How should religious beliefs and organizations figure in our political life? At no time in recent memory has this question aroused the passion it does now, thanks to a president who reiterates the importance of his evangelical Christian beliefs and to the Supreme Court’s recent far-reaching changes in the interpretation of the First Amendment’s guarantee—in the Establishment and Free Exercise Clauses—that “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.”

In truth, we face not one crucial question about religion and government, but four related ones. The first question is what accommodations government should make to the exercise of religion. Should it exempt religious pacifists from military duty, permit Amish parents to withdraw their children from school after the eighth grade, or allow the Native American Church to use peyote during worship? Should any such exemptions from ordinary legal requirements be extended to people with strong nonreligious claims to the same privileges? How far should legislators decide these matters; to what extent should courts do so as part of enforcing the Constitution?

The second question concerns the government’s sponsoring or endorsing particular understandings of religious truth. Should public school teachers lead prayers, or complement the teaching of evolution with ideas of intelligent design? Should winter holiday displays on public property include crèches? Should the Pledge of Allegiance contain the words “under God?” Do these practices amount to forbidden establishments of religion? Related questions arise about the discourse of high officials. Should they, like George W. Bush, keep referring to their own religious convictions and employ rhetorical expressions that appeal to cobelievers?

The third question involves financial assistance. Religious groups offer a wide range of social services, including hospitals, adoption agencies, food kitchens, drug rehabilitation programs, and schools. Should public money assist these endeavors, so long as the state does not favor any particular religion or favor religious efforts over nonreligious ones? If the government can, consistent with the Establishment Clause, pay for these services, should it? If so, what conditions should it set regarding those who receive and provide the services? More particularly, when it pays the bill, should it also bar religious discrimination? Finally, to what degree should religious organizations and those who contribute to them receive relief from property, sales, and income taxes?

The fourth question, although a staple of political philosophy during the last quarter-century, is the least understood. It concerns the explanations citizens and officials offer for their political positions. Should everyone rely on “public reasons,” not private faith, when they take part in the political processes of liberal democracies, or may people rely on whatever premises they find most convincing? The initial thought, that people relying on religious principles necessarily promote their religion, is mistaken. Consider the impoverished lives of animals subjected to factory farming. Suppose someone supports a law to assure a more tolerable existence for animals raised for human consumption because she believes God made other animals not purely for human benefit but as independently valuable creatures deserving a decent level of care. In supporting the law she does not impose her religious views; rather she uses her deep understanding of reality to protect creatures that deserve protection. That people should rely on public reasons does not simply follow from the principle that government should not promote religion.

In Liberty of Conscience, Martha Nussbaum, a philosopher and public intellectual, attempts to provide answers to these questions. Governments, in her view, should make accommodations to the concerns of religious conscience and should extend these to the expression of nonreligious conscience as well. Governments should not engage in any endorsement of religion, however inclusive. Religious groups and believers should participate on an equal basis in programs that aid nongovernmental providers of social services. On the fourth question, about the kinds of discourse that are appropriate in justifying political preferences, Nussbaum indicates in passing that she favors the “public reasons” position I have just summarized.

At the heart of Nussbaum’s approach is her claim that shielding religious conscience is supremely important for any society that respects individuals and treats them with dignity. Conscience for her is not limited to convictions about how people should act; conscience is “the faculty in human beings with which they search for life’s ultimate meaning.” Any society that acknowledges the equality of citizens should respect their consciences equally, a respect extending to nonbelievers. According to Nussbaum, the Free Exercise and Establishment Clauses should be understood mainly to guarantee equal liberty of conscience. This broad approach, she acknowledges, does not eliminate many difficult exercises of judgment, and courts rightly leave some issues to legislative resolution; but courts, she argues, should not abdicate the task of protecting liberty of conscience to the legislative and executive branches.

Much of Liberty of Conscience displays various applications of Nussbaum’s general theory, but we can see better why that theory matters in practice if we understand some of the alternatives she rejects along the way. She opposes the Supreme Court’s present doctrine that the Free Exercise Clause does not protect religious practice against reasonable general laws. Rather, she argues, for example, that those who take peyote as a central aspect of their worship should, under the Free Exercise Clause, have the right to do so, despite a reasonable general law against the use of peyote. She also opposes a focus on a person’s degree of religious liberty that is inattentive to how that liberty compares with the liberty of others. If the law protects refusal of military service on grounds of religious conscience, it should, she argues, protect such refusal on non-religious grounds as well. In respect to the Establishment Clause, she rejects the narrow interpretation that it forbids only the promotion of particular religious faiths, or only coercive government practices, or applies only to the federal government (not states).


In defending her thesis Nussbaum combines philosophy, legal doctrine, and history. Since she does not suppose that judges should be wedded either to the original understanding of various constitutional clauses or to the precise reasoning of earlier cases, her fundamental argument—made with force, eloquence, and occasional oversimplification—need not rest on any specific account of history or prior decisions. Partly to persuade those whose views about constitutional adjudication differ, Nussbaum lines up historical and judicial sources in support; as I shall suggest, some crucial aspects of these accounts are one-sided.

Nussbaum’s own understandings of complex constitutional issues recognize nuances that committed people on both sides often miss. Raised as an Episcopalian, a convert to Reform Judaism who remains seriously religious, she shows how not only skeptics but deeply religious citizens can welcome a political order in which governments refrain from promoting religion; and she offers many appealing proposals for resolutions of highly contested issues. Recognizing strong countermovements, Nussbaum nevertheless asserts that the dominant ideal in American history has been respect for an equal liberty of conscience. Early on, she asserts that the country “respects people’s committed search for a way of life” and “has long understood that liberty of conscience is worth nothing if it is not equal.”

One doubts, however, that Nussbaum really believes liberty of conscience is worthless if it is not absolutely equal. She rightly contends that any country with an established religion does not respect various religious consciences equally; but a country with a weak establishment that grants generous liberty of conscience (as with the Church of England, for example) is surely preferable to outright oppression of all but the favored faith.

Perhaps this rhetorical slip is of little moment, but Nussbaum’s problems with history begin on the previous page, where she writes that the Pilgrims crossed the ocean to recover a space of “liberty and equality” denied them in England, but that the settlers forgot the lesson of equality implicit in the first Thanksgiving. It is true that the early Pilgrims, seeking a haven for their own faith, were tolerant of diverse religious outlooks; but they hardly had a developed philosophy of equal respect for conscience. And the more numerous Puritans who soon followed explicitly rejected such notions of equal respect, less because of the harsh conditions Nussbaum emphasizes than because of a conviction that their “City upon a Hill” should be constructed according to their religious premises.

Nussbaum devotes a substantial chapter to Roger Williams, whose Rhode Island society practiced religious toleration and who defended broad liberty of conscience in The Bloudy Tenent of Persecution, well before John Locke’s influential A Letter Concerning Toleration. Williams believed that people of diverse religious understandings, capable of morally virtuous lives, could build a common life together. Unlike Locke, he perceived the vulnerability of conscience to intolerant public authority, regarding persecution as “soul rape.”

Contrary to the standard view, built on Williams’s extensive use of theological language and his unremitting opinion that true Christians (a small number) should separate their religious practices from the unregenerate masses, Nussbaum suggests that his conclusions about conscience did not rest heavily on his own theological understandings. She goes further in treating John Rawls’s work as articulating and developing Williams’s ideas. Rawls claimed that members of a liberal democracy should not rely on their own religious perspectives and other “comprehensive” views to decide fundamental political issues, but rather should seek shared premises and common ways of discovering facts.

Nussbaum provides little support for her ascription of only minimal importance to Williams’s theological premises—a view that may actually reduce the potential appeal for modern evangelical Christians of his plea for equal liberty of conscience. She gives us no basis to suppose that Williams thought citizens should put aside their Christian understanding in deciding whether their governments should aid the poor or protect seriously impaired newborn infants.


One problem with Nussbaum’s continual references to Williams’s views as a primary source of our traditions is the failure of his systematic writings to exert significant influence in the colonies and early republic. When Nussbaum speaks of “the success of his arguments,”1 it is unclear whether she thinks that outsiders were affected by actual practices within Rhode Island or independently came to embrace ideas similar to his.

Nussbaum’s sketch of the history of anti-Catholic bias relates more directly to her interpretation of the Establishment Clause. That history is relevant to whether concern about state support of institutional religion has a solid ground, apart from concerns about equality.2 In a chapter called “Fearing Strangers,” which also describes appalling treatment of Mormons and Jehovah’s Witnesses, Nussbaum’s theme is that anti-Catholic bias in the mid-nineteenth century, directed against immigrants from southern and eastern Europe, led to an emphasis on separation of church and state by nativists who, all the while, were entirely comfortable with Protestant public schools. Intellectual liberals in the mid-twentieth century, who fueled modern opposition to aid for parochial schools, were also biased against Catholics—failing to recognize the Church’s long history of rational approaches, by way of natural law, to morality and politics, unaware of the rigorous, thoughtful education provided in leading Catholic schools, and neglectful of liberal Catholic thinkers, such as Jacques Maritain and John Courtney Murray.

But Nussbaum’s argument disregards or underemphasizes important facts. Within dominantly Protestant colonies, anti-Catholic sentiment was extremely powerful. No stranger to the Protestant colonists, Catholicism was the widely perceived enemy of true belief. Jonathan Edwards, the leader of the Great Awakening, expressed a common view when he spoke of the Pope as the Antichrist and regarded the Catholic Church as representing evil in the cosmic struggle against good.3 Nineteenth-century prejudice was built on a long, rich history of hostility. Opponents of aid to Catholic schools did not, as Nussbaum suggests, lack evidence about the Church’s hostility toward political liberalism. An 1832 papal encyclical condemned the “absurd…proposition which claims that liberty of conscience must be maintained for everyone. It spreads ruin in sacred and civil affairs….”4 Only with the Second Vatican Council in 1965 did the Church embrace robust liberty of religious conscience.

Although the Church’s rich philosophic traditions were pervasive in its best schools, run by Jesuits and other religious orders, parish schools, which provided most Catholic schooling during most of the twentieth century, were more narrow and dogmatic. Various leading separationists grossly overstated any Roman Catholic threat to American democracy; but in defense of their misgivings about aid to Catholic schools they could point to authoritative Church teachings and actual school practices. These practices have shifted significantly, with the great influence of Vatican II and with lay teachers replacing nuns and priests. When Nussbaum writes, “Most Americans can see by now that Roman Catholicism…favors and teaches independence of mind,” she fails to remark that that altered perception is partly the consequence of a changed reality.

Government aid to religious education is the modern issue for which it matters most whether someone sees equality or institutional separation as the heart of nonestablishment. Governments, uncontroversially, help finance religious groups providing other social services, such as hospitals and adoption agencies, so long as nonreligious providers are treated equally. It is widely agreed that (apart from compensation for military and prison chaplains) public money should not directly pay for religious practices and that organizations receiving public funds must make their services available to all regardless of their religion.

The sticking point—not discussed by Nussbaum—is discrimination in employment. The Bush administration has consistently maintained that religious providers of social services who receive federal support should be able to use religious criteria in choosing their staff members. Congressional resistance to this feature of the administration’s legislative proposals, colloquially called “Charitable Choice,” has resulted in a stalemate, leaving the administration to do most of what it wants by executive order. Although allowing religious groups to choose like-minded workers may serve the members’ religious aspirations, the use of religious criteria creates an obvious inequality among those seeking work, one that is unjustified for positions removed from religious practice and education.

The central national and local controversy over aid to religious activities has been about religious education. In the late nineteenth century Congress nearly approved a constitutional amendment to bar such aid; many states then adopted similar amendments to their own constitutions. In 1947, the Supreme Court, in Everson v. Board of Education, allowed New Jersey to pay for bus transportation for children attending parochial schools. Contrary to Nussbaum’s claim that the theme of equality dominated this discussion, Justice Hugo Black’s opinion for the Court emphasized outright separation of church and government. That equality was not Black’s central idea was confirmed by his later dissent when the Court’s majority accepted New York’s provision of nonreligious textbooks for parochial school students.5 For a justice guided mainly by considerations of equality, textbooks for nonreligious institutions would not have seemed so different from bus transport, but Black thought books involved the state in “the heart” of religious education.

In 2002, the Supreme Court approved Cleveland’s voucher program, which granted substantial reimbursement to parents of children in private schools.6 The great majority of the schools qualifying for aid were religious. Acknowledging the role of public schools in integrating our heterogeneous population, Nussbaum calls the case “genuinely difficult,” one instance of her refreshing recognition that in many Supreme Court cases both sides have strong arguments, and that law’s complexity is largely responsive to variations among circumstances, rather than a device of lawyers to befuddle everyone else. Cleveland’s program set reimbursement at a level likely to prove more attractive to religious schools than to other private schools, which expend more per pupil. And, of course, the financial aid involved government deeply in education by religious institutions. But if one concentrates on the state’s helping parents to have an equal chance to realize their educational aspirations for their children, the Court’s approval of the process is sound.

Judicial resolution of the voucher issue—unlikely to be overruled—has shifted the focus back to legislatures. Thus far, concern that public schools will suffer if private education receives substantial subsidies, pressure by teachers’ organizations, and, perhaps, a worry about the increasing number of strongly evangelical Protestant schools have proved formidable obstacles to the enactment of broad statewide voucher programs.

An interesting constitutional issue remains: Do state constitutions that forbid aid to parochial education violate the federal Free Exercise Clause by treating religion unfavorably? In 2004 the Supreme Court decided by a 7–2 margin that despite a general program to aid college students, a state may leave out those who plan to earn a “degree in theology,” one “devotional in nature or designed to induce religious faith.”7 Treating such studies as preparation for the ministry, the Court concluded that a state could choose not to support them. Although this decision pointed toward the Court’s acceptance of broader state restrictions on aid to religious education, Chief Justice William Rehnquist’s opinion was too opaque to allow a confident prediction. That Nussbaum regards the Court’s decision as definitely mistaken may suggest that her penchant for equality has led her to underrate the value of keeping government out of the core functions of institutional religion.

When it comes to governments themselves engaging in religious practices or teaching religious ideas, the Supreme Court has been fairly strict, forbidding them to sponsor oral prayers, Bible reading, and instruction in creationism within public schools, as well as displays of crèches and the Ten Commandments that convey a religious message. Building on the prior test that a practice is unconstitutional if its purpose or effect aids religion, Justice Sandra Day O’Connor, the crucial swing vote in many religion cases, asked whether the state is endorsing or promoting a religious view. Nussbaum agrees with this approach, which is designed to prevent members of minorities and nonbelievers from feeling like outsiders. She thinks that a faithful application of that approach would accept “In God We Trust” on coins, but would not countenance the use of the words “under God” in the Pledge of Allegiance recited by public school students. She thus disagrees with O’Connor’s own unpersuasive conclusion, in a case in which the Court’s majority sidestepped the constitutional issue,8 that the phrase “under God” does not really endorse a religious view but has faded into a form of ceremonial deism.

In 1990, the Supreme Court’s decision in Employment Division v. Smith9 radically altered the constitutional doctrine about accommodations to religious practice. A quarter of a century earlier, the Court had determined that a state could not bar a woman from receiving unemployment compensation because she was unwilling to accept employment requiring Saturday work, when her unwillingness rested on her conviction as a Seventh Day Adventist that she should not work on that day.10 The Court ruled that the state could not justify this impairment of her religious liberty without a “compelling interest” in doing so.

Nussbaum gives a rich description of Mrs. Sherbert, the plaintiff, and of the employment practices in her locale; and she approves the decision’s standard of judgment, writing that it accords due respect to religious conscience. In stressing that the state treated Saturday worshipers worse than those who worship on Sunday, Nussbaum transforms a supplementary ground of the opinion into its major theme; but she resists any theory that the Free Exercise Clause is only about unequal treatment, claiming instead that it calls for significant protection of religious conscience.

Most scholars agree with Nussbaum’s judgment that the state must accommodate religious convictions, though when she speaks of “a very workable and clear-edged doctrine,” she understates the difficulty of assessing burdens on religion and the weight of state interests. Suppose, for example, it is left to judges to determine whether the government should be able to recover money that a debtor who has just filed for bankruptcy has recently tithed to his church. It is no simple exercise to decide how great is the interference with religious practice and how important is the government’s interest in retrieving the money.

The Supreme Court’s other major case protecting free exercise of religions allowed a group of Old Order Amish to withdraw their children from school after eighth grade despite a state law requiring that they stay in school (or receive comparable schooling at home) until the age of sixteen.11 As Nussbaum recognizes, because withdrawal from school may well stunt the future development and vocational opportunities of some Amish children, this case, with powerful competing arguments, was difficult to decide.

Subsequently, the Court rejected claims of a Jewish psychologist in the Air Force that he should be allowed to wear his yarmulke indoors12 and of Native Americans that a development of government land that would destroy the quiet of a sacred locale should be prevented.13 The Court said that the military deserved extraordinary deference to decide what promotes healthy discipline, and that government development of its own land is not a “prohibition” of free exercise covered by the First Amendment. (Nussbaum suggests that the Court in the second case regarded the relevant burden on the plaintiff as not substantial; but the opinion says that the government could build its road even if it “virtually destroy[ed] the…Indians’ ability to practice their religion.”)

Few observers realized that most members of the majorities in both cases were poised to abandon the compelling interest test altogether for situations in which people violate laws that are not themselves directed against religion. That happened in Employment Division v. Smith, which rejected the claim of members of the Native American Church that they should be allowed to use peyote in their religious services. As Nussbaum recounts, Justice Antonin Scalia manipulated his interpretation of the prior cases in astonishing ways in a vain effort to demonstrate the continuity of this decision with previous holdings of the Court—not an uncommon strategy when the Court shifts constitutional doctrine. Few observers of the law were fooled, and outrage was the response of many scholars and the great majority of significant religious organizations and groups devoted to civil liberties.

The upshot has been two important new federal laws, similar legislation in at least thirteen states, and judicial decisions in various states, all restoring the doctrine that the state cannot impose a substantial burden on religious practice (as the peyote law clearly did) unless it has a compelling (very strong) interest in doing so (as it obviously would were, for example, human sacrifice involved). In other words, the state can interfere with religious practice if the interference is only minor or if the government’s need to interfere is strong (as the courts have consistently held about religious claims not to be subject to tax liabilities). As expressed in legislative form by Congress, this test has been accepted by a unanimous Court, even by Justice Scalia, who has a strong distaste for standards whose application is unclear.

The more difficult question is what to do about analogous cases where people assert claims of conscience that are not based on religion. Nussbaum spends considerable space on two conscientious objector cases, agreeing with the Court that genuine pacifists who are not religious in any ordinary sense should be treated like religious pacifists. (The Court tortured the language of the federal statute to reach this result, rather than stating a constitutional requirement.)

But what should be done about nonreligious drug users and nonreligious parents who claim that it is a matter of conscience to spend Saturdays with children? One might defend a limitation to religious claims if analogous, powerful nonreligious claims are unlikely (e.g., the right to withdraw children from school) or if the granting of nonreligious claims would create great dangers of fraud, as might be the case if all nonreligious claims of conscience were accepted as a basis to use a proscribed drug such as peyote.

Nussbaum expresses herself as “somewhat nervous” about the suggestion of one writer (myself) along those lines. She is hesitant to allow a differentiation between religious and nonreligious claims to use a forbidden substance when such a differentiation would rightly be rejected with respect to a law imposing military conscription. She also suggests that claims to stay home with one’s family on Saturday lack constitutional status.

This precise combination of positions is difficult to defend. With a psychedelic drug such as LSD, a user can plausibly assert a search for meaning in life. If potential protection extends to nonreligious drug users, any threshold requirement that a user show that he is bearing a “substantial burden” will disappear, since we have no way to figure out when a prohibition on nonreligious use imposes a substantial burden. (The state could still prohibit use by everyone if it has a compelling interest in doing so.) And why does Nussbaum rule out the possibility that the desire of parents to spend time with their children may raise a serious issue of conscience? The search for meaning in the lives of many parents depends on involvement with loved family members, and they take their care of their children as their highest moral responsibility. It would take much more than scattered comments to distinguish the nonreligious drug user (potentially protected) from the (unprotected) parent with a nonreligious reason to spend precious weekend days with children.

Nussbaum reverts to the last of our four questions about the place of religion in political life near her book’s end. She draws upon John Rawls’s notion that with respect to the basic premises of liberal democracy there is an overlapping consensus between believers and nonbelievers. In doing so she implies that, at least on fundamental issues, citizens and officials alike should rely on nonreligious grounds for political judgments. Nussbaum’s valid concern about equality of citizens provides a strong basis to conclude that officials should not run for office on the basis of their religious understandings and should not in their political discourses rely on particular religious outlooks or disregard nonbelievers. But it hardly follows, as Nussbaum seems to believe, that ordinary citizens should try to shed their religious convictions when they engage in politics or even that legislators should attempt to do so when they resolve troubling issues like the protection of animals. Such conclusions would require arguments beyond those Nussbaum makes here.14 My own conviction is that the balance of considerations regarding religious conscience points to a less drastic exclusion of religious convictions from political judgment.15

This Issue

May 15, 2008