The Yale Law Journal


Freedom for Religion

04 Dec 2023

abstract. The First Amendment’s religious-freedom provisions are best understood as protecting “freedom for religion”—religious liberty for the benefit of religion, for generous protection of its free exercise by individuals and groups, and for the autonomy of religious institutions. The Supreme Court’s most recent decisions appear headed in that direction.


Here’s a radical thought. The religious-freedom protections of the First Amendment exist for religionfor the benefit of religion, for its affirmative protection, for its liberal free exercise, for the autonomy of religious institutions, and for the purpose of enabling the pursuit of religious truth.

The choice of prepositions is surprisingly important. Religious freedom under the U.S. Constitution is certainly not about protecting individuals, society, or the state from religion. Just the reverse, it is about protecting religious belief and exercise from the competing commands of society or the state. Nor does the phrase “freedom of religion,” quite capture the essence of American religious freedom. It implies, subtly, indifference to religion, gentle skepticism, a posture of agnosticism: people may believe what they will, and government needs to be fine with that (to an extent).

What such a stance slights, I submit, is the original perspective underlying the First Amendment’s Religion Clauses: the conviction that there really is (or well may be) such a thing as “religious truth”; that the right to pursue such truth, and to hold on to it once one has found it, is of supreme value and ultimate concern and, consequently, worthy of the highest degree of protection by government; that the obligations of devotion to God, and of acting in accordance with God’s believed will are, more than just theoretically, of prior and superior obligation to the commands of the state and of human society; and that American governments are premised on recognition of this reality. In a nutshell, the vision underlying the First Amendment’s religious-liberty provisions is that freedom of religious belief and exercise is a fundamental, natural right that precedes the social compact of government and one with which government rightfully possesses no power to interfere. Thus, the commands of society and government in principle must yield to the true commands of God. This, uniquely, sets religious freedom apart from all other constitutional freedoms. Further, since the state, or any mere human authority, cannot be trusted to know and follow God’s will perfectly, the state lacks any rightful authority to prescribe matters of religious doctrine, belief, or conduct or to compel adherence to any religious creed. Government therefore must never dictate religious belief or conduct, and it must give the broadest berth possible to the sincere exercise of religious convictions by individuals and groups.

All of this is for the benefit of religion and religious persons and institutions.

The First Amendment protects freedom for religion—its free exercise by religious persons and groups; the freedom and autonomy of religious groups to make their own institutional judgments concerning doctrine, practice, membership, leadership, and governance; and the correlative freedom not to be compelled by government to observe or exercise any religious belief other than one freely chosen by the religious individual or group.

This is not—not yet, not quite—the theory of religious freedom embraced by the Supreme Court. But the Court may be closing in on such a vision. The Court, in its recent decisions in this area, has been, as it were, circling the runway of such a fundamentally religion-centric conception of religious freedom. But it has not quite landed on a coherent understanding of the First Amendment Religion Clauses. The Court is burdened by, but gradually shedding, the doctrinal clutter and unclarity of the past. Its decisions are beginning to home in on a new, corrected conception of religious liberty—a vision that can be seen in the emerging pattern formed by its recent decisions. That vision is that the Religion Clauses should not be understood as protecting a secular, functionally agnostic “freedom of religion” and certainly not understood as embodying a functionally antagonistic “freedom from religion.” Rather, the Religion Clauses are fundamentally concerned with promoting an openly supportive, welcoming, natural-rights-oriented “freedom for religion.”

In Part I of this Essay, I will sketch—in compressed form1—the straightforward case for this conception of religious liberty as a matter of First Amendment first principles. It is, I submit, the only conception that makes full sense of the reason for having a constitutional provision affirmatively protecting the “free exercise” of religion. It is the only conception fully consistent with the idea of religious liberty as a preconstitutional, literally God-given, “natural” right—the understanding of religious liberty possessed by the generation that adopted the First Amendment. It is the conception that best fits the premise that religious liberty is a substantive freedom, not merely a nondiscrimination rule. It is the conception that makes the most sense of the text of the First Amendment, understood in its historical context and in light of widely shared backdrop premises at the time.2

It is also the view that best accounts for the Court’s recent, important, and (mostly) rightly decided cases in this area and that best explains the flaws and limitations of the Court’s analysis and doctrines in those cases. In Part II, I will examine the Court’s recent religious-liberty decisions and examine how well they match up (or fail to match up) with this model.

I. religious liberty, for religion

A. The Sense and Sensibility of the Text

A useful place to begin is by asking what religious liberty is doing in the Constitution in the first place. Why protect religious freedom in particular? Why have a constitutional guarantee of the “free exercise” of religion? What is the overall sense and sensibility of the First Amendment’s religious-freedom provisions?

As I have argued at much greater length elsewhere, it is surpassingly hard to justify unique constitutional protection for religious liberty on strictly secular grounds.3 At the very least, it is hard to justify religious freedom in any strong sense—that is, as an affirmative, substantive freedom conferring some sphere of immunity from government’s usual powers. If one is indifferent to religion, it is hard to explain why protecting religious freedom would be a central concern of the Founding generation, worthy of so much attention and care in the language of the First Amendment. And if one is skeptical of religion or religious institutions, or outright hostile toward religion, the idea of giving special constitutional protection specifically to religious practice and conduct becomes almost impossible to explain. Put bluntly: if one starts from the premise that religious conviction does not correspond to anything “really” real—that religious faith is an idiosyncratic personal preference, like any other purely personal preference (except perhaps more suspect and less valued than other such personal preferences)—it would not make much sense to indulge religious practice by writing into the Constitution unique legal protection for “the free exercise thereof.”4

The most one might grant is a passive toleration of religion, at least where one considers it harmless to do so. Perhaps one might acquiesce to a provision forbidding state discrimination on the basis of religion. One might even favor adopting something resembling the Establishment Clause—a ban on official religious doctrines enforced by law—for reasons of “separation of church and state,” to protect society from undue religious influence, or as some kind of official legal “truce” among competing religious views or traditions.5 (As I note presently, this is not a proper understanding of our Establishment Clause and does not accurately reflect the reasons why it was included as part of the First Amendment’s package of protections for religious liberty.6 But it is at least possible to imagine a purely secular, even antireligious, reason for having a provision of such a sort.) The secular case for religious liberty can take one only so far.

But protection specifically of “the free exercise” of religion? Affirmative constitutional protection for religious practice, even when such practice is in conflict with other general secular legal norms? (That would seem to be what is embraced by the “exercise” of religious practice being protected as “free.”) That is exceedingly hard to justify on religion-skeptical, secular terms. If one is a committed secularist or atheist, why on earth embrace a constitutional provision explicitly and affirmatively protecting the “free” “exercise” of “religion”?7

There really is no good answer to this riddle. The only plausible answer is that the premise is wrong: it makes little sense to read the Religion Clauses from a secularist perspective in the first place. That is simply a wrong starting point. To read the Religion Clauses through modern or postmodern eyes of religious skepticism and nonreligious secularism is to read them anachronistically. Mark DeWolfe Howe put it well: “Though it would be possible . . . that men who were deeply skeptical in religious matters should demand a constitutional prohibition against abridgements of religious liberty, surely it is more probable that the demand should emanate from those who themselves were believers.”8

By far the more sensible explanation of religious liberty as a set of constitutional freedoms—and, thus, by far the more sensible lens through which to read and understand them—is that the Founding generation wished to protect religion because religious practice was valued, respected, and considered of supreme importance. The reason the Framers singled out religion for special constitutional protection was broad agreement both with religion’s intrinsic importance and with the political proposition that government should not have the power to interfere with, limit, regulate, manage, or control something so supremely important.9

In short, religious liberty was considered something that existed for religion—to protect its free and autonomous exercise, immune from either state impairment or state direction and control. The Free Exercise Clause does the former—protects private religious exercise. The Establishment Clause does the latter—assures that the private sphere of autonomy remains private and is not captured or coerced by government. That, not protecting society or the state from religious influence, was the reason for the Establishment Clause. The two Religion Clauses are, as it were, two sides of the same coin, protecting a single common value of religious liberty from two different directions. It makes no sense of the text, taken as a whole and in its social and linguistic context, to read the Religion Clauses as a semicontradiction: a “proreligion” Free Exercise Clause balanced or mitigated by an “antireligion” Establishment Clause counterweight.10

As a matter of textual logic, historical context, and common sense, the Religion Clauses of the First Amendment only make full sense when understood as having been written against the backdrop of a series of widely shared essentially religious premises. As I have expressed the point in other writing:

Religious freedom only makes entire sense as a social and constitutional arrangement on the supposition that God exists (or very likely exists); that God makes claims on the loyalty and conduct of human beings; and that such claims, rightly perceived and understood, are prior to, and superior to, the claims of any human authority. Simply put: God’s commands—God’s will, God’s purposes—rightfully trump man’s. Freedom of religion, understood as a human legal right, is government’s recognition of the priority and superiority of God’s true commands over anything the state or anyone else requires or forbids.11

If that is an accurate description of the premises underlying American constitutional protection of religious liberty, a great deal follows. It follows from these premises that the pursuit of religious truth and right relationship with God was regarded by the Constitution’s drafters as something of huge, fundamental importance—arguably the most important human pursuit. It follows by implication, that religious conviction is singular and unique in its presumed entitlement to override the claims of any mere human authority, including government. Religion is special in this regard; it is not like other strongly held personal views, political opinions, or anything else.12 Government power must in principle yield to religious conscience (and only to religious conscience, not to secular moral scruples or views). To be sure, human perception and understanding of God and God’s true commands is necessarily imperfect. But this does not mean that government has the right to override religious convictions whenever government disagrees with them or thinks them wrong. Just the opposite. Government is imperfect too—the state is a demonstrably fallible human institution and a dangerously powerful one—and is therefore not to be presumed superior in its ability to discern religious truth. Quite the contrary, the state is to be regarded as incompetent on both theological and political grounds to determine what God does or does not require.13 Different individuals and different religious communities inevitably disagree as to what loyalty, fidelity, and obedience to God may require, as regards religious practice and right human conduct. But the logic of the First Amendment is that these differences must be respected by government and that those differences do not in the least serve to justify government in overriding religious freedom.14 The state simply has no legitimate power to judge such matters. Government may not judge the correctness or extent of religious obligation; it may not regulate or referee matters of religious faith and conduct; and it may not “establish” official religious views or in any fashion prescribe and coerce religious exercise, worship, or belief. Rather, the state must acknowledge and recognize individual and group religious freedom as a natural and inalienable right, a literally God-given entitlement that precedes the social compact and is not surrendered by entering into civil society.15 Religious liberty is not a right conferred by government or judged by government. It is, rather, a fundamental prior limitation on all government power.

This is, I submit, the theory on which American constitutional protection of religious liberty rests.16 It is the theory that best coheres with the original meaning of the words of the text of the Religion Clauses.17 It is the only theory that fully credits the backdrop social and linguistic understandings of the clauses’ words and terms, that gives due weight to the religious society and religious premises that produced the text, and that does not read its terms anachronistically through modern or postmodern glasses.18 It is the theory that best accounts for the history of the Religion Clauses and the social and cultural movements that pressed for their inclusion in the Bill of Rights.19 It is the theory that best explains the decision to have a constitutional religious-liberty provision. And it is the more natural and sensible reading of the provisions we have, taken in their historical and linguistic context.

The Religion Clauses are well designed to achieve the twin purposes of affirmatively protecting religion and disabling government from interfering with its exercise in any way, direct or indirect. They are rather poorly designed if imagined to be an engine of secular skepticism, agnosticism, or suspicion. (Again, the latter reading is hard to reconcile with having a “free exercise” provision as a distinct, separate constitutional freedom in the first place.) The Establishment Clause and the Free Exercise Clause are logically read as side-by-side, harmonious, companion protections of a single liberty. They should not be read as being in tension or contradiction. The two clauses work in tandem: free exercise forbids government from keeping persons from exercising their freely chosen faith; nonestablishment forbids government from making persons exercise a religious faith other than that of their own choosing. Other accounts of the Religion Clauses might explain certain aspects or features of religious liberty: religious liberty as a more limited principle of nondiscrimination (but not a substantive freedom); or as a principle of government neutrality or noninvolvement with religion; or religious liberty as freedom of the state from religion. However, each of these theories ignores some essential feature of the text, is historically anachronistic, is limited in its explanatory power, implausibly suggests a textual self-contradiction, or all of the above.20

As I suggested at the outset, this idea that the Religion Clauses rest on essentially religious premises might sound almost radical to many constitutional scholars today. It is in some respects a lost perspective, so much so that it might not even occur to some modern readers as a viable possibility. Yet if the task of constitutional interpretation is to ascertain correctly and apply faithfully the original meaning of the constitutional text—a proposition some dispute, but one I believe is a necessary first principle of written constitutionalism21—then restoring that lost perspective is the first step to correcting sound adjudication in this area of the law.

B. Reading the Religion Clauses “Faithfully”

The vision of First Amendment religious freedom as being fundamentally for the positive benefit and protection of religion yields some important doctrinal conclusions concerning the proper understanding of the Religion Clauses.22

First, reading the Religion Clauses from the perspective of their original religious premises very strongly suggests that the “free exercise” of religion is an affirmative substantive freedom and not a mere nondiscrimination rule. The Free Exercise Clause is not just about what government targets; it is about what it hits. To the extent the clause might be thought ambiguous in this regard, the idea that religious freedom is for the benefit of the religious believer suggests that the text should be viewed and uncertainty resolved from the believer’s perspective. And viewed from the standpoint of the religious adherent—rather than that of an indifferent government official reading and enforcing stated rules without regard to their effect on religious exercise—a requirement of law “prohibit[s]” an individual’s or institution’s free religious exercise when it operates to punish, penalize, or prevent the religious adherent’s religious conduct.23

The Free Exercise Clause, understood from the perspective of the affected religious adherent, is thus not merely a requirement that government not treat religious practice any more harshly than other comparable conduct, or that it not single out religious exercise or belief for special disabilities. (It is a requirement of nondiscrimination, too, of course, practically as an a fortiori case: if government may not impair the preexisting autonomy of religious individuals and institutions to act in accordance with their faith, it surely may not punish or penalize such groups or persons by subjecting the exercise of religion to unique, or specially targeted, burdens, barriers, or exclusions.24) But when the First Amendment bars government action “prohibiting the free exercise” of religion, it does not mean only that government may not discriminate against religion, single it out for exclusion or disfavor, ban religious activity because it is religious, or otherwise act out of special hostility to religion. That much is a given.

If that were all the “free” “exercise” of religion encompassed, the Constitution’s language would seem inapt as a way of formulating so limited a prohibition on interference with a right. Rather, the clause more naturally connotes that government power must not be used in such a way as to defeat or impair the prior right of free religious exercise. This follows from the nature of religious freedom as a fundamental, preconstitutional natural right: the powers of government come with a reservation and preservation of prior rights in favor of the fundamental freedom of religious exercise in accordance with individual and institutional religious conscience. Freedom for religion means that religion comes first; government comes second. The right to freely exercise freely chosen beliefs confers a sphere of constitutional immunity from government regulation or burdens on religious believers.

As a doctrinal matter, this means that the narrow reading of the Free Exercise Clause, adopted by the Supreme Court in 1990 in the case of Employment Division, Department of Human Resources v. Smith is wrong.25 The so-called “exemptions” reading of the Free Exercise Clause,26 the reading that recognizes that bona fide religious exercise is a constitutional exception to the powers of government, is the correct one.

To be sure, that only gets you so far. There remain serious questions as to where the precise boundaries of this exception fall—whether and when there might be exceptions to the exceptions that the Constitution recognizes in favor of sincere religious exercise. But that is the relevant question. The question is not whether the First Amendment protects the free exercise of religion as a substantive freedom from government regulation as a matter of first principles. It does.

Second, for essentially all the same reasons, the assessment of whether government actions “burden” religious liberty in a cognizable way—so as to count, so to speak, as “prohibiting” free exercise—again should be viewed from the perspective of the religious adherent. Specifically, this assessment should adopt the religious adherent’s sincere understanding of his or her beliefs and of what constitutes a sufficient burden on them to amount to prohibiting their free exercise. Government is not the rightful judge of whether its law has a sufficient effect on the exercise of a sufficiently “important” religious tenet, requirement, or principle. That is a question of what the individual’s (or institution’s) exercise of faith involves and when its free exercise is affected—and that is a question for the religious adherent.27 Government has no right to “balance” religious exercise against government’s burdens on it. The very idea of “free exercise” is that the balance is drawn by the individual’s sincere understanding of his or her own faith—what it requires, forbids, or counsels.28

A third doctrinal corollary of understanding the First Amendment as protecting freedom for religion is the idea of religious institutional autonomy. The right of religious institutions to freely exercise religious convictions parallels and builds upon the right of individuals to do so. The free exercise of religion can occur in groups, communities of faith, and other institutional arrangements. It doesn’t have to, of course: the right to the free exercise of sincere religious convictions does not require association with a specific religious denomination, institution, or organization. Nor is it limited to the exercise of religious beliefs counted as formal “tenets” by any religious body or shared by all members of a particular group.29 But given that the exercise of religion can occur in or by groups, religious liberty has an inescapable institutional-autonomy aspect as well. A religious group gets to define its religious identity as a group, prescribe its own tenets, beliefs, practices, mission, and internal governance rules, and limit its leadership and its membership to individuals who share the mission, identity, and purposes of the group as defined by the group—all free from government interference or control. Freedom for religion means religious institutional autonomy.30

Fourth, and importantly, the view of religious liberty I have described implies a basic, obvious, but strangely neglected limitation on what is covered by this freedom: constitutional protection for the free exercise of religion is limited to religion. If religious liberty under the First Amendment is truly about religious liberty—that is, if the freedom is a freedom for religion, specifically—it means that religious exercise and conscience are constitutionally privileged over nonreligious, secular, personal, or philosophical views. It also means—what amounts to pretty much the same thing—that the constitutional meaning of “religion” has a narrower compass than purely personal, not-actually-religious views. Put bluntly: religion means religion—principles and obligations flowing from a theistic belief in a God (or gods), outside of oneself, not from one’s own internal nonreligious convictions, however sincere they might be.31

This is less generous a conception of “religion” than has animated the modern Supreme Court’s jurisprudence, to be sure. The Court has been soft on the constitutional meaning of religion, abandoning its original sense in favor of a more fuzzy, flexible, functional modern view closer to whatever-an-individual-believes-deeply. This view, reflected in a series of military-service conscientious objector cases in the 1960s and 1970s,32 was perhaps driven by noble liberal instincts. But by watering down the Constitution’s original understanding of “religion,” it has tended ultimately to weaken the Free Exercise Clause’s protection of actual religious freedom by trying to make it protect everything in the world. Broadening the freedom to include everything thins it out, making the freedom itself diminish to the vanishing point. If religious freedom in principle carves out a sphere of substantive immunity from the usual rules enacted by government, a rule of everything-is-religion becomes an intolerable rule of everybody-can-do-whatever-they-want. Just such a fear appears to have fueled the Court’s abandonment in Employment Division v. Smith of the idea of free exercise as such a substantive freedom.33 As the Court put it a century and a half ago, such a rule would make every man “a law unto himself.”34 And that just can’t be. Thus, if religion embraces everything, religious freedom must protect nothing, or very little. If religion is everything, religion is nothing.35

The right answer, and the only sustainable answer, is that religious freedom is limited to religion as understood in its original sense.36

A fifth doctrinal implication concerns the enduringly thorny problem of when a “compelling state interest” overrides religious-freedom claims: if the right of religious free exercise is premised, ultimately, on the notion that God’s true commands and requirements have priority over conflicting commands or requirements of the state, that view establishes boundaries on when courts and other decision makers can justifiably override a claim of religious liberty. Such a claimed trump over religious freedom is usually premised on the notion that the right of free exercise is (impliedly) limited by a (silent, so far as the text is concerned) exception in favor of (assertedly) “compelling” interests of the state.

All theories of religious liberty must wrestle with the problem of extraordinary harm perpetrated in the name of religion. What if a claimed right of religious conduct would impose or threaten essentially intolerable injury to the baseline natural human rights of others or destroy civil society? And all theories asserting that the state sometimes may override claims of religious liberty—on the grounds of indispensable necessity to avert intolerable harm—must wrestle with the problem of the seemingly silent text on this point and a lack of principled constitutional criteria justifying such overrides. Where does such override authority come from?37

Present doctrine does not supply a satisfactory set of answers to these questions. On this point, I submit that taking the premises on which religious freedom rests seriously, and taking equally seriously the problem of the seemingly silent text, suggests that the inquiry should not focus on what the state considers sufficiently important to override religious freedom. Such an approach impermissibly implies that the state is ultimately the supreme arbiter of religious freedom. Instead, the inquiry must focus on what plausibly can be counted as falling within the scope of free exercise of religion” in the first place. That is to say, whether a claim of religious autonomy should lose is properly a matter of what “free exercise” covers—intrinsic limitations, formed by an understanding of the meaning and scope of the Free Exercise Clause’s language—rather than an extrinsic trump on religious freedom on terms created by, defined by, and imposed on an ad hoc basis by an ultimately sovereign state.

A “compelling state interest” is, in this understanding, something of a doctrinal misnomer. In the context of a freedom premised on the ultimate sovereignty of God’s requirements over man’s, the proper focus is on what can plausibly be claimed in the name of God as the legitimate exercise of “religion” and how far such claims legitimately can extend as a matter of first principles, not on when the state is entitled to win on its own terms.38 This change in perspective might end up producing some of the same conclusions as the traditional formulation. But it would likely alter many others.39 Taken seriously, viewing “compelling” overrides as a function of the intrinsic meaning and limits of the “free exercise” principle, rather than as creatures of government policy, moves the fulcrum decidedly in the direction of greater protection of religious liberty.

Sixth, and finally, freedom for religion as an orienting metaphor strongly supports what is in any event the most straightforward textual and historical understanding of the Establishment Clause: that it is a cognate provision protecting religious liberty, fully harmonious with the Free Exercise Clause and not a competitor of it. The Establishment Clause is not at all a religion-fighting principle of constitutional exclusion of or discrimination against religious persons, groups, or views in public programs or benefits. Rather, it is a requirement that government not compel or prescribe the involuntary—the unfreeexercise of religion. The “separation of church and state” that it achieves is not the expunging of religious influence from society or politics but the expelling of government control over matters of religious faith.40

II. the court’s recent cases

How well do the Supreme Court’s more recent religious-liberty cases match up with this model of “freedom for religion” and its doctrinal implications, sketched in Part I? How has the Court been doing in this area over the past decade? Might this, or something like it, be what the Court has been up to in its Religion Clause decisions? In this Part, I take up these questions, and briefly consider some of the Court’s recent leading cases.

It would be going too far to claim that the model presented in Part I has been explicitly the Court’s conception and deliberate direction. But that model may provide a useful lens through which to view certain recent decisions and trends. It explains some, even many, of the Court’s most significant recent religious-freedom cases. More broadly, it provides a plausible explanation of the emerging pattern formed by a seeming kaleidoscope of separate cases. In what follows, I discuss four such apparent patterns.

First, there is the long series of decisions—now dating back several decades—forbidding government exclusion of religious persons or groups from government-benefit programs or expressive forums. These decisions are driven by vital principles of the Free Exercise Clause (and by like principles of the Free Speech Clause) forbidding discrimination based on religious views, identity, or expression; as a corollary, they reject certain unsound readings of the Establishment Clause that had prevailed a half century ago.

A second trend has not yet reached its conclusion: the seeming movement in the direction of abandoning the narrow reading of the Free Exercise Clause adopted in Employment Division v. Smith in 1990 in favor of restoration, in some form, of the more liberty-protective view that the First Amendment confers a substantive freedom for religion—a sphere of immunity or exception from the usual power of the state.

Third, there is the relatively recent line of decisions that recognize the right of religious institutions and groups to decide for themselves, free from government regulation or interference, matters of their own ministry, leadership, personnel, and the standards of conduct required of members. These decisions involve key interpretations of both the Free Exercise Clause and the Establishment Clause—and themselves provide support for abandoning the Smith rule.

Fourth, and overlapping the preceding categories, there is the series of decisions rejecting the unsound and ahistorical reading of the Establishment Clause derived from the infamous “Lemon test,” devised in the Court’s 1971 decision in Lemon v. Kurtzman.41 The Lemon test had sometimes been applied to forbid any government action or policy that produced a direct or indirect effect advancing, accommodating, or even merely endorsing religion. In place of Lemon, the Court has been moving (albeit uncertainly and awkwardly) in the direction of restoring the original understanding of the Establishment Clause as forbidding government prescription or compulsion of religious exercise. These decisions have involved everything from public prayer opening government meetings42 to government displays of religious symbols and monuments43 to, most recently, a public high-school football coach’s public practice of personal postgame prayer at the fifty-yard line.44

Not all of these trends are new to the past decade or so. Some are continuations, or extensions, of principles with long pedigrees. Other trends are incomplete and might not fully and finally resolve in the anticipated direction described. Still others are harder to pin down clearly.

A. The Common Core of Free Exercise: No Discrimination Against Religion

The first principle is clear and the Court has been clear in embracing it: The First Amendment forbids government discrimination against religion. Period. Specifically, the Free Exercise Clause at bare minimum forbids exclusion from, or discrimination against, religious persons, institutions, or views in government programs or forums. The First Amendment’s Free Speech Clause stands for the identical proposition whenever expression or expressive conduct is involved or when government predicates an exclusion from a program or forum on a participant’s religious identity, purpose, viewpoint, or speech.45 No fewer than four decisions of the Court in the past six years—and arguably several more—embrace this principle. And they do so strongly, clearly, unequivocally, and emphatically.

The decisions in Carson v. Makin,46 Espinoza v. Montana Department of Revenue,47 and Trinity Lutheran Church of Columbia, Inc. v. Comer48 all stand for a version of this core proposition. Carson and Espinoza both hold, directly and explicitly, that government may not exclude or discriminate against private religious education in programs of government funding of private school choice or in the provision of other education benefits or services to students and their families.49 Nor may government exclude religious institutions (including churches and church schools) from participation in government programs for which they are otherwise eligible, because of their religious affiliation, expression, identity, or mission. Trinity Lutheran so holds, explicitly, as does, after a fashion, Fulton v. City of Philadelphia—an important, paradigmatic illustration of this principle.50 Trinity Lutheran involved eligibility for state grants to fund playground resurfacing and held that a church-affiliated preschool could not be excluded because of its religious associations, identity, and programs. “The Free Exercise Clause,” the Court held, “‘protect[s] religious observers against unequal treatment’ and subjects to the strictest scrutiny laws that target the religious for ‘special disabilities’ based on their ‘religious status.’”51

Fulton v. City of Philadelphia is a dramatic recent case illustrating this principle in a distinctive factual situation. The essential holding of the case was that government may not exclude a religious social-service group’s participation in a government program for provision of foster-care services because of the group’s religious scruples against placing children in homes of unmarried couples (including both same-sex and opposite-sex unmarried couples), where the government’s policy with respect to eligibility criteria was not genuinely and reliably “neutral” in its application. Catholic Social Services (CSS), a program participant, held such religious views against placement of children in homes of unmarried couples, as a matter of conscientious conviction. The City of Philadelphia contended that CSS’s views violated the City’s nondiscrimination rules. Specifically, the City asserted that allowing CSS to participate violated the city’s bans on sexual-orientation and marital-status discrimination by participating private social service agencies. Chief Justice Roberts’s majority opinion concluded that the City’s policies were not reliably neutral and “generally applicable” in this regard. That was because the city vested in administrative officers the power and discretion to make exceptions to the policies’ general rule. Had the policies in fact been generally applicable, and not subject to ad hoc exceptions—and assuming the city’s policies did not in some other way specifically target religious beliefs or institutions for exclusion from the program—the City’s exclusion of CSS almost certainly would have been upheld under the rule of Employment Division v. Smith.52 However, because the city had vested discretion to make exceptions to its policies in a government official—an official empowered to consider and assess the reasons for CSS’s noncompliance, including its religious reasons—the City’s decision to exclude CSS became subject to strict scrutiny. The Court held that Philadelphia’s policy and practice did not pass such scrutiny.53

The judgment in Fulton was unanimous: Philadelphia’s exclusion of Catholic Social Services was unconstitutional. Justice Alito’s thorough, seventy-seven-page opinion concurring in the judgment only, joined by Justices Thomas and Gorsuch, would have gone further and overruled Smith as being contrary to the original meaning of the Free Exercise Clause.54 There is much to be said for that position—and the Court may be on the verge of adopting it. (I discuss that possible, even likely, future development presently.55) Yet the Fulton majority’s narrower conclusion was significant in its own right: The use of government power so as to exclude religious persons, groups, or organizations from participation in a public program is subject to strict scrutiny, and presumptively unconstitutional, if that power is used or is even capable of being used specifically to exclude religious views and religious conduct. That is an important principle. Religious identity, views, or conduct cannot be the basis for exclusion from a government program or benefit. If the government’s arrangement would in theory permit its officials potentially to act on such a basis, it is not a genuinely and reliably religion-neutral program.56

A version of the government-must-not-discriminate-on-the-basis-of-religious-conduct-or-religious-motivation principle also helps explain, in part, the Court’s shift, midpandemic, in the COVID-19 emergency government shutdown order cases. In cases decided earlier in the pandemic, a sharply divided Court had upheld, over vigorous dissents, state and local orders forbidding or restricting the size of indoor religious worship group gatherings, even though certain commercial businesses had been permitted to keep operating with less stringent occupancy caps or limitations on outdoor activities.57 Later, as experience (arguably) changed, and as the Court’s personnel (definitely) changed, the Court reversed course, embracing strongly the principle that religious worship and assembly must be treated at least as favorably as any other comparable private activity and be subjected to shutdown orders only when those comparable businesses or activities are likewise so subject.58

Finally, one can see this nondiscrimination principle at work in a rather different context, in the Court’s 2022 decision in Kennedy v. Bremerton School District, where the Court upheld the right of a public high-school football coach to kneel for personal prayer, at midfield, after the conclusion of games.59 The Court held that the Free Exercise Clause and Free Speech Clause protect the right of individuals, including government schoolteachers and other school employees, to engage in personal religious observance, free from punishment or prohibition, during times and in settings where employees are or would be permitted to engage in personal conduct of a nonreligious nature.60 Again, the principle is nondiscrimination: government must not discriminate on the basis of the religious nature or viewpoint of an employee’s expression or other activity.

The principle embraced in Kennedy is the same principle powering the outcomes in Carson, Espinoza, Trinity Lutheran, Fulton, and Tandon: government may not discriminate against religion in its programs, policies, or practices. The Free Exercise Clause might well require more than this. But it certainly requires no less. Each of these cases stands for the proposition that the Free Exercise Clause, at minimum, prohibits government discrimination against or exclusion of religious persons and groups, in government’s regulations, in its administration of public programs (including education funding), and in its provision of forums for engaging in expressive activities.

The principle itself is hardly new or novel. It is at least as old as Widmar v. Vincent,61 where the Court held that a state university could not exclude a student religious group’s use of campus facilities that it made generally available to student groups, on account of the religious content of the group’s meetings. The Establishment Clause did not authorize such exclusion or discrimination (an important principle itself), and the Free Speech and Free Exercise Clauses did not permit it.62 The Widmar principle—a principle of equal access for religious speech or expression in a limited public forum provided by government—has been restated, reaffirmed, and applied in various forms more than a half-dozen times in the decades since Widmar:63 It has been applied to public-school districts’ rental of facilities to private community groups;64 to high-school student group activity periods;65 and to after-school, on-premises elementary-school programs run by outside religious groups.66 The rule has by now become thoroughly well established and essentially uncontested, cemented by the Court’s many recent cases reiterating the principle.67 Government may not in its programs discriminate or exclude on the basis of religion, religious expression, religious mission, or religious identity. That rule lies at the very core of the Constitution’s protection of freedom for religion.68

What is arguably new in recent years is the Court’s willingness to extend and apply that familiar principle to less familiar, less intuitively obvious situations. Nondiscrimination against student religious groups seeking equal access to meeting space at state universities—the situation presented in Widmar—was a relatively easy and straightforward application of the principle. Extending that principle to public high-school student clubs was a short and seemingly simple next step. So too with equal access for community religious groups offering after-school elementary-school programs on the same terms as other community groups, and with religious groups’ eligibility to rent public-school facilities, outside regular school hours, for religious meetings or worship services.69 The principle of nondiscrimination against religion transposed readily from one situation to another. In each instance, the principle was applied in a setting easily recognized as involving the right to engage in speech, in a public “forum,” without discrimination on the basis of the religious content of the expression involved.

It takes a bit more imagination to conceptualize education funding programs, government grant programs for school playground resurfacing projects, and eligibility criteria for participation in provision of foster care, as involving the same principle of “equal access” to a “public forum,” without discrimination on the basis of religion. Not much more imagination, perhaps. But it does require a willingness to take the small intuitive leap from a literal to a more metaphorical “forum” in which religious individuals and groups possess the same right to inclusion without discrimination. The Supreme Court in recent years has been willing and able to make that analytic leap in settings where the Court previously had not yet been prepared to do so.

B. Free Exercise as an Affirmative Freedom: Scrapping Smith Soon?

The principle of nondiscrimination against religion is easy. The somewhat harder, more contested proposition is that the Free Exercise Clause confers a sphere of substantive immunity from government regulation, including from even ostensibly neutral and generally applicable rules.70 The position is simply this: not only must government treat religious belief and exercise no less well than it treats comparable nonreligious conduct, but it must, as a constitutional rule, go further and protect religious liberty independently of how it treats other activity. The Free Exercise Clause is an affirmative protection of freedom for religion, carving out religion’s free exercise from the usual powers of government.

If this is correct—and the Supreme Court appears to be leaning in this direction—it follows that its 1990 decision in Employment Division v. Smith was wrong and should be overruled. The Court came to the edge of so ruling in Fulton, stopping just short of a direct ruling to that effect. Justice Alito’s concurrence, joined by two other Justices, would have taken that further step.71 In addition, two other Justices expressed sympathy with overruling Smith.72 A majority of the Court thus appears poised to take that next step, perhaps soon. The only questions giving the Court pause were what weight to give Smith as a matter of stare decisis—a question the Justices have wrestled with in other important contexts, notably in Dobbs v. Jackson Women’s Health Organization, overruling Roe v. Wade’s creation of a constitutional right to abortion73—and what standard should replace Smith if it is repudiated. Other than that, it appears to be a matter of the Court waiting for the right occasion to overrule Smith.

The Fulton majority opinion, written by Chief Justice Roberts, reflected the Chief Justice’s preference for incrementalism and narrowest-basis-possible decisions.74 But, as set forth above, the “narrower” ground there embraced—the nondiscrimination principle—was itself set forth in a fairly aggressive form. Given the vigor with which the Court’s decision in Fulton (and before that in Tandon) interpreted and applied Smith’s preconditions, it might be difficult for the Court to find a case posing a square conflict with Smith. Overruling Smith, if it is to occur, thus might require a case where the government’s action impairing religious liberty is unquestionably “neutral” as between religious and secular acts, treats religious conduct fully as well as any similar, permitted secular activity,75 involves a truly “generally applicable” rule, and provides for no exceptions or discretion as to its application. That might be a long wait.

Justice Barrett’s brief, important concurring opinion in Fulton, written on behalf of herself, Justice Kavanaugh, and Justice Breyer, appeared ready to repudiate Smith in a case requiring such a step: “In my view, the textual and structural arguments against Smith are more compelling [than the historical argument]. As a matter of text and structure it is difficult to see why the Free Exercise Clause—lone among the First Amendment freedoms—offers nothing more than protection from discrimination.”76 That’s a strong statement that Smith’s overripe rule is ready to fall. But the harder question for Barrett (and Kavanaugh) was “what should replace Smith?”77 Where burdens on religious exercise arise from truly neutral and generally applicable rules, Barrett expressed skepticism about swapping Smith’s categorical rule for an equally categorical return to “strict scrutiny,” given that the Court’s jurisprudence concerning other First Amendment rights, like speech and assembly, does not treat incidental or indirect infringements arising from neutral laws as necessarily subject to the highest standard of scrutiny.78

The short answer to this concern, I submit, is that the free exercise of religion is a different freedom from freedom of speech, with different parameters and its own independent standards. The Free Exercise Clause will sometimes duplicate the Free Speech Clause in what it protects, but sometimes not: it is a separate, independent freedom that frequently goes beyond what free speech requires. Free exercise is, as it were, its own thing. While doctrines employed in free-speech jurisprudence may in certain respects be analogous and instructive, they are not determinative. Free exercise is a different freedom, with its own requirements, just as, say, the Second Amendment and the Fourth Amendment are different freedoms with their own separate requirements, rules, doctrines, and standards.79

Justice Barrett posed some further questions to be answered before the Court could safely announce a replacement to Smith. I will offer here some brief answers to Justice Barrett’s questions. Should burdens on religious entities be treated differently from burdens on individual free exercise?80 No. For reasons stated above, free-exercise rights may be exercised by groups but are not limited to institutional prerogatives or to religious beliefs shared by all members of a formal church, denomination, or other organization. The next question Barrett posed was, “Should there be a distinction between indirect and direct burdens on religious exercise?”81 This is genuinely contestable, but the better answer is again no, if one adopts the perspective of the religious believer: as noted above, if the government’s action burdens the adherent’s freedom to exercise sincerely held beliefs, that should be sufficient to trigger the constitutional right. The Free Exercise Clause does not weigh relative degrees or directness of burdens on religious exercise.82

Finally, if “strict scrutiny” applies, what form should it take, given the pre-Smith Court’s different formulations in different cases?83 And if strict scrutiny applies, would pre-Smith cases all come out the same way as they did?84 The question of “overrides” on religious liberty is the fifth proposition I offered above as to the implications of viewing the First Amendment as a protection of freedom for religion.85 This is indeed an area that requires the Court to engage in substantial further thought and reconsideration; Justice Barrett’s question is a good and important one. Suffice it to say here that such overrides, which arise only by textual implication, should be rare, disfavored, and strictly limited to cases of essentially intolerable harm to the baseline fundamental private rights of others outside the religious community. Even under the Court’s pre-Smith “strict scrutiny,” the standard applied was not strict enough and many of the Court’s applications were arguably wrong.86

These are questions that the Supreme Court should and must address if, and when, it overrules Smith. But the important point is that these questions have answers. The Court stands at the threshold of overruling Smith and should do so, just as Justice Barrett notes, to vindicate the constitutional logic and necessary textual meaning of the Religion Clauses as affirmatively protecting freedom for religion. It is now properly a matter of the right occasion and getting the details of Smith’s replacement right.

C. Freedom for Religious Institutions’ Internal Autonomy

As argued above, freedom for religion includes at its core the right of religious groups and institutions to define themselves—to determine the precepts that define the religious community, to prescribe and maintain standards of belief and conduct to which its members are expected to adhere, and to freely make decisions concerning leadership and governance of the faith community.87 The Court has been doing remarkably well on this score over the past dozen years. On the autonomy of religious institutions in matters of internal doctrine, discipline, leadership, and teaching, the Court has, since 2012, embraced a strong, clear vision of religious liberty.

The groundbreaking case was Hosanna-Tabor Evangelical Lutheran Church & School v. EEOC in which the Court unanimously held that the First Amendment forbids any government interference with the employment relationship between a religious body and those it, in good faith, considers its “ministers”those leaders, teachers, and others who, in the words of the Court, “personify” the beliefs of the religious community.88 The decision embraced, in sweeping terms, the right of religious groups to exercise autonomy in matters of “internal governance” and “control over the selection of those who will personify its beliefs.”89 It specifically affirmed “a religious group’s right to shape its own faith and mission through its appointments.”90 The Court made clear that, although earlier lower-court decisions had given this doctrine the somewhat inapt label of “the ministerial exception,” the principle involved was general and firmly established by the First Amendment Religion Clauses: a religious group has the “right to shape its own faith and mission” including through its decisions concerning the selection and retention of teachers and others who speak for or represent the religious community.91 It is clear from the Court’s opinion that this right of autonomy embraces more than just the hiring of “ministers,” narrowly defined.

This principle, the Court held, excluded government power to evaluate decisions by a religious school concerning the employment of teachers involved in carrying out and communicating the religious beliefs, values, and mission of the school.92 Even where government purports to be applying a neutral law of general applicabilitythe cases where Smith had held the Free Exercise Clause not to restrict state powerthe autonomy of religious institutions prevailed. Thus, Hosanna-Tabor seriously conflicted with and undermined Smith from the outset, even though the Court sought to distinguish the cases on the theory that Smith concerned “outward physical acts” rather than decisions concerning a religious group’s internal administration, leadership, and employment. Even where government may regulate employment or leadership decisions of nonreligious organizations, it may not regulate a religious group’s selection of its leaders and exemplars.

The Court grounded its holding in the seemingly Smith-defying proposition that “the text of the First Amendment itself . . . gives special solicitude to the rights of religious organizations.”93 This is important, and bears emphasis: the Court held, unanimously, that the Free Exercise Clause provides additional and independent rights to religious organizations beyond those to which nonreligious groups are entitled as a matter of the freedoms of speech and association.94 Recall the Court’s language, quoted above, rejecting the proposition that religious groups’ rights of expression and association should be controlled by the same standards applied to secular groups as a matter of Free Speech Clause doctrine: “We cannot accept the remarkable view that the Religion Clauses have nothing to say about a religious organization’s freedom to select its own ministers.”95

Our Lady of Guadalupe School v. Morrissey-Berru96reaffirmed and extended Hosanna-Tabor, making doubly clear what Hosanna-Tabor already said in a reasonably straightforward fashion. Although sometimes dubbed a “ministerial exception,” the rule of Hosanna-Tabor did not depend on formal titles, ordination, or specific training. It extended to elementary-school teachers whose roles called upon them to engage in religious instruction, prayer, and communicating religious faith in other ways. “What matters, at bottom, is what an employee does.”97

There remains an outlier pre-Hosanna-Tabor decision of the Court that is at odds with the principles of Hosanna-Tabor and Our Lady of Guadalupe—and with the principles of Widmar v. Vincent98concerning the rights of campus student religious groups. In Christian Legal Society v. Martinez, the Court held, by a vote of 5-4, that a Christian student group at a public university did not have the right to maintain its distinctive religious identity and beliefs by prescribing a statement of faith to which leaders and members were asked to subscribe and adhere. On this basis, the Court upheld the exclusion of the student religious group from a state university forum.99 Taken seriously, Christian Legal Society was a rejection of the freedom-of-speech and expressive-association rights of campus religious groups, contrary to Widmar. In addition, it was a rejection of the Free Exercise Clause rights of religious groups to determine their own identity, leadership, and message in the public-campus context, contrary to Hosanna-Tabor and Our Lady of Guadalupe.100

Simply put: Hosanna-Tabor and Our Lady of Guadalupe hold that government may not regulate or restrict the right of religious groups, institutions, and communities to make religious decisions about the standards of faith and conduct required of their leaders, exemplars, and members. Christian Legal Society had held—just two years before Hosanna-Tabor embraced that unanimous proposition—that government (in the form of a state university law school) may impose precisely such regulations or restrictions on religious groups, as a state-imposed condition on what would otherwise be a constitutional right under the First Amendment. 101 Those two positions are irreconcilable. The decision in Christian Legal Society is drastically out of line with the subsequent decisions in Hosanna-Tabor and Our Lady of Guadalupe. The decision in Christian Legal Society has become a derelict in the constitutional law of religious freedom—an erroneous remnant of a discarded perspective that would reject the freedom of religious groups to define their own identities and the rules of conduct that govern their internal membership and leadership practices, or that would condition participation in a public program or forum on forfeiture of such religious freedom. If Hosanna-Tabor and Our Lady of Guadalupe are right—and they are—Christian Legal Society is wrong and should be overruled.

D. Correcting Establishment Clause Interpretation

On the meaning of the Establishment Clause, the Supreme Court has reached generally sound outcomes in recent years, albeit sometimes with problematic reasoning. Nonetheless, the Court appears to be bearing down on the right answer: that this aspect of religious liberty forbids government coercion in matters of religious belief or exercise, thus protecting the freedom not to exercise religion, except as freely chosen.102 This—not “history” or “tradition,” as is sometimes invoked by the Court in its recent decisions103—is the best explanation for the Court’s cases refusing to enjoin the display of religious symbols (American Legion v. American Humanist Ass’n),104 the saying by individuals of voluntary prayers at government meetings (Town of Greece v. Galloway),105 and, with some awkwardness, a public high-school football coach’s postgame religious practice, sometimes joined voluntarily by students, of public prayer on the fifty-yard line (Kennedy v. Bremerton School District).106

The reason why a community’s display of a large cross as part of a war memorial—the situation at issue in American Legion—does not violate the Establishment Clause is not that the cross has been there a long time (as the Court’s opinion sometimes suggests107). It is because the symbol does not coerce religious exercise. If it did, the memorial’s age surely should not save it. Longstanding violations of the Constitution are not more supportable than new ones. Rather, the constitutionality of the cross hinges on the fact that mere speech, including government speech and symbolic expression, is not itself coercive.108 This is a basic principle of First Amendment free-speech doctrine, and it applies to religious speech the same as expression of any other content. So, too, does it apply to the practice of rotating, private, voluntary invocations at the opening of government meetings.109 As long as no one is forced to participate, and individuals remain free to absent themselves from (or simply decline to participate in) the religious aspect of the proceedings, the Establishment Clause is not violated. It is not a question of tradition; it is a question of coercion. While some might reasonably question the usefulness and etiquette of government’s use of religious symbolism or ceremony as a matter of good policy, mutual respect, and avoiding community divisiveness, symbols and ceremonial invocations that do not involve coercion do not, in general, present serious questions of religious freedom.110

And the reason why a football coach’s postgame prayer at the fifty-yard line does not violate the Establishment Clause is not that such practice is strongly supported by historical practices or tradition (as language in Kennedy obliquely suggests111). It is that no one else is (truly) coerced personally to engage in the coach’s religious acknowledgement. (Again, if they were, that would be a different matter, as the Court’s opinion implicitly acknowledges by its careful delineation of the relevant facts.)112 A public-school football coach’s personal religious expression, in times, places, and situations where school employees, including teachers, are generally permitted to engage in identifiable personal expressive conduct, is no more intrinsically coercive than any other school employee’s expression on any other topic, simply by virtue of its religious content. Indeed, to discriminate against personal religious expression because of its religious content would be, as noted above, a violation of the Free Speech and Free Exercise Clauses.113

Kennedy illustrates two other important points about the Establishment Clause that the Court has started to get right in recent years. First, like several cases before it, Kennedy rejects the notion that the Establishment Clause contradicts in principle and sometimes overrides in practice the Free Speech and Free Exercise Clauses. As noted above, it is a fundamental error to read the Establishment Clause as an “antireligion” provision in the first place, and it compounds the error to say that such a principle defeats what otherwise are the requirements of equal treatment of religion and religious expression. In Kennedy, the Ninth Circuit had held that the Establishment Clause embodied a prohibition on religious activity that “trump[ed]” Coach Kennedy’s free-speech and free-exercise rights.114 The Supreme Court rejected the premise:

But how could that be? It is true that this Court and others often refer to the “Establishment Clause,” the “Free Exercise Clause,” and the “Free Speech Clause” as separate units. But the three Clauses appear in the same sentence of the same Amendment . . . A natural reading of that sentence would seem to suggest the Clauses have “complementary” purposes, not warring ones where one Clause is always sure to prevail over the others.115

The second important point is Kennedy’s definitive rejection of the Court’s notorious three-pronged “Lemon test,” named for the case of Lemon v. Kurtzman in 1971, which has plagued Establishment Clause jurisprudence since its invention.116 Thirty years ago, in one of the first articles I wrote as a (very young!) law professor, I proudly and eagerly—but perhaps a bit prematurely—proclaimed that “Lemon Is Dead.”117 The Kennedy decision at last seals the deal, driving a stake through the heart of the Lemon test. The Court has not yet definitively embraced “noncoercion” as the single test for the Establishment Clause—it continues to refuse to commit itself to any one approach and has a wealth of wayward precedents with which to deal, distinguish, and deconstruct—but its recent cases make clear that a majority of the Court no longer views the Clause as antireligion in its focus, at war with free exercise, or as erecting a strict and hostile “separation” of religion from civic life. The Establishment Clause emphatically does not embody a principle of “freedom from religion.”


The First Amendment Religion Clauses are all about protecting religious liberty for the sake of religion. This means that government may never discriminate against religion, against religious persons or groups, or on the basis of religious identity, affiliation or expression. It means, more than that, that government may not punish, penalize, or prevent the exercise of sincerely held religious beliefs as the religious believer understands them to direct his or her conduct and that government may not do so even through the reliably neutral application of ostensibly neutral and generally applicable laws. Freedom of religion under the First Amendment is thus both a rule of nondiscrimination and a substantive immunity from government regulation. It means that the exceptions to this strongly presumptive liberty—the situations where government’s rules rightfully can prevail over claims to the free exercise of religious liberty as the requirements of faith are understood by religious adherents—are tightly limited to situations where the claimed religious liberty poses an intolerable harm or risk to the fundamental rights of others so as to not plausibly fit within the intrinsic scope and limitations of “free exercise” in the first place; the exceptions are truly exceptional and not merely the rubber-stamp approval of all assertions of supposed important or compelling interests of governments. It means that the same principles govern the religious freedom and autonomy rights of religious groups, organizations, and institutions. And it means that government can no more prescribe or compel religious conduct, rendering its exercise unfree, than it may proscribe or impair such exercise, rendering the individual (or institution) not free to engage in it.

The Supreme Court’s interpretation of the Religion Clauses is in the course of working itself pure—or trying to. In recent years, the Court’s decisions have moved in the direction of approximating these ideals. But the Court has not quite arrived there yet, being hemmed in to some extent by precedents and hamstrung by old notions that die hard. Nonetheless, if the past decade is any indication, the Court appears (for the most part) on course toward adopting a reading of First Amendment religious liberty that recognizes that the terms of that amendment affirmatively embrace freedom for religion.

Distinguished University Chair & Professor of Law, the University of St. Thomas (Minneapolis/St. Paul). Thanks to Tom Berg, Nathan Chapman, and Sherif Girgis for comments on earlier drafts of this Essay.