Schuette v. Coalition to Defend Affirmative Action, 572 U.S. 291 (2014)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

SUPREME COURT OF THE UNITED STATES

_________________

No. 12–682

_________________

BILL SCHUETTE, ATTORNEY GENERAL OF MICHI-GAN, PETITIONER v. COALITION TO DEFEND AF-FIRMATIVE ACTION, INTEGRATION AND IMMI-GRANT RIGHTS AND FIGHT FOR EQUALITYBY ANY MEANS NECESSARY (BAMN), et al.

on writ of certiorari to the united states court of appeals for the sixth circuit

[April 22, 2014]

     Justice Kennedy announced the judgment of the Court and delivered an opinion, in which The Chief Justice and Justice Alito join.

     The Court in this case must determine whether an amendment to the Constitution of the State of Michigan, approved and enacted by its voters, is invalid under the Equal Protection Clause of the Fourteenth Amendment to the Constitution of the United States.

     In 2003 the Court reviewed the constitutionality of two admissions systems at the University of Michigan, one for its undergraduate class and one for its law school. The undergraduate admissions plan was addressed in Gratz v. Bollinger, 539 U. S. 244 . The law school admission plan was addressed in Grutter v. Bollinger, 539 U. S. 306 . Each admissions process permitted the explicit consideration of an applicant’s race. In Gratz, the Court invalidated the undergraduate plan as a violation of the Equal Protection Clause. 539 U. S., at 270. In Grutter, the Court found no constitutional flaw in the law school admission plan’s more limited use of race-based preferences. 539 U. S., at 343.

     In response to the Court’s decision in Gratz, the university revised its undergraduate admissions process, but the revision still allowed limited use of race-based preferences. After a statewide debate on the question of racial preferences in the context of governmental decisionmaking, the voters, in 2006, adopted an amendment to the State Constitution prohibiting state and other governmental entities in Michigan from granting certain preferences, including race-based preferences, in a wide range of actions and decisions. Under the terms of the amendment, race-based preferences cannot be part of the admissions process for state universities. That particular prohibition is central to the instant case.

     The ballot proposal was called Proposal 2 and, after it passed by a margin of 58 percent to 42 percent, the resulting enactment became Article I, §26, of the Michigan Constitution. As noted, the amendment is in broad terms. Section 26 states, in relevant part, as follows:

     “(1) The University of Michigan, Michigan State University, Wayne State University, and any other public college or university, community college, or school district shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of      public employment, public education, or public contracting.

     “(2) The state shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.

     “(3) For the purposes of this section ‘state’ includes, but is not necessarily limited to, the state itself, any city, county, any public college, university, or community college, school district, or other political subdivision or governmental instrumentality of or within the State of Michigan not included in sub-section 1.”

     Section 26 was challenged in two cases. Among the plaintiffs in the suits were the Coalition to Defend Affirmative Action, Integration and Immigrant Rights and Fight for Equality By Any Means Necessary (BAMN); students; faculty; and prospective applicants to Michigan public universities. The named defendants included then-Governor Jennifer Granholm, the Board of Regents of the University of Michigan, the Board of Trustees of Michigan State University, and the Board of Governors of Wayne State University. The Michigan Attorney General was granted leave to intervene as a defendant. The United States District Court for the Eastern District of Michigan consolidated the cases.

     In 2008, the District Court granted summary judgment to Michigan, thus upholding Proposal 2. BAMN v. Regents of Univ. of Mich., 539 F. Supp. 2d 924. The District Court denied a motion to reconsider the grant of summary judgment. 592 F. Supp. 2d 948. A panel of the United States Court of Appeals for the Sixth Circuit reversed the grant of summary judgment. 652 F. 3d 607 (2011). Judge Gibbons dissented from that holding. Id., at 633–646. The panel majority held that Proposal 2 had violated the principles elaborated by this Court in Washington v. Seattle School Dist. No. 1, 458 U. S. 457 (1982) , and in the cases that Seattle relied upon.

     The Court of Appeals, sitting en banc, agreed with the panel decision. 701 F. 3d 466 (CA6 2012). The majority opinion determined that Seattle “mirrors the [case] before us.” Id., at 475. Seven judges dissented in a number of opinions. The Court granted certiorari. 568 U. S. ___ (2013).

     Before the Court addresses the question presented, it is important to note what this case is not about. It is not about the constitutionality, or the merits, of race-conscious admissions policies in higher education. The consideration of race in admissions presents complex questions, in part addressed last Term in Fisher v. University of Texas at Austin, 570 U. S. ––– (2013). In Fisher, the Court did not disturb the principle that the consideration of race in admissions is permissible, provided that certain conditions are met. In this case, as in Fisher, that principle is not challenged. The question here concerns not the permissibility of race-conscious admissions policies under the Constitution but whether, and in what manner, voters in the States may choose to prohibit the consideration of racial preferences in governmental decisions, in particular with respect to school admissions.

     This Court has noted that some States have decided to prohibit race-conscious admissions policies. In Grutter, the Court noted: “Universities in California, Florida, and Washington State, where racial preferences in admissions are prohibited by state law, are currently engaged in experimenting with a wide variety of alternative approaches. Universities in other States can and should draw on the most promising aspects of these race-neutral alternatives as they develop.” 539 U. S., at 342 (citing United States v. Lopez, 514 U. S. 549, 581 (1995) (Kennedy, J., concurring) (“[T]he States may perform their role as laboratories for experimentation to devise various solutions where the best solution is far from clear”)). In this way, Grutter acknowledged the significance of a dialogue regarding this contested and complex policy question among and within States. There was recognition that our federal structure “permits ‘ innovation and experimentation’ ” and “enables greater citizen ‘ involvement in democratic processes.’ ” Bond v. United States, 564 U. S. –––, ––– (2011) (slip op., at 9) (quoting Gregory v. Ashcroft, 501 U. S. 452, 458 (1991) ). While this case arises in Michigan, the decision by the State’s voters reflects in part the national dialogue regarding the wisdom and practicality of race-conscious admissions policies in higher education. See, e.g., Coalition for Economic Equity v. Wilson, 122 F. 3d 692 (CA9 1997).

     In Michigan, the State Constitution invests independent boards of trustees with plenary authority over public universities, including admissions policies. Mich. Const., Art. VIII, §5; see also Federated Publications, Inc. v. Board of Trustees of Mich. State Univ., 460 Mich. 75, 86–87, 594 N. W. 2d 491, 497 (1999). Although the members of the boards are elected, some evidence in the record suggests they delegated authority over admissions policy to the faculty. But whether the boards or the faculty set the specific policy, Michigan’s public universities did consider race as a factor in admissions decisions before 2006.

     In holding §26 invalid in the context of student admissions at state universities, the Court of Appeals relied in primary part on Seattle, supra, which it deemed to control the case. But that determination extends Seattle’s holding in a case presenting quite different issues to reach a conclusion that is mistaken here. Before explaining this further, it is necessary to consider the relevant cases that preceded Seattle and the background against which Seattle itself arose.

     Though it has not been prominent in the arguments of the parties, this Court’s decision in Reitman v. Mulkey, 387 U. S. 369 (1967) , is a proper beginning point for discussing the controlling decisions. In Mulkey, voters amended the California Constitution to prohibit any state legislative interference with an owner’s prerogative to decline to sell or rent residential property on any basis. Two different cases gave rise to Mulkey. In one a couple could not rent an apartment, and in the other a couple were evicted from their apartment. Those adverse actions were on account of race. In both cases the complaining parties were barred, on account of race, from invoking the protection of California’s statutes; and, as a result, they were unable to lease residential property. This Court concluded that the state constitutional provision was a denial of equal protection. The Court agreed with the California Supreme Court that the amendment operated to insinuate the State into the decision to discriminate by encouraging that practice. The Court noted the “immediate design and intent” of the amendment was to “establis[h] a purported constitutional right to privately discriminate.” Id., at 374 (internal quotation marks omitted and emphasis deleted). The Court agreed that the amendment “expressly authorized and constitutionalized the private right to discriminate.” Id., at 376. The effect of the state constitutional amendment was to “significantly encourage and involve the State in private racial discriminations.” Id., at 381. In a dissent joined by three other Justices, Justice Harlan disagreed with the majority’s holding. Id., at 387. The dissent reasoned that California, by the action of its voters, simply wanted the State to remain neutral in this area, so that the State was not a party to discrimination. Id., at 389. That dissenting voice did not prevail against the majority’s conclusion that the state action in question encouraged discrimination, causing real and specific injury.

     The next precedent of relevance, Hunter v. Erickson, 393 U. S. 385 (1969) , is central to the arguments the respondents make in the instant case. In Hunter, the Court for the first time elaborated what the Court of Appeals here styled the “political process” doctrine. There, the Akron City Council found that the citizens of Akron consisted of “ ‘people of different race[s], . . . many of whom live in circumscribed and segregated areas, under sub-standard unhealthful, unsafe, unsanitary and overcrowded conditions, because of discrimination in the sale, lease, rental and financing of housing.’ ” Id., at 391. To address the problem, Akron enacted a fair housing ordinance to prohibit that sort of discrimination. In response, voters amended the city charter to overturn the ordinance and to require that any additional antidiscrimination housing ordinance be approved by referendum. But most other ordinances “regulating the real property market” were not subject to those threshold requirements. Id., at 390. The plaintiff, a black woman in Akron, Ohio, alleged that her real estate agent could not show her certain residences because the owners had specified they would not sell to black persons.

     Central to the Court’s reasoning in Hunter was that the charter amendment was enacted in circumstances where widespread racial discrimination in the sale and rental of housing led to segregated housing, forcing many to live in “ ‘unhealthful, unsafe, unsanitary and overcrowded conditions.’ ” Id., at 391. The Court stated: “It is against this background that the referendum required by [the charter amendment] must be assessed.” Ibid. Akron attempted to characterize the charter amendment “simply as a public decision to move slowly in the delicate area of race relations” and as a means “to allow the people of Akron to participate” in the decision. Id., at 392. The Court rejected Akron’s flawed “justifications for its discrimination,”justifications that by their own terms had the effect of acknowledging the targeted nature of the charter amendment. Ibid. The Court noted, furthermore, that the charter amendment was unnecessary as a general means of public control over the city council; for the people of Akron already were empowered to overturn ordinances by referendum. Id., at 390, n. 6. The Court found that the city charter amendment, by singling out antidiscrimination ordinances, “places special burden on racial minorities within the governmental process,” thus becoming as impermissible as any other government action taken with the invidious intent to injure a racial minority. Id., at 391. Justice Harlan filed a concurrence. He argued the city charter amendment “has the clear purpose of making it more difficult for certain racial and religious minorities to achieve legislation that is in their interest.” Id., at 395. But without regard to the sentence just quoted, Hunter rests on the unremarkable principle that the State may not alter the procedures of government to target racial minorities. The facts in Hunter established that invidious discrimination would be the necessary result of the procedural restructuring. Thus, in Mulkey and Hunter, there was a demonstrated injury on the basis of race that, by reasons of state encouragement or participation, became more aggravated.

     Seattle is the third case of principal relevance here. There, the school board adopted a mandatory busing program to alleviate racial isolation of minority students in local schools. Voters who opposed the school board’s busing plan passed a state initiative that barred busing to desegregate. The Court first determined that, although “white as well as Negro children benefit from” diversity, the school board’s plan “inures primarily to the benefit of the minority.” 458 U. S., at 472. The Court next found that “the practical effect” of the state initiative was to “remov[e] the authority to address a racial problem—and only a racial problem—from the existing decisionmaking body, in such a way as to burden minority interests” because advocates of busing “now must seek relief from the state legislature, or from the statewide electorate.” Id., at 474. The Court therefore found that the initiative had “explicitly us[ed] the racial nature of a decision to determine the decisionmaking process.” Id., at 470 (emphasis deleted).

     Seattle is best understood as a case in which the state action in question (the bar on busing enacted by the State’s voters) had the serious risk, if not purpose, of causing specific injuries on account of race, just as had been the case in Mulkey and Hunter. Although there had been no judicial finding of de jure segregation with respect to Seattle’s school district, it appears as though school segregation in the district in the 1940’s and 1950’s may have been the partial result of school board policies that “permitted white students to transfer out of black schools while restricting the transfer of black students into white schools.” Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701 –808 (2007) (Breyer, J., dissenting). In 1977, the National Association for the Advancement of Colored People (NAACP) filed a complaint with the Office for Civil Rights, a federal agency. The NAACP alleged that the school board had maintained a system of de jure segregation. Specifically, the complaint alleged “that the Seattle School Board had created or perpetuated unlawful racial segregation through, e.g., certain school-transfer criteria, a construction program that needlessly built new schools in white areas, district line-drawing criteria, the maintenance of inferior facilities at black schools, the use of explicit racial criteria in the assignment of teachers and other staff, and a general pattern of delay in respect to the implementation of promised desegregation efforts.” Id., at 810. As part of a settlement with the Office for Civil Rights, the school board implemented the “Seattle Plan,” which used busing and mandatory reassignments between elementary schools to reduce racial imbalance and which was the subject of the state initiative at issue in Seattle. See 551 U. S., at 807–812.

     As this Court held in Parents Involved, the school board’s purported remedial action would not be permissible today absent a showing of de jure segregation. Id., at 720–721. That holding prompted Justice Breyer to observe in dissent, as noted above, that one permissible reading of the record was that the school board had maintained policies to perpetuate racial segregation in the schools. In all events we must understand Seattle as Seattle understood itself, as a case in which neither the State nor the United States “challenge[d] the propriety of race-conscious student assignments for the purpose of achieving integration, even absent a finding of prior de jure segregation.” 458 U. S. at 472, n. 15. In other words the legitimacy and constitutionality of the remedy in question (busing for desegregation) was assumed, and Seattle must be understood on that basis. Ibid. Seattle involved a state initiative that “was carefully tailored to interfere only with desegregative busing.” Id., at 471. The Seattle Court, accepting the validity of the school board’s busing remedy as a predicate to its analysis of the constitutional question, found that the State’s disapproval of the school board’s busing remedy was an aggravation of the very racial injury in which the State itself was complicit.

     The broad language used in Seattle, however, went well beyond the analysis needed to resolve the case. The Court there seized upon the statement in Justice Harlan’s concurrence in Hunter that the procedural change in that case had “the clear purpose of making it more difficult for certain racial and religious minorities to achieve legislation that is in their interest.” 385 U. S., at 395. That language, taken in the context of the facts in Hunter, is best read simply to describe the necessity for finding an equal protection violation where specific injuries from hostile discrimination were at issue. The Seattle Court, however, used the language from the Hunter concurrence to establish a new and far-reaching rationale. Seattle stated that where a government policy “inures primarily to the benefit of the minority” and “minorities . . . con-sider” the policy to be “ ‘in their interest,’ ” then any state action that “place[s] effective decisionmaking authority over” that policy “at a different level of government” must be reviewed under strict scrutiny. 458 U. S., at 472, 474. In essence, according to the broad reading of Seattle, any state action with a “racial focus” that makes it “more difficult for certain racial minorities than for other groups” to “achieve legislation that is in their interest” is subject to strict scrutiny. It is this reading of Seattle that the Court of Appeals found to be controlling here. And that reading must be rejected.

     The broad rationale that the Court of Appeals adopted goes beyond the necessary holding and the meaning of the precedents said to support it; and in the instant case neither the formulation of the general rule just set forth nor the precedents cited to authenticate it suffice to invalidate Proposal 2. The expansive reading of Seattle has no principled limitation and raises serious questions of compatibility with the Court’s settled equal protection jurisprudence. To the extent Seattle is read to require the Court to determine and declare which political policies serve the “interest” of a group defined in racial terms, that rationale was unnecessary to the decision in Seattle; it has no support in precedent; and it raises serious constitu-tional concerns. That expansive language does not provide a proper guide for decisions and should not be deemed authoritative or controlling. The rule that the Court of Appeals elaborated and respondents seek to establish here would contradict central equal protection principles.

     In cautioning against “impermissible racial stereotypes,” this Court has rejected the assumption that “members of the same racial group—regardless of their age, education, economic status, or the community in which they live—think alike, share the same political interests, and will prefer the same candidates at the polls.” Shaw v. Reno, 509 U. S. 630, 647 (1993) ; see also Metro Broadcasting, Inc. v. FCC, 497 U. S. 547, 636 (1990) (Kennedy, J., dissenting) (rejecting the “demeaning notion that members of . . . defined racial groups ascribe to certain ‘minority views’ that must be different from those of other citizens”). It cannot be entertained as a serious proposition that all individuals of the same race think alike. Yet that proposition would be a necessary beginning point were the Seattle formulation to control, as the Court of Appeals held it did in this case. And if it were deemed necessary to probe how some races define their own interest in political matters, still another beginning point would be to define individuals according to race. But in a society in which those lines are becoming more blurred, the attempt to define race-based categories also raises serious questions of its own. Government action that classifies individuals on the basis of race is inherently suspect and carries the danger of perpetuating the very racial divisions the polity seeks to transcend. Cf. Ho v. San Francisco Unified School Dist., 147 F. 3d 854, 858 (CA9 1998) (school district delineating 13 racial categories for purposes of racial balancing). Were courts to embark upon this venture not only would it be undertaken with no clear legal standards or accepted sources to guide judicial decision but also it would result in, or at least impose a high risk of, inquiries and categories dependent upon demeaning stereotypes, classifications of questionable constitutionality on their own terms.

     Even assuming these initial steps could be taken in a manner consistent with a sound analytic and judicial framework, the court would next be required to determine the policy realms in which certain groups—groups defined by race—have a political interest. That undertaking, again without guidance from any accepted legal standards, would risk, in turn, the creation of incentives for those who support or oppose certain policies to cast the debate in terms of racial advantage or disadvantage. Thus could racial antagonisms and conflict tend to arise in the context of judicial decisions as courts undertook to announce what particular issues of public policy should be classified as advantageous to some group defined by race. This risk is inherent in adopting the Seattle formulation.

     There would be no apparent limiting standards defining what public policies should be included in what Seattle called policies that “inur[e] primarily to the benefit of the minority” and that “minorities . . . consider” to be “ ‘in their interest.’ ” 458 U. S., at 472, 474. Those who seek to represent the interests of particular racial groups could attempt to advance those aims by demanding an equal protection ruling that any number of matters be foreclosed from voter review or participation. In a nation in which governmental policies are wide ranging, those who seek to limit voter participation might be tempted, were this Court to adopt the Seattle formulation, to urge that a group they choose to define by race or racial stereotypes are advantaged or disadvantaged by any number of laws or decisions. Tax policy, housing subsidies, wage regulations, and even the naming of public schools, highways, and monuments are just a few examples of what could become a list of subjects that some organizations could insist should be beyond the power of voters to decide, or beyond the power of a legislature to decide when enacting limits on the power of local authorities or other governmental entities to address certain subjects. Racial division would be validated, not discouraged, were the Seattle formulation, and the reasoning of the Court of Appeals in this case, to remain in force.

     Perhaps, when enacting policies as an exercise of democratic self-government, voters will determine that race-based preferences should be adopted. The constitutional validity of some of those choices regarding racial preferences is not at issue here. The holding in the instant case is simply that the courts may not disempower the voters from choosing which path to follow. In the realm of policy discussions the regular give-and-take of debate ought to be a context in which rancor or discord based on race are avoided, not invited. And if these factors are to be interjected, surely it ought not to be at the invitation or insistence of the courts.

     One response to these concerns may be that objections to the larger consequences of the Seattle formulation need not be confronted in this case, for here race was an undoubted subject of the ballot issue. But a number of problems raised by Seattle, such as racial definitions, still apply. And this principal flaw in the ruling of the Court of Appeals does remain: Here there was no infliction of a specific injury of the kind at issue in Mulkey and Hunter and in the history of the Seattle schools. Here there is no precedent for extending these cases to restrict the right of Michigan voters to determine that race-based preferences granted by Michigan governmental entities should be ended.

     It should also be noted that the judgment of the Court of Appeals in this case of necessity calls into question other long-settled rulings on similar state policies. The California Supreme Court has held that a California constitutional amendment prohibiting racial preferences in public contracting does not violate the rule set down by Seattle. Coral Constr., Inc. v. City and County of San Francisco, 50 Cal. 4th 315, 235 P. 3d 947 (2010). The Court of Appeals for the Ninth Circuit has held that the same amendment, which also barred racial preferences in public education, does not violate the Equal Protection Clause. Wilson, 122 F. 3d 692 (1997). If the Court were to affirm the essential rationale of the Court of Appeals in the instant case, those holdings would be invalidated, or at least would be put in serious question. The Court, by affirming the judgment now before it, in essence would announce a finding that the past 15 years of state public debate on this issue have been improper. And were the argument made that Coral might still stand because it involved racial preferences in public contracting while this case concerns racial preferences in university admissions, the implication would be that the constitutionality of laws forbidding racial preferences depends on the policy interest at stake, the concern that, as already explained, the voters deem it wise to avoid because of its divisive potential. The instant case presents the question involved in Coral and Wilson but not involved in Mulkey, Hunter, and Seattle. That question is not how to address or prevent injury caused on account of race but whether voters may determine whether a policy of race-based preferences should be continued.

     By approving Proposal 2 and thereby adding §26 to their State Constitution, the Michigan voters exercised their privilege to enact laws as a basic exercise of their democratic power. In the federal system States “respond, through the enactment of positive law, to the initiative of those who seek a voice in shaping the destiny of their own times.” Bond, 564 U. S., at ––– (slip op., at 9). Michigan voters used the initiative system to bypass public officials who were deemed not responsive to the concerns of a majority of the voters with respect to a policy of granting race-based preferences that raises difficult and delicate issues.

     The freedom secured by the Constitution consists, in one of its essential dimensions, of the right of the individual not to be injured by the unlawful exercise of governmental power. The mandate for segregated schools, Brown v. Board of Education, 347 U. S. 483 (1954) ; a wrongful invasion of the home, Silverman v. United States, 365 U. S. 505 (1961) ; or punishing a protester whose views offend others, Texas v. Johnson, 491 U. S. 397 (1989) ; and scores of other examples teach that individual liberty has constitutional protection, and that liberty’s full extent and meaning may remain yet to be discovered and affirmed. Yet freedom does not stop with individual rights. Our constitutional system embraces, too, the right of citizens to debate so they can learn and decide and then, through the political process, act in concert to try to shape the course of their own times and the course of a nation that must strive always to make freedom ever greater and more secure. Here Michigan voters acted in concert and statewide to seek consensus and adopt a policy on a difficult subject against a historical background of race in America that has been a source of tragedy and persisting injustice. That history demands that we continue to learn, to listen, and to remain open to new approaches if we are to aspire always to a constitutional order in which all persons are treated with fairness and equal dignity. Were the Court to rule that the question addressed by Michigan voters is too sensitive or complex to be within the grasp of the electorate; or that the policies at issue remain too delicate to be resolved save by university officials or faculties, acting at some remove from immediate public scru-tiny and control; or that these matters are so arcane that the electorate’s power must be limited because the people cannot prudently exercise that power even after a full debate, that holding would be an unprecedented restriction on the exercise of a fundamental right held not just by one person but by all in common. It is the right to speak and debate and learn and then, as a matter of political will, to act through a lawful electoral process.

     The respondents in this case insist that a difficult question of public policy must be taken from the reach of the voters, and thus removed from the realm of public discussion, dialogue, and debate in an election campaign. Quite in addition to the serious First Amendment implications of that position with respect to any particular election, it is inconsistent with the underlying premises of a responsible, functioning democracy. One of those premises is that a democracy has the capacity—and the duty—to learn from its past mistakes; to discover and confront persisting biases; and by respectful, rationale deliberation to rise above those flaws and injustices. That process is impeded, not advanced, by court decrees based on the proposition that the public cannot have the requisite repose to discuss certain issues. It is demeaning to the democratic process to presume that the voters are not capable of deciding an issue of this sensitivity on decent and rational grounds. The process of public discourse and political debate should not be foreclosed even if there is a risk that during a public campaign there will be those, on both sides, who seek to use racial division and discord to their own political advantage. An informed public can, and must, rise above this. The idea of democracy is that it can, and must, mature. Freedom embraces the right, indeed the duty, to engage in a rational, civic discourse in order to determine how best to form a consensus to shape the destiny of the Nation and its people. These First Amendment dynamics would be disserved if this Court were to say that the question here at issue is beyond the capacity of the voters to debate and then to determine.

     These precepts are not inconsistent with the well-established principle that when hurt or injury is inflicted on racial minorities by the encouragement or command of laws or other state action, the Constitution requires redress by the courts. Cf. Johnson v. California, 543 U. S. 499 –512 (2005) (“[S]earching judicial review . . . is necessary to guard against invidious discrimination”); Edmonson v. Leesville Concrete Co., 500 U. S. 614, 619 (1991) (“Racial discrimination” is “invidious in all contexts”). As already noted, those were the circumstances that the Court found present in Mulkey, Hunter, and Seattle. But those circumstances are not present here.

     For reasons already discussed, Mulkey, Hunter, and Seattle are not precedents that stand for the conclusion that Michigan’s voters must be disempowered from acting. Those cases were ones in which the political restriction in question was designed to be used, or was likely to be used, to encourage infliction of injury by reason of race. What is at stake here is not whether injury will be inflicted but whether government can be instructed not to follow a course that entails, first, the definition of racial categories and, second, the grant of favored status to persons in some racial categories and not others. The electorate’s instruction to governmental entities not to embark upon the course of race-defined and race-based preferences was adopted, we must assume, because the voters deemed a preference system to be unwise, on account of what voters may deem its latent potential to become itself a source of the very resentments and hostilities based on race that this Nation seeks to put behind it. Whether those adverse results would follow is, and should be, the subject of debate. Voters might likewise consider, after debate and reflection, that programs designed to increase diversity—consistent with the Constitution—are a necessary part of progress to transcend the stigma of past racism.

     This case is not about how the debate about racial preferences should be resolved. It is about who may resolve it. There is no authority in the Constitution of the United States or in this Court’s precedents for the Judiciary to set aside Michigan laws that commit this policy determination to the voters. See Sailors v. Board of Ed. of County of Kent, 387 U. S. 105, 109 (1967) (“Save and unless the state, county, or municipal government runs afoul of a federally protected right, it has vast leeway in the management of its internal affairs”). Deliberative debate on sensitive issues such as racial preferences all too often may shade into rancor. But that does not justify removing certain court-determined issues from the voters’ reach. Democracy does not presume that some subjects are either too divisive or too profound for public debate.

     The judgment of the Court of Appeals for the Sixth Circuit is reversed.

It is so ordered.

     Justice Kagan took no part in the consideration or decision of this case.

SUPREME COURT OF THE UNITED STATES

_________________

No. 12–682

_________________

BILL SCHUETTE, ATTORNEY GENERAL OF MICHI-GAN, PETITIONER v. COALITION TO DEFEND AF-FIRMATIVE ACTION, INTEGRATION AND IMMI-GRANT RIGHTS AND FIGHT FOR EQUALITYBY ANY MEANS NECESSARY (BAMN), et al.

on writ of certiorari to the united states court of appeals for the sixth circuit

[April 22, 2014]

     Justice Breyer, concurring in the judgment.

     Michigan has amended its Constitution to forbid state universities and colleges to “discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.” Mich. Const., Art. I, §26. We here focus on the prohibition of “grant[ing] . . . preferential treatment . . . on the basis of race . . . in . . . public education.” I agree with the plurality that the amendment is consistent with the Federal Equal Protection Clause. U. S. Const., Amdt. 14. But I believe this for different reasons.

     First, we do not address the amendment insofar as it forbids the use of race-conscious admissions programs designed to remedy past exclusionary racial discrimination or the direct effects of that discrimination. Application of the amendment in that context would present different questions which may demand different answers. Rather, we here address the amendment only as it applies to, and forbids, programs that, as in Grutter v. Bollinger, 539 U. S. 306 (2003) , rest upon “one justification”: using “race in the admissions process” solely in order to “obtai[n] the educational benefits that flow from a diverse student body,” id., at 328 (internal quotation marks omitted).

     Second, dissenting in Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701 (2007) , I explained why I believe race-conscious programs of this kind are constitutional, whether implemented by law schools, universities, high schools, or elementary schools. I concluded that the Constitution does not “authorize judges” either to forbid or to require the adoption of diversity-seeking race-conscious “solutions” (of the kind at issue here) to such serious problems as “how best to administer America’s schools” to help “create a society that includes all Americans.” Id., at 862.

     I continue to believe that the Constitution permits, though it does not require, the use of the kind of race-conscious programs that are now barred by the Michigan Constitution. The serious educational problems that faced Americans at the time this Court decided Grutter endure. See, e.g., I. Mullis, M. Martin, P. Foy, & K. Drucker,Progress in International Reading Literacy Study, 2011 International Results in Reading 38, Exh. 1.1 (2012)(elementary-school students in numerous other countries outperform their counterparts in the United States in reading); I. Mullis, M. Martin, P. Foy, & A. Arora, Trends in International Mathematics and Science Study (TIMSS), 2011 International Results in Mathematics 40, Exh. 1.1 (2012) (same in mathematics); M. Martin, I. Mullis, P. Foy, & G. Stanco, TIMSS, 2011 International Results in Science, 38, Exh. 1.1 (2012) (same in science); Organisation of Economic Co-operation Development (OECD), Education at a Glance 2013: OECD Indicators 50 (Table A2.1a) (secondary-school graduation rate lower in the United States than in numerous other countries); McKinsey & Co., The Economic Impact of the Achievement Gap in America’s Schools 8 (Apr. 2009) (same; United States ranks 18th of 24 industrialized nations). And low educational achievement continues to be correlated with income and race. See, e.g., National Center for Education Statistics, Digest of Education Statistics, Advance Release of Selected 2013 Digest Tables (Table 104.20) (White Americans more likely to have completed high school thanAfrican-Americans or Hispanic-Americans), online at http://nces.ed.gov/programs/digest (as visited Apr. 15, 2014, and available in Clerk of Court’s case file); id., Table 219.75 (Americans in bottom quartile of income most likely to drop out of high school); id., Table 302.60 (White Americans more likely to enroll in college than African-Americans or Hispanic-Americans); id., Table 302.30 (middle- and high-income Americans more likely to enroll in college than low-income Americans).

     The Constitution allows local, state, and national communities to adopt narrowly tailored race-conscious programs designed to bring about greater inclusion and diversity. But the Constitution foresees the ballot box, not the courts, as the normal instrument for resolving differences and debates about the merits of these programs. Compare Parents Involved, 551 U. S., at 839 (Breyer, J., dissenting) (identifying studies showing the benefits of racially integrated education), with id., at 761–763 (Thomas, J., concurring) (identifying studies suggesting racially integrated schools may not confer educational benefits). In short, the “Constitution creates a democratic political system through which the people themselves must together find answers” to disagreements of this kind. Id., at 862 (Breyer, J., dissenting).

     Third, cases such as Hunter v. Erickson, 393 U. S. 385 (1969) , and Washington v. Seattle School Dist. No. 1, 458 U. S. 457 (1982) , reflect an important principle, namely, that an individual’s ability to participate meaningfully in the political process should be independent of his race. Although racial minorities, like other political minorities, will not always succeed at the polls, they must have the same opportunity as others to secure through the ballot box policies that reflect their preferences. In my view, however, neither Hunter nor Seattle applies here. And the parties do not here suggest that the amendment violates the Equal Protection Clause if not under the Hunter-Seattle doctrine.

     Hunter and Seattle involved efforts to manipulate the political process in a way not here at issue. Both cases involved a restructuring of the political process that changed the political level at which policies were enacted. In Hunter, decisionmaking was moved from the elected city council to the local electorate at large. 393 U. S., at 389–390. And in Seattle, decisionmaking by an elected school board was replaced with decisionmaking by the state legislature and electorate at large. 458 U. S., at 466.

     This case, in contrast, does not involve a reordering of the political process; it does not in fact involve the movement of decisionmaking from one political level to another. Rather, here, Michigan law delegated broad policymaking authority to elected university boards, see Mich. Const., Art. VIII, §5, but those boards delegated admissions-related decisionmaking authority to unelected university faculty members and administrators, see, e.g., Bylaws of Univ. of Mich. Bd. of Regents §8.01; Mich. State Univ. Bylaws of Bd. of Trustees, Preamble; Mich. State Univ. Bylaws for Academic Governance §4.4.3; Wayne State Univ. Stat. §§2–34–09, 2–34–12. Although the boards unquestionably retained the power to set policy regarding race-conscious admissions, see post, at 25–29 (Sotomayor, J., dissenting), in fact faculty members and administrators set the race-conscious admissions policies in question. (It is often true that elected bodies—including, for example, school boards, city councils, and state legislatures—have the power to enact policies, but in fact delegate that power to administrators.) Although at limited times the university boards were advised of the content of their race-conscious admissions policies, see 701 F. 3d 466, 481–482 (CA6 2012), to my knowledge no board voted to accept or reject any of those policies. Thus, un-elected faculty members and administrators, not voters or their elected representatives, adopted the race-conscious admissions programs affected by Michigan’s constitutional amendment. The amendment took decisionmaking authority away from these unelected actors and placed it in the hands of the voters.

     Why does this matter? For one thing, considered conceptually, the doctrine set forth in Hunter and Seattle does not easily fit this case. In those cases minorities had participated in the political process and they had won. The majority’s subsequent reordering of the political process repealed the minority’s successes and made it more difficult for the minority to succeed in the future. The majority thereby diminished the minority’s ability to participate meaningfully in the electoral process. But one cannot as easily characterize the movement of the decisionmaking mechanism at issue here—from an administrative process to an electoral process—as diminishing the minority’s ability to participate meaningfully in the political process. There is no prior electoral process in which the minority participated.

     For another thing, to extend the holding of Hunter and Seattle to reach situations in which decisionmaking authority is moved from an administrative body to a political one would pose significant difficulties. The administrative process encompasses vast numbers of decisionmakers answering numerous policy questions in hosts of different fields. See Free Enterprise Fund v. Public Company Accounting Oversight Bd., 561 U. S. 477 , ___ (2010) (Breyer, J., dissenting). Administrative bodies modify programs in detail, and decisionmaking authority within the administrative process frequently moves around—due to amendments to statutes, new administrative rules, and evolving agency practice. It is thus particularly difficult in this context for judges to determine when a change in the locus of decisionmaking authority places a comparative structural burden on a racial minority. And to apply Hunter and Seattle to the administrative process would, by tending to hinder change, risk discouraging experimentation, interfering with efforts to see when and how race-conscious policies work.

     Finally, the principle that underlies Hunter and Seattle runs up against a competing principle, discussed above. This competing principle favors decisionmaking though the democratic process. Just as this principle strongly supports the right of the people, or their elected representatives, to adopt race-conscious policies for reasons of inclusion, so must it give them the right to vote not to do so.

     As I have said, my discussion here is limited to circumstances in which decisionmaking is moved from an un-elected administrative body to a politically responsive one, and in which the targeted race-conscious admissions programs consider race solely in order to obtain the educational benefits of a diverse student body. We need now decide no more than whether the Federal Constitution permits Michigan to apply its constitutional amendment in those circumstances. I would hold that it does. Therefore, I concur in the judgment of the Court.

SUPREME COURT OF THE UNITED STATES

_________________

No. 12–682

_________________

BILL SCHUETTE, ATTORNEY GENERAL OF MICHI-GAN, PETITIONER v. COALITION TO DEFEND AF-FIRMATIVE ACTION, INTEGRATION AND IMMI-GRANT RIGHTS AND FIGHT FOR EQUALITYBY ANY MEANS NECESSARY (BAMN), et al.

on writ of certiorari to the united states court of appeals for the sixth circuit

[April 22, 2014]

     Justice Scalia, with whom Justice Thomas joins, concurring in the judgment.

     It has come to this. Called upon to explore the jurisprudential twilight zone between two errant lines of precedent, we confront a frighteningly bizarre question: Does the Equal Protection Clause of the Fourteenth Amendment forbid what its text plainly requires? Needless to say (except that this case obliges us to say it), the question answers itself. “The Constitution proscribes government discrimination on the basis of race, and state-provided education is no exception.” Grutter v. Bollinger, 539 U. S. 306, 349 (2003) (Scalia, J., concurring in part and dissenting in part). It is precisely this understanding—the correct understanding—of the federal Equal Protection Clause that the people of the State of Michigan have adopted for their own fundamental law. By adopting it, they did not simultaneously offend it.

     Even taking this Court’s sorry line of race-based-admissions cases as a given, I find the question presented only slightly less strange: Does the Equal Protection Clause forbid a State from banning a practice that the Clause barely—and only provisionally—permits? Reacting to those race-based-admissions decisions, some States—whether deterred by the prospect of costly litigation; aware that Grutter’s bell may soon toll, see 539 U. S., at 343; or simply opposed in principle to the notion of “benign” racial discrimination—have gotten out of the racial-preferences business altogether. And with our express encouragement: “Universities in California, Flor-ida, and Washington State, where racial preferences in admissions are prohibited by state law, are currently engaging in experimenting with a wide variety of alternative approaches. Universities in other States can and should draw on the most promising aspects of these race-neutral alternatives as they develop.” Id., at 342 (emphasis added). Respondents seem to think this admonition was merely in jest.[1] The experiment, they maintain, is not only over; it never rightly began. Neither the people of the States nor their legislatures ever had the option of directing subordinate public-university officials to cease considering the race of applicants, since that would deny members of those minority groups the option of enacting a policy designed to further their interest, thus denying them the equal protection of the laws. Never mind that it is hotly disputed whether the practice of race-based admissions is ever in a racial minority’s interest. Cf. id., at 371–373 (Thomas, J., concurring in part and dissenting in part). And never mind that, were a public university to stake its defense of a race-based-admissions policy on the ground that it was designed to benefit primarily minorities (as opposed to all students, regardless of color, by enhancing diversity), we would hold the policy unconstitutional. See id., at 322–325.

     But the battleground for this case is not the constitutionality of race-based admissions—at least, not quite. Rather, it is the so-called political-process doctrine, derived from this Court’s opinions in Washington v. Seattle School Dist. No. 1, 458 U. S. 457 (1982) , and Hunter v. Erickson, 393 U. S. 385 (1969) . I agree with those parts of the plurality opinion that repudiate this doctrine. But I do not agree with its reinterpretation of Seattle and Hunter, which makes them stand in part for the cloudy and doctrinally anomalous proposition that whenever state action poses “the serious risk . . . of causing specific injuries on account of race,” it denies equal protection. Ante, at 9. I would instead reaffirm that the “ordinary principles of our law [and] of our democratic heritage” require “plaintiffs alleging equal protection violations” stemming from fa-cially neutral acts to “prove intent and causation and not merely the existence of racial disparity.” Freeman v. Pitts, 503 U. S. 467, 506 (1992) (Scalia, J., concurring) (citing Washington v. Davis, 426 U. S. 229 (1976) ). I would further hold that a law directing state actors to provide equal protection is (to say the least) facially neutral, and cannot violate the Constitution. Section 26 of the Michigan Constitution (formerly Proposal 2) rightly stands.

I

A

     The political-process doctrine has its roots in two of our cases. The first is Hunter. In 1964, the Akron City Council passed a fair-housing ordinance “ ‘assur[ing] equal opportunity to all persons to live in decent housing facilities regardless of race, color, religion, ancestry or national origin.’ ” 393 U. S., at 386. Soon after, the city’s voters passed an amendment to the Akron City Charter stating that any ordinance enacted by the council that “ ‘regulates’ ” commercial transactions in real property “ ‘on the basis of race, color, religion, national origin or ancestry’ ”—including the already enacted 1964 ordinance—“must first be approved by a majority of the electors voting on the question” at a later referendum. Id., at 387. The question was whether the charter amendment denied equal protection. Answering yes, the Court explained that “although the law on its face treats Negro and white, Jew and gentile in an identical manner, the reality is that the law’s impact falls on the minority. The majority needs no protection against discrimination.” Id., at 391. By placing a “special burden on racial minorities within the governmental processes,” the amendment “disadvantage[d]” a racial minority “by making it more difficult to enact legislation in its behalf.” Id., at 391, 393.

     The reasoning in Seattle is of a piece. Resolving to “eliminate all [racial] imbalance from the Seattle public schools,” the city school board passed a mandatory busing and pupil-reassignment plan of the sort typically imposed on districts guilty of de jure segregation. 458 U. S., at 460–461. A year later, the citizens of the State of Washington passed Initiative 350, which directed (with exceptions) that “ ‘no school . . . shall directly or indirectly require any student to attend a school other than the school which is geographically nearest or next nearest the student’s place of residence . . . and which offers the course of study pursued by such student,’ ” permitting only court-ordered race-based busing. Id., at 462. The lower courts held Initiative 350 unconstitutional, and we affirmed, announcing in the prelude of our analysis—as though it were beyond debate—that the Equal Protection Clause forbade laws that “subtly distor[t] governmental processes in such a way as to place special burdens on the ability of minority groups to achieve beneficial legislation.” Id., at 467.

     The first question in Seattle was whether the subject matter of Initiative 350 was a “ ‘racial’ issue,” triggering Hunter and its process doctrine. 458 U. S., at 471–472. It was “undoubtedly. . . true” that whites and blacks were “counted among both the supporters and the opponents of Initiative 350.” Id., at 472. It was “equally clear” that both white and black children benefitted from desegre-gated schools. Ibid. Nonetheless, we concluded that desegre-gation “inures primarily to the benefit of the minority, and is designed for that purpose.” Ibid. (emphasis added). In any event, it was “enough that minorities may consider busing for integration to be ‘legislation that is in their interest.’ ” Id., at 474 (quoting Hunter, supra, at 395 (Harlan, J., concurring)).

     So we proceeded to the heart of the political-process analysis. We held Initiative 350 unconstitutional, since it removed “the authority to address a racial problem—and only a racial problem—from the existing decisionmaking body, in such a way as to burden minority interests.” Seattle, 458 U. S., at 474. Although school boards in Washington retained authority over other student-assignment issues and over most matters of educational policy generally, under Initiative 350, minorities favoring race-based busing would have to “surmount a considerably higher hurdle” than the mere petitioning of a local assembly: They “now must seek relief from the state legislature, or from the statewide electorate,” a “different level of government.” Ibid.

     The relentless logic of Hunter and Seattle would point to a similar conclusion in this case. In those cases, one level of government exercised borrowed authority over an apparently “racial issue,” until a higher level of government called the loan. So too here. In those cases, we deemed the revocation an equal-protection violation regardless of whether it facially classified according to race or reflected an invidious purpose to discriminate. Here, the Court of Appeals did the same.

     The plurality sees it differently. Though it, too, dis-avows the political-process-doctrine basis on which Hunter and Seattle were decided, ante, at 10–14, it does not take the next step of overruling those cases. Rather, it reinterprets them beyond recognition. Hunter, the plurality suggests, was a case in which the challenged act had “target[ed] racial minorities.” Ante, at 8. Maybe, but the Hunter Court neither found that to be so nor considered it relevant, bypassing the question of intent entirely, satisfied that its newly minted political-process theory sufficed to invalidate the charter amendment.

     As for Seattle, what was really going on, according to the plurality, was that Initiative 350 had the consequence (if not the purpose) of preserving the harms effected by prior de jure segregation. Thus, “the political restriction in question was designed to be used, or was likely to be used, to encourage infliction of injury by reason of race.” Ante, at 17. That conclusion is derived not from the opinion but from recently discovered evidence that the city of Seattle had been a cause of its schools’ racial imbalance all along: “Although there had been no judicial finding of de jure segregation with respect to Seattle’s school district, it appears as though school segregation in the district in the 1940’s and 1950’s may have been the partial result of school board policies.” Ante, at 9.[2] That the district’s effort to end racial imbalance had been stymied by Initiative 350 meant that the people, by passing it, somehow had become complicit in Seattle’s equal-protection-denying status quo, whether they knew it or not. Hence, therewas in Seattle a government-furthered “infliction of a specific”—and, presumably, constitutional—“injury.” Ante, at 14.

     Once again this describes what our opinion in Seattle might have been, but assuredly not what it was. The opinion assumes throughout that Seattle’s schools suffered at most from de facto segregation, see, e.g., 458 U. S., at 474, 475—that is, segregation not the “product . . . of state action but of private choices,” having no “constitutional implications,” Freeman, 503 U. S., at 495–496. Nor did it anywhere state that the current racial imbalance was the (judicially remediable) effect of prior de jure segregation. Absence of de jure segregation or the effects of de jure segregation was a necessary premise of the Seattle opinion. That is what made the issue of busing and pupil reassignment a matter of political choice rather than judicial mandate.[3] And precisely because it was a question for the political branches to decide, the manner—which is to say, the process—of its resolution implicated the Court’s new process theory. The opinion itself says this: “[I]n the absence of a constitutional violation, the desirability and efficacy of school desegregation are matters to be resolved though the political process. For present purposes, it is enough [to hold reallocation of that political decision to a higher level unconstitutional] that minorities may consider busing for integration to be legislation that is in their interest.” 458 U. S., at 474 (internal quotation marks omitted).

B

     Patently atextual, unadministrable, and contrary to our traditional equal-protection jurisprudence, Hunter and Seattle should be overruled.

     The problems with the political-process doctrine begin with its triggering prong, which assigns to a court the task of determining whether a law that reallocates policy-making authority concerns a “racial issue.” Seattle, 458 U. S., at 473. Seattle takes a couple of dissatisfying cracks at defining this crucial term. It suggests that an issue is racial if adopting one position on the question would “at bottom inur[e] primarily to the benefit of the minority, and is designed for that purpose.” Id., at 472. It is irrelevant that, as in Hunter and Seattle, 458 U. S., at 472, both the racial minority and the racial majority benefit from the policy in question, and members of both groups favor it. Judges should instead focus their guesswork on their own juridical sense of what is primarily for the benefit of minorities. Cf. ibid. (regarding as dispositive what “our cases” suggest is beneficial to minorities). On second thought, maybe judges need only ask this question: Is it possible “that minorities may consider” the policy in question to be “in their interest”? Id., at 474. If so, you can be sure that you are dealing with a “racial issue.”[4]

     No good can come of such random judicial musing. The plurality gives two convincing reasons why. For one thing, it involves judges in the dirty business of dividing the Nation “into racial blocs,” Metro Broadcasting, Inc. v. FCC, 497 U. S. 547, 603, 610 (1990) (O’Connor, J., dissenting); ante, at 11–13. That task is as difficult as it is unappealing. (Does a half-Latino, half–American Indian have Latino interests, American-Indian interests, both, half of both?[5]) What is worse, the exercise promotes the noxious fiction that, knowing only a person’s color or ethnicity, we can be sure that he has a predetermined set of policy “interests,” thus “reinforc[ing] the perception that members of the same racial group—regardless of their age, education, economic status, or the community in which they live—think alike, [and] share the same political interests.”[6] Shaw v. Reno, 509 U. S. 630, 647 (1993) . Whether done by a judge or a school board, such “racial stereotyping [is] at odds with equal protection mandates.” Miller v. Johnson, 515 U. S. 900, 920 (1995) .

     But that is not the “racial issue” prong’s only defect. More fundamentally, it misreads the Equal Protection Clause to protect “particular group[s],” a construction that we have tirelessly repudiated in a “long line of casesunderstanding equal protection as a personal right.” Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 224, 230 (1995) . It is a “basic principle that the Fifth and Fourteenth Amendments to the Constitution protect persons, not groups.” Id., at 227; Metro Broadcasting, supra, at 636 (Kennedy, J., dissenting).[7] Yet Seattle insists that only those political-process alterations that burden racial minorities deny equal protection. “The majority,” after all, “needs no protection against discrimination.” 458 U. S., at 468 (quoting Hunter, 393 U. S., at 391). In the years since Seattle, we have repeatedly rejected “a reading of the guarantee of equal protection under which the level of scrutiny varies according to the ability of different groups to defend their interests in the representative process.” Richmond v. J. A. Croson Co., 488 U. S. 469, 495 (1989) . Meant to obliterate rather than endorse the practice of racial classifications, the Fourteenth Amendment’s guarantees “obtai[n] with equal force regardless of ‘the race of those burdened or benefitted.’ ” Miller, supra, at 904 (quoting Croson, supra, at 494 (plurality opinion)); Adarand, supra, at 223, 227. The Equal Protection Clause “cannot mean one thing when applied to one individual and something else when applied to a person of another color. If both are not accorded the same protection it is not equal.” Regents of Univ. of Cal. v. Bakke, 438 U. S. 265 –290 (1978) (opinion of Powell, J.).

     The dissent trots out the old saw, derived from dictum in a footnote, that legislation motivated by “ ‘prejudice against discrete and insular minorities’ ” merits “ ‘more exacting judicial scrutiny.’ ” Post, at 31 (quoting United States v. Carolene Products, 304 U. S. 144 –153, n. 4). I say derived from that dictum (expressed by the four-Justice majority of a seven-Justice Court) because the dictum itself merely said “[n]or need we enquire . . . whether prejudice against discrete and insular minorities may be a special condition,” id., at 153, n. 4 (emphasis added). The dissent does not argue, of course, that such “prejudice” produced §26. Nor does it explain why certain racial minorities in Michigan qualify as “ ‘insular,’ ” meaning that “other groups will not form coalitions with them—and, critically, not because of lack of common interests but because of ‘prejudice.’ ” Strauss, Is Carolene Products Obsolete? 2010 U. Ill. L. Rev. 1251, 1257. Nor does it even make the case that a group’s “discreteness” and “insu-larity” are political liabilities rather than political strengths[8]—a serious question that alone demonstrates the prudence of the Carolene Products dictumizers in leaving the “enquir[y]” for another day. As for the question whether “legislation which restricts those political processes which can ordinarily be expected to bring about repeal of undesirable legislation . . . is to be subjected to more exacting judicial scrutiny,” the Carolene Products Court found it “unnecessary to consider [that] now.” 304 U. S., at 152, n. 4. If the dissent thinks that worth considering today, it should explain why the election of a university’s governing board is a “political process which can ordinarily be expected to bring about repeal of undesirable legislation,” but Michigan voters’ ability to amend their Constitution is not. It seems to me quite the opposite. Amending the Constitution requires the approval of only “a majority of the electors voting on the question.” Mich. Const., Art. XII, §2. By contrast, voting in a favorable board (each of which has eight members) at the three major public universities requires electing by majority vote at least 15 different candidates, several of whom would be running during different election cycles. See BAMN v. Regents of Univ. of Mich., 701 F. 3d 466, 508 (CA6 2012) (Sutton, J., dissenting). So if Michigan voters, instead of amending their Constitution, had pursued the dissent’s preferred path of electing board members promising to “abolish race-sensitive admissions policies,” post, at 3, it would have been harder, not easier, for racial minorities favoring affirmative action to overturn that decision. But the more important point is that we should not design our jurisprudence to conform to dictum in a footnote in a four-Justice opinion.

C

     Moving from the appalling to the absurd, I turn now to the second part of the Hunter-Seattle analysis—which is apparently no more administrable than the first, compare post, at 4–6 (Breyer, J., concurring in judgment) (“This case . . . does not involve a reordering of the political process”), with post, at 25–29 (Sotomayor, J., dissenting) (yes, it does). This part of the inquiry directs a court to determine whether the challenged act “place[s] effective decisionmaking authority over [the] racial issue at a different level of government.” Seattle, 458 U. S., at 474. The laws in both Hunter and Seattle were thought to fail this test. In both cases, “the effect of the challengedaction was to redraw decisionmaking authority over racial matters—and only over racial matters—in such a way as to place comparative burdens on minorities.” 458 U. S., at 475, n. 17. This, we said, a State may not do.

     By contrast, in another line of cases, we have emphasized the near-limitless sovereignty of each State to design its governing structure as it sees fit. Generally, “a State is afforded wide leeway when experimenting with the appropriate allocation of state legislative power” and may create “political subdivisions such as cities and counties . . . ‘as convenient agencies for exercising such of the governmental powers of the state as may be entrusted to them.’ ” Holt Civic Club v. Tuscaloosa, 439 U. S. 60, 71 (1978) (quoting Hunter v. Pittsburgh, 207 U. S. 161, 178 (1907) ). Accordingly, States have “absolute discretion” to determine the “number, nature and duration of the powers conferred upon [municipal] corporations and the territory over which they shall be exercised.” Holt Civic Club, supra, at 71. So it would seem to go without saying that a State may give certain powers to cities, later assign the same powers to counties, and even reclaim them for itself.

     Taken to the limits of its logic, Hunter-Seattle is the gaping exception that nearly swallows the rule of structural state sovereignty. If indeed the Fourteenth Amendment forbids States to “place effective decisionmaking authority over” racial issues at “different level[s] of government,” then it must be true that the Amendment’s ratification in 1868 worked a partial ossification of each State’s governing structure, rendering basically irrevocable the power of any subordinate state official who, the day before the Fourteenth Amendment’s passage, happened to enjoy legislatively conferred authority over a “racial issue.” Under the Fourteenth Amendment, that subordinate entity (suppose it is a city council) could itself take action on the issue, action either favorable or unfavorable to minorities. It could even reverse itself later. What it could not do, however, is redelegate its power to an even lower level of state government (such as a city-council committee) without forfeiting it, since the necessary effect of wresting it back would be to put an additional obstacle in the path of minorities. Likewise, no entityor official higher up the state chain (e.g., a county board) could exercise authority over the issue. Nor, even, could the state legislature, or the people by constitutional amendment, revoke the legislative conferral of power to the subordinate, whether the city council, its subcommittee, or the county board. Seattle’s logic would createaffirmative-action safe havens wherever subordinate offi-cials in public universities (1) traditionally have enjoyed “effective decisionmaking authority” over admissions policy but (2) have not yet used that authority to prohibit race-conscious admissions decisions. The mere existence of a subordinate’s discretion over the matter would work a kind of reverse pre-emption. It is “a strange notion—alien to our system—that local governmental bodies can forever pre-empt the ability of a State—the sovereign power—to address a matter of compelling concern to the State.” 458 U. S., at 495 (Powell, J., dissenting). But that is precisely what the political-process doctrine contemplates.

     Perhaps the spirit of Seattle is especially disquieted by enactments of constitutional amendments. That appears to be the dissent’s position. The problem with §26, it suggests, is that amending Michigan’s Constitution is simply not a part of that State’s “existing” political process. E.g., post, at 4, 41. What a peculiar notion: that a revision of a State’s fundamental law, made in precisely the manner that law prescribes, by the very people who are the source of that law’s authority, is not part of the “political process” which, but for those people and that law, would not exist. This will surely come as news to the people of Michigan, who, since 1914, have amended their Constitution 20 times. Brief for Gary Segura et al. as Amici Curiae 12. Even so, the dissent concludes that the amendment attacked here worked an illicit “chang[ing] [of ] the basic rules of the political process in that State” in “the middle of the game.” Post, at 2, 4. Why, one might ask, is not the amendment provision of the Michigan Constitution one (perhaps the most basic one) of the rules of the State’s political process? And why does democratic invocation of that provision not qualify as working through the “existing political process,” post, at 41?[9]

II

     I part ways with Hunter, Seattle, and (I think) the plurality for an additional reason: Each endorses a version of the proposition that a facially neutral law may deny equal protection solely because it has a disparate racial impact. Few equal-protection theories have been so squarely and soundly rejected. “An unwavering line of cases from this Court holds that a violation of the Equal Protection Clause requires state action motivated by discriminatory intent,” Hernandez v. New York, 500 U. S. 352 –373 (1991) (O’Connor, J., concurring in judgment), and that “official action will not be held unconstitutional solely because it results in a racially disproportionate impact,” Arlington Heights v. Metropolitan Housing Development Corp., 429 U. S. 252 –265 (1977). Indeed, we affirmed this principle the same day we decided Seattle: “[E]ven when a neutral law has a disproportionately adverse effect on a racial minority, the Fourteenth Amendment is violated only if a discriminatory purpose can be shown.” Crawford v. Board of Ed. of Los Angeles, 458 U. S. 527 –538 (1982).

     Notwithstanding our dozens of cases confirming the exception-less nature of the Washington v. Davis rule, the plurality opinion leaves ajar an effects-test escape hatch modeled after Hunter and Seattle, suggesting that state action denies equal protection when it “ha[s] the serious risk, if not purpose, of causing specific injuries on account of race,” or is either “designed to be used, or . . . likely to be used, to encourage infliction of injury by reason of race.” Ante, at 9, 17 (emphasis added). Since these formulations enable a determination of an equal-protection violation where there is no discriminatory intent, they are inconsistent with the long Washington v. Davis line of cases.[10]

     Respondents argue that we need not bother with the discriminatory-purpose test, since §26 may be struck more straightforwardly as a racial “classification.” Admitting (as they must) that §26 does not on its face “distribut[e] burdens or benefits on the basis of individual racial classifications,” Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701, 720 (2007) , respondents rely on Seattle’s statement that “when the political process or the decisionmaking mechanism used to address racially conscious legislation—and only such legislation—is singled out for peculiar and disadvantageous treatment,” then that “singling out” is a racial classification. 458 U. S., at 485, 486, n. 30. But this is just the political-process theory bedecked in different doctrinal dress. A law that “neither says nor implies that persons are to be treated differently on account of their race” is not a racial classification. Crawford, supra, at 537. That is particularly true of statutes mandating equal treatment. “[A] law that prohibits the State from classifying individuals by race . . . a fortiori does not classify individuals by race.” Coalition for Economic Equity v. Wilson, 122 F. 3d 692, 702 (CA9 1997) (O’Scannlain, J.).

     Thus, the question in this case, as in every case in which neutral state action is said to deny equal protection on account of race, is whether the action reflects a racially discriminatory purpose. Seattle stresses that “singling out the political processes affecting racial issues for uniquely disadvantageous treatment inevitably raises dangers of impermissible motivation.” 458 U. S., at 486, n. 30. True enough, but that motivation must be proved. And respondents do not have a prayer of proving it here. The District Court noted that, under “conventional equal protection” doctrine, the suit was “doom[ed].” 539 F. Supp. 2d 924, 951 (ED Mich. 2008). Though the Court of Appeals did not opine on this question, I would not leave it for them on remand. In my view, any law expressly requiring state actors to afford all persons equal protection of the laws (such as Initiative 350 in Seattle, though not the charter amendment in Hunter) does not—cannot—deny “to any person . . . equal protection of the laws,” U. S. Const., Amdt. 14, §1, regardless of whatever evidence of seemingly foul purposes plaintiffs may cook up in the trial court.

*  *  *

     As Justice Harlan observed over a century ago, “[o]ur Constitution is color-blind, and neither knows nor tolerates classes among citizens.” Plessy v. Ferguson, 163 U. S. 537, 559 (1896) (dissenting opinion). The people of Michigan wish the same for their governing charter. It would be shameful for us to stand in their way.[11]

SUPREME COURT OF THE UNITED STATES

_________________

No. 12–682

_________________

BILL SCHUETTE, ATTORNEY GENERAL OF MICHI-GAN, PETITIONER v. COALITION TO DEFEND AF-FIRMATIVE ACTION, INTEGRATION AND IMMI-GRANT RIGHTS AND FIGHT FOR EQUALITYBY ANY MEANS NECESSARY (BAMN), et al.

on writ of certiorari to the united states court of appeals for the sixth circuit

[April 22, 2014]

     Chief Justice Roberts, concurring.

     The dissent devotes 11 pages to expounding its own policy preferences in favor of taking race into account in college admissions, while nonetheless concluding that it “do[es] not mean to suggest that the virtues of adopting race-sensitive admissions policies should inform the legal question before the Court.” Post, at 57 (opinion of Sotomayor, J.). The dissent concedes that the governing boards of the State’s various universities could have implemented a policy making it illegal to “discriminate against, or grant preferential treatment to,” any individ-ual on the basis of race. See post, at 3, 34–35. On the dissent’s view, if the governing boards conclude that drawing racial distinctions in university admissions is undesirable or counterproductive, they are permissibly exercising their policymaking authority. But others who might reach the same conclusion are failing to take race seriously.

     The dissent states that “[t]he way to stop discrimination on the basis of race is to speak openly and candidly on the subject of race.” Post, at 46. And it urges that “[r]ace matters because of the slights, the snickers, the silent judgments that reinforce that most crippling of thoughts: ‘I do not belong here.’ ” Ibid. But it is not “out of touch with reality” to conclude that racial preferences may themselves have the debilitating effect of reinforcing precisely that doubt, and—if so—that the preferences do more harm than good. Post, at 45. To disagree with the dissent’s views on the costs and benefits of racial preferences is not to “wish away, rather than confront” racial inequality. Post, at 46. People can disagree in good faith on this issue, but it similarly does more harm than good to question the openness and candor of those on either side of the debate.[1]*

SUPREME COURT OF THE UNITED STATES

_________________

No. 12–682

_________________

BILL SCHUETTE, ATTORNEY GENERAL OF MICHI-GAN, PETITIONER v. COALITION TO DEFEND AF-FIRMATIVE ACTION, INTEGRATION AND IMMI-GRANT RIGHTS AND FIGHT FOR EQUALITYBY ANY MEANS NECESSARY (BAMN), et al.

on writ of certiorari to the united states court of appeals for the sixth circuit

[April 22, 2014]

     Justice Sotomayor, with whom Justice Ginsburg joins, dissenting.

     We are fortunate to live in a democratic society. But without checks, democratically approved legislation can oppress minority groups. For that reason, our Constitution places limits on what a majority of the people may do. This case implicates one such limit: the guarantee of equal protection of the laws. Although that guarantee is traditionally understood to prohibit intentional discrimination under existing laws, equal protection does not end there. Another fundamental strand of our equal protection jurisprudence focuses on process, securing to all citizens the right to participate meaningfully and equally in self-government. That right is the bedrock of our democracy, for it preserves all other rights.

     Yet to know the history of our Nation is to understand its long and lamentable record of stymieing the right of racial minorities to participate in the political process. At first, the majority acted with an open, invidious purpose. Notwithstanding the command of the Fifteenth Amendment, certain States shut racial minorities out of the political process altogether by withholding the right to vote. This Court intervened to preserve that right. The majority tried again, replacing outright bans on voting with literacy tests, good character requirements, poll taxes, and gerrymandering. The Court was not fooled; it invalidated those measures, too. The majority persisted. This time, although it allowed the minority access to the political process, the majority changed the ground rules of the process so as to make it more difficult for the minority, and the minority alone, to obtain policies designed to foster racial integration. Although these political restructurings may not have been discriminatory in purpose, the Court reaffirmed the right of minority members of our society to participate meaningfully and equally in the political process.

     This case involves this last chapter of discrimination: A majority of the Michigan electorate changed the basic rules of the political process in that State in a manner that uniquely disadvantaged racial minorities.[1] Prior to the enactment of the constitutional initiative at issue here,all of the admissions policies of Michigan’s public colleges and universities—including race-sensitive admissions poli-cies[2]—were in the hands of each institution’s governing board. The members of those boards are nominated by political parties and elected by the citizenry in statewide elections. After over a century of being shut out of Michigan’s institutions of higher education, racial minorities in Michigan had succeeded in persuading the elected board representatives to adopt admissions policies that took into account the benefits of racial diversity. And this Court twice blessed such efforts—first in Regents of Univ. of Cal. v. Bakke, 438 U. S. 265 (1978) , and again in Grutter v. Bollinger, 539 U. S. 306 (2003) , a case that itself concerned a Michigan admissions policy.

     In the wake of Grutter, some voters in Michigan set out to eliminate the use of race-sensitive admissions policies. Those voters were of course free to pursue this end in any number of ways. For example, they could have persuaded existing board members to change their minds through individual or grassroots lobbying efforts, or through general public awareness campaigns. Or they could have mobilized efforts to vote uncooperative board members out of office, replacing them with members who would share their desire to abolish race-sensitive admissions policies. When this Court holds that the Constitution permits a particular policy, nothing prevents a majority of a State’s voters from choosing not to adopt that policy. Our system of government encourages—and indeed, depends on—that type of democratic action.

     But instead, the majority of Michigan voters changed the rules in the middle of the game, reconfiguring the existing political process in Michigan in a manner that burdened racial minorities. They did so in the 2006 election by amending the Michigan Constitution to enact Art. I, §26, which provides in relevant part that Michigan’s public universities “shall not discriminate against, or grant preferential treatment to, any individual or group on the basis of race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting.”

     As a result of §26, there are now two very different processes through which a Michigan citizen is permitted to influence the admissions policies of the State’s universities: one for persons interested in race-sensitive admissions policies and one for everyone else. A citizen who is a University of Michigan alumnus, for instance, can advocate for an admissions policy that considers an applicant’s legacy status by meeting individually with members of the Board of Regents to convince them of her views, by joining with other legacy parents to lobby the Board, or by voting for and supporting Board candidates who share her position. The same options are available to a citizen who wants the Board to adopt admissions policies that consider athleticism, geography, area of study, and so on. The one and only policy a Michigan citizen may not seek through this long-established process is a race-sensitive admissions policy that considers race in an individualized manner when it is clear that race-neutral alternatives are not adequate to achieve diversity. For that policy alone, the citizens of Michigan must undertake the daunting task of amending the State Constitution.

     Our precedents do not permit political restructurings that create one process for racial minorities and a separate, less burdensome process for everyone else. This Court has held that the Fourteenth Amendment does not tolerate “a political structure that treats all individuals as equals, yet more subtly distorts governmental processes in such a way as to place special burdens on the ability of minority groups to achieve beneficial legislation.” Washington v. Seattle School Dist. No. 1, 458 U. S. 457, 467 (1982) (internal quotation marks omitted). Such restructuring, the Court explained, “is no more permissible than denying [the minority] the [right to] vote, on an equal basis with others.” Hunter v. Erickson, 393 U. S. 385, 391 (1969) . In those cases—Hunter and Seattle—the Court recognized what is now known as the “political-process doctrine”: When the majority reconfigures the political process in a manner that burdens only a racial minority, that alteration triggers strict judicial scrutiny.

     Today, disregarding stare decisis, a majority of the Court effectively discards those precedents. The plurality does so, it tells us, because the freedom actually secured by the Constitution is the freedom of self-government—because the majority of Michigan citizens “exercised their privilege to enact laws as a basic exercise of their democratic power.” Ante, at 15. It would be “demeaning to the democratic process,” the plurality concludes, to disturb that decision in any way. Ante, at 17. This logic embraces majority rule without an important constitutional limit.

     The plurality’s decision fundamentally misunderstands the nature of the injustice worked by §26. This case is not, as the plurality imagines, about “who may resolve” the debate over the use of race in higher education admissions. Ante, at 18. I agree wholeheartedly that nothing vests the resolution of that debate exclusively in the courts or requires that we remove it from the reach of the electorate. Rather, this case is about how the debate over the use of race-sensitive admissions policies may be resolved, contra, ibid.—that is, it must be resolved in constitution-ally permissible ways. While our Constitution does not guarantee minority groups victory in the political process, it does guarantee them meaningful and equal access to that process. It guarantees that the majority may not win by stacking the political process against minority groups permanently, forcing the minority alone to surmount unique obstacles in pursuit of its goals—here, educational diversity that cannot reasonably be accomplished through race-neutral measures. Today, by permitting a majority of the voters in Michigan to do what our Constitution forbids, the Court ends the debate over race-sensitive admissions policies in Michigan in a manner that contravenes constitutional protections long recognized in our precedents.

     Like the plurality, I have faith that our citizenry will continue to learn from this Nation’s regrettable history; that it will strive to move beyond those injustices towards a future of equality. And I, too, believe in the importance of public discourse on matters of public policy. But I part ways with the plurality when it suggests that judicial intervention in this case “impede[s]” rather than “advance[s]” the democratic process and the ultimate hope of equality. Ante, at 16. I firmly believe that our role as judges includes policing the process of self-government and stepping in when necessary to secure the constitutional guarantee of equal protection. Because I would do so here, I respectfully dissent.

I

     For much of its history, our Nation has denied to many of its citizens the right to participate meaningfully and equally in its politics. This is a history we strive to put behind us. But it is a history that still informs the society we live in, and so it is one we must address with candor. Because the political-process doctrine is best understood against the backdrop of this history, I will briefly trace its course.

     The Fifteenth Amendment, ratified after the Civil War, promised to racial minorities the right to vote. But many States ignored this promise. In addition to outright tactics of fraud, intimidation, and violence, there are countless examples of States categorically denying to racial minorities access to the political process. Consider Texas; there, a 1923 statute prevented racial minorities from participating in primary elections. After this Court declared that statute unconstitutional, Nixon v. Herndon, 273 U. S. 536 –541 (1927), Texas responded by changing the rules. It enacted a new statute that gave political parties themselves the right to determine who could participate in their primaries. Predictably, the Democratic Party specified that only white Democrats could participate in its primaries. Nixon v. Condon, 286 U. S. 73 –82 (1932). The Court invalidated that scheme, too. Id., at 89; see also Smith v. Allwright, 321 U. S. 649 (1944) ; Terry v. Adams, 345 U. S. 461 (1953) .

     Some States were less direct. Oklahoma was one of many that required all voters to pass a literacy test. But the test did not apply equally to all voters. Under a “grandfather clause,” voters were exempt if their grand-fathers had been voters or had served as soldiers before 1866. This meant, of course, that black voters had to pass the test, but many white voters did not. The Court held the scheme unconstitutional. Guinn v. United States, 238 U. S. 347 (1915) . In response, Oklahoma changed the rules. It enacted a new statute under which all voters who were qualified to vote in 1914 (under the unconstitutional grandfather clause) remained qualified, and the remaining voters had to apply for registration within a 12-day period. Lane v. Wilson, 307 U. S. 268 –271 (1939). The Court struck down that statute as well. Id., at 275.

     Racial minorities were occasionally able to surmount the hurdles to their political participation. Indeed, in some States, minority citizens were even able to win elective office. But just as many States responded to the Fifteenth Amendment by subverting minorities’ access to the polls, many States responded to the prospect of elected minority officials by undermining the ability of minorities to win and hold elective office. Some States blatantly removed black officials from local offices. See, e.g., H. Rabinowitz, Race Relations in the Urban South, 1865–1890, pp. 267, 269–270 (1978) (describing events in Tennessee and Virginia). Others changed the processes by which local officials were elected. See, e.g., Extension of the Voting Rights Act, Hearings before the Subcommittee on Civil and Constitutional Rights of the House Committee on the Judiciary, 97th Cong., 1st Sess., pt. 1, pp. 2016–2017 (1981) (hereinafter 1981 Hearings) (statement of Professor J. Morgan Kousser) (after a black judge refused to resign in Alabama, the legislature abolished the court on which he served and replaced it with one whose judges were appointed by the Governor); Rabinowitz, supra, at 269–270 (the North Carolina Legislature divested voters ofthe right to elect justices of the peace and county commissioners, then arrogated to itself the authority to select justices of the peace and gave them the power to select commissioners).

     This Court did not stand idly by. In Alabama, for example, the legislature responded to increased black voter registration in the city of Tuskegee by amending the State Constitution to authorize legislative abolition of the county in which Tuskegee was located, Ala. Const. Amdt. 132 (1957), repealed by Ala. Const. Amdt. 406 (1982), and by redrawing the city’s boundaries to remove all the black voters “while not removing a single white voter,” Gomillion v. Lightfoot, 364 U. S. 339, 341 (1960) . The Court intervened, finding it “inconceivable that guaranties embedded in the Constitution” could be “manipulated out of existence” by being “cloaked in the garb of [political] realignment.” Id., at 345 (internal quotation marks omitted).

     This Court’s landmark ruling in Brown v. Board of Education, 347 U. S. 483 (1954) , triggered a new era of political restructuring, this time in the context of education. In Virginia, the General Assembly transferred control of student assignment from local school districts to a State Pupil Placement Board. See B. Muse, Virginia’s Massive Resistance 34, 74 (1961). And when the legislature learned that the Arlington County school board had prepared a desegregation plan, the General Assembly “swiftly retaliated” by stripping the county of its right to elect its school board by popular vote and instead making the board an appointed body. Id., at 24; see also B. Smith, They Closed Their Schools 142–143 (1965).

     Other States similarly disregarded this Court’s mandate by changing their political process. See, e.g., Bush v. Orleans Parish School Bd., 187 F. Supp. 42, 44–45 (ED La. 1960) (the Louisiana Legislature gave the Governor the authority to supersede any school board’s decision to integrate); Extension of the Voting Rights Act, Hearings on H. R. 4249 et al. before Subcommittee No. 5 of the House Committee on the Judiciary, 91st Cong., 1st Sess., 146–149 (1969) (statement of Thomas E. Harris, Assoc. Gen. Counsel, American Federation of Labor and Congress of Industrial Organizations) (the Mississippi Legislature removed from the people the right to elect superintendents of education in 11 counties and instead made those positions appointive).

     The Court remained true to its command in Brown. In Arkansas, for example, it enforced a desegregation order against the Little Rock school board. Cooper v. Aaron, 358 U. S. 1, 5 (1958) . On the very day the Court announced that ruling, the Arkansas Legislature responded by changing the rules. It enacted a law permitting the Governor to close any public school in the State, and stripping local school districts of their decisionmaking authority so long as the Governor determined that local officials could not maintain “ ‘a general, suitable, and efficient educational system.’ ” Aaron v. Cooper, 261 F. 2d 97, 99 (CA8 1958) (per curiam) (quoting Arkansas statute). The then-Governor immediately closed all of Little Rock’s high schools. Id., at 99–100; see also S. Breyer, Making Our Democracy Work 49–67 (2010) (discussing the events in Little Rock).

     The States’ political restructuring efforts in the 1960’s and 1970’s went beyond the context of education. Many States tried to suppress the political voice of racial minorities more generally by reconfiguring the manner in which they filled vacancies in local offices, often transferring authority from the electorate (where minority citizens had a voice at the local level) to the States’ executive branch (where minorities wielded little if any influence). See, e.g., 1981 Hearings, pt. 1, at 815 (report of J. Cox & A. Turner) (the Alabama Legislature changed all municipal judgeships from elective to appointive offices); id., at 1955 (report of R. Hudlin & K. Brimah, Voter Educ. Project, Inc.) (the Georgia Legislature eliminated some elective offices and made others appointive when it appeared that a minority candidate would be victorious); id., at 501 (statement of Frank R. Parker, Director, Lawyers’ Comm. for Civil Rights Under Law) (the Mississippi Legislature changed the manner of filling vacancies for various public offices from election to appointment).

II

     It was in this historical context that the Court intervened in Hunter v. Erickson, 393 U. S. 385 (1969) , and Washington v. Seattle School Dist. No. 1, 458 U. S. 457 (1982) . Together, Hunter and Seattle recognized a fundamental strand of this Court’s equal protection jurisprudence: the political-process doctrine. To understand that doctrine fully, it is necessary to set forth in detail precisely what the Court had before it, and precisely what it said. For to understand Hunter and Seattle is to understand why those cases straightforwardly resolve this one.

A

     In Hunter, the City Council of Akron, Ohio, enacted a fair housing ordinance to “assure equal opportunity to all persons to live in decent housing facilities regardless of race, color, religion, ancestry, or national origin.” 393 U. S., at 386 (internal quotation marks omitted). A majority of the citizens of Akron disagreed with the ordinance and overturned it. But the majority did not stop there; it also amended the city charter to prevent the City Council from implementing any future ordinance dealing with racial, religious, or ancestral discrimination in housing without the approval of the majority of the Akron electorate. Ibid. That amendment changed the rules of the political process in Akron. The Court described the result of the change as follows:

“[T]o enact an ordinance barring housing discrimination on the basis of race or religion, proponents had to obtain the approval of the City Council and of a majority of the voters citywide. To enact an ordinance preventing housing discrimination on other grounds, or to enact any other type of housing ordinance, proponents needed the support of only the City Council.” Seattle, 458 U. S., at 468 (describing Hunter; emphasis deleted).

     The Court invalidated the Akron charter amendment under the Equal Protection Clause. It concluded that the amendment unjustifiably “place[d] special burdens on racial minorities within the governmental process,” thus effecting “a real, substantial, and invidious denial of the equal protection of the laws.” Hunter, 393 U. S., at 391, 393. The Court characterized the amendment as “no more permissible” than denying racial minorities the right to vote on an equal basis with the majority. Id., at 391. For a “State may no more disadvantage any particular group by making it more difficult to enact legislation in its behalf than it may dilute any person’s vote or give any group a smaller representation than another of comparable size.” Id., at 392–393. The vehicle for the change—a popular referendum—did not move the Court: “The sovereignty of the people,” it explained, “is itself subject to . . . constitutional limitations.” Id., at 392.

     Justice Harlan, joined by Justice Stewart, wrote in his concurrence that although a State can normally allocate political power according to any general principle, it bears a “far heavier burden of justification” when it reallocates political power based on race, because the selective reallocation necessarily makes it far more difficult for racial minorities to “achieve legislation that is in their interest.” Id., at 395 (internal quotation marks omitted).

     In Seattle, a case that mirrors the one before us, the Court applied Hunter to invalidate a statute, enacted by a majority of Washington State’s citizens, that prohibited racially integrative busing in the wake of Brown. As early as 1963, Seattle’s School District No. 1 began taking steps to cure the de facto racial segregation in its schools. 458 U. S., at 460–461. Among other measures, it enacted a desegregation plan that made extensive use of busing and mandatory assignments. Id., at 461. The district was under no obligation to adopt the plan; Brown charged school boards with a duty to integrate schools that were segregated because of de jure racial discrimination, but there had been no finding that the de facto segregation in Seattle’s schools was the product of de jure discrimination. 458 U. S., at 472, n. 15. Several residents who opposed the desegregation efforts formed a committee and sued to enjoin implementation of the plan. Id., at 461. When these efforts failed, the committee sought to change the rules of the political process. It drafted a statewide initiative “designed to terminate the use of mandatory busing for purposes of racial integration.” Id., at 462. A major-ity of the State’s citizens approved the initiative. Id., at 463–464.

     The Court invalidated the initiative under the Equal Protection Clause. It began by observing that equal protection of the laws “guarantees racial minorities the right to full participation in the political life of the community.” Id., at 467. “It is beyond dispute,” the Court explained, “that given racial or ethnic groups may not be denied the franchise, or precluded from entering into the political process in a reliable and meaningful manner.” Ibid. But the Equal Protection Clause reaches further, the Court stated, reaffirming the principle espoused in Hunter—that while “laws structuring political institutions or allocating political power according to neutral principles” do not violate the Constitution, “a different analysis is required when the State allocates governmental power nonneutrally, by explicitly using the racial nature of a decision to determine the decisionmaking process.” 458 U. S., at 470. That kind of state action, it observed, “places special burdens on racial minorities within the governmental process,” by making it “more difficult for certain racial and religious minorities” than for other members of the community “to achieve legislation . . . in their interest.” Ibid.

     Rejecting the argument that the initiative had no racial focus, the Court found that the desegregation of public schools, like the Akron housing ordinance, “inure[d] primarily to the benefit of the minority, and [was] designed for that purpose.” Id., at 472. Because minorities had good reason to “consider busing for integration to be ‘legislation that is in their interest,’ ” the Court concluded that the “racial focus of [the initiative] . . . suffice[d] to trigger application of the Hunter doctrine.” Id., at 474 (quoting Hunter, 393 U. S., at 395) (Harlan, J. concurring)).

     The Court next concluded that “the practical effect of [the initiative was] to work a reallocation of power of the kind condemned in Hunter.” Seattle, 458 U. S., at 474. It explained: “Those favoring the elimination of de facto school segregation now must seek relief from the state legislature, or from the statewide electorate. Yet authority over all other student assignment decisions, as well as over most other areas of educational policy, remains vested in the local school board.” Ibid. Thus, the initiative required those in favor of racial integration in public schools to “surmount a considerably higher hurdle than persons seeking comparable legislative action” in different contexts. Ibid.

     The Court reaffirmed that the “ ‘simple repeal or modification of desegregation or antidiscrimination laws, without more, never has been viewed as embodying a presump-tively invalid racial classification.’ ” Id., at 483 (quoting Crawford v. Board of Ed. of Los Angeles, 458 U. S. 527, 539 (1982) ). But because the initiative burdened future attempts to integrate by lodging the decisionmaking authority at a “new and remote level of government,” it was more than a “mere repeal”; it was an unconstitutionally discriminatory change to the political process.[3] Seattle, 458 U. S., at 483–484.

B

     Hunter and Seattle vindicated a principle that is as elementary to our equal protection jurisprudence as it is essential: The majority may not suppress the minority’s right to participate on equal terms in the political process. Under this doctrine, governmental action deprives minor-ity groups of equal protection when it (1) has a racial focus, targeting a policy or program that “inures primarily to the benefit of the minority,” Seattle, 458 U. S., at 472; and (2) alters the political process in a manner that uniquely burdens racial minorities’ ability to achieve their goals through that process. A faithful application of the doctrine resoundingly resolves this case in respondents’ favor.

1

     Section 26 has a “racial focus.” Seattle, 458 U. S., at 474. That is clear from its text, which prohibits Michigan’s public colleges and universities from “grant[ing] preferential treatment to any individual or group on the basis of race.” Mich. Const., Art. I, §26. Like desegregation of public schools, race-sensitive admissions policies “inur[e] primarily to the benefit of the minority,” 458 U. S., at 472, as they are designed to increase minorities’ access to institutions of higher education.[4]

     Petitioner argues that race-sensitive admissions policies cannot “inur[e] primarily to the benefit of the minority,” ibid., as the Court has upheld such policies only insofar as they further “the educational benefits that flow from a diverse student body,” Grutter, 539 U. S., at 343. But there is no conflict between this Court’s pronouncement in Grutter and the common-sense reality that race-sensitive admissions policies benefit minorities. Rather, race-sensitive admissions policies further a compelling state interest in achieving a diverse student body precisely because they increase minority enrollment, which necessarily benefits minority groups. In other words, constitutionally permissible race-sensitive admissions policies can both serve the compelling interest of obtaining the educational benefits that flow from a diverse student body, and inure to the benefit of racial minorities. There is nothing mutually exclusive about the two. Cf. Seattle, 458 U. S., at 472 (concluding that the desegregation plan had a racial focus even though “white as well as Negro children bene-fit from exposure to ‘ethnic and racial diversity in the classroom’ ”).

     It is worth emphasizing, moreover, that §26 is relevant only to admissions policies that have survived strict scrutiny under Grutter; other policies, under this Court’s rulings, would be forbidden with or without §26. A Grutter-compliant admissions policy must use race flexibly, not maintain a quota; must be limited in time; and must be employed only after “serious, good faith consideration of workable race-neutral alternatives,” 539 U. S., at 339. The policies banned by §26 meet all these requirements and thus already constitute the least restrictive ways to advance Michigan’s compelling interest in diversity in higher education.

2

     Section 26 restructures the political process in Michigan in a manner that places unique burdens on racial minorities. It establishes a distinct and more burdensome political process for the enactment of admissions plans that consider racial diversity.

     Long before the enactment of §26, the Michigan Constitution granted plenary authority over all matters relating to Michigan’s public universities, including admissions criteria, to each university’s eight-member governing board. See Mich. Const., Art. VIII, §5 (establishing the Board of Regents of the University of Michigan, the Board of Trustees of Michigan State University, and the Board of Governors of Wayne State University). The boards have the “power to enact ordinances, by-laws and regulations for the government of the university.” Mich. Comp. Laws Ann. §390.5 (West 2010); see also §390.3 (“The government of the university is vested in the board of regents”). They are “ ‘constitutional corporation[s] of independent authority, which, within the scope of [their] functions, [are] co-ordinate with and equal to . . . the legislature.’ ” Federated Publications, Inc. v. Board of Trustees of Mich. State Univ., 460 Mich. 75, 84, n. 8, 594 N. W. 2d 491, 496, n. 8 (1999).

     The boards are indisputably a part of the political process in Michigan. Each political party nominates two candidates for membership to each board, and board members are elected to 8-year terms in the general statewide election. See Mich. Comp. Laws Ann. §§168.282, 168.286 (West 2008); Mich. Const., Art. VIII, §5. Prior to §26, board candidates frequently included their views on race-sensitive admissions in their campaigns. For example, in 2005, one candidate pledged to “work to end so-called ‘Affirmative-Action,’ a racist, degrading system.” See League of Women Voters, 2005 General Election Voter Guide, online at http://www.lwvka.org/guide04/regents/html (all Internet materials as visited Apr. 18, 2014, and available in Clerk of Court’s case file); see also George, U-M Regents Race Tests Policy, Detroit Free Press, Oct. 26, 2000, p. 2B (noting that one candidate “opposes affirmative action admissions policies” because they “ ‘basically sa[y] minority students are not qualified’ ”).

     Before the enactment of §26, Michigan’s political structure permitted both supporters and opponents of race-sensitive admissions policies to vote for their candidates of choice and to lobby the elected and politically accountable boards. Section 26 reconfigured that structure. After §26, the boards retain plenary authority over all admissions criteria except for race-sensitive admissions policies.[5] To change admissions policies on this one issue, a Michigan citizen must instead amend the Michigan Constitution. That is no small task. To place a proposed constitutional amendment on the ballot requires either the support of two-thirds of both Houses of the Michigan Legislature or a vast number of signatures from Michigan voters—10 percent of the total number of votes cast in the preceding gubernatorial election. See Mich. Const., Art. XII, §§1, 2. Since more than 3.2 million votes were cast in the 2010 election for Governor, more than 320,000 signatures are currently needed to win a ballot spot. See Brief for Gary Segura et al. as Amici Curiae 9 (hereinafter Segura Brief). Moreover, “[t]o account for invalid and duplicative signatures, initiative sponsors ‘need to obtain substantially more than the actual required number of signatures, typically by a 25% to 50% margin.’ ” Id., at 10 (quoting Tolbert, Lowenstein, & Donovan, Election Law and Rules for Using Initiatives, in Citizens as Legislators: Direct Democracy in the United States 27, 37 (S. Bowler, T. Donovan, & C. Tolbert eds., 1998)).

     And the costs of qualifying an amendment are significant. For example, “[t]he vast majority of petition efforts . . . require initiative sponsors to hire paid petition circulators, at significant expense.” Segura Brief 10; see also T. Donovan, C. Mooney, & D. Smith, State and Local Politics: Institutions and Reform 96 (2012) (hereinafter Donovan) (“In many states, it is difficult to place a measure on the ballot unless professional petition firms are paid to collect some or all the signatures required for qualification”); Tolbert, supra, at 35 (“ ‘Qualifying an initiative for the statewide ballot is . . . no longer so much a measure of general citizen interest as it is a test of fundraising ability’ ”). In addition to the cost of collecting signatures, campaigning for a majority of votes is an expensive endeavor, and “organizations advocating on behalf of marginalized groups remain . . . outmoneyed by corporate, business, and professional organizations.” Strolovitch & Forrest, Social and Economic Justice Movements and Organizations, in The Oxford Handbook of American Political Parties and Interest Groups 468, 471 (L. Maisel & J. Berry eds., 2010). In 2008, for instance, over $800 million was spent nationally on state-level initiative and referendum campaigns, nearly $300 million more than was spent in the 2006 cycle. Donovan 98. “In several states, more money [is] spent on ballot initiative campaigns than for all other races for political office combined.” Ibid. Indeed, the amount spent on state-level initiative and referendum campaigns in 2008 eclipsed the $740.6 million spent by President Obama in his 2008 presidential campaign, Salant, Spending Doubled as Obama Led Billion-Dollar Campaign, Bloomberg News, Dec. 27, 2008, online at http://www.bloomberg.com/apps/news?pid=newsarchive&sid=anLDS9WWPQW8.

     Michigan’s Constitution has only rarely been amended through the initiative process. Between 1914 and 2000, voters have placed only 60 statewide initiatives on the Michigan ballot, of which only 20 have passed. See Segura Brief 12. Minority groups face an especially uphill battle. See Donovan 106 (“[O]n issues dealing with racial and ethnic matters, studies show that racial and ethnic minorities do end up more on the losing side of the popular vote”). In fact, “[i]t is difficult to find even a single statewide initiative in any State in which voters approved policies that explicitly favor racial or ethnic minority groups.”[6] Segura Brief 13.

     This is the onerous task that §26 forces a Michigan citizen to complete in order to change the admissions policies of Michigan’s public colleges and universities with respect to racial sensitivity. While substantially less grueling paths remain open to those advocating for any other admissions policies, a constitutional amendment is the only avenue by which race-sensitive admissions policies may be obtained. The effect of §26 is that a white graduate of a public Michigan university who wishes to pass his historical privilege on to his children may freely lobby the board of that university in favor of an expanded legacy admissions policy, whereas a black Michigander who was denied the opportunity to attend that very university cannot lobby the board in favor of a policy that might give his children a chance that he never had and that they might never have absent that policy.

     Such reordering of the political process contravenes Hunter and Seattle.[7] See Seattle, 458 U. S., at 467 (the Equal Protection Clause prohibits “ ‘a political structure that treats all individuals as equals,’ yet more subtly distorts governmental processes in such a way as to place special burdens on the ability of minority groups to achieve beneficial legislation” (citation omitted)). Where, as here, the majority alters the political process to the detriment of a racial minority, the governmental action is subject to strict scrutiny. See id., at 485, n. 28. Michigan does not assert that §26 satisfies a compelling state interest. That should settle the matter.

C

1

     The plurality sees it differently. Disregarding the language used in Hunter, the plurality asks us to contort that case into one that “rests on the unremarkable principle that the State may not alter the procedures of government to target racial minorities.” Ante, at 8. And the plurality recasts Seattle “as a case in which the state action in question . . . had the serious risk, if not purpose, of causing specific injuries on account of race.” Ante, at 8–9. According to the plurality, the Hunter and Seattle Courts were not concerned with efforts to reconfigure the political process to the detriment of racial minorities; rather, those cases invalidated governmental actions merely because they reflected an invidious purpose to discriminate. This is not a tenable reading of those cases.

     The plurality identifies “invidious discrimination” as the “necessary result” of the restructuring in Hunter. Ante, at 8. It is impossible to assess whether the housing amendment in Hunter was motivated by discriminatory purpose, for the opinion does not discuss the question of intent.[8] What is obvious, however, is that the possibility of invidious discrimination played no role in the Court’s reasoning. We ordinarily understand our precedents to mean what they actually say, not what we later think they could or should have said. The Hunter Court was clear about why it invalidated the Akron charter amendment: It was impermissible as a restructuring of the political process, not as an action motivated by discriminatory intent. See 393 U. S., at 391 (striking down the Akron charter amendment because it “places a special burden on racial minorities within the governmental process”).

     Similarly, the plurality disregards what Seattle actually says and instead opines that “the political restriction in question was designed to be used, or was likely to be used, to encourage infliction of injury by reason of race.” Ante, at 17. Here, the plurality derives its conclusion not from Seattle itself, but from evidence unearthed more than a quarter-century later in Parents Involved in Community Schools v. Seattle School Dist. No. 1, 551 U. S. 701 (2007) : “Although there had been no judicial finding of de jure segregation with respect to Seattle’s school district, it appears as though school desegregation in the district in the 1940’s and 1950’s may have been the partial result of school board policies that ‘permitted white students to transfer out of black schools while restricting the transfer of black students into white schools.’ ”[9] Ante, at 9 (quoting Parents Involved, 551 U. S., at 807–808 (Breyer, J., dissenting) (emphasis added). It follows, according to the plurality, that Seattle’s desegregation plan was constitutionally required, so that the initiative halting the plan was an instance of invidious discrimination aimed at inflicting a racial injury.

     Again, the plurality might prefer that the Seattle Court had said that, but it plainly did not. Not once did the Court suggest the presence of de jure segregation in Seattle. Quite the opposite: The opinion explicitly suggested the desegregation plan was adopted to remedy de facto rather than de jure segregation. See 458 U. S., at 472, n. 15 (referring to the “absen[ce]” of “a finding of prior de jure segregation”). The Court, moreover, assumed that no “constitutional violation” through de jure segregation had occurred. Id., at 474. And it unmistakably rested its decision on Hunter, holding Seattle’s initiative invalid because it “use[d] the racial nature of an issue to define the governmental decisionmaking structure, and thus impose[d] substantial and unique burdens on racial minorities.” 458 U. S., at 470.

     It is nothing short of baffling, then, for the plurality to insist—in the face of clear language in Hunter and Seattle saying otherwise—that those cases were about nothing more than the intentional and invidious infliction of a racial injury. Ante, at 8 (describing the injury in Hunter as “a demonstrated injury on the basis of race”); ante, at 8–9 (describing the injury in Seattle as an “injur[y] on account of race”). The plurality’s attempt to rewrite Hunter and Seattle so as to cast aside the political-process doctrine sub silentio is impermissible as a matter of stare decisis. Under the doctrine of stare decisis, we usually stand by our decisions, even if we disagree with them, because people rely on what we say, and they believe they can take us at our word.

     And what now of the political-process doctrine? After the plurality’s revision of Hunter and Seattle, it is unclear what is left. The plurality certainly does not tell us. On this point, and this point only, I agree with Justice Scalia that the plurality has rewritten those precedents beyond recognition. See ante, at 5–7 (opinion concurring in judgment).

2

     Justice Breyer concludes that Hunter and Seattle do not apply. Section 26, he reasons, did not move the relevant decisionmaking authority from one political level to another; rather, it removed that authority from “unelected actors and placed it in the hands of the voters.” Ante, at 5 (opinion concurring in judgment). He bases this conclusion on the premise that Michigan’s elected boards “delegated admissions-related decisionmaking authority to unelected university faculty members and administrators.” Ibid. But this premise is simply incorrect.

     For one thing, it is undeniable that prior to §26, board candidates often pledged to end or carry on the use of race-sensitive admissions policies at Michigan’s public universities. See supra, at 18. Surely those were not empty promises. Indeed, the issue of race-sensitive admissions policies often dominated board elections. See, e.g., George, Detroit Free Press, at 2B (observing that “[t]he race for the University of Michigan Board of Regents could determine . . . the future of [the University’s] affirmative action policies”); Kosseff, UM Policy May Hang On Election, Crain’s Detroit Business, Sept. 18, 2000, p. 1 (noting that an upcoming election could determine whether the University would continue to defend its affirmative action policies); University of Michigan’s Admissions Policy Still an Issue for Regents’ Election, Black Issues in Higher Education, Oct. 21, 2004, p. 17 (commenting that although “the Supreme Court struck down the University of Michigan’s undergraduate admissions policy as too formulaic,” the issue “remains an important [one] to several peo-ple running” in an upcoming election for the Board of Regents).

     Moreover, a careful examination of the boards and their governing structure reveals that they remain actively involved in setting admissions policies and procedures. Take Wayne State University, for example. Its Board of Governors has enacted university statutes that govern the day-to-day running of the institution. See Wayne State Univ. Stat., online at http://bog.wayne.edu/code. A number of those statutes establish general admissions procedures, see §2.34.09 (establishing undergraduate admissions procedures); §2.34.12 (establishing graduate admissions procedures), and some set out more specific instructions for university officials, see, e.g., §2.34.09.030 (“Admissions decisions will be based on a full evaluation of each student’s academic record, and on empirical data reflecting the characteristics of students who have successfully graduated from [the university] within the four years prior to the year in which the student applies”); §§2.34.12.080, 2.34.12.090 (setting the requisite grade point average for graduate applicants).

     The Board of Governors does give primary responsibility over day-to-day admissions matters to the university’s President. §2.34.09.080. But the President is “elected by and answerable to the Board.” Brief for Respondent Board of Governors of Wayne State University et al. 15. And while university officials and faculty members “serv[e] an important advisory role in recommending educational policy,” id., at 14, the Board alone ultimately controls educational policy and decides whether to adopt (or reject) program-specific admissions recommendations. For example, the Board has voted on recommendations “to revise guidelines for establishment of honors curricula, including admissions criteria”; “to modify the honor point criteria for graduate admission”; and “to modify the maximum number of transfer credits that the university would allow in certain cases where articulation agreements rendered modification appropriate.” Id., at 17; see also id., at 18–20 (providing examples of the Board’s “review[ing] and pass[ing] upon admissions requirements in the course of voting on broader issues, such as the implementation of new academic programs”). The Board also “engages in robust and regular review of administrative actions involving admissions policy and related matters.” Id., at 16.

     Other public universities more clearly entrust admissions policy to university officials.      The Board of Regents of the University of Michigan, for example, gives primary responsibility for admissions to the Associate Vice Provost, Executive Director of Undergraduate Admissions, and Directors of Admissions. Bylaws §8.01, online at http://www.regents.umich.edu/bylaws. And the Board of Trustees of Michigan State University relies on the President to make recommendations regarding admissions policies. Bylaws, Art. 8, online at http://www.trustees.msu.edu/bylaws. But the bylaws of the Board of Regents and the Board of Trustees “make clear that all university operations remain subject to their control.” Brief for Respondents Regents of the University of Michigan, the Board of Trustees of Michigan State University et al. 13–14.

     The boards retain ultimate authority to adopt or reject admissions policies in at least three ways. First, they routinely meet with university officials to review admissions policies, including race-sensitive admissions policies. For example, shortly after this Court’s decisions in Gratz v. Bollinger, 539 U. S. 244 (2003) , and Grutter, 539 U. S., at 306, the President of the University of Michigan appeared before the University’s Board of Regents to discuss the impact of those decisions on the University. SeeProceedings 2003–2004, pp. 10–12 (July 2003), onlineat http://name.umdl.umich.edu/ACW7513.2003.001. Six members of the Board voiced strong support for the University’s use of race as a factor in admissions. Id., at 11–12. In June 2004, the President again appeared before the Board to discuss changes to undergraduate admissions policies. Id., at 301 (June 2004). And in March 2007, the University’s Provost appeared before the Board of Regents to present strategies to increase diversity in light of the passage of Proposal 2. Proceedings 2006–2007, pp. 264–265 (Mar. 2007), online at http://name.umdl.umich.edu/ACW7513.2006.001.

     Second, the boards may enact bylaws with respect to specific admissions policies and may alter any admissions policies set by university officials. The Board of Regents may amend any bylaw “at any regular meeting of the board, or at any special meeting, provided notice is given to each regent one week in advance.” Bylaws §14.03. And Michigan State University’s Board of Trustees may, “[u]pon the recommendation of the President[,] . . . determine and establish the qualifications of students for admissions at any level.” Bylaws, Art. 8. The boards may also permanently remove certain admissions decisions from university officials.[10] This authority is not merely theoretical. Between 2008 and 2012, the University of Michigan’s Board of Regents “revised more than two dozen of its bylaws, two of which fall within Chapter VIII, the section regulating admissions practices.” App. to Pet. for Cert. 30a.

     Finally, the boards may appoint university officials who share their admissions goals, and they may remove those officials if the officials’ goals diverge from those of the boards. The University of Michigan’s Board of Regents “directly appoints [the University’s] Associate Vice Provost and Executive Director of Undergraduate Admissions,” and Michigan State University’s Board of Trustees elects that institution’s President. Brief for Respondents Regents of the University of Michigan, the Board of Trustees of Michigan State University et al. 14.

     The salient point is this: Although the elected and politically accountable boards may well entrust university officials with certain day-to-day admissions responsibilities, they often weigh in on admissions policies themselves and, at all times, they retain complete supervisory authority over university officials and over all admissionsdecisions.

     There is no question, then, that the elected boards in Michigan had the power to eliminate or adopt race-sensitive admissions policies prior to §26. There is also no question that §26 worked an impermissible reordering of the political process; it removed that power from the elected boards and placed it instead at a higher level of thepolitical process in Michigan. See supra, at 17–22. This case is no different from Hunter and Seattle in that respect. Just as in Hunter and Seattle, minorities in Michigan “participated in the political process and won.” Ante, at 5 (Breyer, J., concurring in judgment). And just as in Hunter and Seattle, “the majority’s subsequent reordering of the political process repealed the minority’s successes and made it more difficult for the minority to succeed in the future,” thereby “diminish[ing] the minority’s ability to participate meaningfully in the electoral process.” Ibid. There is therefore no need to consider “extend[ing] the holding of Hunter and Seattle to reach situations in which decisionmaking authority is moved from an administrative body to a political one,” ibid. Such a scenario is not be-fore us.

III

     The political-process doctrine not only resolves this case as a matter of stare decisis; it is correct as a matter of first principles.

A

     Under our Constitution, majority rule is not without limit. Our system of government is predicated on an equilibrium between the notion that a majority of citizens may determine governmental policy through legislation enacted by their elected representatives, and the overriding principle that there are nonetheless some things the Constitution forbids even a majority of citizens to do. The political-process doctrine, grounded in the Fourteenth Amendment, is a central check on majority rule.

     The Fourteenth Amendment instructs that all who act for the government may not “deny to any person . . . the equal protection of the laws.” We often think of equal protection as a guarantee that the government will apply the law in an equal fashion—that it will not intentionally discriminate against minority groups. But equal protection of the laws means more than that; it also secures the right of all citizens to participate meaningfully and equally in the process through which laws are created.

     Few rights are as fundamental as the right to participate meaningfully and equally in the process of government. See Yick Wo v. Hopkins, 118 U. S. 356, 370 (1886) (political rights are “fundamental” because they are “preservative of all rights”). That right is the bedrock of our democracy, recognized from its very inception. See J. Ely, Democracy and Distrust 87 (1980) (the Constitution “is overwhelmingly concerned, on the one hand, with procedural fairness in the resolution of individual disputes,” and on the other, “with ensuring broad participation in the processes and distributions of government”).

     This should come as no surprise. The political process is the channel of change. Id., at 103 (describing the importance of the judiciary in policing the “channels of political change”). It is the means by which citizens may both obtain desirable legislation and repeal undesirable legislation. Of course, we do not expect minority members of our society to obtain every single result they seek through the political process—not, at least, when their views conflict with those of the majority. The minority plainly does not have a right to prevail over majority groups in any given political contest. But the minority does have a right to play by the same rules as the majority. It is this right that Hunter and Seattle so boldly vindicated.

     This right was hardly novel at the time of Hunter and Seattle. For example, this Court focused on the vital importance of safeguarding minority groups’ access to the political process in United States v. Carolene Products Co., 304 U. S. 144 (1938) , a case that predated Hunter by 30 years. In a now-famous footnote, the Court explained that while ordinary social and economic legislation carries a presumption of constitutionality, the same may not be true of legislation that offends fundamental rights or targets minority groups. Citing cases involving restrictions on the right to vote, restraints on the dissemination of information, interferences with political organizations, and prohibition of peaceable assembly, the Court recognized that “legislation which restricts those political processes which can ordinarily be expected to bring about repeal of undesirable legislation” could be worthy of “more exacting judicial scrutiny under the general prohibitions of the Fourteenth Amendment than are most other types of legislation.” Id., at 152, n. 4; see also Ely, supra, at 76 (explaining that “[p]aragraph two [of Carolene Products footnote 4] suggests that it is an appropriate function of the Court to keep the machinery of democratic government running as it should, to make sure the channels of political participation and communication are kept open”). The Court also noted that “prejudice against discrete and insular minorities may be a special condition, which tends seriously to curtail the operation of those political pro-cesses ordinarily to be relied upon to protect minorities, and which may call for a correspondingly more search-ing judicial inquiry.” Carolene Products, 304 U. S., at 153, n. 4, see also Ely, supra, at 76 (explaining that “[p]aragraph three [of Carolene Products footnote 4] suggests that the Court should also concern itself with what majorities do to minorities, particularly mentioning laws ‘directed at’ religious, national and racial minorities and those infected by prejudice against them”).

     The values identified in Carolene Products lie at the heart of the political-process doctrine. Indeed, Seattle explicitly relied on Carolene Products. See 458 U. S., at 486 (“[W]hen the State’s allocation of power places unusual burdens on the ability of racial groups to enact legisla-tion specifically designed to overcome the ‘special condition’ of prejudice, the governmental action seriously‘curtail[s] the operation of those political processes ordinarily to be relied upon to protect minorities’ ” (quoting Carolene Products, 304 U. S., at 153, n. 4)). These values are central tenets of our equal protection jurisprudence.

     Our cases recognize at least three features of the right to meaningful participation in the political process. Two of them, thankfully, are uncontroversial. First, every eligible citizen has a right to vote. See Shaw v. Reno, 509 U. S. 630, 639 (1993) . This, woefully, has not always been the case. But it is a right no one would take issue with today. Second, the majority may not make it more difficult for the minority to exercise the right to vote. This, too, is widely accepted. After all, the Court has invalidated grandfather clauses, good character requirements, poll taxes, and gerrymandering provisions.[11] The third feature, the one the plurality dismantles today, is that a majority may not reconfigure the existing political process in a manner that creates a two-tiered system of political change, subjecting laws designed to protect or benefit discrete and insular minorities to a more burdensome political process than all other laws. This is the political-process doctrine of Hunter and Seattle.

     My colleagues would stop at the second. The plurality embraces the freedom of “self-government” without limits. See ante, at 13. And Justice Scalia values a “near-limitless” notion of state sovereignty. See ante, at 13 (opinion concurring in judgment). The wrong sought to be corrected by the political-process doctrine, they say, is not one that should concern us and is in any event beyond the reach of the Fourteenth Amendment. As they see it, the Court’s role in protecting the political process ends once we have removed certain barriers to the minority’s participation in that process. Then, they say, we must sit back and let the majority rule without the key constitutional limit recognized in Hunter and Seattle.

     That view drains the Fourteenth Amendment of one of its core teachings. Contrary to today’s decision, protecting the right to meaningful participation in the political process must mean more than simply removing barriers to participation. It must mean vigilantly policing the political process to ensure that the majority does not use other methods to prevent minority groups from partaking in that process on equal footing. Why? For the same reason we guard the right of every citizen to vote. If “[e]fforts to reduce the impact of minority votes, in contrast to direct attempts to block access to the ballot,” were “ ‘second-generation barriers’ ” to minority voting, Shelby County v. Holder, 570 U. S. ___, ___ (2013) (Ginsburg, J., dissenting) (slip op., at 5), efforts to reconfigure the political process in ways that uniquely disadvantage minority groups who have already long been disadvantaged are third-generation barriers. For as the Court recognized in Seattle, “minorities are no less powerless with the vote than without it when a racial criterion is used to assign governmental power in such a way as to exclude particular racial groups ‘from effective participation in the political proces[s].’ ”[12] 458 U. S., at 486.

     To accept the first two features of the right to meaningful participation in the political process, while renouncing the third, paves the way for the majority to do what it has done time and again throughout our Nation’s history: afford the minority the opportunity to participate, yet manipulate the ground rules so as to ensure the minority’s defeat. This is entirely at odds with our idea of equality under the law.

     To reiterate, none of this is to say that the political-process doctrine prohibits the exercise of democratic self-government. Nothing prevents a majority of citizens from pursuing or obtaining its preferred outcome in a political contest. Here, for instance, I agree with the plurality that Michiganders who were unhappy with Grutter were free to pursue an end to race-sensitive admissions policies in their State. See ante, at 16–17. They were free to elect governing boards that opposed race-sensitive admissions policies or, through public discourse and dialogue, to lobby the existing boards toward that end. They were also free to remove from the boards the authority to make any decisions with respect to admissions policies, as opposed to only decisions concerning race-sensitive admissions policies. But what the majority could not do, consistent with the Constitution, is change the ground rules of the political process in a manner that makes it more difficult for racial minorities alone to achieve their goals. In doing so, the majority effectively rigs the contest to guarantee a particular outcome. That is the very wrong the political-process doctrine seeks to remedy. The doctrine “hews to the unremarkable notion that when two competitors are running a race, one may not require the other to run twice as far or to scale obstacles not present in the first runner’s course.” BAMN v. Regents of Univ. of Michigan, 701 F. 3d 466, 474 (CA6 2012).

B

     The political-process doctrine also follows from the rest of our equal protection jurisprudence—in particular, our reapportionment and vote dilution cases. In those cases, the Court described the right to vote as “ ‘the essence of a democratic society.’ ” Shaw, 509 U. S., at 639. It rejected States’ use of ostensibly race-neutral measures to prevent minorities from exercising their political rights. See id., at 639–640. And it invalidated practices such as at-large electoral systems that reduce or nullify a minority group’s ability to vote as a cohesive unit, when those practices were adopted with a discriminatory purpose. Id., at 641. These cases, like the political-process doctrine, all sought to preserve the political rights of the minority.

     Two more recent cases involving discriminatory restructurings of the political process are also worthy of mention: Romer v. Evans, 517 U. S. 620 (1996) , and League of United Latin American Citizens v. Perry, 548 U. S. 399 (2006) (LULAC).

     Romer involved a Colorado constitutional amendment that removed from the local political process an issue primarily affecting gay and lesbian citizens. The amendment, enacted in response to a number of local ordinances prohibiting discrimination against gay citizens, repealed these ordinances and effectively prohibited the adoption of similar ordinances in the future without another amendment to the State Constitution. 517 U. S., at 623–624. Although the Court did not apply the political-process doctrine in Romer,[13] the case resonates with the principles undergirding the political-process doctrine. The Court rejected an attempt by the majority to transfer decision-making authority from localities (where the targeted minority group could influence the process) to state government (where it had less ability to participate effec-tively). See id., at 632 (describing this type of political restructuring as a “disability” on the minority group). Rather than being able to appeal to municipalities for policy changes, the Court commented, the minority was forced to “enlis[t] the citizenry of Colorado to amend the State Constitution,” id., at 631—just as in this case.

     LULAC, a Voting Rights Act case, involved an enactment by the Texas Legislature that redrew district lines for a number of Texas seats in the House of Representatives. 548 U. S., at 409 (plurality opinion). In striking down the enactment, the Court acknowledged the “ ‘long, well-documented history of discrimination’ ” in Texas that “ ‘touched upon the rights of . . . Hispanics to register, to vote, or to participate otherwise in the electoral process,’ ” id., at 439, and it observed that that the “ ‘political, social, and economic legacy of past discrimination’ . . . may well [have] ‘hinder[ed] their ability to participate effectively in the political process,’ ” id., at 440. Against this backdrop, the Court found that just as “Latino voters were poised to elect their candidate of choice,” id., at 438, the State’s enactment “took away [their] opportunity because [they] were about to exercise it,” id., at 440. The Court refused to sustain “the resulting vote dilution of a group that was beginning to achieve [the] goal of overcoming prior electoral discrimination.” Id., at 442.

     As in Romer, the LULAC Court—while using a different analytic framework—applied the core teaching of Hunter and Seattle: The political process cannot be restructured in a manner that makes it more difficult for a traditionally excluded group to work through the existing process to seek beneficial policies. And the events giving rise to LULAC are strikingly similar to those here. Just as redistricting prevented Latinos in Texas from attaining a benefit they had fought for and were poised to enjoy, §26prevents racial minorities in Michigan from enjoying a last-resort benefit that they, too, had fought for through the existing political processes.

IV

     My colleagues claim that the political-process doctrine is unadministrable and contrary to our more recent equal protection precedents. See ante, at 11–15 (plurality opinion); ante, at 7–17 (Scalia, J., concurring in judgment). It is only by not acknowledging certain strands of our jurisprudence that they can reach such a conclusion.

A

     Start with the claim that Hunter and Seattle are no longer viable because of the cases that have come after them. I note that in the view of many, it is those precedents that have departed from the mandate of the Equal Protection Clause in the first place, by applying strict scrutiny to actions designed to benefit rather than burden the minority. See Gratz, 539 U. S., at 301 (Ginsburg, J., dissenting) (“[A]s I see it, government decisionmakers may properly distinguish between policies of exclusion and inclusion. Actions designed to burden groups long denied full citizenship stature are not sensibly ranked with measures taken to hasten the day when entrenched discrimination and its aftereffects have been extirpated” (citation omitted)); id., at 282 (Breyer, J., concurring in judgment) (“I agree . . . that, in implementing the Constitution’s equality instruction, government decisionmakers may properly distinguish between policies of inclusion and exclusion, for the former are more likely to prove consistent with the basic constitutional obligation that the law respect each individual equally” (citation omitted)); Adarand Constructors, Inc. v. Peña, 515 U. S. 200, 243 (1995) (Stevens, J., dissenting) (“There is no moral or constitutional equivalence between a policy that is designed to perpetuate a caste system and one that seeks to eradicate racial subordination. Invidious discrimination is an engine of oppression, subjugating a disfavored group to enhance or maintain the power of the majority. Remedial race-based preferences reflect the opposite impulse: a desire to foster equality in society”); Wygant v. Jackson Bd. of Ed., 476 U. S. 267 –302 (1986) (Marshall, J., dissenting) (when dealing with an action to eliminate “pernicious vestiges of past discrimination,” a “less exacting standard of review is appropriate”); Fullilove v. Klutznick, 448 U. S. 448 –519 (1980) (Marshall, J., concurring in judgment) (race-based governmental action designed to “remed[y] the continuing effects of past racial discrimination . . . should not be subjected to conventional ‘strict scrutiny’ ”); Bakke, 438 U. S., at 359 (Brennan, White, Marshall, and Blackmun, JJ., concurring in judgment in part and dissenting in part) (“racial classifications designed to further remedial purposes” should be subjected only to intermediate scrutiny).

     But even assuming that strict scrutiny should apply to policies designed to benefit racial minorities, that view is not inconsistent with Hunter and Seattle. For nothing the Court has said in the last 32 years undermines the principles announced in those cases.

1

     Justice Scalia first argues that the political-process doctrine “misreads the Equal Protection Clause to protect ‘particular group[s],’ ” running counter to a line of cases that treat “ ‘equal protection as a personal right.’ ” Ante, at 9 (opinion concurring in judgment) (quoting Adarand, 515 U. S., at 230). Equal protection, he says, protects “ ‘persons, not groups.’ ” Ante, at 10 (quoting Adarand, 515 U. S., at 227). This criticism ignores the obvious: Discrimination against an individual occurs because of that individual’s membership in a particular group. Yes, equal protection is a personal right, but there can be no equal protection violation unless the injured individual is a member of a protected group or a class of individuals. It is membership in the group—here the racial minority—that gives rise to an equal protection violation.

     Relatedly, Justice Scalia argues that the political-process doctrine is inconsistent with our precedents because it protects only the minority from political restructurings. This aspect of the doctrine, he says, cannot be tolerated because our precedents have rejected “ ‘a reading of the guarantee of equal protection under which the level of scrutiny varies according to the ability of different groups to defend their interests in the representative process.’ ” Ante, at 10 (quoting Richmond v. J. A. Croson Co., 488 U. S., 469, 495 (1989) (plurality opinion)). Equal protection, he continues, “ ‘cannot mean one thing when applied to one individual and something else when applied to a person of another color.’ ” Ante, at 10 (quoting Bakke, 438 U. S., at 289–290) (opinion of Powell, J.).

     Justice Scalia is troubled that the political-process doctrine has not been applied to trigger strict scrutiny for political restructurings that burden the majority. But the doctrine is inapplicable to the majority. The minority cannot achieve such restructurings against the majority, for the majority is, well, the majority. As the Seattle Court explained, “ ‘[t]he majority needs no protection against discriminat[ory restructurings], and if it did, a referendum, [for instance], might be bothersome but no more than that.’ ” 458 U. S., at 468. Stated differently, the doctrine protects only the minority because it implicates a problem that affects only the minority. Nothing in my opinion suggests, as Justice Scalia says, that under the political-process doctrine, “the Constitution prohibits discrimination against minority groups, but not against majority groups.” Ante, at 10, n. 7. If the minority somehow managed to effectuate a political restructuring that burdened only the majority, we could decide then whether to apply the political-process doctrine to safeguard the political right of the majority. But such a restructuring is not before us, and I cannot fathom how it could be achieved.

2

     Justice Scalia next invokes state sovereignty, arguing that “we have emphasized the near-limitless sovereignty of each State to design its governing structure as it sees fit.” Ante, at 13 (opinion concurring in judgment). But state sovereignty is not absolute; it is subject to constitutional limits. The Court surely did not offend state sovereignty by barring States from changing their voting procedures to exclude racial minorities. So why does thepolitical-process doctrine offend state sovereignty? The doctrine takes nothing away from state sovereignty that the Equal Protection Clause does not require. All it says is that a State may not reconfigure its existing political processes in a manner that establishes a distinct and more burdensome process for minority members of our society alone to obtain legislation in their interests.

     More broadly, Justice Scalia is troubled that the political-process doctrine would create supposed “affirmative-action safe havens” in places where the ordinary political process has thus far produced race-sensitive admissions policies. Ante, at 13–14. It would not. As explained previously, the voters in Michigan who opposed race-sensitive admissions policies had any number of options available to them to challenge those policies. See supra, at 34–35. And in States where decisions regarding race-sensitive admissions policies are not subject to the political process in the first place, voters are entirely free to eliminate such policies via a constitutional amendment because that action would not reallocate power in the manner condemned in Hunter and Seattle (and, of course, present here). The Seattle Court recognized this careful balance between state sovereignty and constitutional protections:

“[W]e do not undervalue the magnitude of the State’s interest in its system of education. Washington could have reserved to state officials the right to make all decisions in the areas of education and student assignment. It has chosen, however, to use a more elaborate system; having done so, the State is obligated to operate that system within the confines of the Fourteenth Amendment.” 458 U. S., at 487.

The same is true of Michigan.

3

     Finally, Justice Scalia disagrees with “the proposition that a facially neutral law may deny equal protection solely because it has a disparate racial impact.” Ante,at 15 (opinion concurring in judgment). He would acknowledge, however, that an act that draws racial distinctions or makes racial classifications triggers strict scrutiny regardless of whether discriminatory intent is shown. See Adarand, 515 U. S., at 213. That should settle the matter: Section 26 draws a racial distinction. As the Seattle Court explained, “when the political process or the decisionmaking mechanism used to address racially conscious legislation—and only such legislation—is singled out for peculiar and disadvantageous treatment, the governmental action plainly rests on ‘distinctions based on race.’ ” 458 U. S., at 485 (some internal quotation marks omitted); see also id., at 470 (noting that although a State may “ ‘allocate governmental power on the basis of any general principle,’ ” it may not use racial considerations “to define the governmental decisionmaking structure”).

     But in Justice Scalia’s view, cases like Washington v. Davis, 426 U. S. 229 (1976) , and Arlington Heights v. Metropolitan Housing Development Corp., 429 U. S. 252 (1977) , call Seattle into question. It is odd to suggest that prior precedents call into question a later one. Seattle (decided in 1982) postdated both Washington v. Davis (1976) and Arlington Heights (1977). Justice Scalia’s suggestion that Seattle runs afoul of the principles established in Washington v. Davis and Arlington Heights would come as a surprise to Justice Blackmun, who joined the majority opinions in all three cases. Indeed, the Seattle Court explicitly rejected the argument that Hunter had been effectively overruled by Washington v. Davis and Arlington Heights:

“There is one immediate and crucial difference between Hunter and [those cases]. While decisions such as Washington v. Davis and Arlington Heights considered classifications facially unrelated to race, the charter amendment at issue in Hunter dealt in explicitly racial terms with legislation designed to benefit minorities ‘as minorities,’ not legislation intended to benefit some larger group of underprivileged citizens among whom minorities were disproportionately represented.” 458 U. S., at 485.

And it concluded that both the Hunter amendment and the Seattle initiative rested on distinctions based on race. 458 U. S., at 485. So does §26.[14]

B

     My colleagues also attack the first prong of the doctrine as “rais[ing] serious constitutional concerns,” ante, at 11 (plurality opinion), and being “unadministrable,” ante, at 7 (Scalia, J., concurring in judgment). Justice Scalia wonders whether judges are equipped to weigh in on what constitutes a “racial issue.” See ante, at 8. The plurality, too, thinks courts would be “with no clear legal standards or accepted sources to guide judicial decision.” Ante, at 12. Yet as Justice Scalia recognizes, Hunter and Seattle provide a standard: Does the public policy at issue “inur[e] primarily to the benefit of the minority, and [was it] designed for that purpose”? Seattle, 458 U. S., at 472; see ante, at 8. Surely this is the kind of factual inquiry that judges are capable of making. Justice Scalia, for instance, accepts the standard announced in Washington v. Davis, which requires judges to determine whether discrimination is intentional or whether it merely has a discriminatory effect. Such an inquiry is at least as difficult for judges as the one called for by Hunter and Seattle. In any event, it is clear that the constitutional amendment in this case has a racial focus; it is facially race-based and, by operation of law, disadvantages only minorities. See supra, at 15–16.

     “No good can come” from these inquiries, Justice Scalia responds, because they divide the Nation along racial lines and perpetuate racial stereotypes. Ante, at 9. The plurality shares that view; it tells us that we must not assume all individuals of the same race think alike. See ante, at 11–12. The same could have been said about desegregation: Not all members of a racial minority in Seattle necessarily regarded the integration of public schools as good policy. Yet the Seattle Court had little difficulty saying that school integration as a general matter “inure[d] . . . to the benefit of” the minority. 458 U. S., at 472.

     My colleagues are of the view that we should leave race out of the picture entirely and let the voters sort it out. See ante, at 13 (plurality opinion) (“Racial division would be validated, not discouraged, were the Seattle formulation . . . to remain in force”); ante, at 9 (Scalia, J., concurring in judgment) (“ ‘[R]acial stereotyping [is] at odds with equal protection mandates’ ”). We have seen this reasoning before. See Parents Involved, 551 U. S., at 748 (“The way to stop discrimination on the basis of race is to stop discriminating on the basis of race”). It is a sentiment out of touch with reality, one not required by our Constitution, and one that has properly been rejected as “not sufficient” to resolve cases of this nature. Id., at 788 (Kennedy, J., concurring in part and concurring in judgment). While “[t]he enduring hope is that race should not matter[,] the reality is that too often it does.” Id., at 787. “[R]acial discrimination . . . [is] not ancient history.” Bartlett v. Strickland, 556 U. S. 1, 25 (2009) (plurality opinion).

     Race matters. Race matters in part because of the long history of racial minorities’ being denied access to the political process. See Part I, supra; see also South Carolina v. Katzenbach, 383 U. S. 301, 309 (1966) (describingracial discrimination in voting as “an insidious and pervasive evil which had been perpetuated in certain parts of our country through unremitting and ingenious defiance of the Constitution”). And although we have made great strides, “voting discrimination still exists; no one doubts that.” Shelby County, 570 U. S., at __ (slip op., at 2).

     Race also matters because of persistent racial inequality in society—inequality that cannot be ignored and that has produced stark socioeconomic disparities. See Gratz, 539 U. S., at 298–300 (Ginsburg, J., dissenting) (cataloging the many ways in which “the effects of centuries of law-sanctioned inequality remain painfully evident in our communities and schools,” in areas like employment, poverty, access to health care, housing, consumer transactions, and education); Adarand, 515 U. S., at 273 (Ginsburg, J., dissenting) (recognizing that the “lingering effects” of discrimination, “reflective of a system of racial caste only recently ended, are evident in our workplaces, markets, and neighborhoods”).

     And race matters for reasons that really are only skin deep, that cannot be discussed any other way, and that cannot be wished away. Race matters to a young man’s view of society when he spends his teenage years watching others tense up as he passes, no matter the neighborhood where he grew up. Race matters to a young woman’s sense of self when she states her hometown, and then is pressed, “No, where are you really from?”, regardless of how many generations her family has been in the country. Race matters to a young person addressed by a stranger in a foreign language, which he does not understand because only English was spoken at home. Race matters because of the slights, the snickers, the silent judgments that reinforce that most crippling of thoughts: “I do not belong here.”

     In my colleagues’ view, examining the racial impact of legislation only perpetuates racial discrimination. This refusal to accept the stark reality that race matters is regrettable. The way to stop discrimination on the basis of race is to speak openly and candidly on the subject of race, and to apply the Constitution with eyes open to the unfortunate effects of centuries of racial discrimination. As members of the judiciary tasked with intervening to carry out the guarantee of equal protection, we ought not sit back and wish away, rather than confront, the racial inequality that exists in our society. It is this view that works harm, by perpetuating the facile notion that what makes race matter is acknowledging the simple truth that race does matter.

V

     Although the only constitutional rights at stake in this case are process-based rights, the substantive policy at issue is undeniably of some relevance to my colleagues. See ante, at 18 (plurality opinion) (suggesting that race-sensitive admissions policies have the “potential to become . . . the source of the very resentments and hostilities based on race that this Nation seeks to put behind it”). I will therefore speak in response.

A

     For over a century, racial minorities in Michigan fought to bring diversity to their State’s public colleges and universities. Before the advent of race-sensitive admissions policies, those institutions, like others around the country, were essentially segregated. In 1868, two black students were admitted to the University of Michigan, the first of their race. See Expert Report of James D. Anderson 4, in Gratz v. Bollinger, No. 97–75231 (ED Mich.). In 1935, over six decades later, there were still only 35 black students at the University. Ibid. By 1954, this number had risen to slightly below 200. Ibid. And by 1966, to around 400, among a total student population of roughly 32,500—barely over 1 percent. Ibid. The numbers at the University of Michigan Law School are even more telling. Duringthe 1960’s, the Law School produced 9 black graduates among a total of 3,041—less than three-tenths of 1 percent. See App. in Grutter v. Bollinger, O. T. 2002, No. 02–241, p. 204.

     The housing and extracurricular policies at these institutions also perpetuated open segregation. For instance, incoming students were permitted to opt out of rooming with black students. Anderson, supra, at 7–8. And some fraternities and sororities excluded black students from membership. Id., at 6–7.

     In 1966, the Defense Department conducted an investigation into the University’s compliance with Title VI of the Civil Rights Act, and made 25 recommendations for increasing opportunities for minority students. Id., at 9. In 1970, a student group launched a number of protests, including a strike, demanding that the University increase its minority enrollment. Id., at 16–23. The University’s Board of Regents responded, adopting a goal of 10 percent black admissions by the fall of 1973. Id., at 23.

     During the 1970’s, the University continued to improve its admissions policies,[15] encouraged by this Court’s 1978 decision in Bakke. In that case, the Court told our Nation’s colleges and universities that they could consider race in admissions as part of a broader goal to create a diverse student body, in which students of different backgrounds would learn together, and thereby learn to live together. A little more than a decade ago, in Grutter, the Court reaffirmed this understanding. In upholding the admissions policy of the Law School, the Court laid to rest any doubt whether student body diversity is a compelling interest that may justify the use of race.

     Race-sensitive admissions policies are now a thing of the past in Michigan after §26, even though—as experts agree and as research shows—those policies were making a difference in achieving educational diversity. In Grutter, Michigan’s Law School spoke candidly about the strides the institution had taken successfully because of race-sensitive admissions. One expert retained by the Law School opined that a race-blind admissions system would have a “very dramatic, negative effect on underrepresented minority admissions.” Grutter, 539 U. S., at 320 (inter-nal quotation marks omitted). He testified that the school had admitted 35 percent of underrepresented minority students who had applied in 2000, as opposed to only 10 percent who would have been admitted had race not been considered. Ibid. Underrepresented minority students would thus have constituted 4 percent, as opposed to the actual 14.5 percent, of the class that entered in 2000. Ibid.

     Michigan’s public colleges and universities tell us the same today. The Board of Regents of the University of Michigan and the Board of Trustees of Michigan State University inform us that those institutions cannot achieve the benefits of a diverse student body without race-sensitive admissions plans. See Brief for Respondents Regents of the University of Michigan, the Board of Trustees of Michigan State University et al. 18–25. During proceedings before the lower courts, several university officials testified that §26 would depress minority enrollment at Michigan’s public universities. The Director of Undergraduate Admissions at the University of Michigan “expressed doubts over the ability to maintain minority enrollment through the use of a proxy, like socioeconomic status.” Supp. App. to Pet. for Cert. 285a. He explained that university officials in States with laws similar to §26 had not “ ‘achieve[d] the same sort of racial and ethnic diversity that they had prior to such measures . . . without considering race.’ ” Ibid. Similarly, the Law School’s Dean of Admissions testified that she expected “a decline in minority admissions because, in her view, it is impossible ‘to get a critical mass of underrepresented minorities . . . without considering race.’ ” Ibid. And the Dean of Wayne State University Law School stated that “although some creative approaches might mitigate the effects of [§26], he ‘did not think that any one of these proposals or any combination of these proposals was reasonably likely to result in the admission of a class that had the same or similar or higher numbers of African Americans, Latinos and Native Americans as the prior policy.’ ” Ibid.

     Michigan tells a different story. It asserts that although the statistics are difficult to track, “the number of underrepresented minorities . . . [in] the entering freshman class at Michigan as a percentage changed very little” after §26. Tr. of Oral Arg. 15. It also claims that “the statistics in California across the 17 campuses in the University of California system show that today the underrepresented minority percentage is better on 16 out of those 17 campuses”—all except Berkeley—than before California’s equivalent initiative took effect. Id., at 16. As it turns out, these statistics weren’t “ ‘even good enough to be wrong.’ ” Reference Manual on Scientific Evidence 4 (2d ed. 2000) (Introduction by Stephen G. Breyer (quoting Wolfgang Pauli)).

     Section 26 has already led to decreased minority enrollment at Michigan’s public colleges and universities. In 2006 (before §26 took effect), underrepresented minorities made up 12.15 percent of the University of Michigan’s freshman class, compared to 9.54 percent in 2012—a roughly 25 percent decline. See University of Michigan—New Freshman Enrollment Overview, Office of the Registrar, online at http://www.ro.umich.edu/report/10enrolloverview.pdf and http://www.ro.umich.edu/report/12enrollmentsummary.pdf.[16] Moreover, the total number of college-aged underrepresented minorities in Michigan has increased even as the number of underrepresented minorities admitted to the University has decreased. For example, between 2006 and 2011, the proportion of black freshmen among those enrolled at the University of Michigan declined from 7 percent to 5 percent, even though the proportion of black college-aged persons in Michigan increased from 16 to 19 percent. See Fessenden and Keller, How Minorities Have Fared in States with Affirmative Action Bans, N. Y. Times, June 24, 2013, online at http://www.nytimes.com/interactive/2013/06/24/us/affirmative-action-bans.html.

UNIVERSITY OF MICHIGAN

Black Students[17]

     A recent study also confirms that §26 has decreased minority degree attainment in Michigan. The University of Michigan’s graduating class of 2012, the first admitted after §26 took effect, is quite different from previous classes. The proportion of black students among those attaining bachelor’s degrees was 4.4 percent, the lowest since 1991; the proportion of black students among those attaining master’s degrees was 5.1 percent, the lowest since 1989; the proportion of black students among those attaining doctoral degrees was 3.9 percent, the lowest since 1993; and the proportion of black students among those attaining professional school degrees was 3.5 percent, the lowest since the mid-1970’s. See Kidder, Restructuring Higher Education Opportunity?: African American Degree Attainment After Michigan’s Ban on Affirmative Action, p. 1 (Aug. 2013), online at http://papers.ssrn.com/sol3/abstract=2318523.

     The President and Chancellors of the University of California (which has 10 campuses, not 17) inform us that “[t]he abandonment of race-conscious admissions policies resulted in an immediate and precipitous decline in the rates at which underrepresented-minority students applied to, were admitted to, and enrolled at” the university. Brief for President and Chancellors of the University of California as Amici Curiae 10 (hereinafter President and Chancellors Brief). At the University of California, Los Angeles (UCLA), for example, admission rates for underrepresented minorities plummeted from 52.4 percent in 1995 (before California’s ban took effect) to 24 percentin 1998. Id., at 12. As a result, the percentage of underrepresented minorities fell by more than half: from 30.1 percent of the entering class in 1995 to 14.3 percent in 1998. Ibid. The admissions rate for underrepresented minorities at UCLA reached a new low of 13.6 percent in 2012. See Brief for California Social Science Researchers and Admissions Experts as Amici Curiae 28.

     The elimination of race-sensitive admissions policies in California has been especially harmful to black students. In 2006, for example, there were fewer than 100 black students in UCLA’s incoming class of roughly 5,000, the lowest number since at least 1973. See id., at 24.

     The University of California also saw declines in minor-ity representation at its graduate programs and professional schools. In 2005, underrepresented minorities made up 17 percent of the university’s new medical students, which is actually a lower rate than the 17.4 percent reported in 1975, three years before Bakke. President and Chancellors Brief 13. The numbers at the law schools are even more alarming. In 2005, underrepresented minorities made up 12 percent of entering law students, well below the 20.1 percent in 1975. Id., at 14.

     As in Michigan, the declines in minority representation at the University of California have come even as the minority population in California has increased. At UCLA, for example, the proportion of Hispanic freshmen among those enrolled declined from 23 percent in 1995 to 17 percent in 2011, even though the proportion of Hispanic college-aged persons in California increased from 41 percent to 49 percent during that same period. See Fessenden and Keller.

UCLA

Hispanic Students[18]

     And the proportion of black freshmen among those enrolled at UCLA declined from 8 percent in 1995 to 3 percent in 2011, even though the proportion of blackcollege-aged persons in California increased from 8 percent to 9 percent during that same period. See ibid.

UCLA

Black Students[19]

     While the minority admissions rates at UCLA and Berkeley have decreased, the number of minorities enrolled at colleges across the county has increased. See Phillips, Colleges Straining to Restore Diversity: Bans on Race-Conscious Admissions Upend Racial Makeup at California Schools, Wall Street Journal, Mar. 7, 2014,p. A3.

BERKELEY AND UCLA[20]

     The President and Chancellors assure us that they have tried. They tell us that notwithstanding the university’s efforts for the past 15 years “to increase diversity on [the University of California’s] campuses through the use of race-neutral initiatives,” enrollment rates have “not rebounded . . . [or] kept pace with the demographic changes among California’s graduating high-school population.” President and Chancellors Brief 14. Since Proposition 209 took effect, the university has spent over a half-billion dollars on programs and policies designed to increase diversity. Phillips, supra, at A3. Still, it has been unable to meet its diversity goals. Ibid. Proposition 209, it says, has “ ‘completely changed the character’ of the university.” Ibid. (quoting the Associate President and Chief Policy Advisor of the University of California).

B

     These statistics may not influence the views of some of my colleagues, as they question the wisdom of adopting race-sensitive admissions policies and would prefer if our Nation’s colleges and universities were to discard those policies altogether. See ante, at 2 (Roberts, C. J., concurring) (suggesting that race-sensitive admissions policies might “do more harm than good”); ante, at 9, n. 6 (Scalia, J., concurring in judgment); Grutter, 539 U. S., at 371–373 (Thomas, J., concurring in part and dissenting in part); id., at 347–348 (Scalia, J., concurring in part and dissenting in part). That view is at odds with our recognition in Grutter, and more recently in Fisher v. University of Texas at Austin, 570 U. S. ___ (2013), that race-sensitive admissions policies are necessary to achieve a diverse student body when race-neutral alternatives have failed. More fundamentally, it ignores the importance of diversity in institutions of higher education and reveals how little my colleagues understand about the reality of race in America.

     This Court has recognized that diversity in education is paramount. With good reason. Diversity ensures that the next generation moves beyond the stereotypes, the assumptions, and the superficial perceptions that students coming from less-heterogeneous communities may harbor, consciously or not, about people who do not look like them. Recognizing the need for diversity acknowledges that, “[j]ust as growing up in a particular region or having particular professional experiences is likely to affect an individual’s views, so too is one’s own, unique experience of being a racial minority in a society, like our own, in which race unfortunately still matters.” Grutter, 539 U. S., at 333. And it acknowledges that “to cultivate a set of leaders with legitimacy in the eyes of the citizenry, it is necessary that the path to leadership be visibly open to talented and qualified individuals of every race and ethnicity.” Id., at 332.

     Colleges and universities must be free to prioritize the goal of diversity. They must be free to immerse their students in a multiracial environment that fosters frequent and meaningful interactions with students of other races, and thereby pushes such students to transcend any assumptions they may hold on the basis of skin color. Without race-sensitive admissions policies, this might well be impossible. The statistics I have described make that fact glaringly obvious. We should not turn a blind eye to something we cannot help but see.

     To be clear, I do not mean to suggest that the virtues of adopting race-sensitive admissions policies should inform the legal question before the Court today regarding the constitutionality of §26. But I cannot ignore the unfortunate outcome of today’s decision: Short of amending the State Constitution, a Herculean task, racial minorities in Michigan are deprived of even an opportunity to convince Michigan’s public colleges and universities to consider race in their admissions plans when other attempts to achieve racial diversity have proved unworkable, and those institutions are unnecessarily hobbled in their pursuit of a diverse student body.

*  *  *

     The Constitution does not protect racial minorities from political defeat. But neither does it give the majority free rein to erect selective barriers against racial minorities. The political-process doctrine polices the channels of change to ensure that the majority, when it wins, does so without rigging the rules of the game to ensure its success. Today, the Court discards that doctrine without good reason.

     In doing so, it permits the decision of a majority of the voters in Michigan to strip Michigan’s elected university boards of their authority to make decisions with respect to constitutionally permissible race-sensitive admissions policies, while preserving the boards’ plenary authority to make all other educational decisions. “In a most direct sense, this implicates the judiciary’s special role in safeguarding the interests of those groups that are relegated to such a position of political powerlessness as to command extraordinary protection from the majoritarian political process.” Seattle, 458 U. S., at 486 (internal quotation marks omitted). The Court abdicates that role, permitting the majority to use its numerical advantage to change the rules mid-contest and forever stack the deck against racial minorities in Michigan. The result is that Michigan’s public colleges and universities are less equipped to do their part in ensuring that students of all races are “better prepare[d] . . . for an increasingly diverse workforce and society . . .” Grutter, 539 U. S., at 330 (internal quotation marks omitted).

     Today’s decision eviscerates an important strand of our equal protection jurisprudence. For members of historically marginalized groups, which rely on the federal courts to protect their constitutional rights, the decision can hardly bolster hope for a vision of democracy that preserves for all the right to participate meaningfully and equally in self-government.

     I respectfully dissent.