting, and if less than fifty percent of the eligible electorate registered or voted in 1964, 1968, or 1972.10 Conventional wisdom regards this coverage as an ingenious means of “reverse-engineering” protection in the most racist jurisdictions at the heart of Jim Crow.11 And with an emphasis on the lack of change to this portion of the coverage formula, the section 5 simulacrum is seen as a 1965 Rand-McNally road map, drawn before completion of the Interstate Highway System, much less the Internet.

There are three related flaws at the heart of this cartoon. The first is the presumption of indefinite life: the paper upon which a road map is printed may eventually deteriorate, but it does not evaporate on a specified date. Similarly, the simulacrum seems devoted to the notion that section 5 will for all practical purposes linger indefinitely.12 But this does not describe the actual preclearance regime under challenge. The provision before the Court must be reconsidered in 2021, and expires in 2031.13 Congress was either justified in crafting a preclearance provision for this period or it was not. There is no basis for considering the speculative political calculus of a future Congress, much less the factual predicate of a future decade, in assessing the constitutionality of the actual statute in question.

The second is the presumption of stasis: a printed edition of a road map is incapable of incorporating new information. Again, this is not the case for coverage under the actual preclearance regime. Though the coverage formula was developed in 1965, it was designed from the outset to be adaptive, with the extent of coverage changing in response to changing circumstances. The two-part standard above (covering areas where voting was contingent upon the use of a test or device, and which had less than fifty percent registration or turnout) is only the start of the formula. To that are added the section 3 bail-in and the section 4 bailout provisions. Hundreds of jurisdictions that once had a responsibility to preclear election laws no longer must do so; many jurisdictions not covered in 1965 were later covered by extension of the statute, and eighteen jurisdictions have been (temporarily) covered by judicial order.14 As facts on the ground demonstrate that jurisdictions should be subject to preclearance or free from preclearance, coverage is designed to expand and contract.

This answers the objection that Congress should have developed a new formula in 2006 based on new information, and that any pre-existing formula is necessarily outdated. In reality, the bail-in and bailout mechanisms mean that the “old” formula continually encompasses new information. A jurisdiction is now covered if it was once covered (old formula) and has not bailed out (new information); a jurisdiction is not covered now if it was not covered (old formula) and has not been bailed in (new information). Bail-in and bailout are necessary components of the coverage determination, but are largely absent from the section 5 simulacrum.15 Looking at the coverage determination holistically, it is apparent that the list of covered jurisdictions is just as much the product of current determinations as it is the product of decisions from 1965.16

The third flaw is related to the last, and to the objection that Congress should in 2006 have found a new way to distinguish covered from non-covered jurisdictions based on present conditions, and that its failure to “update” was mere political cowardice. There have been several suggestions that distinctions in 2006 data should be reflected in 2006 coverage.17 At oral argument in Shelby County, for example, the Chief Justice seemed to suggest that an appropriate 2006 formula would have to consider distinctions ostensibly parallel to the original coverage formula. Just as the original formula focused on registration and turnout, so did the Chief Justice, comparing the relative registration and turnout of white and African American voters in Mississippi and Massachusetts.18 He seemed to say that the preclearance regime treats Mississippi “worse” than Massachusetts, when—buying into the simulacrum—conditions on the ground today indicate that it should be the other way around.

Proponents have also defended the preclearance regime based on distinctions in 2006 data: most prominently, the comparative number of actions brought under section 2 of the Act, which prohibits vote denial and dilution on the basis of race and language minority status nationwide (rather than in select jurisdictions) and places the burden of proof on affected plaintiffs (rather than jurisdictions seeking a change in policy). The rates of successful section 2 litigation were substantially higher in covered jurisdictions than in non-covered jurisdictions; the respondents in Shelby County used this disparity, among others, to justify Congress’s decision to perpetuate coverage distinctions in 2006.19

Both of these measures—and all other numerical distinctions based on voting-related outcomes between covered and non-covered jurisdictions20—ask fundamentally the wrong question. The issue is not whether Congress had sufficient reason to believe that covered jurisdictions were meaningfully different from non-covered jurisdictions in 2006. The real issue is whether Congress had sufficient reason to believe, based on past practice, that covered jurisdictions would be meaningfully different from non-covered jurisdictions without a preclearance regime. In the Chief Justice’s terms, the question is whether 2006 evidence showed a likelihood of greater harm to minority participation in Mississippi than Massachusetts—in the absence of section 5. Only this latter question gets to the heart of the extent to which the coverage of the preclearance regime is tailored to Congress’s objectives.

The problem is that the most probative evidence for this relevant inquiry effectively requires evaluating Mississippi in 2006 without section 5. This is a counterfactual beyond anyone’s capacity to model or measure accurately. As a result, Congress necessarily had to turn to proxies.21 I have argued elsewhere that the absence of bailout, coupled with present evidence of harm, essentially serves as a proxy at least as meaningful as other contemporary data.22 The logic is as follows. The original formula covered jurisdictions where democracy was effectively broken—that is, where elections in 1964, 1968, or 1972 were held without the participation of at least half of the eligible electorate. Any jurisdiction where participation was not limited “on account of” race or language minority status could bail out, leaving areas covered only where participation was limited on account of race or ethnicity, often through vicious intentional discrimination. Other jurisdictions that were found to have engaged in intentional discrimination on the basis of race or language minority status could be covered through the bail-in mechanism. After August 4, 1984, any jurisdiction could bail out of coverage with a ten-year record of “good behavior” preserving and working to improve the political participation of minorities.23 Together, this means that jurisdictions not covered in 2006 either were not broken24 or had been broken and demonstrated improvement. In these jurisdictions, Congress determined that remedies other than preclearance (e.g., affirmative litigation under section 2) might suffice to address occasional lapses or new difficulties. In contrast, jurisdictions that were covered in 2006 and unable to bail out had been broken on account of race, and had not yet demonstrated a sufficient departure from the patterns of the past.25 That fact alone is a powerful indication that, absent section 5, there would be a meaningful difference between non-covered jurisdictions and covered jurisdictions in the potential efficacy of alternative remedial schemes. When added to 15,000 pages of legislative records articulating current threats to minority participation in covered jurisdictions, even given the presence of section 5, the proxy becomes stronger still.26

The analysis above places a substantial amount of weight on the meaningful potential for bailout when a jurisdiction has demonstrated an absence of recent action impairing political rights on account of race. That weight is proper; bailout is not merely an afterthought to coverage, but an integral design feature of its scope. It amounts to one of the major differences between a section 5 simulacrum revolving around a fixed original formula and the actual statute before the Court.

II. Covered Procedures or Practices

Another element of the section 5 simulacrum is the centrality of the problems that inspired preclearance: shifting legal regimes developed by recalcitrant registrars and precinct officials to shut minority citizens out of the voting process.27 Extreme (and ever-changing) racial and ethnic restrictions on access to a valid ballot were unquestionably the primary predicates for the preclearance portion of the Voting Rights Act. Now, although restrictions with disproportionate impact have not vanished, most racial and ethnic minorities in covered jurisdictions are able to cast a valid ballot. This must be the basis of the assertion at oral argument by Shelby County’s counsel that “the problem to which the Voting Rights Act was addressed is solved.”28 If the principal visual association for section 5 is a Southern registrar (backed by state-supported violence) inventing a new procedure in order to effectively stymie all of his jurisdiction’s African American voters, the 2006 provision looks as outdated in substance as it does in coverage.29

To the extent that the section 5 simulacrum ever represented reality, however, the real statute is certainly no longer merely about wholesale access to the ballot. The authors of the Act recognized the powerful creativity of the drive to retain political power, and built a preclearance regime encompassing “any” procedure with respect to voting that either denies or abridges the right to vote on account of race or color.30 This certainly includes the Jim Crow tactics of the 1960s: blunderbuss acts aimed at minority citizens across the board. But it also includes more subtle procedures that dilute political power by affecting not most minority voters, but a substantial and disproportionate number of those voters. And at least since Allen v. State Board of Elections,31 the provision’s scope has also been interpreted to encompass “second-generation” abridgment of political power.32 Such abridgment includes not only procedures that constrict access to the polls, but also districting and annexation decisions that limit or degrade racial or ethnic minorities’ meaningful representation even when each individual can cast a valid ballot.33

Congress has amended the preclearance provisions of the Voting Rights Act several times since Allen—including amendments to overrule interpretive decisions of the Court—but in doing so, it has embraced, not overruled, Allen’s central premise, recognizing that the illegitimate dilution of electoral power can take many forms. Longstanding principles of statutory interpretation provide that reenacted statutory provisions are presumed to adopt and incorporate prior judicial constructions of statutory text.34 Each time Congress reenacted the preclearance regime, it embraced Allen.35 And each time, the Court validated the reenactment as constitutional, including its application to dilution of minority representation, and not merely to problems with registration and turnout.36 In 2006, Congress expressly reaffirmed its intent to address these “second generation barriers.”37 No less than the preclearance provisions of 1970, 1975, and 1982, the preclearance provision of 2006 confronts evils far more expansive than those reflected in the 1960s news broadcasts of the simulacrum.

The expansive substantive breadth of the real section 5 renders it more tailored, not less, to preventing unconstitutional deprivations of electoral power in a contemporary environment. Regrettably, efforts to deprive minorities of political power on account of race or ethnicity through second-generation structures like redistricting are still stubborn features of the present political environment.38 Indeed, it may be logistically easier now to effectuate second-generation abridgment than it is to effectuate the first-generation, 1960s-era equivalent. For example, the growth of mail-in registration facilitated by the National Voter Registration Act39 may make it more difficult to reliably discern the race or ethnicity of an individual prospective registrant,40 but the Census Bureau delivers the racial or ethnic composition of a precinct to any official’s door.

III. Alternative Remedial Schemes

The section 5 simulacrum also depicts preclearance as one particularly aggressive treatment protocol among other remedies just as capable of treating the relevant illness with fewer side effects. Intentional discrimination based on race or ethnicity is unconstitutional, and can be addressed by a civil rights lawsuit under 42 U.S.C. § 1983. And even where direct proof of unconstitutional intentional discrimination is difficult to amass, section 2 of the Act provides a cause of action to confront electoral practices that, in the “totality of circumstances,” result in a “denial or abridgement” of the right to vote.41 These are powerful responsive litigation tools to combat discrimination. The Department of Justice and privately funded civil rights groups have ample motivation to deploy these tools, and tout their successes in using them around the country. The simulacrum sets section 5 alongside section 2 and section 1983—and thus arrayed on the remedial shelf, the benefits of preclearance appear roughly equivalent and its intrusion on local governance therefore less necessary.42

This perspective on preclearance may well be fostered by judicial experience with the common posture of section 5 litigation. Because section 5 exists, courts no longer see the repeated cycles that gave rise to the original need for preclearance: dilution, followed by responsive litigation, followed by changes fostering dilution in different ways. Courts also do not generally see the “first drafts” of policies that are modified in the shadow of the existing preclearance process: either policies that are not promulgated because they would be blocked by section 5 before they are implemented, or policies that are promulgated but revised after dialogue with the Department of Justice (before or after a formal objection). For the tiny portion of preclearance cases that result in litigation, courts are asked to decide how the statute’s terms should be applied to a contested policy, which is an inquiry in the heartland of judicial competence. They are not generally asked to weigh whether alternative regimes might have resolved the alleged harms equally effectively, which is quintessentially a question of policy. The judiciary’s lived experience with section 5 cases may well influence its vision of the place of preclearance in the remedial toolkit.43

In reality, section 5 is not merely the more aggressive equivalent of responsive litigation. It is well understood that the preclearance regime prevents iterative cycles of discrimination and lawsuit that responsive litigation cannot address: it stops the game of Whac-A-Mole in which responsive litigation is never able to catch up with determined wrongdoers. Scholars are also beginning to recognize that the preclearance regime establishes a forum for productive policy dialogue—a structured negotiation between submitting jurisdictions and the Department of Justice regarding means to mitigate apparent discrimination44—that responsive litigation may provide less readily or less effectively.45

But even when preclearance and responsive litigation are trained on the same contested electoral rules and arrive at essentially the same conclusions with respect to their legality, there are enormous practical distinctions between the regimes. Responsive litigation depends on an ability to amass, process, and present substantial information—demographic and electoral data, formal legislative records and legislators’ informal comments, and historical context, among others. Courts are, correctly, loath to offer even preliminary relief before data sufficient to establish a case have been assembled, particularly during the period shortly before an election.46 More prominent disputes—for example, statewide redistricting battles—are likely to draw substantial governmental and private resources and expertise to help prepare responsive litigation, but given finite resources, there is a greater risk that smaller jurisdictions like towns, villages, constable districts, and school boards will be overlooked.47 And even in the larger jurisdictions, assembling sufficient data to present a credible case takes significant time.

Changes to electoral policy several months before an election—much less those implemented on the eve of an election—may not afford adequate opportunity to assemble a reasonably robust litigation response when the claim involves racial vote dilution. There are, to be sure, a large number of disputes that head to the courts in the weeks before an election,48 but these generally depend on questions of law rather than questions of fact.49 In contrast, when electoral changes close to an election are discriminatory, the issue is entirely fact-based—and the election will often take place before sufficient proof of the wrong can be assembled. The officials elected under the improper regime are then empowered to make policy until the plaintiffs are able to prove harm and the courts order a remedial election—or, with greater delay, a regularly scheduled future election under remedial procedures.50 Preclearance is designed to stop problems before they have meaningful impact on local communities; responsive litigation is far more likely to fail this test. These contextualized understandings of the limitations of litigation, and their consequences, are largely absent from the simulacrum of section 5.

IV. Racial Essentialism and Racial Entitlements

The elements of the simulacrum discussed above generally involve conventional wisdom about how the preclearance regime was intended to work. Another central element of the section 5 simulacrum revolves around the perception of a presumably unintended consequence: the fear that the provision fosters racial essentialism. The concern arises most centrally in the redistricting context. The notion is that section 5 requires covered jurisdictions to create “black districts” or “Latino districts” by lumping together voters with little in common other than their race or ethnicity.51 Indeed, the section 5 simulacrum provides cause for racial minorities to demand tailored districts that are privileged by federal law over districts demanded by other interest groups, and which remain locked into future maps. This is the conception of section 5 that might most plausibly be driving Justice Scalia’s notion of the provision as providing “racial entitlements.”52

With respect to the essentialism concern, the flaw in the section 5 simulacrum is the assumption of causality. There may well be racial essentialism in covered jurisdictions, but to the extent that it exists, the fault cannot properly be laid at the feet of the actual federal statute. Section 5 prohibits abridgment of voting power on account of race or ethnicity. It therefore has teeth in the redistricting context only to the extent that racial or ethnic minority blocs actually exhibit consistent voting preferences that would, if unprotected, be abridged. The predicate condition for abridgement is a set of preferences distinct from those of the surrounding community. As an illustration, imagine a jurisdiction in which all voters prefer precisely the same types of candidates, and imagine further that you were intent on abridging the voting power of a subgroup of these citizens. How would you draw district lines or shape an at-large district to accomplish the goal?

Thus, in order for districts to “abridge” voting power on account of race or ethnicity, racial or ethnic groups whose power is at risk must have consistent, and distinct, political preferences. This is known in voting rights circles as racially polarized voting. If voting is not racially polarized—if many minority voters prefer the same types of candidates as the majority, or if minority voters are politically diverse in the same ways that the majority community is politically diverse—a jurisdiction’s chosen district scheme cannot abridge “the” racial community’s voting power because there is no such entity as “the” racial community for purposes of political preference. Section 5 requires no remedy here. If, in contrast, voting is racially polarized—if minority voters are politically unified in a manner distinct from the surrounding community—then it is possible to comprehend (and preclude) practices that abridge this voting power. But then the force of section 5 is based on actual political preferences, not essentialist assumptions.

It is true that, in practice, some jurisdictions may draw “black districts” or “Latino districts” based solely on demographics, indulging racial essentialism and citing section 5 as a politically palatable excuse. But the misapplication of a statute is not a constitutional flaw of the statute itself. And the judiciary has refused to indulge these misunderstandings.53 Courts have consistently emphasized that preclearance decisions are not properly premised on racial statistics alone, overturning the contrary assumptions of covered jurisdictions in the process.54 Rather, retrogression is based on pragmatic political power, as evinced by registration rates, turnout rates, and the degree to which racial or ethnic minority voters actually demonstrate consistent political preferences. The precondition for section 5 to impose meaningful constraints on local decisions is that minorities are already politically unified. Section 5 cannot be blamed for creating unwarranted assumptions, because it shapes government behavior only where the contested facts in question are already present.

Occasionally, the further objection is raised that districts drawn to preserve the political power of polarized minorities may not indulge assumptions about polarization or create polarization, but will perpetuate or strengthen it.55 For example, districts drawn to provide opportunity to effectuate the political preferences of polarized minorities will tend to elect, at least initially, candidates responsive to the interests of that minority group. And, the argument goes, that candidate will encourage further polarization among the electorate in order to cement his or her chances of re-election.

While it is always possible that such a phenomenon might be observed in individual districts, it is not clear why increased polarization necessarily follows from districts drawn to effectuate the political influence of minority populations that have been unified in the past. Imagine a district with a polarized Latino community comprising fifty-five percent of the district’s electorate. This community may have a history of political cohesion, but it is neither unitary nor homogeneous; indeed, rejecting racial essentialism requires an understanding that different Latino voters will have slightly different sets of salient concerns. For every candidate interested in maintaining the ethnic polarization of the electorate in order to win that Latino bloc, another should emerge with an interest in crossing ethnic lines to win the Latino vote as well as some of the non-Latino vote, and still another should emerge with an interest in winning the non-Latino vote as well as some of the Latino vote. Still others should emerge with an interest in amassing winning coalitions of the electorate that are not measured in ethnic terms at all. The winning strategy will depend on the cleavages that the voters find most salient. To the extent that the winning coalition divides along ethnic lines, it is not the district configuration but the voters’ own preferences that are the cause. The real section 5 therefore does not create or foster a polarized electorate. Rather, in an electorate where racial polarization is presently a fact of life, it simply allots a modicum of representation to the minority.56

This function of the preclearance regime leads directly to the racial entitlements argument. The real statute does provide a benefit to politically unified minority communities. It creates a legal obligation to draw districts preserving the political power of these communities. No federal law creates a similar obligation for others.

This function of the law as a special benefit is central to the section 5 simulacrum—which also de-emphasizes the essentially remedial context of the provision.57 As discussed above, the preclearance regime only binds jurisdictions where democracy was broken on account of race or language minority status, and where the preceding decade has not been free of practices that continue to abridge the franchise on account of race or language minority status. It preserves minority representation in areas where intentional discrimination was the norm, until individualized determinations, jurisdiction by jurisdiction, make clear that the local norm no longer persists. The remedy is therefore both temporally and geographically limited. The remedy is substantively limited as well. Section 5 prevents intentional discrimination and backsliding; that is, in areas where the process does not specifically target minorities, it prevents only electoral changes that leave racial and language minority communities with less representation than the status quo ante.58 If this is an “entitlement,” it is an entitlement to preserve a modicum of representation in jurisdictions that aimed to squelch that representation and have not—yet—demonstrated a sufficiently robust record of change.

V. Race-Conscious Decisionmaking and Federalism

Thus far, this Essay has noted elements of the section 5 simulacrum at roughly increasing degrees of generality: a coverage formula perceived as outdated, confronting problems perceived as solved, through a device perceived as redundant, in a manner perceived as entrenching racial divisions. The final notions discussed herein are also the most expansive. The section 5 simulacrum is squarely in the crosshairs of the Court’s concerns over race-conscious government action and federalism.

The preclearance regime is a federal limit that stops Texas, but not Indiana, from implementing, for example, a restrictive photo identification requirement.59 It does so because Texas egregiously discriminated on the basis of race and ethnicity in the past and has not demonstrated a sufficient break from these practices in the present. The combination has prompted a federal response that tests the impact of the state’s policy on racial and language minorities’ effective political power before it is implemented, rather than in the more usual course of responsive litigation.60 Thus, in certain circumstances, the preclearance regime imposes a serious race-conscious federal constraint on the autonomy of state actors.

For the section 5 simulacrum, these features are the legislation’s centerpiece: the ears on an editorial cartoon of President Obama or the mold of a cartoon President Reagan’s hair. For advocates, they are precisely the reason to celebrate section 5; for opponents, they are a primary cause for attack.

But the more that these elements acontextually define the preclearance provision, the more they thrust it into a two-front proxy war. Over the last three decades, the Court has shown increased skepticism about race-conscious government decisionmaking in various contexts.61 Over the same period, the Court has also become more critical of the federal government’s perceived overreaches, particularly when Congress impinges directly on the autonomy of state actors.62 The section 5 simulacrum, springing from a 1965 vision of the proper roles of race and federal power, is on the wrong end of two formidable jurisprudential freight trains.

Above, I argued that the real section 5 differs meaningfully from its caricatured equivalent. Here, the real section 5 is a powerful race-conscious assertion of federal power. But, crucially, it is not merely so—or, at least, not in the ways that have concerned the Court in the past.

The critical distinction is the source of section 5’s authority. Most of the Court’s recent constitutional race cases address Fourteenth Amendment challenges to states’ race-conscious decisions. The federalism cases are more varied; some concern Congress’s Article I power (either the extent of that power or the means by which Congress may conscript state entities to execute it), while some concern federal attempts to open state treasuries or redefine constitutional harm under the Fourteenth Amendment. None of these recent cases, however, examines Congress’s attempt to enforce the narrow mandate of the Fifteenth Amendment, which provides an unquestioned predicate for section 5.63

The Fifteenth Amendment expressly grants Congress the power—and the responsibility—to ensure that states do not deny or abridge the right to vote on account of race.64 There are remarkably few ways to enforce that mandate that involve neither attention to race nor imposition on state autonomy. If the Court’s general skepticism about race-consciousness and federal impositions on state autonomy (and its consequent willingness to closely scrutinize such measures) is legitimate, it must be grounded in a sense that such action runs contrary to the constitutional command in its normal course. Here, in the discrete context of access to political power, the Fifteenth Amendment affirmatively authorizes what the Court normally finds constitutionally troublesome.

The Fifteenth Amendment also occupies a distinct position in the constitutional structure. Consider its predecessor. Whatever the 1868 conception of the Fourteenth Amendment may have been, our modern understanding of its equal protection mandate plainly embraces a prohibition on racial discrimination in voting, just as in every other public enterprise.65 If the Fifteenth Amendment accomplished nothing more, it would have become a mere constitutional artifact. I use the subjunctive tense, because the mooting of a constitutional amendment is, at the very least, an unusual idea. Constitutional provisions have been expressly repealed (e.g., the Eighteenth Amendment) or specifically superseded (e.g., the election of the Vice President by selecting the runner-up in presidential ballots), but to be simply subsumed by an earlier enactment, to fade away from meaning?66 Such a result is anomalous, and possibly wrong. This offers good reason to believe that the Fifteenth Amendment provides content beyond that contained in the Fourteenth Amendment. Although a full exploration of the Fifteenth Amendment’s distinctive substance presents numerous conceptual challenges well beyond the scope of this Essay, the enforcement power granted to Congress may be a good place to start.67 This power cannot be merely identical to power already granted by its constitutional predecessor.

Both text and structure therefore indicate that in Congress’s exercise of its Fifteenth Amendment enforcement power, race-conscious federal requirements of states should prompt deference, not close scrutiny. When Congress enforces the Fifteenth Amendment, it acts pursuant to an enumerated power, not contrary to a constitutional prohibition. The appropriate inquiry is therefore not whether the 2006 reenactment is suboptimally tailored, but rather whether it is sufficiently tailored; there is no constitutional cause for a least-restrictive-means test here.

To be sure, the deference due Congress is not unlimited. The Court should properly police a rough congruence between Congress’s means and the constitutional end. Congress may not, for example, enforce the prohibition against racial discrimination in the franchise by prohibiting the purchase of electronic voting machines—at least, not without a record establishing that such machines have been or are likely to be used to foster racial discrimination.68 But this last caveat makes the relevant point: once Congress establishes a relevant constitutional harm, and establishes a reasonable tie between that harm and its preferred deterrent or remedy, there is no constitutional reason for the Court to import its skepticism from other constitutional contexts into legislation premised on the Fifteenth Amendment. Further pushback would only defend the objects of congressional attention from the Constitution.


Section 5 is a statutory provision of tremendous symbolic importance. But in a judicial arena, there is a danger in allowing it to be a symbol first and a statute second. Such an approach turns the dispute into a challenge to an editorial cartoon of the Voting Rights Act, factually and legally decontextualized and thereby distorted by each passing pen. The oral argument in Shelby County, like the prevailing popular narrative, betrayed apparent reliance on a section 5 simulacrum. It would be a mistake to assume that these snippets of instinct and conventional wisdom will find their way into the U.S. Reports. But it would also be a mistake to pay them no heed.

Ultimately, the preclearance provisions are real pieces of federal legislation. They were forged in a messy legislative process, both like and unlike other negotiations creating other legislation. They are premised on specific enumerated constitutional powers, and arise out of a real historical context that is both legally and politically relevant. They are designed to address problems that the pragmatic realities of litigation make resistant to other remedial provisions. They are applied by courts and agencies to real electoral policies in individual circumstances characterized by different demographic and political realities, and they produce tangible results that affect real people in measurable ways. All of that is more than enough real substance for the grist of a constitutional challenge. There is no need to litigate the validity of a simulacrum instead.

Justin Levitt is a Visiting Associate Professor of Law at Yale Law School and an Associate Professor at Loyola Law School in Los Angeles. He received his B.A. from Harvard University in 1995, his M.P.A. from the Harvard Kennedy School in 2002 and his J.D. from Harvard Law School in 2002. He would like to thank Heather Gerken for her helpful comments refining the frame of the piece, and Joey Fishkin, Dale Ho, and Spencer Overton for their insightful suggestions. All errors, of course, are the author’s own.

Preferred citation: Justin Levitt, Section 5 as Simulacrum, 123 Yale L.J. Online 151 (2013),