The Radical Uncertainty of Free Exercise Principles: A Comment on Fulton v. City of Philadelphia
Elwood and Eleanor Davis Professor Emeritus of Law, George Washington University Law School and David R. and Sherry Kirschner Berz Research Professor of Law and Religion, George Washington Law School
As the Supreme Court’s 2020‒21 Term moved into mid-June, many lawyers, scholars, and concerned citizens waited anxiously for the Court’s disposition of Fulton v. City of Philadelphia. The case involved a conflict between Catholic Social Services (CSS) of Philadelphia and the City of Philadelphia, over whether CSS had a right under the Free Exercise Clause to refuse to evaluate same sex married couples for eligibility to be foster parents. The contract between CSS and the city included requirements of non-discrimination with respect to sexual orientation of prospective foster parents. CSS objected to this requirement, claiming that the rule burdened its organizational religious beliefs that marriage is reserved for unions of one man and one woman.
The U.S. Court of Appeals for the Third Circuit ruled in favor of Philadelphia. A central premise of its ruling was the ongoing validity of the Supreme Court’s 1990 decision in Employment Division v. Smith, which held that the Free Exercise Clause does not confer rights to religion-based exemptions from laws that are religion-neutral and generally applicable to the relevant parties. The Third Circuit agreed with the city that Smith precluded a constitutional right of exemption from its prohibition on discrimination against same-sex married couples, because all social welfare agencies (religious or not) in the foster care system had to abide by that prohibition.
In its certiorari petition, CSS explicitly urged the Court to consider whether Smith should be overruled, and the Court included that question in the grant of review. CSS asserted, as many lawyers had for years, that Smith had unconstitutionally left religious liberty at the mercy of legislators and government administrators. Courts, claimed CSS, should play a larger role.
The Supreme Court heard argument in Fulton on the day after Election Day, 2020, just a few days after Justice Amy Coney Barrett took the oath of her new office. Seven and half months later, the Supreme Court surprised every Court watcher with a unanimous decision in favor of CSS. The Court was not unanimous, however, on the issues involving the content of free exercise principles. Chief Justice John Roberts wrote the Court opinion for six justices, resting its holding on the narrow and questionable ground that Philadelphia’s non-discrimination policies were not “generally applicable,” and therefore were outside of the protective ambit of Smith. That conclusion led the Court to apply more rigorous scrutiny to the city’s treatment of CSS, and the Court determined, with almost no analysis, that the exclusion violated CSS’s free exercise rights.
Over the course of two separate opinions, joined by all of them, three justices—Justices Samuel Alito, Neil Gorsuch, and Clarence Thomas—concurred in the judgment only. The Gorsuch opinion accused the majority of relying on disingenuous arguments to avoid the question of whether Smith should be overruled. And Justice Alito, in a sweeping seventy-seven pages, argued emphatically that Smith should indeed be overruled.
Justice Roberts’s Court opinion offered no answer to Justice Alito’s extended attack on Smith. The only answer from anyone in the Court’s majority appeared in a very brief, concurring opinion by Justice Barrett. It was not a surprise that the appointment of Justice Barrett proved auspicious. Her predecessor, Justice Ruth Bader Ginsburg, had supported the decision in Smith throughout her tenure on the Court. Justice Barrett indeed diverged from the Ginsburg line, but not in the ways her supporters had hoped, or her opponents had feared.
Justice Barrett, joined by Justice Kavanaugh, leaned toward agreement that the text and structure of the Free Exercise Clause did not support Smith. But her opinion expressed doubt concerning Justice Alito’s view of the history relevant to the Free Exercise Clause, and identified a set of crucial questions, pointedly not answered by Justice Alito, about the uncertainty that would follow from overruling Smith. Justice Barrett saw no reason to leap into the free exercise thicket in a case where all nine justices agreed that CSS should prevail.
Fulton invites consideration of virtually all the questions and nuances of free exercise law that have occupied judges, lawyers, and scholars for the past three decades. In the space we have here, we want to highlight several features of Fulton.
As widely noted, the Court was unanimous on the outcome in favor of CSS. Unanimity is rare in Religion Clause adjudication, and its causes and consequences deserve attention. Part I of this Comment compares the last surprisingly unanimous Religion Clause decision, Hosanna-Tabor Evangelical Lutheran Church & School v. EEOC, with the decision in Fulton. As we will explain, unanimity in Hosanna-Tabor was wide, deep, and rooted in longstanding Religion Clause principles. In contrast, unanimity in Fulton was shallow, perhaps even pretextual, a mask incapable of hiding deep disagreements.
Part II explores the threads by which Smith’s future hangs. Part II.A. describes the hostile reaction to Smith over the past thirty years in legal, academic, religious, and political circles. This hostility manifested itself in the briefing in Fulton and appeared in full flower in Justice Alito’s concurring opinion.
Part II.B. explains why Justice Alito’s view of the text and history of the Free Exercise Clause is deeply wrong. The framers of the First Amendment designed the Clause to protect religious communities from government interference with their worship practices. Justice Alito’s far broader conception of the clause as protective of all religiously motivated practices is at the root of his mistakes.
Part II.C. explores the consequences of his mistaken view. If the Free Exercise Clause strenuously protects all religiously motivated practices, and if believers get to self-identify which of their practices are religiously motivated, the clause becomes a ticket to override virtually all government policy. Identifying Justice Alito’s mistakes thus sheds light on why Smith was correct, and why so many people have trouble seeing that. Our critique of Justice Alito also helps explain why Justices Barrett and Kavanaugh drew back from embracing his view. Justices Barrett and Kavanaugh saw, perhaps through a glass darkly, that Justices Alito, Gorsuch, and Thomas had no answer to the questions that inevitably arise from such a broad constitutional conception of religious exercise. In that process of perception lies the salutary story of why the Court lacked five votes to overrule Smith, and the explanation of why the current law in some form remains likely to endure.
Part II.D., focusing on the Court’s COVID-19 cases and Fulton itself, explores the apparent contraction of what qualifies as a law of general applicability. The questions here are subtle, and deserve attention, because these narrowing moves threaten to undermine the framework on which Employment Division v. Smith has been built.
I. Free Exercise of Religion and the Mysteries of Unanimity
Nine years ago, in Hosanna-Tabor, a fully unanimous Court embraced the “ministerial exception” to laws governing the employment relationship. The case involved a fourth-grade teacher, Cheryl Perich, who had a ministerial title and some responsibilities for teaching the faith. After being dismissed, Perich filed a lawsuit against her employer, alleging retaliation for seeking civil redress for discrimination based on disability. The context most assuredly invited a liberal critique of any legal theory that cut off civil-rights claims by employees.
Nevertheless, all nine justices agreed that Perich served a ministerial function within the school, and that both First Amendment Religion Clauses barred the claim against her employer. The decision rested on a longstanding principle, dating to the nineteenth century in American law, that courts are constitutionally incompetent to decide exclusively ecclesiastical questions. The fitness of a class of persons, or of a particular person, for ministry is such a question. Accordingly, once a court determines that an employee functions as a minister—that is, has substantial responsibility for teaching or communicating the faith—any inquiry into the wrongfulness of her dismissal must end. The matter is for the religious employer alone.
Every justice agreed to this basic account, and to the application of the relevant principles to Perich. Moreover, and crucial to our account of Hosanna-Tabor, the ministerial exception invites no balancing of interests. It operates just like other Establishment Clause limitations, such as the prohibition on public-school-sponsored prayer. No state interest, however important or precisely served, can overcome Establishment Clause limitations.
The refusal to balance interests in ministerial exception cases is made explicit in Chief Justice Roberts’s opinion for the Court. This move is highly distinctive in Religion Clause adjudication. In Free Exercise cases, before and after Smith, the government’s interest in imposing a restriction is always of relevance. And one can imagine that an interest-balancing approach would matter significantly in a ministerial exception case. In the case of Cheryl Perich, in particular, her responsibility for teaching the faith was relatively slight. She did not preach at large worship gatherings. The government’s side of the ledger, had it counted, would register the strong interest in barring retaliation against employees who go to public agencies or courts with discrimination claims. Because some employees are in positions for which the ministerial character is uncertain, a ruling that they can be the subject of retaliatory firings deters their complaints and limits rightful enforcement of the law.
Hosanna-Tabor, therefore, involves unanimity at a level that is wide, deep, and heavily pedigreed. Its approach covers the adjudication of all exclusively ecclesiastical questions, whether they involve personnel, property, or internal structures of church governance. Courts must abstain from adjudicating these disputes, and government interests, however strong, play no part in their resolution.
This is not a doctrine of free exercise-based church autonomy, in which interests are uneasily balanced. It has a wider base than that—both Religion Clauses. It also has a narrower ambit—exclusively ecclesiastical questions, rather than the vaguer notion of internal church affairs. And its methodology is categorical. Courts are wholly incompetent to decide those questions.
Contrast the situation in Fulton, which offers only the veneer of unanimity. First, the unanimity is in the result, not the reasoning. Six justices, represented in Chief Justice Roberts’s opinion for the Court, asserted that the relevant norms of non-discrimination were not generally applicable within the meaning of Smith. This conclusion rested on the spectacularly specious ground that the commissioner of social services had discretion, which his office had never exercised, to make exceptions to the city contract’s requirements of non-discrimination in screening prospective foster parents. Even though the non-discrimination norms in fact had been uniformly and consistently applied to every social welfare agency, religious or secular, the majority hunted for a way to push the case out of the ambit of Smith.
Once that was done, the Court still had to determine whether the city’s policy could survive the strict judicial review that followed. In a remarkably superficial passage, the majority reasoned that the very possibility of exceptions meant that the government’s interest in denying exceptions could not be sufficiently important. As anyone can see, it is utterly unpersuasive to diminish the city’s interest in denying exceptions to non-discrimination norms by reference to the never-exercised power to grant such exceptions. Something explains this thin and weak reasoning, but none of the justices who joined the Court opinion makes the attempt.
The three justices who did not join the Court’s opinion had a ready explanation for the Court’s maneuvers. Justice Gorsuch, joined by Justices Alito and Thomas, attacked the majority’s reasoning on the general applicability of Philadelphia’s non-discrimination norms. He asserted, with good cause, that the majority had gone far beyond the bounds of normal legal reasoning to find a way not to address the attack on Smith, which the Court in its certiorari grant had agreed to entertain. Justice Alito, joined by Justices Gorsuch and Thomas, accepted the invitation with gusto, and concluded that Smith should be overruled.
So, unlike in Hosanna-Tabor, where all nine justices accepted the essential premises of ecclesiastical abstention and its lesser included element of the ministerial exception, Fulton revealed extreme tension among groups of justices about the basic premises that underlie the free exercise guaranty. Moreover, this unanimity of result, coupled with the obvious tension between the Alito-Gorsuch-Thomas group and all the others, has produced understandable speculation about the initial assignment of the opinion, and possible migration of the justices during the opinion writing process.
We believe that the initial line-up of the justices contained three groups—three (Alito-Gorsuch-Thomas) who were eager to both overrule Smith and rule for CSS; three (Roberts-Kavanaugh-Barrett) who were willing to rule for CSS but reluctant or unwilling to overrule Smith; and three (Breyer-Kagan-Sotomayor) who were inclined to rule for the city under the existing case law. The questions from the justices at oral argument are consistent with this appraisal. If these divisions held, there would have been no majority opinion. We would have had a splintered three-three-three ruling. CSS would have won, and the opinion for the Roberts group would have been the narrowest opinion in support of the result, and therefore would have been controlling.
All this suggests that somewhere along the way, a deal was struck to eliminate any dissenting opinions. In exchange, the likely dissenters got a very narrow Court opinion that resolved none of the deeper questions about conflicts between religious freedom and anti-discrimination law as it protects LGBTQ persons.
Highly questionable arguments drove the Court’s decision that the city’s policies were not generally applicable. The opinion refers to a provision (section 3.21) in the City’s standard contract with agencies that provide foster care services to the effect that those agencies, including CSS, will not “reject . . . prospective foster parents . . . based on their race, ethnicity, color, sex, sexual orientation, gender identity, religion, [or] national origin unless an exception is granted by the Commissioner or the Commissioner’s designee, in his/her sole discretion.” This grant of discretion, the Court reasons, invites case-by-case exceptions, and therefore destroys the general applicability of the anti-discrimination norm.
Although Smith itself had suggested that, in the unemployment context, a regime of discretionary exceptions for “good cause” should not be viewed as generally applicable, the unemployment context is one in which such exceptions are routine and on-going. Once “good cause” exceptions are made, the scheme no longer applies to all parties in the same way. In contrast, the city commissioner in Philadelphia had never made an exception under 3.21 for discrimination on forbidden grounds against prospective foster parents, and the city asserted that the commissioner lacked authority under other provisions of the contract and under local law to make such exceptions. So, as a matter of consistent practice, the city treated the non-discrimination norm with respect to prospective foster parents as generally applicable. Notably, this argument about discretion played almost no part in the presentation by CSS to the Court, and yet it wound up at the heart of the majority opinion.
The emphasis in the Court opinion on the contract, which Philadelphia can revise, reveals that unanimity in Fulton is Court-wide and an inch deep. The absence of any rigor in applying the standard of review confirms that assessment. Justices Sotomayor, Kagan, and Breyer may well have believed that the city’s interests in avoiding stigmatic injury to same sex couples, and material injury to LGBTQ children in need of placement, were sufficiently compelling to justify denial of an exemption from conditions of public service. Obviously, they did not say so. On any plausible accounting of the concerns of individual justices, unanimity in Fulton is a translucent veneer.
II. The Struggle over First Principles of Free Exercise
It is easy to forget that the core protections of the Free Exercise Clause remain solid and unchallenged. The government may not target for regulation or prohibition the content of worship by a particular faith. That content, historically a feature of regulation in England, involves exclusively ecclesiastical questions, about which a secular government has no legitimate interest.
The fights over free exercise rights in the twentieth and twenty-first centuries, however, are rarely about that core. Instead, the fights are about general laws and policies that do not target specific faiths but have an impact on religiously motivated practices outside of worship. Fulton presents a paradigm example. The City of Philadelphia did not direct the Archdiocese of Philadelphia to make available the sacrament of marriage to same-sex couples. Entitlement to that religious status is entirely within the province of each faith community. Rather, the city prohibited discrimination among lawfully married couples in the certification of potential foster parents. That is a public project, publicly regulated, and publicly financed. The city was advancing the interests of all of its people, not targeting Catholics or others for their view of marriage.
A. The Near-Silence in Defense of Smith
As Part I of this paper suggests, the fights within the Court over the past, present, and future of free exercise principles remain open and fierce. One puzzling aspect of these fights is the attitude of the current justices toward the correctness of Employment Division v. Smith. When Justice Anthony Kennedy retired, the last of the five justices who joined in Smith left the Court. We know that at least three current justices want to overrule it, and at least two more are somewhat skeptical of it. Of the other four, not one has publicly embraced Smith.
Moreover, the reticence on the Court to defend Smith is mirrored in the larger society of scholars, lawyers, and concerned citizens. From the beginning, Smith has taken intense critical fire, on a variety of grounds. The Court’s most liberal members at the time, Justices William Brennan, Thurgood Marshall, and Harry Blackmun, vigorously dissented in Smith. Critics have asserted that Smith is wrong as a matter of textual interpretation and original understanding. They have offered versions of constitutional history to undermine it. They assert that it cannot be squared with precedent, or with the Court’s treatment of other First Amendment rights. And they bewail the Court’s pronouncement of the Smith rule without warning to litigants that the Court was considering a major departure from prior decisions.
These criticisms originally came from the right, left, and center of the ideological spectrum. The coalition that organized the push for the Religious Freedom Restoration Act (“RFRA”) included representatives from all of these perspectives and included the voice of highly respected civil liberties organizations. Congress enacted RFRA in 1993 by a nearly unanimous vote. We have noticed over many years that political actors, like governors, state attorneys general, and city officials are reluctant or unwilling to defend Smith, because so many religious people and institutions have long criticized it.
In the period since about 2000, claims that religious freedom justifies refusal to serve same sex couples have engendered tremendous and intense opposition, but rarely has that opposition been framed as a defense of Smith. Instead, the arguments have taken the form of asserting that the government has a compelling interest in eradicating LGBTQ discrimination, and thus would prevail even if Smith were overruled.
Fulton proceeded in the Supreme Court along these lines. The grant of certiorari in Fulton explicitly invited the Court to revisit Smith. As one would expect, the briefs for Catholic Social Services and many of its amici strenuously urged the Court to overrule Smith. Nevertheless, the City of Philadelphia and many of its amici refused to address the merits of Smith. They argued instead that, even if Smith was wrong, courts should reject claims of religious freedom to discriminate based on sex or sexual orientation. 
In the Fulton briefing, we were the voice that no one would admit they wanted to hear. Together with Professors William Marshall and Frederick Gedicks, we filed a brief on the side of the city, arguing that Smith is correct and should be reaffirmed. All four of us understand that the regime of Smith can be insensitive to religious minorities, but we still believe that, all things considered, Smith is better than any other proposed approach to religious accommodations. By our count, there were only three other amicus briefs that explicitly argued for retention of Smith. Among the forty-seven amicus briefs filed on behalf of the city, the authors of forty-four chose to ignore that question.
Despite that overwhelming quantitative imbalance, the majority of the Court resisted the urging to overrule Smith. Thirty years of political and scholarly criticism, extended and repeated legislative efforts to overturn the decision, and briefing heavily stacked in one direction in Fulton convinced only three justices. Three justices expressed considerable reticence about replacing Smith, and three others remained silent on the question. Something, not yet in view, explains the reluctance or unwillingness of six justices to overrule Smith.
B. The Deep Flaws in Justice Alito’s Critique of Smith
Justice Alito’s opinion, which ranges over the entire waterfront of constitutional argument—text, structure of the First Amendment, nineteenth-century history of religious freedom in the U.S., and an evaluation of the relevant precedent—is riddled with false narratives, internal contradictions, and errors of history. No one on the Court tried to answer his opinion in full, but unanswered is not the same as correct.
From the outset, Justice Alito’s aggressive effort to shape the narrative leads him astray. He promises a “fresh look at what the Free Exercise Clause demands.” In the slip opinion published on June 17, 2021, Justice Alito reviewed the law as it stood at the time of Smith (1990) and asserted that Sherbert (1963) “had been in place for nearly four decades when Smith was decided.” A page later, he re-asserted that the Sherbert test provided the governing rule in free exercise cases for “the next 37 years.” We are certain his skill at arithmetic is better than this, but we cannot help but notice the direction of his error, exaggerating the Sherbert rule’s longevity.
More broadly, his focus on the period from Sherbert to the eve of Smith as the place to begin a “fresh look” at the demands of the Free Exercise Clause is highly revealing. He might logically have begun with the text and history of the clause. Or, he might have started with the Supreme Court’s first major engagement with the clause in Reynolds v. United States (1878), which rejected the idea that the clause privileged religiously motivated action (in that case, plural marriage). Instead, he emphasizes the period between 1963 and 1990 as a way to paint a narrative in which Sherbert-Yoder is normative, and Smith is an aberration.
His account, however, is quite backwards. Our amicus brief in Fulton more accurately described the flow of Free Exercise jurisprudence from 1878 onward. The Court had consistently rejected claims of free exercise exemptions until the early 1960’s. Only in the period from the early 1960’s to 1990 did the Court purport to apply the compelling interest test to any claims of religious exemption. Moreover, as we explain later in this section, even in that period the Court frequently worked around that test.
The most accurate description of free exercise exemption claims through history is that the Court denied all of them between 1878 and 1963; appeared occasionally sympathetic to them from 1963 through 1990; and then again repudiated them under the Smith rule from 1990 onward. By our count, that would be a total of 116 years in which free exercise exemptions were not constitutionally mandatory, and 27 years in which, quite sporadically, such claims succeeded. If we start counting from the ratification of the First Amendment in 1791, the count would be 203 years without any notion of free exercise exemptions, and 27 years of a doctrine ostensibly favorable to exemptions. The period of Sherbert-Yoder was the aberration. Smith returned the law of the Free Exercise Clause to normalcy.
Justice Alito’s other errors are more subtle, but hardly less conspicuous to the careful reader. Time after time, his own footnotes contradict his assertions.
Start with his explanation of the text. Justice Alito claims that he will analyze the “normal and ordinary” meaning of the Free Exercise Clause. One of his first moves in searching for that meaning is to claim that the word “religion” requires no discussion for purposes of this case. At that point, he drops a quite extraordinary footnote, which reads: “Whatever the outer boundaries of the term ‘religion’ as used in the First Amendment, there can be no doubt that CSS’s contested policy represents an exercise of ‘religion.’”
The question in Fulton, however, is not whether the Catholic view of who qualifies for its sacrament of marriage is a matter of religion. No one disputes that. Rather, the question is whether the CSS policy of who may serve as a foster parent represents an “exercise of religion” in the constitutional sense. This policy is not a reflection of the internal judgment of the Church about its sacraments, a judgment that would exclude couples of other faiths. The policy is aimed at the general population of Philadelphia, Catholic or not. Whether such a policy qualifies constitutionally as an exercise of religion is a matter open to serious historical doubt, as Justice Alito’s own sources strongly indicate.
Justice Alito’s turn to eighteenth-century dictionaries further illustrates the embarrassing gap between the text of the Free Exercise Clause and the citations he uses to support his interpretation of that text. He notes, with multiple sources, that the term “exercise” had both a broad primary definition (“[p]ractice” or “outward performance”) and a narrower secondary one (an “[a]ct of divine worship whether publick or private”). Which of these should control the interpretation of the Free Exercise Clause? The first appears to be its non-specific use as “practice or performance,” as in musical or physical exercise. The second definition, however, appears in explicit connection with religion and has a singular meaning: “Act of divine worship.” As we explain further below, this religion-centered definition coincides perfectly with what late eighteenth-century Americans understood as the exercise of religion.
How does Justice Alito escape the conclusion that a worship-centered definition of religious exercise should control the meaning of the Constitution? He turns, briefly and parenthetically, to precedent from the middle of the twentieth century—in particular, Cantwell v. Connecticut. This is just sleight of hand. First, the Cantwell opinion, penned in 1940, makes absolutely no reference to the original meaning of the First Amendment. It thus carries no weight as an originalist interpretation. Second, the context of Cantwell is street preaching by members of the Jehovah’s Witnesses. This is dissemination of the word of God and may fall within the original meaning of the Free Exercise Clause. Even if that conclusion is correct, however, it does not come close to proving that all religiously motivated conduct should qualify as religious exercise within the “normal and ordinary” meaning of the Free Exercise Clause. Justice Alito is cheating in his textualist story, and even the cheating does not get him where he wants to go.
The best reading of Justice Alito’s footnotes belies his assertion that CSS’s practice involves the exercise of religion. On narrow, dictionary-driven textualist grounds, the phrase “Free Exercise of Religion” should be limited to acts of worship, and (as we explain below) the closely related practices of public preaching and proselytizing.
Justice Alito’s analysis of constitutional history reveals the same tendentious qualities as his analysis of text. The relevant history fully supports the notion that “exercise of religion” referred to acts of worship, and certainly did not encompass all religiously motivated conduct. Once again, the footnotes sharply contradict Justice Alito’s assertions in the text.
Justice Alito traces “free exercise” back to an act by the Maryland Assembly in 1649 and says that by 1789 “every State except Connecticut had a constitutional provision protecting religious liberty.” On the sound assumption that the federal Free Exercise Clause reflects the widespread pattern of such protections, Justice Alito’s originalist argument rests on his reading of those state constitutional provisions.
Here, however, his argument goes off the rails. Instead of analyzing the content of religious liberty encompassed by those constitutional provisions, Justice Alito focuses solely on a limitation frequently imposed on that liberty. Many state constitutions expressly provided that the right of religious liberty does not protect conduct that would endanger “the public peace” or “safety.” “If, as Smith held,” Justice Alito writes, “the free-exercise right does not require any religious exemptions from generally applicable laws, it is not easy to imagine situations in which a public-peace-or-safety carve out would be necessary.”
Once again, Justice Alito goes out of his way to avoid the core question. Instead of stressing the importance of carve-outs, Justice Alito should first and foremost have attended to the substance of religious liberty itself. Even a cursory reading of Justice Alito’s long footnote, cataloguing state constitutional guaranties in the founding period, shows that these provisions focused exclusively on freedom of worship and belief.
To take just one example, the 1780 Massachusetts Constitution reads: “It is the right as well as the duty of all men in society, publicly, and at stated seasons, to worship the SUPREME BEING, the Great Creator and Preserver of the Universe. And no subject shall be hurt, molested, or restrained, in his person, liberty, or estate, for worshipping GOD in the manner and season most agreeable to the dictates of his own conscience; or for his religious profession or sentiments; provided he doth not disturb the publick peace, or obstruct others in their religious worship.”
The concern for public order in the 1780 Massachusetts Constitution derives from the threat of unruly forms of worship, not from the broader category of all religiously motivated conduct. The Massachusetts provision echoes other state constitutions. All locate the right in worship, and many extend it to religious “profession or sentiments.”
The history of mid-eighteenth-century religious conflict in the colonies provides the context necessary to fully understand the references to peace and order. The First Great Awakening brought widespread religious revivals throughout the colonies. Many of the revivalists were itinerant preachers who faced hostility from some “settled” clergy (those who had congregations) over doctrinal and practical differences. These itinerant preachers, foremost among them George Whitefield, preached theatrical sermons, calling for believers to experience a “new birth” of Christ in their hearts. According to these preachers, this new birth is necessary to avoid the very real peril of hell, which they described in vivid detail. The sermons led some listeners to collapse, others to speak in tongues, and still others to cry out in fear for their souls. This type of worship stood in sharp contrast to the learned sermons and staid services of Congregationalist clergy in New England, or Anglican priests in the southern colonies.
Many ministers and others objected to the “wild” form of religiosity, which they found to be more spectacle than divine worship. Others feared that they would lose their congregants to the itinerant preachers. And still others resented the question from congregants or itinerant preachers about whether the settled ministers had received the “new birth.”
In the years before the Revolution, some colonial legislatures attempted to protect the interests of settled ministers by restricting the activities of itinerant ministers. Virginia made the most intense effort by requiring all preachers to obtain a license from a board of Anglican clergy. James Madison wrote that he was greatly affected by the sight of Baptist evangelists jailed for violating the act.
Conflicts over these evangelical revivals fully explain the carve-outs that Justice Alito finds so important. Even if a state constitutional provision protected the right of itinerant preachers to offer a public worship service—often held outside, because towns or settled clergy regularly denied them the use of a church—the concern remained that worshippers might disturb public peace and order.
By focusing only on the carve-out, Justice Alito ignores the interpretive question at the heart of any serious textualist or originalist inquiry: What did the “free exercise of religion” mean at the Founding? Justice Alito simply assumes the conclusion he wants to reach and ignores obvious contradictory evidence. That evidence shows that the constitutional understanding of religious exercise, at the Founding, is far removed from the stance of CSS toward married same sex couples as foster parents, no matter how religiously motivated the CSS policy may have been. Justice Alito’s stipulation that CSS’s policies are religious exercise, within the original meaning of the Free Exercise Clause, is wrong.
This is not the place to fully develop an alternative account of the textual and historical meaning of the “free exercise of religion.” We can, however, sketch one that is much more plausible than that offered by Justice Alito. This narrative starts with the 1689 Act of Toleration, which granted most dissenting Protestant communities the right to worship and hold beliefs that differed significantly from the practices and doctrines of the Church of England. The Act required dissenting Protestants to take an oath of allegiance to the Crown and confirm their belief in certain doctrines of the faith, most prominently the Holy Trinity and the divine inspiration of the Bible. But it allowed them to worship in their own congregations and according to their own beliefs about baptism, communion, and church order—doctrines that had been at the heart of most disputes between the dissenters and the Church of England. The Act also denied protection to any worship that threatened public peace or order—a clause that many state constitutions retained after Independence.
During the eighteenth century, dissenting Protestant communities in the American colonies frequently invoked the rights granted by the Act of Toleration. Although many colonial governments complied with the provisions, or even offered much broader liberties for worship, others resisted. At first, Massachusetts even refused to permit Anglican missionaries and churches, but the Crown soon ended that recalcitrance. Virginia proved the most obstinate opponent of the Act. As late as the Revolutionary Era, Virginia’s government denied that the Act applied to the colony. Licensing authorities almost uniformly refused to permit dissenting ministers to preach or form congregations in Virginia.
The liberty sought by these dissenting Protestant faiths, then, focused on the rights of worship and public preaching. As their rhetoric shifted from toleration to religious freedom, the groups sought equality with whatever denomination enjoyed favored status in the colony. Laws that required licensing of dissenting clergy or registration of religious meeting houses failed the test of equality.
The debate over public funding for ministers or houses of worship is the closest potential analogy for modern arguments for religious exemptions. Opponents of such funding, particularly Baptists, argued that the imposition of religious taxation treated them unequally (because they refused to provide or receive any compelled support) and impermissibly involved the state in matters outside the scope of its temporal authority. The dissenting arguments blocked efforts to impose religious taxes in some states (most notably Virginia) but failed in others (primarily Connecticut and Massachusetts).
Theophilus Parsons, a drafter of the Massachusetts Constitution, later justified the taxation of dissenters by distinguishing between the spiritual and civil functions of religion. All are free to believe, worship, or publicly preach according to their faith. But state support for certain Protestant faiths is proper because instruction in core Protestant doctrines—especially belief in a future state of rewards and punishments—will produce citizens who are more likely to obey laws and public morality. Such funding serves the purely civil function of ensuring peace and order, and benefits even those who do not seek or qualify for state funding of their ministries.
The eighteenth-century history thus suggests (in accord with Justice Alito’s citations) that “exercise of religion” focused on belief, worship, and public preaching. It does not support his conclusion that the “exercise of religion” encompasses any act that claimants believe is required by their faith. The Massachusetts example proves most telling: The law protects the “spiritual” domain, not a pluralistic idea of religiously motivated conduct in the civil or secular domain.
We recognize that this Protestant-centered view of the “exercise of religion” fails to respect the beliefs and practices of many religious adherents in our pluralist nation. We do not argue that this interpretation is the best normative reading of the Free Exercise Clause today. Instead, the argument we are making is that the Constitution drafters of the eighteenth century understood the restriction on laws “prohibiting the free exercise [of religion]” as a barrier to government interference with worship, and perhaps with preaching the Word through religious instruction and proselytizing. The drafters’ meaning takes on immeasurably greater importance as the society became more religiously pluralistic. Reaching beyond the original constitutional limits to protect all religiously motivated conduct in a pluralistic society is unmanageable. As Justice Scalia in Smith explained: “because we value and protect that religious divergence, we cannot afford the luxury of deeming presumptively invalid, as applied to the religious objector, every regulation of conduct that does not protect an interest of the highest order.”
Justice Alito makes yet one more, all-too-common mistake in his analysis. He complains that Smith made the Free Exercise Clause into “essentially an anti-discrimination provision.” Although it is correct to say that both the Establishment Clause and the Free Exercise Clause have anti-discrimination components, Smith goes beyond this and recognizes the Free Exercise Clause as protecting the right to choose forms of worship. As Justice Scalia wrote, “It would doubtless be unconstitutional, for example, to ban the casting of ‘statues that are to be used for worship purposes,’ or to prohibit bowing down before a golden calf.” The focus of these examples is not sectarian discrimination. Rather, they refer to direct regulation of worship qua worship, protected as such by the Free Exercise Clause.
C. The Consequences of Justice Alito’s Flawed Interpretation
Justice Alito’s overbroad reading of the Free Exercise Clause leads inexorably to a series of constitutional troubles, one piled on top of the other. First, it sweeps in a vast range of human behavior that is subject to law, including employment arrangements, family relations, and the duties of those who contract with the public. This range is expanded yet further by the recent tendency to advance claims of religious complicity, as a means of resisting duties to others who exercise privacy rights, reproductive rights, and marital rights in ways that some religious people oppose. Complicity claims arise in relationships, and contemporary life is thick with relationships that invite one party to object to the behavior of others. Masterpiece Cakeshop, Burwell v. Hobby Lobby, and Fulton all arose out of such claims of complicity.
Second, Justice Alito here and elsewhere subscribes to an extremely generous and subjective notion of what constitutes a substantial burden on religious exercise. If religious exercise is confined to worship and preaching, courts can measure objectively the extent to which a government policy interferes with that exercise or imposes substantial costs on it. In contrast, the notion of burden as an internal, subjective imposition on conscience cannot be second-guessed or measured. Thomas v. Review Board pushed the idea of burden in this subjective direction, and Justice Alito’s opinion in Hobby Lobby amplified it. Other than an attack (rarely attempted by the government) on the sincerity of belief, the government has almost no way to successfully argue that an asserted burden on religious belief and practice is insubstantial.
Third, and directly related to the second, Justice Alito’s opinion asserts that the standard of review for exemption claims that “comes most readily to mind is the standard that Smith replaced: A law that imposes a substantial burden on religious exercise can be sustained only if it is narrowly tailored to serve a compelling government interest.” These requirements—compelling interests, and the demand for narrow tailoring—are independently difficult to satisfy. Interests of the government may be vital, yet may fail the compelling interest test by falling just short, or by being undercut by exceptions, however narrow, that weaken the government’s claim about the weight of the interest.
Requirements of narrow tailoring likewise are difficult to satisfy because the government frequently has alternatives, albeit less efficient or more expensive, to accomplish its ends. At times, the broader alternative may be far more effective at achieving the government’s goals. For example, a complete ban on importation of a prohibited substance, used in religious sacraments, is likely to work far better than monitoring by government agents of the actual deployment of the substance by the religious group.
The combination of these three points—the breadth of what counts as an exercise of religion; the ease of satisfying the “substantial burden” test; and the difficulty of satisfying the tests of “narrow tailoring—compelling interest”—is what makes Justice Alito’s approach so sweeping and unmanageable.
We know from extensive experience during the regime of Sherbert-Yoder that courts will not stay on this path, even if they pretend to do so. In the 1980s, faced with a doctrine that made it nearly impossible for the government to deny an exemption, the Court found multiple ways to work around the promise of presumptive immunity for religious claims. By 1990, the Court had worked around the Sherbert-Yoder standard far more often than it had applied that standard with the promised rigor. The methods included weakening the doctrine in government-controlled enclaves such as prisons and the military; refining the concept of burdens to exclude difficult cases, such as those brought by Native Americans with respect to sacred lands, and weakening the compelling interest test to make it easier to satisfy. Lower courts did likewise. All of these moves produced significant departures from rule-of-law values of consistency and predictability in application of the law. Why should we expect anything different now?
Fourth, Justice Alito’s broad conception of free exercise, coupled with his assertion of a strict judicial standard to govern such claims, reveals the deep flaws in his structural argument about the constitutional treatment of speech, press, and association compared to religion. With respect to coverage, rights of speech, press, and association all relate to human activities of communication. This is a broad subject indeed, but it is miniscule when compared with the entire universe of human behavior, all of which may be touched by religious conviction.
When a news organization, for example, enters into employment relationships, or constructs a building in which to create its product, no one asserts that the First Amendment imposes strict standards on the government’s regulation of such activity. With respect to such matters, if government regulates news organizations to the same extent as other, comparable entities—that is, if the regulations are speech-neutral and generally applicable—the First Amendment has no work to do. By contrast, those who want to overturn Smith in free-exercise cases assert that the incidental impacts of any regulation of religiously motivated actors must be justified under strict, government-limiting standards. If free exercise meant worship activities and not the entirety of religiously motivated acts, the structural gap between religion and other First Amendment-protected activity would shrink dramatically.
Furthermore, with respect to standards of review, no justice or scholar has ever contended that a single, highly potent review standard should govern every possible dispute arising from regulation of communicative activity. When government regulates the content of communication, it must answer to the highest constitutional concerns. In contrast, when government regulates the time, place, and manner of expression, the relevant standards are more relaxed. When the regulation takes the form of control of expressive conduct, and the government has an interest in the conduct independent of the message it sends, the standards are quite lenient indeed. Most generally, free speech principles do not protect speech against “incidental burdens” from generally applicable, speech-neutral laws.
These latter categories—regulation of conduct generally, speech-neutrally, and independent of the message it sends—is a precise analogue for generally applicable standards that apply to all behavior of a certain kind, religiously motivated or not. When Justice Alito and others call for strict scrutiny in such cases, they are demanding special treatment for religion, not treatment equal to that afforded other constitutional rights.
Many of the concerns expressed in Justice Barrett’s brief concurring opinion can be instructively linked to one or more of the problems that emerge from Justice Alito’s overbroad conception of what the Constitution means by the exercise of religion. Her opinion asks why the Free Exercise Clause, alone among First Amendment rights, should be limited to concerns about discrimination. The answer is that the clause is not so limited. It protects the activities of worship and preaching the Word as strenuously as it protects political advocacy, but it does not protect everything done in the name of religion.
Justice Barrett is “skeptical about swapping Smith’s categorical antidiscrimination approach for an equally categorical strict scrutiny regime, particularly when this Court’s resolution of conflicts between generally applicable laws and other First Amendment rights—like speech and assembly—has been much more nuanced.” This skepticism is appropriate, but far too weak, because Justice Alito’s conception of the exercise of religion is so much broader than any judicial conception of what counts as the exercise of these other rights.
Justice Barrett goes on to raise a series of concerns, at least some of which arise entirely from unasked and unanswered questions about the breadth of free exercise rights. For instance, she asks whether “entities like Catholic Social Services—which is an arm of the Catholic Church—[should] be treated differently than individuals?” A better understanding of the constitutional meaning of the exercise of religion would lead to a related, but different question—whether Catholic Social Services is exercising religion in the constitutional sense when it provides social services to the general public. Of course, CSS is motivated by religion, and understands itself in religious terms, but that cannot transform all of its daily activities into the kind of religious exercise protected by the First Amendment.
Justice Barrett’s final question about review standards is especially pointed: “[I]f the answer is strict scrutiny, would pre-Smith cases rejecting free exercise challenges to garden-variety laws come out the same way?” This question is telling, in two respects. First, she is acknowledging—as Justice Alito, CSS, and its amici do not—the pervasive pattern of work-arounds in pre-Smith law. Second, her reference to “garden variety laws” shows an acknowledgment of the scope of exemptions encompassed by Justice Alito’s broad notion of free exercise. They would include generally applicable tax laws, labor laws, social welfare policy, land use control, and other areas of regulation not aimed at the experience of worship.
Justice Alito’s opinion invited these questions, and Justices Barrett, Kavanaugh, and Breyer deserve praise for raising them. The opinion of Justice Barrett, who is a former clerk for Justice Scalia, builds on the legacy of his opinion in Smith. As a result, and despite the few lonely voices speaking up in defense of Smith, its principles endure. The lower courts remain bound by it. The Supreme Court has neither the incentive nor sufficient interest among the current justices to take it up again quickly. Without a workable alternative, the Smith framework for adjudication of free exercise cases will remain in place. And if religious exercise is understood in Justice Alito’s broad terms, there is no workable alternative.
D. The Scope of “General Applicability”
The one major inroad into the regime of Smith is the apparent narrowing of what counts as a generally applicable law. Two major developments point in this direction: (1) the Court’s treatment, in its emergency docket, of orders limiting attendance at houses of worship during the COVID-19 pandemic; and (2) the Court’s analysis in Fulton of the role of administrative discretion in undermining the general applicability of the law. Taken to their logical ends, these developments suggest that the category of generally applicable laws will shrink to the vanishing point, and Smith will become irrelevant. The experience of courts with the pre-Smith law, however, suggests a different fate.
The COVID cases and the significance of secular exceptions. As recently as the summer of 2020, the Court’s most prominently expressed view about the concept of general applicability remained its opinion in Church of the Lukumi Babalu Aye, Inc. v. Hialeah. Failures of general applicability were to be found only in invidious and hostile departures from even-handed treatment of all religions and their secular counterparts. Such departures demonstrate deliberate discrimination against religion generally, or (as in Lukumi) a particular religion.
When the first COVID cases arising from state limitations on attendance at houses of worship appeared at the Court in July of 2020, a narrow majority led by Chief Justice Roberts adhered to the Lukumi approach. In South Bay United Pentecostal Church v. Newsom, the Court refused to issue an emergency stay of court orders imposing a limit on attendance, on the ground that the state had treated houses of worship the same as the appropriate comparators. As Chief Justice Roberts explained, the right comparators included “lectures, concerts, movie showings, spectator sports and theatrical performances, where large groups of people gather in close proximity for extended periods of time.” And, as Chief Justice Roberts continued, “the Order exempts or treats more leniently only dissimilar activities, such as operating grocery stores, banks, and laundromats, in which people neither congregate in large groups nor remain in close proximity for extended periods.” The notions of similarity and dissimilarity were driven, quite appropriately, by the state’s expert assessment of public health risks.
By November of 2020, however, Justice Ginsburg had passed away and had been replaced by Justice Barrett. The 5‒4 majority that had controlled the outcome in the summer of 2020 became a 5‒4 majority in support of a far more aggressive attitude toward rejecting findings of general applicability. In Roman Catholic Diocese of Brooklyn v. Cuomo, and later in Tandon v. Newsom, the new majority asserted that religious gatherings had a presumptive right to be treated as well as the state treated the most favored secular gatherings, such as retail establishments and factories. The state’s justifications for the different treatment, grounded in expert opinion on the risk of spreading COVID-19, vanished in this approach.
This new formulation of what constituted a failure of general applicability threatened the Smith regime. If any secular exception, however well-grounded in considerations of policy or legislative decisions about coverage of a law, undermines general applicability, religious claimants would be able to successfully seek exemptions from a wide variety of laws of many different kinds.
It is far too soon to know whether this “most favored nation” approach to general applicability will take firm hold across the entire range of religious exemption claims, but we see reasons to be skeptical of that likelihood. First, and true to our conviction that restrictions on worship are at the core of the free exercise guaranty, the COVID-19 cases all involve limitations on group worship, and therefore implicate the most acute constitutional concern. Second, as a matter of process, the COVID-19 cases all arose in the Court’s emergency docket, and so were not subject to full briefing and oral argument. This may help explain why none of the justices who joined in the Court’s opinion in Fulton, including Justices Kavanaugh and Barrett, even mentions the COVID decisions in Fulton.
More substantively, note that where strict review of religious exemption claims is triggered by the presence of a secular exception, the government will inevitably lose. The very presence of any exception demonstrates that the government interest in denying a religious exception falls short of compelling. Perhaps the majority in Fulton, determined to avoid overruling Smith, chose not to advance a theory of general applicability that would make Smith mostly irrelevant, and simultaneously tilt the constitutional scales heavily toward every religious claimant.
Fulton and the relevance of administrative discretion. If we have correctly sized up the silence in Fulton with respect to the “most favored nation” approach to general applicability, what is to be made of the Fulton Court’s reliance on administrative discretion as the factor that undermines general applicability in this case? As we discussed in Part I, the discretion on which the Court relies in Fulton had never been exercised in the context of permitting social welfare agencies to refuse to screen particular classes of prospective foster parents. Such agencies, for example, could not screen only parents whose faith matched the agency’s faith commitments, although such preferences are not uncommon.
The analogy on which the Court relies in Fulton is to the unemployment compensation context. In Sherbert v. Verner, the state agency was engaged in making decisions about whether a claimant had “good cause” to refuse available work. As the Court in Smith viewed that context, it lent itself to individualized governmental assessment of the reasons for the relevant conduct. . . . [A] distinctive feature of unemployment compensation programs is that their eligibility criteria invite consideration of the particular circumstances behind an applicant’s unemployment. . . . [O]ur decisions in the unemployment cases stand for the proposition that where the State has in place a system of individual exemptions, it may not refuse to extend that system to cases of “religious hardship” without compelling reason.
The system in Fulton looked nothing like the system described in Sherbert. Philadelphia had no process in place for evaluating the reasons why private contracting agencies might want exceptions from non-discrimination requirements. The city had created no such exceptions as a matter of substance. The contractual reference to the “Commissioner’s discretion” appeared to be a piece of administrative boilerplate, not a product of a policy judgment about the possibilities of “good cause” to ignore anti-discrimination rules that govern the screening of prospective parents.
As we argued in Part I, Fulton’s treatment of this grant of discretion was a convenience, a way to allow the Court to rule in favor of CSS without addressing the question of whether Smith should be overruled. But this move, if not cabined, threatens to do as much harm as that overruling might do. In the world of administrative agencies—especially at the state and local level—the use of waivers, enforcement discretion, and other ways of creating exceptions to policy is commonplace. Indeed, if enforcement discretion alone undermines general applicability, the entirety of the criminal law will no longer qualify as generally applicable.
In addition, large bodies of law administered by judges, rather than agencies, similarly incorporate devices for the exercise of discretion as a way of mitigating harsh, unjust, or constitutionally troubling results. Such discretion may operate to reinforce general legal principles of fairness, or to protect important interests, some secular and others religious. For example, judges may choose to construe statutes in ways that avoid conflicts with religion, and they may similarly choose to apply general doctrines of equity to avoid hardships produced by religious beliefs, especially in matters involving care and custody of children.
At times, and quite appropriately, administrative and judicial discretion may be exercised in favor of claims of religious accommodation. Accepting an absence from school or work on grounds of religious obligation, for example, may be fair and appropriate. Constitutional norms do not preclude such recognition when they are not designed to advance particular faiths and when they cause no significant harm to others.
But treating every grant of discretion, whether or not exercised, as fatal to a claim of general applicability turns the possibility of permissive religious accommodation into a surprise mandate of accommodation. In Fulton, the Court treated the grant of discretion to the commissioner as the beginning and end of the case in favor of CSS. The mere existence of discretion supposedly made the norm of non-discrimination not generally applicable, and simultaneously defeated any argument from the city that it had a compelling interest in denying the sought-after religious exemption. This was a way for the Court to put Fulton behind it, but its circularity is obvious, and should not be a general and frequent basis for free exercise adjudication going forward.
* * *
Fulton is a signal to lower court judges and litigants that the regime of Smith, in general terms, survived challenge and remains intact. If judges see it in that light, they will be loath to make Smith disappear through the device of finding most laws and regulations not generally applicable. To be sure, free exercise claimants will aggressively push these theories. Judges will occasionally find cases that seem sympathetic on the side of free exercise claimants, and search for ways to make use of such devices in order to rule in favor of religion. We expect that many of these cases will involve clashes between conservative religious groups or individuals, and those who assert rights of LGBTQ equality, a context in which both Masterpiece Cakeshop and Fulton unfortunately have created licenses to lean towards religious claims.
But judges will also see many religious claims that they do not find sympathetic, as well as cases in which the risk of materially undermining important government interests will loom large. These cases will appear in the context of criminal law, tort law, child welfare law, marriage and divorce law, labor and employment law, and administrative regulation of almost every kind. This is the state of affairs against which Justice Scalia warned in Smith, a warning to which Justices Barrett, Kavanaugh, and Breyer appeared sensitive in Fulton.
Lower courts, and the Supreme Court itself, repeatedly worked around this threat in the years between Yoder and Smith. This time around, we expect that lower courts will find ways to resist the move toward shrinking the category of generally applicable laws, because the alternative will be to allow religious claimants to be, again and again, a law unto themselves. We recognize the instability of the current moment in free exercise jurisprudence, but we remain confident that wiser heads, whether or not more plentiful, will continue to prevail.
[*] The authors filed an amicus brief in Fulton, along with two other religion-law scholars, Brief of Professors Ira C. Lupu, Frederick Mark Gedicks, William P. Marshall, and Robert W. Tuttle as Amici Curiae in Support of Respondents, Fulton v. City of Philadelphia, 141 S. Ct. 1868 (2021) (No. 19–123). We were ably represented in the filing of this amicus brief by David Flugman, Caitlin Halligan, and other fine lawyers at the law firm of Selendy & Gay PLLC.
[†] F. Elwood & Eleanor Davis Professor Emeritus of Law, George Washington University Law School.
[‡] David R. and Sherry Kirschner Berz Research Professor of Law and Religion, George Washington University Law School.
 Fulton v. City of PhiladelphiaFulton v. City of Philadelphia, 141 S. Ct. 1868 (2021).
 Fulton v. City of Philadelphia, 922 F.3d 140 (3d Cir. 2019).
 Employment Div. v. Smith, 494 U.S. 872 (1990).
 Fulton, 141 S. Ct. at 1877.
 Id. at 1881–82
 Id. at 1926 (Gorsuch, J., joined by Thomas and Alito, JJ., concurring in the judgment).
 Id. at 1883 (Alito, J., joined by Thomas and Gorsuch, JJ., concurring in the judgment).
 Id. at 1882 (Barrett, J., joined by Kavanaugh, J. and in all but the first paragraph by Breyer, J., concurring).
 See, e.g., City of Boerne v. Archbishop Flores, 521 U.S. 507 (1997).
 Fulton 141 S. Ct. at 1882 (Barrett, J., concurring).
 Hosanna-Tabor Evangelical Lutheran Church & School v. EEOC, 565 U.S. 171 (2012).
 Watson v. Jones, 80 U.S. (13 Wall.) 679 (1871).
 We explore in depth what led to unanimity in Hosanna-Tabor in Ira C. Lupu & Robert W. Tuttle, The Mystery of Unanimity in Hosanna-Tabor Evangelical Lutheran Church & School v. EEOC, 20 Lewis & Clark L. Rev. 1265 (2017) [hereinafter Lupu & Tuttle, The Mystery].
 Id. at 1276–77.
 See, e.g., Engel v. Vitale, 370 U.S. 421 (1962).
 Hosanna-Tabor, 565 U.S. at 196.
 See, e.g., Presbyterian Church v. Mary Elizabeth Blue Hull Memorial Presbyterian Church, 393 U.S. 440 (1969).
 See, e.g., Serbian E. Orthodox Diocese v. Milivojevich, 426 U.S. 696 (1976).
 The Court’s decision in Jones v. Wolf, 443 U.S. 595 (1979), offers an occasional counterweight to the broad idea of “ecclesiastical abstention” reflected in the ministerial exception. That decision permits courts to decide cases that may be resolved solely by “secular law.” In the context of employment disputes, the Supreme Court and lower courts have given little room for that alternative method of resolving cases. See Our Lady of Guadalupe School v. Morrissey-Berru, 140 S. Ct. 2049 (2020). One important exception to that trend are certain claims of a hostile work environment based on sex. For discussion, see Ira C. Lupu & Robert W. Tuttle, #MeToo Meets the Ministerial Exception: Sexual Harassment Claims by Clergy and the First Amendment’s Religion Clause, 25 Wm. & Mary J. of Gender, Race, and Soc. Just. 249 (2019). See also Rachel Casper, When Harassment at Work is Harassment at Church: Hostile Work Environment and the Ministerial Exception, 25 U. of Pa. J. L. & Soc. Change (forthcoming 2021).
 For discussion and critique of a variety of theories of church autonomy, see Lupu & Tuttle, The Mystery, at 1296–99, and sources cited therein.
 In Our Lady of Guadalupe School v. Morrissey-Berru, 140 S. Ct. 2049 (2020), the Court divided 7–2 on application of the ministerial exception to particular elementary school teachers, but no justice rejected the basic principle on which the exception rests.
 Fulton v. City of Philadelphia, 141 S. Ct. 1868, 1881‒82 (2021). For discussion of the superficiality of the Court’s treatment of this question, see Holly Hollman, Court Requires Religious Exemption But Leaves Many Questions Unanswered, SCOTUSblog (June 22, 2021).
 See Josh Blackman, Was There a Double Flip in the November Sitting, Volokh Conspiracy (June 17, 2021) (speculating that the initial assignment of Fulton was to Justice Alito, and the initial assignment of Texas v. California, eventually authored by Justice Breyer, was to Justice Roberts).
 See Amy Howe, Argument Analysis: Justices Sympathetic to Faith-Based Foster-Care Agency in Anti-Discrimination Dispute, SCOTUSblog (Nov. 4, 2020) (analyzing the oral argument in Fulton).
 United States v. Marks, 430 U.S. 188 (1976) (when the Court lacks a majority opinion, the narrowest opinion in support of the result represents the controlling law of the case).
 Fulton, 141 S. Ct. at 1878; Supplemental Appendix to City Respondents’ Brief on the Merits at SA16, Fulton, 141 S. Ct. 1868 (No. 19–123).
 Fulton, 141 S. Ct. at 1878.
 See Employment Div. v. Smith, 494 U.S. 872, 884 (1990).
 The other, even less plausible argument rested on the Court’s aggressive reconstruction of Philadelphia’s Fair Practices Ordinance, which prohibits discrimination in “public accommodations opportunities” based on (among other grounds) an individual’s sexual orientation, constituted a relevant, generally applicable law. Fulton, 141 S. Ct. at 1879–81. The Court construed the ordinance as not covering foster parenting as a “public accommodation,” because the relevant service of certification as a foster parent involved a high degree of selectivity. Whatever the merits of this reading, which we doubt, the Court’s adoption of it flew in the face of carefully reasoned contrary findings by both the U.S. District Court for the Eastern District of Pennsylvania and the U.S. Court of Appeals for the Third Circuit. Both had proximity to local law, and both had determined that the Fair Practices Ordinance does indeed cover the opportunity to serve as a foster parent. See Fulton v. City of Philadelphia, 320 F. Supp. 3d 661, 677 (E.D. Pa. 2018) (stating that “the Parties’ intent that the Fair Practices Ordinance apply to CSS’s services is manifest by the clear and unequivocal terms of the Services Contract”); Fulton v. City of Philadelphia, 922 F.3d 140, 148 (3d Cir. 2019) (applying the Fair Practices Ordinance to foster parents).
 See, e.g., Act of Uniformity, 14 Car. 2 c.4 (1662) (requiring all preachers, professors, and teachers to take an oath affirming the theology and liturgy of the Anglican Book of Common Prayer, on pain of losing their position).
 Church of the Lukumi Babalu Aye, Inc. v. Hialeah, 508 U.S. 520 (1993) is a rare exception, because it involved a bundle of regulations that targeted the worship practice of animal sacrifice by the Santerians.
 Smith itself, as well as the RFRA decision in Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418 (2006) both involved generally applicable, non-targeted policies that had an impact on sacramental practices.
 Smith, 494 U.S. at 907–21 (Blackmun, J., joined by Brennan and Marshall, JJ., dissenting). In Fulton, Justice Alito’s opinion canvasses the many justices who have expressed disagreement with Smith. Fulton, 141 S. Ct. at 1888–89 (Alito, J. concurring).
 See Fulton, 141 S. Ct. at 1900 n.34 (Alito, J. concurring).
 See id.
 See Stephanie Barclay & Mark Rienzi, Constitutional Anomalies or As-Applied Challenges? A Defense of Religious Exemptions, 59 B.C. L. Rev. 1595 (2018). Professor William Marshall was an early, contrary, and important voice in defending Smith on the grounds that it respected the equality of religious and non-religious speech. William P. Marshall, In Defense of Smith and Free Exercise Revisionism, 58 U. Chi. L. Rev. 308 (1991). Justice Alito made no effort to answer Professor Marshall’s quite persuasive argument.
 Laycock, supra note 36.
 42 U.S.C. § 2000bb, 2000bb-4.
 See Douglas Laycock, Free Exercise and the Religious Freedom Restoration Act, 62 Fordham L. Rev. 883, 895–96 (1994); see also Fulton, 141 S. Ct. at 1893 (Alito, J., concurring) (noting then-Rep. Chuck Schumer’s introduction of the bill that became RFRA and President Bill Clinton’s signing of it).
 The leading cases include Masterpiece Cakeshop v. Colorado Civil Rights Commission, 138 S. Ct. 1719 (2017); and Arlene’s Flowers v. Washington, 443 P.3d 1203 (Wa. 2019), cert. denied, No. 19–333, 2021 WL 2742795 (July 2, 2021). See also Brush & Nib Studio LLC v. City of Phoenix, 247 Ariz. 269 (2019); 303 Creative LLC v. Elenis, No. 19–1413, 2021 U.S. App. LEXIS 22449 (10th Cir., July 26, 2021). For a sample of the commentary, pro and con, on the Masterpiece Cakeshop litigation and result, see Masterpiece Cakeshop, Ltd. v. Colorado Civil Rights Commission, SCOTUSblog (last visited Sept. 30, 2021).
 Brief of City Respondents at 47, Fulton, 141 S. Ct. 1868 (No. 19–123). The city’s brief devoted the bulk of its argument regarding Smith to the proposition that considerations of stare decisis favored retention of Smith. Id. at 48–52. The city offered not a single word in defense of Smith’s correctness.
 Brief of Professors Ira C. Lupu, Frederick Mark Gedicks, William P. Marshall, and Robert W. Tuttle as Amici Curiae in Support of Respondents, Fulton, 141 S. Ct. 1868 (No. 19–123).
 Professor Eugene Volokh filed a brief on behalf of neither side. Brief of Professor Eugene Volokh as Amici Curiae in support of Neither Party, Fulton, 141 S. Ct. 1868 (No. 19–123) (arguing that the Court should not overrule Smith, and applauding statutory schemes, such as RFRA, that authorize religious exemptions). On the side of Philadelphia, only two briefs besides ours urged a reaffirmation of Smith. One, from the National League of Cities and other organizations concerned with municipal government, stressed the detrimental impact on their operations that overruling would produce. Brief of the National League of Cities et al. as Amici Curiae in Support of Respondents, Fulton, 141 S. Ct. 1868 (No. 19–123). The other was from the Freedom from Religion Foundation and other atheist organizations. Brief of the Freedom from Religion Foundation, Center for Inquiry, and the American Humanist Association as Amici Curiae in Support of Respondents, Fulton, 141 S. Ct. 1868 (No. 19–123).
 Fulton, 141 S. Ct. at 1889 (Alito, J., concurring).
 Fulton, No. 19‒123, slip op. at 11–12 (June 17, 2021) (Alito, J., concurring).
 Id. at 13. See also id. at 1 (“in . . . Smith, [citation omitted], the Court abruptly pushed aside nearly 40 years of precedent . . .”). In the version of Fulton now posted on LEXIS, these three errors about the time gap between Sherbert and Smith have been corrected. 141 S. Ct. at 1883, 1889–90. As of the date of publication, the errors remained in the slip opinion published at supremecourt.gov.
 Reynolds v. United States, 98 U.S. 145 (1878). Although Justice Alito repudiates the Free Exercise approach taken in Reynolds, Fulton, 141 S. Ct. at 1913 (Alito, J., concurring), he coyly asserts that his “discussion does not suggest that Reynolds should be overruled.” Id. at 1913 n.75. Why not? What compelling interest is served by denying an exemption from anti-bigamy laws to consenting adults who enter a plural marriage for religious reasons? Moves like this are an excellent reminder of the endless possibility of manipulation of strict scrutiny in Free Exercise exemption cases.
 Brief for Professors Lupu et al., supra, note 46, at 7–16. For a related but not identical treatment of the arc of Free Exercise law, see James M. Oleske, Free Exercise (Dis)honesty, 2019 Wisc. L. Rev. 689 (2019).
 This rejection included Minersville School Dist. v. Gobitis, 310 U.S. 586 (1940), which denied a claim by Jehovah’s Witness children that the Free Exercise Clause entitled them to an exemption from the duty to salute the American flag in school. When West Virginia Bd. of Educ. v. Barnette, 319 U.S. 624 (1943) overruled Gobitis, it did so on the ground that the speech clause protected all school children from compulsion to salute the flag, regardless of their reason for objection. See Barnette, 319 U.S. at 634–35, 635 n.15. We put the flag salute story in a wider perspective in Ira C. Lupu & Robert W. Tuttle, Secular Government, Religious People 183–87 (2014). In Justice Alito’s telling of the flag salute story, he distorts the narrative to make it seem as if Barnette rested on the religious beliefs of the children. Fulton, 141 S. Ct. at 1913 (Alito, J., concurring).
 Fulton, 141 S. Ct. at 1895 (Alito, J., concurring). Alito has a lengthy footnote explaining broad judicial departures from the “normal and ordinary” meaning of the term “Congress” within the First Amendment. Id. at 1895 n.27 (noting that the Fourteenth Amendment makes the First Amendment applicable to all functions of state government, and the First Amendment applies to federal administration of the law as well as law-making by Congress). Apparently, “normal and ordinary” can become quite extraordinary and abnormal in judicial hands.
 Id. at 1895 n.29.
 Id. at 1895 n.31.
 Cantwell v. Connecticut, 310 U.S. 296 (1940). In his Fulton opinion, Justice Alito writes, “The Court long ago declined to give the First Amendment’s reference to ‘exercise’ this narrow reading. See, e.g., Cantwell v. Connecticut, 310 U. S. 296, 303–04 (1940).” Fulton, 141 S. Ct. at 1895 (Alito, J., concurring). This is the entirety of his explanation of why the original meaning of “free exercise of religion” extends beyond acts of Divine worship.
 Cantwell rests on freedom of speech, 310 U.S. at 307–11, as well as free exercise of religion, so it becomes less important to decide if street preaching is such an exercise. Jesse Cantwell would have been equally protected by the First Amendment if he had been engaged in political or social advocacy, unrelated to religion.
 Fulton, 141 S. Ct. at 1900 (Alito, J., concurring) (citing Michael McConnell, The Origins and Historical Understanding of Free Exercise of Religion, 103 Harv. L. Rev. 1410, 1445 (1989)).
 Donald S. Lutz, The State Constitutional Pedigree of the U.S. Bill of Rights, 22 Publius 19, 19–20, 27–29 (1992).
 Justice Alito lists many such provisions in his Fulton concurrence. See Fulton, 141 S. Ct. at 1902 n.43 (Alito, J., concurring).
 Id. at 1903.
 Id. at 1902 n. 43. See Northwest Ordinance of 1787, 32 J. of the Cont’l Cong. 334 (July 13, 1787). The Northwest Ordinance enumerated fundamental rights that each territory must respect. Article One of these fundamental rights protects religious liberty. “No person demeaning himself in a peaceable and orderly manner shall ever be molested on account of his mode of worship or religious sentiments in the said territory.” Id. at 340. See also James H. Hutson, Church and State in America: The First Two Centuries 137 (2007) (“On one subject there was unanimity: Governments must not interfere in the spiritual realm, in men’s beliefs and modes of worship.”)
 The central challenge for church-state policies during this period arose from a half-century of exploding diversity within Protestantism. Evangelicalism—reflected in the “First Great Awakening”—challenged traditional forms of religiosity from New England’s Congregationalists to Virginia’s Anglicans. Evangelicalism focused on a religion of the heart, rather than one directed primarily at adherence to creeds or ritual practices. This shift splintered congregations and denominations, while creating or greatly expanding the reach of “dissenting” faith communities (Methodists and Baptists in particular). Although some clergy welcomed the religious vitality of this piety, many others expressed deep concern that its disdain for the settled religious order threatened the peace of the civil community. See generally, Frances Fitzgerald, The Evangelicals: The Struggle to Shape America 13–24 (2017); James H. Hutson, Church and State in America: The First Two Centuries 73–92 (2008); Thomas S. Kidd, The Great Awakening: The Roots of Evangelical Christianity in Colonial America (2007); Patricia U. Bonomi, Under the Cope of Heaven: Religion, Society and Politics in Colonial America 123–167 (rev. ed. 2003); Mark Noll, The Rise of Evangelicalism: The Age of Edwards, Whitefield and the Wesleys (2003).
 See Fulton, 141 S. Ct. at 1902 n.43 (Alito, J., dissenting).
 See supra note 67.
 Hutson, supra note 67; Kidd, supra note 67, passim; Bonomi, supra note 67, at 129–60; Noll, supra note 67, at 128–32, 149–50.
 Kidd, supra note 67, at chs. 4–5; Bonomu, supra note 67, at 157–59.
 Fitzgerald, supra note 67, at 21; Hutson, supra note 67, at 77–79; Bonomi, supra note 67, at 142, 150.
 Kidd, supra note 67, at ch. 5; Noll, supra note 67, at 129–30 (on Rev. Gilbert Tennent, The Danger of an Unconverted Ministry, Considered in a Sermon on Mark 6:34 (1740)).
 Hutson, supra note 67, at 82–89; Bonomi, supra note 67, at 181–84.
 Hutson, supra note 67, at 90.
 Kidd, supra note 67, chs. 5–7; Bonomi, supra note 67, at 147–49.
 Legislatures remained free, at the time of the Founding and still today, to accommodate religious concerns that lay beyond the constitutional guaranty of free exercise of religion. The most famous and obvious example is that of the Quakers, who refused to swear oaths, and refused to bear arms in defense of themselves or their communities. Laws that excused Quakers from oath requirements, conscription, or militia duty recognized their concerns of religious conscience, but the constitutions (state and federal) did not require that. Justice Alito recites the story of legislative accommodations of Quakers and other sects, but he insists that these measures are evidence that the constitution required such exemptions. Justice Scalia correctly saw them as permissive, and not constitutionally mandatory. City of Boerne v. Archbishop Flores, 521 U.S. 507, 541 (1997) (Scalia, J., concurring).
 The Toleration Act, 1 W. & M. c 18 (1689). See Hutson, supra note 67, at 48–56.
 Notably, however, the Act required ministers of dissenting congregations to obtain licenses from the local magistrate, register their place of worship with the local Anglican bishop, and keep the doors to their place of worship open during any meetings of the congregation. The Act granted no right of religious liberty to Roman Catholics or Unitarians. And it required all subjects to attend some place of worship on the Sabbath. Id.; see Ellis M. West, The Free Exercise of Religion: Its Original Constitutional Meaning 44 (2019).
 Rhode Island (Roger Williams) and Pennsylvania (William Penn) were founded by strident advocates of religious liberty and incorporated general provisions for freedom of worship and belief into their founding documents. Nicolas P. Miller, The Religious Roots of the First Amendment: Dissenting Protestants and the Separation of Church and State 60–63 (2012).
 West, supra note 79, at 46; Miller, supra note 80, at 101; Hutson, supra note 67, at 79–81; Bonomi, supra note 67, at 66, 164–66.
 In 1691, the Crown granted a new Royal Charter for Massachusetts, which incorporated the Act of Toleration. Bonomi, supra note 67, at 61. The colony, however, maintained a restrictive practice of funding for recognized religious congregations. Hutson, supra note 67, at 81–82.
 Hutson, supra note 67, at 82–90; Bonomi, supra note 67, at 181–85.
 Hutson, supra note 67, at 90 (“Large numbers of Virginians . . . were comfortable with the definition of liberty of conscience that had been popularized by eighteenth-century British dissenters: freedom from forcible interference with public worship coupled with acquiescence in state intervention in other aspects of religion.”) (The quote focuses specifically on Presbyterians, however, who received significantly better treatment in Virginia than Baptists. See id. at 87.)
 Id. at 91–92. See also West, supra note 79, at 85–86.
 Miller, supra note 80, at 106–08.
 West, supra note 79, at 62–67; Miller, supra note 80, at 144–47; Hutson, supra note 67, at 117, 121–25.
 See Mass. Const. of 1780, arts. II, III.
 Barnes v. Inhabitants of the First Church of Falmouth, 6 Mass. 401 (1810).
 Id. at 408–10.
 Our account of the eighteenth-century history and meaning of the Free Exercise Clause is buttressed by Professor Rakove’s excellent recent study of the subject. Jack N. Rakove, Beyond Belief, Beyond Conscience: The Radical Significance of the Free Exercise of Religion (2020). See especially chaps. 2–3.
 This failure simply demonstrates the inadequacy of originalist interpretation—and the inevitable temptation to shape that interpretation to support a desired outcome. Peter J. Smith, Originalism and Level of Generality, 51 Ga. L. Rev. 1, 10 (noting that in practice the application of originalism “appears ad hoc, largely unconstrained, and thus susceptible to the same kind of results-oriented decision-making that originalists have long decried”). The same should be said of the Court’s emphasis on “history and tradition” as the basis for denying that government displays and religious speech—which invariably reflect Christian beliefs—violate the Establishment Clause. See, e.g., Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067 (2019) (upholding display of Latin Cross as a secular memorial to those who died serving the U.S. in World War I); Town of Greece v. Galloway, 572 U.S. 565 (2014) (upholding a pattern of predominantly Christian prayer to open Town Council meetings).
 See generally West, supra note 79 (arguing that the best originalist interpretation of the Free Exercise Clause limits the scope of its protection to a domain of “religious worship and practice” that is distinct from civil authority’s power over laws “protecting and promoting the earthly wellbeing of persons.” Id. at 197).
 Employment Div. v. Smith, 494 U.S. 872, 888 (1990). This statement is followed by a long list of the categories of law that would be vulnerable to exemption claims under Justice Alito’s approach. See id. at 889.
 Fulton v. City of Philadelphia, 141 S. Ct. 1868, 1897 (2021) (Alito, J., concurring). Justice Barrett repeats this charge. Id. at 1883 (Barrett, J., concurring).
 Bd. of Educ. of Kiryas Joel Sch. Dist. v. Grumet, 512 U.S. 687 (1994); Larson v. Valente, 456 U.S. 228 (1982).
 Church of the Lukumi Babalu Aye, Inc. v. Hialeah, 508 U.S. 520 (1993).
 Smith, 494 U.S. at 877–78.
 See generally Douglas NeJaime & Reva B. Siegel, Conscience Wars: Complicity-Based Conscience Claims in Religion and Politics, 124 Yale L.J. 2516, 2518–19 (2015) (analyzing the political and legal contexts in which people seek exemptions from laws that make “objectors complicit in the assertedly sinful conduct of others.”).
 Masterpiece Cakeshop v. Colorado Civil Rights Comm’n, 138 S. Ct. 1719 (2017).
 Burwell v. Hobby Lobby, 573 U.S. 682 (2014).
 Robin West describes these claims as ways that dissenters from egalitarian and feminist norms seek to exit the social contract concerning equal opportunities for all. Robin West, Freedom of the Church and Our Endangered Civil Rights: Exiting the Social Contract, in The Rise of Corporate Religious Liberty, (Zoe Robinson, Chad Flanders & Micah Schwartzman eds., 2015).
 Sherbert v. Verner, 374 U.S. 398 (1963), involved just such a question. Justice Brennan describes the loss of unemployment compensation as a result of being unavailable for work on Saturday (Adele Sherbert’s Sabbath day) as the conceptual equivalent of a fine imposed upon her for Saturday worship. Id. at 404.
 Thomas v. Review Bd., Ind. Empl. Sec. Div., 450 U.S. 707 (1981).
 Hobby Lobby, 573 U.S. at 723–26 (accepting the assertion by the owners of Hobby Lobby that the contraceptive mandate substantially burdens their religious belief).
 This problem is well-analyzed in Frederick Mark Gedicks, Substantial Burdens: How Courts May (and Why They Must) Judge Burdens on Religion Under RFRA, 85 Geo. Wash. L. Rev. 94 (2017); see also Ira C. Lupu & Robert W. Tuttle, Secular Government, Religious People 241–44 (2014) (discussing how the ruling in Thomas effectively makes each religious exemption claimant a judge in her own cause).
 Fulton v. City of Philadelphia, 141 S. Ct. 1868, 1924 (2021) (Alito, J. concurring). He leaves open the possibility that “this test [might be] rephrased or supplemented with specific rules.” Id. This is a neat invitation to weasel out of the test when it produces results with which a justice is unhappy, as so often happened under the Sherbert-Yoder regime. Moreover, Justice Alito ignores the fact that Smith replaced a bundle of standards, of which this was the among the strictest. Note as well that Justice Alito subtly shifts away from a test of “least restrictive means” to one of “narrow tailoring,” even though RFRA (designed by Congress to reinstate pre-Smith law) uses the test of least restrictive means. 42 U.S.C. § 2000bb-1. The test of least restrictive means is even more difficult to satisfy than that of “narrow tailoring,” because the government can always find a means, albeit more expensive or less efficient, less restrictive of religious liberty than the one challenged. See Hobby Lobby, 573 at 764–68 (Ginsburg, J., dissenting). See also Ira C. Lupu, Hobby Lobby and the Dubious Enterprise of Religious Exemptions, 38 Harv. J. L. & Gender 35, 86–90 (2015) (analyzing treatment of “least restrictive means” in Hobby Lobby and elsewhere).
 In Hobby Lobby, Justice Alito implies that only the interest in avoiding racial discrimination, and not other kinds of discrimination, is compelling. See Hobby Lobby, 573 U.S. at 733.
 For this kind of reasoning about the connection between under-inclusion and the weight of government interests, see Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 432–33 (2006); Hobby Lobby, 573 U.S. at 727. Both are RFRA decisions, but the statutory standard under RFRA tracks the free exercise standard for which Justice Alito contends in Fulton.
 See O Centro, 546 U.S. 418 (applying RFRA to exempt the group from a broad ban on importation of huasca tea, because the government could take the more narrowly tailored step of monitoring the group’s use of the hallucinogenic substance).
 In the Fulton briefing, lawyers calling for Smith to be overruled conveniently ignored these work-around cases, because the pattern disturbs their preferred (but false) narrative that Smith abruptly abandoned a consistent posture of strict scrutiny of free exercise exemption claims. But we called attention to them in our amicus brief in Fulton, supra note 44, at 12–16. Justice Barrett (a former Justice Scalia clerk) was definitely attuned to them. Fulton, 141 S. Ct. at 1883 (Barrett, J., concurring). For additional discussion, see Lupu, supra note 107, at 51–53; McConnell, supra note 33, at 1110 (describing Supreme Court’s free exercise law on the eve of Smith as a “Potemkin doctrine.”).
 O'Lone v. Estate of Shabazz, 482 U.S. 342 (1987).
 Goldman v. Weinberger, 475 U.S. 503 (1986).
 Lyng v. Northwest Indian Cemetery Protective Ass’n, 485 U.S. 439 (1988); see also Bowen v. Roy, 476 U.S. 693 (1986); Tony & Susan Alamo Foundation v. Sec’y of Labor, 471 U.S. 290 (1985).
 United States v. Lee, 455 U.S. 252 (1982); Bob Jones Univ. v. United States, 461 U.S. 574 (1983). In both decisions, the Court failed to inquire into the availability of means less restrictive of religious liberty.
 See James E. Ryan, Smith and the Religious Freedom Restoration Act: An Iconoclastic Assessment, 78 Va. L. Rev. 1407, 1416–37 (1992). Indeed, with the benefit of hindsight, we can see that almost all of the successful free exercise exemption claims in this period involved both unemployment compensation and questions involving worship. Yoder and Thomas are the outliers. Both involve religiously motivated conduct outside of worship, and both have been the source of great controversy. Perhaps, intuitively, the Supreme Court was leaning toward a notion of free exercise as worship without ever so declaring.
 See, e.g., United States v. Stevens, 559 U.S. 460 (2010); Boos v. Barry, 485 U.S. 312 (1988).
 See, e.g., Ward v. Rock Against Racism, 491 U.S. 781 (1989).
 See United States v. O’Brien, 391 U.S. 367 (1968). In cases evaluating content-neutral restrictions on conduct that involves symbolic speech, the Court has consistently ruled that government actions were constitutionally permissible despite incidental limitations on some expressive activity. In addition to O’Brien, see, e.g., Barnes v. Glen Theatre, Inc., 501 U.S. 560, 567 (1991); City of Erie v. Pap’s A.M., 529 U.S. 277, 296 (2000).
 See, e.g., Cohen v. Cowles Media, 501 U.S. 663, 669–70 (1991) (“[G]enerally applicable laws do not offend the First Amendment simply because their enforcement against the press has incidental effects on its ability to collect and report the news.”) (collecting cases); Branzburg v. Hayes, 408 U.S. 665, 682 (1972) (“It is clear that the First Amendment does not invalidate every incidental burdening of the press that may result from the enforcement of civil or criminal statutes of general applicability.”).
 The gap between protection of religious speech and all other speech under such an approach is constitutionally unacceptable. See William P. Marshall, What is the Matter with Equality? An Assessment of the Equal Treatment of Religion and Nonreligion in First Amendment Jurisprudence, 75 Ind. L.J. 193 (2000).
 Fulton v. City of Philadelphia, 141 S. Ct. 1868, 1883 (2021) (Barrett, J., concurring).
 Several of her questions about review standards emerge from the same concern about whether courts will consistently ask the same, strict questions about regulatory impacts on religion: “What forms of scrutiny should apply? Compare Sherbert v. Verner, 374 U. S. 398, 403 (1963) (assessing whether government’s interest is ‘compelling’), with Gillette v. United States, 401 U.S. 437, 462 (1971) (assessing whether government’s interest is ‘substantial’)” and “Should there be a distinction between indirect and direct burdens on religious exercise? Cf. Braunfeld v. Brown, 366 U. S. 599, 606–07 (1961) (plurality opinion).” Fulton 141 S. Ct. at 1883 (Barrett, J., concurring).
 Justice Barrett added the citation: “Cf. Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, 565 U. S. 171 (2012).” Id. As we discuss in Part I, Hosanna-Tabor does not declare a general doctrine of privilege or autonomy for religious entities. See supra notes 13–25 and accompanying text. Rather, the decision reinforces the concept of judicial abstention on exclusively ecclesiastical questions. Perhaps Justice Barrett intends the “Cf.” signal as a very shorthand way to communicate that distinction.
 Religiously affiliated hospitals, serving the general public and heavily supported by public money, are in the same boat as religious charities. They should not be free to claim mandatory free exercise clause protection for their refusal to provide necessary services to women or to transgender patients. Their status may soon get tested in the Supreme Court. See Dignity Health v. Minton, 39 Cal. App. 5th 1155 (Cal. App. 1 Dist., 2019), petition for cert. filed (U.S. Mar. 13, 2020) (No. 20–1135). For a sophisticated discussion of related issues, see Elizabeth Sepper, Zombie Religious Institutions, 112 Nw. U. L. Rev. 929 (2018).
 “See Smith, 494 U. S., at 888–889.” Fulton, 141 S. Ct. at 1883 (Barrett, J., concurring). At the cited pages from Smith, Justice Scalia listed the many free exercise work-around cases from the 1980s.
 Soon after Fulton, the Court denied certiorari in two cases that had raised the issue of overruling Smith. Ricks v. Idaho Contractors Bd., 435 P.3d 1 (Idaho 2018) cert. denied, No. 19–66, 2021 WL 2637837 (June 28, 2021); Arlene’s Flowers, Inc. v. Washington, 441 P.3d 1203 (Wash. 2019), cert. denied, No. 19–333, 2021 WL 2742795 (July 2, 2021). Justices Alito, Gorsuch, and Thomas noted that they would grant certiorari in Arlene’s Flowers.
 Church of the Lukumi Babalu Aye, Inc. v. Hialeah, 508 U.S. 520 (1993).
 South Bay United Pentecostal Church v. Newsom, 140 S. Ct. 1613 (2020).
 Id. at 1613 (Roberts, C.J., concurring).
 Id.; see also Calvary Chapel Dayton Valley v. Sisolak, 140 S. Ct. 2603 (2020) (upholding similar Nevada orders with respect to gatherings at houses of worship).
 Roman Catholic Diocese of Brooklyn v. Cuomo, 141 S. Ct. 63 (2020).
 Tandon v. Newsom, 141 S. Ct. 1294 (2021).
 For defense of the “most favored nation” theory, see, e.g., Douglas Laycock & Steven T. Collis, Generally Applicable Law and the Free Exercise of Religion, 95 Neb. L. Rev. 1 (2016). For critique of this approach, see James Oleske, Lukumi at Twenty: Legacy of Uncertainty for Religious Liberty and Animal Welfare Laws, 19 Animal L. Rev. 295 (2013). For a constitutionally expansive and more mixed evaluation of the approach, see Nelson Tebbe, The Principle and Politics of Equal Value, 121 Colum. L. Rev. (forthcoming 2021). See also Christopher G. Lund, A Matter of Constitutional Luck: The General Applicability Requirement in Free Exercise Jurisprudence, 26 Harv. J. L. & Pub. Pol’y 627 (2003).
 Note the Court’s treatment of the decision in Danville Christian Academy, Inc. v. Beshear, 981 F.3d 505 (6th Cir. 2020), which involved restrictions on in-person attendance at elementary and secondary schools, religious and secular. The state had not imposed equally demanding restrictions on pre-schools or institutions of higher education, so a “most favored nation” argument was available in this case, but the Court did not deploy it. Danville Christian Academy Inc. v. Beshear, 141 S. Ct. 527 (2020).
 Fulton v. City of Philadelphia, 141 S. Ct. 1868, 1881–82 (2021). In O Centro, a RFRA case, the Supreme Court used this move, in which underinclusive coverage of a policy is fatal to the government’s case for compelling interest when the policy is challenged for not exempting religiously motivated conduct. See Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 433 (2006). For discussion of why the Court did not rely on this argument in Hobby Lobby, despite its availability, see Lupu, supra note 107, at 82–86.
 See Rogers v. U.S. Dept. of Health & Human Servs., 466 F. Supp. 3d 625 (D.S.C. 2020). In Rogers, the plaintiffs claimed that the federal and state defendants violated the Establishment Clause by contracting with a faith-based foster care agency that refused to place children in any homes that did not conform to the requirements of the agency’s religious beliefs and practices. The court denied defendants’ motions to dismiss the case and the matter remains in discovery. Although federal and state laws and regulations prohibit contractors from discriminating based on religion, the then-Secretary of HHS and Governor of South Carolina granted waivers to permit the contract at issue. For further discussion of Rogers, including the impact of Fulton on the litigation, see Robert W. Tuttle, Foster Care and the Unsettled Religion Clauses, 60 Fam. Ct. Rev. (forthcoming 2022).
 Employment Div. v. Smith, 494 U.S. 872, 884 (1990).
 In the context of anti-discrimination laws, we urge drafters of policy instruments to remove, and to explicitly renounce, the possibility of discretionary exceptions. Leaving such discretion in place invites a replay of Fulton.
 In 303 Creative LLC v. Elenis, 6 F.4th 1160 (10th Cir. 2021), petition for cert. filed, 2021 WL 4459045 (U.S. Sept. 24, 2021) (No. 21‒476), Chief Judge Tymkovich argued in dissent that the discretion lodged in the Colorado Civil Rights Commission by its processes of enforcement and adjudication made the state’s law not “generally applicable.” Id. at 1206–09 (Tymkovich, C.J., dissenting). The discretion to which he pointed was no more than the agency’s normal process of investigating and adjudicating claims, finding some but not others to be violations of the state’s anti-discrimination laws.
 See, e,g., NLRB v. Catholic Bishop of Chicago, 440 U.S. 490 (1979) (construing National Labor Relations Act to exclude jurisdiction over church-operated schools that teach secular and religious subjects).
 See, e.g., Trans World Airlines, Inc. v. Hardison, 432 U.S. 63 (1977) (recognizing duty of religious accommodation in the private workplace under Title VII of the 1964 Civil Rights Act, so long as the accommodation produces no more than de minimis harm to employer or other employees).