MCCONNELL'S “FIRST AMENDMENT APPROACH”: NONESTABLISHMENT AND PRIVATIZATION

Both religion and sexuality, suggests McConnell, involve choice but also go beyond it, encompass both opinion and conduct, possess both public and private aspects, and, in short, “are central aspects of personal identity” (McConnell Reference McConnell, Olyan and Nussbaum1998, 234). If the pluralist solution to religious difference has been to avoid a public position on the merits of contending religious views, might the solution to deep divisions about the morality of same-sex relationships be refusal to take a public position on this topic as well? A “First Amendment” position would treat both conflicting views “as conscientious positions, worthy of respect, much as we treat both atheism and faith as worthy of respect” (McConnell Reference McConnell, Olyan and Nussbaum1998, 235). Sketching the jurisprudence of the religion clauses, McConnell distinguishes between the privatization approach, often associated with separationism as it consigns religion-sensitive issues to the private sphere, and the equal access approach, often associated with neutrality between religion and nonreligion because it allows “competing groups to participate in the public sphere on equal terms.” In the context of sexuality, privatization “would insist that all activities directly related to the formation of opinion about homosexuality be confined to private institutions, where there should be no interference with either beliefs (orientation) or conduct. Equal access would call for equal treatment in the public domain for all private views, but would “be careful not to convey the impression that the government is expressing a view” (McConnell Reference McConnell, Olyan and Nussbaum1998, 237). Applying these jurisprudential categories to sexual orientation and practice, McConnell's test is whether governmental stances in various areas convey moral approval or disapproval, on the one hand, or whether they merely allow those endorsing or disapproving various types of sexual expression to participate on equal terms in the public sphere like those of diverse religious beliefs, on the other.

Although McConnell (Reference McConnell, Olyan and Nussbaum1998, 239–255) applies this formulation interestingly to a wide variety of issue areas, my focus here is the institution of civil marriage. To McConnell, the case for same-sex marriage as a free exercise or equal access claim is weak, because “most combinations of human beings are ineligible for matrimony.” In Reynolds v. United States (98 U.S. 145, 1870), Reynolds was seeking neither benefits nor the recognition of polygamous marriage, but only to be left alone. “In other words, Reynolds unsuccessfully sought what homosexuals already have: the right to live with the person(s) of their choice, as if married, without hindrance from the state … It is one thing to say that the government may not interfere with a religious (or sexual) practice in the privacy of the home, and quite a different thing to say that the government must adjust the definition of a public institution to conform to the doctrines or desires of a minority” (McConnell Reference McConnell, Olyan and Nussbaum1998, 249).

As a disestablishment claim, on the other hand, the recognition of same-sex marriage would not solve the “establishment” problem for McConnell, but only broaden the “establishment” to give favored status to two “churches” (McConnell Reference McConnell, Olyan and Nussbaum1998, 250). By two “churches,” McConnell is not referring to civil marriage for same-sex couples augmenting religious marriage for opposite-sex couples; some religions will unite same-sex couples, and opposite-sex couples may of course choose civil marriage without religious accouterments. He is implying, rather, that the civil recognition betokened by marriage for same-sex couples would augment the civil recognition that in most jurisdictions is currently awarded only to opposite-sex couples. Because the civil institution of marriage accords respect and benefits to those pursuing a specific sort of choice and conception of the good, true disestablishment, McConnell argues, would require eliminating marriage as a public institution. Unions would be privately formed and celebrated, just as like-minded individuals associate to form religious groups and institutionalize them. In a persuasive account, Gordon Babst argues that the ban on same-sex marriage can be attributed to a de facto shadow establishment that has enshrined a sectarian definition of marriage. Similarly to McConnell, he suggests that apart from simply recognizing same-sex marriages, another alternative “would be to let individual couples decide for themselves within their communities of faith, or otherwise, what marriage signifies for them and their communities, rather than have a definition imposed on them by the State” (Babst Reference Babst2002, 83; see also 84). This solution would not preclude the expression of a secular public interest through law on divorce, adoption, inheritance, and other worldly interests.

McConnell is himself suggesting that there is no neutral position regarding same-sex marriage. “Limitation of marriage to heterosexual unions necessarily implies that homosexual unions lack the qualities for which marriage is legally recognized and favored, while extending marriage to homosexual unions would necessarily imply that homosexual unions have those socially favored qualities” (McConnell Reference McConnell, Olyan and Nussbaum1998, 250). Equating same-sex unions with traditional marriage will appear to many, for better or for worse, as a promotion or upgrade, and hence as a moral endorsement of these unions. As a further complicating factor, some experiencing same-sex attraction already feel marginalized, and those who cannot or will not enter long-term commitments may feel demoted by comparison with those who can and do. McConnell therefore suggests that a defense of same-sex marriage is best grounded on its ability “to publicly reaffirm the values of faithfulness and monogamy, while subordinating the more contentious moral question of homosexuality per se” (McConnell Reference McConnell, Olyan and Nussbaum1998, 251). He implies, then, that same-sex marriage would be an endorsement of the values that traditional marriage represents to many people, rather than an endorsement of same-sex relationships in and of themselves. Overall, he suggests that if same-sex rights supporters' concern is with the real effects of discrimination rather than ideological victories, they should not attempt to make antidiscrimination laws into moral statements. Otherwise, “it will be apparent that their real purpose is … to impose their beliefs through the power of the state” (McConnell Reference McConnell, Olyan and Nussbaum1998, 255).

Although McConnell's attempt to provide a “First Amendment” solution to moral disagreements on the value of same-sex relationships is ingenious in its conceptualization, I do not believe that it works. First, the institution of marriage itself carries a normative status. As put by Josephson (Reference Josephson2005, 271), “Marriage posits a specific desirable form for intimacy and family life — despite contemporary reality — and reinforces that form through legal, economic, political, and social privileges.” In other words, the contours of the institution are not neutral, but instead condition expectations and represent an endorsement of a particular and preferred view of how citizens should, ideally, conduct their lives. To traditionalist opponents of same-sex marriage, the inclusion of same-sex couples detracts from the value of a bedrock human institution by subjecting its definition to the vagaries of popular culture. To skeptics about marriage itself, however, this inclusion bolsters the sway of a rigid social institution at the expense of nonparticipants.

To put this differently, while traditionalist opponents fear that the inclusion of same-sex couples will devalue marriage, skeptics fear that it will add too much value, increase the institution's hegemony, and devalue them by comparison. Even the disestablishment of marriage as a public institution and the relegation of its spiritual and ethical meaning to the private sphere would not instantiate neutrality. Where traditionalist opponents want to maintain the status quo concerning marriage's participants, same-sex couples and their allies also seek no change in the substance of civil marriage, but only desire inclusion on the same terms as opposite-sex couples. Neither group wishes to engage in the merely private pursuit of what they understand as the value of marriage; both want the endorsement that accompanies only a public institution. As matters stand, traditionalists not only enjoy in most jurisdictions exclusive public endorsement of their interpretation of marriage, but also control the states.

Second and more important, there is an asymmetry in McConnell's “First Amendment” solution to conflicts regarding same-sex attraction. He is correct, of course, that the pluralist solution to religious difference has been to avoid a public position on the merits of contending religious views. When the government avoids a public position and avoids endorsing one or some religious views over others, however, it is attempting to be neutral as to the substance of the competing views in question. McConnell's solution, on the other hand, is like an attempt at neutrality between the idea that a particular religion should be accorded recognition equal to that accorded to other religions, and the idea that a particular religion, because it offends many people, need not be thus recognized and respected. David A. J. Richards suggests that same-sex attraction is a form of conscience central to ethical identity, but that it has traditionally been silenced, “condemned as a kind of ultimate heresy or treason against essential moral values” (Richards Reference Richards1999, 90; see also 92, 70, 126–127; and Richards Reference Richards2005, 107–108). From this perspective, opposition to identity claims as a form of conscience suggests that this identity “is as unworthy of respect as a traditionally despised religion like Judaism; the practice of that form of heresy may thus be abridged, and certainly persons may be encouraged to convert from its demands or, at least, be supinely and ashamedly silent” (Richards Reference Richards1999, 93; see also Richards Reference Richards2005, 108–109).

If we apply McConnell's solution to this analogy, it is as if those on one side were to argue that Judaism be recognized as on a par with other religious beliefs, while those on the other side argued that Judaism is an offensive belief system that need not be respected. Moreover, the government could then be neutral and avoid a moral position in this dispute. Although Jews would be protected against private violence and possibly against employment discrimination, there would be many conscience-based exemptions. Finally, the government would have to ensure that its efforts were not intended to effect a cultural education and transformation of attitudes. It would not be surprising if people became desensitized to instances of discrimination that would immediately be recognized as such in an alternative context. Political correctness would impel a nominal respect for Judaism, but cultural reality would convey heavy-handed reminders that such respect emanated not from conviction but from grace.

What McConnell is suggesting, then, is not neutrality or equality among religious positions, but neutrality among attitudes that citizens might take toward religious positions. He implicitly admits this point when he mentions the massive educational effort that would be needed to discipline “aggressive, undercivilized” military troops to accept openly gay comrades, concluding that the government would have to move beyond mere civil toleration to inculcate full acceptance of their moral legitimacy (McConnell Reference McConnell, Olyan and Nussbaum1998, 243). Openly gay individuals currently serve in the armed forces of most industrialized countries, with or without benefit of “massive” educational efforts. This point also suggests, however, that, when a dominant consensus exists, whether represented by those offended by a religion or by same-sex relationships, governmental neutrality between that consensus and the minority position has the effect, regardless of intention, of endorsing the status quo. Finally, where religious groups may consider their own beliefs and practices superior to those of other religious groups and may even criticize others, their own beliefs and practices generally possess an objective status or reality for them apart from their opinions of other groups. For individuals or groups who believe that acting on same-sex attraction is immoral, however, in McConnell's formulation expression of these attitudes is often central, rather than incidental, to their own “religious practice.”

Overall, then, McConnell's “First Amendment” approach does not instantiate a neutral public stance regarding controversy over same-sex attraction in general or same-sex marriage in particular. Although he intends to accommodate conflicting belief systems in an ostensibly pluralist fashion, the unavoidable privileging of the status quo renders his proposed solution far less evenhanded than he imagines. The fact that formal equality does not always instantiate substantive equality is well known. McConnell's discussion of same-sex marriage admirably illustrates this point. As long as civil marriage exists as a public institution, it is tantamount to an establishment, and thus functions as a gatekeeper. Public institutions and policies, or their absence, will always shape social attitudes. Because the privatization approach to First Amendment jurisprudence is not particularly compatible with the instantiation of civic equality regarding marriage, perhaps the equal access approach may be more promising. Positive action, rather than simply noninterference, may be required for the achievement of civil equality. It is to this issue that I now turn.

EQUAL ACCESS APPROACHES: FORMAL NEUTRALITY

Formal neutrality as it pertains to religion means that the government should neither extend special benefits to nor impose special burdens on organizations or activities because of their religious or nonreligious nature. This stance is akin to what Vincent Phillip Muñoz has called official noncognizance of religious belief and practice. Religion is a matter of individual obligation, and consequently lies outside the recognition or cognizance of civil society. “A government noncognizant of religion, in other words, must be blind to religion. It cannot use religion or religious preferences as a basis for classifying citizens” (Muñoz Reference Muñoz2003, 23). Both religious exemptions, which single out particular religious practices for protection, and nonpreferentialism, which treats all religions equally but favors religion over nonreligion, violate religious liberty, because this liberty conflicts with “the authority of the state to make these classifications in the first place” (Muñoz Reference Muñoz2003, 24, n.18; see also 28). That is, the government is still taking authoritative notice of religion. In establishment cases, noncognizance “requires that religious organizations and individuals stand in formal equality with nonreligious citizens and individuals.” Therefore, the government may not “exclude individuals and organizations from generally available benefits” (Muñoz Reference Muñoz2003, 30) based on religious affiliation. In free exercise cases, under noncognizance, the state may neither support nor exclude religion or religious individuals in their practices, regardless of benefit or burden, simply because these practices are religious (Muñoz Reference Muñoz2003, 31). In short, where religion partakes of public benefits, under formal neutrality it does so because its mission fulfills a secular purpose or social function that accords with public policy. Its religious nature is irrelevant.

Under formal neutrality, then, religious individuals or entities may enjoy benefits unavailable under the rubric of strict separation or privatization, despite their religiosity, but they cannot enjoy benefits and exemptions often proffered under strict separation because of their religiosity. That is, they may enjoy access to benefits equal to those accorded to nonreligious entities, but they are deprived of benefits that might otherwise flow from respect for their religious nature. An example of formal neutrality that benefits religion along with nonreligion can be found in Zelman v. Simmons-Harris (536 U.S. 639, 2002), in which the Supreme Court allowed the Cleveland school district to offer vouchers to students in failing schools that could be redeemed either in other districts' public schools or in private schools, both religious and nonreligious. Such a decision can be argued to support the free exercise of religion in a way that neither endorses nor penalizes religious choice. A classic example of formal neutrality that burdens religion as well as nonreligion may be found in Employment Division v. Smith, in which the Supreme Court ruled that the state of Oregon could deny unemployment benefits to members of the Native American Church for using peyote in religious ceremonies without demonstrating a compelling state interest, as the penalty for religious practice was only the incidental effect of a neutral and generally applicable law. The majority held in part that consideration of a religious exemption from the effect of such laws “would enmesh judges in an impermissible inquiry into the centrality of particular religious beliefs or practices to a faith” (494 U.S. 872, 1990), at 873). The overall effect, however, was to disregard the arguable centrality of peyote to religious practice in particular.

Formal neutrality or noncognizance can be applied in various ways to public policy concerning same-sex marriage. First, if we consider same-sex couples and opposite-sex couples as equivalent to two different “churches,” in McConnell's sense, neither type of couple wishing to marry could be excluded from the generally available benefits, including both material advantages and communal respect that accompany civil marriage. Both types would be regarded as “sharing in the same goods of love and commitment, reinforced by a similar bond of sexual intimacy” (McConnell Reference McConnell, Olyan and Nussbaum1998, 249), although in different ways. Same-sex marriage would not represent any sort of special right or “religious” exemption; it would mean admission to the institution of marriage on the same terms as traditional couples. For same-sex couples who consider the long-term commitment of marriage as the public expression of conscientious belief, marriage would also function as an expression of ethical identity. Government could not use “religion” or “religious preference” as a basis for classifying citizens for purposes of marriage. In this sense, government would be neutral.

If from one perspective formal neutrality facilitates the inclusion of same-sex couples in civil marriage, from another, however, this stance can hinder it. Insofar as government supports and prefers the values traditionally represented by marriage, it would include same-sex couples not because this represents a free exercise of conscientious belief, but because marriage fulfills a secular purpose or social function that accords with public policy. More specifically, under formal neutrality, both same-sex and opposite-sex couples may receive the benefits of marriage, just as, in Zelman v. Simmons-Harris, school vouchers may be redeemed at both public and private schools, including religious schools. Insofar as the government is religion-blind, parents who wish to redeem vouchers at religious schools benefit from formal neutrality. Religion-blindness cuts both ways, however. Just as it aids the free exercise of religion or conscientious belief in Zelman, it hinders free exercise in Smith. That is, when benefits to religion are derived from the judgment that according these benefits facilitates a secular purpose or approved social function, rather than the free exercise of belief, it is the secular public purpose that is in the driver's seat. The protections afforded to any second “church” are weak.

In Smith, religion-blindness meant that generally applicable laws against the use of peyote took precedence over the religious function of peyote in Native American religious ceremonies. Regarding same-sex marriage, similarly, the state-sanctioned values instantiated by marriage take precedence over what marriage might mean from a religious or ethically secular standpoint to those wishing to marry. In Monsma's terms, formal neutrality displays weak nonestablishment of “churches” because it retains marriage as a civil institution, as well as weak free exercise because its religion-blindness would justify the inclusion of same-sex couples only insofar as this would serve a secular purpose or approved social function, not because the ability to marry fulfills the free exercise of conscientious belief. Protections are nonexistent, additionally, for individuals experiencing same-sex attraction or same-sex couples who do not marry. Some may have conscience-based objections to what they see as a rigid institution. They are nevertheless deprived of the benefits of marriage, both material and symbolic, much as those who subscribe to no religion are often deprived of the respect accorded to those who are religious.

Moreover, when secular public purposes are the independent variable, so to speak, diverse religions may not in fact receive the equal treatment that would apply under nonpreferentialism, which accords identical treatment to all religions but ignores nonreligious belief systems. The beliefs and practices of some religions may be deemed more congruent with public purposes than others. The requirement that Mormons in Utah renounce polygamy, which they regarded as a religiously-mandated practice as do fundamentalist Mormons today, comes to mind as a historical example. And we would expect widespread agreement that if the practice were human sacrifice, public purposes should properly trump the free exercise of religion. Consider, however, the use of sacramental wine in Christian church services, which was exempted during Prohibition because this use was thought not unduly to conflict with the aims of Prohibition. I must entertain the conjecture that Christianity, as both a majority and a mainstream religion, received a pass denied to Native Americans. In specific application, then, the effects of formal neutrality may benefit some religions and religious practices while burdening others; that is precisely why it is “formal.” Regarding same-sex marriage, although opposite-sex and same-sex marriage may be two “churches,” traditional marriage is both a majority and a mainstream practice. Same-sex marriage is neither. Thus, people may more easily view traditional marriage as congruent with public purposes. Formal neutrality may require not only negative liberty but also positive action. Nevertheless, it may still be applied in ways that, in Gutmann's terms, constitute a public expression of civic inequality.

EQUAL ACCESS APPROACHES: POSITIVE NEUTRALITY

A second approach to equal access is represented by positive or substantive neutrality. On Monsma's view, “Consequences that discriminate are as oppressive to religion as intent to discriminate.” Therefore, if a generally applicable law even unintentionally “makes it harder for a person of devout faith to follow the tenets of his or her faith, that person's free exercise of religion has been hindered” (Monsma Reference Monsma and Monsma2002, 266). Positive neutrality resembles a pluralistic model of McConnell's that I term pluralist accommodation. McConnell understands the religion clauses to “guarantee a pluralistic republic in which citizens are free to exercise their religious differences without hindrance from the state (unless necessary to important purposes of civil government), whether that hindrance is for or against religion” (McConnell Reference McConnell1992, 168). The proper question for him is not whether a law or practice advances religion, but rather whether its purpose or effect will “foster religious uniformity or otherwise distort the process of reaching and practicing religious convictions” (McConnell Reference McConnell1992, 175; see also 169). Because individual believers must judge for themselves the dictates of conscience concerning their religious obligations, “the government must be ‘religion-blind’ except when it accommodates religion — i.e., removes burdens on independently adopted religious practice” (McConnell Reference McConnell1992, 177). As with formal neutrality, the government cannot favor or disfavor religion. Unlike formal neutrality, however, if public support is going to religious as well as secular nonprofit organizations, for example, religious organizations need not in McConnell's view adhere to general rules that burden their religious practices as the price of access to public programs (McConnell Reference McConnell1992, 184; see also 134, 185–186).

McConnell's model, then, is pluralist because it encourages a wide variety of religious expression and practice; it is accommodationist because accommodation, even involving public funds, is a tool for promoting pluralism, and discouraging religious uniformity. Like formal neutrality, positive neutrality promotes weak nonestablishment. But whereas formal neutrality does so because it is religion-blind and any support for religion is incidental, with positive neutrality the government may positively and knowingly support religious belief and practice if not doing so would have the effect, in Laycock's terms, of discouraging religious belief, practice, and observance over nonbelief, nonpractice, and nonobservance. Where formal neutrality requires weak free exercise policies, however, because free exercise cannot be an intrinsic value but only one that is incidental to the achievement of secular public purposes, positive neutrality promotes strong free exercise policies. That is, the very purpose of weak nonestablishment is to allow strong support for the free exercise of religion.

Regarding Zelman, for example, the redemption of school vouchers at private religious schools would be justifiable, as in formal neutrality, as an instance of weak nonestablishment, and because this is deemed good public policy from a religion-blind standpoint. Under positive neutrality, however, the policy would also be justifiable because it facilitates the free exercise of religion for parents who wish to send their children to private religious schools without incurring a financial penalty. Formal neutrality would allow the redemption of vouchers at religious schools under a generally applicable law that is deemed congruent with beneficial public purposes irrespective of their tendency to support religious practice. Thus, it would exemplify a weak free exercise policy. Positive neutrality, however, would promote such redemption as a double positive, so to speak: it fulfills an admirable secular purpose in providing avenues by which students may find alternatives to failing public schools, and it removes obstacles that might otherwise discourage or deter the free exercise, in McConnell's terms, of “independently adopted religious practice.” Thus, it would exemplify strong free exercise policy.

Smith as decided, by contrast, is an embodiment of formal neutrality. A generally applicable law prohibiting the use of consciousness-altering substances need not be justified by a compelling state interest despite religious users' penalty through the forfeiture of unemployment benefits. Although this disposition may complicate the free exercise of religion, public policy must be religion-blind, and officially noncognizant of this result. Under positive neutrality, on the other hand, the use of peyote might still carry penalties even for its religious use, but with a different justification. Such a case would require the demonstration of a compelling state interest if the religious use of peyote were to be prohibited, although a compelling state interest would not be required for the prohibition of other uses. The law at issue in Smith had the effect, though not the intent, of fostering religious uniformity, in McConnell's terms. If “the government must be “religion-blind” except when it accommodates religion — i.e., removes burdens on independently adopted religious practice” (McConnell Reference McConnell1992, 177), it must therefore be cognizant of the religious burden here imposed by the law. The default position, then, would be one of accommodation absent a compelling state interest to the contrary, in order to protect the free exercise of religion. Like formal neutrality, positive neutrality does not conflict with the absolute prohibition of particular practices that through the political process are deemed harmful or in conflict with good public policy, even if these practices emanate from religious belief. But where formal neutrality may forbid a practice blind to its impact on religion, positive neutrality might forbid a practice despite awareness or cognizance of the prohibition's impact on religion or conscientious belief. To put this differently, where positive neutrality accords benefits not in spite of but because of their facilitation of religious free exercise, it accords burdens not because their effect on religious practice is civilly immaterial, but in spite of the fact that religious free exercise will be deterred.

Regarding public policy toward same-sex marriage, at the outset positive neutrality appears more promising than formal neutrality. If, in McConnell's analogy, traditional marriage represents one “church,” confining marriage to opposite-sex couples tends to “foster religious uniformity or otherwise distort the process of reaching and practicing religious convictions” (McConnell Reference McConnell1992, 175). That is, same-sex couples who wish to marry may desire not only the respect and material benefits that accompany civil marriage, but also the ability to live according to conscientious convictions that may place a high value on civil marriage. Denied access to this institution, they will not “convert” to another “religion” and decide to marry opposite-sex partners. They will, however, at least implicitly carry the status of outsiders, of those not entitled to the recognition accorded to those who practice other “faiths,” because “religious uniformity” is the price of admission to the status of insider. Pluralist accommodation, then, requires that government accommodate and remove burdens “on independently adopted religious practice” absent a compelling state interest in refusing to do so.

Moreover, unlike the case of formal neutrality or noncognizance, the recognition of same-sex marriage as a second “church” would not be dependent upon its fulfilling a secular purpose or social function that accords with public policy. Rather, the intrinsic value of facilitating the free exercise of religion and conscientious belief would ground the recognition of a second “church.” The result would combine the weak nonestablishment and the strong free exercise hallmarks of positive neutrality. That is, marriage would still be a civil institution and thus an establishment, but it would be open to couples to whom it has heretofore been closed, thereby expanding the scope for the free exercise of religion and/or conscientious belief.

Even positive neutrality, however, has pitfalls. As noted, under formal neutrality diverse religions may not receive the equal treatment associated with nonpreferentialism, because the protection of religious practice is only incidental to the congruence of this protection with public purposes. Under positive neutrality, however, religious beliefs and practices may be favored at the expense of conscientious beliefs and practices that are nonreligious in nature. If government should be religion-blind except when accommodating religion, such accommodation may burden those whose beliefs and practices are central to their self-definition and ethical identities, but do not carry the same status as religious belief and practice. Accommodation may remove burdens from religious belief and practice, but it may do so at the expense of denying benefits to some of those whose beliefs and practices, although independently adopted in McConnell's sense, have not achieved the status of “religion.” Accordingly, this development may, in Laycock's terms, have the effect of encouraging religious belief, practice, and observance, while simultaneously discouraging nonbelief, nonpractice, and nonobservance, or at least of increasing their costs.

This point may be illustrated by President George W. Bush's faith-based initiative. Although many faith-based organizations have long possessed separate, nonprofit arms that provide social services that are formally walled off from the organizational promotion of religion, the current program allows religious groups to maintain their religious character and to use religious criteria in selecting employees (Goodstein Reference Goodstein1997, A16; see also Stevenson Reference Stevenson2002, A28). In one sense, public funding characterizes formal neutrality: the government is noncognizant of or blind to religion, as it does not distinguish between religious and nonreligious allegiances in awarding funds. As McConnell notes about public support in general, “Indeed, to deny equal support to a college, hospital, or orphanage on the ground that it conveys religious ideas is to penalize it for being religious” (McConnell Reference McConnell1992, 169). Aid to religion is incidental to support for the secular public purpose that an organization might pursue. On the other hand, noncognizance cuts both ways. Just as formal neutrality will not allow exemptions from penalties for the use of peyote in a religious context; neither will it allow government to grant exemptions from generally applicable laws to social service providers simply because these providers are religious. If there is no secularly based right to discriminate on the basis of religion or sexual orientation, for example, there is similarly no religiously based right to discriminate.

Alternatively, the faith-based initiative may be interpreted to exemplify positive neutrality. Aid to religious social service organizations accommodates religion, removes burdens on religious practice, and avoids encouraging religious uniformity. On the other hand, although this pluralist accommodation may aim at exempting religious organizations from seemly burdensome nondiscrimination laws, these exemptions render it more costly for individuals of the “wrong” faith tradition or sexual orientation to find employment or to be open about their identities. Accommodations do not coerce individuals to undertake positive actions that they would prefer to eschew. Nevertheless, they publicly sanction the private placement of burdens on independently adopted religious or sexual practices by individuals.

In 2002, for example, a psychological therapist at a United Methodist Children's Home in Decatur, Georgia, was fired on the discovery that she was a lesbian, as the home objected to her nonconformity with its religious doctrines. Although neither federal nor Georgia law protects against discrimination in hiring based on sexual orientation, the woman's lawyer contended that civil rights laws “protect against religious discrimination that takes the form of requiring an employee to lead the kind of life and subscribe to the kinds of beliefs that assert there is only one true and virtuous path.” The case was settled out of court in favor of the therapist (Scott and Badertscher, Reference Scott and Badertscher2003). While positive neutrality may be religion-blind toward faith-based organizations, it is anything but religion-blind, at least indirectly, when it allows organizations to use litmus tests in hiring individuals who provide social services that are not themselves intrinsically religious in nature. If employees are based in an area where most opportunities are not only in faith-based settings, but in settings that are also Christian and heterosexist, the scope of their civil enjoyments and opportunities is significantly narrowed. Thus, although positive neutrality may encourage the exercise of religious difference by communities of faith, it may simultaneously hinder individual free exercise by burdening “independently adopted religious practice.” In Gutmann's terms, by making distinctions between insiders and outsiders, such policies constitute a public expression of civic inequality.

BEYOND NEUTRALITY

If the inclusion of same-sex couples in the institution of civil marriage is analogous to adding a second “church” to a religious establishment, this change does not help those who eschew any “church.” That is, the status of individuals or same-sex couples who for whatever reason do not wish to marry is similar to the status of those in an ostensibly religious nation who subscribe to no religion. If the inclusion of same-sex couples in marriage fosters “religious” diversity and removes burdens from a nontraditional sort of “religious” belief, practice, and observance, it may simultaneously, however, burden individuals and/or couples whose convictions involve nonbelief, nonpractice, and nonobservance. Just as religious belief and practice is not for everyone, marriage is not for everyone, despite the exhortations of some enthusiasts. Jonathan Rauch, for example, suggests that the growing prevalence of domestic partnerships and civil unions, complete with material benefits, competes with marriage and thereby devalues it. “If the institution is undermined, “The culprit … is not the presence of same-sex couples; it is the absence of same-sex marriage.” Rather than simply “a lifestyle option,” marriage should be expected of all committed couples and privileged as “better than other ways of living” (Rauch Reference Rauch2005, 91; see also Sullivan Reference Sullivan1996, 7, 99–100, 106–116). “Marriage is for everyone — no exclusions, no exceptions” (Rauch Reference Rauch2005, 6). He could hardly make the point more emphatically.

Skeptics who deplore the rigidity of marriage, on the other hand, extend their skepticism to same-sex marriage. Extending marriage to the previously excluded, observes Valerie Lehr, will not help those in need of benefits who cannot or will not marry, such as those whose “families” comprise a network of close friends. “That is, the extension of marriage rights might well make it harder for us to form the ‘families’ that we choose by extending the reach of family as defined and regulated currently” (Lehr Reference Lehr1999, 33). She prefers contesting the norm of state regulation rather than extending it (Lehr Reference Lehr1999, 34). The legal recognition of same-sex marriage extends a benefit to some, but it extends a forced choice between the rigidity of marriage and exclusion of a valuable status to others. As Josephson puts it, skeptics “share a serious concern that same-sex marriage will establish a new form of ascriptive citizenship that appears to include sexual minorities, but in fact excludes most LGBT [lesbian, gay, bisexual, transgendered] persons” (Josephson Reference Josephson2005, 274). The recognition of same-sex marriage still endorses a preferred way of life. By extending this option to more couples, however, this move ironically renders more problematic the status of those who cannot or do not wish to participate. The more inclusive marriage becomes, the more conspicuous will be those excluded, just as agnostics and atheists stand out among a majority of self-reported believers. Regarding both religion and sexuality, the larger the number of insiders, the more likely it is that outsiders will be regarded as deviant from the norm.

Equal access under the rubric of either formal or positive neutrality, then, can bolster majoritarian interpretations and expressions of either religion or sexuality at the expense of unpopular or unconventional beliefs and practices. Strict separation advocate Gregg Ivers, arguing the impossibility of neutrality in the application of legal principles, compares the definition and interpretation of particular principles to the design of a golf course. Various players bring diverse levels and types of skill and preparation to their games. “Moreover, golf-course architects have in mind certain types of players when they design courses. Not all players are expected to perform well on all golf courses. An architect purposely gears a certain course to particular strengths and weaknesses … How else does one explain the common refrain of golfers struggling through an endless round that ‘the course just wasn't set up for my game’?” (Ivers Reference Ivers, Monsma and Soper1998, 169).

Traditionally, marriage has been a public institution which, like a golf course, is “set up” to favor opposite-sex couples who desire to undertake long-term commitments, and its material benefits reflect this fact. Their “game” is therefore favored by the mix of strengths that they bring to it. Do we disestablish marriage as a public institution altogether; that is, do we dismantle the golf course because some are outsiders and experience a sense of isolation? If so, same-sex and opposite-sex couples, as well as individuals, who do not marry, will all “play golf” in the rough, or where they can since there are no established courses. The only rules will be those that they themselves devise. Or, alternatively, do we keep the game of golf, but admit players to the current courses who wish to play by current rules, and perhaps design other courses so that we have a mixture of courses that tap a variety of players' strengths?

Those who would retain the institution of marriage as a civil establishment could mitigate the effect of this establishment by broadening it to include all couples who wish to play this particular “course.” Other “courses” might of course be designed, but their justification and details are beyond the scope of this essay. Although positive neutrality may favor religious practice and observance over nonpractice and nonobservance, and some kinds of practice and observance over others, the solution is a reorientation of positive neutrality. That is, until or unless demonstrated to be contrary to a compelling state interest, the government should avoid exclusions that have the effect of fostering uniformity in the couples who believe in the long-term commitment that marriage represents and who seeks to bear witness to this belief by the “practice” of marriage. The exclusion of same-sex couples from marriage will not, in McConnell's terms, “distort” the process of reaching convictions concerning their personal commitments, but it surely “distorts” — actually, prohibits — the ability of these couples to “practice” them in the same manner as opposite-sex couples. Although they may live as if married, “without benefit of clergy” or county registrars in most jurisdictions, they must do so without the tangible benefits and intangible respect accorded to traditional couples. Because the desire to marry may stem from couples' conscientious convictions about the kinds of lives they are called upon to live, the government must be sexual orientation-blind, as it were, except when it accommodates differences in sexual orientation by removing burdens on independently adopted personal commitments.

Trade-offs exists here, however, and it would be disingenuous to conclude this article without a frank admission of this fact. First, I believe in the legal separation of religion and state. As the discussion of skeptics about marriage indicates, making concessions to establishment in the realm of marriage may appear to those uninterested in marriage much the same as concessions to religious establishment appear to agnostics, atheists, and the generally nonreligious. A greater degree of free exercise may aid minority religions and sexual orientations, but it may also increase the hegemony of religion in general or marriage in general to the detriment of those who wish to participate in neither type of institution. Laycock suggests that under full separation, only religious unions would be marriages, while civilly sanctioned relationships would be civil unions for all, both same-sex and opposite-sex couples. Religious and secular dissolutions of marriage would be similarly separated. His point is that the conflation of religious and legal relationships contributes to the acrimony of the debate over same-sex marriage. “We have so combined a religious institution with a legal one that millions of Americans share the … view that the ‘sanctity’ of marriage somehow depends on law, not faith” (Laycock Reference Laycock, Laycock, Picarello and Wilson2008, 207; see 201–207). Separating religion and the state in administering the institution of marriage would augment the free-exercise rights of same-sex couples who wish to marry. It might lessen the seeming hegemony of both religion and marriage, however, by separating a seemingly monolithic institution into two, the religious part of which would itself be administered by many diverse faith communities. Overall, moral value inheres in legal separation. We should not lose sight of this fact despite the virtues of positive neutrality in the current political climate.

Second, some commentators have expressed concern that the widespread institutionalization of civil marriage for same-sex couples may be accompanied by a threat to the religious freedom of both individuals and faith communities. Some may not wish to provide goods or services to same-sex couples who according to these providers' conscientious beliefs are engaging in sinful behavior. Although detailed discussion of these issues is beyond the scope of this article, I want briefly to address the fear that religious communities refusing to perform same-sex marriage might be “punished” by the loss of their tax exemptions if sexual orientation were added to generally applicable nondiscrimination laws. Douglas Kmiec argues, however, that tax exemptions are not subsidies, where the provision of public funds entitles the government to some control over their expenditure. The public benefit conferred by religious communities that unite couples in marriage, moreover, does not make these communities “state actors,” as they are not the exclusive providers of marriage rites (Kmiec Reference Kmiec, Laycock, Picarello and Wilson2008, 108–116). New legal questions arise with any change in the law. Although a host of issues may present themselves with the widespread recognition of same-sex marriage, this possibility ought not to deter those who support the change.

Meanwhile, as put by Daniel Brudney, “It is worth reflecting on the fact that people who otherwise seem hostile to state institutions, who deem them corrupt, wicked, or at best a necessary evil, nevertheless deeply want the state to endorse their point of view” as representative of “the people.” He suggests that this desire underlies the arguments of both proponents and opponents of same-sex marriage. “That dispute is increasingly not about the provision of concrete legal rights and benefits … but about whether the term ‘marriage’ is to be applied to a relationship — and applied not by a minister, priest, rabbi, or imam but by an agent of the state” (Brudney Reference Brudney2005, 832). In other words, lack of interference with private beliefs and practices is insufficient. What is wanted is positive action or approval, the provision of a context or public framework within which all couples may bear witness to their conscientious beliefs, religious, or otherwise.

The effect of disestablishment or privatization of marriage would be comparable to the closing of swimming pools in some southern communities during the civil rights movement to avoid being forced to admit blacks, or to the elimination of after-school clubs in some public schools to escape having to allow gay clubs on campus. Not only could no one swim, black or white, and not only could no students organize, gay or straight. In addition, the context within which the disfavored groups lived remained as before. The beliefs and practices, so to speak, of blacks and gays were in no way facilitated simply because those of whites and straights were also curtailed. The status of blacks and gays in the larger society was in no way improved even though whites and straights received “equal treatment.” In each case, both the favored and disfavored groups were burdened in a formal sense. But given the social context of the larger society, the beliefs, practices, identities, and commitments of only the disfavored groups were burdened in a substantive sense.

The institution of civil marriage is meant to serve a social function and public purpose. I am not suggesting here that such a purpose is illegitimate. Rather, what is illegitimate is its selective and unjustified application in ways that render citizens civically unequal. Whether or not alternative institutions should exist that accord material benefits to those who do not marry is a separate question. Civil marriage in its present form is a public institution, and as such, it is sanctioned and encouraged by the state. Yet some couples are excluded, although they wish simply to make the same kind of formal, long-term commitment available to couples who are included. Although marriage may or may not include a religious component, it always involves a civil component, requiring a license to be legitimate. Moreover, religious marriages are performed by clergy who function not only as religious authorities, but also in the United States as officers vested with state authority on these occasions. If no civil institution of marriage existed and if all marriages were religious, same-sex couples would be excluded by some religious traditions, which is their right. Such exclusion, however, would not instantiate a public expression of civic inequality. If, in McConnell's terms, same-sex marriage proponents are asking the government to “adjust the definition of a public institution,” marriage's status as a public institution is exactly the reason that its constituency should be broadened to include same-sex couples. If legitimate reasons exist for maintaining the institution of civil marriage as an establishment, legitimate reasons also exist for administering it in ways that promote the free exercise of conscientious belief.