CHAPTER 6

RELIGIOUS EXEMPTIONS FROM SECULAR PUBLIC POLICIES

THE AMERICAN POLITY IS a mighty engine of public policies generated by all levels of government, policies that affect us all in myriad ways. Each of these policies is usually framed in general terms so the policy will apply equally to everyone subject to its terms. This goal of equal treatment under the law, however, is vastly complicated by a singular fact about America: It is the most diverse Western society on earth—regardless of how one defines diversity, or what kinds of diversity are under discussion.1

Among the many aspects of American diversity, perhaps the most distinctive (compared with other modern democracies) is our rich array of religious beliefs and practices.² These beliefs and practices are deeply valued not only by their adherents but also by the many more secular Americans who believe that religious values are good for our society.* Unlike most other types of diversity, religious belief and practice are both constitutionally protected. This protection is conferred by two parts of the First Amendment—the Free Exercise Clause and the Establishment Clause—that are in significant conflict or tension with each other, especially with regard to accommodation decisions.* (The Fourteenth Amendment’s Equal Protection Clause also applies, but to a lesser extent.) This tension sometimes occurs when the government’s very efforts to accommodate a religious practice (e.g., when it funds chaplains or exempts religious groups from taxes, land use controls, equal employment, and other generally applicable laws) might be deemed excessive government entanglement in the religious sphere under the applicable legal tests for prohibited establishments.

The Supreme Court has elaborated certain doctrines designed to bring the constitutional values enshrined in the two clauses into a proper balance with other constitutional values and social interests, almost all of which are essentially secular, sometimes explicitly so (e.g., the Constitution’s ban on any religious test for office³). These balancing doctrines seek to preserve the conflicting values and interests while also protecting precious rule-of-law norms: coherence, consistency, predictability, and even-handedness. This delicate balancing act presents a “hard issue,” in this book’s parlance. Few experts in the field think that the Court has gotten it right. Many believe that it should defer more to the balance that legislatures have already struck in the statute or rule under challenge. One thing seems abundantly clear: The combination of a more expansive regulatory state, and more diverse views about which behaviors are appropriate or offensive, is engendering more—and more complex—social and legal conflicts.

The chapter consists of five parts. The first presents the larger context in which the issue arises—specifically, our religious diversity and rights culture. The second part explores three key concepts—separation, accommodation, and neutrality—that the Court and other discussants use when analyzing accommodation claims. Next, I discuss judicial methodology, focusing on two legal doctrines—the “standard of review” and the test for “establishment”—that the Court uses to resolve such disputes. (I do so as briefly and non-technically as possible.)

The next part, “Emerging Accommodation Issues,” has three sections. The first analyzes the Court’s controversial 2014 decision in the Hobby Lobby case holding that certain closely-held, for-profit employers could claim religious exceptions (or exemptions; I use the terms interchangeably) from generally applicable laws—there, the Affordable Care Act mandate that employer-provided insurance cover contraceptives.4 The second section discusses the proliferating demands for exceptions in the wake of the Court’s 2015 Obergefell decision upholding same-sex marriage.⁵ Here, sexual and gender minorities’ claims to equal treatment clash with claims by those who refuse to provide services to these minorities, citing religious objections and Hobby Lobby. And most recently, civil rights disputes have arisen over which bathrooms and locker rooms transgender people may use, disputes that presage many others. Finally, I discuss how we might best accommodate the conflicting interests going forward.

CONTEXT

Two fundamental features of American life intensify the conflicts between religion and secular public policies (and between one religion and others). The first is its religious diversity (as well as its growing secularism). The second is our distinctive rights culture—mostly legal but also informal—which complicates such conflicts by pressing them into adversarial and uncompromising forms.

Religious Diversity*

America is an anomaly among modern postindustrial societies: It is more religious, its religiosity is more diverse, and its separation of state and religion is more rigorous. This religiosity, diversity, and separation—which contrast sharply with the European experience on all three points—did not come immediately or naturally. The zealous Puritans who founded American society were soon reduced to a minority even in the Plymouth colony, and their descendants engaged in a constant and mostly unsuccessful struggle with secularism, indifference, and declining church attendance. The only doctrine they valued was their own—or, as the nineteenth-century minister Henry Ward Beecher put it, “Orthodoxy is my doxy and heterodoxy is your doxy, if your doxy is not like my doxy.” Many colonists, themselves anti-establishment dissenters, nonetheless did not shrink from persecuting, expelling, even killing those whose worship differed from theirs—an intolerance widely practiced in Massachusetts, Maryland, Virginia, New Amsterdam, and other colonies.

What brought religious diversity to America was not so much toleration as immigration, which the colonies desired primarily for its economic value to thinly settled frontier communities and their entrepreneurial merchants and real estate speculators. Toleration for the strange beliefs and practices of other Christians, not to speak of non-Christians and nonbelievers, came only slowly and even then, incompletely. Americans, literally repelled by the religious wars (which actually were about much more than religion⁶) that had butchered and benighted so many of their ancestors in Europe, gradually found their way to the live-and-let-live acceptance of heterodox beliefs and disparate practices that mark our constitutional tradition—and, today, our dominant social ethos. The dramatic declines in anti-Catholicism and anti-Semitism during the last half of the twentieth century were important signs of this attitudinal change. If any more proof of the new ethos were desired, one need only look to President Bush’s speech to the nation only two days after the September 11 terrorist attacks, praising Arab Americans and Islam while declaring war on their violent co-religionists. President Obama also celebrated Islam in his Cairo speech, not long after his inauguration. Yet intolerance (occasionally, even violence) toward atheists, agnostics, Muslims, and other non-Christians still intrudes into the public square at times—perhaps never more so than in Donald Trump’s presidential campaign. At the same time, however, public opinion polls find that attitudes toward Muslims are more positive than ever, even after incidents like the June 2016 mass shooting by an Islamic terrorist in Orlando.7

America’s religious diversity also reflects the fierce, fractious ethos of its Protestantism, an ethos refined through almost four centuries of sectarian struggle as well as by American society’s distinctive racial, class, ethnic, and regional cleavages. This history has precipitated a vast number of sects and denominations (a term loosely used here to mean large, confessionally-related groups). This heterodox pattern was apparent as early as the seventeenth century with the migration of Huguenots, Quakers, Mennonites, and other dissenting groups to the American colonies. Nor is this fragmentation diminishing. During the twentieth century, the Pentecostal movement produced more than three hundred distinct denominations, and a significant dissident faction has recently arisen within the Southern Baptist Convention, by far the largest Protestant group. Indeed, further fragmentation seems likely as the forces of secularism, pluralism, and bureaucratization transform the nature, function, and relationships of religious groups. Almost fifty years ago, Peter Berger described them as “marketing agencies” working in a “system of free competition very similar to that of laissez-faire capitalism,” one in which “[r]eligious contents become subjects of ‘fashion.’ ” On the supply side, denominations seek more distinctive niches or “brand identities” with which to hold and attract members. On the demand side, people in what Wade Clark Roof calls the “questing culture” pursue a bewildering variety of customized mixes of meditation, therapy, counseling, self-help, behavior modification, fellowship, and a search for “spirituality without religion.” The power of this competitive, individuating impulse helps to explain why so little has come of formal ecumenical efforts among American churches, most of which have involved restoring relations broken by the Civil War more than a century earlier.

These centrifugal forces extend far beyond Protestantism. American Judaism, for example, long ago split into four major streams—and a fifth, “humanistic Judaism,” is now seeking recognition. Some of these, like Reconstructionism, have no real counterparts in other countries, while others such as Orthodoxy are divided into so-called black hat, Lubavitcher, modern, and many other sects led by particular rabbis, with the greatest growth occurring in the more conservative sects. As this example suggests, the fragmentation is by no means confined to liberal, congregational-oriented, and decentralized groups; it also extends to the more hierarchical Catholic and Orthodox Christian churches and especially to the evangelical and Pentecostal sects, which are the fastest growing of all in the United States and are rapidly expanding overseas. Indeed, the largest gains in the evangelical sector have occurred in nondenominational groups, which have tripled as a percentage of the total since 1965. The degree of fragmentation is further underscored by the fact that once we set aside a few large denominations—Catholics, Baptists, and Methodists—no single religious group exceeds 5 percent of the population.

Even within the most hierarchical of American religions, diverse views abound. Today’s Catholic Church has been described, perhaps hyperbolically, as “a federation of internally divided quasi denominations.… [P]eople who identify as Catholic are more liberal on sexual morals than Protestants as a whole. Birth rates and opinions on abortions are virtually the same. Like Protestantism, the American Catholic Church today seems to be many denominations, loosely united.”8 Thus, Catholics split fairly evenly between “traditional” and “progressive” views on basic issues of faith, and much the same is true among other putatively conservative groups:

For example … one-third of committed evangelicals and 41% of committed Catholics believe that legal abortions should be available to women in at least some circumstances other than rape, incest, or to save the life of the mother, and well over half of the members of these two groups support the distribution of birth control information in public schools. Thirty-three percent of committed evangelicals—and nearly as high a percentage of Mormons—believe that government regulation is necessary for the public good, and 28 percent of both groups think that the federal government does a better job than it is often given credit for. Similarly, about one-fourth of committed black Protestants believe that the government cannot afford to do more for the needy, and almost one-third feel that African Americans are largely responsible for their own economic circumstances.9

American Protestantism’s intra-denominational divisions over public policies, then, are greater than their inter-denominational ones over religious practices and doctrines.

The religious practices and views within most religious groups in America are far more heterogeneous—liturgically, doctrinally, organizationally, demographically, and in their geographic distributions—than are the same groups elsewhere in the world, and this diversification is accelerating. For each church that amalgamates with others today, many more are born afresh or separate from their parent congregations. An estimated 1,600 religions and denominations exist in the United States, a far cry from the “three-religion country” proclaimed by Will Herberg in the 1950s. About half of them were founded since 1965. Indeed, as Richard Ostling points out, each of two dozen denominations has as many local congregations today as made up the whole of American religion in 1776. Here, as elsewhere in American life, technological and market forces have played their part in spawning diverse forms of worship—tele-churches, mega-churches, and pastoral teachings on the Internet.

Reflecting international trends, the mainline Protestant denominations—predominantly white, relatively affluent, ecumenical, affiliated with the National Council of Churches, and leaders in college and seminary education—have experienced marked decline since the 1960s. The American Episcopal Church, for example, has lost more than a quarter of its members in the last thirty years. Some religions and denominations, of course, have lost more than others. Catholicism is maintaining its numbers, aided by immigration; about 3 percent of its seminarians are Vietnamese. In contrast, Judaism’s share declined during the 1990s and that of the Baptists and Methodists did not change. At the same time, the shares of fundamentalist, evangelical, Pentecostal, and charismatic sects (especially Mormons, the Assemblies of God, and the Church of God in Christ) have grown dramatically; ecstatic Christianity is resurgent. About 10 percent of adults raised in another tradition now identify as evangelical Protestants.

Christianity, if increasingly fragmented and prone to denominational shifting, still utterly dominates Americans’ religious affiliations. More Christians live here than in any other country in the world; Christianity predominates in virtually every U.S. county. Roughly seven in ten Americans identified with some Christian denomination in 2014. Protestants made up 46.5 percent of those identifying as Christian, while Catholics made up 23.9 percent. This large share of Christian identifiers, however, had declined by 8 percent since 2007. This is partly because religions other than Christianity have grown rapidly. Approximately one million Hindus live in the United States, compared with only 70,000 in 1977; most of the world’s diverse Hindu traditions are practiced here, as are the Sikh, Jain, and other offshoots of Hinduism. The Muslim population is now roughly 3 million. Like Protestants, they are divided by ethnicity, race, and language; one-third are from South and East Asia, 30 percent are African American, and 25 percent are Arabs, with significant political tensions existing within and among these groups. There soon will be more Muslims in the United States than in the small Arab countries—although Islamic State attacks in Europe in 2015 will make it harder to resettle Muslim refugees here.10 As with Hinduism, all branches of Islamic worship exist in the United States. More than 4 million practicing Buddhists live here, representing the full range of Asian Buddhist traditions, along with many indigenous American ones. Only twenty-one Buddhist study centers existed here from 1900 to 1964, but more than a thousand exist today; in 2001, American-born Buddhists consecrated an immense stupa in a Rocky Mountain valley. There are many Zoroastrian temples here, and more Baha’i live in the United States than in Iran. Even pagan religions seem to be flourishing; in 2000, the Pagan Educational Network estimated 600,000 practitioners here.

But even this does not quite capture the religious diversity in America because it fails to capture the steady devolution of worshipful energy, liturgical innovation, and even governing authority from national religious organizations to their regional and local units. (Secular organizations have also experienced such devolution.) Even the Jesuits, a relatively disciplined and hierarchical Catholic group, have come under the spell of these fragmenting, individualizing conditions of modern American religious life. This devolution of even highly centralized religious authority rests on practical, structural, and ideological considerations. The Episcopal Diocese of New York, for example, conducts its Sunday worship in fourteen languages, and Catholics, who are increasingly Spanish- and Asian-language speakers, come from twenty-three different countries of origin. Another devolutionary factor is Americans’ traditional personalization of their religious practices, whether as members of organized denominations or not—a remarkable phenomenon famously analyzed by philosopher William James more than a century ago in The Varieties of Religious Experience.

The United States is unique in its vast proliferation of sects, which usually break away from more mainstream denominations, and “new religions” or cults, usually centered around charismatic leaders. In Alan Wolfe’s study of middle-class suburbanites, he found that “this strong strain of individualism … helps explain why, as religious as Americans are, they also distrust organized religion: in 1990, as few as 23 percent of the American population expressed a great deal of confidence in religious institutions.… Americans would be more comfortable living next to blacks than to religious sectarians.” Even traditional sectarianism cannot wholly satisfy this radical religious individualism, which Wolfe sees as distinctively privatistic, voluntaristic, nondogmatic, and separate from organized religion.

Indeed, sociologists have argued that all religions in America, including those with strong hierarchical organizations, become de facto congregationalist in form and practice. Even the Catholic Church, which for centuries resisted devolving power over diocesan decisions, may have to cede more authority to parish and lay groups in the United States in the wake of the pedophilia scandals covered up by bishops and other central authorities. In Protestant congregations, a liberal-conservative divide often forms, and “special-purpose groups,” which reflect members’ diverse interests but also create more divisions, abound. Even more fragmented are the apparently growing number of “home churches” where a single family, sometimes joined by a few others, constitutes its own unique congregation and liturgy. This chronic dissatisfaction with religious institutions bears a striking resemblance to the populist, anti-hierarchical, maverick, questing, and competitive impulses of Americans in other spheres. Their mistrust of mainline religions probably springs from many of the same social, psychological, and ideological sources as their suspicion of large corporations, political parties, professional expertise, and government.

Americans do profess strong religious convictions. Fully 96 percent of them say that they believe in God—about the same as fifty years ago—with 90 percent believing in heaven, 65 percent in the Devil, and 75 percent in angels that affect human affairs. More than 40 percent of Americans have been telling pollsters for six decades that they attend religious services each week. Some observers doubt the accuracy of this figure, but its consistency over time is striking. In any event, church attendance is higher, often much higher, than in any industrialized country other than Ireland and Poland, as are other indicia of religiosity. The United States is the only advanced Western industrial society with a strong fundamentalist movement. In 1996, religion accounted for 16 percent of all TV programming in the United States and appeared on 257 stations, up from 1 percent and nine stations in 1974. Sales of “Christ-honoring products” had quadrupled to $4 billion since 1980. Even back in 2001—before the vast growth in Internet use, 25 percent of adult users, about 28 million people, had gone online to find religious and spiritual material—more than the number who had visited gambling sites, participated in online auctions or traded stocks online, and a sharp increase from the number in 2000.

All this religious fervor among Americans makes their theological ignorance all the more remarkable. They appear to know little about basic religious ideas and facts. Although 93 percent of their homes contain at least one Bible and a third claim to read it at least once a week, 54 percent cannot name the authors of the Gospels, 63 percent do not know what a Gospel is, 58 percent cannot name five of the Ten Commandments, and 10 percent think Joan of Arc was Noah’s wife! Indeed, a recent survey found an astonishing number of born-again Christians whose views seem to flatly contradict the Bible.

The number of atheists, agnostics, and “no preference” (or “nones”) in the United States seems to be rising.* According to Pew, those who self-identify as such have almost doubled since 1965, and a large, growing minority of younger Americans call themselves “spiritual” rather than religious. Pew finds that even among the religious, a large drop has occurred in traditional religious beliefs, practices, and commitment. For example, only 46 percent of U.S. Catholics said in 1993 that they regularly attended mass, down from 74 percent in 1958. This marked decline is occurring even among white evangelical Christians.¹¹

These changes are strikingly generational. As Millennials enter adulthood, they become less religiously affiliated than older generations. Nearly a quarter of Generation Xers now claim to have no particular religion or describe themselves as atheists or agnostics, up four points in seven years. Baby boomers also are now more likely to identify as “nones.” Religious intermarriage is on the rise: Among those who married since 2010, nearly four in ten are in mixed marriages; for pre-1960 marriages, the share was 19 percent.12 Indeed, Americans seem more tolerant of other religions than of atheism.* Pew found that 49 percent would be “unhappy if a family member were to marry someone who doesn’t believe in God,” whereas only about 15 percent would be unhappy having someone from the other party join their family.¹³

Americans increasingly support bringing religious voices into the political sphere. About half (49 percent) of those polled in 2014 supported having houses of worship express views on political and social questions (up from 40 percent in 2012), and 32 percent felt that religious leaders should endorse political candidates (up from 24 percent in 2010). Perhaps they support more religion in politics because so many fear that religion’s influence is diminishing; 72 percent believe this—and they may be right. Indeed, America’s great cultural divide is not among different religions. It is between believers and nonbelievers.

Whether all of these social facts amount to paradox, a deeper consistency, poor survey techniques, or arrant hypocrisy is uncertain. But Americans clearly profess respect for the religious diversity around them,† as Wolfe finds:

When we consider how many people have died in the name of religion over the years, the acceptance of so many different kinds of belief in America is remarkable. One is tempted to call it real diversity, not because the idea of diversity in inappropriate to race, gender, or sexual orientation, but because religion claims to speak to what really and truly matters in life. I confess that at some level I did not fully understand the non-judgmentalism of middle-class Americans.… [but] what comes along with it [is] a strong commitment to the principle that a wide variety of religious views ought to be allowed to flourish.

Protecting religious diversity in American life might seem straightforward. Historically, Steven Smith notes, “the ferment that caused religious diversity to flourish—and that is largely responsible for the condition of religious freedom we enjoy today—was a product of pluralism; it owed little or nothing to judicial review, or to the legal elaboration and enforcement of any constitutional ‘principle of religious freedom.’ ”14 Religious minorities, many of them recent immigrants with traditions (such as Santeria¹⁵) that seem exotic to Americans, now have vast breathing space where they can cultivate their distinctive beliefs and practices. We saw that Catholicism, our largest single denomination, accounts for a declining share of the population. And other conditions also favor religious pluralism: the religion clauses, dominant social norms, and officials’ responsiveness to even small groups of deep religious faith.

Rights Culture

For centuries, American and foreign commentators have noted that American culture (not just law) is more steeped in the notion of individual rights than perhaps any other. In the United States, individuals assert these rights strongly not only against government but also against each other. Law professor Mary Ann Glendon traced the rigidities, hyper-individualism, and other consequences of this “rights talk.”¹⁶ Several points about this rights culture are important for present purposes.

•  The proliferation and assertion of religion-based rights by A necessarily reduces the rights-free space in which B, C, D … can exercise their freedom to act without affecting A’s rights. Unless rights-holders are willing to redefine their rights claims to avoid a clash (as through accommodation, discussed below), conflict is inevitable.

•  Where the conflict concerns religion-based rights—those thought necessary to worship God and protect a religious community against outside influences—rights-holders are likely to be most determined to exercise them in the most uncompromising fashion.

•  The assertion of these claims occurs within a system of what lawyer-sociologist Robert Kagan calls “adversarial legalism.”17 This system has many rule-of-law virtues, as Kagan explains, but its costs are also many. It encourages a rigid, aggressive, legalistic approach to resolving social disputes. This means, among other things, conflict resolution that is costly, protracted, contentious, party-dominated (in practice, lawyer-dominated), technical and procedural, winner-take-all and thus crushing to the loser. It is also a poor way to make complex public policy, as I have explained elsewhere.¹⁸ (Some disputes over religious rights are narrow and localized, but others—for example, the Hobby Lobby and Obergefell cases discussed below—have far broader implications affecting many people beyond the litigants and the nature of important public programs.)

•  In the cultural, legal, and political wars, a common argument is that those who seek an exemption from a general requirement for one or another religious practice are claiming special rights, placing them “above the law” by denying the rights of others who oppose the exemption. What is and is not lawful, of course, depends on the legal principles that govern a specific conflict. Here, the point is that such an argument resonates strongly with popular demands for legal equality. I analyze this question later in the chapter.

•  Our constitutional structure—particularly separation of powers and federalism—makes cooperation and conciliation essential to government’s effective functioning.¹⁹ The rights culture, however, tends to cut against this, impeding both political and social capital formation. Although conflict, like the diversity that inevitably causes it, advances social learning, too much of it can generate dangerous levels of gridlock, mistrust, despair, and even violence.20

THREE KEY CONCEPTS: SEPARATION, ACCOMMODATION, AND NEUTRALITY*

We can best map this contested terrain by exploring what divides the disputants. The polar positions are “separation” and “accommodation.” Each of these positions, however, ultimately depends on how each defines and applies the idea of “neutrality” (discussed in the next section), which the Court requires the government to observe when it regulates religion. For example, Supreme Court justices William Brennan and Thurgood Marshall were strong separationists in Establishment Clause cases but vigorously favored accommodation in Free Exercise cases.

Separationists tend to think that religion is a potentially divisive force when brought into public life; they wish to consign it to the private realm where, like other forms of expression and activity, law should protect it on the same neutral terms that liberalism applies to these other forms. Invoking Jefferson’s famous metaphor, which the Court adopted only as recently as 1947, separationists believe that both religion and politics gain when the law erects between them “a wall … which must be kept high and impregnable.” They also draw on Jefferson’s equally famous distinction between belief, which the government cannot regulate, and conduct, which in this view it can sometimes regulate. In 1990, the Court affirmed this distinction in Employment Division, Department of Human Resources of Oregon v. Smith after having criticized it in the Wisconsin v. Yoder and Sherbert v. Verner cases. These three decisions are discussed below.

Acommodationists believe that religion generates enormous public and private benefits (a claim many separationists concede—up to a point) but insist that it can only retain its meaning-creating, value-conferring force and integrity if religious persons have leeway in which to pursue distinctive beliefs and practices. Nevertheless, accommodation in the religious realm is more elusive than in other areas of public policy, especially where religious claims conflict with equally strong claims grounded in largely secular moralities for which universalizing rights-talk seems appropriate—for example, abortion (the “right to choose” versus the “right to life”), and equal treatment (the right to avoid condoning conduct that a religion considers sinful versus the right to equal treatment). In such cases, both sides may view compromise as an accommodation with evil. (Think of William Lloyd Garrison’s denunciation of the 1850 Compromise as a “pact with the Devil.”) Even short of that extreme position, accommodation will often seem to sacrifice important values—which sometimes may indeed be true.

The law, accommodationists say, should create a strong presumption favoring even those practices that seem illiberal or in some cases are illegal—although just how strong that presumption should be is a question that, as we shall see, divides Congress and has set it against the Court. Separationists reply that such accommodations do not just tolerate religion but “establish” it in violation of the Constitution—that accommodation is not neutral as between religion and nonreligion or perhaps even as among religions. So what does neutrality mean?

Neutrality. Americans value two kinds of neutrality: neutrality (1) among religions, and (2) between the religious values of a largely pious people and the secular values of a liberal civic culture. But this is still too simple. Stanley Fish observes that any effort to define neutrality without appealing to some non-neutral value is “mission impossible” because one can only decide hard cases by appealing to a non-neutral principle. (Of course, the Court will gussy up the decisive principle so that it appears neutral.)21 In speaking of neutrality, then, I take Fish’s point: Some understandings of neutrality are more plausible and persuasive than others in specific cases, but they are not neutral in a more universal or rigorous sense. Neutrality’s meaning also depends on how one frames the particular comparisons in the neutrality analysis—an instance of a more general difficulty encountered when we reason by analogy. Judgment, not logic, determines which features of the supposedly analogous cases should be regarded as salient to the comparison.

Neutrality, then, is a norm whose meaning and application to actual conflicts can be highly contestable.* Just to get the discussion rolling, I define neutrality (following Professor Douglas Laycock) in substantive terms: The government should not use burdens or benefits to influence people’s religious choices.²² Legally, this has come to mean that a law must accommodate any religious practice or claim that does not threaten “compelling governmental interests.” (The Religious Freedom Restoration Act [RFRA]²³ and its state-level counterparts essentially adopt this standard, as we shall see.) This legal definition underscores the public value of preserving a broad private sphere of religious autonomy even when a religious practice may offend general public policies and civic sensibilities. In this view of neutrality, government may not coerce a religious practice. Neither may it sponsor or endorse the practice, favor one religion over another, or favor religion over non-religion (or vice versa).† But in choosing among competing standards of judicial review (analyzed in the next section), the Court must put its thumb on one side of the scale or the other. By holding that a government interest can carry the day even if it is less than “compelling,” the Court could make it harder for a challenger to demand that the government grant a religious accommodation.

Separationists and accommodationists have conceptual, empirical, normative, and political differences about neutrality. To begin with, separationists and accommodationists disagree about what neutrality means, whether it is really possible in religiously diverse America, and if so whether it is a defensible public value or legal standard in this context. Because these three issues are not always distinguished, separationists like political theorist Stephen Macedo, a liberal defender of civic values against the inroads of religion, and accommodationists like law professor Stephen Carter, a religious conservative who wants to broaden religion’s role in public life, can end up agreeing on a number of points even though they present their positions as strongly antagonistic. In what follows, I use their work to sharpen and illuminate the debate.

Their first commonality is that, as Carter approvingly puts it, “religion has no sphere. It possesses no natural bounds. It is not amenable to being pent up. It sneaks through cracks, creeps through half-open doors … and it flows over walls.… Rushing past boundaries is what religion does.” Macedo acknowledges this leakage but decries it, warning that it poses a mortal danger to the ideal of civic liberalism. They also agree that state neutrality—basing government policy solely on secular ideas without regard to its effects on religion—is illusory. Religion is inevitably political in that it helps shape our public and private values. To Carter, “[n]eutrality is a theory about freedom of religion in a world that does not and cannot actually exist” because the state cannot act without taking account of religion. A pretense of neutrality, Carter argues, is used to override deep convictions for little state gain, as when the Supreme Court invoked neutrality to uphold military discipline against an Orthodox Jew for wearing a yarmulke while in uniform. Macedo too doubts the possibility of strict neutrality but wants to contain this leakage lest it contaminate both sides, especially the liberal polity, which needs inoculation against certain sect-values. The state, he argues, must promote civic virtues, not comprehensive ideals of the good life, but even this civic promotion affects religious beliefs and practices in nonneutral ways.

Carter also claims that the neutrality norm favors big, influential religions over small, defenseless ones. Interestingly, separationists say the same about accommodations, although some Free Exercise decisions granting accommodations to small sects seem to refute this theory.* Carter’s larger hostility claim, however, is very weak. Traditionally, many federal and state laws gave special consideration to religions in general and small sects in particular. Smith denied an accommodation, but the RFRA statute was enacted to overrule such denials. It is true that courts tend to be less accommodating than legislatures, almost always upholding the government even under the “compelling interest” test that Smith rejected but the REFRA restored. This is discussed later in the chapter.

Most separationists like Macedo agree with accommodationists like Carter on religion’s virtues. Indeed, Macedo maintains that the separatist project “depends upon the support of religious reasons and religious communities—a support that can be encouraged by a liberal public philosophy but not altogether justified by it.” Liberalism, he says, depends on the reasons, norms, and moral convictions generated by religious communities, and also on the political and moral education that such communities provide. Even Catholicism, whose earlier illiberalism he strongly condemns, has strengthened the American polity, in his view. Its natural law doctrine checks the moral excesses of democratic majoritarianism; its “subsidiarity” principle supports devolution of power; and it rightly insists that human values transcend the political, justifying why one should not invest politics with all one’s moral energies.

Beyond these convergences, however, separationists and accommodationists disagree sharply. To someone like Carter, the state that Macedo wants to protect poses a far greater threat to religion than religion does to the state, and the state’s putatively civic projects, which Macedo wants to promote, are in fact pervaded by assumptions of value and fact that can be hostile to religion. To Carter, these assumptions cohere to constitute a comprehensive, secular worldview that not only competes with religion in defining the meaning of life but remorselessly deploys its monopoly of coercive power, together with a conviction of the state’s superior rectitude, to establish and maintain its dominance. To Macedo, in contrast, this competition is disciplined by the ground rules of separation, and the hegemony of an activist state is required to sustain his vision of “liberalism with spine,” which goes beyond promoting tolerance, freedom, order, and prosperity. It must also secure “the preconditions of active citizenship,” including the state’s “educative” interests in citizens’ character, in order to pursue society’s collective ends. Some religions, notably fundamentalist groups that insist on subordinating the state’s educational and moral authority to that of the group and family, are anathema to this vision and must be overridden in the name of civic liberalism. Which values civic liberalism includes, of course, is itself a hotly disputed issue.

I am tempted to suggest that despite these sharp rhetorical conflicts, and except at the extremes, most separationists and accommodationists differ only in degree. Militants of both stripes doubtless will reject this pacific suggestion. Unlike Macedo, for example, many separationists categorically oppose use of vouchers in religious schools; unlike Carter, many accommodationists demand school prayer. Nor do I deny that such differences can produce quite disparate views about particular cases or policy disputes, as with school choice proposals. I suggest, rather, that in the end the precise location of the lines that the law must draw depends on principles that come down to matters of degree, of more or less.

Carter, for example, concedes that there are limits “beyond which no claim of religious freedom will be recognized,” citing the case of religiously-mandated murder. One could cite more difficult cases like church-sanctioned child and spousal abuse, racial segregation, and female genital mutilation, but the point would be the same: Lines must be drawn and someone with temporal authority must draw them. But having acknowledged this, Carter then complains that it is the courts that eventually will do this, and that they inevitably “center their concern on the needs of the state, not the needs of the religionist.” But Congress and the states, as we have seen, often do precisely the opposite. Courts, moreover, must review these legislative choices. The real issue for the courts, then, is the legal standard for reviewing them.

And here is where the casus belli between separationists and accommodationists dissolves into a difference of degree, albeit a consequential one, involving the standard of judicial review in religion cases discussed in some detail in the next section. Both Macedo and Carter support tests under which courts draw lines by identifying, weighing, and balancing competing interests much as legislatures do. They differ only over whether the state’s interest must be “compelling” or merely “important” and whether the burden must be “the least restrictive means” or merely “reasonable.” Outcomes in particular cases may turn on these differences.24

Diverse religious beliefs “are as infinite as the imagination,” and once the genies of opt-outs and exceptions to general rules (or standard school curricula) are out of the bottle, they cannot easily be confined. Among some academics, this was a leading argument against the RFRA in which Congress resurrected the more religion-accommodating “compelling interest” standard that Smith had jettisoned. (Congress had no such qualms; the RFRA passed almost unanimously!) The Court’s concern about the RFRA reflects a similar and legitimate anxiety about line-drawing in an area fraught with constitutional constraints imposed by the religion clauses and federalism’s vertical division of power. These constraints are the subject to which I now turn.

JUDICIAL METHODOLOGY: STANDARD OF JUDICIAL REVIEW, AND THE TESTS FOR “ESTABLISHMENT”*

The constitutional setting for religious accommodation issues consists of the Establishment and Free Exercise clauses of the First Amendment: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” Two judicial doctrines, developed over the last ninety years or so, constitute what we might think of as “meta-doctrines” that frame the way the Court then goes on to decide specific religious accommodation disputes. These meta-doctrines are: (1) the standard of judicial review used to decide Free Exercise Clause cases, and (2) the tests used to decide Establishment Clause cases.† My brief explanation of them sets the stage for the next section of this chapter, where I shall examine how the Court has resolved specific disputes under these clauses in various areas of social conflict.

Standard of Judicial Review. In deciding cases under the Religion Clauses (and many other constitutional provisions), courts must decide how much deference they will accord to the legislative (and administrative) judgments embodied in the governmental rule or practice that is being challenged as unconstitutional. Depending on the kind of constitutional claim that is being asserted against the government, the Court has applied one of three standards of review. The default standard is called “rationality” review. The most lenient or relaxed (pro-government) standard, it requires only that the governmental action be reasonably related to a legitimate government objective. Age-based claims, for example, are reviewed under this standard. The second standard is “intermediate scrutiny,” under which the law being challenged must further an “important” governmental interest by means “substantially related” to that interest. Examples include most gender classifications and undocumented immigrant children’s right to attend public schools.25 The most stringent (pro-claimant) standard of review is “strict scrutiny”; here, the government’s action must further a “compelling” governmental interest and do so in a way that is least restrictive of the right that the claimant asserts. (This is called “narrow tailoring.”) Strict scrutiny review applies to government actions that impinge on a “fundamental right” (religion, for example) or that treat people differently based on certain “suspect classifications,” of which the main ones are race, national origin, and—central to this discussion—religion.

In religion cases, the Court has developed its standard of review through a series of cases involving “exotic” or minority religious practices that believers, invoking the Free Exercise Clause, defended against criminal prosecutions. In Reynolds v. United States, decided in 1878, Mormons challenged a federal statute making polygamy a crime. In this, its first application of the Clause, the Court did not spend much time worrying about the standard of review in upholding the statute. It sufficed that polygamy had always been treated as “an offence against society.… [I]t is impossible to believe that the constitutional guaranty of religious freedom was intended to prohibit legislation in respect to this most important feature of social life.”²⁶

The modern evolution of the standard of review in religion cases begins with the 1963 decision in Sherbert v. Verner,²⁷ involving a state’s denial of unemployment benefits—there, to a Seventh Day Adventist who refused to work on Saturday. In Sherbert, the Court required the accommodation; denying it would “substantially burden” her religious freedom and the state had no strong reason not to grant it. This “Sherbert test” is essentially a strict scrutiny standard of review.*

The next evolutionary step came in 1990 with a decision discussed earlier: Employment Division, Department of Human Resources of Oregon v. Smith. There, the Court upheld a state law denying unemployment benefits to a man fired for using peyote as part of a religious ceremony, which constituted “misconduct” under the state’s illegal drug law. Because the law was neutral, generally applicable, and not targeted at any particular religion, the Court upheld it, seeing no threat to free exercise in the denial of an exemption for peyote use, and thus no need for heightened review. For such a law, Smith held, the state need not grant a religious exception.

Religious groups strongly denounced this “Smith test” for allowing the state too much leeway to punish believers. They persuaded Congress not only to pass laws granting religious exceptions but also to enact a more general reform, the RFRA, in 1993. This statute mostly rejected the Smith test and revived the Sherbert test—even as applied to neutral, generally applicable rules. The RFRA formulated this test as follows: “Government shall not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability, except as provided in subsection (b) of this section.” Subsection (b) provided that “Government may substantially burden a person’s exercise of religion only if it demonstrates that application of the burden to the person (1) is in furtherance of a compelling governmental interest; and (2) is the least restrictive means of furthering that compelling governmental interest.” This is sometimes called an “enhanced Sherbert test”; subsection (b) (2) replaced the “narrowly tailored” element of the test with an arguably more demanding, proexception test of the “least restrictive means.”

But even the RFRA—crafted by a Congress that strongly (indeed, almost unanimously) rejected Smith—was not to be the final word. Four years later, in 1997, the Court in City of Boerne v. Flores28 struck down the RFRA as applied to state laws. This delighted those, like law professor Marci Hamilton, who had condemned the RFRA for giving religion special privileges while countenancing genuine harms to others.²⁹ Hamilton argued this before the Supreme Court, which held that by legally basing the RFRA on its Fourteenth Amendment remedial powers against the states, Congress had in effect redefined a substantive constitutional right—something that only a court or a constitutional amendment could do. Accordingly, the RFRA applies only to federal laws; the states are bound by the Religion Clauses as interpreted by the Court, not by the RFRA. As to state laws, Smith would continue to uphold general, neutral criminal laws that did not grant a religious exception even if those laws burdened religion.*

This evolution of the standard of review in Free Exercise cases was still not over, for two reasons:

•  Many states (twenty-one as of late 2016) passed their own RFRAs; other states are considering them; and some, including California, have rejected them. While intended to echo the federal RFRA, state RFRAs differ from one another.³⁰ Some of them, for example, dilute the substantial burden requirement, require more onerous scrutiny of the governmental interest, and expand the notion of “religious exercise.”³¹ (North Carolina and Michigan were also considering RFRA legislation in 2016.) Many of these state RFRAs are embroiled in controversy about whether they protect religious objections to same-sex marriage and gay rights statutes. I discuss these disputes in the next section.

•  In 2000, three years after City of Boerne, Congress—invoking its power under the Spending and Commerce Clauses—enacted the Religious Land Use and Institutionalized Persons Act (RLUIPA). Like the RFRA, the RLUIPA was strongly promoted by religious groups seeking to restore the special protections threatened by City of Boerne. The RLUIPA governs state and local government decisions that affect land use and institutionalized persons in ways that burden religion. It not only requires strict scrutiny of such decisions, but also allows successful challengers to recover their attorney fees, which Professor Hamilton believes encourages governments to concede.32

Thus, the Sherbert test lives on in both the RFRA and RLUIPA, limiting government actions that affect religious exercise.

EMERGING ACCOMMODATION ISSUES

In 2014, the Court decided a highly controversial RFRA case, Burwell v. Hobby Lobby Stores, Inc.,³³ whose bearing on difficult religious accommodation issues will be bitterly contested and litigated for years to come. The next year, the Court upheld same-sex marriage in Obergefell v. Hodges, decided on due process and equal protection, not RFRA, grounds. Both have already generated many accommodation demands, most invoking various civil rights laws. In this section, I briefly discuss these two decisions and some of the hard issues (some would call them “culture wars”) arising in their wake. Before doing so, however, it is important to keep the larger context in mind. As scholars who closely examine the disputes arising in this area note, the vast majority of RFRA-type claims have nothing to do with such hotbutton issues; instead, they involve the religious practices of vulnerable minorities (e.g., Amish, Muslims, prisoners)—exemption claims that, as Professor Christopher Lund puts it, “liberals should welcome.”³⁴ Thus advocates for religious accommodation should think twice before upending federal and state RFRA laws based on atypical high-profile cases like Hobby Lobby.³⁵

Hobby Lobby. At issue were regulations issued by the U.S. Department of Health and Human Services (HHS) under the Patient Protection and Affordable Care Act of 2010 (ACA). These regulations required non-exempt employers to provide coverage for the twenty contraceptive methods approved by HHS, including four that “may have the effect of preventing an already fertilized egg from developing any further by inhibiting its attachment to the uterus.” HHS granted exemptions to religious employers (churches and other religious non-profits) with religious objections to the contraceptive coverage mandate. Under such exemptions, insurers must “exclude contraceptive coverage from the employer’s plan and provide plan participants with separate payments for contraceptive services without imposing any cost-sharing requirements on the employer, its insurance plan, or its employee beneficiaries.”

The owners of two closely-held business corporations who have sincere Christian beliefs that life begins at conception sued HHS under RFRA and the Free Exercise Clause, challenging this mandate insofar as it required them to provide the four contraceptives to which they had religious objections.* The Court did not decide the case on constitutional grounds; instead, it interpreted the terms of the federal RFRA statute, applying its “substantial burden” and “least restrictive alternative” tests to Hobby Lobby’s facts. For the Court majority, Justice Alito first held that RFRA’s protections for “persons” extended to for-profit firms, noting that for-profit firms were not required “to pursue profit at the expense of everything else, and many do not do so.” Hobby Lobby was owned and controlled by members of a single family, so their beliefs were the same as the company’s, obviating concerns about determining the company’s sincere beliefs.

Applying RFRA’s tests, the Court held that the ACA “substantially burdened” Hobby Lobby by requiring it to “engage in conduct that seriously violates [the family’s] religious beliefs.” It reasoned that noncompliance with the ACA or dropping the insurance altogether would risk severe economic harm resulting from penalties, taxes, and competitive disadvantage if the firm dropped the insurance altogether. Rejecting HHS’s claim that the link between providing contraceptive coverage and destroying an embryo was too attenuated to constitute a substantial burden under RFRA, the Court held that its Free Exercise precedents precluded inquiry into the reasonableness of any sincerely-held belief, and Hobby Lobby’s was clearly sincere. But even assuming that the ACA mandate served a compelling government interest, the Court held, it failed RFRA’s “least restrictive means” test: HHS could have chosen more accommodating alternatives, provided the offensive contraceptives to women at no cost to religiously objecting employers, or extended to Hobby Lobby–type firms the ACA’s exception for houses of worship and accommodation for other religious nonprofits allowing them to self-certify their religious objection and requiring their insurers to exclude contraceptive coverage from the policy. HHS, the Court found, had failed to provide a good reason for not extending such exceptions to a firm like Hobby Lobby.

Justice Ginsburg’s powerful dissenting opinion, joined in most respects by the three other liberal justices, rejected the majority’s assurance (reinforced by Justice Kennedy’s concurring opinion) that its decision could be limited to Hobby Lobby’s facts. Congress, she argued, only intended RFRA to overrule Smith—not to effect the radical change that the majority opinion would produce; religious exceptions were proper only for individuals and nonprofit firms, which is why Title VII had never granted them to for-profit corporations. She argued that the ACA mandate satisfied both RFRA tests. Its insurance scheme served compelling governmental interests, and the tenuous link between firms’ religious convictions and the contraceptive coverage mandate did not substantially burden firms’ free exercise rights. Nor would the less restrictive alternatives cited by the majority actually serve Congress’s objectives.* Most troubling, she predicted, the majority’s logic would eviscerate numerous civil rights laws by inviting anyone who could cite a sincerely held religious objection to claim—and receive—an exception to those laws. And future courts, she observed, could only decide such claims by making judgments about the merits of particular religious claims, judgments that courts could not constitutionally make. Many other commentators have joined Justice Ginsburg in warning about Hobby Lobby’s invitation to demands for religious exceptions in many areas other than contraception coverage, with some seeing the decision as a threat “to the entire regulatory state.”36 And while Hobby Lobby formally applies only to federal RFRA claims, it may well affect how the states interpret their own RFRA statutes.³⁷ Finally, as law professor Ira (Chip) Lupu explains, the decision “appears to provide constitutional legitimation as well,” insofar as RFRA purported to restore constitutional norms (i.e., the Sherbert strict scrutiny test) that the Supreme Court had abandoned (in Smith).³⁸

Obergefell and Its Aftermath. One year after its Hobby Lobby decision, the Court in Obergefell v. Hodges upheld the constitutionality of same-sex marriage, which it found to be a fundamental right protected by the Due Process Clause. Unquestionably among the most important Court rulings in modern history, Obergefell directly affects the lives, identities, and family arrangements of millions of people historically subject to harsh discrimination and blighted existence in many areas of American life. The decision’s basis in the federal Constitution precludes state-level erosion of that right.

By dismantling the legal barrier to marriage equality for gays, Obergefell may have paved the way to broader protections for all LGBT people in civil rights statutes more generally. Few federal civil rights laws cover sexual and gender orientation minorities,* nor do most state civil rights laws cover them (although more states are already doing so).³⁹ Notably, as Lupu observes, racial minorities won legal equality in employment, education, and other public services well before they gained it in marriage, the most intimate realm, in Loving v. Virginia.⁴⁰ In contrast, LGBT people have won this harder-to-achieve equality first, which may mean that gaining their legal equality in areas less intimate and more conventional than marriage will come more easily and swiftly.41

But Obergefell is a two-edged sword. On the one hand, it supports expanding LGBT rights into the workplace and areas other than marriage—often with the strong support of business groups and other prominent institutions that sometimes threaten to boycott recalcitrant jurisdictions. The NBA,⁴² the NCAA,⁴³ PayPal,⁴⁴ Salesforce,⁴⁵ and the Atlantic Coast Conference,⁴⁶ for example, have all exerted pressure on North Carolina in the wake of its anti-LGBT legislation. But on the other hand, Obergefell is galvanizing strong, religiously-based objections to these protections and fueling demands for constitutional, statutory, or regulatory exceptions in some states. And because most civil rights laws now include religion as a protected attribute, employers and other objectors can invoke their own free-exercise rights to buttress their demand for exception by citing the state or federal RFRAs as well as Hobby Lobby’s extension of religious exceptions into the commercial and corporate spheres. Further, the RFRAs’ key tests are ambiguous enough to make such demands legally plausible. As Lupu notes, “RFRAs are a Rorschach test, on which everyone can project their hopes and fears, and cannot be proven wrong until the courts resolve particular questions. Even when courts do so, the resolutions are highly fact-specific, so RFRA-based judicial precedents are rarely generalizable.”⁴⁷

Indeed, this process is already well under way. At the federal level, a proposed “First Amendment Defense Act” would bar the government from taking any “discriminatory action” against people who believe that same-sex marriage is wrong on moral or religious grounds,⁴⁸ while an Equality Act bill would extend civil rights protections to LGBTs in various contexts. The states and localities naturally present a much more diverse and arguably important field for these battles because state law governs many more of the contexts in which LGBTs seek equal treatment. Just how important the federal-state difference is will partly depend on how states and localities resolve the following issues. For example:

•  Hobby Lobby interpreted only the federal RFRA and so is not binding on states, although many states will likely follow it in reading “persons” in their own RFRAs to protect some for-profit entities.

•  A state or locality may not have an RFRA statute (or if it does, the statute may differ from the federal one). In either case, the state or locality may still choose to look for guidance to the federal RFRA, to Hobby Lobby’s interpretation of it, or both as it formulates new laws or interprets existing ones. (Many state RFRAs were enacted long before marriage equality was on the horizon.)

Lupu’s description of this complex pattern remained accurate as of July 2016:

[T]wenty-two states (and D.C.) have jurisdiction-wide laws that forbid discrimination based on LGBT status in employment, housing, and/or public accommodations. Twenty-one states now have RFRAs modeled on the federal Act. The overlap between these two sets is four—Connecticut, Illinois, New Mexico, and Rhode Island are currently the only states that have RFRAs and state-wide anti-discrimination laws that include LGBT status. In addition, another eight states have the combination of broad anti-discrimination laws and state constitutional provisions that have been construed in ways akin to the pre-Smith, pre-RFRA regime of free exercise adjudication. Accordingly, under current law, only twelve states—four with RFRAs plus eight with constitutional norms—present the possibility of strong conflict between statewide LGBT anti-discrimination laws and statewide religious freedom laws.

Local law offers another dimension to the possible clashes. In a considerable number of states that have RFRAs, are considering RFRAs, or have strong constitutional protections for religious exercise, local jurisdictions have enacted LGBT anti-discrimination laws. These include (among many others) Phoenix, Arizona, where proposed amendments to the state’s religious freedom law produced a political uproar, nearly costing the state the 2015 Super Bowl; and Atlanta, Georgia, where a proposed state level RFRA generated considerable controversy, and eventually died as a result of concern that it would promote discrimination against members of same sex couples and others.49

As of then, Lupu reported, the much-publicized disputes between LGBTs and various wedding florists and bakers had been resolved in favor of the LGBTs; the specific legal rationales reflected (1) a state’s own religious freedom law, which may or may not include an RFRA (and one inflected by Hobby Lobby), and (2) its anti-discrimination law, which may or may not extend to LGBTs or to the specific contexts or markets (e.g., housing, goods and services, child welfare) in which disputes arise. Actual resolution of such disputes will turn on these two legal variables. Obergefell is marriage-specific, so it remains unclear how it will affect emerging disputes in other realms in which religious claimants demand accommodation. One such realm is access by transgender people to bathrooms, locker rooms, and other public facilities based on their chosen sexual identity rather than their gender at birth. In July 2016, Massachusetts became the nineteenth state to protect transgender people in this way.⁵⁰ Another important arena, which has received less attention so far, is child welfare services.⁵¹

In May 2016, the Obama administration directed that Title IX of the federal Civil Rights Act, which prohibits discrimination on the basis of sex in federally funded programs or activities, requires schools and other recipients to allow transgender students to use bathrooms aligned with their current gender identity rather than their gender at birth. Almost half the states as well as some private groups promptly sued to challenge this directive, and in August 2016 a federal district judge in Texas enjoined it.⁵² Although some of the private challenges are free exercise claims based on religious requirements of modesty in undressing before members of the opposite sex, most challenges are based on other theories—that “sex” discrimination under Title IX does not cover gender identity; that the directive failed to comply with the Administrative Procedure Act; and that it overrides constitutional protections of state sovereignty.53 Quite apart from the legal issue, the administration rolled out the directive in a way that has already sowed public misunderstanding, opposition, and partisan wrangling.⁵⁴

Another potential dispute could involve challenges to the tax-exempt status of religious groups that refuse to accept same-sex marriage. The stakes in this challenge would be high; the groups value tax-exempt status, while others see it as a government imprimatur that a group’s activities are consistent with public policy and values. Such a challenge would rely on Bob Jones University v. United States,⁵⁵ a 1983 ruling in which the Supreme Court allowed the government to revoke the tax-exempt status of a religious university that barred interracial dating by students. (The Court did not address “churches or other purely religious institutions.”) Since then, the government has not revoked any religious entity’s tax exemption on such grounds, and the Obama administration said that it would not do so in the wake of Obergefell. In my view, tax exemption should not be denied to an otherwise-eligible religious group simply because it limits access to its activities and facilities based on fundamental, good-faith religious tenets inconsistent with a still-controversial government policy. Tax exemption aside, other disputes will surely arise over whether governments can deny otherwise-eligible religious groups access to public funds or other benefits if they discriminate against LGBTs. In the final section, I suggest some general principles to guide our approach to disputes of this kind.

GOING FORWARD

Conflicts between the religiously devout and other people whose beliefs and behaviors offend them go back to the earliest human communities. Today, the almost limitless ways of living differently from the majority in our society—as well as the vast increase in the number and scope of government policies that regulate people’s behaviors—multiply these conflicts. Liberal norms of tolerance and conciliation can help to reduce these conflicts’ intensity to manageable levels. In an earlier book, I advocated two informal “punctilios”—more candor, and thicker skins—that can ease (not solve) these diversity-generated disputes.56 But I also noted that such norms will often be inadequate to avert conflicts between fervent religious convictions and equally fervent demands (some partly theological) for equal treatment, dignity, and respect. Indeed, they may even exacerbate them.

When these norms and punctilios fail to manage the conflict, we are left with law, the essential democratic tool of last resort albeit one with intrinsic limitations.⁵⁷ Those who contest hard issues, of course, always negotiate solutions in the law’s shadow. When law represents and then enforces workable solutions between conflicting interests, it can legitimate those solutions and help implement them. Still, the fraught politics of many accommodation issues and the new contexts in which evolving legal principles must be applied produce much uncertainty and conflict. Moreover, some techniques to encourage compromise on these issues—for example, finding more money, splitting the difference, making side payments to offset losses, redefining the conflict, and pushing it to a different governmental level—are often unavailable here.

How, then, can a liberal, diverse, constitutionally constrained democratic society hope to resolve these proliferating disputes? Many thoughtful students of religious accommodation issues have considered this question, and LGBT claims give it a renewed urgency. I begin with Professor Steven Heyman’s point that these conflicts proceed from the fundamentally different “identities and ways of life” of religious traditionalists and LGBT people (many of whom, of course, are religious).⁵⁸ Finding his view useful but somewhat question-begging, I then present a number of other approaches.

Heyman first rejects two decision rules in which either the most powerful group or the political majority gets its way. Mere dominance, he argues, “does not mean that it should be allowed to promote its own identity and way of life at the expense of another group.” He then rejects RFRA’s pro-religion tilt because it treats LGBT status not as an inherently valuable way of life but merely as a placeholder for a “governmental interest.” This, he suggests, may be appropriate in a case like Smith where the government’s interest is to enforce a general public welfare law (there, narcotics control), but not in a clash that pits two ways of life against one another. Trying to balance values that are really incommensurable, he claims, does not sufficiently protect rights, and a tolerant “live-and-let-live” approach provides no guidance as to which should prevail when the disputants interact in public spaces or when the state is regulating public services or benefits. Heyman’s preferred approach is “mutual recognition” by both sides of the full and equal citizenship of the other and the right of each to hold and live according to their own beliefs.

But suppose the identity and way of life of one group are inconsistent with that of the other? Heyman posits some “easy” cases—the law cannot compel religious objectors to perform same-sex marriages,* nor must churches change their teachings or use their property for acts that offend them. He then poses a “hard” one, mentioned earlier, in which a vendor refuses to perform a wedding-related service (e.g., photography, cake, flowers) for LGBT couples, one that it willingly provides (and that the law does not permit it to withhold) from mixed-sex couples on the basis of certain characteristics. Heyman would resolve this conflict in favor of LGBTs because the vendor is defining its identity in a way that “denies the legitimacy of” the couple’s identity.

This conclusion, however, is conclusory, side-stepping the central question of what legitimacy means (an issue also discussed in the previous chapter on affirmative action). Perhaps the vendors in such cases are not denying the couple’s marital “legitimacy” but are instead saying in effect: “You have the right to marry but because I oppose and am offended by it, I simply don’t want to deal with you, so go in peace and find another provider who isn’t offended.” Law professor Douglas Laycock made a similar point in a brief filed in Obergefell: “Burdens on religious exercise arise only when the state demands that religious organizations or believers recognize or facilitate a marriage in ways that violate their religious commitments.… [G]ay couples ought to be free to be married, and religious dissenters ought to be free to refuse to recognize these marriages.”59

Consider further that civil rights law allows A to discriminate against B so long as A’s decision is not based on specific prohibited grounds like race, sex, disability, and in most states, religion. A may refuse to deal with B if A is motivated by dislike of B’s appearance, perceived immorality, rude conduct, or any other factor not prohibited by the statute. Distinguishing those reasons from a religious one, of course, can be difficult, as they commonly overlap and intent is an elusive state of mind. The law, by treating these reasons differently, seeks to strike a delicate balance between A’s interest in autonomy and B’s interest in equal treatment. But in many other areas of the law, an actor’s intent likewise determines outcomes. Here, A’s possible religious motivation simply adds a constitutionally-inflected wrinkle to the balancing.

In short, Heyman’s mutual recognition approach does not tell us how to apply the RFRA balancing factors, much less how to resolve hard cases. In what follows, I propose other principles or considerations that may provide more guidance in such cases.

Dignity. Any analysis of a proposed accommodation (under RFRA’s various prongs, say) should assess how it would affect the dignity of those affected. Although considering dignitary effects should not be controversial in principle—indeed, dignity is as fundamental a norm in many well-functioning foreign legal systems as “due process” and “equal protection” are in ours—it (like legitimacy) is easier to assert than to define.

Champions of LGBT rights stress the dignitary harms that may result when, for example, a wedding baker cancels a tasting upon finding out that the couple is gay⁶⁰—a harm magnified by the ancient obloquy LGBTs have had to endure and by the fragility of their recent gains. Yet religious persons may also have strong claims of dignitary harm if their accommodation demands are denied. As Sherif Girgis writes:

[I]n many disputes, both sides could claim with equal force that a decision against them would morally stigmatize them. Grant that exemptions from baking same-sex wedding cakes tell gay couples that intimacies central to their identity are immoral. What about denying the bakers’ claims? Won’t that tell them—and traditional Muslims, Orthodox Jews, and Christians—that beliefs central to their identity are bigoted? If exemptions from performing abortions tar women who’ve had them, coercing prolife doctors must brand them enemies of women’s equality. On most serious issues, any side might feel deeply stigmatized by rival actions or policies.61

The fact that dignitary claims may exist on both sides does not necessarily diminish their significance. In the end, decision makers must assess each such claim and weigh it against competing ones. For example, I imagine that merely selling a product to be used in a gay wedding to which one objects is not “participation” in any obvious sense and thus is at most a minor indignity, if that. But other factors could alter that assessment. (One might also note that a group’s power to obtain the legal accommodation in the first place implies that its members’ dignity is already well respected by the state.)

While decision makers should weigh dignitary harms on both sides of conflicts over religious freedom, they might be wary of overemphasizing such harms. As Girgis explains, a claim of stigma can actually be self-fulfilling: “The more that we—or officials, in weighing complicity claims—say that a policy or belief expresses disdain for a group, the more it will take on that social meaning. Lawmakers or judges trying to fight the harm might thus extend it.”

Choice. Policy makers and courts can ease some conflicts over religious freedom by taking individual choice into account when fashioning the legal rules applied to such disputes. Americans place a very high value on choice—in families, markets, religion, and most other social settings—for well-understood reasons. Choice protects individual autonomy and freedom from state coercion that would compromise their will or conscience. In many disputes, A’s choice can only be upheld by rejecting B’s; there, coercion is unavoidable. But in a society whose social and legal architecture is usually flexible enough to accommodate competing values, we can sometimes minimize coercion by invoking the more general principle of choice.

If choice is to be a weighty factor in analyzing a proposed accommodation, we must determine how much choice actually exists. Consider the now-familiar example of a same-sex couple denied a wedding cake or other service by a vendor with a genuine, strong religious objection to such marriages. The jurisdiction’s antidiscrimination statute, of course, may have resolved the conflict between the choice of the couple and the choice of the vendor in favor of one or the other. But even if the statute does prohibit this kind of discrimination, the vendor may claim a constitutionally-required or RFRA-based exception (citing Hobby Lobby). In this case, the decision maker’s task is more difficult. It might take choice into account by determining whether the couple can obtain the same or comparable service from another vendor. (In most but not all cases, comparability of price, quality, and convenience should be easy to determine.) If so, the couple’s desire for the service can be met without having to compel the objecting vendor to provide it. Choice, of course, is no panacea: Either decision will disappoint one side or the other. Even so, maximizing choice might reduce that harm or make it more palatable.

School curriculum and prayer provide other familiar contexts where valuing choice can reduce coercion and conflict yet still accommodate the competing values (e.g., neutrality among religions and between religion and secularism; and state responsibility for the welfare of children whose parents are incompetent, neglectful, or abusive). Although the specific facts will matter a lot—simple or categorical answers are seldom possible—some cases are easier than others. Small children’s conduct and religious views are largely controlled by the parents, presumably (not always) dictated by what they think is in their child’s best interests. In Yoder, the parents’ choice grew out of a deep, responsible Amish tradition in which that community replaced the state in many areas of life.62 Reasonable people can assess these situations differently, of course, just as the justices did in Yoder. But the inevitable interest-balancing should accord great, perhaps decisive, weight to parental choices about their children’s welfare.

School cases often involve young adults. Here, the value of their own choices should be important, perhaps dispositive, in resolving disputes; young adults are far better equipped than younger children to exercise informed choice. The Court’s recent prayer decisions have blurred the relevance of maturity. Lee v. Weisman63 barred a school-sponsored clerically-led prayer at a middle school graduation ceremony, and Santa Fe Independent School District v. Doe⁶⁴ barred a student-delivered prayer over the school’s public address system before high school football games. Reasonable people may disagree about whether the state was sufficiently neutral on the facts of these cases and whether each group of students was mature enough to choose, but I am troubled by how the Court defined choice and coercion in these cases. The students were free to remain seated or silent if they did not wish to participate, yet the Court still held that they were deprived of choice and harmed psychologically.

This reasoning, I think, trivializes the ideas of coercion and harm, presumes that young adults are more psychologically fragile than the vast majority probably are, and seeks to protect them from possible peer disapproval if they hold to their convictions and choose not to participate. If the school maintains the requisite neutrality and merely recognizes and fairly accommodates many students’ observance on certain occasions, that should suffice. The mere possibility that non-observant students might feel some social discomfort is not coercion—it is inevitable in a religiously diverse society—and there should be no constitutional right to avoid it. By limiting their exposure to any risk of such discomfort, the Court may actually increase their vulnerability to it when they are exposed, as they surely will be, to other unfamiliar practices later in life. (The bizarre controversy over “trigger warnings” in the college context comes to mind.)⁶⁵ As the dissenting justices in Lee urged, the idea of unconstitutional coercion should be limited to the threat of official sanctions, intimidation, or clear psychological harm.

In cases involving young children or in which parental choice arguably threatens a child’s life or public safety, courts must strike a somewhat different balance and deny a religious accommodation. Prince v. Massachusetts,66 a 1944 decision, is the classic example. There, the Court upheld the state’s power to vaccinate a child against communicable disease despite religious and other objections, invoking a traditional distinction between a belief and conduct. Belief is almost always protected,⁶⁷ but conduct is increasingly subject to regulation based on a majoritarian view of the child’s interests rather than that of parents with exotic or deviant beliefs. Still, the vast majority of parents have a sustained and informed moral, physical, and financial commitment to their child that no state can match and that warrants great deference. Such deference is a price we rightly pay for the advantages of family autonomy, which typically (not always, alas) serves the child’s interests well enough—especially compared with many foster care alternatives. The state, of course, must always prohibit parental abuse and neglect, but it should only overrule parents’ religious choices when they create serious, well-established, and unacceptable risks to their children—for example, female genital mutilation.⁶⁸

Conduct versus Status. Some demands for accommodation proceed from an aversion to having to act in a religiously repellent way, while others base their claim on a religious objection to the target’s very status. Professor Lupu analyzes how this conduct-status distinction might affect accommodation claims under RFRA. He imagines an employer with a genuine religious objection to the worker’s same-sex marriage claiming an exemption from the Family Marriage and Leave Act (FMLA), which requires firms to grant unpaid leave to workers with seriously ill family members. In Hobby Lobby, the Court granted the accommodation under the RFRA balancing test but here, Lupu argues, no exemption should be granted. Whereas the ACA would have forced Hobby Lobby to act in ways that promote the objectionable contraception, the FMLA would require the employer to recognize a worker’s objectionable marriage status. To Lupu, it is more harmful to victims of discrimination if it pertains to many aspects of their family life, not just to contraception:

[T]he objection is broader and far more troubling [than in Hobby Lobby] because its target is a relationship, an ongoing status protected in many ways by law. If the employer, acting on religious grounds, can treat a marriage as invalid, the employer can presumptively exclude the spouse from all benefits that federal law requires private employers to provide to employees’ spouses.… It is difficult to see how this kind of claim of a religious burden could be limited to same sex marriages. Why would it not also extend to other marriages to which the employer had religious objection (e.g., inter-racial, inter-faith, purely secular), or to parent-child relationships to which the employer had religious objections (e.g., a child born out of wedlock, or through some form of assisted reproduction)? An employer might assert that any of these relationships are unnatural, disordered, contrary to God’s plan, or evil. These kinds of objections, to status rather than acts, extend far beyond any singular act of “sinful” behavior, and are sweepingly hostile to the life plans of those whose family connections are being denied.69

Even so, Lupu notes, the Court’s abject deference in Hobby Lobby to the firm’s own subjective characterization of its religious conscience may render the conduct-status distinction useless in an RFRA analysis—a difficulty that Justice Ginsburg emphasized in an early decision allowing a public university to deny official recognition to a student group that excluded LGBT students.⁷⁰

Least Restrictive Alternative. Another path to accommodation is already built into the law: RFRA’s requirement that a secular government rule that conflicts with a specific religious practice can only be upheld if the rule is the “least restrictive” way to achieve a secular government purpose that is “compelling.” This implies that a religious exception can only qualify to the extent that it is specific and clearly limited. As we saw, RFRA took this prong from the Sherbert test (which the Smith decision had briefly supplanted). It was dispositive in the Court’s 2015 ruling in Holt v. Hobbs, which cited Hobby Lobby in striking down a prison’s overly broad restriction on inmates’ beards.71 As noted in my summary of the Hobby Lobby ruling, the Court there found that the ACA failed this test: The law already had a less restrictive alternative for religious nonprofits—i.e., an exception that HHS could easily have extended to Hobby Lobby without further impairing the government’s interest in covering contraceptives. This least restrictive alternative test has the important virtue of minimizing the clash between secular and religious interests.

But the other way to limit these conflicts is for the law to narrow the scope of the religious exemption. As Lupu explains, the Court has often held that a religious exception can be no broader than is necessary to protect those roles and activities that have a “distinctively religious” character. Much of what goes on in for-profit businesses, and even in religious organizations, will lack this religious connection. In addition, striking the correct balance between allowing government to pursue compelling public values and allowing a specific religious exception should take into account the full range of potential third-party effects, positive or negative.* Several other factors can affect the least-restrictive-alternative analysis. Expanding the choices available to either or both sides is one. Another is the scope of the exception: The narrower it is, the more likely it is to be the least restrictive of secular interests. An exception’s scope can vary along several dimensions: the number of people it covers, the magnitude of its various effects on them, the breadth of its subject matter, its duration, and so forth. Whether it is indeed the least restrictive alternative or not depends on how it scores in each of these dimensions compared with possible alternative exceptions. Narrower exceptions tend to have a clearer purpose and meaning, and a more predictable effect. (For this reason among others, some statutes require public notice of exceptions.)

Lupu provides an example: the exemption from Title VII for coreligionist hiring by religious entities. Such specific exceptions “are typically not subject to interest-balancing. More significantly, they clearly identify their beneficiaries and specify the norms against which the exemptions may be invoked. So, for example, if non-profit religious entities were to be fully exempted from a federal law prohibiting LGBT discrimination in employment, the exemption would be clear and absolute.… [It] protects only entities with primary religious purposes, and hence presumptively excludes for-profit firms. Moreover, that exemption does not extend to exclusions based on race, sex, national origin, or any forbidden ground of discrimination other than religion.”72

Decentralizing Accommodation. As we have seen, states and localities sometimes enact religious accommodations in their own constitutions, RFRAs, civil rights statutes, and other laws. Roderick Hills makes a strong argument that this decentralization is socially desirable and should be protected.⁷³ His premise is that many accommodation issues constitute “reasonable and deep disagreements” about liberty that cannot be reconciled in principle or empirically. Both are worthy of respect because, as in a Hobby Lobby–type situation, “social and legal consensus provides no commonly acknowledged baseline against which the disputing parties can measure their rights, [thus] one side of the dispute can plausibly claim that any resolution of the [dispute] constitutes an invasion of their fundamental rights. There is no neutral ground between coercing the organization into violating its religious scruples or allowing that organization to coerce its constituents (employees, customers, contractors, etc.) into obeying the organization’s religious commitments.… Each side has plausible arguments to back their position, giving them all the more reason to dig in their heels.”

Hills urges that subnational jurisdictions not be preempted by national rules on accommodation but be permitted to modify or waive them. The Court already did some of this work in City of Boerne, spawning a multiplicity of state RFRAs. “The federal system as a whole,” he contends, “extends equal concern and respect to rival and reasonable conceptions of religious liberty by giving each conception a larger area in which it can be acknowledged as authoritative. Such federalism broadens what philosopher Jeremy Waldron calls the “right of rights”—that is, the right of citizens to say what their rights mean, where disagreement about the rights’ content is reasonably disputed.74

The de minimis Principle. The English and American common law developed the principle of de minimis non curat lex, usually translated as “the law does not concern itself with trifles.” This is a norm of judicial economy but more important, it also recognizes a more widely-relevant precept: In an increasingly crowded, diverse, sharp-elbowed society like twenty-first-century America, each individual should be expected to absorb some limited degree of intrusion, imposition, even offense if we are to live together without endless conflict and litigation. How limited? This is where the de minimis idea comes in. Like many vital principles, it is neither self-defining nor self-applying but requires case-by-case judgment. It does not deny the genuine, deeply-held feelings and values that the de minimis principle may decline to remedy. It simply means that one cannot look to the law to prevent or assuage the offense.

The de minimis principle often arises in the law of nuisance where A’s and B’s activities interfere with one another’s use and enjoyment of their neighboring properties. For example, A may be offended by the appearance of B’s house, the junk that B leaves in the yard, the smells that emanate from B’s kitchen, and so forth, and we may conclude that we would react the same way. But up to some level of offense (often uncertain in advance), A cannot look to the law to protect his sensibility. If A has chosen to live in a crowded urban community, A is expected to absorb the kinds of intrusions that “go with the territory.” The fact that an offense A experiences arises out of a religious commitment rather than a secular one (e.g., privacy, esthetic taste) should not make the de minimis principle inapplicable to A’s offense, though it might affect how one assesses the de minimis level in particular cases. What constitutes a de minimis burden is often in the eye of the beholder. It ultimately turns on the decision maker’s judgment, informed by analogous decisions in the past.

In the religious accommodation context, the de minimis principle would ordinarily come into play in applying an RFRA statute or a Sherbert-type strict scrutiny test to a government requirement affecting religion. The Supreme Court recently considered such a case, Zubik v. Burwell.75 It involved a regulation issued under the ACA requiring employers to provide health insurance to their employees, including contraceptive coverage. The Obama administration proposed a number of different processes and criteria under which religious employers might obtain an exemption to the contraceptive element of the coverage, some more onerous than others. Ultimately, it imposed a requirement that those claiming a religious exemption submit a form to their insurer or to the government (which will then transmit it to the insurer) stating their religious objections. Religious employers argued that having to submit the form to the government violated RFRA because it would substantially burden their religious exercise. To them, it was not de minimis. Rather than decide this issue, the Court punted, remanding the case in hopes that a compromise would be reached under which the groups would not have to submit the form. In my view, having to submit a simple form in order to obtain the exemption should be treated as de minimis—as a reasonable way to satisfy the contending interests in a notoriously complex and controversial program, while fully respecting the groups’ religious views.

Free Speech to Reduce Conflict. Law professor Andrew Koppelman has proposed allowing religious objectors to gay marriage and other secular policies to forthrightly express their objections without incurring liability under antidiscrimination laws that prohibit conduct that creates a “hostile environment” for protected minorities.⁷⁶ Under his approach, business owners could articulate their views but could not threaten to discriminate or treat any individual customer worse than others. “[E]ven if they have no statutory right to refuse to facilitate ceremonies they regard as immoral, they are unlikely to be asked to participate in those ceremonies. On the contrary, same-sex couples will almost all want nothing to do with them.” In this way, the conflict between the groups can be avoided to the extent that they are kept far apart from one another and the disputed service is not demanded.

Koppelman notes that his approach would vindicate important First Amendment values in disputes in which A experiences and expresses A’s moral distress at the conduct or beliefs of B, a member of a minority protected by equal-treatment and hostile-environment rules. One such value is the protection of religious disagreement, which, he reminds us, was a core concern of the Framers. Another is to promote mutual transparency among persons by encouraging their speech. A third value is the social learning that ethical confrontations can engender, which is particularly important when traditionally despised practices are at issue. Koppelman emphasizes that his approach can reduce and manage such conflicts better but is unlikely to eliminate them.

Compromise. A combination of these approaches can facilitate compromise. In March 2015, a Republican legislature and governor passed a law that received national attention, now known as the “Utah compromise.” Its gist is to extend the state’s civil rights law to bar housing and employment discrimination against LGBTs based on their gender identity, sexual orientation, and gay marriage. At the same time, it prohibits employer reprisals against religious objectors who express their opposition to these practices so long as their speech does not interfere with the company’s business. It allows public employees, like marriage license clerks, who object on religious grounds to recuse themselves so long as the service is not denied to the couple. As for religious nonprofit organizations, they already enjoy broad exemption from the state’s employment discrimination law, including from its prohibition of race and sex discrimination. The compromise thus allows religious nonprofits to refuse to hire LGBTs, just as it can refuse to hire other minorities.

Like most compromises, this one does not satisfy all participants and will be hard to replicate. It permits vendors to refuse service to LGBTs. Indeed, although the Mormon Church supported the compromise, it also affirmed “the rights of a religious community to set its standards and live according to them” by adopting new rules barring young children of gay couples from Church membership and disciplining members in same-sex marriages.77 That said, the negotiations and compromise in Utah were widely praised by both sides and by commentators throughout the nation. Whether its protections of objectors will mollify them in other states, and whether LGBTs will accept other practices that they feel treat them as second-class citizens, remains to be seen.

A compromise that leaves both sides frustrated can engender bitterness among committed, intransigent opponents; this can make the compromise harder to sustain, legitimate, and institutionalize. In some cases, this determined opposition may succeed in partially reversing it, as with abortion rights, or forestall its full realization, as with gun control. Even so, some reformers believe that religious accommodation can be a powerful tool to mollify opposition by showing it respect and by relieving it from what it sees as offensive impositions. Professor Robin Wilson, for example, maintains that providing Hobby Lobby–type religious exemptions has facilitated the enactment of same-sex marriage laws by many states and has even furthered women’s access to health services. (It is telling that same-sex marriage proposals are never enacted unless they include exceptions for religious objectors and cover clergy and houses of worship.⁷⁸ On the other hand, the enacted laws have not given exceptions to wedding vendors or public employees.)

Unsurprisingly, commentators differ over political strategy, predictions about how an accommodation is likely to affect third parties, and judgments about how American society does—and should—work through its cultural and legal conflicts. Some fear that religious groups’ success in winning such accommodations will just inflame the culture wars and further embolden the groups’ efforts to reverse the change.⁷⁹ Lupu predicts that both things may happen simultaneously: Obergefell will energize the movement to expand civil rights protection of LGBTs and this effort will stoke religious resistance to that movement.80 Needless to say, outcomes will depend on the breadth, specificity, and exact wording of a proposed exception.

CONCLUSION

This chapter’s analysis of the law and policy concerning accommodation of religion in the face of secular policies leads to important conclusions. In hard cases, the deep tensions—between protecting religious liberty through the Free Exercise Clause, preventing excessive government entanglement with religion through the Establishment Clause, and nourishing First Amendment values—yields few obvious answers. If government continues to expand its regulation of social conditions, it inevitably coerces more people, intensifying the conflicts between religious and secular values and extending those conflicts to more areas. The issue of accommodation, therefore, is bound to become more prominent and complex—especially as America’s social diversity continues to grow.

Ensuring that accommodation respects the integrity and humanity of all participants while protecting their fundamental rights poses an immense social and legal challenge. With rights at the core of our public discourse, accommodations are bound to compromise them as they are redefined and tailored in light of particular conflicts. To some this trimming and compromise dilute rights’ essence—a view that I believe misconceives how a liberal, diverse, competitive polity like ours must govern itself. In the realm of hard issues, the notion of a clear, absolute right is a fantasy—and a dangerous one at that. We wisely put a thumb on the scale to protect one value or another. But when we do so, we actually weigh and compromise them in order to preserve what the Court long ago called our system of ordered liberty.⁸¹

* By secular, I mean an attitude or orientation that is not grounded primarily in religious convictions.

* Some constitutional law scholars argue that there is really but one religion clause and that this fact affects how it is interpreted. See, e.g., Akhil Reed Amar, “The Supreme Court, 1999 Term Foreword: The Document and the Doctrine,” 114 Harvard Law Review 26, 120 (2000). Others argue that the Free Exercise Clause is essentially redundant because of the large zone of protection afforded to religious exercise by other constitutional provisions, especially the Fourteenth Amendment. See Mark Tushnet, “The Redundant Free Exercise Clause?” 33 Loyola University of Chicago Law Review 71 (2001); Kenji Yoshino, “Covering,” 111 Yale Law Journal 769, 927–30 (2002).

* Much of this section on religious diversity is taken more or less verbatim from my book Diversity in America: Keeping Government at a Safe Distance (2003), 264–71, with some updates. Sources can be found in the endnotes to that discussion. Readers not interested in exploring this topic can skip to the next section without losing the thread of the larger analysis.

* Rodney Stark, a leading religion scholar, disputes this. Citing low response rates, he doubts the accuracy of the widely cited Pew surveys from which most of this paragraph’s statistics are taken. And he contends that many of the “nones” were already atheists when earlier surveys were done but only now are acknowledging it. Stark, The Triumph of Faith: Why the World Is More Religious Than Ever (2015).

* The state constitutions of seven states ban atheists from holding public office, and Maryland’s provides that belief in God is a requirement for jurors and witnesses. These provisions are almost certainly unconstitutional. Torcaso v. Watkins, 367 U.S. 488 (1961).

† Religious diversity as an ideal (as distinguished from it as a fact) is very recent, as is the ideal of immigrant diversity, with its religious implications. Only in 1965 did Congress end the national origin immigration quotas and Vatican II issue Dignitatis Humanae, urging religious freedom for all.

* Much of this section is taken more or less verbatim from my book, Diversity in America: Keeping Government at a Safe Distance (2003), 274–76. Sources can be found in the endnotes to that discussion. Readers not interested in exploring this topic can skip to the next section without losing the thread of the larger analysis.

* Other liberal democracies with strong secularizing trends but no Establishment Clause face similar dilemmas. See, e.g., Stephen Castle, “Refusal to Run a Church Ad Has Britain in an Uproar,” New York Times, November 24, 2015, A4.

† As the “In God We Trust” motto on U.S. currency suggests, the neutrality norm recognizes exceptions. They are usually rationalized in terms of the Framers’ usage, hoary tradition, or de minimis offense. For a non-exhaustive list of this kind of “ceremonial deism,” see John T. Noonan, Jr., The Lustre of Our Country (1998), 236–37.

* For example, in Wisconsin v. Yoder, 406 U.S. 205 (1972), the Court granted an exception to the Old Order Amish, hardly a political powerhouse even in Pennsylvania. It gave one to an even smaller, more obscure sect in Gonzales v. O Centro Espirita, 546 U.S. 418 (2005). Smith did go the other way, denying an exception for even a minor, peyote-eating Native American sect, but as I explain, Congress in the RFRA emphatically rejected Smith.

* Readers not interested in exploring the legal architecture of “establishment” can skip to the next section without losing the thread of the larger analysis.

† Two other meta-doctrines are standing and incorporation. Standing prescribes who may sue an Establishment Clause violation. Incorporation holds that the Religion Clauses limit the states as well as the federal government.

* Even so, the Court once upheld a federal action that the dissenting justices thought “promise[d] to destroy an entire religion.” Lyng v. Northwest Indian Cemetery Protective Assn., 485 U.S. 439, 476 (1988).

* Notably, some states have refused to use a Smith framework to interpret the Free Exercise provisions of their own constitutions, instead providing more expansive, Sherbert-type protection. Paul Benjamin Linton, “Religious Freedom Claims and Defenses under State Constitutions,” 7 University of St. Thomas Journal of Law and Public Policy 103 (2013).

* Pharmacies are another venue in which religious objections to contraceptives are being litigated. In June 2016, the Court declined to hear such a case, over a strong dissent by three justices. Stormans, Inc. v. Wiesman, 136 S. Ct. 2433.

* The Obama administration, eager to avoid a confrontation over contraceptive coverage, developed a rule that it hoped would be an end run around Hobby Lobby. Under the rule, employers like Hobby Lobby can state their objections to HHS, which will transmit it to the insurer, which must then provide the coverage at no additional cost to the firm. I discuss this rule, which remains under challenge (Zubik v. Burwell, 136 S.Ct. 1557 [2016]) in the final section of this chapter.

* The exceptions are the Violence Against Women Act, and President Obama’s broadening of an executive order to protect LGBTs from discrimination by government contractors.

* A court has so held in the high-visibility case of Kentucky count clerk Kim Davis. Miller v. Davis, 2015 U.S. Dist. LEXIS 105822 (E.D. KY, Aug. 12, 2015). North Carolina and some other states have adopted “Kim Davis” rules that protect same-sex marriage licenses but allow religiously objecting officials to recuse themselves from participating.

* Even before RFRA, the Court emphasized the third-party burdens aspect in a 1985 ruling finding that a statute favoring Sabbath-observing workers regardless of any adverse effects on other workers and the employer violated the Establishment Clause. Estate of Thornton v. Caldor, Inc., 472 U.S. 703 (1985).