Oral arguments were recently held in the controversial case of Schuette v. Coalition to Defend Affirmative Action.[1] At issue is the constitutionality of a 2008 amendment to the Michigan Constitution that prohibits public universities from “discriminat[ing] against, or grant[ing] preferential treatment to, any individual on the basis of race, sex, color, ethnicity, or national origin.”[2] The challengers of this amendment view it as a brazen attack on affirmative action. In addition, they also view it as a clear violation of the Equal Protection Clause under the so-called Hunter Doctrine. At least in their view, this doctrine requires that strict scrutiny be applied to any law that moves a decision involving race from one level of government to another, or restructures the political process in a racial fashion—and they certainly do not think that this amendment serves a compelling interest.[3] The supporters of this amendment, including Jennifer Gratz of Gratz v. Bollinger fame, consider the idea that an amendment that prevents the allocation of preferences or benefits on the basis of race could violate the Equal Protection Clause to be completely preposterous. In their words, “[I]t does not violate equal protection to require equal treatment.”[4]

At oral argument, however, the Court did not seem to see the case as this simple. On the contrary, many of the Justices, especially Justices Sotomayor and Ginsburg, seemed troubled by the implications of Michigan’s amendment for traditionally disadvantaged minorities.[5] At the same time, several Justices seemed perplexed by the potentially wide scope of a doctrine that, in its most expansive form, subjects every law that moves a decision from one level of government to another to strict scrutiny. For instance, if Michigan’s constitutional amendment is subject to strict scrutiny, is Title VII subject to strict scrutiny? Is the Fair Housing Act? Is 42 U.S.C. § 1983? All three of these laws similarly move a decision involving race—the decision of how much governmental protection to provide against racial discrimination—from one level of government (the states) to another (the federal government).[6] These types of potential consequences led to a great deal of time being consumed by questions looking to answer that classic legal question: where do we draw the line?[7]

Many lines were suggested as the proper trigger for strict scrutiny, including when a restructuring has a disparate impact on a protected minority group, when a restructuring takes the form of a constitutional amendment, and when a restructuring is motivated by discriminatory intent. And yet, no one mentioned the most obvious, tenable line of all: when a restructuring is executed via an initiative or another form of direct democracy.

Direct Democracy and Minority Rights

Although disputed, there is a robust body of political science literature finding that direct democracy is often employed to the detriment of minority rights. Overall, this literature demonstrates that voters overwhelmingly support initiatives that restrict public accommodations laws, impede school desegregation, and generally discriminate against homosexuals and illegal immigrants.[8] On this point, it bears noting the factual context of the two cases that are traditionally associated with the Hunter Doctrine, Hunter v. Erickson[9] and Washington v. Seattle School District No. 1 (Seattle Schools).[10] The former involved a direct initiative at the local level that repealed a fair housing ordinance, and the latter involved a direct initiative that stripped Seattle School District No. 1 of the authority to implement a school busing plan designed to remedy de facto segregation.

These cases are not mere relics from another generation, as some of the supporters of Michigan’s constitutional amendment have sought to portray them. On the contrary, the Court came face to face with another potentially discriminatory direct initiative just last Term in Hollingsworth v. Perry.[11] In addition—and perhaps of more salience to the Roberts Court—the Court has recently struck down a number of initiatives restricting First Amendment rights, often insinuating in the process that the restrictions in question are targeted at disfavored speakers or operate in a discriminatory fashion, albeit in a different sense.[12]

The Textual Hook

Nevertheless, the Court is unlikely to be swayed by political theory, or even a long line of cases demonstrating that direct democracy often leads to undesirable results. Since 1912, and its decision in Pacific States Telephone & Telegraph v. Oregon,[13] the Court has been completely unwilling to countenance any argument that direct democracy, in any of its forms, violates the Republican Form of Government Clause.[14] In addition, and more relevant for this discussion, it has been unwilling to acknowledge that the fact that a particular law was passed via direct democracy should ever impact its analysis.[15] This is likely why no one bothered to mention the relevance of the initiative process during oral arguments in Schuette.

In dismissing the relevance of direct democracy, the Court, however, seems to be overlooking a crucial portion of the Constitution. Under Article VI, Section 3, all “Senator and Representatives . . . and the Members of the several State Legislatures, and all executive and judicial Officers, . . . of the Several States, shall be bound by Oath or Affirmation, to support this Constitution.”[16] Although this may seem like a meaningless formality, the Court has never treated it as such. On the contrary, the Court has consistently stated that this oath to support the Constitution is a key reason why laws passed by either Congress or state legislatures should receive a presumption of constitutionality and thus ordinarily be subject only to rational-basis review.[17] The Court, moreover, has never stated that ordinary citizens are themselves presumed to adhere to the Constitution.


The key question, of course, is what the implications of this potential textual oversight are. It seems highly implausible that the Court will hold that all laws passed via direct democracy are unconstitutional or no longer subject to a presumption of constitutionality. This would violate the spirit of Pacific States and ignore a century of practice. It would also ignore the other interests that are served by rational-basis review of state laws, most prominently federalism. Perhaps the answer, as suggested by Julian Eule, could be for the Court to take a so-socalled “hard look” at laws, such as Michigan’s constitutional amendment, that are passed via direct democracy.[18] This solution, however, engenders more questions than answers. For instance, how should we weigh the relative importance of federalism and Article VI to the presumption of constitutionality? Should we just split the difference and utilize intermediate scrutiny? Should we downplay federalism when state governments are not involved and use strict scrutiny? Should it depend on the nature of the law being examined? And in the absence of definitive answers to these questions, there are legitimate concerns that such an approach would be highly manipulable.[19]

I propose a simpler solution, hopefully less susceptible to charges of judicial activism. Instead of requiring that some level of heightened scrutiny be applied to initiatives passed via direct democracy, courts should shift the burden of persuasion to the supporters of the law and require that they affirmatively disprove discriminatory intent. Rather than simply importing scrutiny through a backdoor, this solution is particularly appropriate in the context of direct democracy. As demonstrated by the examples above, most potentially unconstitutional initiatives are facially neutral. The electorate is not going to pass a law directly preventing minorities from attending its schools, but it might coincidently pass a law preventing school districts from assigning children on a nongeographic basis just as their local school districts decide to implement busing. Accordingly, although this is certainly a contested point, many of these cases are going to come down to intent.

The issue with this eventuality is that intent is somewhat incoherent in the context of direct democracy. As distinguished from a legislature, which has at most a few hundred members, in some states, millions vote for and against initiatives and each citizen likely has her own motivations and intent in doing so. Moreover, unlike a legislative bill, there is no ready source of definitive legislative history available in the context of an initiative, other than perhaps media reports. The upshot of all of these difficulties is that proving discriminatory intent, which is already difficult in the best of circumstances, is essentially impossible in the context of direct democracy. Thus, it should come as no surprise that the district court in Schuette cursorily dismissed allegations of discriminatory intent. In fact, after noting that “[e]xamining intent in the context of a ballot initiative presents a unique problem due to the sheer number of individuals whose intent is relevant,” it determined that it was bound by Sixth Circuit precedent holding that “a district court cannot inquire into the electorates motivations in an equal protection clause context” unless the “only possible rationale” for a facially neutral measure is that it is “racially motivated.”[20]

In many ways, this seems backwards. Not only are minorities stripped of the usual protections of the Article VI oath—which, even if some find meaningless, the Court does not—they are also forced to surmount a practically (and in some instances legally) higher burden to prove discriminatory intent. Moreover, this is all because someone else has chosen an arguably constitutionally disfavored method of lawmaking. It is much more equitable to force the supporters of the amendment to shoulder this burden because they have chosen a method of legislation that is susceptible to such issues in proving intent, and thus retain the right to choose a different method. In addition, they will likely find it much easier to disprove discriminatory intent than the challengers will find proving it. They, after all, as the primary drafters and supporters of the measure, will have the most insight into the true intent and purpose of the measure and will also be aware of the most relevant evidence bearing on this inquiry.


The Michigan electorate should be required to pay now or pay later to prove that the laws of its state do not violate the Constitution. If it chooses to pursue a state constitutional amendment through direct initiative, it will trade the scrutiny of the legislative process, mediated by those who have taken an Article VI oath to uphold the Federal Constitution, for the burden of affirmatively proving that its measure is constitutional.[21] If it chooses instead to lobby its state legislature to propose a state constitutional amendment, it will face no such burden later in the process. Only by requiring this choice can we really be sure that “equal treatment” is actually the same as “equal protection” in this case.