maize13855464.0001.001 in

    13. Justice Ginsburg’s Umbrella

    I. Introduction

    Near the end of her dissent in Shelby County v. Holder[1], Justice Ginsburg suggested a simple analogy to illustrate why the regional protections of the Voting Rights Act (VRA) were still necessary. She wrote that “[t]hrowing out preclearance when it has worked and is continuing to work to stop discriminatory changes is like throwing away your umbrella in a rainstorm because you are not getting wet.”[2]

    The image went viral in the aftermath of the decision. It appeared in media accounts, academic commentary, fundraising appeals, and sundry blogs. And for good cause. The image crisply captured why the VRA’s supporters[3] believed the preclearance regime remained necessary and why they thought scrapping it would be so damaging. It is still raining, they had been urging, and the umbrella the VRA offers continues to provide critical protection. Throw out that umbrella, the argument went, and lots of people are sure to get soaked.[4]

    Curiously, the Shelby County majority seemed to agree. Chief Justice Roberts’s opinion for the Court held § 4(b) of the VRA unconstitutional and thereby rendered the preclearance regime inoperative.[5] But while the Chief Justice discarded the umbrella Justice Ginsburg deemed so important, he never disputed the consequences she said would follow from doing so. Indeed, the reasons he provided for shutting down the preclearance regime suggested that he, too, expected that many people would get wet as a result of the decision.

    It turns out that Chief Justice Roberts and Justice Ginsburg disagreed about a different point entirely. To belabor the analogy—something, be warned, this chapter will do repeatedly—the Justices disagreed about whether getting wet was worse than carrying an umbrella. For the Chief Justice, carrying an umbrella, at least one like the VRA’s preclearance regime, is an extremely costly and damaging activity. By contrast, Justice Ginsburg viewed getting wet as the more damaging experience. She recognized that carrying an umbrella may be inconvenient and even costly but, in her view, well worth the bother. Keeping dry should be the priority.

    The “umbrella” at issue in Shelby County was, without doubt, an unusual one. With its regionally applicable, burden-shifting requirements, the VRA’s preclearance regime has long been understood to be an “exceptional” and “extraordinary” statute.[6] Disagreement among the Justices about its continued use might accordingly be minimized or even dismissed as a regime-specific dispute. And yet, I will argue that the different ways in which the majority and dissent in Shelby County valued getting wet and staying dry exposes a more foundational and far-reaching disagreement.

    Specifically, this chapter presses the idea that Chief Justice Roberts’s willingness to discard Justice Ginsburg’s umbrella reveals a distinct conception of federal antidiscrimination law. It is a conception that sees the existing regime to be a source of unjust enrichment to its beneficiaries. Under this view, the regime does not simply make victims of undeniable discrimination whole but instead places a host of interested parties, victims included, in a decidedly better position than they would have been had the discrimination never occurred. For this reason, the regime is viewed to be a costly and damaging enterprise that should be limited at every opportunity.

    Notably, this conception of federal antidiscrimination law does not deny the persistence of discrimination, and indeed, discrimination of the old-school, unconstitutional variety. To be sure, adherents of this view continue to be concerned that the linkage between challenged conduct and invidious intent has become too attenuated in some, and perhaps most, cases.[7] But their more pressing worry is that the regime today does more harm than the discrimination it presently addresses, even when that discrimination is indisputably unconstitutional or otherwise invidious in nature.

    Put differently, the issue disputed in Shelby County, and in a host of other contemporary civil rights cases, is not about whether people are still getting wet but whether it is worth it to keep them dry. And for a majority of the present Court—and a majority for some time now—keeping dry is no longer cost-justified.

    * * *

    This chapter proceeds as follows. Part I explores why the Shelby County majority discarded Justice Ginsburg’s umbrella. Specifically, it argues that Chief Justice Roberts did not mistake dry conditions for a problem solved but instead implemented a considered preference for getting wet. Part II situates this preference in a larger jurisprudence, in which a majority of the Court has limited the use of the umbrella provided by federal antidiscrimination law even as it acknowledged the rain to be ongoing. Part III argues that these cases represent a distinct strand in the Court’s long-standing antipathy to federal antidiscrimination law, one that seeks to limit the regime based on the belief that it is more costly than the discrimination it presently addresses.

    II. Was It Raining in Shelby County?

    Much of the debate preceding the Court’s decision in Shelby County focused on the conditions for political participation in covered jurisdictions. No one disputed that these conditions had improved markedly since Congress first crafted the statute and that the VRA itself was largely responsible for these improvements.[8] What was disputed was the extent to which these improvements were dependent on the VRA’s continued operation and the degree of backsliding that would occur if the regime were scrapped. In other words, the dispute concerned whether or not the rain had stopped.

    Justice Ginsburg’s umbrella analogy captured what supporters of the VRA had been arguing—namely, that the improved conditions in covered jurisdictions existed only because the preclearance regime actively blocked misconduct where it applied. Under this view, the VRA was not only responsible for improved conditions in covered jurisdictions, but its continued operation was essential to maintaining those conditions. As Justice Ginsburg explained, do not mistake dry conditions under the umbrella for a sunny day.[9]

    The regime’s critics countered that conditions in covered jurisdictions looked better because they were better and that preclearance no longer had much to do with it.[10] After all, some rainy days turn into sunny ones, and when they do, putting away the umbrella makes a lot of sense. It was, notably, this view that animated Chief Justice Roberts’s suggestion in 2009 that the preclearance regime might be nothing more than an elephant whistle, shooing away a nonexistent threat.[11]

    And yet, the Chief Justice’s opinion in Shelby County did not pursue this line of argument. True, it dismissed a defense based on deterrence as analytically flawed, explaining that deterrence could always be invoked to justify the regime even if evidence on the ground suggested the risk of backsliding was negligible.[12] What the opinion did not do, however, was take issue with Justice Ginsburg’s argument that severe backsliding would occur absent the preclearance regime. Indeed, Chief Justice Roberts declined to dispute Justice Ginsburg’s characterization of the evidence. He did not question the scope of unconstitutional conduct she described or the consequences she said would follow from the Court’s ruling in Shelby County.

    Justice Ginsburg described that evidence in detail, and her description made clear that she thought the evidence left no doubt that it was still raining in places like Shelby County, Alabama. She cited numerous examples in which covered jurisdictions violated both the VRA and the Constitution.[13] She observed, moreover, that contemporary unconstitutional conduct in covered jurisdictions remained remarkably widespread even as the evidence showed that the preclearance regime worked to deter and block a good deal of misconduct in covered jurisdictions.[14] Put differently, the evidence showed how preclearance, much like a real umbrella, operated imperfectly as a shield against the rain and that this imperfect protection provided a good indication of what would follow should the umbrella be discarded.

    The Shelby County majority likely viewed the evidence Justice Ginsburg cited more equivocally. Much of it involved dilution claims stemming from redistricting disputes of the sort that once prompted the Chief Justice to lament this “sordid business, this divvying us up by race.”[15] More broadly, the Shelby County majority no doubt suspected that many of the examples cited by Justice Ginsburg and collected in the congressional record sounded more in discriminatory effect than intent or simply tracked a jurisdiction’s inability to disprove animus rather than its affirmative existence.

    But insofar as the Justices in the majority held these suspicions, Chief Justice Roberts opted not to voice them.[16] Rather than take issue with Justice Ginsburg’s characterization of the evidence, the Chief Justice concluded that the discrimination she described as she described it was legally insufficient to justify the statute’s continued regional application.[17] As explanation, he observed that this discrimination was not as severe as it was when Congress first crafted the regime in 1965; that it had not led Congress to alter the statute’s preexisting coverage formula; and that it encompassed subjects different from the ones that Congress listed in the coverage formula when it first subjected places to the regime’s requirements.[18]

    I have explained elsewhere why these observations, all of which are true, should have been insufficient to render preclearance obsolete—and indeed should have been irrelevant—under applicable doctrine that the Shelby County majority did not purport to displace.[19] For present purposes, however, the doctrinal inadequacy of these observations matters less than what they expose about the Court’s toleration for unremedied or inadequately remedied discrimination, including, notably, discrimination that violates the Constitution. And it turns out that the Court is willing to tolerate quite a bit.

    For example, Chief Justice Roberts observed that the discrimination documented in the 2006 record was not as severe as the discrimination that first led Congress to enact the VRA. As he noted, the record evidence did not “show[] anything approaching the ‘pervasive,’ ‘flagrant,’ ‘widespread,’ and ‘rampant’ discrimination that faced Congress in 1965, and that clearly distinguished the covered jurisdictions from the rest of the Nation at that time.”[20]

    The Chief Justice was certainty correct about this. The 2006 record documented a host of ugly incidents but nothing that rose to a level equivalent to the systemic, brazen defiance of constitutional norms that defined the pre-VRA South. Even with the aggressive backsliding now under way in places like Texas and North Carolina, no one expects conditions to deteriorate to the level that prompted Congress to enact the statute in the first place.

    That’s good news, as far as it goes, but it does not explain why contemporary, persistent, and prevalent unconstitutional discrimination in covered jurisdictions should not be remedied in the manner in the manner Congress had selected. The Warren Court had recognized Congress to possess close to plenary authority when crafting remedies for unconstitutional racial discrimination in voting.[21] And even when the Rehnquist Court pulled back, requiring a tight connection between remedies and unconstitutional conduct, it never suggested that some constitutional violations were more worthy of remedy than others.[22] That Court’s concern was with remedies that targeted conduct that was not itself unconstitutional rather than with gradations among constitutional injuries.

    Shelby County, by contrast, appears to stake out a distinction between discrimination of the extreme Jim Crow variety, and the more contained type of unconstitutional conduct we see today. And it suggests that Congress may not select what it reasonably believes is the most effective way to remedy unconstitutional racial discrimination when that discrimination falls short of the type that defined Alabama in 1965.

    Chief Justice Roberts, however, makes clear that the discrimination Justice Ginsburg described was insufficient not simply because Jim Crow–era discrimination was worse. He explained that it was also flawed because it involved problems that were different in kind from those that first prompted Congress to enact the regime. He wrote, “The dissent relies on ‘second-generation barriers,’ which are not impediments to the casting of ballots, but rather electoral arrangements that affect the weight of minority votes.”[23] In other words, the discrimination that Justice Ginsburg described differed from the type of discrimination captured by the original coverage formula, which was “based on voting tests and access to the ballot, not vote dilution.”[24] This difference, apparently, renders contemporary discrimination legally insufficient to justify Congress’s decision to retain the original coverage formula.[25]

    Shelby County rejected the idea that Congress might have rationally concluded that current conditions justified preserving a formula crafted forty years earlier. It thus suggested that Congress needed to revise the coverage formula or craft a new one. Preserving the formula was not a constitutional option. Congress’s decision to preserve the formula could not be “shaped” by the record; only a decision to revise it could be. See Katz, supra note 19, at 330.

    Much like the observation that contemporary discrimination is not as rough as the Jim Crow variety, the Chief Justice’s suggestion that second-generation barriers are off-point demands more explanation than he provided. It is true, of course, that the coverage formula invalidated by Shelby County made no mention of so-called second-generation problems, such as the practice of manipulating district lines to inhibit minority influence, and was based instead on the use of tests and devices and low voter participation.[26] But that fact hardly means that the practices grouped as “second-generation” are unrelated to the concerns Congress meant to target when it crafted the coverage formula. In fact, just the opposite is true.

    So-called second-generation practices predate the VRA by decades and stand with the white primary, the literacy test, the poll tax, and other tactics that were used concurrently in the Jim Crow South to ensure that African American citizens lacked the ability to cast “meaningful”[27] ballots and to “strip” them “of every vestige of influence” in selecting public officials.[28] True, Congress relied on the use of tests and devices as the “trigger” for the original coverage formula, but it did so not in order to limit the statute’s reach. Instead, it selected the specified trigger because it captured with remarkable accuracy the places that engaged in the broader range of conduct (including “second-generation” conduct) that had rendered the Fifteenth Amendment a nullity throughout the pre-VRA South.[29] The statutory trigger linked tests and devices to low participation, but the statute’s target was never so limited.[30]

    The Supreme Court itself recognized as much in 1969.[31] Justice Harlan disagreed at the time[32], and Justices Thomas and Scalia would do so later.[33] But a majority of the Court has repeatedly recognized congressional intent for the VRA to apply to these practices and confirmed Congress’s power to deploy the VRA in this way. Chief Justice Roberts’s opinion in Shelby County nevertheless suggested otherwise but did not explain why.

    What is clear, however, is what the opinion as written accomplished. It brought the preclearance regime to an immediate and, perhaps permanent[34], halt, even as it countenanced evidence of widespread and ongoing discrimination. This discrimination, to be sure, fell short of the Jim Crow norm and did not directly involve those “tests or devices” listed in the original coverage formula. It was discrimination, nevertheless, and a good deal of it ran afoul of the Constitution. Critically, Chief Justice Roberts never suggested otherwise. Indeed, he seemed to agree with Justice Ginsburg when she wrote that it was still raining in covered jurisdictions and that it would continue to rain, predictably, for some time to come. At a minimum, the Chief Justice said nothing that called her forecast into question.

    Therein lies Shelby County’s significance. The decision displays the Court’s willingness to discard an umbrella on a rainy day with full knowledge that rain will continue. It may have been raining harder in the past, and the present rainstorm may (or may not) differ in other ways from what came before. Regardless, the Shelby County majority opted to toss out an umbrella in the middle of a rainstorm, fully aware of what it was doing.

    The umbrella at issue in Shelby County was an unusual one, and it is certainly arguable that the Court’s willingness to discard it rested on its distinctive features. Long considered strong medicine, the VRA’s preclearance requirement reversed the presumption of validity that typically attaches to legislative and administrative action, and presumed instead that public officials in places subject to the requirement were engaged in discrimination unless and until they could convince a federal official otherwise.[35]

    It turned out, however, that this defining aspect of the preclearance regime was less controversial in Shelby County than the regime’s limited geographic reach. The fact that the preclearance obligation existed in some places but not others has always bristled[36], but it had also been thought to contribute to the regime’s legitimacy. Far from a blanket obligation, preclearance had long been seen as a targeted remedy, applying only in places where the need for it was most acute. Indeed, geography was one reason Justice Kennedy once cited the VRA as the paradigmatic example of congruent and proportional legislation.[37]

    That, of course, was nearly two decades ago, and times change, as the Chief Justice has reminded us.[38] Ultimately, it was the regime’s limited geographic application that contributed more directly to its downfall in Shelby County than its burden-shifting requirements.[39] A majority of the Justices found themselves unconvinced that places subject to the statute were sufficiently different from other places to justify their being subjected to the statute’s distinct requirements. And it was this skepticism that may best explain the Court’s willingness to scrap the regime.

    If so, the Shelby County majority’s willingness to discard preclearance in the face of persistent, documented discrimination might be dismissed as a regime-specific move to secure a desired end. Whether the Court was more troubled by the regime’s geographic selectivity or the burden-shifting obligations it imposed, it was convinced that preclearance had to go. Under this view, the holding is consequential, to be sure, but only because the specific statute the decision incapacitated was itself a consequential one, in terms of both its real world effect and the salient place it occupied in the public’s imagination.

    And yet, this reading of Shelby County is not, in my view, the best reading of the decision. Rather than simply charting a one-time path to a desired destination, the Court’s willingness to discard an umbrella in the rain is better understood within a broader jurisprudence, described below.

    III. Is It Raining Elsewhere?

    Far from unique, Shelby County’s tolerance for ongoing discrimination represents a common stance in modern civil rights law. In numerous cases, the Court has limited federal antidiscrimination measures such as the VRA, Title VII, and the Age Discrimination in Employment Act (ADEA) in the face of uncontested evidence of discrimination of the sort the statutes at issue were designed to address. These cases, moreover, all generated dissenting opinions, often written by Justice Ginsburg, which disputed both the holdings and the analytical moves used to reach them. Situated within Justice Ginsburg’s Shelby County construct, these dissenting opinions all argued that an umbrella should be used in the rain, while a majority, time and again, sided with getting wet.

    Consider a few eclectic but representative examples.

    Nassar and Gross: Two days before the Court handed down Shelby County, it held that an employee alleging retaliation under Title VII needed to show that the complaints he lodged about status-based discrimination not only contributed to his being denied a coveted transfer but also were the but-for cause of that denial. University of Texas Southwestern Medical Center v. Nassar[40] relied heavily on the Court’s 2009 ruling in Gross v. FBL Financial Services, Inc., which applied the same rule to an employee alleging age discrimination under the ADEA.[41] Both decisions deemed the employees involved to be ineligible for the more forgiving “motivating factor” analysis Congress set forth in § 703m of the Civil Rights Act of 1991.[42] The Court split 5–4 in both cases, with dissenting opinions by Justices Ginsburg and Stevens challenging the way the majority understood relevant precedent, the purpose and structure of the 1991 CRA, and applicable agency action.[43]

    Amid this disagreement, however, all of the Justices seemed to agree—or, at least, no one denied—that the plaintiff-employees who brought both cases had been subjected to intentional discrimination of the sort the statutes at issue targeted[44]—namely, that Naiel Nassar’s complaints about disparate treatment based on his Middle Eastern descent contributed to his being denied a transfer[45], and that Jack Gross’s age contributed to the restructuring of job responsibilities he challenged.[46] In both cases, then, the Justices seemed well aware and willing to accept that intent-based discrimination had occurred. They split over whether the employees should be entitled to relief given this discrimination.

    Put differently, no one doubted that the employees who brought these cases had been caught in the rain. What they disagreed about was whether they were entitled to the umbrella provided by Title VII and the ADEA. And a majority held they were not.

    Coleman: A year before Nassar, the Court struck down a provision of the Family Medical Leave Act (FMLA) that guaranteed twelve weeks of unpaid medical leave to eligible employees suffering from serious medical conditions.[47] Justice Ginsburg’s dissenting opinion in Coleman v. Court of Appeals of Maryland argued that the sex-neutral “self-care” provision constituted an essential part of a comprehensive statutory regime that included the family care provisions that the Court had already upheld.[48] She explained that Congress crafted this regime to address the pervasive discrimination women confronted in the workplace stemming from pregnancy-related issues and more general sex stereotypes about family care responsibilities.[49]

    Justice Ginsburg did not invoke umbrella imagery in Coleman, but she might easily have employed it. Her argument, at bottom, was that the umbrella provided by the FMLA would have a gaping hole in it without the statute’s self-care provision. The self-care provision, she explained, “serves to blunt the force of stereotypes of women as primary caregivers by increasing the odds that men and women will invoke the FMLA’s leave provisions in near-equal numbers.”[50]

    A majority of the Court, however, was unmoved. Justice Kennedy’s plurality opinion held that the relationship between the self-care provision and the discrimination Justice Ginsburg described was too complex and attenuated to satisfy constitutional scrutiny.[51] And yet, much like Chief Justice Roberts’s opinion in Shelby County, Justice Kennedy’s opinion in Coleman did not question the prevalence of the discrimination Justice Ginsburg described.[52] Like the Chief Justice, moreover, Justice Kennedy opted to discard the umbrella Congress crafted to address that discrimination.

    Ledbetter: In 2007, a majority of the Court held that an employee’s claim for sex-based wage discrimination was time barred because she filed suit long after the employer’s initial discriminatory wage decision. Justice Alito’s majority opinion in Ledbetter v. Goodyear Tire & Rubber Co.[53] held that the statute of limitations ran from that initial decision rather than from the issuance of subsequent paychecks, the amount of which reflected the initial discrimination.[54]

    Justice Ginsburg’s dissent argued that the statute was better read to allow Ledbetter’s suit to proceed, an argument she might easily have bolstered by invoking the umbrella image she employed in Shelby County. Indeed, Justice Ginsburg’s dissent made clear that Goodyear had been raining on Lilly Ledbetter for a very long time and that Title VII should be available to provide her relief. Justice Ginsburg closed her opinion calling for a statutory amendment to reverse the majority’s ruling, a call Congress heeded.[55]

    The majority in Ledbetter was not persuaded by Justice Ginsburg’s argument, but it never questioned that Goodyear had intentionally discriminated against Ledbetter based on sex by paying her less than both similarly situated and less-qualified male colleagues. Justice Alito expressed no doubt about this point. But in his view, Ledbetter’s failure to use the Title VII umbrella at the beginning of the storm precluded her from using it later.

    Bossier Parish: Like Ledbetter, Reno v. Bossier Parish School Board was reversed by subsequent statutory amendment (albeit one that was later ruled to be unconstitutional).[56] Back in 2000, Justice Scalia’s majority opinion held that the VRA permitted implementation of a districting plan in which African American voters constituted a majority in none of the plan’s twelve electoral districts. It was alleged and, somewhat surprisingly, stipulated that the School Board had refused to draw a majority-minority district because it wanted to prevent an African American candidate from being elected to the board.[57]

    Justice Scalia’s majority opinion held that the redistricting plan could be implemented notwithstanding this discriminatory purpose. The opinion explained that § 5 of the VRA did not block implementation of electoral changes enacted with discriminatory intent. Instead, it blocked only that subset of electoral changes enacted with “retrogressive” intent—that is, the intent to make things worse for the minority group in question. Mere animus would not suffice.[58] The opinion, moreover, suggested any rule to the contrary would raise a serious constitutional question.[59]

    Justice Souter’s dissent argued that electoral changes enacted with a discriminatory, albeit not “retrogressive” purpose, fell within the § 5 proscription. As he put it, blocking implementation of unconstitutional conduct of this sort—the rain Justice Ginsburg subsequently described—was precisely what Congress had designed § 5 to address and what Congress, in his view, had ample power to mandate.[60]

    Thirteen years later, Shelby County made clear that Congress lacks this power. The decision viewed Congress’s 2006 decision to adopt Justice Souter’s Bossier Parish reading as evidence of constitutional overreach. It thereby suggested Congress had no power to include within the statutory proscription conduct that was unconstitutional.[61]

    Bossier Parish, Ledbetter, Coleman, Gross, and Nassar are, without doubt, distinguishable from one another on numerous grounds. Yet they share a defining characteristic that makes them representative examples of a more general stance in federal civil rights law. Like Shelby County, these decisions all circumscribe the federal regime in contexts in which the occurrence of intentional, invidious, and even unconstitutional conduct is left unquestioned. Placed within Justice Ginsburg’s Shelby County framework, these cases all involved rain; the Court knew it, and a majority was nevertheless steadfast that an umbrella should not be used.

    Admittedly, likening the discrimination observed in these cases to rain is a contestable move. My premise is that discrimination may be distinguished from liability, at least in certain contexts, and that we learn something by making this distinction. That premise accordingly rejects the idea that discrimination is necessarily or most usefully understood as a legal conclusion that is coextensive with liability. Instead, it posits that people like Naiel Nassar and Lilly Ledbetter found themselves in the rain even though the Court ruled against them. They lost despite the fact that it was raining and decidedly not because the Court thought the sky was clear.

    Understanding the cases in this manner—that is, by parsing discrimination from liability—brings into focus a distinct strand of civil rights jurisprudence. To be sure, judicial skepticism toward the federal civil rights regime is nothing new, and the Court has long sought to scale back federal antidiscrimination law. Decisions that do so in the face of uncontested evidence of intentional discrimination are undoubtedly part of this effort. And yet, my claim is that they are a distinct component of it. Unlike those cases that deny relief by deeming challenged conduct to be nondiscriminatory, these decisions discard the umbrella even as the need for it persists. The suggestion is that the federal civil rights project, while hardly complete, is no longer worth pursuing. Rather than a mission accomplished, it is a mission abandoned.

    Of course, not every decision circumscribing the federal civil rights regime falls decisively into one group or the other. Some deny the rain, or at least express skepticism about it, but also voice mistrust about using the available umbrella should the rain alleged actually be falling.[62] Elsewhere, however, the distinction is clear, with a growing number of decisions displaying a willingness to discard the umbrella in the rain knowing full well people will get wet as a result.

    Shelby County’s willingness to immobilize § 5 of the VRA without disputing the discrimination Justice Ginsburg described is part of this latter group of decisions. Far from unique, Shelby County stands with a host of other decisions that acknowledge discrimination persists and yet posit that core elements of the federal civil rights project are no longer worth pursuing. The next section explores why this sensibility drives so much of contemporary antidiscrimination law.

    IV. On Unjust Enrichment and Harmless Error

    Decisions that limit federal antidiscrimination law typically view the regime’s broader application as deeply problematic. Among the concerns most often cited is the worry that an expansive approach to the regime encourages frivolous lawsuits, exposing employers and other defendants to wasteful litigation costs and spurring inefficient defensive decision making.[63] Curb the regime, it is argued, lest undeserving plaintiffs be unjustly enriched at the expense of diligent defendants and, in many circumstances, the rest of us.

    Animating this concern is the suspicion that frivolous claims outnumber legitimate ones and that the discrimination federal antidiscrimination law was crafted to address is largely a thing of the past. Unsurprisingly, decisions that find challenged conduct to be nondiscriminatory highlight this sensibility, with the dispute at hand seen either to involve a frivolous claim or to suggest circumstances in which one might find expression.[64]

    Less expected, decisions that deny relief in the face of uncontested discrimination also voice concern that frivolous claims are rising as genuine discrimination declines. These decisions acknowledge the rain but deem it insufficiently worrisome to warrant use of the umbrella at issue.[65] More pressing is the need to check the regime and guard against its unjustified application.

    Under this view, victims of documented discrimination might be understood or even dismissed as unfortunate, but unavoidable, collateral damage sacrificed for the greater good. And yet, it is not the unpleasantness of the image that keeps it off the pages of U.S. reports but instead the belief that the victims of discrimination in these cases have not been significantly damaged at all. True, they have been caught in the rain, but these decisions suggest that getting wet may not be as damaging as some seem to think. In fact, they suggest it might not be damaging at all.

    With this suggestion, Shelby County and cases like it shift the terrain on which civil rights disputes have long been fought. Rather than contest allegations or evidence of discrimination, they dismiss discrimination itself as inconsequential. They reject Justice Ginsburg’s belief that getting caught in the rain is the source of enduring damage and, in its place, insert the idea of harmless error into civil rights jurisprudence.

    That idea, in turn, has led the Court to view much of federal antidiscrimination law as providing a windfall to its beneficiaries. Far from making victims of discrimination whole, the regime is seen as leaving them in a decidedly better position than they would have been had they never gotten wet. The umbrella Justice Ginsburg thinks provides vital protection is seen instead to be a source of unjust enrichment to those it shields. The resulting project consequently becomes one dedicated to limiting use of the umbrella whenever possible, rain notwithstanding.

    Hence, the recent mixed-motive decisions requiring plaintiffs to show “but-for” causation work hard to make sure that getting wet is not the vehicle for getting ahead. These decisions hold that if, absent the discrimination alleged, the plaintiff would have been denied the disputed promotion or transfer, the discrimination itself should not be the source of liability.[66] Because Title VII’s “motivating” factor rule allows for liability in such circumstances[67], Nassar and Gross rejected it, finding the plaintiff-employees ineligible for both the acknowledgment of wrongdoing a liability ruling embodies and the attorneys’ fees that accompany it. Of no moment was the fact that the rejected approach barred injunctive relief when the desired transfer or promotion would have otherwise been denied.[68]

    As telling, and perhaps even more so, is a little noted aspect of Chief Justice Roberts’s opinion in Shelby County, in which he described the 2006 amendment overruling Bossier Parish to “prohibit laws that could have favored [minority voters] but did not do so because of a discriminatory purpose.”[69] The words “could have favored” are revealing. Far from unartful drafting, they suggest that the Bossier Parish School Board did not injure minority voters when it adopted a districting plan avowedly designed to prevent the election of an African American representative. Instead, Shelby County suggests that the school board’s unconstitutional conduct only blocked adoption of a plan “that could have favored” black voters. The broader suggestion is that unconstitutional discrimination does not necessarily deny minority voters an equal opportunity to participate in the political process but instead may simply deprive them of favored or preferential treatment.

    That suggestion is a remarkable one, and one that documents the extent to which the locus of civil rights jurisprudence has shifted. A longstanding worry in this realm has been the concern that the prohibition on disparate impact would devolve into a mandate for affirmative action and prompt potential defendants to adopt preferential policies in order to shield themselves from liability.[70] To guard against this result (and its apparent conflict with explicit statutory language), the Court has long refused to read bans on disparate impact expansively.

    But now, this concern about preferential treatment is also shaping the Court’s approach to discriminatory intent. In a growing number of cases, it has read the VRA, Title VII, and other federal civil rights measures narrowly in contexts where animus was evident (or at least evidence of it went unchallenged), and it has done so because more expansive statutory readings were thought to yield unwarranted preferential treatment. Notably absent from these cases is a well-intentioned defendant laboring to comply with a statutory mandate. Instead, the Court has come to see federal antidiscrimination law itself as the source of damaging preferences. Even the prohibition on invidious intent, the core tenet of federal civil rights law, has evolved into a problem and hence a target. It is what needs to be constrained, if not eliminated, while the conduct the regime once targeted is dismissed as harmless and those once understood as victims are transformed into the unjustly enriched.

    V. Conclusion

    The Court has long sought to scale back the federal civil rights regime and has typically done so by characterizing challenged conduct as nondiscriminatory. This chapter tracks a distinct line of cases that are undeniably part of the larger effort but that limit the regime while recognizing discrimination rather than denying it. These decisions throw out an umbrella in a rainstorm, knowing full well it is raining and that the rain will continue. They accordingly posit that the rain does less damage than the umbrella, at least in certain circumstances, and that the Court is institutionally able to figure out the circumstances in which the umbrella should be discarded.

    About the Author

    Ralph W. Aigler Professor of Law, University of Michigan Law School. Many thanks to Cali Cope-Kasten for excellent research assistance.

    Notes

    1. Shelby Cty. v. Holder, 570 U.S. , 133 S. Ct. 2612 (2013).return to text

    2. Id. at 2650 (Ginsburg, J., dissenting).return to text

    3. This includes me. See Brief of Amici Curiae Ellen D. Katz and the Voting Rights Initiative in Support of Respondents, Shelby Cty. v. Holder, 133 S. Ct. 2612, 2013 WL 457386 (2013).return to text

    4. See, e.g., Brief for the Federal Respondent, Shelby Cty. v. Holder, 133 S. Ct. 2612, 2013 WL 315242 (2013); Brief for Respondent-Intervenors Earl Cunningham et al., Shelby Cty. v. Holder, 2013 WL 315241, 133 S. Ct. 2612 (2013); Brief for Respondent-Intervenors Bobby Pierson et al., Shelby Cty. v. Holder, 133 S. Ct. 2612, 2013 WL 325379 (2013).return to text

    5. Section 4(b) of the VRA “covered” jurisdictions if they utilized a “test or device” as a prerequisite to voting and had low levels of voter participation on specified dates between 1964 and 1972. See 42 U.S.C. § 1973b(b) (2006), transferred to 52 U.S.C. § 10303 [hereinafter “§ 4(b)”]. Once covered, jurisdictions could no longer use their test or device and could not implement any electoral changes without first showing that the proposed change would be nondiscriminatory. See 42 U.S.C. § 1973c (2006), transferred to 52 U.S.C. § 10303, [hereinafter “§ 5”]. This preclearance obligation applied only to jurisdictions covered by Section 4(b). See id. As a result, eliminating § 4(b) dissolved all existing obligations to seek preclearance.return to text

    6. See, e.g., South Carolina v. Katzenbach, 383 U.S. 301, 334 (1966) (“[Section 5] may have been an uncommon exercise of congressional power…but the Court has recognized that exceptional conditions can justify legislative measures not otherwise appropriate.”); Reno v. Bossier Parish Sch. Bd., 520 U.S. 471, 501 (1997) (“Section 5…was highly controversial because it imposed novel, extraordinary remedies in certain areas where discrimination had been most flagrant.”).return to text

    7. See, e.g., Nw. Austin Mun. Util. Dist. No. One [NAMUDNO] v. Holder, 557 U.S. 193, 226–29 (2009) (Thomas, J., concurring).return to text

    8. See, e.g., Shelby Cty., 133 S. Ct. at 2625–26; Brief for Petitioner at 9–12, Shelby Cty., 133 S. Ct. 2612 (2013).return to text

    9. SeeShelby Cty., 133 S. Ct. at 2650–51 (Ginsburg, J., dissenting); Brief for the Hon. Congressman John Lewis as Amicus Curiae in Support of Respondents and Intervenor-Respondents at 24–38, Shelby Cty. v. Holder, 2013 WL 476051, 133 S. Ct. 2612 (2013) (describing the importance of § 5).return to text

    10. Transcript of Oral Argument at 30–35, 65–68, Shelby Cty. v. Holder, 133 S. Ct. 2612 (2013).return to text

    11. Transcript of Oral Argument at 28, NAMUDNO, 557 U.S. 193 (2009).return to text

    12. See Shelby Cty., 133 S. Ct. at 2627.return to text

    13. See id. at 2639–43, 2646–47 (Ginsburg, J., dissenting).return to text

    14. See id. at 2639–42 (Ginsburg, J., dissenting).return to text

    15. League of United Latin American Citizens [LULAC] v. Perry, 548 U.S. 399, 511 (2006) (Roberts, C.J., concurring). Meanwhile, Justices Thomas and Scalia think these claims are not judicially cognizable. SeeLULAC, 548 U.S. at 511–12 (Scalia, J., dissenting in part); Holder v. Hall, 512 U.S. 874, 891–930, 944–46 (1994) (Thomas, J., concurring).return to text

    16. Nor did Justice Thomas. His concurring opinion argued that the majority’s analysis left the preclearance requirement not simply moot but unconstitutional as well. Still, he did not delve into the evidence. Shelby Cty., 133 S. Ct. at 2631–32 (Thomas, J., concurring).return to text

    17. See Shelby Cty., 133 S. Ct. at 2619, 2629–31.return to text

    18. Shelby Cty., 133 S. Ct. at 2629–31.return to text

    19. See Ellen D. Katz, What Was Wrong with the Record?, 12 Election L.J. 329 (2013).return to text

    20. Shelby Cty., 133 S. Ct. at 2629.return to text

    21. Katzenbach v. Morgan, 384 U.S. 641 (1966); South Carolina v. Katzenbach, 383 U.S. 301 (1966).return to text

    22. City of Boerne v. Flores, 521 U.S. 507 (1997).return to text

    23. Shelby Cty., 133 S. Ct. at 2629.return to text

    24. Id.return to text

    25. Id. In a related objection, the majority opinion observed that when Congress opted to “reenact[]” the original coverage formula when it reauthorized the VRA in 2006, it “did not use the record it compiled to shape a coverage formula grounded in current conditions.” Id. Part of this problem concerned the substance of the discrimination documented, discussed above. Part, however, was procedural, reflecting the Court’s unhappiness with Congress’s decision to preserve the coverage formula rather than revise it or craft a new one.return to text

    26. See Katz, supra note 19, at 331.return to text

    27. Terry v. Adams, 345 U.S. 461, 476 (1953) (Frankfurter, J., concurring).return to text

    28. Id. at 470. See generally Quiet Revolution in the South: The Impact of the Voting Rights Act, 1965–1990 (Chandler Davidson and Bernard Grofman eds., 1994).return to text

    29. See Brief for the Federal Respondent at 48–50, Shelby Cty., 133 S. Ct. 2612 (2013).return to text

    30. Allen v. State Bd. of Elections, 393 U.S. 544, 566–67 (1969) (Preclearance was meant “‘to be all-inclusive of any kind of practice,’” and to be given “the broadest possible scope.”).return to text

    31. Id. at 567.return to text

    32. Id. at 582–91 (Harlan, J., dissenting in part).return to text

    33. See Holder v. Hall, 512 U.S. 874, 892–903 (Thomas, J., and Scalia, J., concurring).return to text

    34. The decision, notably, falls short of holding that Congress cannot reach practices like racial vote dilution with a remedy like preclearance. The counterfactual suggestion that Congress did not intend a statute of this breadth, theoretically, allows a legislative response. Replacements have been proposed, see https://www.congress.gov/bill/114th-congress/house-bill/2867; http://beta.congress.gov/bill/113th-congress/house-bill/3899, that are of more limited reach than the preclearance regime they purport to replace. Should a bill of this sort become law, the new regime will likely be challenged for the reasons Justice Thomas provides in his concurrence. See Shelby Cty. v. Holder, 133 S. Ct. at 2612, 2631–32 (Thomas, J., concurring).return to text

    35. See South Carolina v. Katzenbach, 383 U.S. 301, 327 (1966).return to text

    36. See id. at 358–61 (Black, J., dissenting).return to text

    37. See City of Boerne v. Flores, 521 U.S. 507, 532–33 (1997).return to text

    38. See Shelby Cty., 133 S. Ct. at 2625–26; NAMUDNO, 557 U.S. at 203–04, 211.return to text

    39. See Shelby Cty., 133 S. Ct. at 2623–24, 2628–31.return to text

    40. Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570 U.S. , 133 S. Ct. 2517 (2013).return to text

    41. See Gross v. FBL Fin. Servs., Inc., 557 U.S. 167 (2009).return to text

    42. 42 U.S.C. § 2000e-2(m) (“Except as otherwise provided in this subchapter, an unlawful employment practice is established when the complaining party demonstrates that race, color, religion, sex, or national origin was a motivating factor for any employment practice, even though other factors also motivated the practice.”).return to text

    43. Nassar, 133 S. Ct. at 2534–47 (Ginsburg, J., dissenting); Gross, 557 U.S. at 180–90 (Stevens, J., dissenting).return to text

    44. See Jackson v. Birmingham Bd. of Educ., 544 U.S. 167, 173 (2005) (describing retaliation as a “form of intentional [status-based] discrimination”).return to text

    45. Nassar, 133 S. Ct. at 2532.return to text

    46. Gross, 557 U.S. at 170–72.return to text

    47. Coleman v. Ct. of Appeals of Md., 566 U.S. , 132 S. Ct. 1327 (2012) (striking down 29 U.S.C. § 2612(a)(1)(D) (2009)).return to text

    48. Id. at 1342 (Ginsburg, J., dissenting). See also Nev. Dep’t of Human Res. v. Hibbs, 538 U.S. 721 (2003) (upholding provisions of the FMLA).return to text

    49. Coleman, 132 S. Ct. at 1340–49 (Ginsburg, J., dissenting).return to text

    50. Id. at 1348 (Ginsburg, J., dissenting) (internal quotations omitted) (quoting Brief for Nat’l P’ship for Women & Families et al. as Amici Curiae Supporting Petitioner 26, Coleman v. Ct. of Appeals of Md., 132 S. Ct. 1327 (2012)).return to text

    51. Coleman, 132 S. Ct. at 1334–37.return to text

    52. See id. at 1334 (“In enacting the FMLA, Congress relied upon evidence of a well-documented pattern of sex-based discrimination in family-leave policies. States had facially discriminatory leave policies that granted longer periods of leave to women than to men.”); Shelby Cty., 133 S. Ct. at 2619 (“[V]oting discrimination still exists; no one doubts that.”).return to text

    53. 550 U.S. 618 (2007), superseded by statute, Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5 (codified in scattered sections of 29 U.S.C. and 42 U.S.C.).return to text

    54. Ledbetter, 550 U.S. at 639–42.return to text

    55. See id. at 661 (Ginsburg, J., dissenting). See also The Lilly Ledbetter Fair Pay Act of 2009, Pub. L. No. 111-2, 123 Stat. 5 (providing that the 180 statutes of limitations for filing an equal pay lawsuit regarding pay discrimination resets each time the employer issues a paycheck that reflects the pay discrimination).return to text

    56. See Reno v. Bossier Parish Sch. Bd. [Bossier Parish II], 528 U.S. 320 (2000), superseded by statute, Fannie Lou Hamer, Rosa Parks, and Coretta Scott King Voting Rights Act Reauthorization and Amendments Act of 2006, Pub. L. No. 109-246, 120 Stat. 577 § 2(b)(6) (2006) (declaring that Bossier Parish II “misconstrued Congress’ original intent in enacting the Voting Rights Act of 1965 and narrowed the protections afforded by section 5 of such Act”); Shelby Cty., 133 S. Ct. at 2626–27.return to text

    57. See Jurisdictional Statement at 8, Bossier Parish II, 528 U.S. 320 (2000) (“There was evidence that several Board members preferred the [redistricting plan with all white-majority districts] because they did not want black representation on the Board. Board member Barry Musgrove said that ‘the Board was “hostile” toward the idea of a black majority district.’”). It was alleged, and assumed for purposes of the litigation, that § 2 of the VRA required the creation of such a district. Bossier Parish I held even if that was true, a § 2 violation was not grounds to deny preclearance under § 5. See Reno v. Bossier Parish Sch. Bd. [Bossier Parish I], 520 U.S. 471, 483 (1997) (“[A] violation of § 2 is not grounds in and of itself for denying preclearance under § 5.”).return to text

    58. The decision meant the Bossier Parish School Board could implement its redistricting plan because the evidence did not suggest the Board had been seeking to make political participation more difficult for African American voters in the parish. Because no African American candidate had been elected to the school board, the Board members’ intent to keep such a candidate from winning did not involve an intent to make things worse for black voters in Bossier Parish. See Bossier Parish II, 528 U.S. at 340–41.return to text

    59. Bossier Parish II, 528 U.S. at 336.return to text

    60. Id. at 361, 363–67 (Souter, J., dissenting).return to text

    61. Shelby Cty., 133 S. Ct. at 2626–27.return to text

    62. An example is the 2009 decision in Bartlett v. Strickland, 556 U.S. 1 (2009), in which the Court held that § 2 of the VRA does not require the aggregation of minority voters too few in numbers to constitute a majority in a single-member district. Under certain conditions, minority voters comprising less than 50 percent of a district’s electorate are able to elect candidates of choice by forming coalitions with like-minded white voters. Bartlett, 556 U.S. at 3, 14–15. Justice Souter’s dissent in Bartlett argued that § 2 should be used to foster such coalitions, id. at 27 (Souter, J., dissenting), but Justice Kennedy’s plurality opinion held that the very prospect of such coalitions meant the failure to draw districts that fostered them could not be discriminatory. That is, a refusal to draw the district alleged was not discriminatory within the meaning of § 2. Id. at 14–15 (plurality opinion). Justice Kennedy then went on to explain why reading § 2 to require such districts gave rise to a host of concerns that counseled against adopting that reading, regardless of how the challenged conduct might be characterized. Id. at 20–22. See also Ricci v. DeStefano, 557 U.S. 557, 579–84 (2009) (plurality opinion).return to text

    63. See, e.g., St. Mary’s Honor Ctr. v. Hicks, 509 U.S. 502 (1993); Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989).return to text

    64. See, e.g., Hicks, 509 U.S. 502; Wards Cove, 490 U.S. 642.return to text

    65. See supra notes 40–62 and accompanying text.return to text

    66. See supra notes 40–46 and accompanying text.return to text

    67. 42 U.S.C. § 2000e-2(m).return to text

    68. See Univ. of Tex. Sw. Med. Ctr. v. Nassar, 570 U.S. , 133 S. Ct. 2517 (2013); Gross v. FBL Fin. Servs., Inc., 557 U.S. 167 (2009).return to text

    69. 133 S. Ct. at 2626–27 (emphasis added).return to text

    70. See, e.g., Ricci v. DeStefano, 557 U.S. 557 (2009) (plurality opinion); Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989); Furnco Constr. Corp. v. Waters, 438 U.S. 567 (1978). See also Johnson v. Transp. Agency of Santa Clara County, 480 U.S. 616, 670 (1987) (Scalia, J., dissenting).return to text