Plaintiffs-appellants would like to amend the Massachusetts Constitution to allow public financial support to be directed toward private, religiously affiliated schools. Plaintiffs attempted to propose their amendment through the Massachusetts initiative procedure, but two distinct provisions of the Massachusetts Constitution prevented initiatives on this subject. They now challenge these subject-matter exclusions from the initiative process on federal Free Speech, Free Exercise, and Equal Protection grounds. In the end, plaintiffs' arguments fail, and although our analysis diverges at points, we affirm the district court's grant of summary judgment.1

Plaintiffs are parents of children enrolled in religiously affiliated schools who sought to amend Amendment Article 18 of the Massachusetts Constitution (the "Anti-Aid Amendment"), which prohibits public financial support for private primary or secondary schools.2 Mass. Const. amend. art. 18. Article 48 of the Massachusetts Constitution provides that, in addition to the amendment procedure available to the state legislature, the Constitution may also be amended by popular initiative. Mass. Const. amend. art. 48, pt. 1. Following the required procedure, plaintiffs submitted an initiative petition, for certification, to the Massachusetts Attorney General to modify the Anti-Aid Amendment by adding a sentence stating that nothing in the Anti-Aid Amendment shall prevent the Commonwealth from providing loans, grants, or tax benefits to students attending private schools, regardless of the schools' religious affiliation. The Attorney General, however, denied certification of the proposed initiative, because Article 48 prohibits amendment of the Anti-Aid Amendment by initiative (the "Anti-Aid Exclusion") and because the petition explicitly relates to "religious institutions," another matter expressly excluded from the initiative process by Article 48 (the "Religious Exclusion").

Section Two of Article 48 limits Massachusetts' initiative process by listing the "Excluded Matters," which are not subject to popular action by initiative, including, inter alia, appointment or compensation of judges; the powers, creation or abolition of the courts; and specific appropriation of state money. Mass. Const. amend. art. 48, pt. 2, § 2. The pertinent provision of Article 48, referred to as the Anti-Aid Exclusion, states that "[n]either the eighteenth [Anti-Aid] amendment of the constitution... nor this provision for its protection, shall be the subject of an initiative amendment," while the Religious Exclusion mandates that "[n]o measure that relates to religion, religious practices or religious institutions... shall be proposed by an initiative petition." Id. Plaintiffs challenge the validity of both of these exclusions under the U.S. Constitution.

The first issue before us is whether the Massachusetts Constitution's limitations on the initiative process violate the First Amendment free speech rights of prospective initiative proponents. Appellants argue that the exclusions to the state initiative process, which prevent them from pursuing amendments regarding religion or state aid to private institutions, should be considered content-based restrictions on core political speech subject to strict scrutiny.

The difficulty with the appellants' argument is that a state initiative procedure, although it may involve speech, is also a procedure for generating law, and is thus a process that the state has an interest in regulating, apart from any regulation of the speech involved in the initiative process. In other words, the challenged exclusions constitute regulations "aimed at non-communicative impact, but nonetheless having adverse effects on communicative opportunity." Laurence H. Tribe, American Constitutional Law § 12-2 at 790 (2d ed.1988). See, e.g., United States v. O'Brien, 391 U.S. 367, 382, 88 S. Ct. 1673, 20 L. Ed. 2d 672 (1968) (rejecting draft card burner's claim that a statute prohibiting the destruction of draft cards violated his First Amendment rights, reasoning that the law punished him for the "noncommunicative impact of his conduct," although the court recognized the symbolic value of burning a draft card). Unlike regulations that are "aimed at communicative impact," regulations that aim at preventing some harm independent of speech—in this case, the use of the initiative process for the passage of certain types of laws believed to be unsuited to that process—are not presumed unconstitutional, and are not subjected to strict scrutiny. Tribe, American Constitutional Law § 12-2, at 790. See, e.g., City of Erie v. Pap's A.M., 529 U.S. 277, 291, 120 S. Ct. 1382, 146 L. Ed. 2d 265 (2000) (upholding a ban on nude dancing, because "the ordinance does not attempt to regulate the primary effects of the expression, i.e., the effect on the audience of watching nude erotic dancing, but rather the secondary effects, such as the impacts on public health, safety, and welfare," which are unrelated to expression). Regulations of this type are, at most, subject to intermediate scrutiny, under which they will be upheld if the "harmful consequences of this particular form of expressive behavior, quite apart from any ideas it might convey, outweigh the good." Tribe, American Constitutional Law § 12-2, 791. See, e.g., Grayned v. Rockford, 408 U.S. 104, 115-16, 92 S. Ct. 2294, 33 L. Ed. 2d 222 (1972) (upholding ordinance barring noisy demonstrations near schools, because the government has sufficiently "weighty reasons" to restrict this type of expressive activity). Applying this balancing, we uphold Massachusetts' exclusions to its initiative process for the reasons explained below.

Before arriving at this explanation, we will first examine the arguments of the parties—a task that is particularly difficult in this case, because the parties have planted themselves firmly at opposite poles, with plaintiffs arguing for strict scrutiny and Massachusetts arguing that only minimal rationality review is appropriate. In the end, we find that the law requires our analysis to proceed by a middle path in this apparent battle of absolutes. We hold that Massachusetts' exclusions to its initiative process are narrowly drawn to further a significant state interest, and thus survive intermediate scrutiny.

1. The Communicative Value of the Initiative Process

The first step in our free speech analysis must be to determine whether citizens' use of the initiative process constitutes expressive conduct, permitting appellants to invoke the First Amendment to challenge the Massachusetts initiative exclusions. See, e.g., Texas v. Johnson, 491 U.S. 397, 403, 109 S. Ct. 2533, 105 L. Ed. 2d 342 (1989) (citing Spence v. Washington, 418 U.S. 405, 409-11, 94 S. Ct. 2727, 41 L. Ed. 2d 842 (1974)). We do not find that there is any serious debate as to this point. A state initiative process provides a uniquely provocative and effective method of spurring public debate on an issue of importance to the proponents of the proposed initiative. The Supreme Court has made clear that the process involved in proposing legislation by means of initiative involves core political speech. See Meyer v. Grant, 486 U.S. 414, 108 S. Ct. 1886, 100 L. Ed. 2d 425 (1988) (overturning state's prohibition on using paid petition circulators); Buckley v. Am. Constitutional Law Found., Inc., 525 U.S. 182, 119 S. Ct. 636, 142 L. Ed. 2d 599 (1999) (overturning various registration requirements for petition circulators). In Meyer, the Supreme Court recognized that "the solicitation of signatures for a petition involves protected speech." 486 U.S. at 422, n. 5, 108 S. Ct. 1886. Furthermore, the mere fact that plaintiffs "remain free to employ other means to disseminate their ideas does not take their [preferred means of] speech through [the initiative process] outside the bounds of First Amendment protection." Id. at 424, 108 S. Ct. 1886. Clearly, plaintiffs have been prevented from engaging in the sort of activity that implicates the First Amendment. This conclusion, however, in no way ends our analysis; it only opens the door for us to apply constitutional freedom of speech principles to the limitations Massachusetts places on its initiative process.

We have recognized that "a fine line separates permissible regulation of state election processes from impermissible abridgement of First Amendment rights," Pérez-Guzmán v. Gracia, 346 F.3d 229, 239 (1st Cir.2003), and the same is true of regulation of state initiative procedures. In Pérez-Guzmán, 346 F.3d at 239-47, we invalidated Puerto Rico's requirement that petition signatures needed for registering a new political party to appear on the general election ballot be notarized, holding that it violated the First Amendment of the federal Constitution. In so doing, we stated that "we afford exacting scrutiny to severe restrictions on ballot access." Id. at 239. We began our analysis "with an assessment of the severity of the restriction," id., and having found it to be severe, we applied strict scrutiny, id. at 243-44.

Plaintiffs argue that we should apply a similar two-step analysis here. However, plaintiffs' suggested analysis makes an end-run around the most difficult part of their case. The district court in this case found that speech was only incidentally affected by the Massachusetts subject matter exclusions. Boyette v. Galvin, 311 F. Supp. 2d 237, 240 (D.Mass.2004). Although the district court recognized that speech was involved, it concluded that the primary goal of the exclusion was to prevent certain types of laws from being passed by means of the popular initiative process, and not to limit what people could say or how they could say it. Id. at 240-41. By contrast, the common denominator in Pérez-Guzmán and other cases cited by plaintiffs is a direct restriction on the communicative aspect of the political process. In Pérez-Guzmán, like in Meyer, the state regulated how people could promulgate their political views, in their respective attempts to put a new party on the ballot, Pérez-Guzmán, 346 F.3d at 230-31, and to circulate petitions for a proposed initiative, Meyer, 486 U.S. at 414, 108 S. Ct. 1886. Strict scrutiny applied in these cases precisely because they involved direct regulation of the petition process itself.

We believe that the present case calls for a lower level of scrutiny. We know of no general principle that, in addition to constitutional amendment or lawmaking via a process instituted by the state legislature, a state must provide an opportunity for its residents to propose constitutional amendments or laws on all subjects by means of an initiative process. While we accept that use of the initiative process can facilitate dissemination of initiative proponents' views, the next step in a free speech analysis is to determine whether or not the regulation in question aims at regulating speech, or whether it has some other primary end, such that any effect on speech is purely incidental. As we alluded to at the outset of this analysis, the First Amendment generally provides greater protection against laws that are "aimed at communicative impact" of the conduct they regulate than from laws "aimed at non-communicative impact, but nonetheless having adverse effects on communicative opportunity." Tribe, American Constitutional Law § 12-2, at 790. The primary goal of state initiative procedures is to create an avenue of direct democracy whereby citizens can participate in the generation of legislation—that is, the act of creating law. Laws such as those considered in Meyer and its progeny were aimed at directly regulating the means that initiative proponents could use to reach their audience of potential petition signers. In contrast, we find that subject matter exclusions like those regulating the Massachusetts initiative process aim at preventing the act of generating laws and constitutional amendments about certain subjects by initiative. While they eliminate a valuable avenue of expression about those subjects, the speech restriction is no more than an unintended side-effect of the exclusions. It is because of this sometimes overlooked, but nevertheless fundamental principle in constitutional free speech doctrine that we must reject appellants' proposed analysis. We turn now to Massachusetts' proposed alternative analysis.

2. Massachusetts' Need to Regulate the Lawmaking Act

The communicative power of an initiative stems precisely from the fact that it is not just speech; it is a process that can lead to the creation of new laws or constitutional amendments. Massachusetts urges us to hold that its restrictions on the amendment process do not regulate speech qua speech, and thus do not trigger strict scrutiny under the First Amendment.

Government actions that are aimed at some goal other than restricting the conveyance of ideas are generally permissible, even if they incidentally inhibit free speech. See, e.g., Arcara v. Cloud Books, Inc., 478 U.S. 697, 106 S. Ct. 3172, 92 L. Ed. 2d 568 (1986) (upholding the closure of an adult bookstore because prostitution was taking place on the premises). Arcara is a prototypical example of this type of case, because the law in question only regulated a specific type of conduct—prostitution—which did not implicate speech. However, even a law seemingly entirely removed from speech can have effects on speech. In Arcara, for example, the prohibition on prostitution resulted in the closure of a book store. Enforcement of a prohibition with such incidental effects does not, however, implicate the First Amendment, and this type of law need only survive rationality review.

Plaintiffs do not cite to any precedent for the proposition that, under the Free Speech Clause of the First Amendment, a state may not restrict the subjects that can be addressed through its initiative process. The D.C. Circuit addressed a similar free speech challenge to a restriction on an initiative process in the case of Marijuana Policy Project v. United States, 304 F.3d 82 (D.C.Cir.2002). In that case, the D.C. Circuit held that a statute precluding the use of the D.C. ballot initiative process to lower drug penalties did not unconstitutionally restrict free speech rights of medical marijuana advocates, but only shifted the forum of debate from the District of Columbia to Congress. Id. at 85-86. The court explained that "although the First Amendment protects public debate about legislation, it confers no right to legislate on a particular subject." Id. at 85. Massachusetts argues that we should adopt this reasoning to apply rational basis review in the instant case.

3. Regulating Conduct with Speech and Nonspeech Elements

We cannot agree with the D.C. Circuit's finding that subject-matter exclusions from the initiative process "restrict[ ] no speech," id. at 85, nor with its conclusion that this type of selective carve-out "implicates no First Amendment concerns," id. at 83. For the same reasons, we also reject Massachusetts' argument that we should apply only rational basis review to the Anti-Aid and Religious Exclusions. This case is not like Arcara, where the Supreme Court criticized the New York Court of Appeals for having applied the analysis established in United States v. O'Brien, 391 U.S. 367, 88 S. Ct. 1673, 20 L. Ed. 2d 672 (1968), because the Court did not consider the regulated conduct—prostitution—expressive. See Arcara, 478 U.S. at 705, 106 S. Ct. 3172 (concluding that "unlike the symbolic draft card burning in O'Brien, the sexual activity carried on in this case manifests absolutely no element of protected expression"). Rather, we would continue along the analysis laid out by the Supreme Court in Texas v. Johnson, 491 U.S. at 403, 109 S. Ct. 2533. Having determined that Meyer, 486 U.S. at 422, 108 S. Ct. 1886, indicates that a state initiative process manifests elements of protected expression, under Johnson, "we next decide whether the State's regulation is related to the suppression of free expression." 491 U.S. at 403, 109 S. Ct. 2533. Here, as we have already explained, the Massachusetts exclusions in question regulate which types of laws or amendments can be passed by initiative, without any reference to who may speak or what message they may convey. Thus, because "the State's regulation is not related to expression,... the less stringent standard [the Supreme Court] announced in United States v. O'Brien for regulations of non-communicative conduct controls." Id. (citing O'Brien, 391 U.S. at 377, 88 S. Ct. 1673 (finding that, although burning a draft card can be expressive conduct, the federal government's interest in preventing the destruction of draft cards is sufficient to uphold defendant's conviction)).

The standard enunciated in O'Brien governs "when `speech' and `nonspeech' elements are combined in the same course of conduct." O'Brien, 391 U.S. at 376, 88 S. Ct. 1673. See also Clark v. Community for Creative Non-Violence, 468 U.S. 288, 104 S. Ct. 3065, 82 L. Ed. 2d 221 (1984) (applying O'Brien scrutiny to the application of a ban on camping on the Mall in Washington, D.C., to demonstrators who sought to sleep overnight there to protest the plight of homeless people). While we agree with the D.C. Circuit that this type of regulation of a state initiative process is not aimed at regulating speech, we cannot see how, given the Supreme Court's analysis in Meyer, subject-matter exclusions from a state initiative process "restrict[ ] no speech." Marijuana Policy Project, 304 F.3d at 85. To the contrary, since expression is affected by the regulations of the state initiative process, we apply the intermediate scrutiny standard set out in O'Brien.

Under the O'Brien standard, conduct combining "speech" and "non-speech" elements can be regulated if four requirements are met: (1) the regulation "is within the constitutional power of the Government;" (2) "it furthers an important or substantial governmental interest;" (3) "the governmental interest is unrelated to the suppression of free expression;" and (4) "the incidental restriction on alleged First Amendment freedoms is no greater than is essential to the furtherance of that interest." O'Brien, 391 U.S. at 377, 88 S. Ct. 1673. We have no difficulty finding that the Massachusetts exclusions meet the second requirement, as Massachusetts certainly has a substantial interest in maintaining the proper balance between promoting free exercise and preventing state establishment of religion. Neither do we doubt that Massachusetts has a substantial interest in restricting the means by which these fundamental rights can be changed. We have already stated that the exclusions aim at preventing certain uses of the initiative process, not at stemming expression, and thus meet the third O'Brien requirement.

As for the first requirement, that the regulation be within the constitutional power of the government, we find that the only serious, non-speech-related constitutional challenges to Massachusetts' power to regulate the subjects that may be reached by its initiative process are the Free Exercise and Equal Protection arguments, which we reject in this opinion. Having now concluded that Massachusetts' interest in protecting the fundamental free exercise and freedom from state-established religion is substantial and its method otherwise constitutionally permissible, we finally consider the fourth O'Brien requirement: whether the incidental restrictions on would-be initiative proponents' First Amendment freedoms are greater than essential to the furtherance of that interest. Since we see no other way in which Massachusetts could achieve its interest in safeguarding these fundamental freedoms in its Constitution from popular initiative, we recognize that the restriction on speech is no more than is essential. Thus, we conclude that Massachusetts' Anti-Aid and Religious Exclusions do not violate the First Amendment free speech guarantee.

We now consider whether the Religious Exclusion violates the Free Exercise Clause of the First Amendment.3 The Free Exercise Clause guarantees that "Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof," U.S. Const. amend. I (emphasis added), and it has been applied to the States through the Fourteenth Amendment. See Cantwell v. Connecticut, 310 U.S. 296, 303, 60 S. Ct. 900, 84 L. Ed. 1213 (1940).

The protections provided by the Free Exercise Clause may be broken down into a number of conceptual categories, none of which are implicated by the Religious Exclusion. First and foremost, the Free Exercise Clause entails an absolute prohibition on government infringement on the "freedom to believe." Torcaso v. Watkins, 367 U.S. 488, 492-93, 81 S. Ct. 1680, 6 L. Ed. 2d 982 (1961). Because the prohibition is absolute, laws which infringe on individuals' freedom of belief are per se unconstitutional. Id. (refusing to consider the state's asserted justifications for a law that infringed on citizens' freedom of belief); see also West Virginia State Bd. of Educ. v. Barnette, 319 U.S. 624, 63 S. Ct. 1178, 87 L. Ed. 1628 (1943) (holding that a state could not compel students in public schools to salute the flag, where their religion forbade it). However, that prohibition is inapposite here because the Religious Exclusion does not hinge on the religious beliefs of initiative proponents. See McDaniel v. Paty, 435 U.S. 618, 627, 98 S. Ct. 1322, 55 L. Ed. 2d 593 (1978) (finding that a Tennessee law that precluded ministers from eligibility as constitutional convention delegates did not infringe on freedom of belief, but nevertheless invalidating it under strict scrutiny).

In McDaniel v. Paty, the Supreme Court examined a state law preventing a minister from serving as a constitutional convention delegate under strict scrutiny, because although it did not directly burden his religious beliefs, it directly burdened his religious "status, acts, and conduct." McDaniel, 435 U.S. at 626-27, 98 S. Ct. 1322. This protection is related to, though distinct from, the per se prohibition on laws infringing on belief. Arguing from McDaniel, plaintiffs maintain that a state violates the Free Exercise Clause when it creates a general political process, but excludes some from access to that process on the basis of religion. However, plaintiffs fail to explain how this proposition applies to their case. The Religious Exclusion prevents anyone from proposing new laws or constitutional amendments relating to religion through the initiative process. It does not exclude religious people, or people of a certain religion, from proposing laws or amendments. In other words, a religious individual of any particular faith, like any other citizen, can propose a new law or amendment on any subject that is open to amendment by initiative.

Moreover, the Religious Exclusion we are asked to scrutinize does not distinguish based on religious status.4 Like the scholarship program at issue in Locke v. Davey, the Religious Exclusion does not deny plaintiffs "the right to participate in the political affairs of the community," 540 U.S. at 720, 124 S. Ct. 1307 (distinguishing Locke from McDaniel, 435 U.S. 618, 98 S. Ct. 1322, 55 L. Ed. 2d 593); nor does it "require[ ] [plaintiffs] to choose between their religious beliefs and receiving a government benefit." Id. at 720-21, 98 S. Ct. 1322 (distinguishing Hobbie v. Unemployment Appeals Comm'n of Fla., 480 U.S. 136, 107 S. Ct. 1046, 94 L. Ed. 2d 190 (1987); Thomas v. Review Bd. of Ind. Employment Security Div., 450 U.S. 707, 101 S. Ct. 1425, 67 L. Ed. 2d 624 (1981); Sherbert v. Verner, 374 U.S. 398, 83 S. Ct. 1790, 10 L. Ed. 2d 965 (1963)). The exclusion applies equally to measures proposed by any group or individual, regardless of their religious affiliation or lack thereof.

Having concluded that the Religious Exclusion does not discriminate on the basis of religious belief or status, we also briefly note that plaintiffs make no colorable argument that the exclusion prohibits any religious act or conduct. It does not, for example, preclude performing rites required by their religion, see Church of the Lukumi Babalu Aye v. City of Hialeah, 508 U.S. 520, 533, 113 S. Ct. 2217, 124 L. Ed. 2d 472 (1993) (outlawing animal sacrifice central to the Santería religion), or single out their religion's method of worship, see Fowler v. Rhode Island, 345 U.S. 67, 73 S. Ct. 526, 97 L. Ed. 828 (1953) (preventing Jehovah's Witnesses from meeting in public parks while other denominations were allowed to hold services). The Supreme Court has stated its reluctance to strike down "legislation which imposes only an indirect burden on the exercise of religion, i.e., legislation which does not make unlawful the religious practice itself." Braunfeld v. Brown, 366 U.S. 599, 606, 81 S. Ct. 1144, 6 L. Ed. 2d 563 (1961). Certainly, the amendments that plaintiffs want to propose may be motivated by their religious beliefs, but they do not claim that working to pass those amendments is an aspect of practicing their religion.

Finally, plaintiffs ask us to consider whether the passage of the Religious Exclusion was motivated by animus toward religion. The Supreme Court has considered the existence of animus motivating a law's proponents when determining whether the law violates the Free Exercise Clause of the First Amendment. See, e.g., Locke, 540 U.S. at 725, 124 S. Ct. 1307 (finding nothing in the history, text or application of the Washington scholarship program in question "that suggests animus towards religion"). In the Establishment Clause context, the Supreme Court has "often stated the principle that the First Amendment forbids an official purpose to disapprove of a particular religion or of religion in general." Church of the Lukumi, 508 U.S. at 532, 113 S. Ct. 2217 (noting that "[t]hese cases, however, for the most part have addressed governmental efforts to benefit religion or particular religions, and so have dealt with a question different, at least in its formulation and emphasis, from the issue here"). Although plaintiffs present significant evidence of animus against Catholics in Massachusetts in 1855 when the Anti-Aid Amendment was passed, they fail to show that religious animus motivated the passage of the Religious Exclusion in 1918. Plaintiffs rely heavily on one statement made by the Religious Exclusion's sponsor, indicating that he would "protect the initiative and referendum against the religious fanatics and against the professional religionists." However, we need not reach the question of whether this type of fervor against religiously motivated political action would require that the amendment be struck down, because plaintiffs present no evidence that the other members of the Constitutional Convention of 1917-1918 acted from similar motivations. Given the wide margin by which the Religious Exclusion passed, and the significant Catholic representation at the Convention, we see no evidence that animus against religion was a motivating factor behind the Exclusion's passage.

Furthermore, plaintiffs cite to no case in which evidence of animus toward religion was itself sufficient to invalidate a government action, without the animus being tied to some resulting infringement on freedom of belief or on religious status, acts or conduct. While we must apply strict scrutiny when "the object of a law is to infringe upon or restrict practices because of their religious motivation," plaintiffs here have not shown that the Religious Exclusion results in any restriction of their religious practices. Id. at 533, 113 S. Ct. 2217 (citing Employment Div., Dept. of Human Resources of Ore. v. Smith, 494 U.S. 872, 878-79, 110 S. Ct. 1595, 108 L. Ed. 2d 876 (1990)) (emphasis added) (considering evidence of animus toward the Santería religion where the ordinance prohibiting ritual slaughter of animals did not, on its face, target the Santería religion, but effectively outlawed one of the religion's "principal forms of devotion").

Plaintiffs also argue that the Massachusetts Exclusions violate the protections afforded by the Equal Protection Clause. U.S. Const. amend. XIV. Plaintiffs argue that the Religious Exclusion violates equal protection guarantees because it infringes on the fundamental right to religious free exercise, disadvantages a suspect class, and fails the more searching rational basis review required in Romer v. Evans, 517 U.S. 620, 116 S. Ct. 1620, 134 L. Ed. 2d 855 (1996). They also challenge the Anti-Aid Exclusion under a suspect classification theory and a disparate impact theory. Because of the complexity of the various equal protection arguments presented, we will address the arguments one at a time, to the extent that is possible. In the end, we affirm the district court's grant of summary judgment against plaintiffs' equal protection claims.

1. Alleged Violation of the Equal Protection Fundamental Right to Free Exercise of Religion

Before moving to what we see as the substance of plaintiffs' equal protection claims, we first address their argument that the Massachusetts Exclusions restricts their fundamental right to free exercise of religion. Where a plaintiff's First Amendment Free Exercise claim has failed, the Supreme Court has applied only rational basis scrutiny in its subsequent review of an equal protection fundamental right to religious free exercise claim based on the same facts. Locke, 540 U.S. at 721, n. 3, 124 S. Ct. 1307 (citing Johnson v. Robison, 415 U.S. 361, 375, n. 14, 94 S. Ct. 1160, 39 L. Ed. 2d 389, (1974)).5 Because we held, above, that the Religious Exclusion does not violate the Free Exercise Clause, we apply rational basis scrutiny to the fundamental rights based claim that this exclusion violates equal protection. For the reasons stated throughout this opinion, we find that the Massachusetts Exclusions pass such review.

2. Plaintiffs' Claim that the Religious and Anti-Aid Exclusions Implicate a Suspect Classification

The central equal protection issue presented is whether the Massachusetts Religious Exclusion and Anti-Aid Exclusion impermissibly distort the political process to the disadvantage of religious individuals. Because we find that the Religious Exclusion does not draw distinctions based on a suspect classification, we hold that it does not violate the Equal Protection Clause.

Plaintiffs claim that the Religious Exclusion and Anti-Aid Exclusion draw distinctions on the basis of religion, which they argue is a suspect classification for purposes of equal protection analysis.6 However, even assuming that religious classification should be treated as suspect, we do not see how the Religious Exclusion and Anti-Aid Exclusion draw distinctions among Massachusetts citizens based on a suspect classification. The Religious Exclusion prohibits initiative petitions that concern "religion, religious practices or religious institutions." The Anti-Aid Exclusion precludes amendment by initiative of the Anti-Aid Amendment, which, in addition to containing Massachusetts' free exercise clause, prevents state funding for private institutions. On their face, the Exclusions simply carve out particular subject matters from the initiative process. They do not require different treatment of any class of people because of their religious beliefs. They do not give preferential treatment to any particular religion. See Larson v. Valente, 456 U.S. 228, 246, 102 S. Ct. 1673, 72 L. Ed. 2d 33 (1982) (holding, on First Amendment grounds, that "a state law granting a denominational preference" must be treated as suspect, and subjected to strict scrutiny). In short, this is not the classic violation of equal protection in which a law creates different rules for distinct groups of individuals based on a suspect classification. See, e.g., Strauder v. West Virginia, 100 U.S. 303, 25 L. Ed. 664 (1879) (overturning state law which limited jury service to white men only).

The Supreme Court has, nevertheless, sometimes struck down facially neutral laws, which it recognized were crafted to avoid facial discrimination. See, e.g., Hunter v. Erickson, 393 U.S. 385, 387-91, 89 S. Ct. 557, 21 L. Ed. 2d 616 (1969) (invalidating an amendment to the Akron, Ohio city charter, which, in addition to the usual vote by the city council, required approval by a majority of the city's voters for any ordinance regulating real estate transactions "on the basis of race, color, religion, national origin or ancestry," even though the law "on its face treats [African-American] and white, Jew and gentile in an identical manner"); Washington v. Seattle School Dist. No. 1, 458 U.S. 457, 102 S. Ct. 3187, 73 L. Ed. 2d 896 (1982) (striking down a Washington initiative that prohibited school boards from requiring children to be bused to more distant public schools, while at the same time making exceptions for every imaginable impetus for busing other than racial integration). Unlike the Massachusetts Exclusions, the law in Hunter evinces a clear, solely detrimental effect on a suspect class. In Hunter, the law's one obvious result was to make it more difficult to pass laws prohibiting racial discrimination. Hunter, 393 U.S. at 391, 89 S. Ct. 557. Similarly, in Washington, although the initiative did not say so on its face, children could be bused to a more distant school for virtually any reason except racial integration. 458 U.S. at 471, 102 S. Ct. 3187. In contrast, the Massachusetts Religious Exclusion prevents both initiatives that would disfavor as well as those that might benefit religion, and the Anti-Aid Exclusion, in addition to preventing amendment of the clause precluding funding to private institutions, also prohibits amendment of Massachusetts' free exercise clause. Unlike the laws invalidated in Hunter and Washington, it is undeniable that the Massachusetts Exclusions both hinder and help the causes of the alleged suspect class.7

Certainly any form of invidious discrimination because of religion is forbidden. But "the Establishment Clause and the Free Exercise Clause[ ] are frequently in tension." Locke, 540 U.S. at 718, 124 S. Ct. 1307. States may properly refuse to enact laws that they reasonably believe may tend to establish religion. While the free exercise of religion is guaranteed, state support of religion is, in general, disfavored. And we know of no constitutional principle that prevents a state from determining that sensitive measures that relate to religion, religious practices, or religious institutions should not be made or initiated by the public initiative process but rather only via the legislature. A state might fear that such measures, if presented as public referenda, might be more likely to fuel religious strife or to result in enactments unfair to religious (or non-religious) minorities. Bearing these distinctions in mind, we are disinclined to extend the race-based Hunter and Washington lines of cases to this context, which so closely mirrors the mandates found in our federal Establishment Clause and Free Exercise Clause.

3. Plaintiffs Have Not Shown Discriminatory Intent

Plaintiffs further argue that the Anti-Aid Exclusion violates the Equal Protection Clause on a disparate impact theory. However, this argument fails because plaintiffs have not shown a discriminatory purpose behind the exclusion. "[A] law, neutral on its face and serving ends otherwise within the power of government to pursue," is not invalid under the Equal Protection Clause simply because it may disproportionately affect a suspect class. Washington v. Davis, 426 U.S. 229, 242, 96 S. Ct. 2040, 48 L. Ed. 2d 597 (1976). "An unwavering line of cases from [the Supreme Court] holds that a violation of the Equal Protection Clause requires state action motivated by discriminatory intent; the disproportionate effects of state action are not sufficient to establish such a violation." Hernandez v. New York, 500 U.S. 352, 372-73, 111 S. Ct. 1859, 114 L. Ed. 2d 395 (1991). Plaintiffs present evidence that widespread anti-Catholic prejudice was a motivating factor behind passage of the original Anti-Aid Amendment in 1855, a fact which defendants do not dispute. However, the Anti-Aid Amendment was largely overhauled in 1917, with the support of 85 of the 94 Catholic delegates to the Constitutional Convention, and soon afterwards, the Anti-Aid Exclusion was passed with similarly broad support. No evidence has been offered that the exclusion was motivated by the same Anti-Catholic animus that impelled the passage of the original Anti-Aid Amendment. Plaintiffs cannot mix and match the intent behind one amendment and place it with the impact of a later, distinct amendment. Thus, without reaching the question of whether the Anti-Aid Exclusion has a disparate effect on religious individuals, we reject plaintiffs' equal protection disparate impact argument because plaintiffs fail to show the required discriminatory intent.

4. Rational Basis Review Under Romer v. Evans

Having rejected plaintiffs' arguments that the Massachusetts Exclusions should be subjected to strict scrutiny, we conclude by considering whether the exclusions survive rational basis review. "[I]f a law neither burdens a fundamental right nor targets a suspect class, we will uphold the legislative classification so long as it bears a rational relation to some legitimate end." Romer v. Evans, 517 U.S. 620, 631, 116 S. Ct. 1620, 134 L. Ed. 2d 855 (1996) (striking down an amendment to the Colorado Constitution that would have precluded protection from discrimination on the basis of sexual orientation as failing rational basis review). In this case, we have no difficulty finding that Massachusetts' goal of preventing the establishment of religion is a legitimate one. Additionally, the chosen means clearly bear a rational relation to that end. Thus, the instant case is easily distinguished from Romer, where the Court found that passage of the Colorado amendment could only be explained by "`a bare ... desire to harm a politically unpopular group,'" which "`cannot constitute a legitimate governmental interest.'" Id. at 634, 116 S. Ct. 1620 (quoting Dep't of Agric. v. Moreno, 413 U.S. 528, 534, 93 S. Ct. 2821, 37 L. Ed. 2d 782 (1973)).

For the foregoing reasons, the decision of the district court is affirmed.

Affirmed.