Political religion vs non-establishment: Reflections on 21 st -century political theology: Part 2

IV Liberal alternatives: Replacing separation with equal liberty

The prospect outlined in the Abstract has spurred the development of an alternative constitutional theory of the religion clauses. The ‘equal liberty’ approach also decenters the separation frame. ¹ Unlike the integrationists, however, proponents seek to recast the establishment clause in terms of an equality analysis geared to the protection of vulnerable minorities and individuals against discrimination within an overarching political secular frame.

The strongest attack on separation from the equal liberty perspective is in the work of Eisgruber and Sager. ² Their critique is innovative in that it rejects not only the separation metaphor but also its underlying premise, namely, that religion is so special that it should be constitutionally privileged in some respects and disadvantaged in others. ³ The religion clauses have long been read this way, unsurprisingly since the Constitution does treat religion as distinctive. But the separation frame has apparently led to incoherent jurisprudence and conceptual confusion. ⁴ The integrationists’ alternative is to drop the rhetoric of separation while retaining the premise that religion is special – intrinsically valuable and instrumentally so for democratic republics. Thus they seek to expand privileges for religious groups with respect to immunities from regulation (one-sided separation) while insisting on equal treatment by the state (neutrality as non-preferentialism) with respect to presence in the public sphere and aid. ⁵ Eisgruber and Sager reject both the separation frame and the idea that religion’s intrinsic characteristics make it special in a constitutionally relevant sense, as historical anachronisms. ⁶ They retain separation in the narrow sense of ruling out establishment of an official national Church or religious text. But separation rhetoric has already served this purpose (disestablishment) and is no longer needed.

On the ‘equal liberty’ view, religion should not be given special constitutional treatment unless there are concerns about discrimination and inequality. Religious actors neither have a presumptive constitutional right to exemption from valid laws, nor should their organizations be disadvantaged in the disbursement of government funds when they carry out a function supported by the government. ⁷ Moreover the separation metaphor cannot tell us what should be done in specific cases and thus it invites contradictory applications and disingenuous distinctions. Typically the issue is not whether but how the state should be permitted to affect religion or vice versa. ⁸ Unlike the integrationists, however, Eisgruber and Sager apparently abandon both sides of the ‘wall’ metaphor’s prohibitions expressed in the separationist slogan, ‘no aid, no hindrance’. ⁹ The equal liberty standard purports to capture what is best in our constitutional jurisprudence and to provide determinate, coherent criteria for deciding cases. It entails three propositions: no member of the political community ought to be devalued or privileged due to his or her spiritual commitments; we have no constitutional reason to treat religion as deserving special benefits or as subject to special disabilities; all persons should be given an equally capacious range of constitutional liberty including free speech, personal autonomy and associative freedom. ¹⁰

The idea is that invidious discrimination on the basis of religion or irreligion is no better or worse than it is on the basis of race, ethnic origin, gender, sexual orientation, etc. ¹¹ Religion’s special status in our constitutional tradition derives from the vulnerability of people to discrimination on that basis, particularly salient at the time of the founding, not from the special value of religion. ¹² Individuals’ standing in the community should not depend on their religion; religious minorities should not be subject to unequal treatment by the state. Eisgruber and Sager thus embrace O’Connor’s endorsement analysis with respect to the government sponsorship of religious messages, interpreting it as an anti-discrimination principle. State endorsement means that equal concern and respect are denied to some religious and all non-religious citizens. ¹³ Implicit disparagement of minorities is unavoidable when the government sponsors religious expression because it casts some groups as socio-cultural (religious) insiders. ¹⁴ Here equal treatment of individuals is the key constitutional criterion. The gambit of multiple establishment, i.e. equal inclusion of the major religious symbols in the public spheres of the state, civil society, public schools, etc., is rejected because the idea that it could ever exclude exclusion is as unconvincing as the claim that the presence of only secular symbols somehow excludes the religious. ¹⁵

But, the way Eisgruber and Sager apply their equal liberty principle to state financial aid, be it in the form of tax exemptions, school vouchers, or funding of religious welfare institutions, is neither convincing nor consistent. Equal liberty apparently requires that when making funds available, the government avoid preferring, endorsing, or affiliating itself with a particular religious doctrine or message. ¹⁶ Since religion is not special, religious organizations may receive funding when serving the same public function as secular organizations, say in running hospitals or schools. There is one proviso: when government sends resources to religious groups (charities, schools) through whatever indirect route (tax breaks or vouchers) it must ensure citizens have a meaningful secular alternative (i.e. choice), and it must not privilege any religion. The criteria of genuine choice – no favoritism among religions and ‘neutrality’ between religion and non-religion – are the main equal liberty concerns they discuss. ¹⁷ They thus seem to embrace the non-preferentialism of the integrationists, on the ground that religion is not special and thus should be treated the same as non-religious organizations serving similar functions.

As critics have pointed out it is not entirely clear on non-discrimination grounds why a meaningful secular alternative is required if religion is not special. ¹⁸ Nor is it clear whether direct aid to such religious institutions would be acceptable given secular alternatives. If religion is not special, why insist on these conditions? I agree with Greenawalt: equal liberty cannot do all of the work here. ¹⁹ There are important reasons other than concerns about discrimination against minorities, for government to stay out of the realm of religious truth – and these are at stake when any aid or funding is given even if it is indirect. ²⁰ Moreover, Eisgruber and Sager do seem to rely on the special nature of religious affiliation and belief when addressing the question of why no constitutional fairness issues arise regarding state endorsement of secular symbols, by insisting on a secular alternative given ‘indirect’ funding of religious schools and in defending the ministerial exception. They mention four distinctive features of religious affiliation on which they, despite disclaimers, partly base their constitutional analysis: (1) religious belief and affiliation are important components of individual and group identity; (2) religious memberships constitute an expansive web that invariably constructs in-groups and out-groups; (3) the stakes of inclusion can be experienced as high: exclusion can have dramatic consequences such as being eternally damned or shunned; and (4) religious doctrines have a comprehensive character insofar as they speak to ultimate questions of life’s meaning, origins and value whereas secular principles and institutions are self-consciously incomplete. ²¹ These features of religious association are dispositive in their endorsement analysis and in their insistence on a secular alternative in funding cases. Apparently religion is special after all.

But there is a more disturbing feature to the way Eisgruber and Sager parse equal liberty with respect to funding. In their concern to ensure non-discrimination among religious and non-religious groups, they neglect to apply these principles to vulnerable individuals and minorities within religious groups or under their control. Having abandoned separation analysis and accepted tax breaks, and at least ‘indirect’ state funding for religious organizations including schools, it is astounding that they do not make that conditional on compliance with constitutional non-discrimination principles vis-à-vis vulnerable members of religious organizations or with respect to their employees! ²² Indeed, they begin their book by noting that organized religions run big enterprises from hospitals to drug rehabilitation centers to schools and universities with many employees who are not co-religionists. Moreover, they ridicule the idea that religion or choices about religion could remain unaffected by the state. ²³ ‘The real question is not whether government should affect religious choices, but how it should do so …’ ²⁴ Yet there is no serious discussion of establishment clause requirements regarding the regulation of self-regulation when state aid is given, in light of the principle of equal liberty!

When it comes to exemptions for religious groups (already enjoying tax breaks) who discriminate in employment against women, against non-adherents, on the basis of sexual orientation, or with respect to race on ‘religious grounds’, they invoke general constitutional protections of personal liberty and freedom of association (the third prong of their equal liberty analysis) that allegedly ‘run to the benefit of all members of our constitutional community’, as justification. ²⁵ This may be convincing with respect to choice of clergy, but one wonders what the ‘equal liberty’ justification for the infamous ‘ministerial exception’ articulated by the courts with respect to Title VII of the Civil Rights Act (the federal law prohibiting sex discrimination) might be absent the understanding that religion is special. ²⁶

It is surprising that advocates of an equal liberty standard make no objection to a long line of highly controversial cases that permit discrimination by religious employers on the basis of religion while relying on freedom-of-association justifications to defeat establishment clause challenges. Eisgruber and Sager seem to accept the court ruling that the National Labor Relations Board is without authority to certify a union as a bargaining agent for lay teachers in schools operated by a Church. ²⁷ They defend the broad interpretation of the religion-specific exemption granted under Title VII to religious organizations to discriminate with respect to all their employees, even those with no connection to religious functions, upheld by the court in a 1987 case in which an African-American janitor who worked for 16 years in a gymnasium open to the public, run by the Mormon Church, was then fired for not being a member of the Church. ²⁸ Apparently associational autonomy protects minority religious groups from determinations by ‘outsiders’ about how they treat their employees! What is not considered is the violation of the freedom of religious conscience of those very employees who brought suit, not to mention their individual claims to equal liberty. ²⁹ They avoid discussing the notorious case in which the Supreme Court (rightly) rejected a ‘freedom of religion’ challenge to the withdrawal of tax-exempt status from a private religious university due to racial discrimination on the grounds that it was not a charitable institution because it violated fundamental principles of constitutional public law and public policy. ³⁰ One wonders how they would apply their equal liberty analysis to the recent extension of the ‘ministerial exemption’ to teachers in religious schools, denying them coverage of the Americans with Disabilities Act that forbids the firing of an employee due to illness or for lodging an official complaint. ³¹ Indeed, if religion is not special, why grant standing to taxpayers in cases of ‘indirect’ funding to religious schools at all? ³²

Eisgruber and Sager do say that states are free to impose conditions when it aids religious schools or charities but they argue that this is not constitutionally required on their equal liberty approach. ³³ They thus reject the idea that ‘indirect’ funding by the state of religion should give it a lever to induce religious employers to conform with federal and constitutional anti-discrimination principles.

This just will not do. To be consistent, if one replaces the separationist ‘no aid, no hindrance’, ‘religion is special’ approach with an equality analysis that applies to individuals, then constitutional equal protection against discriminatory treatment for those within or under the jurisdiction of religious groups would have to follow. Otherwise, if the state ‘indirectly’ aids religion on ‘neutral’ grounds, it would unacceptably privilege religious groups and become complicit with (endorse) the messages their discriminatory practices convey. ³⁴ Unconditional aid would not only entail a special form of state support of religion that the establishment clause was designed to prevent (distinct from its treatment of non-religious groups it aids) – it adds insult to injury by taxing individuals to subsidize religious enterprises that may discriminate against them! Whether the issue is tax exemptions for religious properties used entirely for religious purposes, or subsidies for religious organizations that carry out public purposes, government aid cannot be unconditional on a coherent equal liberty standard of non-establishment.

Indeed the distinction between direct and indirect aid does not help: quite the contrary. As Justice Kagan put it in her dissent in a recent case, tax breaks or other forms of ‘indirect’ funding, e.g. through school vouchers or school tuition organizations, achieve the same thing as ‘direct’ appropriations of monies in that they provide financial support to select individuals or organizations and thus either way religion is financed by government. ³⁵ Money is money. Denial of taxpayer standing to challenge government’s monetary support of religion because it is ‘indirect’ effectively de-constitutionalizes the establishment clause, by insulating its financing of religious activity from legal challenge. ³⁶ While there may be a legitimate public purpose behind such laws, to find out whether this is the case, the individual must have her or his day in court. And to survive an establishment clause challenge, under consistent equal liberty analysis, indirect state aid to religious ‘charitable’ enterprises must subject them to the same constitutional anti-discrimination principles and policy requirements that other private recipients have to follow.

Indeed, since the incorporation of the establishment clause under the 14th amendment, the survival and propagation of religion can no longer be considered a valid public purpose of the state. Decisions to the contrary are flawed because they operate on anachronistic historical premises. In a telling dissent in the 1970 case in which the Supreme Court upheld a tax exemption for a religious organization whose properties were used exclusively for religious worship, Justice Douglas argued that the mere fact that for over 200 years states provided such exemptions is irrelevant because this preceded the mid--20th-century rights revolution. ³⁷ While the state does aid other non-profit organizations, that is because they carry out a public function the state itself would otherwise carry out, such as care for the sick, the aged, the orphaned, or the poor. But churches used for religious purposes do not fit into this class because religion is not a public purpose under the federal constitution and the state (the states) is (are) prohibited from carrying out religious functions! ³⁸

This clarifies the flaw in both the equal liberty and the integrationist versions of ‘neutrality as non-preferentialism’. If fostering religion is not a valid public purpose under the establishment clause, then there are no grounds for granting religious institutions tax exemptions merely because they are religious. Nor are there grounds for giving only religious charitable institutions immunity from constitutional equal protection principles that otherwise apply when private institutions receive state aid for providing desirable public services. The regulatory, welfare state does frame education, care for the sick, the aged, the poor, support for culture, and so forth as public purposes and it does deem private organizations, religious or secular, that perform them as providing public services. But pace integrationist ideology, and Eisgruber and Sager’s version of the equal liberty approach, it is incumbent upon the state to influence all ‘charitable’ organizations receiving exemptions, in the right way. The point was made clear in Justice Brennan’s concurrence in Lemon: ‘[W]hen a sectarian institution accepts state financial aid it becomes obligated under the Equal Protection clause of the Fourteenth Amendment not to discriminate.’ ³⁹

Justice Brennan also put his finger on the constitutional conundrum religious organizations face: if they demand and receive tax exemptions and other forms of ‘indirect’ public funding by virtue of performing a public function, the Church institution, be it a school or a charity, may become public for more purposes than the Church may wish. It may then realize that its victory on the establishment clause front has meant abandonment of a convincing free exercise claim to absolute autonomy. ⁴⁰ Indeed. Today’s integrationists are well aware of this prospect and that is why they try to ‘nuke’ the establishment clause under the banner of freedom of religion. They trade on habit and ‘tradition’ with regard to immunities and privileges that organized religion enjoyed in the past and still does under regimes of official legal pluralism elsewhere, hoping to shift our constitutional democracy in that direction. The ideology invoked is that religion should be treated fairly; the discursive frame of this alleged neutrality claim is, as we have seen, non-preferentialism. But neutrality as evenhandedness between religion and non-religion does not exist as a constitutional principle in the USA, nor should it. The correct equal liberty response to the integrationists should be that freedom of religion does not entail either an unconditional constitutional right to state aid or immunity from the regulation of self-regulation.

V Separation and reflexive regulation: Squaring the circle

We are indeed in the epoch of the regulatory state. Political secularism and the concomitant separation on the level of ends, institutions, legitimacy, and symbolically, is as important for preserving liberal constitutional democracy in the interventionist as it was for the laissez-faire epochs. It is a category mistake to assume that separation means no relation between that which has been differentiated. Yet criteria are needed for determining when an accommodation, a regulation or no regulation is appropriate. Separation cannot do that work.

For this, equal liberty is helpful, but it must supplement, rather than replace, the separation baseline, and it, in turn, must be supplemented with other political principles to orient establishment clause jurisprudence. As a first step, I propose a radicalized equal liberties approach that applies to individuals wherever they are situated. Equal liberty should not be construed as a non-preferential ‘neutrality’ principle applying to religious and non-religious groups. As an anti-discrimination standard it must have as its ultimate referent the individual, not the group, and apply directly to the person whose liberty (of conscience) and equality are at stake.

Indeed the reductionist attempt to avoid the inevitably messy process of adjudicating the religion clauses by denying that religion is special and by relying on general non-discrimination principles to protect religious minorities, risks upending the dualistic constitutional approach that is crucial to guaranteeing religious freedom and civil liberty to all. It plays into the hands of those masking populism and multiple establishments as ecumenism and pluralism. But constitutional dualism assumes that there is something special about religion that renders individual religious conscience and religious minorities distinctly vulnerable and religious organization particularly dangerous in a liberal democratic constitutional republic. The debate over whether and in what way religion or religious conscience is special is ongoing with no knockdown arguments on either side. ⁴¹ I go with pragmatic and historical arguments for according religious conscience and religious organization specific constitutional treatment within an overall frame of political secularism. Here is why. As Eisgruber and Sager note, religious doctrines are comprehensive in a distinctive way in that they refer to the sacred regarding values and rites, which are not seen as humanly conferred but rather are deemed intrinsic and obligatory. Although they are hardly immune to change or internal contestation, religious convictions and practices do reference injunctions that allegedly come from external (transcendent) sources and are in that sense not experienced by the believer as a matter of choice. This is so regardless of the sociological facts of conversion, and shifting ‘choice’ of religious denominations. These beliefs, rites and reasons are fully accessible only to members of the community they help constitute. The latter construct social identity groups that inevitably create insiders and outsiders, members and non-members. Such ‘nomos communities’ typically have gatekeepers with jurisdictional power over insiders regarding salvation, membership and good standing. ⁴² Religious authorities can invoke the value of what the religious believe exists outside and beyond the self, shoring up their interpretive and gatekeeping power, ensuring compliance with the rules and the internal (often gendered) hierarchies of religious communities. Religious authorities’ pronouncements on contentious public issues, especially those having to do with gender and family law, may carry a special weight, which could be reinforced if endorsed by public authorities. In short, religious groups and organizations involve power relations internally with respect to members and externally in competition with one another over recruitment, and over state largesse and power.

This is why, historically and in the current context, religious conscience still requires special constitutional protection, why the democratic state cannot deem a religious identity or any ecumenical conglomeration of religious identities as coterminous with the identity of ‘we the people’ and so must be neutral with respect to religious doctrine, symbols, or creed. This is also why the regulation of self-regulation is a desideratum of any liberal constitutional republic that takes the rights, concerns and voices of all its citizens, including minorities and vulnerable members within or employed by religious groups, seriously. It is unsurprising that the constitutions of every state along with international and regional human rights conventions give religion special mention. A liberal democratic constitutional republic draws its authority from immanent sources (‘the people’, correct procedures, the rule of law, etc.) and construes its members as political equals and judges of the values that guide their lives. Thus it can neither invoke reasons to justify policies that are not accessible to all nor adopt a law or policy that endorses a social identity that is less inclusive than the citizenry as a whole, and no religious identity fits that bill for any modern society. ⁴³ To do so would require some citizens to accept justifications of public law and policy that deny them recognition as full members of the polity. ⁴⁴ The constitutional dualism of the American religion clauses thus provides special protection for religious conscience and free exercise along with special restrictions to preclude establishment – be it through state endorsement, collaboration, coercion, or funding. Principled non-establishment along these lines protects the civic quality, civil identity, public purposes, autonomy and inclusiveness of the political community’s institutions along with the integrity of religious conscience and association. The anti-discrimination aspect of equal liberty gets at only one aspect of this.

In short, the freedom of association and equality among groups is insufficient. Martha Nussbaum’s focus on freedom of individual conscience rather than freedom of religion takes a crucial step in radicalizing the equal liberty approach. She explicitly rejects group-based non-preferentialism. ⁴⁵ Her claim is that religious liberty is ultimately based on respect for the faculty in human beings with which they search for life’s meaning. ⁴⁶ The ‘respect conscience principle’ is universalistic insofar as the faculty of conscience is held to be present in all human beings. Equal liberty with regard to conscience requires that government accord a protected space around it so that people can act as their consciences dictate provided they do not violate the rights of others or a compelling state interest. Unlike Eisgruber and Sager, however, Nussbaum’s equal liberty analysis assumes religious conscience and practice require special constitutional treatment for many of the pragmatic and historical reasons noted above. Yet she would extend constitutional protection and accommodation to cover non-religious conscience as far as administratively possible. ⁴⁷

Nussbaum’s equal liberty approach, moreover, supplements choice and associational freedom with a capabilities perspective, helping us to get at the problem of adaptive preferences endemic in religious association. ⁴⁸ Liberty of conscience and consent pertain to some but not all of the core capabilities at stake in religion–state relations. ⁴⁹ Schools and families are agencies of primary socialization. From the child’s perspective religious association is not voluntary, but ascribed via these agencies. If, in the name of religious norms, the young are socialized into gender hierarchies and the opportunities of girls to develop their capacities are radically restricted then equal liberty is violated despite a legal guarantee of exit and ‘choice’. In these and other cases, what she calls the principle of moral constraint requires non-violation of the central capabilities of individuals by religious groups. Accordingly, ‘we should refuse to give deference to religion when its practices harm people in the areas covered by the major capabilities’. ⁵⁰ Here, in short, the regulation of self-regulation is called for. Whether or not one agrees with her capabilities approach to justice, Nussbaum is on the right track in proposing the principles of non-subordination, non-domination and non-curtailment of capabilities to supplement liberal understandings of equal liberty, as criteria for line-drawing around regulatory issues. Her nuanced analysis of the dilemma that exists between acknowledging the intrinsic value of respecting people’s capabilities as participants in a religious way of life and taking seriously the risks posed particularly to the development of women’s capabilities in traditional religious cultures, is impressive.

Nevertheless, she errs in downplaying the work that separation as a discourse does. She cedes too much when she states that the rhetoric of separation implies that the goal of the establishment clause is to purify the public square of all reference to religion, in effect ‘establishing’ secularism as a theory of government. ⁵¹ This is nonsense both historically and conceptually. Like Eisgruber and Sager, Nussbaum proposes replacing separation talk with equal liberty analysis (of conscience and capabilities) regarding establishment clause issues, guided by the ‘endorsement test’. ⁵² She too is concerned primarily with equal liberty of conscience for all religious believers, and thus she construes the establishment clause as ‘just another way of shoring up free exercise, together with other capabilities of citizens’. ⁵³ Accordingly, the motivation behind the establishment clause is to prevent citizens from being violated in conscience and practice by the pressures of a dominant religious group with political and legal power behind it. The motivation behind free exercise is to prevent belief and worship from being impeded or burdened by public action. Thus non-establishment is ‘really about both free exercise and full equality as citizens’. ⁵⁴

Nussbaum is well aware of the contemporary assault on the establishment clause by prominent justices on the Supreme Court and her aim is to counter these moves. She roundly rejects Justice Thomas’ jurisdictional argument against the incorporation of the clause under the 14th amendment on the grounds that it was only a federalism provision rather than a matter of securing substantive individual rights. ⁵⁵ As indicated, she also convincingly criticizes the late Justice Rhenquist’s doctrine of non-preferentialism, the theory that non-establishment precludes only preferential endorsement and aid to a particular religion but does not forbid the granting of aid to all religions. ⁵⁶ And she argues cogently against the reduction of the establishment clause to issues of coercion, an approach embraced by Justices Kennedy and Scalia. ⁵⁷ She approvingly notes Souter’s argument that the coercion theory of non-establishment would make the establishment clause a ‘nullity’ supplying nothing beyond free exercise. ⁵⁸ Since her main concern is to protect the equal liberty of conscience, capabilities and standing of minorities and minorities within minorities, she endorses, as we have seen, Justice O’Connor’s endorsement analysis interpreted along these lines. ⁵⁹

However, these are not the only motivations or principles behind non-establishment. Nussbaum’s focus on equality even in the radicalized form of liberty of conscience and capabilities development, is still too reductionist. For, as indicated above, there are democratic political principles and public purposes at stake in non-establishment that are irreducible to liberty of conscience, capabilities development, non-coercion, or free exercise. Equal liberty analyses screen these out because they concern the civil political community, the political and constitutional regime as a whole and the individual qua citizen rather than qua religious believer. But non-establishment is linked to democratic and civil republican political values insofar as it helps secure the civil (politically secular) democratic space for the pursuit of public purposes justifiable to religious, non-religious and anti-religious citizens. ⁶⁰ Precluding the united power of church and state on any level, the use of public funds for religious purposes, the endorsement of religious doctrines or symbols by the state, and state formulation or enforcement of religious law directly or via delegation to religious authorities, non-establishment as separation helps ensure that the democratic political process and laws will be public, inclusive and political rather than political-theological, exclusionary and sectarian. The non-establishment principle linked to separation in these respects, and the overall constitutional dualism of the American religion clauses, were and are correctly seen as mechanisms for de-dramatizing religious conflict by lowering its stakes – by taking public power, the identity of the state or nation, public law, authority and largesse off the table of competition among religious factions. Separation along these lines frees religious groups to concentrate on winning and keeping adherents thanks to their message and the convictions of their members while preventing their corruption of and by the state.

Non-establishment is thus more than an adjunct to free exercise and draws on principled democratic-political and republican reasons that equal liberty analyses cannot easily articulate. Accordingly, there can be no general right of exemption for religious groups from legitimate laws in a constitutional democracy, without impugning democratic legitimacy and the rule of law. Nor should immunities and privileges for corporate religious bodies treated as alternative sovereigns in the body politic, be granted on the dubious claim that this is the just way to accommodate religious minorities. The principle of non-establishment precludes the devolution of the democratic constitutional state’s authority to legislate or to enforce public or private, personal law, to religious groups (thereby creating multiple establishments). Ceding of state coercive law-making authority to a religious group typically reinforces the internal power of religious elites, makes exit and the internal rectification of inequalities harder, and is unlikely to enhance the freedoms of vulnerable or dissenting members within the group. ⁶¹ Nussbaum is well aware of all this. Nevertheless, she states that there is nothing in her approach that militates against a separate system of personal law so long as provisions are made to protect individual liberties and capabilities. ⁶²

Therein lies the flaw with the liberty-of-conscience approach. Apart from the enormous practical difficulties that oversight of separate legal codes would entail, a system of plural decentralized law-making would ultimately vitiate non-establishment as an independent constitutional principle. It would also be counterproductive: in the name of accommodating and integrating religious minorities into the polity, the proliferation of multiple establishments would invariably have the effect of casting the civil law, especially regarding marriage, reproduction and the family, as aligned to the religion of the majority. ⁶³ Indeed, typically multiple establishments go together with formal or de facto establishment of the majority’s religion. Segmental legal pluralism does not make religious minorities feel that the state is equally home to all its citizens. Instead it leads to them to devalue the democratic constitutional state, to cast doubt on political secularism generally, by reifying and politicizing cleavages along religious lines and by disintegrating a plural society rather than helping to integrate such a society. Religious legal pluralism may comport with a minimalist regime of twin tolerations but not one based on non-establishment and equal citizenship. Thus, an overly narrow reading of the principles and purposes of non-establishment for the sake of free exercise is ultimately self-defeating. As the American example shows, there are other ways to accommodate religious plurality and to protect religious minorities.

Liberal, democratic constitutionalism is a fragile historical achievement. It is based on the principle of equal citizenship, concern for the public affairs of the polity and the exercise of public power for public purposes. It erects institutional mechanisms such as the separation of powers, basic rights and freedoms, to ensure that public power is accountable, representative, responsive and inclusive, to the voice and interests of all. Political secularism and non-establishment are indispensable for such a project. These are hardly the ends or principles that structure the domain of religious organization and thus democracy demands that the exercise of public power including over personal law be kept out of the hands of those exercising religious power and it requires oversight of the latter to ensure compliance with basic constitutional principles.

It is now widely accepted that government is responsible for public welfare, ensuring the provision of social services like education, medical care and social security to all. We certainly should invoke equal liberty values radicalized to include capabilities equality and supplemented with critical-republican principles of non-domination, as criteria for state accommodation and regulation of religion when these are called for. The regulatory state may encourage private actors to supplement its provision of public services but if it indirectly funds or aids such provision, through tax exemptions or otherwise, this must be conditional on compliance with basic constitutional norms. Otherwise state aid would make the state complicit with what it aids, undermining its public purposes. Freedom of religion should not be a cover for oppression, privilege, or discrimination. These matters cannot be left to the vagaries of the political process. To secure individual rights within and not only for religious association, political secularism and separation must be secured constitutionally. But equality does not come easy: it requires intervention and regulation by the state if private powers violate it.

None of this means democratic states require congruence between religious associations’ internal structures, or norms and democratic organizational principles. ⁶⁴ As indicated, the two have different purposes and ends. Constitutional dualism based on the principles of non-establishment, freedom of conscience and religious exercise precludes both Erastrian co-optation of religion by the state and theocratic co-optation of the state by religious organizations. Once we acknowledge the importance of political secularism to liberal constitutional democracy in every epoch, and once we give up the old canard of strict separation, it should become clear that the appropriate response to the inevitable plurality of religious and of non-religious conscience must involve robust constitutional principles of non-establishment and free exercise. We can then focus on the day-to-day task of line-drawing.

Elsewhere I have argued that we should think in terms of a three-pronged regulatory approach. ⁶⁵ The self-regulation of religious communities (autonomy) is unproblematic so long as a right to exit exists and principles of equal citizenship and basic capabilities of the young and vulnerable co-religionists are not at stake. The constitutional principles of liberty of conscience, free exercise and freedom of religious association require respect for religious rituals and practices of minorities and majorities provided basic rights are not violated. Accommodation, moreover, can be required under a free exercise clause when it is necessary to realize these principles and to protect minority faiths against unfair and unnecessarily burdensome regulations even if they come in the shape of facially neutral laws. Direct top-down regulation and outright prohibition are necessary when practices labeled religious involve harm to members, especially the more vulnerable among them, or to outsiders, and when they contravene important public purposes. No separation is possible in such contexts. The democratic sovereign state has always regulated religion via its criminal and civil law, setting the parameters of the permissible. Thus the USA has long prohibited human sacrifice, child marriage, or polygamy, retaining its civil monopoly over publicly enforceable personal law, and it requires all children to be educated to meet state requirements. ⁶⁶ Any regulation can be democratically contested and learning is always possible. But religious communities may claim deference only on condition that an overriding public interest and the criteria of equal individual liberty and capabilities development are not violated. The third regulatory approach involves reflexive law: the regulation of self-regulation, a mode that is increasingly important in the contemporary context. ⁶⁷ Assuming the baseline of the welfare state, delegation of service provision to private actors must be subject to indirect regulation to ensure that they comport with constitutional principles. Indirect aid such as tax exemptions must be conditioned on such compliance. No corporate group that systematically discriminates against or denies equal liberty and capabilities development to vulnerable members in its internal hierarchies, employment decisions and so forth should receive tax exemptions, subsidies, or any other form of aid just because it is labeled religious. It should be under the same anti-discrimination rules as other providers aided by the state. Nor should these groups be delegated law-making power in the sense of legal pluralism. The regulation of self-regulation must not involve sharing the coercive power or authority of the state with religious groups. But neither may it abolish their autonomy to live under their religious norms within certain limits. None of this means individuals would lose their freedom of religious association, their liberty of conscience, or their free exercise rights. But they cannot have it both ways: with state aid comes regulation to ensure that public power does not lend authority to institutions that violate public purposes.

The Americans got it right when they constitutionalized two religion clauses securing free exercise of religious conscience and prohibiting the establishment of religion. Their response to religious pluralism was to create a constitutional democracy whose public power, authority and largesse are off limits to the religious groups proliferating in society and are controlled by none of them. This is one of the greatest innovations in the American constitutional experiment. The American model of political secularism predicated on a robust constitutionalized non-establishment principle, the flip-side of strong protection for the freedom of religious conscience, has long been an exception to the typical ways of constructing state–religion relations. Mistakenly seen as a strict separation model predicated on privatization of religion, the constitutional dualism of the American approach has not been imitated elsewhere. But robust non-establishment principles remain a sine qua non to secure liberal constitutional democracy as well as real liberty of conscience within, for and outside religious associations. In a 21^st-century world in which religion is both flourishing and has become highly politicized, and in which constitution-making and reform are an important political stake in many national and transnational contexts, the core principles of non-establishment, liberty of conscience and free exercise, should be reinforced and defended everywhere. The American version of constitutional dualism should perhaps become a model rather than an outlier.