Coalition to Defend Affirmative Action, Integration & Immigrant Rights & Fight for Equality by Any Means Necessary v. Regents of the University

SUTTON, Circuit Judge,

dissenting.

I join Judge Gibbons’ dissent and write separately to make a few additional points.

Today’s lawsuit transforms a potential virtue of affirmative action into a vice. If there is one feature of affirmative-action programs that favors their constitutionality, it is that they grow out of the democratic process: the choice of a majority of a State’s residents to create race-conscious admissions preferences at their public universities not to benefit a majority race but to facilitate the educational opportunities of disadvantaged racial minorities. Such democratically enacted programs, like all democratically enacted laws, deserve initial respect in the courts, whether the particulars of a program satisfy the Fourteenth Amendment, see Grutter v. Bollinger, 539 U.S. 306, 343 , 123 S.Ct. 2325 , 156 L.Ed.2d 304 (2003), or violate it, see Gratz v. Bollinger, 539 U.S. 244, 275-76 , 123 S.Ct. 2411 , 156 L.Ed.2d 257 (2003).

Yet this lawsuit turns these assumptions on their head. Democracy, it turns out, has nothing to do with it. Plaintiffs insist that the Fourteenth Amendment’s guarantee of “equal protection of the laws” imposes two new rules on the policy debates surrounding affirmative action in higher education. Rule one: States not only may establish race-conscious affirmative-action programs, but they must do so to comply with the Fourteenth Amendment. Rule two: even if the Fourteenth Amendment does not mandate that States establish affirmative-action programs at their public universities, it bars them from eliminating such programs through amendments to their constitutions.

A.

The first theory has little to recommend it, so little that the notion of mandatory affirmative action will come as a surprise to all Justices of the United States Supreme Court, past and present, who have labored to determine whether state universities may ever enact such race-conscious programs under the United States Constitution. No Justice has taken the position that this recurring and vexing debate has all been a distraction, one that overlooked the hidden truth that the Fourteenth Amendment necessarily permits affirmative-action programs because it demands them.

Plaintiffs nonetheless insist that, “to the extent that [Proposal 2] ... bar[s] race or gender conscious programs that would be permissible under the Fourteenth Amendment, it violates the Equal Protection Clause.” Coalition First Br. at 39; see also, e.g., Oral Arg. at 7:36-45. Yet the words of the one amendment (prohibiting the State from “discriminat[ing] ... on the basis of race”) cannot violate the words of the other (“nor shall any State deny to any person ... the equal protection of the laws”).

That is especially true in the context of classifications based on race, which are presumptively unconstitutional and which *506 must run the gauntlet of strict scrutiny to survive. See Gratz, 539 U.S. at 270 , 123 S.Ct. 2411 . If racial preferences are only occasionally and barely constitutional, it cannot be the case that they are always required. A State that wishes to treat citizens of all races and nationalities equally “is free as a matter of its own law” to do so. Oregon v. Hass, 420 U.S. 714, 719 , 95 S.Ct. 1215 , 43 L.Ed.2d 570 (1975). A first premise for resolving this case is, and must be, that a State does not deny equal treatment by mandating it.

B.

The claimants’ other theory is of a piece. Having argued that the people of Michigan may not resort to the political process to eliminate racial preferences because the Fourteenth Amendment demands them, the claimants alternatively insist that the “political process doctrine” of the Fourteenth Amendment separately prohibits the State from eliminating such programs already in existence by way of a state constitutional amendment. Coalition First Br. at 27. That is not much of an alternative, as it comes to the same end. More fundamentally, the argument misapprehends what States may do as a matter of “politics” and “process.” Under the realm of politics, as just shown, the people of a State may choose to end rather than continue affirmative-action programs. Under the realm of process, the people of a State are free to use amendments to their constitution — the same charter of state government that delegated power to create affirmative-action programs in the first place— as the vehicle for making the change.

It is not that easy, plaintiffs insist. Even if States may do all of these things, the political-process doctrine prohibits States from altering their constitutions in a way that places “special burden[s] on racial minorities within the governmental process.” Hunter v. Erickson, 393 U.S. 385, 391 , 89 S.Ct. 557 , 21 L.Ed.2d 616 (1969); see Coalition First Br. at 28. Yes, of course. But Michigan did no such thing.

By any reasonable measure, Proposal 2 does not place “special burdens” on racial minorities. It bans “discriminating] against, or granting] 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. That is not a natural way to impose race-based burdens. The words of the amendment place no burden on anyone, and indeed are designed to prohibit the State from burdening one racial group relative to another. All of this furthers the objectives of the Fourteenth Amendment, the same seed from which the political-process doctrine sprouted.

That the people of Michigan made this change through their Constitution, as opposed to state legislation or a new policy embraced by the governing boards at the three state universities, does not impose a “special burden” on any racial minority. There is nothing unusual about placing an equal-protection guarantee in a constitution. That is where individual-liberty guarantees often go, and that after all is where the national framers placed the federal counterpart. States need not place equal-protection guarantees at the structural location of the plaintiffs’ choosing, be it at the governing boards of each university, the faculty of each university or the admissions office of each university. Instead of neutralizing the political process, that approach would skew it.

What else at any rate could the people of Michigan have done? Keep in mind that Proposal 2 applies not just to “public education” but also to “public employ *507 ment” and to “public contracting.” Neither the governing boards of the universities nor the universities’ faculties nor for that matter each university’s admissions committee oversees “public employment” or “public contracting” for the entire State. And the Michigan Constitution prohibits the legislature from interfering in “the university’s sphere of educational authority.” Federated Publ’ns, Inc. v. Bd. of Trustees of Mich. State Univ., 460 Mich. 75 , 594 N.W.2d 491, 497 (1999). That left the State just one option for addressing racial preferences in all three areas: a statewide constitutional amendment. This time-hallowed option places no special burden on proponents of affirmative action other than the customary burdens placed on anyone seeking to pass a constitutional amendment.

The charge that the Federal Constitution prohibits States from banning racial preferences through amendments to their constitutions also fails to account for one of the most fervent criticisms of state constitutions: They are too easy to change. The referendum option facilitates change, as it makes altering the constitution as easy as (if not easier than) altering legislation. After obtaining the requisite number of signatures (10% of the votes cast in the last gubernatorial election, see Mich. Const, art. XII, § 2), the proponents of change need to obtain just 51% of a popular vote and are spared the need to obtain the consent of the Governor and each house of the legislature. That is why the former Chief Justice of the California Supreme Court, faced with an equally variable state constitution, decries the “perpetual instability of California’s state constitutional law.” See Ronald M. George, Keynote Address, 62 Stan. L. Rev. 1515 , 1516 (2010).

Nothing prevents proponents of affirmative action from borrowing a page from the same playbook in a future state referendum — unless, that is, 51 % of Michigan voters do not support the change. But if the caveat applies, the answer is not to resort to the political-process doctrine, the goal of which is to promote neutral democratic means and ends, not to ban them. The short of it is that a bare requirement of a 51% popular vote as a vehicle for constitutional change is as good as it gets — at least from the perspective of proponents of future change. States may create higher, super-majority impediments to change. But they cannot create super-minority facilitations of it.

Of the 51 constitutions in this country, moreover, the most difficult to change is the Federal Constitution, which requires three-quarters of the States to ratify an amendment. Now that is a difficult charter to change. Is it really possible that the same Federal Constitution that is nearly impervious to change precludes a State from housing its elimination of racial preferences in a constitution that requires just a 51% popular vote for passage? Doubtful.

Odder still, the United States Constitution generally does not meddle in the way the States structure their governments. Sailors v. Bd. of Educ. of Kent Cnty., 387 U.S. 105, 109 , 87 S.Ct. 1549 , 18 L.Ed.2d 650 (1967). If they want a unicameral legislature, they can have it. See Neb. Const, art. Ill, § 1. And if they want direct democracy through constitutional and legislative referenda, they can have it. Pac. States Tel. & Tel. Co. v. Oregon, 223 U.S. 118, 151 , 32 S.Ct. 224 , 56 L.Ed. 377 (1912). If even the “republican form of government” guarantee is unenforceable, see U.S. Const, art. IV, § 4; Luther v. Borden, 48 U.S. (7 How.) 1, 42 , 12 L.Ed. 581 (1849), it is surely the case that the Fourteenth Amendment does not dictate the level of government at which a State *508 must enact a statewide ban on race discrimination.

It is worth considering plaintiffs’ alternative to this straightforward interpretation of the Fourteenth Amendment, the one every other court to face this issue has embraced, including our own when the plaintiffs sought to prevent Proposal 2 from going into effect in 2007. See Coal. to Defend Affirmative Action v. Granholm, 473 F.3d 237 , 250 (6th Cir.2006); Coal. for Econ. Equity v. Wilson, 122 F.3d 692 , 707 (9th Cir.1997); Coral Constr., Inc. v. City & Cnty. of San Francisco, 50 Cal.4th 315 , 113 Cal.Rptr.3d 279 , 235 P.3d 947, 960 (2010). Instead of allowing the people of Michigan to end racial preferences through a statewide popular vote that amends the State Constitution, they insist the Fourteenth Amendment permits this change in one, and only one, way: multiple elections over multiple years. Their theory contains several premises and several steps. One: any change may not come through legislation because the Michigan Constitution puts the governing boards at each of the three public universities (Michigan, Michigan State and Wayne State) in charge of educational policy, including apparently admissions policies, at each university. Two: there are eight members of the Board at each university, and two of them stand for election every two years in statewide elections. Three: these statewide elections are the only neutral way to permit opponents of racial preferences to establish such an admissions policy. Four: after eight years, the opponents of racial preferences will have had a chance to replace all members of the three Boards and presumably by then, if not a few years before, would have the votes to end racial preferences at each university. And all of this explains only the rules for Michigan’s three public universities. Anyone wishing to change admissions policies at Michigan’s other institutions of higher education faces an equally elaborate process. See Mich. Const, art. VIII, § 6 (authorizing the elected governor to appoint members of regional university boards to staggered eight-year terms); id. § 7 (authorizing locally elected boards to regulate community and junior colleges).

Whatever else one might say about the path on which this interpretation of the Fourteenth Amendment takes us, it does not follow Occam’s razor in getting there. Yet needless complexity is the least of the problems raised by this theory. How would the supporters of Proposal 2 end racial preferences in public contracting and public employment, which the university boards do not oversee? Does plaintiffs’ theory mean opponents of racial preferences must do both — win a statewide referendum and win twenty-four statewide individual elections? Must proponents of affirmative action do the same to reverse a contrary policy? Even if we ignore public contracting and employment, who says a single 10% petition drive and a single 51% popular vote make life more difficult for proponents of change than twenty-four statewide elections (among others) for twenty-four individuals over an eight-year period? Where are the empirical studies to prove the point? Should not the proponents of invalidating a state constitutional amendment be expected to prove the premise before the court accepts this invitation? Could the supporters of Proposal 2 amend the Michigan Constitution to remove control of the universities from their governing boards and place it in the hands of the legislature? What then? Would statewide legislation banning (or reinstating) racial preferences be permissible? There are many questions here, and all of them counsel against adopting this disfig *509 uring interpretation of the Fourteenth Amendment.

I do not doubt that Proposal 2 places a burden on proponents of affirmative action: They no longer have access to it, and they must amend the constitution to get it back. But the Fourteenth Amendment insists only that all participants in the debate have an equal shot. It does not ensure victory for one side or the other in this or that policy debate. It would be paradoxical if something called the “political process doctrine” insulated one side of a vigorous policy debate from a timeless rule of politics: win some, lose some. Winning some and losing some, as it happens, is just what has occurred across the country with statewide ballots seeking to eliminate racial preferences. While some initiatives have prevailed (California, Michigan, Nebraska and Washington), others have failed (Colorado, Missouri and Oklahoma). See Tim Hoover, Amendment 46 Fizzling Out, Denver Post, Nov. 7, 2008 at B4.

Another oddity of this theory is that it would apply even if the Michigan Constitution eliminated affirmative-action programs in another way. In 1963, the people of Michigan passed an earlier amendment to their Constitution, one that prohibited race discrimination by governmental entities. See Mich. Const, art. I, § 2. In view of this prohibition, a Michigan resident surely would have the right to bring a claim that the State Constitution’s existing prohibition on race-based classifications bars a system of racial preferences in admissions, contracting and employment. If there is one thing that the closely divided decisions in Regents of Univ. of Cal. v. Bakke, 438 U.S. 265 , 98 S.Ct. 2733 , 57 L.Ed.2d 750 (1978), Gratz and Grutter illustrate, it is that the Michigan Supreme Court could reasonably invalidate, or reasonably uphold, racial preferences under the State Constitution’s existing equal-protection guarantee. A decision invalidating racial preferences, however, would have precisely the same effect as Proposal 2, establishing that the Constitution bars racial preferences and placing the onus on proponents of racial preferences to alter the Constitution. The claimants have no answer to this point. If Proposal 2 violates the political-process doctrine, so too would a decision by the Michigan Supreme Court that comes to the same end through a permissible interpretation of the 1963 equal-protection guarantee.

The same is true of a state court’s decision to apply strict scrutiny to racial preferences. Most state constitutions, including all of the ones in our circuit, mirror this aspect of the Federal Constitution: They require strict scrutiny of governmental classifications based on race. See, e.g., D.F. v. Codell, 127 S.W.3d 571, 575 (Ky.2003); Harvey v. State, Dep’t of Mgmt. & Budget, Bureau of Retirement Servs., 469 Mich. 1 , 664 N.W.2d 767, 770 (2003); State v. Thompson, 95 Ohio St.3d 264 , 767 N.E.2d 251 , 255-56 (2002); State v. Tester, 879 S.W.2d 823, 828 (Tenn.1994). One might think these constitutional safeguards would be secure from second-guessing by federal courts, but they are not under plaintiffs’ theory. For they impose special burdens on some policies that might benefit racial minorities in the admissions process without imposing the same burdens on policies that benefit groups categorized along non-suspect lines, such as children of alumni, athletes or band members. Under each of these state constitutions, a member of a racial minority who wants a governmental privilege must identify a compelling state interest that supports its provision. If plaintiffs are correct, that reality dooms the States’ equal protection clauses, even though the federal Equal Protection *510 Clause requires no less. That cannot be right.

The Court’s decision in Romer v. Evans, 517 U.S. 620 , 116 S.Ct. 1620 , 134 L.Ed.2d 855 (1996), which did not concern racial classifications, holds nothing to the contrary. Colorado enacted a constitutional amendment prohibiting the State and its municipalities from enacting laws banning discrimination on the basis of “homosexual, lesbian or bisexual orientation, conduct, practices or relationships.” Id. at 624 , 116 S.Ct. 1620 . In invalidating the amendment, the Court noted that the amendment “impostes] a broad and undifferentiated disability” (the inability to seek protection from discrimination at the state or local level) “on a single named group” (gays and lesbians). Id. at 632 , 116 S.Ct. 1620 . The amendment “was inexplicable by anything but animus toward the class it affects” and therefore “lack[ed] a rational relationship to legitimate state interests.” Id. By contrast, Proposal 2 serves a rational interest, indeed a compelling one: eliminating racial classifications in admissions, public employment and public contracting.

The Court’s decisions in Hunter and Seattle, which did concern racial classifications, also hold nothing to the contrary. The laws invalidated in both cases were designed to disadvantage one minority group — African-Americans—and no other. In Hunter , the City of Akron in 1964 amended its charter to require that any housing ordinance forbidding discrimination “on the basis of race, color, religion, national origin or ancestry” be approved in a city-wide referendum before going into effect, an amendment designed to hinder only the political goals of the African-American community. See 393 U.S. at 391 , 89 S.Ct. 557 ; id. at 394, 395 , 89 S.Ct. 557 (Harlan, J., concurring). In Seattle, the State of Washington in 1978 enacted a law forbidding school districts from requiring children to be bused to schools distant from their homes, but made exceptions for busing designed for virtually every purpose except racial integration. 458 U.S. at 462-63, 102 S.Ct. 3187 . It thus “remove[d] the authority to address a racial problem — and only a racial problem — from [school boards].” Id. at 474, 102 S.Ct. 3187 .

The same cannot be said of Proposal 2. In the first place, Proposal 2 removes racial preferences, not anti-discrimination measures. To the extent Proposal 2 has any effect on the political structures through which a group may acquire special treatment in university admissions, it is a leveling one. The law imposes no “special burden[s]” on anyone, Hunter, 393 U.S. at 391 , 89 S.Ct. 557 , but instead eliminates “pernicious” “racial classifications,” Gratz, 539 U.S. at 270 , 123 S.Ct. 2411 . If ever there were a neutral, non-special burden, that is it. The Equal Protection Clause freely permits governments to ban racial discrimination, as here, but it does not freely permit them to ban all bans on racial discrimination, as in Hunter and Seattle.

In the second place, Proposal 2 prohibits discrimination not just on the basis of race but also on the basis of sex, ethnicity and national origin. To the extent it disadvantages anyone, it disadvantages groups that together account for a majority of Michigan’s population, not this or that racial minority. It “make[s] little sense to apply ‘political structure’ equal protection principles where the group alleged to face special political burdens itself constitutes a majority of the electorate.” Coal. for Econ. Equity, 122 F.3d at 704.

Nor is it even clear which groups — men or women, this racial group or that one— Proposal 2 helps and hurts, or when each group will be affected. Perhaps there was a time when a ban on gender-based prefer *511 enees favored men. Perhaps the opposite is true today, as female high school students increasingly outperform their male classmates. See, e.g., Rob Mank, Men Far More Likely to Benefit from Affirmative Action in College Admissions, CBS News, Sept. 26, 2011, http://tmm.cbsnews.com/ 8301-5035UM 62-201116U6-5035U- html. A ban on racial preferences likewise may favor some racial groups today and others tomorrow. The one thing we know for sure is that race-based programs need not, indeed may not, last into perpetuity, as the Court assumed that all such programs would come to an end within 25 years— namely by 2028. See Grutter, 539 U.S. at 343 , 123 S.Ct. 2325 . It is the essence of democracy to allow the people to use their own judgment in deciding when a complicated policy has outlived its usefulness, and nothing in Grutter required the States to wait until 2028 to make that choice.

It is no answer to say that Michigan may adopt a statewide policy regarding racial preferences if, and only if, they adopt statewide policies on other admissions policies — from how much weight to give Advanced Placement courses to how many zoology students to admit to how to treat children of alumni to how to treat football players, oboists or thespians. The Equal Protection Clause reflects our collective judgment that generalizations based on race are dubious in the near term and destructive in the long term, making it appropriate to treat racial proxies, which are presumptively unconstitutional, differently from other more-pedestrian distinctions, which are presumptively constitutional. Gratz, 539 U.S. at 270 , 123 S.Ct. 2411 . It does not bar Michigan from recognizing the same.

Any doubt that Hunter and Seattle support rather than undermine the constitutionality of Proposal 2 is removed by Seattle, the last of the two decisions. In Seattle, Justice Powell, no stranger to affirmative-action debates, raised the concern that the majority’s reasoning meant that, “if the admissions committee of a state law school developed an affirmative-action plan that came under fire, the Court apparently would find it unconstitutional for any higher authority to intervene unless that authority traditionally dictated admissions policies.” 458 U.S. at 499 n. 14, 102 S.Ct. 3187 (Powell, J., dissenting). No worries, the majority responded: The problem with Washington’s anti-busing initiative was “the burden it imposefd] on minority participation in the political process,” id. at 480 n. 23, 102 S.Ct. 3187 , a consideration that made Justice Powell’s hypothetical “entirely unrelated to this case” because it had “nothing to do with the ability of minorities to participate in the process of self-government.” Id. If the Court thought that the removal of an affirmative-action policy was “entirely unrelated” to the concerns in Seattle, then I am hard-pressed to understand why the same is not true in this instance — and just as hard-pressed to understand how anyone can insist our hands are tied in today’s case. The companion political-process case to Seattle, handed down the same day, confirmed the point. The “Equal Protection Clause,” it made clear, “is not violated by the mere repeal of race-related legislation or policies that were not required by the Federal Constitution in the first place.” Crawford v. Bd. of Educ. of Los Angeles, 458 U.S. 527, 538-39 , 102 S.Ct. 3211 , 73 L.Ed.2d 948 (1982). That is all that happened here. The majority seeing it differently, I respectfully dissent.