University of Notre Dame

Religious Liberty and Judicial Deference

November 3, 2022

View PDF



Mark L. Rienzi*

Many of the Supreme Court’s most tragic failures to protect constitutional rights—cases like Plessy v. Ferguson, Buck v. Bell, and Korematsu v. United States—share a common approach: an almost insuperable judicial deference to the elected branches of government.  In the modern era, this approach is often called “Thayerism,” after James Bradley Thayer, a nineteenth-century proponent of the notion that courts should not invalidate actions of the legislature as unconstitutional unless they were clearly irrational.  Versions of Thayerism have been around for centuries, predating Thayer himself.

The Supreme Court took a decidedly Thayerian approach to the First Amendment in the first flag salute case, Minersville School District v. Gobitis.  That approach was short-lived, as Gobitis was swiftly overruled in West Virginia State Board of Education v. Barnette.  Rather than deferring to political actors, Barnette treated the Constitution as placing certain rights “beyond the reach of majorities” and establishing them as “legal principles” that must be “applied by the courts.”  Barnette’s approach to rights—rejecting a Thayerian “duty of deference” for First Amendment rights—has largely triumphed, even in other individual rights contexts.

But a curious anomaly persists.  Unlike in other areas of the law, the discredited Thayerian approach to the First Amendment from Gobitis was eventually adopted into the modern free exercise standard embraced by the Supreme Court in Employment Division v. Smith.  As a result, many free exercise claims have been decided with precisely the kind of rational basis deference we long ago abandoned for other constitutional rights.

This Article examines the relationship between religious liberty claims and Thayerian judicial deference.  With the Supreme Court poised to reconsider Smith, this focus on deference differs from the standard scholarly and judicial approach, which tends to emphasize the debate over religious exemptions.  Focusing instead on deference shows how Smith is an outlier, out of step not only with prior religious liberty cases but also with our broader approach to the enforcement of constitutional rights.  Likewise, when religious liberty is viewed through the lens of deference, it becomes clear that, even without overruling Smith, the Supreme Court has been moving away from Thayerian judicial deference across a wide range of religious liberty disputes over the past decade.  These deference-rejecting decisions cast the Religion Clauses as the “the heart of our pluralistic society,” that help “foster a society in which people of all beliefs can live together in harmony.”  Those high goals are only attainable if religious liberty consists of judicially enforceable rights, rather than occasions for deference to the majoritarian governments that the Bill of Rights is supposed to constrain.


Eighty years ago, the Supreme Court decided a pair of cases about whether the government could force children from a minority religious group to pledge allegiance to the American flag.  The two cases reached opposite results, with one allowing the forced pledge and one invalidating it.  The key difference between the two was a shift in the Court’s view on the question of deference: How much should judges defer to political actors?

In the first case, Minersville School District v. Gobitis (1940), the Court ruled that the government could force the Jehovah’s Witness children to salute the flag.  The Gobitis Court fully understood the children’s religious objection, and claimed to view the minority’s rights of conscience as “so subtle and so dear.”  But the Court thought it had a “duty of deference” that obligated it to yield to the local majority’s view that coercion would instill patriotic impulses in children.

The deferential approach taken in Gobitis was not new.  As James Bradley Thayer explained in 1893, there had long been a strain of judicial thinking that urged courts to defer to rational legislative decisions, even on constitutional questions.  “Thayerism,” as the approach became known, can be seen in a host of infamous constitutional rights cases, including Plessy v. Ferguson, Buck v. Bell, and Korematsu v. United States.  Justice Felix Frankfurter, who wrote the Court’s opinion in Gobitis, was an acolyte of Thayer and thought Thayer’s 1893 essay setting forth this deferential approach “was the most important thing ever written about the Constitution.”

The Court’s embrace of judicial deference in Gobitis was short-lived.  Just three years later, in West Virginia State Board of Education v. Barnette, the Court announced essentially an anti-Thayerian approach to the First Amendment.10   Eschewing deference, Barnette endorsed “the right to differ as to things that touch the heart of the existing order” and said that, even on a majority vote, “no official, high or petty” could force minorities to embrace the majority’s orthodoxy in religion or other matters.11   In the Barnette view, protecting minority beliefs and practices from majority coercion is the very point of the Bill of Rights.  Where important minority rights are threatened by majoritarian government, courts cannot merely defer.

I want to suggest that this conflict over deference is actually at the heart of much of our modern religious liberty jurisprudence.  To be sure, the conventional wisdom is that the deferential Thayerian approach to the First Amendment set forth in Gobitis is dead, and that the Barnette approach of judicial enforcement of constitutional rights controls.12   That conventional wisdom is mostly correct: Barnette’snondeferential understanding of how courts and the Constitution protect rights is broadly embraced by Justices and commentators across the ideological spectrum.13   Barnette even transcends the First Amendment and is often invoked as a key precedent for understanding how constitutional rights work in other important individual rights contexts.14   This approach to rights is widely understood as providing essential judicial protection for minority rights and pluralism.15 

But that approach has not yet fully extended to religious liberty.  Nearly fifty years after Barnette, the Supreme Court actually relied on Gobitis in 1990 when it embraced a restrictive approach to the Free Exercise Clause in Employment Division v. Smith.16   While the Court had previously taken a Barnette-style approach in religious liberty cases, Smith relied instead on Gobitis as correctly stating the general rule of deference, namely that the Free Exercise Clause offers little protection against many general laws imposed by legislative majorities.17 

Most scholars of the Smith decision have justifiably focused on the issue of religious exemptions.18   That makes sense, because Smith framed exemptions as the central issue.19   But I argue here that much can be learned from looking at Smith through the lens of the competing approaches to deference that motivated the Gobitis-to-Barnette reversal.  At a time when the Supreme Court seems poised to reconsider Smith, this analysis of free exercise law through the lens of deference can help both to elucidate why Smith was wrong, and why it has remained so out of step with the Court’s treatment of virtually all other individual rights.

Close attention to the question of deference also provides the best explanation for the past decade of Supreme Court religious liberty decisions.  A string of recent decisions—involving a wide variety of religious liberty claims, a diverse group of religious plaintiffs, and often broad cross sections of the Court coming to surprising agreement—suggests that the Supreme Court is rejecting the deferential Thayerian approach to religious liberty.  The Court has repeatedly emphasized the First Amendment’s role as the “guarantee [that] lies at the heart of our pluralistic society”20 —something the First Amendment simply could not do under a deferential Gobitis understanding of rights.  And while both Gobitis and Smith seemed to fear judicial enforcement of the Religion Clauses, the Court’s recent cases, including last Term’s unanimous decision in Fulton v. City of Philadelphia,21  suggest that the Court views such enforcement as both necessary and salutary for the Clause to help “foster a society in which people of all beliefs can live together harmoniously.”22   In the process, these cases have brought the Court’s religious liberty jurisprudence into closer alignment with its treatment of other fundamental rights.

This Article has five parts.  Part I explores the deference-based Thayerian understanding of constitutional rights that led to the Court’s willingness to defer to school boards about the forced flag salute in Gobitis.  Part II then discusses the prompt rejection of this deferential approach to individual rights in Barnette’s overruling of Gobitis.  Part III explores how the Court’s leading Free Exercise precedent, Smith, is best understood as embracing the Thayerian deference of Gobitis.  Part IV analyzes the Court’s efforts over the past decade to reorient the law of religious liberty away from the narrow, deferential approach and toward the path of judicial protection for minority rights and pluralism described in Barnette.  Part V concludes by discussing the prospects for the Court’s ultimate success in fully eradicating the impact of Thayerian judicial deference on religious liberty and fully embracing the First Amendment as a strong and enforceable protection for peaceful pluralism amidst differences.


Gobitis and Barnette reached opposite answers on the question of whether governments can impose a forced flag salute and pledge on unwilling students.  That outcome was and remains important, both for the particular students and communities affected, and for a free society more broadly.

That difference in outcomes is attributable to an even more important difference between Gobitis and Barnette on the question of deference.  The cases differ sharply as to how and whether judges should defer to political actors.  As will be discussed later, understanding Gobitis and Barnette through this lens sheds important light on the modern approach to religious liberty, both because the Court would later adopt the Gobitis approach of deference in Smith, and because the more recent religious liberty cases are best explained as a broad rejection of such deference.

At the time of Gobitis, thinking about the judicial role was heavily influenced by James Bradley Thayer’s paper “The Origin and Scope of the American Doctrine of Constitutional Law.”23   Alexander Bickel deemed Thayer’s article a “singularly important piece of American legal scholarship,” because of its influence on Holmes, Brandeis and the Justice who would write Gobitis, Felix Frankfurter.24   Frankfurter would later call Thayer’s argument the “most important single essay” about American constitutional law and “the great guide for judges.”25 

A.   Thayer’s Theory: “Whatever Choice Is Rational Is Constitutional.”

Thayer argued that courts presented with constitutional questions actually were not supposed to decide whether a law is unconstitutional “upon a just and true construction.”26   Instead of deploying their own constitutional analysis, Thayer argued that judges should apply a version of what today we might call a “clear mistake” rule or “rational basis” test.  Under it, courts can only invalidate an act of the political branches as unconstitutional “when those who have the right to make laws have not merely made a mistake, but have made a very clear one,—so clear that it is not open to rational question.”27   Thayer argued that because “the constitution often admits of different interpretations” and “there is often a range of choice and judgment,” courts should approach constitutional questions with the approach that “whatever choice is rational is constitutional.”28 

Thayer supported this approach largely out of separation-of-powers concerns.  He argued that state and federal constitutions had carefully separated legislative power from judicial power to ensure “a government of laws, and not of men.”29   Federal judges had not been given the authority to sit as a third branch of the legislature—they were not a “Council of Revision” to look over laws as they were enacted, but instead had been given only the power to decide particular cases that came before them.30   Courts needed to be careful not to invalidate a law “merely because it is concluded that upon a just and true construction the law is unconstitutional.”31   Legislators, rather than judges, had to be given room to balance the “complex, ever-unfolding exigencies of government.”32   Judicial balancing of such concerns must be avoided, lest it turn the court “into a board for answering legislative conundrums.”33 

Thayer thought this deferential approach was beneficial not only for the courts but also for the legislative process.  He believed that judicial review might diminish the likelihood of serious constitutional consideration by the legislature and, ultimately, the people.34   Thayer feared that a robust allowance for judicial enforcement of the Constitution would leave legislatures indifferent to questions of constitutionality: “if we are wrong, they say, the courts will correct it.”35   His goal in limiting judicial review to a search for irrationality was to expound “the clear limits of judicial power; so that responsibility may be brought sharply home where it belongs,” namely to the people and the legislature.36 

Thayer did not focus his argument on individual rights.  He was writing at a time before the Supreme Court had “incorporated” the Bill of Rights, and his article focused largely on structural aspects of the Constitution.37   But his followers—including both Learned Hand and Justice Frankfurter—would apply Thayer’s rule to the Bill of Rights.  Hand, for example, reached the “conclusion that courts should defer to legislative judgments even when First Amendment claims were at stake.”38   While giving the 1958 Oliver Wendell Holmes lecture at Harvard Law School, Hand argued that Thayer’s rule of rationality should apply to cases brought under the Bill of Rights.39   Ronald Dworkin described Hand’s extension of Thayer’s views into the area of individual rights as “the strongest doctrine of [judicial] restraint ever defended by a major judicial figure.”40 

How strong was that doctrine of judicial restraint?  It was strong enough that it eventually led both Hand and Frankfurter to doubt the legitimacy of the Court’s unanimous decision in Brown v. Board of Education.41   For Hand, that doubt was expressed publicly, in his Holmes Lectures at Harvard.42   There, he argued that Brown looked like it was just a judicial rebalancing of interests that the legislature had already conducted, rather than application of a constitutional principle against racial classifications.43   If segregation was simply an available rational choice of the legislature, then the Court would be impermissibly acting as a “third legislative chamber” to reverse it.44 

In his biography of Hand, Gerald Gunther argues that Hand’s interpretation of Brown “came directly from Felix Frankfurter.”45   This may seem odd, given that Frankfurter signed onto the unanimous decision in Brown.  But Frankfurter had reason to want Brown to be interpreted narrowly: he feared that extending Brown to invalidate bans on interracial marriage would imperil the Court’s legitimacy and jeopardize desegregation.46   Frankfurter therefore argued, and Hand eventually agreed, that Brown was not a broad statement of constitutional principle after all, but rather a limited decision that only concerned the field of public education.47   This move created room to say that Brown would not necessarily invalidate bans on interracial marriage.  It also meant that, in Thayerian terms, Brown was illegitimate because the Court should have left such context-dependent balancing to the legislature, “and this had to be condemned by Hand.”48 

B.   Thayerism in Plessy, Buck, and Korematsu

By design, Thayer’s theory of judicial review leaves only a very small role for judges to enforce the Constitution.  In practice, such an approach means that the Constitution will provide very little protection for individual rights.  If courts are forbidden from consulting their own best understanding of what the Constitution requires and must instead accept all government actions based on all nonirrational interpretations of the Constitution, then the Constitution will almost never provide enforceable protection against majoritarian government power.

It is therefore unsurprising that Thayerism is evident in many cases that are rightly regarded as judicial failures to protect minority rights against the power of majoritarian government.  Consider, for example, the Court’s infamous decision in Plessy v. Ferguson, which was decided just three years after Thayer’s essay.49   Faced with a constitutional provision that guaranteed the “equal protection of the laws,” the Court found that the constitutional question “reduces itself to the question whether the statute of Louisiana is a reasonable regulation.”50 

To make matters worse, the Plessy Court emphasized that this reasonableness test “must necessarily” recognize “a large discretion on the part of the legislature.”51   The reasonableness test must further allow the state to act based on the established “usages” and “customs” of the people to further the “promotion of their comfort.”52   Applying this standard, the Court said it could not say that segregation was “unreasonable”53 —thus condemning the country and the Constitution to the next half century of Jim Crow laws.

Plessy vividly illustrates the problem Thayerism creates for the judicial protection of constitutional rights.  If constitutional rights can be reduced to a mere reasonableness test, then those rights offer very little protection.  That is especially true where that reasonableness test allows for “large discretion” on the part of the elected branches of government.54   Legislatures can, of course, still choose to protect rights when they wish; but courts will only rarely see fit to require protection when the majoritarian legislature chooses not to provide it.55 

Thayerism can also be seen in Justice Holmes’s opinion for the Court in Buck v. Bell, in which it denied protection for an eighteen-year-old “feeble-minded” woman, who was forcibly sterilized by the State of Virginia.56   The state legislature had authorized forced sterilizations of “mental defectives” and the Court refused to second-guess the legislature’s prerogative to require sterilization of what it called “the probable potential parent of socially inadequate offspring.”57 

The Court grounded its holding in the reasonableness standard from Jacobson v. Massachusetts,58  which held that states could impose health restrictions amid “the pressure of great dangers” so long as they were “reasonable.”59   The Buck Court thought it reasonable for the State to sterilize those who are “manifestly unfit” so as “to prevent our being swamped with incompetence.”60   The Court also suggested that the legislature’s line drawing about whom to sterilize should not be second-guessed.61 

Thayerian deference is also the centerpiece of Korematsu.62   Although Korematsu begins with the claim that the Court must impose “the most rigid scrutiny” because laws restricting civil rights by racial group are “immediately suspect,” the Court’s actual analysis applied Thayerian deference.63   At every turn, the Court emphasized its own inability or unwillingness to second-guess the judgments of Congress or the military authorities.64   As Thayer had said, “whatever choice is rational is constitutional,”65  and the Korematsu Court thought Japanese internment was at least rational in the circumstances.

The common theme in all of these cases is that the Court allowed a significant infringement on liberty without any significant scrutiny of the government’s claimed reasons for the infringement.  As we shall see below, Justice Frankfurter’s opinion for the Court in Gobitis took a similarly deferential approach to the Bill of Rights, in which the Court could only intervene to protect constitutional liberties if it found the legislature’s law irrational.

C.   Thayerism in Gobitis

1.   Background: “I love my country and I love God more.”

Ten-year-old Billy Gobitas loved his country.  But he loved God first and believed that complying with his school’s requirement to pledge allegiance to the American flag was forbidden by God.66   So, in 1935, he wrote a letter to school officials explaining that he could not salute the flag.  Considering it a form of idol worship, Billy cited the Book of Exodus.67   He explained that God enjoined the people not to make “any graven image, nor bow down to them.”68   Echoing the Madisonian formulation of religious duties preceding political ones,69  Billy emphasized that he did love his country, but had to obey God: “I do not salute the flag not because I do not love my country, but I love my country and I love God more and I must obey His commandments.”70 

As a result of his refusal, Billy and his family endured harassment, expulsion, and boycotts of the family store.71   The family eventually sued the school board, arguing that being forced to say the pledge violated their constitutional right to the free exercise of religion.72 

The family won in both the trial court and the court of appeals, with both courts insisting that government can only force someone to violate his or her religious beliefs if it can prove such coercion is necessary to the protection of important governmental interests.73   Neither court suggested that judges should simply defer to any rational choice by the legislature.74   Eventually, however, the family lost their case at the Supreme Court, where the Justices felt compelled to defer to the views of the school board.75 

2.   Deference trumps constitutional enforcement

The Court’s decision against the religious liberty claim in Gobitis reflects the narrow, Thayerian view of the role of courts in protecting even constitutionally enumerated rights.  Three themes emerge:

a.   Legislatures over Courts

The Gobitis decision was chiefly driven by the Thayerian belief that legislatures, rather than courts, should have primary responsibility for protecting constitutional rights.  This principle was evident in Gobitis itself and was expounded at greater length in Justice Frankfurter’s Barnette dissent, in which he defended the Gobitis approach.

To the Gobitis Court, the Jehovah’s Witness plaintiffs should have addressed their plea for protection to the political branches of the government.  The Justices explained that “the courtroom is not the arena for debating issues of educational policy.”76   Rather it is for legislators, not judges, “to choose among competing considerations in the subtle process of securing effective loyalty to the traditional ideals of democracy.”77   They maintained that it is for legislators, not judges, to think about how to simultaneously respect “individual idiosyncracies [sic] among a people so diversified in racial origins and religious allegiances.”78 

Like Thayer, the Gobitis Court thought that leaving such issues to the legislature was a matter of both proper judicial role and good training in democratic impulses.  As to judicial role, the Court believed that granting judicial protection for religious minorities under the First Amendment “would in effect make us the school board for the country,” which the Court thought was beyond its constitutional authority.79   Protection for “the most precious interests of civilization” therefore needs to be found in the legislature rather than through seeking “vindication in courts of law.”80 

Like Thayer before him, Justice Frankfurter viewed the lack of judicial authority over the protection of individual rights as a good thing.  Leaving most of the protection of constitutional rights to the political branches would force the people to “fight out the wise use of legislative authority,” and would “serve[] to vindicate the self-confidence of a free people” better than transferring the contest “to the judicial arena.”81 

This view of the proper role for courts in turn dictated Justice Frankfurter’s understanding of the single available test the Court could apply for constitutionality.

b.   Deference to Rational Legislative Choices

Because it viewed legislatures as the primary guardians of individual liberty, the Gobitis Court asserted that courts should almost always defer to legislative policy choices.  Gobitis acknowledged a few rare exceptions—chiefly if the laws were targeted against a particular group (i.e., “directed against doctrinal loyalties of particular sects”)82  or when the political branches are somehow broken (i.e., when “the effective means of inducing political changes” are not “free from interference”).83   But it otherwise endorsed judicial deference to legislatures.84 

Like Thayer, Gobitis described this deference to the legislature in terms consistent with what today we would call “rational basis” review.  So long as “the end is legitimate,” the courts should not “deny the legislature the right to select appropriate means.”85   Legislatures should not be “barred from determining the appropriateness of various means” and courts should not “stigmatize legislative judgment” by putting certain choices “beyond the pale of legislative power.”86   Rather than “exercise censorship over the conviction of legislatures,” courts should instead defer to legislative choices.87 

The Court also emphasized that it could not second-guess the governing majority’s choice of means.  It believed that it could not “deny the legislature the right to select appropriate means for [the] attainment” of national unity.88   Indeed, the Court thought it would improperly “stigmatize legislative judgment” and “amount to no less than the pronouncement of pedagogical and psychological dogma” for a court to do so.89   Nor could the Court allow for an exemption only for those students with a conscientious objection as it “might cast doubts in the minds of the other children” and thereby weaken the show of unity the majority sought to create.90   Everyone must conform; no exceptions.

Justice Frankfurter expounded this point further when defending Gobitis in his Barnette dissent.  There, he took the position that, even for rights expressly described in the Bill of Rights, “[i]n no instance is this Court the primary protector of the particular liberty that is invoked.”91   Thus, “even though legislation relates to civil liberties,” the Court has a “duty of deference” to political actors who make the laws.92   The only question the Gobitis approach deems appropriate for constitutional provisions is essentially modern rational basis review: “[W]hether legislators could in reason have enacted such a law” to pursue “a legitimate . . . end.”93 

Justice Frankfurter emphasized his belief that what we would today call “rational basis review” was the only constitutional test available to the Court.  Frankfurter thought that judges were authorized to check only for “the absence of a rational justification for the legislation.”94   But he professed “know[ing] of no other test which this Court is authorized to apply in nullifying legislation.”95   Like Thayer before him, Frankfurter thought if a law was rational, the Court must stand aside and defer to the legislative majority that enacted it.96 

As discussed below, Justice Frankfurter believed this limitation on the judicial role would have salutary effects because of what he saw as inherent dangers in judicial enforcement of constitutional rights.

c.   Fear of Anarchy

Gobitis also explained that, notwithstanding the First Amendment’s protection for free speech and religious exercise, “[c]onscientious scruples” could not be permitted to “relieve[] the individual from obedience.”97   The Court believed that to protect the Jehovah’s Witnesses under the First Amendment would mean that “the freedom to follow conscience has itself no limits” and would undermine, rather than further, the pluralism that “underlies [the] protection of religious toleration.”98 

Justice Frankfurter continued this argument in his Barnette dissent, where he explained that judicial protection of the Jehovah’s Witnesses would elevate individual conscience above the law.99   Doing so would court anarchy, as the religious objector “might refuse to contribute [to] taxes.”100   Justice Frankfurter offered the example of forced Bible reading in schools, explaining how judicial enforcement of the First Amendment might lead to challenges by “parents of the Catholic and Jewish faiths and of some Protestant persuasions” against the required use of the King James Bible.101   Such an approach to the First Amendment would deny to the majority the ability to enact requirements that “seem essential for the welfare of the state” because they “may offend the consciences of a minority.”102   To Justice Frankfurter, that would wrongly suggest “that the consciences of a minority are more sacred and more enshrined in the Constitution than the consciences of a majority.”103 

*     *     *

On these grounds, the Gobitis Court held that it was permissible for governments to punish members of a religious minority for their refusal to engage in speech and conduct demanded by the majority.  As the Court saw it, neither the First Amendment nor any other law gave judges the authority to intervene or to second-guess the legislature’s balancing of interests.  As Thayer had prescribed a half-century earlier, Gobitis said courts must defer.


Gobitis embraced a Thayerian judicial approach in which courts defer to rational government action, even in individual rights cases.  Had Gobitis survived and been broadly adopted, it would have left the Constitution and the courts largely out of the project of protecting rights.  Those with minority views or practices would instead be consigned to hoping that the legislature—controlled by the majority—would choose to grant protections for the minority.

Barnette, however, firmly rejected both the conclusion of Gobitis (that forced flag salutes were permissible) and, more importantly for our purposes, its reasoning about judicial deference.  This Part will discuss Barnette’s rejection of Thayerian deference in the First Amendment context and how it will come to set the standard for modern constitutional rights jurisprudence, in which the Constitution provides meaningful—and judicially enforceable—protections for constitutional rights, even in the face of rational restrictions imposed by the majority.  Examining Barnette’s rejection of Thayerian deference and its broad impact on the law will then set the stage for understanding both the oddity of the Court’s reembrace of Gobitis in Smith (Part III below) and the Court’s more recent return to Barnette principles across a variety of religious liberty contexts (Part IV).

A.   Barnette and Judicial Protection of Rights

The lone dissenter in Gobitis, Harlan Fiske Stone, thought the Court’s Thayerian deference to the legislature was its “surrender of the constitutional protection of the liberty of small minorities to the popular will.”104   Instead of having judicially enforceable constitutional rights to believe and act according to different ideas than the majority, Gobitis sanctioned what was essentially might-makes-right majority control.  This ability of the majority to control the speech and actions of the minority was to be largely unchecked by the courts.  So long as the majority’s goal was “legitimate” and the law was “general,” Gobitis said courts would not interfere.105   This was Thayer’s theory brought to bear on the First Amendment, and it left a targeted minority without enforceable constitutional protections.

The Court’s treatment of the Jehovah’s Witnesses in Gobitis sparked a wave of increased anti-Witness violence.106   As Noah Feldman has observed, the decision was understood by many Americans as announcing “open season on the Witnesses.”107   Mobs attacked Witnesses across the country, including beatings, draggings through the street, and forced marches out of town.108   Witness meeting houses were looted and burned.  Some Witnesses were force-fed castor oil and publicly soiled themselves; a Nebraska man was castrated.109 

All told, the Department of Justice received more than three hundred complaints of mob violence against Witnesses, spanning forty-four states, in 1940 alone.110   Law enforcement sometimes looked the other way, believing they had the Supreme Court’s blessing: “They’re traitors—the Supreme Court says so.”111   The ACLU called the violence against a religious minority “unparalleled in America since the attacks on the Mormons.”112   In some states, governments moved to take away the children of Witness families, to make them wards of the state.113 

By 1942, however, the tide had already begun to turn.  The narrow view of the judicial role in enforcing the Bill of Rights was rejected by three of the Justices from the Gobitis decision.  While dissenting in Jones v. Opelika,114  Justices Black, Douglas, and Murphy disavowed their Gobitis votes and explained how such deference to political actors was incompatible with a free and pluralistic approach to the First Amendment.115   Where Gobitis had disclaimed any significant role for courts enforcing the Bill of Rights in opposition to a general law, the dissenters now rejected that approach, finding instead that “the historic Bill of Rights, has a high responsibility to accommodate itself to the religious views of minorities, however unpopular and unorthodox those views may be.”116   Where Gobitis had left the right to free exercise dependent on the grace of the majority in the legislature, the dissenters instead found that the “First Amendment does not put the right freely to exercise religion in a subordinate position.”117   Moreover, they emphasized the proper role of courts in cases touching on the Bill of Rights: “[I]t is the duty of this Court” to ensure that the legislative majority has not “impair[ed] . . . cherished freedoms in reaching its objective.”118   This is the opposite of the deferential Thayer/Gobitis approach.

The whole Court eventually revisited the question of forced flag salutes in another Jehovah’s Witness case in 1943, West Virginia State Board of Education v. Barnette.119   The case concerned two young Jehovah’s Witness girls, ten-year-old Gathie and eight-year-old Marie Barnette.120   The Barnette family had made clear it was willing to compromise.  While they could only pledge allegiance to God, the children were willing to say “I respect the flag of the United States and acknowledge it as a symbol of freedom and justice to all.”121   If that were not enough, they were also willing to pledge “allegiance and obedience to all the laws of the United States that are consistent with God’s law, as set forth in the Bible.”122   But the state would not budge—either the two girls would pledge allegiance exactly as the school board told them to, or they would be expelled and their parents punished.123 

The Barnettes sued, asserting that the forced pledge “amounts to a denial of religious liberty.”124   In light of the disavowal in Jones, the three-judge district court panel said it did “not feel that it is incumbent upon us to accept [Gobitis] as binding authority.”125   In particular, where the Supreme Court had already “impaired” Gobitis as an authority, the panel did not think it “should deny protection to rights which we regard as among the most sacred” in the Constitution.126 

The panel then specifically rejected the narrow Thayer/Gobitis approach to judicial review in cases concerning constitutional liberties.  It explained that constitutional rights “would not be worth the paper” they are written on if courts were to defer whenever legislatures saw fit to regulate.127   The “bill of rights is not a mere guide for the exercise of legislative discretion,” but instead “is a part of the fundamental law of the land, and is to be enforced as such by the courts.”128   The panel was particularly concerned about the “tyranny of majorities over the rights of individuals or helpless minorities” if courts were to “abdicate the most important duty which rests on them under the Constitution.”129   And rather than defer to other branches, the panel thought that the “delicate and difficult task” of “apprais[ing] the substantiality of the reasons advanced in support of the regulation of the free enjoyment of the rights” falls to the courts.130 

The Supreme Court agreed with this rejection of deference and embraced a much stronger role for the Bill of Rights—and the Court—in protecting minority rights.  The Court made clear that the limitations on government power apply to all parts of the government, “[b]oards of [e]ducation not excepted.”131   And unlike Thayer, Gobitis, Plessy, Buck,and Korematsu,the Barnette Court did not think it should simply defer.  Rather, while acknowledging that such boards of course have “important” and “highly discretionary functions,” the Court emphasized that those functions must be performed “within the limits of the Bill of Rights.”132 

Barnette expressly rejected theThayer/Gobitis approach of mere rationality review.133   That test, Barnette explained, was appropriate for ordinary regulations that do not touch on constitutional liberties.  For example, regulation of “a public utility may well include, so far as the due process test is concerned, power to impose all of the restrictions which a legislature may have a ‘rational basis’ for adopting.”134   But Barnette adamantly rejected such a standard for First Amendment rights, holding that “freedoms of speech and of press, of assembly, and of worship may not be infringed on such slender grounds.”135   Such constitutional liberties could only be restricted to prevent “grave and immediate” dangers.136 

Nor was the Barnette Court convinced that the importance of nationalism or the sensitivity of the task of educating children somehow exempted the government from following the Constitution or the Court from enforcing it.  To the contrary, the Court explained these are “reason[s] for scrupulous protection of Constitutional freedoms,” rather than against.137   The Court explained that the Constitution protects the “right to differ” not only as to “things that do not matter much” but also “as to things that touch the heart of the existing order.”138 

The Justices also made clear that they understood the dangers of failing to provide judicial enforcement of the First Amendment to protect minorities against coercion by the majority.  “Those who begin coercive elimination of dissent soon find themselves exterminating dissenters.  Compulsory unification of opinion achieves only the unanimity of the graveyard.”139   And the Court recognized that the First Amendment’s protection for pluralism is a key to averting such problems: “It seems trite but necessary to say that the First Amendment to our Constitution was designed to avoid these ends by avoiding these beginnings.”140 

Barnette thus reflected a vastly different understanding of the Court’s role in enforcing the Bill of Rights.  Where Gobitis professed powerlessness in the face of even barely rational majority will, Barnette explained that the “very purpose of a Bill of Rights” was to place certain matters “beyond the reach of majorities and officials and to establish them as legal principles to be applied by the courts.”141   Fundamental rights “may not be submitted to vote” and “depend on the outcome of no elections.”142   And legislative majorities should not simply be given deference so long as their laws are at least rational or legitimate.  Rather, it is “the function of this Court” to “apply the Bill of Rights . . . where the invasion of rights occurs.”143   It is difficult to imagine a more direct rejection of Gobitis’s application of Thayer’s rule to the Bill of Rights. 

In the decision’s most famous passage, the Court strongly rejects the notion of majority-imposed orthodoxy: “If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion or force citizens to confess by word or act their faith therein.”144   Where Gobitis had featured judicial deference to the legislature, Barnette now embraced a judicial duty to enforce the Constitution and protect minority rights.

B.   Barnette’s Victory over the Deference of Thayer and Gobitis

There is no real dispute about the relationship between Barnette and Gobitis on the flag-salute question: Barnette overruled Gobitis.  The Barnette Court itself was explicit on this point.145   And, in most cases, both Supreme Court Justices and lower courts have had little difficulty agreeing that Gobitis has been overruled and Barnette is the law.146 

Barnette is thus the unquestioned constitutional standard on the question of forced flag salutes.  But the rejection of deference embodied in Barnette has extended far beyond the flag salute context.  Barnette—and, in particular, Barnette’s nondeferential approach to constitutional rights—has a revered spot in the constitutional canon.  It is influential not just in First Amendment cases, but also in other significant rights cases.  Today, no one argues that courts should apply only Thayerian rationality review when protecting fundamental rights.147 

1.   Barnette’s Triumph in First Amendment Law

Given Barnette’s express statement that it “overruled” Gobitis, it is not surprising that the Barnette understanding of the First Amendment also controlled in the aftermath of the two cases.  In the nearly fifty years between Barnette and Employment Division v. Smith, Gobitis was never cited as a correct statement of how the First Amendment should operate.

Barnette, on the other hand, was treated as the controlling decision.  In fact, in the years after the two decisions, the Court’s overruling of Gobitis was frequently cited as an example of the Court’s willingness to reverse prior constitutional decisions when it realized they were incorrect.  For example, just the Term after Barnette, the Court cited Barnette in Smith v. Allwright, overruling prior precedent to eliminate race-based qualifications in primary elections.148   Although Barnette’s First Amendment holding was not at issue, the Court relied on it for the proposition that “when convinced of former error,” the Court had “freely exercised its power to reexamine the basis of its constitutional decisions” in what has “long been accepted practice.”149   Such use of Barnette remains common to the present day.150 

This was certainly understood to be true as to the merits of the First Amendment claims.  Thus, shortly after Barnette, the Court cited it in United States v. Ballard for the proposition that “freedom of religious belief” is “basic in a society of free men” and “embraces the right to maintain theories of life and of death and of the hereafter which are rank heresy to followers of the orthodox faiths.”151   The Court relied on this explanation of religious liberty when allowing an immigrant conscientious objector to military service to nonetheless become an American citizen, citing both Ballard and Barnette for the Court’s understanding that the “struggle for religious liberty” had led to a “victory for freedom of thought recorded in our Bill of Rights” and “recognizes that in the domain of conscience there is a moral power higher than the State.”152 

To be sure, the Court’s post-Barnette decisions did not suggest that religious parties must always win.  For example, just a year after Barnette the Court decided Prince v. Massachusetts, in which it held that the state could make it illegal for children to engage in street preaching.153   Prince acknowledged Barnette’s protection for the free exercise rights of children against the “preponderant” power of the state.154   But it emphasized that this right was not without limits.  Rather, the Court could recognize the government’s strong interests in avoiding “the crippling effects of child employment” and protecting them from harm on the streets were sufficient to allow Massachusetts to outlaw child street preaching, even though such activities could not be forbidden for adults.155 

The triumph of Barnette, then, is not so much about religious claimants always winning, but an approach to constitutional rights that requires political actors to demonstrate to courts that they have very strong reasons before restricting rights.  This is the opposite of the Thayer/Gobitis approach in which courts would only ask “whether legislators could in reason have enacted such a law” and would otherwise defer.156 

In some free exercise cases, this more protective Barnette approach eventually took the shape of what we today would call “strict scrutiny.”157   In Sherbert v. Verner, the Court held that a burden on religious exercise is only permissible if it is justified by a “compelling state interest.”158   The Court emphasized that Thayerian, Gobitis-style rational basis review had no place under the First Amendment: “It is basic that no showing merely of a rational relationship to some colorable state interest would suffice; in this highly sensitive constitutional area, ‘[o]nly the gravest abuses, endangering paramount interests, give occasion for permissible limitation.’”159 

The same day it decided Sherbert, the Court also invoked Barnette in an Establishment Clause case rejecting a program of Bible reading in public schools, noting that the “majority” could not “use the machinery of the State”160  to coerce because “fundamental rights may not be submitted to [a] vote; they depend on the outcome of no elections.”161   This brought to fruition one of Justice Frankfurter’s fears expressed in his Barnette dissent, namely that the abandonment of Gobitis-style thinking about the Bill of Rights would undermine Bible reading in public schools.162 

2.   Barnette’s Broader Triumph in Constitutional Law

Barnette’s clear controlling status over Gobitis in Religion Clause cases is mirrored by the treatment of Barnette in other areas of the law as well.  Even outside of the First Amendment context, Barnette is often invoked as an important example of how judicial protection of constitutional rights is supposed to work.

For example, in the joint opinion in Planned Parenthood of Southeastern Pennsylvania v. Casey, the Court explains that it “is conventional constitutional doctrine that where reasonable people disagree the government can adopt one position or the other.”163   But the opinion then invokes Barnette as the example to demonstrate that this rule only applies in “a state of affairs in which the choice does not intrude upon a protected liberty.”164   In other words, because the Court decided that abortion was a constitutional right, it could not be regulated on Gobitis terms (i.e., the legislative majority can choose, so long as its action satisfies a bare reasonableness standard), but instead on Barnette terms that largely disable the majority from invading a protected right.  The Thayerian approach of only invalidating laws if they fail rational basis does not apply to constitutional rights.

Likewise in Obergefell v. Hodges,the Court relied heavily on Barnette’s understanding of the relationships between majority power and fundamental rights in finding a right to same-sex marriage.165   The Court explained that a plaintiff “can invoke a right to constitutional protection when he or she is harmed, even if the broader public disagrees and even if the legislature refuses to act.”166   As support for this view of rights, the Court relied on Barnette’sexplanation that the idea of the Bill of Rights was to place certain subjects beyond the reach of popular majorities, so that they could be enforced by courts.  The Court quoted Barnette, stating: “[F]undamental rights may not be submitted to vote; they depend on the outcome of no elections.”167 

Of course, the constitutional holdings in Casey and Obergefell both prompted vigorous dissenting opinions.  But the dissenters were principally arguing about whether abortion and same-sex marriage qualify as fundamental rights protected by the Constitution.168   No Justice, in either case, challenged the assertions that a Barnette-style approach is the proper course for protecting constitutional rights where there is a constitutionally protected liberty interest.  No Justice argued for the deferential Thayer/Gobitis rule that laws can only be invalidated under rational basis scrutiny even if there is a substantive constitutional right at stake. 

Casey and Obergefell thus demonstrate the largely undisputed triumph of the Barnette understanding of judicial protection of constitutional rights.  Barnette—and not the Thayerian deference of Gobitis, Plessy, Buck, and Korematsu—represents our standard constitutional approach to protecting rights.  The point does not even prompt debate, even in the most contentious cases.169 

3.   Barnette’s Triumph in the Academy

Constitutional theorists have likewise had little difficulty concluding that the Barnette approach to judicial enforcement of rights—rather than the deferential Thayerian approach of Gobitis—controls.  This is hardly surprising, given that Barnette is “among the most renowned cases in American history,” and is considered “part of the essential fabric of American constitutional law.”170   While Gobitis is “well known and widely excoriated amongst civil libertarians,”171  Barnette is “celebrated” as a “hallmark[] of American liberty by both the left and the right.”172   While Gobitis is denigrated as “arguably the worst Supreme Court majority opinion in a First Amendment case,” Barnette is “almost universally regarded as one of the very best First Amendment opinions ever produced.”173 

This iconic status is attributable not only to its substantive outcome, but more broadly to its approach to minority rights in a pluralistic democracy.  That is why Cass Sunstein, for example, recently wrote that “[i]f we had to preserve just one Supreme Court opinion to show some other civilization what American constitutional law is all about,” he’d select Barnette because of how “foundational” it is in “help[ing] orient large areas of the law.”174 

Barnette can be Sunstein’s exemplar to show “what American constitutional law is all about” because Barnette’s approach to the judicial enforcement of constitutional rights is so widely accepted as correct.  As Susan Estrich and Kathleen Sullivan have explained, Barnette’s “‘withdrawal’ of fundamental liberties from the political arena is basic to constitutional democracy as opposed to rank majoritarianism, and nowhere is such ‘withdrawal’ more important than in controversies where moral convictions and passions run deepest.”175 

To be sure, scholars have long wrestled with, and continue to debate, questions related to the legitimacy of judicial review.176   Much of this scholarship has focused on what Alexander Bickel termed the “Counter-Majoritarian Difficulty,” namely, the problem of how judicial review could be legitimate in a majoritarian system.177   But that “difficulty” has seemed most acute in the context of judicial enforcement of unwritten constitutional rights.

The countermajoritarian difficulty has turned out to be much less difficult in the context at issue in Gobitis and Barnette, namely, where positive law rights have been written into the Constitution.  While judicial review may seem countermajoritarian when the Court is declaring or discovering new rights, it is markedly less so when the Court is acting as it did in Barnette: simply enforcing the rights that the people already chose to protect in the document.  As Kurt Lash has explained, in such cases the popular will expressed by protecting a right in the Constitution itself “resolves the difficulty by grounding judicial review in the more deeply democratic law of the people,” namely, the Constitution.178   This fact explains why John Hart Ely could observe in Democracy and Distrust that enforcement of positive law rights included in the Constitution “seems to enjoy virtually universal contemporary acceptance.”179 

Given this broad acceptance of judicially enforceable constitutional rights, it is not surprising that Richard Posner recently observed that “the ‘rational basis’ criterion of constitutionality, a legacy of Thayer”—that is, the legal theory at the heart of Gobitis—“has dropped away.”180   Everyone knows, it seems, that courts are supposed to protect constitutional rights, and not merely defer to rational majoritarian decisions of the legislature.


Given Barnette’s place in the constitutional pantheon—not just for its broadly accepted outcome but as an exemplar of “what American constitutional law is all about”—one would not expect courts to rely on Gobitis and its deferential Thayerian approach to the Bill of Rights.  Why would anyone wish to return to what Justice Stone had called, in his now-vindicated Gobitis dissent, the “surrender of the constitutional protection of the liberty of small minorities to the popular will”?181 

But there is one area of constitutional law in which the judicial deference of Gobitis has retained significant lasting influence: religious liberty.  This is because, nearly fifty years after Gobitis was expressly overruled by Barnette, Gobitis received a very consequential revival in Justice Scalia’s majority opinion in Employment Division v. Smith.

The Smith decision is of course most widely understood as a rejection or limitation on the idea of religious exemptions.  But an important and underappreciated aspect of Smith is its approach to the question of deference, and particularly its reembrace of the deferential Gobitis view of the First Amendment.

A.   Smith, Thayer, and Gobitis

Smith concerned a Native American man who was denied unemployment benefits because he was fired for ingesting peyote during a religious ceremony.182   Although there was no criminal prosecution involved, the Supreme Court focused its decision on whether Oregon’s criminal law was permitted “to include religiously inspired peyote use within the reach of its general criminal prohibition on use of that drug.”183   Neither party had asked the Supreme Court to change the legal standards governing free exercise claims.184   Nor had the Court granted certiorari to consider that issue.  Accordingly, the question of the proper standard for free exercise claims was neither briefed nor addressed at oral argument.185 

Just a few years earlier, the Court had explained why mere rational basis was an inappropriate standard to apply to First Amendment religious liberty claims.  First, the Court explained that deferring to the legislature on a standard of bare reasonableness “has no basis in precedent.”186   Second, the Court stated that such deference “relegates a serious First Amendment value to the barest level of minimal scrutiny that the Equal Protection Clause already provides.”187 

Nevertheless, without prompting by the parties, Smith contains a lengthy discussion of what legal standard should apply where a party’s exercise of religion conflicts with a general law enacted by the majority.  Smith established a new standard for many such cases: So long as the government is applying a “neutral” and “generally applicable” law, the First Amendment provides no heightened protection.188   Instead, the Court would only apply deferential, rational basis review.  The Court embraced this standard in large part to avoid the prospect of judges deciding whether society’s interests justified imposing the burdens.189   The Court argued that, precisely because of the nation’s religious diversity and pluralism, holding the government to the Sherbert compelling interest standard “would be courting anarchy.”190 

Smith reflects the Thayerian judicial deference approach in several important ways.  First, Smith, like Gobitis,established a rule that will frequently allow laws enacted by the majority to restrict the minority’s exercise of religion.  So long as the law is “neutral [and] generally applicable”191  (Smith’s framing) or “not directed against doctrinal loyalties of particular sects”192  (Gobitis’s framing), the cases suggested no special scrutiny would apply, and all rational laws would be upheld.  In this regard, Smith joined Gobitis in adopting the Thayer/Hand approach that, even for rights protected by the Bill of Rights, courts should defer to rational legislative judgments.

Second, both cases therefore leave only a very limited role for judges in the protection of constitutional rights.  Both cases insist that it is usually the job of the legislature, rather than the courts, to protect the minority.  Thus, like Gobitis, Smith leaves minorities at the mercy of democratic majorities for their rights.  Smith found that the “relative disadvantage” minorities experience in the political process is an “unavoidable consequence of democratic government” that “must be preferred” to judicial balancing.193   In Thayer’s terms, deference to the legislature is required so that a court does not become “a board for answering legislative conundrums.”194 

Finally, both cases emphasized that an alternative rule—one in which judges apply constitutional scrutiny to determine whether the government can force someone to violate his or her religion—would undermine pluralism and suggest that “conscience has itself no limits”195  (Gobitis) or that protecting conscience would “court[] anarchy”196  (Smith).

These parallels between Smith and Gobitis are no accident.  To the contrary, they are the result of Smith’s direct reliance on Gobitis.  In particular, Smith relied on Gobitis—rather than Barnette—as authority for how religious liberty claims should be decided.  Quoting Gobitis, the Smith Court explained that “mere possession of religious convictions which contradict the relevant concerns of a political society does not relieve the citizen from the discharge of political responsibilities.”197   The Court then further cited Gobitis as an example of the principle that courts will not provide religious exemptions from neutral and generally applicable laws.198   Curiously, Smith did not even note that Gobitis had been overruled.199 

Despite Justice Scalia’s general embrace of originalism, the Smith decision was not driven by any argument about the original public meaning of the text of the Free Exercise Clause.  When arguing that the clause should not lead to religious exemptions, the most the Court says about the text or originalism is that “we do not think the words must be given that meaning.”200   The Court then suggested that its own reading of the text was at least “permissible.”201   The Court’s textual ambivalence was buttressed, however, with its more firmly stated view about what its own “decisions reveal” about the best interpretation.202   Those decisions—most prominently Gobitis—led the Court to conclude that application of strict judicial scrutiny would supplant legislative decisionmaking and create anarchy.203   Thus while Smith was somewhat ambivalent about text, it was quite adamant about judicial role and the likely practical consequences of taking a more protective view of religious liberty.204 

Concurring only in the judgment, Justice O’Connor explained why she could not endorse the Court’s Thayerian, Gobitis-inspired reasoning.  Writing for herself and the three dissenters (Justices Brennan, Marshall, and Blackmun), Justice O’Connor noted that “the First Amendment was enacted precisely to protect the rights of those whose religious practices are not shared by the majority and may be viewed with hostility.”205   Citing Barnette as overruling Gobitis, Justice O’Connor explained that “[t]he history of our free exercise doctrine amply demonstrates the harsh impact majoritarian rule has had on unpopular or emerging religious groups such as the Jehovah’s Witnesses and the Amish.”206   Rejecting the rational basis approach embraced by Gobitis and now Smith, Justice O’Connor argued that “[t]he compelling interest test reflects the First Amendment’s mandate of preserving religious liberty to the fullest extent possible in a pluralistic society” and that the Smith/Gobitis approach of mere rational basis review “denigrate[d] ‘[t]he very purpose of a Bill of Rights.’”207 

The three dissenters likewise rejected the notion that the Religion Clauses allowed the courts to accept “the repression of minority religions” as an “unavoidable consequence of democratic government.”208   To the contrary, echoing Barnette, they asserted that judicial protection for minority religious beliefs and practices was “an essential element of liberty” and that the Founders had “drafted the Religion Clauses precisely in order to avoid” such intolerance.209 

B.   Smith’s Aftermath

Smith provoked a strong reaction.210   Scholars, activists, and lawmakers from across the political spectrum and from a broad range of religious groups came together to enact heightened protections for religious liberty in the Religious Freedom Restoration Act (RFRA).211   That story has already been well chronicled elsewhere.212 

For present purposes, it is enough to note that, by enacting the RFRA, Congress sought to reimpose the compelling interest test that Smith had jettisoned.213   While the Supreme Court had downgraded many free exercise claims to mere rational basis review, Congress thought that strict scrutiny “is a workable test” which helps strike “sensible balances between religious liberty and competing prior governmental interests.”214   But the Supreme Court soon found that Congress could not reimpose those tests as a constitutional matter, and thus held RFRA inapplicable to state and local governments.215   Otherwise, the Court held that RFRA could impose the compelling interest test as a statutory matter against the federal government,216  and that a successor statute to RFRA, the Religious Land Use and Institutionalized Persons Act (RLUIPA),217  could do the same for prison and land-use cases at the state and local levels.218   In addition, many states interpreted their own state constitutions to follow the Sherbert compelling interest approach rather than the Smith rational basis approach.219   Others adopted state-level RFRA statutes.220 

These post-Smith changes had two principal effects that are relevant to our discussion.  First, because so many religious liberty claims were controlled by other sources of law, the Supreme Court would go more than twenty years after Smith without having to even consider applying Smith’s rational basis rule to a neutral and generally applicable statute.  Federal cases were decided under RFRA221  and RLUIPA,222  or involved statutes that fell outside of Smith’s rule because they obviously lacked neutrality and general applicability.223   Litigators brought far fewer free exercise claims,224  and many such claims could be resolved under state statutes and constitutions that applied the pre-Smith rule.225 

Second, the vast expansion of subconstitutional protections (both the federal statutes and the state constitutions and state statutes) provided an opportunity for courts across the country, and occasionally the Supreme Court itself, to engage in precisely the judicial balancing that Smith and Gobitis feared would lead to anarchy, and that Thayer feared was a separation-of-powers problem.  As will be discussed in more detail in the next Part, this has allowed courts to demonstrate both their ability to apply the compelling interest test and that the results of the test do not appear to produce the feared problems.226 


Since 2012, the Supreme Court has decided a series of religious liberty cases, touching virtually every aspect of religious liberty law.  These cases have concerned religious exercises by members of a wide variety of faith groups (Muslims,227  Lutherans,228  Jews,229  Catholics,230  and Wiccans,231  among others).  They have involved a wide variety of types of religious liberty claims (church autonomy,232  Free Exercise Clause,233  Establishment Clause,234  and federal statutes)235  and a wide variety of factual contexts (schools,236  nursing homes,237  prisons,238  Abercrombie & Fitch stores,239  town meetings,240  war memorials,241  and more).  In virtually all of them, the Court has ruled in favor of the religious party, practice, or monument.

Scholars and critics have offered a variety of theories about these cases.  Some have argued that the Court is engaged in the process of interpreting the religion clauses to favor conservative and Christian causes.  For example, Dean Chemerinsky and Professor Gillman argue that recent cases show the clauses “being interpreted to allow powerful religious groups to harm innocent third parties and to establish a privileged status within the political system, to the detriment of true religious liberty and diversity.”242   Others argue that the Court is engaged in an “unusual dialogue” with the “religious right” to forge a “new church-state landscape,”243  and to do so in a way that is “bad for civil rights, especially for rights of women, LGBTQ individuals and people of color.”244 

These theories fail to account for the sheer variety of different parties the Court has ruled for (many are women, many non-Christian, and many are racial minorities).  And they fail to account for the often supermajority support for the winning parties.  That supermajority support has often included Justices Ginsburg, Kagan, Sotomayor, and Breyer—unlikely coconspirators in forging a new conservative Christian monopoly on religious liberty to hurt minority rights.

The better explanation for the Court’s recent religious liberty decisions is that a broad cross-section of the Court is fully embracing a Barnette-style approach to religious liberty, in which the First Amendment provides strong, judicially enforceable protections for religious minorities and religious pluralism.  These decisions firmly reject every aspect of the deferential Thayerian approach to constitutional rights used in Gobitis—its view of judicial role, its deference to merely rational legislative choices, and its fear that judicially enforceable rights are irreconcilable with ordered liberty.  The Court has thus emphatically demonstrated that it views the religion clauses as important and enforceable protectors of religious liberty and peaceful pluralism.245 

This Part will discuss how a rejection of deference explains the Court’s decisions across four different groups of religious liberty cases: church autonomy, free exercise, Establishment Clause, and statutory rights.

A.   Church Autonomy Cases

The Court’s modern path toward rejecting judicial deference began in earnest in 2012 with Hosanna-Tabor Evangelical Lutheran Church & School v.EEOC.246   Hosanna-Tabor concerned a fourth-grade teacher who had been fired by a Lutheran elementary school.247   The teacher claimed she had been fired in violation of a general law duly enacted by the legislative majority: the Americans with Disabilities Act (ADA).248   The school claimed a First Amendment right to terminate the teacher for religious reasons, regardless of the ADA.249   The case thus required the Court to determine whether the First Amendment’s protections for religious autonomy could trump the requirements imposed by a general and majority-supported law.

Had Hosanna-Tabor been decided under the deferential Thayer/Hand/Gobitis approach, it would have been an easy case.  The Americans with Disabilities Act is a broad and general law, duly enacted by a legislative majority.250   The ADA is not “directed against [the] doctrinal loyalties”251  of any particular group, but instead reflects an obviously rational majoritarian conclusion that people should not be fired for having disabilities.252   Under a Gobitis or Smith approach, the Court would have deferred to the legislature—even if that harmed a group with minority beliefs or practices—so long as the Court decided that the law was rationally pursuing a legitimate goal.253   And if the legislature had not seen fit to create an exemption for religious schools hiring elementary school teachers, that would have been the end of the analysis.  Indeed, the Court may well have just repeated its hands-off lines from Gobitis that judicial interference with such a legislative decision would “amount to no less than the pronouncement of pedagogical and psychological dogma in a field where courts possess no marked and certainly no controlling competence.”254 

In briefing to the Supreme Court, no party expressly embraced or endorsed Gobitis itself.  However, the United States government asked the Court to apply Smith’s Gobitis-inspired rule and find no special constitutional protection for religious minorities in the face of a general statute enacted by the majority.  In particular, the government argued that the ADA is “generally applicable” in that it “applies to all employers with more than 15 employees,” that it does not “single out” religious groups for disfavor, and that the school therefore “cannot claim that the Free Exercise Clause provides it an exemption from the ADA’s generally applicable prohibition on retaliation.”255 

Not a single Justice accepted this argument.  Instead, the Court unanimously held that the religion clauses of the First Amendment control, even in the face of a general legislative enactment like the ADA.  The Court found the government’s position “untenable,” “remarkable,” and “hard to square with the text of the First Amendment itself, which gives special solicitude to the rights of religious organizations.”256   The Court refused to apply Smith’s rational basis ruleto what it deemed “an internal church decision that affects the faith and mission of the church itself.”257 

This was not because the Court found the interests advanced by the ADA to be weaker than the majority’s interest in promoting nationalism in Gobitis,or its interest in enforcing drug laws in Smith.  Rather, the Court acknowledged that the interest in enforcing employment discrimination laws was not merely legitimate or rational, but “undoubtedly important.”258   But the Court refused to defer to the majoritarian legislative process about how to balance that interest against the interest of religious groups in choosing who they will work with to teach their faith and carry out their mission.  To the contrary, invoking both the Free Exercise Clause and the Establishment Clause, all nine Justices agreed that “the First Amendment has struck the balance for us” and requires an exemption for the school.259   This is the opposite of the deference approach taken in Gobitis and Smith.  It is the opposite of the approach recommended by Thayer and is observable in cases like Plessy, Buck, and Korematsu.

Eight years later, the Court built on its Hosanna-Tabor decision in Our Lady of Guadalupe School v. Morrissey-Berru.260   Our Lady again concerned the ministerial exception rooted in both religion clauses.261   A 7 –2 majority ruled that nondiscrimination laws like the Americans with Disabilities Act and the Age Discrimination in Employment Act cannot constrain the freedom of religious groups to choose the teachers who will pass on the faith to children.262   This is true even if the school does not rely on an overtly religious reason for its employment decision, and if the teacher lacks a religious-sounding title or special religious training.263   Rather “[w]hat matters, at bottom, is what an employee does.”264   If the employee has religiously important duties like teaching the faith, then the government cannot interfere in the employment decision, even if the employee also has many other secular duties.265   The Constitution protects religious autonomy, even if the group’s values do not conform to those of the governing majority.  Again, the Court refused to simply defer to even broadly supported general laws.

Dissenting, Justices Sotomayor and Ginsburg did not seem to disagree about many of the key religious liberty points.  They acknowledged the ministerial exception as “extraordinarily potent” and reaffirmed their support of the Court’s 9–0 endorsement of the ministerial exception in Hosanna-Tabor.266   Their disagreement concerned not so much the law but what they called “disputed facts” in a “context-specific” analysis.267   Thus, the opinions in Our Lady show all nine Justices recognizing the importance of allowing religious groups to make employment decisions in accordance with their beliefs, even when those decisions implicate weighty societal interests like nondiscrimination.

The broad agreement among the Justices that the Constitution creates judicially enforceable church autonomy rights is important and reminiscent of Barnette.  All nine Justices appear to accept the premise that the First Amendment protects religious liberty, and that courts cannot simply defer even to important and obviously rational statutes like Title VII.  Rather than just defer to the legislature’s rational decisionmaking (the approach of Gobitis, Smith, Thayer, and Hand), the Justices instead treated the religion clauses as strong, judicially enforceable protections for religious groups.

B.   Free Exercise Cases

The church autonomy cases above were decided based on the full breadth of the religion clauses, taking into account both free exercise and establishment principles.  Similar developments have occurred, however, even in cases decided solely on free exercise grounds.  Three groups of free exercise cases stand out: marriage, Blaine Amendments,and COVID-19 lockdowns.

1.   Marriage

The Court’s discussion of free exercise rights in the context of same-sex marriage reflects a Barnette approach to the enforcement and role of the First Amendment.  This has occurred in a range of cases, some of which are directly about religious liberty (Masterpiece Cakeshop, Ltd. v. Colorado Civil Rights Commission268  and Fulton v. City of Philadelphia)269  and some of which only address religious questions in passing (Obergefell v. Hodges270  and Bostock v. Clayton County).271 

First, the Court in Masterpiece Cakeshop addressed whether the Free Exercise Clause provided protection for a religious baker who objected to being forced to create a cake for a same-sex wedding.272   A seven-Justice majority ruled in favor of the baker, finding that the Colorado Civil Rights Commission had demonstrated “religious hostility” and had therefore failed to provide “the religious neutrality that the Constitution requires.”273 

To be sure, this resolution is, in one sense, technically reconcilable with the Gobitis/Smith approach to rights.  One can think of Colorado’s action as “directed against doctrinal loyalties”274  (Gobitis’s framing) or not “neutral”275  (Smith’s framing) toward the baker’s religious beliefs.  If that were all the opinion said, Masterpiece Cakeshop might be a counterexample in our discussion.

But Masterpiece Cakeshop also discussed the reach of the Free Exercise Clause in ways that suggest it is far more protective than Gobitis or Smith would seem to allow.  First, the seven Justices in the majority explained that the case concerned the baker’s “right . . . to exercise fundamental freedoms under the First Amendment.”276   Both Smith and Gobitis had argued that the scope of the Free Exercise Clause did not include a right to exemption from general laws; Masterpiece Cakeshop’sclassification of the baker’s actions as exercise of a “fundamental freedom under the First Amendment” is far more generous.

Second, the majority gave a specific example that is difficult to reconcile with the deferential Gobitis/Smith approach: exemptions for clergy who refuse to perform same-sex marriages.  The Court explained: “When it comes to weddings, it can be assumed that a member of the clergy who objects to gay marriage on moral and religious grounds could not be compelled to perform the ceremony without denial of his or her right to the free exercise of religion.”277 

The Court’s assumption—joined by seven Justices—reflects a Barnette approach in that it recognizes important substantive protections from the First Amendment that cannot be overridden with a mere rational basis.  Under a Gobitis or Smith approach, one might think a sufficiently neutral and general requirement to perform weddings (say, a requirement that all persons licensed by the state to perform weddings must not discriminate) could force clergy to perform.  Yet the Court instead treats the Free Exercise Clauseas providing important, judicially enforceable protections, even against such laws.  The Thayer/Hand approach of deferring, even in Bill of Rights cases, to nonirrational laws is nowhere to be found.278 

The Masterpiece Cakeshop holding is also consistent with the Court’s discussion of religious freedom in its Obergefell decision.  As noted above, Obergefell relied on the Barnette understanding of judicially enforceable minority rights in finding a constitutional right to same-sex marriage.279   And although Obergefell was not a First Amendment case, it discussed the religious opposition to same-sex marriage in terms that suggest strong First Amendment protection—much stronger than would be allowed under the deference approach.

For example, Obergefell explained that the constitutional problem in that case arose not simply because people held “decent and honorable religious or philosophical” objections to same-sex marriage.280   The constitutional problem arose only when one particular view of marriage became “enacted law and public policy” in a way that “put the imprimatur of the State itself” on a view of marriage that “demeans or stigmatizes” those left out.281   The law should not make “[o]utlaw[s]” or even “outcast[s]” of those living out a contrary view of marriage.282 

The Obergefell Court went on to emphasize the importance of First Amendment rights in protecting people and organizations who held a traditional view of marriage.  First, the Court stressed that its decision should not be taken to “disparage[]” those with contrary religious views.283   Second, it emphasized that “[t]he First Amendment ensures that religious organizations and persons are given proper protection as they seek to teach the principles . . . [and] continue the family structure” they have so long revered.”284 

Viewed through the lens of Obergefell, Masterpiece Cakeshop can be understood as the Court’s initial delivery on Obergefell’s earlier promise that the First Amendment would provide protection for religious dissenters on marriage.  Taken together, the cases demonstrate the Court’s faith that a strong First Amendment is an important and enforceable part of helping people of vastly different beliefs live together in peace.  That is a role the First Amendment could not play under a deferential Gobitis/Smith approach to rights.  But it fits perfectly with a Barnette approach.

Secondly, although it was not a religious liberty case, the Court’s Title VII decision in Bostock v. Clayton County similarly embraced a strong Free Exercise doctrine that appears irreconcilable with a Gobitis/Smith approach of Thayerian deference.285   In Bostock, the Court found that Title VII’s ban on sex discrimination included discrimination based on sexual orientation or gender identity, including against employees in same-sex marriages.286   In the course of its opinion, the Court explained that religious employers would not need to be forced to violate their religion.287   This is because the guarantee of free exercise “lies at the heart of our pluralistic society.”288   No one could ever have called the view of the First Amendment depicted in Gobitis “the heart of our pluralistic society.”

Finally, last Term in Fulton v. City of Philadelphia, the Court decided in favor of foster parents and their religious foster agency that would not perform foster care certifications for same-sex couples.289   The Justices voiced a range of views about Smith itself—with three voting to reverse Smith as incompatible with the First Amendment (Justices Alito, Thomas, and Gorsuch),290  three openly musing about how best to replace Smith (Justices Barrett, Kavanaugh, and Breyer),291  and three finding that the Smith questions need not be reached at all (Chief Justice Roberts, and Justices Kagan and Sotomayor).292   But all nine Justices agreed that the religious parties in Fulton should win, and even the Justices who did not openly call for Smith’s reversal gave the decision an interpretation that vastly minimizes the range of cases in which the Smith/Gobitis rule of deference to rational decisions would control.293 

Fulton, in fact, sets forth three different categories of free exercise cases that trigger strict scrutiny and thereby avoid Smith’s deferential rule.  First, the Court reiterated its Masterpiece Cakeshop holding that strict scrutiny would apply where government actors “proceed[] in a manner intolerant of religious beliefs” or “restrict[] practices because of their religious nature.”294   Although the Fulton Court found it “more straightforward” to resolve the case under general applicability, Fulton’s invocation of the Masterpiece Cakeshop path to avoiding deference is significant because some lower courts had previously read the decision as only applying to discrimination by “adjudicatory bodies” rather than the rest of government.295 

Second, the Court explained that strict scrutiny applies where the government “permit[s] secular conduct that undermines the government’s asserted interests in a similar way” to proposed religious conduct.296   Thus, even if the legislature makes rational distinctions, Fulton makes clear that the Court often will not defer to them.

Third, and most importantly for Fulton itself, the Court found that strict scrutiny applies—and therefore the courts cannot simply defer—where a law invites the government to “consider the particular reasons for a person’s conduct” by providing “a mechanism for individualized exemptions.”297   The Court found that Philadelphia’s foster-care system failed this test because the City retained the ability to grant waivers but would not grant one for the religious agency, thus triggering strict scrutiny.298 

To be sure, Fulton also included considerable discussion about whether the Court should reconsider Smith.  Justice Barrett and Justice Kavanaugh openly wondered why the Free Exercise Clause “lone among the First Amendment freedoms” should offer so little protection, and Justice Breyer joined them in noting the range of issues the Court would need to confront to move beyond Smith.299   Justice Alito, joined by Justices Thomas and Gorsuch, argued that Smith’s “severe holding” was “ripe for reexamination.”300   Yet, even without overruling Smith, the Court has at least circumscribed the sphere in which Smith’srule applies, explaining several different paths to avoiding Smith’s Thayerian deference.

2.   State Funding Cases

The Court has also recently applied the Free Exercise Clause in a trio of cases concerning limitations in state funding laws that exclude religious groups.  These laws are sometimes known as “Blaine Amendments” due to similar restrictions that developed in the nineteenth century as a result of anti-Catholic bigotry and were included in many state constitutions.301   Three recent cases required the Court to consider the impact of the Free Exercise Clause on such laws.

First, in Trinity Lutheran Church of Columbia, Inc. v. Comer, a broad 7–2 majority found that Missouri had violated the Free Exercise Clause by excluding religious nonprofits from participating in a generally available government program.302   Missouri had excluded a Lutheran preschool from participating in a state-run program to provide nonprofits with funds for rubberized playground surfaces made from recycled tire scraps.303 

In defending its exclusion, Missouri echoed Gobitis in arguing that its policy should only be invalidated if it failed rational basis review.304   Missouri relied heavily on Smith, arguing that under Smith, courts should simply allow religious groups to “drive policy through the political process” rather than giving them special constitutional protection.305   Because Smith had relegated religious groups to the legislative process, Missouri said the Supreme Court should do the same in Trinity Lutheran.

The Court refused.  Instead, seven Justices rejected the appeal to Smith as controlling.  First, the Courtrelied on Hosanna-Tabor as showing that Smith’s rule does not necessarily mean that “any application of a valid and neutral law of general applicability is necessarily constitutional under the Free Exercise Clause.”306   Second, the Court found that Smith’s rule did not exclude judicial protection where a party’s “religious status” resulted in “special disabilities,” a principle the Court drew not only from Smith but from both Establishment and Free Exercise Clause cases.307   Where the government has violated this rule, its actions are subject not to mere rational basis, but to “the most exacting scrutiny.”308 

None of this is consistent with a Gobitis approach to constitutional rights, or with the Thayer/Hand approach of deferring to rational legislative choices.  The Court was not willing to merely defer to the legislative choices of the majority and was not willing to apply mere rational basis review.  Instead, the Constitution functioned as a serious check on majority power.  Even if the harm to the religious party was likely only “a few extra scraped knees,” such discrimination was “odious to our Constitution all the same, and cannot stand.”309 

The Court reached a similar result in Espinoza v. Montana Department of Revenue, where it rejected a Montana state constitutional provision that excluded religious schools from participating in public programs.310   The Court rejected Blaine Amendments as “born of bigotry” and having a “shameful pedigree,” leading to discrimination that is “condemn[ed]” by the First Amendment.311   The Court applied strict scrutiny to find Montana’s exclusion of religious schools invalid under the Free Exercise Clause.312 

In 2022, the Court extended this rationale in Carson v. Makin, again rejecting a provision barring public funds for private schools solely because the schools are religious.313   Maine argued its program was distinct from the one in Espinoza on the grounds that (1) Maine’s tuition assistance program was designed to stand in as a substitute to traditional public schooling and (2) Maine’s restrictions on religious schools was “use-based” rather than “status-based.”314   The Court rejected both arguments.  Even if Maine’s political branches wanted to exclude religious schools, the state still was not free to violate “free exercise principles governing any such public benefit program.”315   Furthermore, the Court held that a status versus use distinction is no “less offensive to the Free Exercise Clause.”316 

In one sense, it is possible to view Trinity Lutheran, Espinoza, and Carson as somewhat reconcilable with Smith because the laws at issue are not “neutral” toward religion.  Still, the decisions are far more in line with the nondeferential vision of the First Amendment set forth in Barnette rather than Gobitis and SmithTrinity Lutheran, Espinoza, and Carson alltreat the First Amendment as a legal principle to be enforced by courts, which cannot simply defer to rational legislative judgments.  None suggests that there is “no other test which this Court is authorized to apply” than rational basis, as asserted by Justice Frankfurter.317   None seems willing to accept condemning religious minorities to a “relative disadvantage” in their religious exercise as the “unavoidable consequence of democratic government”—namely majoritarian control—that “must be preferred” to judicially created exemptions.318   To the contrary, these cases treat majority-imposed restrictions on religious character—even for entities that are seeking to participate in public funding programs—as “odious to our Constitution.”319 

3.   COVID-19 Lockdowns

Over the past two years, the Court has also had occasion to consider the Free Exercise Clause in a series of cases concerning COVID-19 lockdowns.  Many lower courts had upheld limits on religious worship gatherings in reliance on the deferential standards set forth in both Smith and Jacobson v. Massachusetts.320   On an emergency application by Catholic and Jewish houses of worship, however, the Court issued a per curiam opinion in Roman Catholic Diocese of Brooklyn v. Cuomo granting relief against New York’s limits on indoor worship gatherings.321   While many other activities in New York had been subject to percentage-based occupancy limits—which meant that larger buildings could accommodate more people—the State had limited gatherings for religious worship to only ten or twenty-five people in certain zones, regardless of the size of the building.322 

The lower courts, in reliance on Smith and Jacobson and in light of the public health emergency posed by COVID-19, had upheld the restrictions using only rational basis review.323   The Supreme Court, however, applied strict scrutiny because it found the restrictions were not neutral and generally applicable.324   It found that New York’s restrictions struck “at the very heart of the First Amendment’s guarantee of religious liberty.”325   The Court found that New York’s rules were:

far more restrictive than any COVID-related regulations that have previously come before the Court, much tighter than those adopted by many other jurisdictions hard-hit by the pandemic, and far more severe than has been shown to be required to prevent the spread of the virus at the applicants’ services.326 

While this analysis was technically conducted in line with Smith, the Court curiously did not rely on Smith (it cited Lukumi instead).327   Lower courts have already suggested that Diocese of Brooklyn “represented a seismic shift in Free Exercise law” and compels application of strict scrutiny far more broadly than lower courts had previously thought.328 

Cass Sunstein has called the decision in Diocese of Brooklyn “our anti-Korematsu,” arguing that it is “a strong signal of judicial solicitude for constitutional rights and of judicial willingness to protect against discrimination, even under emergency circumstances in which life is on the line.”329   As discussed above, Korematsu is a Thayerian decision characterized by deference to the government in an emergency.  Sunstein, however, emphasized the Court’s unwillingness to defer to the political branches: “The most noteworthy feature of the per curiam opinion is the absence of deference to state officials in a context in which deference might well be expected.”330   This is the opposite of the Thayerism observable in Korematsu, prompting Sunstein’s label of Diocese of Brooklyn as the “anti-Korematsu.”

The Court followed Diocese of Brooklyn with a per curiam decision about a California COVID-19 restriction on in-home worship in Tandon v. Newsom.331   Again the Court rejected deference to government officials.  Instead, it explained that strict scrutiny would apply—and the government would bear the burden of proving its claims—“whenever they treat any comparable secular activity more favorably than religious exercise.”332 

Notably absent from Tandon was any suggestion that the Court would or should merely defer to government choices, so long as they are rational.  To the contrary, Tandon is quite clear that even if a government distinction might be rational, it is only permissible if the government carries its burden of satisfying strict scrutiny.333   As with Diocese of Brooklyn, Tandon confirms that—even in the face of a pandemic—the Court will enforce the Free Exercise Clause and not simply defer to rational decisions of political actors.

*     *     *

As with the church autonomy cases, then, the Court’s discussions of the Free Exercise Clause in the marriage, Blaine Amendment, and COVID-19 contexts cannot be reconciled with the Thayerian deference of Gobitis.  Rather than deferring to merely rational laws imposed by political actors, these cases suggest an understanding of that clause as an important, and judicially enforceable, protection for a “fundamental” right.  And this nondeferential, judicially enforceable protection applies even in the face of important general laws, and even in emergency circumstances that might otherwise be expected to prompt deference.  “[E]ven in a pandemic, the Constitution cannot be put away and forgotten,” but remains to be enforced by the courts.334   That strong, judicial protection enables the Free Exercise Clause to function as an important “guarantee” of pluralism, even in the face of contrary majoritarian decisions on our most contentious issues.  The Gobitis-driven alternative of a First Amendment characterized by judicial deference to majoritarian legislature could hardly be said to “lie[] at the heart of our pluralistic society.”335 

C.   Establishment Clause Cases

From the outset, Justice Frankfurter was concerned about the relationship between the Establishment Clause and his Gobitis approach to the First Amendment.  If courts were not limited to the Gobitis approach of deferring to the legislature, Justice Frankfurter feared that public schools would be unable to require school days to begin with readings from the Bible.336 

Of course, Justice Frankfurter eventually lost that fight, both in Barnette itself and on the issue of Bible reading in public schools.  By the early 1960s, in reliance on Barnette, the Court rejected requirements of daily Bible reading, just as Justice Frankfurter had feared.337   In particular, the Court emphasized in School District of Abington Township v. Schempp that the mere “consent of the majority” was not enough to overcome the First Amendment, which had put such issues “beyond the reach of majorities” and “establish[ed] them as legal principles to be applied by the courts.”338 

The Establishment Clause cases of the past decade have featured some division and fracturing over exactly how the Court should implement that principle, and how to determine what exactly amounts to an impermissible establishment of religion.  But in most circumstances there appears to be unanimity that the test is not simply to apply Thayerian deference to any reasonable legislative judgment.

First, in Town of Greece v. Galloway, a divided Court rejected an Establishment Clause challenge to a legislative prayer practice in Greece, New York.339   While the Justices were divided 5–4 as to whether the Establishment Clause had been violated, they were unanimous in embracing a view of the First Amendment that positions the Constitution—and the Court—as an important guarantor of minority rights in a religiously pluralistic society.340   This is, of course, the Barnette view of the First Amendment rather than the deferential Thayer/Gobitis approach.

Writing for the majority, Justice Kennedy explained that the Town’s legislative prayers did not violate the First Amendment in part because they did not “denigrate nonbelievers or religious minorities” which “would present a different case.”341   Echoing Barnette, the majority explained that the government cannot “prescribe a religious orthodoxy,” and that legislative prayer practices should instead “strive for the idea that people of many faiths may be united in a community of tolerance and devotion.”342   The majority further emphasized that coercion by the government would be unacceptable, but that “[o]ur tradition assumes that adult citizens, firm in their own beliefs, can tolerate and perhaps appreciate a ceremonial prayer delivered by a person of a different faith.”343   The prayers at issue had been said by members of a variety of different faiths including Christian ministers, a Jewish layman, a Wiccan priestess, and a Baha’i temple chairman.344   In short, the Court acknowledged that the legislative majority was barred from imposing religious orthodoxy, but did not think such imposition had occurred.

Justice Kagan’s dissent likewise embraced the Barnette idea of the First Amendment as an important, and judicially enforceable, protection for religious pluralism.  Writing for the four dissenting Justices, Kagan emphasized the Constitution’s “remarkable” and “momentous offering” to people of all faiths: “that however those individuals worship, they will count as full and equal American citizens.”345   Justice Kagan dissented because she believed the town’s practices violated “the breathtakingly generous constitutional idea that our public institutions belong no less to the Buddhist or Hindu than to the Methodist or Episcopalian.”346   The dissenters emphasized that they did not think town meetings must “become a religion-free zone,” but rather that “pluralism and inclusion in a town hall can satisfy the constitutional requirement of neutrality.”347 

What is notable here is that all nine Justices in Town of Greece were operating from a set of common premises, all of which are the antithesis of Thayerian deference.  All treated the Constitution as having an important role to play protecting minorities and religious pluralism—a role disavowed by Gobitis and Smith, but embraced by Barnette.  All appeared to presume that judicial enforcement of the First Amendment is a central part of that protection.  All agreed that the government cannot impose orthodoxy.  None suggested that the courts should only be looking for mere legislative rationality when the legislative majority has enacted a general law.  In short, while the Town of Greece decision was 5–4, there appeared to be 9–0 agreement on an anti-Thayerian approach to the Establishment Clause.

Five years later, in American Legion v. American Humanist Association, the Court decided that it did not violate the Establishment Clause for a Maryland town to keep a forty-foot cross on public property as a memorial for soldiers who died in World War I.348  As in Town of Greece, the Court split 7–2.349   Likewise, the opinions confirmed that all nine of the Justices view the Religion Clauses as designed to provide important protection for religious minorities.

For example, the majority explained its view that the “Religion Clauses of the Constitution aim to foster a society in which people of all beliefs can live together harmoniously.”350   Further, the Court also explained that tearing down or moving the cross “would not be neutral and would not further the ideals of respect and tolerance embodied in the First Amendment.”351   The Court’s view of the Religion Clauses here is, once again, irreconcilable with the approach set forth in Gobitis, which offered no protection to religious minorities against majority-imposed orthodoxy.  The First Amendment could not adequately protect pluralism in this way if it gives minorities no protection against even barely rational majoritarian power.

Justices Ginsburg and Sotomayor dissented, making clear that theydisagreed with the case’s outcome.  But they took no issue with the claim that the Religion Clauses are designed to allow people of all beliefs to live together in harmony.  Indeed, the dissenters asserted that the Establishment Clause is “designed to preserve individual liberty and civic harmony” by keeping the government neutral between religions.352   And they lamented even “the indirect coercive pressure upon religious minorities to conform” to any majority-imposed religious orthodoxy.353 

The majority and dissent in American Legion, therefore, disagreed about the correct outcome of the case because they disagreed about the application of these principles.  But all nine Justices agreed that the Religion Clauses are designed to allow for people of varying beliefs to live together in harmony and without forced conformity imposed on religious minorities.  In other words, while reaching different conclusions on the facts of the case, the Justices exhibited broad agreement in thinking about religious liberty as a Barnette-style right to be enforced by the Court for the protection of minorities.  A Gobitis approach of Thayerian deference to all nonirrational legislative actions would have offered no such protection.

In 2022, the Supreme Court struck down two separate Establishment Clause defenses made by government entities.  First, in Shurtleff v. City of Boston, the Court on free speech grounds ruled that Boston could not reject a request to fly the Christian flag due to its religious nature.354   For years, Boston had frequently allowed private groups to fly a flag of their choosing on a flagpole outside of City Hall.355   The city permitted hundreds of requests, and did not deny a single one until a group called Camp Constitution requested to fly a Christian flag.356   The city argued it was avoiding a violation of the Establishment Clause, as flying a flag at City Hall would be an expression of government speech.357   The Court unanimously rejected such a stance, holding it is essential to prevent the government from denying private speech solely on the grounds that the speech is religious.358   Justice Kavanaugh, in concurrence, affirms a Barnette understanding of the Religion Clauses by stating the government may not treat religious individuals and organizations “as second-class.”359 

Second, the Court ruled in Kennedy v. Bremerton School District that the Establishment Clause does not preclude a public school coach from participating in individual, noncoercive prayer.360   Bremerton, in seeking to avoid violating the Establishment Clause, curtailed Kennedy’s right to pray at the fifty-yard line following a football game.361   A 6–3 majority held that the Free Exercise and Establishment Clauses of the First Amendment had complimentary, rather than adversarial roles, where one must dominate the other.362   While public school coaches and teachers have public-facing roles, the Court noted the importance of protecting these individuals’ right to noncoercive, private expressions of faith.363   In doing so, the Court invoked a Barnette-style approach to protecting religious expression, with the First Amendment providing judicially enforceable protection against “government attempts to regulate religion and suppress dissent.”364   The Court emphasized the importance of these rights particularly for minorities—emphasizing the example of a Muslim teacher who could be fired “for wearing a headscarf” or a Jewish teacher for wearing a yarmulke—and explained that “learning how to tolerate speech or prayer of all kinds is ‘part of learning how to live in a pluralistic society,’ a trait of character essential to ‘a tolerant citizenry.’”365   Together, Town of Greece, American Legion, Shurtleff,and Kennedy all show the Court treating the First Amendment’s religious freedom protections—both the Establishment Clause and the Free Exercise Clause—as important, judicially enforceable protections against the majority.  None advances the Gobitis view that religious restrictions should be permissible upon a mere showing of rational basis.

There is one exception to this trend: the Court’s 2018 decision in Trump v. Hawaii.366   In Trump, a 5–4 majority upheld a presidential order preventing the entry into the United States of certain foreign nationals of eight countries.367   The majority found that the decision whether to admit foreign nationals implicated “a fundamental act of sovereignty” that is “within the core of executive responsibility.”368   Accordingly, the Court found that ordinary constitutional principles did not apply, and instead the Court would only look for a “facially legitimate and bona fide” reason, and would not “look behind the exercise of that discretion.”369 

Based on this extraordinarily deferential approach, the Court “assume[d]” that, at most, it could apply “rational basis review” to determine whether the government’s actions were plausibly related to legitimate government interests.370   The majority emphasized that this deferential approach was not ordinary Establishment Clause review, but rather a special standard for the particular circumstance at issue.371   Based on this standard, the court upheld the restriction.372   This is the closest the Court has come to Thayerian deference in the religious liberty area in the past decade.

Writing in dissent for herself and Justice Ginsburg, Justice Sotomayor argued both that the restrictions should have been subjected to more rigorous scrutiny and that they failed even rational basis review.373   Justice Sotomayor criticized the majority for applying rational basis review “without explanation or precedential support” where “a more stringent standard of review” was required.374   Further, she emphasized that use of rational basis scrutiny amounted to “throw[ing] the Establishment Clause out the window” and “forgo[ing] any meaningful constitutional review.”375 

The majority, of course, disagreed with Justice Sotomayor’s criticism of their use of rational basis.376   But they did not disagree with the contention that rational basis is extraordinarily deferential.  To the contrary, they emphasized that “it should come as no surprise that the Court hardly ever strikes down a policy as illegitimate under rational basis scrutiny.”377   The majority did not apply rational basis because it understood that to be the usual Establishment Clause standard, but because it understood the particular situation to be an exceptional circumstance, warranting deference akin to Thayerism.378 

*     *     *

The trio of Establishment Clause cases decided over the past decade thus confirms the general victory of an anti-Thayerian approach to the First Amendment over a deferential Gobitis-styleapproach.  Even where the Justices disagree on outcomes, they all appear to be operating from the same premise that the Religion Clauses are supposed to be judicially enforceable protections and, in virtually all circumstances, to call for something greater than mere rationality review.  Even in the lone case to use rational basis, the Justices appear to agree that such deference is generally inappropriate for constitutional claims.

D.   Federal Statutory Cases

Interspersed with the constitutional cases described above, the Court also decided a host of federal statutory religious liberty cases during the past decade.  These cases involved a wide range of statutes: RFRA,379  RLUIPA,380  and Title VII of the 1964 Civil Rights Act.381   And they involved a wide range of factual circumstances: healthcare mandates,382  prison conditions,383  employment,384  interstate travel,385  and capital punishment.386   Across these diverse contexts and statutes, an obvious trend emerged—every religious plaintiff won, many in unanimous decisions.  As relevant to our discussion, three themes emerge from these statutory cases that make them relevant to our constitutional analysis: courts can apply strict scrutiny; anarchy does not result from judicial scrutiny; and the Court continues to resist appeals to Smith and deference to political branches.

First, these statutory cases demonstrate the Court’s ability to apply strict scrutiny.  Smith had feared that a nondeferential approach like strict scrutiny would require courts to “weigh the social importance of all laws against the centrality of all religious beliefs”387  and would “court[] anarchy.”388   But these statutory cases confirm that courts can and do apply strict scrutiny without trying to “weigh” such imponderables without sowing anarchy.  Nor do they turn themselves into “board[s] for answering legislative conundrums” as Thayer had feared.389 

For example, the Court has decided three RFRA cases concerning the federal contraceptive coverage mandate.390   In the most prominent of the trio, Burwell v. Hobby Lobby Stores, Inc.,the Court had no difficulty applying strict scrutiny—and did so without “weighing” the centrality of all religious beliefs against the importance of all laws.391   Instead, the Court simply found that the burdened religious exercise was sincere and looked to see whether the government was using the least restrictive means of achieving a compelling interest.392   The Court assumed that the government had a compelling interest in distributing contraceptives, but found that the government did not need to force unwilling employers to pay for the coverage.393   It found that “[t]he most straightforward way of doing this would be for the Government to assume the cost” for anyone unable to obtain them because of an employer’s religious objection.394 

To be sure, the defendants and dissenters predicted that ruling for the religious parties would “lead to a flood of religious objections.”395   But empirical research shows that no such flood occurred at all—to the contrary, religious liberty claims and victories remain scarce.396   Thus the Court expressed no trepidation about resulting anarchy when it recently held in Little Sisters of the Poor Saints Peter and Paul Home v. Pennsylvania that the entire federal government needs to consider RFRA in all of its actions,397  or when it unanimously held in Tanzin v. Tanvir that Muslim men, who claimed they were wrongfully included on the no-fly list, could pursue RFRA damages against officials in their personal capacities.398 

The Court has likewise shown no hesitation about applying strict scrutiny under RLUIPA.  In Holt v. Hobbs, the Court applied strict scrutiny to require the Arkansas prison system to allow a Muslim inmate to grow a half-inch beard for religious reasons.399   The unanimous Court refused to simply defer to the government, particularly where most other prison systems allowed such beards without jeopardizing prison safety.400   And the Holt precedent was recently used by Justice Kagan to explain why strict scrutiny required Alabama to refrain from carrying out an execution without allowing the prisoner access to clergy.401   There is no indication that either allowing short beards for religious reasons or allowing access to clergy in the death chamber has resulted in anarchy or safety problems.402   Thus although these cases did not involve direct applications of Smith, they provided the Court with repeated opportunities to apply and discuss strict scrutiny, which was one of the primary concerns of the Smith majority.

The Court recently ruled in another prisoner’s rights case—Ramirez v. Collier—that individuals during an execution ought to have the right to have their minister lay hands on their body while audibly praying.403   As in Holt, the Court applied strict scrutiny under RLUIPA.404   Texas requested deference on execution policy determinations, citing security in the execution chamber, among other reasons, as a governmental interest to not have a minister physically touching Ramirez.405   However, an 8–1 court did not find such a reason sufficient.  Complete bans on audible prayer and touch were not the least restrictive means to fulfill such an interest.406   Failure to accommodate an individual’s religious beliefs in their last hours presents irreparable harm that compensation to one’s estate could not remedy.407 

One statutory case did involve a direct appeal to Smith.  In EEOC v. Abercrombie & Fitch Stores, Inc., the popular clothing store had refused to hire a Muslim teenager because her religious observance of wearing a hijab or headscarf would conflict with the store’s “no-headwear” policy.408   Attempting to defend the exclusion against a Title VII claim of religious discrimination, Abercrombie invoked Smith to argue that “generally applicable rules that happen to burden religion” are unproblematic.409 

Rejecting this invocation of Smith, the Court explained that Title VII does not “limit disparate-treatment claims to only those employer policies that treat religious practices less favorably than similar secular practices.”410   The Court left open the possibility that such a rule “may” work in “other contexts,” but found that Title VII “does not demand mere neutrality with regard to religious practices.”411   Title VII instead protects religious exercise by giving “favored treatment” and often requires that otherwise neutral polices “give way to the need for an accommodation.”412 

*     *     *

While these nonconstitutional cases do not directly test whether the Court is using the Thayerian deferential approach revived in Smith, they allow us to see the Court repeatedly protecting religious exercise.  More specifically, these statutes repeatedly place the Court in the exact position Thayer, Hand, Gobitis, and Smith all suggested would be most problematic: applying heightened scrutiny to general laws enacted by the majority.  The Court exhibits no discomfort with this assigned task, nowhere suggests that its efforts court anarchy, and is willing to insist both to governments and to private employers that they must often accommodate religious exercise from burdensome general rules.


Understanding the Court’s religious liberty cases through the lens of deference provides an important insight into what Justice Kagan rightly called our Constitution’s “momentous offering” to the millions who “have come to this country from every corner of the world to share in the blessing of religious freedom.”413   That blessing is built on the First Amendment’s Religion Clauses, including the constitutional promise of free exercise, which the Court recently recognized as at “the heart of our pluralistic society.”414 

For decades it has been obvious that, at least as a general matter, there are no takers on the Court or in the academy for application of mere rational basis review when governments infringe fundamental rights.  Thayer and Hand proposed such a deferential approach, but there is little doubt that the Barnette approach controls.  Barnette replaced Gobitis and its Thayerian deference to legislative majorities with an understanding of the Bill of Rights as a set of “legal principles to be applied by the courts.”415   No one seems to want to turn back, and the occasional case in which the Court applies rational basis now prompts dissenting Justices to liken the practice to throwing the First Amendment “out the window” and “forgo[ing] any meaningful constitutional review.”416 

Barnette’s rejection of Gobitis and its embrace of the judicial enforcement of rights controls in virtually all areas of modern constitutional law.  Few would dispute Cass Sunstein’s characterization of Barnette as “what American constitutional law is all about,”417  at least when it comes to the enforcement of rights.  This is why Barnette is invoked even outside of the First Amendment when the Court wrestles with how to protect rights.  To borrow a line from Barnette, “[t]he very purpose” of constitutional rights is to identify circumstances in which courts should not defer to political actors and, instead, enforce constitutional principles that are binding on majoritarian government.418   

For years, Smith’s surprising reliance on Gobitis rather than Barnette created an anomaly: How could the “the heart of our pluralistic society” be a Free Exercise Clause so hobbled by deference to merely rational government action?  How could the “momentous offering” of religious liberty be so stingy as to often leave minority rights unprotected as the “unavoidable consequence” of majority power?419   How could that version of the Bill of Rights plausibly protect the minority from majoritarian power?

When we consider religious liberty law through the lens of deference, we can see that the Court’s recent religious liberty cases have begun to provide answers to these questions.  Working in virtually every area of religious liberty law, the Court has repeatedly embraced a Barnette-style understanding of religious liberty, rejecting the notion of Thayerian deference.  Outside of Trump v. Hawaii,the Court has not deferred to merely rational judgments of the political branches.420   There is broad agreement on the Court that rationality review is insufficient for protecting important religious liberty rights.  And there appears to be little concern on the Court that judicial application of strict scrutiny—the spectre of which drove Smith’s embrace of Gobitis—is actually sowing anarchy.  In fact, the Justices seem convinced of the opposite: that serious enforcement of free exercise is a crucial component of the Constitution’s ability to promote social peace in a pluralistic society.  As Michael McConnell has argued at length elsewhere, the most logical result of a pluralistic approach to the Free Exercise Clause is not Smith but religious exemptions.421 

So what will become of Smith?  Only two possibilities seem plausible.  One is that the Court will soon reverse Smith, and thereby eliminate the last remaining vestige of the Gobitis-style approach to fundamental rights.  This would certainly be consistent with the cases of the past decade, and several Justices in Fulton appear quite open to that path.  It would not be a surprise to see the Court reverse Smith and embrace a more protective understanding of the Free Exercise Clause, even if there is some uncertainty on the Court about how a replacement for Smith would play out in various circumstances.  In fact, Justices Barrett and Kavanaugh openly suggested in Fulton that perhaps Smith’s rule should not be replaced by a single categorical rule.422  

The second possibility is that, even with the Gobitis approach so clearly defunct, Smith will survive a bit longer, perhaps governing a much narrower range of cases than previously thought.  Scholars have long pointed out that Smith’s focus on laws that are “neutral” and “generally applicable”423  leaves room for argument about whether most laws qualify for Smith’s rational basis rule or not.424   Perhaps the Court will deal with the incongruence of Smith by finding that Smith is more the exception than the rule.  For example, the Court could nominally preserve Smith,but decide, as it did in Trump v. Hawaii, that it only governs a very small, exceptional set of cases for whicha rational basis rule can apply.  The Court’s recent decisions in Tandon and Fulton certainly accomplish some of that work, eschewing deference any time the government allows secular exemptions (Tandon) or reserves discretion to create exemptions (Fulton).

Either way, one thing seems clear: After a decade of developing all aspects of its religious liberty jurisprudence, the Court has left Smith looking rather isolated—a lonely island of Gobitis-style deference in what is otherwise a Barnette-controlledsea of judicially-enforced constitutional rights.

We live in deeply divided times.  The Court seems well aware that we need a fully operational First Amendment that allows free people with divergent views to live and work together in peace.  While the Justices will of course disagree on particular applications, they share broad agreement that the Religion Clauses exist to “foster a society in which people of all beliefs can live together harmoniously.”425   The Religion Clauses can only play that role if the Court continues its movement away from Thayerian deference and toward the notion that the entire Bill of Rights, including the Religion Clauses, imposes legal principles that can and should be enforced by courts.

© 2022 Mark L. Rienzi.  Individuals and nonprofit institutions may reproduce and distribute copies of this Article in any format at or below cost, for educational purposes, so long as each copy identifies the author, provides a citation to the Notre Dame Law Review, and includes this provision in the copyright notice.

* Professor of Law, Catholic University of America, Columbus School of Law; President, Becket Fund for Religious Liberty.  Thank you to all who read, commented, or provided research assistance for this Article.  Disclaimer: I have represented religious parties or amici in virtually all of the post-2011 Supreme Court’s religious liberty cases discussed herein.  All opinions expressed are my own, rather than my employers’ or clients’.

See Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 599–600 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

Id. at 594.

Barnette, 319 U.S. at 667 (Frankfurter, J., dissenting).

See Gobitis, 310 U.S. at 599.

See James B. Thayer, The Origin and Scope of the American Doctrine of Constitutional Law, 7 HARV. L. REV. 129, 140 (1893).

Plessy v. Ferguson, 163 U.S. 537 (1896), overruled by Brown v. Bd. of Educ., 347 U.S. 483 (1954).

Buck v. Bell, 274 U.S. 200, 205–08 (1927).

Korematsu v. United States, 323 U.S. 214 (1944).

NOAH FELDMAN, SCORPIONS: THE BATTLES AND TRIUMPHS OF FDR’S GREAT SUPREME COURT JUSTICES 31 (2010); see also HARLAN B. PHILLIPS, FELIX FRANKFURTER REMINISCES: RECORDED IN TALKS WITH DR. HARLAN B. PHILLIPS 300–01 (Harlan B. Phillips ed., Anchor Books 1962) (1960) (in which Frankfurter calls Thayer’s 1893 article the “most important single essay” about American constitutional law and “the great guide for judges”).

10 See W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943).

11 Id.

12 See infra Section II.B.

13 See infra Section II.B.

14 See, e.g., Obergefell v. Hodges, 576 U.S. 644, 677 (2015) (quoting Barnette, 319 U.S. at 638, for the idea that “certain subjects” were meant to be withdrawn “from the vicissitudes of political controversy . . . plac[ing] them beyond the reach of majorities and officials and . . . establish[ing] them as legal principles to be applied by the courts”); Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 851 (1992) (citing Barnette, 319 U.S. 624,for the proposition “that a State may not compel or enforce one view” on an issue when it would “intrude upon a protected liberty”), overruled by Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228, 2284 (2022).

15 See, e.g., John D. Inazu, A Confident Pluralism, 88 S. CAL. L. REV. 587, 591–92 (2015) (quoting Barnette, 319 U.S. at 642, for the proposition that “[a] confident pluralism maintains that we can” live with differences; “in fact, we must embrace a ‘right to differ’ from state and majoritarian norms”); Holning Lau, Pluralism: A Principle for Children’s Rights, 42 HARV. C.R.-C.L. L. REV. 317, 349 (2007) (“The Court planted the seed of the pluralism principle in West Virginia State Board of Education v. Barnette.”).

16 See Emp. Div. v. Smith, 494 U.S. 872 (1990).

17 See id. at 879.  Surprisingly, Smith relied on Gobitis without indicating that it had been overruled.  See id.

18 See, e.g.,Michael W. McConnell, Free Exercise Revisionism and the Smith Decision, 57 U. CHI. L. REV. 1109, 1114–28 (1990) (discussing the textual, historical, and precedential arguments for and against exemptions); Michael W. McConnell, The Origins and Historical Understanding of Free Exercise of Religion, 103 HARV. L. REV. 1409, 1415 (1990) [hereinafter McConnell, Origins and Historical Understanding] (arguing that the historical evidence, on balance, supports exemptions); Philip A. Hamburger, A Constitutional Right of Religious Exemption: An Historical Perspective, 60 GEO. WASH. L. REV. 915, 916 (1992) (arguing that the evidence does not support a constitutional right to religious exemptions).

19 See Smith, 494 U.S. at 879.

20 Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1754 (2020).  Six Justices joined the opinion in Bostock (Justice Gorsuch, joined by Chief Justice Roberts and Justices Breyer, Sotomayor, Ginsburg, and Kagan).  Justices Alito and Kavanaugh joined the similar statement in American Legion v. American Humanist Association that the “Religion Clauses of the Constitution aim to foster a society in which people of all beliefs can live together harmoniously.”  Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2074 (2019).

21 Fulton v. City of Phila., 141 S. Ct. 1868 (2021).

22 Am. Legion, 139 S. Ct. at 2074.


24 Id.  Richard Posner would later describe Justice Frankfurter colorfully as having “advocated Thayerism with a noisy passion unequaled by any other Thayerian.”  Richard A. Posner, The Rise and Fall of Judicial Self-Restraint, 100 CALIF. L. REV. 519, 530 (2012).

25 Edward A. Purcell, Jr., Learned Hand: The Jurisprudential Trajectory of an Old Progressive, 43 BUFF. L. REV. 873, 885 (1995) (reviewing GERALD GUNTHER, LEARNED HAND: THE MAN AND THE JUDGE (1994)) (quoting PHILLIPS, supra note 9, at 300–01).

26 Thayer, supra note 5, at 144.

27 Id.

28 Id.  Edward Purcell has argued that Thayer intended his deferential standard to apply only when the courts are revising the work of coordinate branches of the federal government (since that was Thayer’s chief focus) and not to actions by the states.  See Purcell, supra note 25, at 886.  As Purcell notes, this is not necessarily how Thayer’s theory was understood by his adherents at the time.  See id.  And as Gobitis demonstrates, it was not how Frankfurter and his colleagues applied the theory.  As Michael Perry has observed, “[e]ven Frankfurter failed to note the distinction—or to heed it, as his dissent in Barnette makes clear.”  Michael J. Perry, Is Capital Punishment Unconstitutional?  And Even If We Think It Is, Should We Want the Supreme Court to So Rule?, 41 GA. L. REV. 867, 872 n.13 (2007).

29 Thayer, supra note 5, at 134 (quoting MASS. CONST. pt. I, art. XXX).

30 Id. at 136 n.1, 136–37.

31 Id. at 144.

32 Id.

33 Id. at 146.

34 See id. at 155–56; see also BICKEL, supra note 23, at 40.

35 Thayer, supra note 5, at 155–56.

36 Id. at 156.

37 Of course, as the Court has often acknowledged, the structural and separation of powers aspects of the Constitution were in fact designed to protect individual liberty.  See, e.g., Bond v. United States, 564 U.S. 211, 221–22 (2011).

38 Purcell, supra note 25, at 874; see LEARNED HAND, THE BILL OF RIGHTS: THE OLIVER WENDELL HOLMES LECTURES, 1958, at 56 (1958) (“I do not think that the interests mentioned in the First Amendment are entitled in point of constitutional interpretation to a measure of protection different from other interests . . . .”).

39 See HAND, supra note 38, at 56; see also BICKEL, supra note 23, at 46–49.


41 Brown v. Bd. of Educ., 347 U.S. 483 (1954).

42 See HAND, supra note 38, at 54–55.

43 See id. (suggesting that what Brown did was “‘overrule’ the ‘legislative judgment’ of states by its own reappraisal of the relative values at stake”).

44 Id. at 55.

45 GUNTHER, supra note 25, at 666; see also Purcell, supra note 25, at 921–22.

46 See GUNTHER, supra note 25, at 666–71.

47 See id. at 667–71.

48 Id. at 671.

49 Plessy v. Ferguson, 163 U.S. 537 (1896), overruled by Brown v. Bd. of Educ., 347 U.S. 483 (1954).

50 Id. at 548, 550 (emphasis added).

51 Id. at 550.

52 Id.

53 Id. at 550–51.

54 Id. at 550.

55 Plessy’s embrace of Thayerism in the Fourteenth Amendment context is particularly troublesome given that the Amendment’s supporters clearly intended for it to provide strong, enforceable constitutional protections against legislative infringements.  For example, in language that seems to foreshadow Barnette, then-Congressman James Garfield explained about the possibility of his political opponents taking away statutory civil rights:

The civil rights bill is now a part of the law of the land.  But every gentleman knows it will cease to be a part of the law whenever the sad moment arrives when that gentleman’s party comes into power.  It is precisely for that reason that we propose to lift that great and good law above the reach of political strife, beyond the reach of the plots and machinations of any party, and fix it in the serene sky, in the eternal firmament of the Constitution, where no storm of passion can shake it and no cloud can obscure it.

CONG. GLOBE, 39th Cong., 1st Sess. 2462 (1866) (emphasis added).  Congressman Garfield would surely have been surprised to learn that a later Supreme Court would think the only thing they had raised “above the reach of political strife” and fixed into the “eternal firmament of the Constitution” was a bare reasonableness test.

56 Buck v. Bell, 274 U.S. 200, 205–08 (1927).

57 Id. at 205, 207 (“In view of the general declarations of the legislature and the specific findings of the Court, obviously we cannot say as matter of law that the grounds do not exist, and if they exist they justify the result.”).

58 See id. at 207.

59 Jacobson v. Massachusetts, 197 U.S. 11, 29 (1905) (“[T]he individual in respect of his liberty may at times, under the pressure of great dangers, be subjected to such restraint, to be enforced by reasonable regulations, as the safety of the general public may demand.”); see id. at 31 (“Whatever may be thought of the expediency of this statute, it cannot be affirmed to be, beyond question, in palpable conflict with the Constitution.”).

60 Buck, 274 U.S. at 207 (“Three generations of imbeciles are enough.”).

61 See id. at 208 (“But the answer is that the law does all that is needed when it does all that it can, indicates a policy, applies it to all within the lines, and seeks to bring within the lines all similarly situated so far and so fast as its means allow.”).

62 Korematsu v. United States, 323 U.S. 214 (1944), abrogated by Trump v. Hawaii, 138 S. Ct. 2392 (2018).

63 Id. at 216.

64 See id. at 218 (“[W]e cannot reject as unfounded the judgment of the military authorities . . . .” (quoting Hirabayashi v. United States, 320 U.S. 81, 99 (1943))); id. (“We cannot say that the war-making branches of the Government did not have ground for believing that . . . .” (quoting Hirabayashi, 320 U.S. at 99)); id. at 219 (“[W]e could not reject the finding of the military authorities . . . .”); id. at 224 (“We cannot—by availing ourselves of the calm perspective of hindsight—now say that at that time these actions were unjustified.”).

65 Thayer, supra note 5, at 144.

66 See Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 591–92 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

67 Letter from Billy Gobitas to Minersville, Pa. Sch. Dirs. (Nov. 5, 1935), [].  The family name of “Gobitas” was misspelled in court reports as “Gobitis.”

68 Id.

69 See James Madison, Memorial and Remonstrance Against Religious Assessments [ca. 20 June] 1785, NATIONAL ARCHIVES: FOUNDERS ONLINE (Jun. 20, 1785), [].

70 Letter from Billy Gobitas to Minersville, Pa. Sch. Dirs., supra note 67.  

71 See KEVIN SEAMUS HASSON, BELIEVERS, THINKERS, AND FOUNDERS: HOW WE CAME TO BE ONE NATION UNDER GOD 22–24 (2016); James F. Van Orden, “Jehovah Will Provide”: Lillian Gobitas and Freedom of Religion, 29 J. SUP. CT. HIST. 136, 141 (2004).

72 See Minersville Sch. Dist. v. Gobitis, 108 F.2d 683, 684 (3d Cir. 1939); Gobitis v. Minersville Sch. Dist., 21 F. Supp. 581, 583 (E.D. Pa. 1937).

73 See Gobitis, 108 F.2d at 692; Gobitis, 21 F. Supp. at 584.

74 See Gobitis, 108 F.2d 683; Gobitis, 21 F. Supp. 581

75 See Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 598 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

76 Id.

77 Id.

78 Id.

79 Id.

80 W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 671 (1943) (Frankfurter, J., dissenting).  In 1944, Justice Frankfurter would adopt a similarly deferential approach to the political branches in Korematsu, in which he viewed the constitutionality of Japanese internment as a decision for the legislative and executive branches: “That is their business, not ours.”  Korematsu v. United States, 323 U.S. 214, 225 (1944) (Frankfurter, J., concurring).

81 Gobitis, 310 U.S. at 600; cf. Barnette, 319 U.S. at 671 (Frankfurter, J., dissenting) (“Only a persistent positive translation of the faith of a free society into the convictions and habits and actions of a community is the ultimate reliance against unabated temptations to fetter the human spirit.”).

82 Gobitis, 310 U.S. at 594.

83 Id. at 600.

84 See id.

85 Id. at 598, 595.

86 Id. at 597–98.

87 Id. at 599.

88 Id. at 595.

89 Id. at 597.

90 Id. at 600.

91 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 648 (1943) (Frankfurter, J., dissenting).

92 Id. at 667.

93 Id. at 647.

94 Id. at 666.

95 Id.

96 Stephen Gard has aptly described Justice Frankfurter’s test as one that “negatively define[s] the freedom of speech guaranteed by the first amendment as that speech which no reasonable person could conceive of a reason to suppress.”  Stephen W. Gard, The Flag Salute Cases and the First Amendment, 31 CLEV. ST. L. REV. 419, 433 (1982).  Gard also explains how “in the hands of Justice Frankfurter this test operated like a rachet to contract progressively the scope of constitutionally protected liberty” because once the Court had decided Gobitis, of course it would be reasonable for legislatures to think a forced flag salute was permissible.  Id.

97 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 594 (1940), overruled by Barnette, 319 U.S. 624.

98 Id.

99 See Barnette, 319 U.S. at 655 (Frankfurter, J., dissenting).

100 Id. at 657 (quoting Hamilton v. Regents of the Univ. of Cal., 293 U.S. 245, 268 (1934) (Cardozo, J., concurring)).

101 Id. at 659.

102 Id. at 662.

103 Id.

104 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 606 (1940) (Stone, J., dissenting), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

105 See id. at 594–98 (majority opinion).

106 See SHAWN FRANCIS PETERS, JUDGING JEHOVAH’S WITNESSES: RELIGIOUS PERSECUTION AND THE DAWN OF THE RIGHTS REVOLUTION 8–13 (2000) (“The U.S. Supreme Court’s notorious ruling in the Gobitis flag-salute case, handed down in June 1940, helped to ignite some of the worst anti-Witness violence of the period.”); id. at 13 (“From 1938 to 1946, when the persecution of Jehovah’s Witnesses was reaching almost epidemic proportions in some parts of the United States, the Court handed down twenty-three opinions covering a total of thirty-nine Witness-related cases.”);DAVID T. SMITH, RELIGIOUS PERSECUTION AND POLITICAL ORDER IN THE UNITED STATES 119–22 (2015); see also Garrett Epps, America’s New Lesson in Tolerance, THE ATLANTIC (Sept. 1, 2016), [].

107 FELDMAN, supra note 9, at185.  The violence prompted Eleanor Roosevelt to ask, “[m]ust we drag people out of their homes to force them to do something which is in opposition to their religion?”  ELEANOR ROOSEVELT, MY DAY: THE BEST OF ELEANOR ROOSEVELT’S ACCLAIMED NEWSPAPER COLUMNS, 1936–1962, at 46 (David Emblidge ed., 2001).

108 See PETERS, supra note 106, at 811.

109 See id. at 9, 91–95.


111 Murad Hussain, Defending the Faithful: Speaking the Language of Group Harm in Free Exercise Challenges to Counterterrorism Profiling, 117 YALE L.J. 920, 963 n.223 (2008).

112 ACLU, supra note 110.

113 Gard, supra note 96, at 425 (“In fact, however, as a result of official efforts to enforce the requirement, parents had been subjected to the threat of imprisonment, and children were subjected to the threat of being made wards of the state and of being removed from the custody of their parents.”).

114 Jones v. Opelika, 316 U.S. 584 (1942), vacated, 319 U.S. 103 (1943).

115 See id. at 623 (Black, Douglas & Murphy, JJ., dissenting) (arguing that the Gobitis approach “tends to suppress the free exercise of a religion practiced by a minority group”).

116 Id. at 624.

117 Id.

118 Id. at 611 (Murphy, J., dissenting); see id. at 611–12 (“[T]he protection of the Constitution must be extended to all, not only to those whose views accord with prevailing thought but also to dissident minorities . . . .”).

119 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

120 See Gregory L. Peterson, E. Barrett Prettyman, Jr., Shawn Francis Peters, Bennett Boskey, Gathie Barnett Edmonds, Marie Barnett Snodgrass & John Q. Barrett, Recollections of West Virginia State Board of Education v. Barnette, 81 ST. JOHN’S L. REV. 755, 755 (2007).

121 Barnette, 319 U.S. at 628 n.4.

122 Id.

123 See id. at 628–29.

124 Barnette v. W. Va. State Bd. of Educ., 47 F. Supp. 251, 252 (S.D. W. Va. 1942) (“There is, therefore, but one question for our decision, viz.: Whether children who for religious reasons have conscientious scruples against saluting the flag of the country can lawfully be required to salute it.”).

125 Id. at 253.

126 Id.

127 Id. at 254.

128 Id.

129 Id.  The court further explained why serious constitutional review required judicial analysis of any claimed threat, rather than mere deference to a legislative judgment.  See id. at 253–54 (“There is not a religious persecution in history that was not justified in the eyes of those engaging in it on the ground that it was reasonable and right and that the persons whose practices were suppressed were guilty of stubborn folly hurtful to the general welfare.”).

130 Id. at 254 (quoting Schneider v. State, 308 U.S. 147, 161 (1939)).

131 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 637 (1943).

132 Id.

133 See id. at 639.

134 Id.

135 Id.

136 Id.

137 Id. at 637 (emphasis added).

138 Id. at 642.

139 Id. at 641.

140 Id.

141 Id. at 638 (emphasis added).

142 Id.

143 Id. at 639–40.  Barnette also noted that the Court had a continuing obligation to enforce First Amendment protections, even as the nation changed from one in which “liberty was attainable through mere absence of governmental restraints” in the eighteenth century to having “expanded and strengthened governmental controls” in the twentieth century.  Id.

144 Id. at 642.

145 Id. (“The decision of this Court in Minersville School District v. Gobitis and the holdings of those few per curiam decisions which preceded and foreshadowed it are overruled . . . .”).

146 See, e.g., Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 569 (1993) (Souter, J., concurring in part and concurring in the judgment) (“And Gobitis¸ after three Justices who originally joined the opinion renounced it for disregarding the government’s constitutional obligation ‘to accommodate itself to the religious views of minorities,’ . . . was explicitly overruled in [Barnette].” (quoting Jones v. Opelika, 316 U.S. 584, 624 (1942) (Black, Douglas & Murphy, JJ., dissenting))); Wooley v. Maynard, 430 U.S. 705, 714 (1977) (“In overruling its prior decision in [Gobitis] . . . .”); Parents for Priv. v. Barr, 949 F.3d 1210, 1231 n.17 (9th Cir. 2020) (“The Supreme Court, however, overruled Gobitis three years later in [Barnette].”).  A Westlaw search indicates that Barnette’s“fixed star” line has been quoted by more than 200 published opinions, including fourteen by the Supreme Court.

147 Ronald Dworkin observed that Judge Hand’s views about judicial restraint—based on applying Thayer’s rule to the protection of constitutional rights—“are not much studied in law schools now, or treated as very important.”  DWORKIN, supra note 40, at 343; see also id. at 12 (noting that Hand’s approach “was once an open possibility, [but] history has long excluded it; practice has now settled that courts do have a responsibility to declare and act on their best understanding of what the Constitution forbids”).  As then-Judge Posner explained in 2012, “[t]he ‘rational basis’ criterion of constitutionality, a legacy of Thayer, has dropped away.”  Posner, supra note 24, at 534.

148 See Smith v. Allwright, 321 U.S. 649, 665 & n.10 (1944).

149 Id. at 665.

150 See Ramos v. Louisiana, 140 S. Ct. 1390, 1415 (2020) (Kavanaugh, J., concurring).

151 United States v. Ballard, 322 U.S. 78, 86 (1944).

152 See Girouard v. United States, 328 U.S. 61, 68 (1946).

153 See Prince v. Massachusetts, 321 U.S. 158, 170 (1944).

154 Id. at 165, 165–66 (citing W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943)).

155 Id. at 168.

156 Barnette, 319 U.S. at 647 (Frankfurter, J., dissenting).  Justice Frankfurter continued:

Therefore, in denying to the states what heretofore has received such impressive judicial sanction, some other tests of unconstitutionality must surely be guiding the Court than the absence of a rational justification for the legislation.  But I know of no other test which this Court is authorized to apply in nullifying legislation.

Id. at 666(emphasis added).

157 There is a robust, ongoing debate over whether protection for fundamental rights should be subject to tiers of scrutiny at all, or whether it should be absolute.  See, e.g.,Joel Alicea & John D. Ohlendorf, Against the Tiers of Constitutional Scrutiny, 41 NAT’L AFFS. 72 (2019).  That dispute is beyond the scope of this Article.  For present purposes, it is enough to recognize that whether protection is absolute or is subject to some judicial balancing as under strict or intermediate scrutiny, all such systems fall on the Barnette side of the divide, in that they impose actual, judicially-enforced limits beyond merely deferring to rational legislative judgments as to constitutional rights.

158 Sherbert v. Verner, 374 U.S. 398, 406 (1963).

159 Id. (quoting Thomas v. Collins, 323 U.S. 516, 530 (1945)).

160 Sch. Dist. v. Schempp, 374 U.S. 203, 226, 225–26 (1963).

161 Id. at 226 (quoting Barnette, 319 U.S. at 638).  De’Siree Reeves argues persuasively, based on original research into Justice Brennan’s papers, that the confluence of Sherbert and Schempp on this issue was a result of Justice Brennan’s efforts to focus the Religion Clauses on minority rights.  See De’Siree N. Reeves, Missing Link: The Origin of Sherbert and the Irony of Religious Equality, 15 STAN. J. C.R. & C.L. 201, 236–48 (2019).

162 Barnette, 319 U.S. at 659 (Frankfurter, J., dissenting). 

163 Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 851 (1992).

164 Id.

165 Obergefell v. Hodges, 576 U.S. 644, 681 (2015).

166 Id. at 677.

167 Id. (quoting Barnette, 319 U.S. at 638).

168 See id. at 686 (Roberts, C.J., dissenting); id. at 714 (Scalia, J., dissenting); id. at 721 (Thomas, J., dissenting); Casey, 505 U.S. at 913 (Stevens, J., concurring in part and dissenting in part); id. at 923 (Blackmun, J., concurring in part, concurring in the judgment in part, and dissenting in part); id. at 951 (Rehnquist, C.J., concurring in the judgment in part and dissenting in part); id. at 979 (Scalia, J., concurring in the judgment in part and dissenting in part). 

169 Nor is this treatment of rights limited to the Fourteenth Amendment.  Rather, the Court has emphasized in a variety of contexts that it views the Constitution as reflecting an enforceable judgment or balancing of interests which legislatures and later courts are not free to revise.  See, e.g.,United States v. Stevens, 559 U.S. 460, 470 (2010) (“The First Amendment itself reflects a judgment by the American people that the benefits of its restrictions on the Government outweigh the costs.”); District of Columbia v. Heller, 554 U.S. 570, 635 (2008) (asserting that the Second Amendment “is the very product of an interest balancing by the people” and not alterable by “future legislatures”); Crawford v. Washington, 541 U.S. 36, 61 (2004) (noting that Confrontation Clause “reflects a judgment, not only about the desirability of reliable evidence (a point on which there could be little dissent), but about how reliability can best be determined”).

170 William P. Marshall, Progressive Constitutionalism, Originalism, and the Significance of Landmark Decisions in Evaluating Constitutional Theory, 72 OHIO ST. L.J. 1251, 1251 (2011); see also David L. Hudson, Jr. & Jacob David Glenn, Fixed Stars: Famous First Amendment Phrases and Their Indelible Impact, 15 CHARLESTON L. REV. 189, 190 (2020) (deeming Barnette’s “fixed star” passage among “the most enduring passages in First Amendment jurisprudence”).

171 Bruce Ackerman, Liberating Abstraction, 59 U. CHI. L. REV. 317, 325 (1992).

172 Marshall, supra note 170, at 1251.

173 Michael Stokes Paulsen, The Unconscionable War on Moral Conscience, 91 NOTRE DAME L. REV. 1167, 1170–71 (2016) (reviewing ROBERT P. GEORGE, CONSCIENCE AND ITS ENEMIES: CONFRONTING THE DOGMAS OF LIBERAL SECULARISM (2013)).

174 Cass R. Sunstein, The Supreme Court’s Five Greatest Moments: Will June Bring the Next Standout Ruling?, BLOOMBERG (June 1, 2015), [].

175 Susan R. Estrich & Kathleen M. Sullivan, Abortion Politics: Writing for an Audience of One, 138 U. PA. L. REV. 119, 131 (1989) (“The inclusion of the free exercise clause attests to this point.”).  A Barnette-style approach to the judicial enforcement of the Constitution is also behind, for example, Dean Erwin Chemerinsky’s argument that “the political process cannot be relied on to comply voluntarily with the Constitution” so that “it is likely the courts or nothing for enforcing and upholding the Constitution.”  ERWIN CHEMERINSKY, WE THE PEOPLE: A PROGRESSIVE READING OF THE CONSTITUTION FOR THE TWENTY-FIRST CENTURY 20 (2018); see id. at 19 (“More generally, there is little incentive for the political process to protect unpopular minorities, such as racial or political minorities.”).

176 See, e.g., BICKEL, supra note 23; DWORKIN, supra note 40.

177 BICKEL, supra note 23, at 16 (“The root difficulty is that judicial review is a counter-majoritarian force in our system.”); see JOHN HART ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEW 7 (1980) (“[R]ule in accord with the consent of a majority of those governed is the core of the American governmental system.”).

178 Kurt T. Lash, Originalism, Popular Sovereignty, and Reverse Stare Decisis, 93 VA. L. REV. 1437, 1446 (2007) (“Any legislative action that diverges from this higher law is an inferior expression of the people’s will and deserves invalidation.”).

179 ELY, supra note 177, at 8 (noting the argument that the Bill of Rights can be thought of as a set of “side constraints” on majorities to prevent tyranny, and these constraints are “more democratic” because they “have been imposed by the people themselves”).

180 Posner, supra note 24, at 534; see also Steven G. Calabresi, Originalism and James Bradley Thayer, 113 NW. U. L. REV. 1419, 1424–27 (2019) (noting the harmful effects of Thayerian restraint, including in Plessy v. Ferguson, Debs v. United States, Buck v. Bell, Gobitis,and Korematsu:“Suffice it to say that Thayerian restraint has unquestionably led to some truly terrible case law”).

181 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 606 (1940) (Stone, J., dissenting), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

182 See Emp. Div. v. Smith, 494 U.S. 872, 874 (1990).

183 Id. at 874.

184 As Justice Souter explained in Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 571–72 (1993) (Souter, J., concurring in part and concurring in the judgment), Smith was decided without “full-dress argument” on whether strict scrutiny or rational basis should apply.  Instead, the parties—including the State of Oregon—had treated the strict scrutiny rule as part of the “settled free exercise principles” controlling the case.  Id. (“[N]either party squarely addressed the proposition the Court was to embrace, that the Free Exercise Clause was irrelevant to the dispute.”).

185 Relying in part on Justice Blackmun’s papers, Nathan Lewin has recently argued that the process by which the Court decided Smith actually runs afoul of the “principle of party presentation” the Supreme Court has unanimously endorsed in United States v. Sineneng-Smith, 140 S. Ct. 1575 (2020).  See Nathan Lewin, Did the Supreme Court Forget How It Curtailed Religious Freedom 30 Years Ago?, NEWSWEEK (May 14, 2020) (quoting Sineneng-Smith, 140 S. Ct. at 1577), [].

186 Hobbie v. Unemployment Appeals Comm’n., 480 U.S. 136, 141, 141–42 (1987) (quoting Bowen v. Roy, 476 U.S. 693, 727 (1986) (O’Connor, J., concurring in part and dissenting in part)).

187 Id. at 141–42 (quoting Roy, 476 U.S. at 727 (O’Connor, J., concurring in part and dissenting in part)).  In her opinion in Roy, Justice O’Connor had reviewed the cases and concluded that they demonstrated that “[o]nly an especially important governmental interest pursued by narrowly tailored means can justify exacting a sacrifice of First Amendment freedoms as the price for an equal share of the rights, benefits, and privileges enjoyed by other citizens.”  Roy, 476 U.S. at 728 (O’Connor, J., concurring in part and dissenting in part).  As the Court pointed out in Hobbie, five Justices in Roy shared Justice O’Connor’s views on this point.  See Hobbie, 480 U.S. at 141.

188 Smith, 494 U.S. at 881.

189 See id. at 890.

190 Id. at 888.

191 Id. at 881.

192 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 594 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

193 Smith, 494 U.S. at 890.

194 Thayer, supra note 5, at 146.

195 Gobitis, 310 U.S. at 594.

196 Smith, 494 U.S. at 888.

197 Id. at 879 (quoting Gobitis, 310 U.S. at 594–95).

198 See id.

199 See id.

200 Id. at 878.

201 Id.

202 Id.

203 See id. at 878­88.

204 See id. at 888–90.

205 Id. at 902 (O’Connor, J., concurring in the judgment).

206 Id.

207 Id. at 903, 902–03 (quoting W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 638 (1943)).

208 Id. at 909 (Blackmun, J., dissenting) (citing id. at 890 (majority opinion)).

209 Id.

210 See, e.g.,Douglas Laycock, The Remnants of Free Exercise, 1990 SUP. CT. REV. 1, 1 (“Anger at the result was compounded by anger at the procedure.  The Court sharply changed existing law without an opportunity for briefing or argument, and it issued an opinion claiming that its new rules had been the law for a hundred years.”).

211 See 42 U.S.C. §§ 2000bb to 2000bb-4 (2018).

212 See generally Douglas Laycock & Oliver S. Thomas, Interpreting the Religious Freedom Restoration Act, 73 TEX. L. REV. 209 (1994).

213 See 42 U.S.C. §§ 2000bb to 2000bb-4 (2018).

214 42 U.S.C. § 2000bb(a)(5) (2018).

215 See City of Boerne v. Flores,521 U.S. 507, 536 (1997).

216 Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 424 (2006).

217 42 U.S.C. §§ 2000cc to 2000cc-5 (2018).

218 See Cutter v. Wilkinson, 544 U.S. 709, 715 (2005).

219 See generally Christopher C. Lund, Religious Liberty After Gonzales: A Look at State RFRAs, 55 S.D. L. REV. 466, 467 (2010) [hereinafter Lund, Religious Liberty After Gonzales]; Christopher C. Lund, RFRA, State RFRAs, and Religious Minorities, 53 SAN DIEGO L. REV. 163, 169 (2016) [hereinafter Lund, RFRA, State RFRAs, and Religious Minorities].

220 See generally Lund, Religious Liberty After Gonzales, supra note 219, at 467; Lund, RFRA, State RFRAs, and Religious Minorities, supra note 219, at 164.

221 See Gonzales, 546 U.S. at 424.

222 See Cutter, 544 U.S. at 714.

223 Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 534 (1993).

224 See Douglas Laycock, Comment, Theology Scholarships, the Pledge of Allegiance, and Religious Liberty: Avoiding the Extremes but Missing the Liberty, 118 HARV. L. REV. 155, 201 & n.281 (2004) (“Filing rates for free exercise claims plummeted after Smith, and these claims had lower success rates than the larger number of claims decided before Smith.” (footnote omitted) (citing Amy Adamczyk, John Wybraniec & Roger Finke, Religious Regulation and the Courts: Documenting the Effects of Smith and RFRA, 46 J. CHURCH & STATE 237, 250 tbl.1 (2004), as “reporting 310 claims decided in the nine-and-a-quarter years before Smith, compared to thirty-eight claims decided in three-and-a-half years after Smith”)).

One minor exception is that the Court did briefly address a free exercise claim in a footnote in Christian Legal Society Chapter of the University of California, Hastings College of the Law v. Martinez, 561 U.S. 661, 697 n.27 (2010).  Martinez concerned whether a Christian student group at a public law school should be permitted to restrict its leadership to Christians who shared the group’s beliefs.  See id. at 669–73.  In a testament to the power of Smith, the Christian group had downplayed its free exercise claim so much that the Court could dispose of it in a footnote explaining that Smith protects “otherwise valid” restrictions from attack.  Id. at 697 n.27.

225 To be sure, the mere fact that these cases could be decided under state RFRAs did not guarantee success—many such claims failed.  See Lund, Religious Liberty After Gonzales, supra note 219, at 467.

226 See Mark L. Rienzi, The Case for Religious Exemptions—Whether Religion Is Special or Not, 127 HARV. L. REV. 1395, 1411 (2014) (reviewing BRIAN LEITER, WHY TOLERATE RELIGION? (2013); ANDREW KOPPELMAN, DEFENDING AMERICAN RELIGIOUS NEUTRALITY (2013)) (“The problem with this argument is that we have now had enough experience under religious exemption regimes to know that they do not create anarchy.  For example, we have lived under a religious exemption regime at the federal level for more than twenty years.  And roughly half of all states apply similar regimes based on either state court decisions or state statutes resembling the federal regime.  No serious claim can be made that these systems have produced anarchy, or anything close to it.  Indeed, it is far more likely that none of us even notices when we travel from an exemption state to a nonexemption state and back.” (footnotes omitted)).

227 See, e.g., Tanzin v. Tanvir, 141 S. Ct. 486, 489 (2020); EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768, 770 (2015); Holt v. Hobbs, 574 U.S. 352, 355–56 (2015).

228 See, e.g., Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012, 2017 (2017); Hosanna-Tabor Evangelical Lutheran Church & Sch. v. EEOC, 565 U.S. 171, 177–80 (2012).

229 See, e.g., Agudath Isr. of Am. v. Cuomo, 141 S. Ct. 889 (2020) (mem.).

230 See Little Sisters of the Poor Saints Peter & Paul Home v. Pennsylvania, 140 S. Ct. 2367, 2375 (2020); Our Lady of Guadalupe Sch. v. Morrissey-Berru, 140 S. Ct. 2049, 2055 (2020).

231 See, e.g., Town of Greece v. Galloway, 572 U.S. 565, 572 (2014).

232 See, e.g., Hosanna-Tabor, 565 U.S. at 180; Our Lady, 140 S. Ct. at 2055.

233 See, e.g., Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246, 2251 (2020); Trinity Lutheran, 137 S. Ct. at 2017.

234 See, e.g., Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2074 (2019); Town of Greece, 572 U.S. at 572.

235 See, e.g., Little Sisters, 140 S. Ct. at 2373; Zubik v. Burwell, 578 U.S. 403, 406–07 (2016); Holt v. Hobbs, 574 U.S. 352, 356 (2015); Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 691 (2014).

236 See, e.g., Hosanna-Tabor, 565 U.S. at 178–79.

237 See, e.g., Little Sisters, 140 S. Ct. at 2375.

238 See, e.g., Hobbs, 574 U.S. at 355–56.

239 See, e.g., EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768, 770 (2015).

240 See, e.g., Town of Greece v. Galloway, 572 U.S. 565, 569–70 (2014).

241 See, e.g., Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2074 (2019).

242 Erwin Chemerinsky & Howard Gillman, Symposium: The Unfolding Revolution in the Jurisprudence of the Religion Clauses, SCOTUSBLOG (Aug. 6, 2020, 10:36 AM), [].

243 Linda Greenhouse, Opinion, Religious Crusaders at the Supreme Court’s Gates, N.Y. TIMES (Sept. 12, 2019), [].

244 Leslie Griffin, Symposium: Religions’ Wins Are Losses, SCOTUSBLOG (Aug. 4, 2020, 1:43 PM), [].

245 To be sure, the claim here is not that the Court has only recently discovered the notion that the First Amendment in general, and the religion clauses in particular, serve the cause of pluralism.  Barnette’s poetic language about “[t]he very purpose of a Bill of Rights” makes clear that accommodation of differences has long been understood as a critical aspect of the First Amendment.  See W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 638 (1943); see also Zorach v. Clauson, 343 U.S. 306, 313–14 (1952) (noting that accommodation “follows the best of our traditions”).  My point here is simply that, over the past decade, the Justices have put particular emphasis on the work the religion clauses do to protect pluralism, and that they could not do that work with only a narrow, deferential approach to free exercise as seen in Gobitis or Smith.

246 Hosanna-Tabor Evangelical Lutheran Church & Sch. v. EEOC, 565 U.S. 171 (2012).

247 See id. at178–79.

248 See id. at 179.

249 See id. at 180.

250 See 42 U.S.C. §§ 12101–12213 (2018).

251 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 594 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

252 See 42 U.S.C. § 12101(b)(4) (2018).

253 See Gobitis, 310 U.S. at 597–98 (where “the end is legitimate” the court will not “stigmatize legislative judgment” about how to achieve it).  In his Barnette dissent, Justice Frankfurter defended the Gobitis approach and explained that the Court should only focus on “whether legislators could in reason have enacted such a law.”  Barnette, 319 U.S. at 647 (Frankfurter, J., dissenting).

254 Gobitis, 310 U.S. at 597–98.

255 Brief for the Federal Respondent at 22, Hosanna-Tabor Evangelical Lutheran Church & Sch. v. EEOC, 565 U.S. 171 (2012) (No. 10-553).

256 Hosanna-Tabor Evangelical Lutheran Church & Sch. v. EEOC, 565 U.S. 171, 189 (2012).

257 Id. at 190.

258 Id. at 196.

259 Id. at 184, 196.

260 Our Lady of Guadalupe Sch. v. Morrissey-Berru, 140 S. Ct. 2049 (2020).

261 See id. at 2060.

262 See id. at 2069.

263 See id. at 2063–64, 2068.

264 Id. at 2064.

265 See id.

266 Id. at 2072 (Sotomayor, J., dissenting).

267 Id. at 2073, 2075.

268 Masterpiece Cakeshop, Ltd. v. Colo. C.R. Comm’n, 138 S. Ct. 1719 (2018).

269 Fulton v. City of Phila., 141 S. Ct. 1868 (2021).

270 Obergefell v. Hodges, 576 U.S. 644 (2015).

271 Bostock v. Clayton Cnty., 140 S. Ct. 1731 (2020).

272 See Masterpiece Cakeshop, 138 S. Ct. at 1723.

273 Id. at 1724.

274 Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 594 (1940), overruled by W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624 (1943).

275 Emp. Div. v. Smith, 494 U.S. 872, 881 (1990).

276 Masterpiece Cakeshop, 138 S. Ct. at 1723.

277 Id. at 1727.

278 Indeed, it appears that all nine Justices may have agreed with this principle, as the two dissenters did not take issue with the majority’s statement about clergy.  See id. at 1748 (Ginsburg, J., dissenting).

279 See supra text accompanying notes 165–67.

280 Obergefell v. Hodges, 576 U.S. 644, 672 (2015).

281 Id.

282 Id. at 667.

283 Id. at 672.

284 Id. at 679–80.

285 See Bostock v. Clayton Cnty., 140 S. Ct. 1731 (2020).

286 See id. at 1742, 1754.

287 See id. at 1754.

288 Id.

289 See Fulton v. City of Phila., 141 S. Ct. 1868, 1873 (2021).

290 See id. at 1924 (Alito, J., concurring in the judgment) (“Smith was wrongly decided.”).

291 See id. at 1883 (Barrett, J., concurring) (“We need not wrestle with these questions in this case, though, because the same standard applies regardless whether Smith stays or goes.”).

292 See id. at 1881 (majority opinion) (“Because the City’s actions are therefore examined under the strictest scrutiny regardless of Smith,we have no occasion to reconsider that decision here.”).

293 See id. at 1882.

294 Id. at 1877.

295 Id.; see, e.g., State v. Arlene’s Flowers, Inc., 441 P.3d 1203, 1218 (Wash. 2019), cert. denied, 141 S. Ct. 2884 (2021).

296 Fulton, 141 S. Ct. at 1877.

297 Id. (quoting Emp. Div. v. Smith, 494 U.S. 872, 884 (1990)).

298 See id. at 1878.  The Court further found that home certifications involve “a customized and selective assessment” and a “sensitive process” that agencies “understandably approach” from “different angles.”  Id. at 1880.

299 Id. at 1881 (Barrett, J., concurring); see id. at 1882–83.

300 Id. at 1883 (Alito, J., concurring in the judgment).  Justice Alito feared that the Court’s decision could “vanish” leaving the parties “back where they started” if Philadelphia simply eliminated the contract’s waiver provision.  Id. at 1887.  On remand, however, Philadelphia conceded defeat under the Free Exercise Clause and has since resumed working with Catholic Social Services.  See Julia Terruso, Philadelphia Reaches $2 Million Settlement with Catholic Foster-Care Agency, Aiming to Prevent Future Challenges to LGBTQ Rights, PHILA. INQUIRER(Nov. 22, 2021), [].

301 See Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246, 2259 (2020) (noting that Blaine Amendments were “born of bigotry” and carry a “shameful pedigree” (quoting Mitchell v. Helms, 530 U.S. 793, 828–29 (2000) (plurality opinion))).

302 See Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012, 2025 (2017).

303 See id. at 2017.

304 See Brief of Respondent at 4, Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012 (2017) (No. 15-577).

305 See id. at 37, 37–39.

306 Trinity Lutheran, 137 S. Ct. at 2021 n.2.

307 Id. at 2019, 2019–20 (first quoting Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, 508 U.S. 520, 533, 542 (1993); then citing McDaniel v. Paty, 435 U.S. 618, 628 (1978); and then citing Everson v. Bd. of Educ., 330 U.S. 1, 67 (1947)).

308 Id. at 2021 (citing Lukumi, 508 U.S. at 546).

309 Id. at 2025, 2024–25.

310 See Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246, 2262 (2020).

311 Id. at 2259, 2262.

312 Id. at 2260.

313 Carson v. Makin, 142 S. Ct. 1987, 2002 (2022).

314 See id. at 1998.

315 Id. at 2000.

316 Id. at 2001.

317 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 666 (1943) (Frankfurter, J., dissenting).

318 Emp. Div. v. Smith, 494 U.S. 872, 890 (1990).

319 Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246, 2255 (2020) (quoting Trinity Lutheran Church of Colombia, Inc. v. Comer, 137 S. Ct. 2012, 2025 (2017)).

320 Josh Blackman has collected many of these cases.  See Josh Blackman, The “Essential” Free Exercise Clause, 44 HARV. J.L. & PUB. POL’Y 637, 647–61 (2021).

321 See Roman Cath. Diocese of Brooklyn v. Cuomo, 141 S. Ct. 63, 66 (2020).

322 See id. at 66–67.

323 See, e.g., Roman Cath. Diocese of Brooklyn v. Cuomo, 495 F. Supp. 3d 118, 127, 129–31 (E.D.N.Y. 2020) (“[N]early every court to consider the issue has followed suit and applied a rational basis analysis to free exercise challenges to COVID-related restrictions on religious gatherings.”), rev’d and remanded sub nom. Agudath Isr. of Am. v. Cuomo, 983 F.3d 620, 625 (2d Cir. 2020).

324 See Diocese of Brooklyn, 141 S. Ct. at 67.

325 Id. at 68.

326 Id. at 67.

327 Id.

328 Calvary Chapel Dayton Valley v. Sisolak, 982 F.3d 1228, 1232 (9th Cir. 2020).

329 Cass R. Sunstein, Our Anti-Korematsu, 2021 AM. J.L. & EQUAL. 221, 222, 232.

330 Id. at 225.  Sunstein argues that this rejection of deference to the political branches is “of potentially enduring importance.”  Id. at 222.

331 See Tandon v. Newsom, 141 S. Ct. 1294, 1297 (2021).

332 Id. at 1296.

333 Id. at 1297–98 (“This is the fifth time the Court has summarily rejected the Ninth Circuit’s analysis of California’s COVID restrictions on religious exercise.  It is unsurprising that such litigants are entitled to relief.  California’s Blueprint System contains myriad exceptions and accommodations for comparable activities, thus requiring the application of strict scrutiny.” (first citing Harvest Rock Church, Inc. v. Newsom, 141 S. Ct. 889 (2020) (mem.); then citing S. Bay United Pentecostal Church v. Newsom, 141 S. Ct. 716 (2021); then citing Gish v. Newsom, 141 S. Ct. 1290 (2021) (mem.); and then citing Gateway City Church v. Newsom, 141 S. Ct. 1460 (2021) (mem.))).

334 Roman Cath. Diocese of Brooklyn v. Cuomo, 141 S. Ct. 63, 68 (2020).

335 Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1754 (2020).

336 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 659 (1943) (Frankfurter, J., dissenting).  Frankfurter’s concern here likely was not the result of personal religious fervor, given that he described himself as “reverent agnostic.”  PHILLIPS, supra note 9, at 291.

337 See Sch. Dist. v. Schempp, 374 U.S. 203, 226 (1963).

338 Id. at 225–26 (quoting Barnette, 319 U.S. at 638).  The Bible reading in Schempp was required for the school, but individual students were permitted to refrain from participation; the Court still invalidated it, citing BarnetteSee id. at 205–07.

339 See Town of Greece v. Galloway, 572 U.S. 565, 568–70 (2014).

340 See id. at 591; id. at 603 (Alito, J., concurring); id. at 610 (Breyer, J., dissenting); id. at 630 (Kagan, J., dissenting).

341 Id. at 583 (majority opinion).

342 Id. at 581, 584.

343 Id. at 584, 583–84.

344 See id. at 572.

345 Id. at 615 (Kagan, J., dissenting).

346 Id. at 616.

347 Id.

348 See Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2074 (2019).

349 See id. at 2074, 2090, 2092, 2103.

350 Id. at 2074.

351 Id. at 2090.

352 Id. at 2104 (Ginsburg, J., dissenting).

353 Id. at 2105 (quoting Engel v. Vitale, 370 U.S. 421, 431 (1962)).

354 See Shurtleff v. City of Bos., 142 S. Ct. 1583, 1587 (2022)

355 See id. at 1587.

356 See id.

357 Id. at 1593.

358 See id. at 1587.

359 Id. at 1595 (Kavanaugh, J., concurring).

360 See Kennedy v. Bremerton Sch. Dist., 142 S. Ct. 2407, 2433 (2022).

361 See id. at 2417–18.

362 See id. at 2421.

363 See id. at 2423.

364 Id. at 2421.

365 Id. at 2425, 2430–31 (quoting Lee v. Weisman ex rel. Weisman, 505 U.S. 577, 590 (1992)).  The Court went on to emphasize that “[r]espect for religious expressions is indispensable to life in a free and diverse Republic” and protected by both the free speech and free exercise clauses.  Id. at 2432–33.

366 Trump v. Hawaii, 138 S. Ct. 2392 (2018).

367 See id. at 2403, 2423.

368 Id. at 2407, 2418 (quoting United States ex rel. Knauff v. Shaughnessy, 338 U.S. 537, 542 (1950)).

369 Id. at 2419 (quoting Kleindienst v. Mandel, 408 U.S. 753, 769, 770 (1972)).  The Court emphasized that this deferential review was the same level of review it had applied in a prior free speech case concerning a denial of entry.  See id. (citing Mandel, 408 U.S. at 753).

370 Id. at 2420.

371 See id. at 2418 (“The case before us differs in numerous respects from the conventional Establishment Clause claim.”).

372 See id. at 2423.

373 See id. at 2441 (Sotomayor, J., dissenting).

374 See id.

375 Id. at 2441 n.6.

376 See id. at 2419 (majority opinion).

377 Id. at 2420.

378 See id. at 2418.  For an argument that Trump v. Hawaii would have had a better chance of triggering heightened scrutiny had the case been argued under the Free Exercise Clause, rather than the Establishment Clause, see Brief Amicus Curiae of the Becket Fund for Religious Liberty in Support of Neither Party at 5–6, Trump v. Hawaii, 138 S. Ct. 2392 (2018) (No. 17-965) (“The lower courts’ use of the wrong Clause and the wrong test have led them to decide important questions of First Amendment rights and national security by relying on inferences about the state of mind of a single government official. . . . Because the Proclamation’s constitutionality under the First Amendment has not properly been litigated below, the case should be remanded, and Respondents should be given the chance to litigate their thus far undeveloped Free Exercise claim.”).

379 See Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 688–90 (2014).

380 See Holt v. Hobbs, 574 U.S. 352, 356 (2015).

381 See EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768, 771 (2015).

382 See Hobby Lobby Stores, 573 U.S. at 688–91.

383 See Holt, 574 U.S. at 355.

384 See Abercrombie & Fitch, 575 U.S. at 770.

385 See Tanzin v. Tanvir, 141 S. Ct. 486, 489 (2020).

386 See Ramirez v. Collier, 142 S. Ct. 1264, 1268 (2022).

387 Emp. Div. v. Smith, 494 U.S. 872, 890 (1990).  To Smith, avoiding this task was precisely why rational basis “must be preferred”—even if it leaves religious minorities at a disadvantage:

It may fairly be said that leaving accommodation to the political process will place at a relative disadvantage those religious practices that are not widely engaged in; but that unavoidable consequence of democratic government must be preferred to a system in which each conscience is a law unto itself or in which judges weigh the social importance of all laws against the centrality of all religious beliefs.


388 Id. at 888 (“Any society adopting such a system would be courting anarchy, but that danger increases in direct proportion to the society’s diversity of religious beliefs . . . .”).

389 Thayer, supra note 5, at 146.

390 Little Sisters of the Poor Saints Peter and Paul Home v. Pennsylvania, 140 S. Ct. 2367, 2372–73 (2020); Zubik v. Burwell, 578 U.S. 403, 405–06 (2016); Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 688–91 (2014).

391 See Hobby Lobby, 573 U.S. at 691–92.

392 See id. at 726–31.  The dissenters too recognized that strict scrutiny did not allow them to “question the centrality of a particular religious exercise.”  Id. at 748 (Ginsburg, J., dissenting).

393 See id. at 726–31 (majority opinion).

394 Id. at 728.  The Court further explained that the regulatory method of compliance the federal government had created for religious nonprofits would also be less restrictive of Hobby Lobby’s religious liberty.  See id. at 730.  That method was subsequently enjoined when challenged by nonprofits, see Zubik, 578 U.S. at 409, and revised to become optional by the federal government after the Supreme Court “directed” the government to “accommodat[e]” objections to that process.  Little Sisters, 140 S. Ct. at 2383 (alteration in original) (quoting Zubik, 578 U.S. at 408).

395 Hobby Lobby, 573 U.S. at 732.

396 See, e.g., Luke W. Goodrich & Rachel N. Busick, Sex, Drugs, and Eagle Feathers: An Empirical Study of Federal Religious Freedom Cases, 48 SETON HALL L. REV. 353, 356 (2018) (“Contrary to predictions that Hobby Lobby would open the floodgates of religious liberty litigation, these cases remain scarce, making up only 0.6% of the federal docket.  And contrary to predictions that religious people would be able to wield Hobby Lobby as a trump card, successful cases are even scarcer . . . .”); Stephanie H. Barclay & Mark L. Rienzi, Constitutional Anomalies or As-Applied Challenges?  A Defense of Religious Exemptions, 59 B.C. L. REV. 1595, 1639 (2018) (“Our findings do not indicate that government win rates have undergone a dramatic change since Hobby Lobby.”); cf. Stephen Cranney, Are Christians More Likely to Invoke RFRA—and Win—than Other Religions Since Hobby Lobby?, 72 MERCER L. REV. 585, 593 (2021) (reviewing datasets and concluding that RFRA “is primarily used to protect less privileged minority religions”).

397 See Little Sisters, 140 S. Ct. at 2383 (finding “Congress’ intent beyond dispute” that federal agencies must consider RFRA in the implementation of federal law).  Prior to the Court’s holding in Little Sisters, some lower courts had asserted that agencies lacked authority to independently interpret and apply RFRA.  See, e.g.,Pennsylvania v. Trump, 351 F. Supp. 3d 791, 822 (E.D. Pa. 2019), aff’d sub nom. Pennsylvania v. President U.S., 930 F.3d 543 (3d Cir. 2019), rev’d and remanded sub nom. Little Sisters, 140 S. Ct. at 2367.

398 See Tanzin v. Tanvir, 141 S. Ct. 486, 489 (2020).  In Tanzin, the Court was considering the claim of three Muslim men who alleged they had wrongfully been placed on the federal “No Fly List” because they had refused to serve as government informants on members of their religious congregation.  See id.

399 See Holt v. Hobbs, 574 U.S. 352, 357–58 (2015) (noting that Smith had “largely repudiated the method of analysis used in prior free exercise cases like” Yoder and Sherbert, but that RLUIPA is designed to provide “expansive protection for religious liberty”).

400 See id. at 368–69.  The Court emphasized that it did not believe the application of strict scrutiny would hamper prison security, because strict scrutiny still “affords prison officials ample ability to maintain security” while also respecting religious exercise.  Id. at 369.

401 See Dunn v. Smith, 141 S. Ct. 725, 725 (2021) (mem.) (Kagan, J., concurring in denial of application to vacate injunction)

402 See id. (“But past practice, in Alabama and elsewhere, shows that a prison may ensure security without barring all clergy members from the execution chamber.  Until two years ago, Alabama required the presence of a prison chaplain at an inmate’s side.”); accord id. at 727 (Kavanaugh, J., dissenting from denial of application to vacate injunction) (asserting that if states wish to avoid RLUIPA litigation, they “should figure out a way to allow spiritual advisors into the execution room, as other States and the Federal Government have done”).

403 Ramirez v. Collier, 142 S. Ct. 1264, 1284 (2022).

404 See id. at 1277.

405 See id. at 1280.

406 See id.

407 See id. at 1282.

408 EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768, 770 (2015).

409 Brief for Respondent at 25, 24–26, EEOC v. Abercrombie & Fitch Stores, Inc., 575 U.S. 768 (2015) (No. 14-86).

410 Abercrombie, 575 U.S. at 775.

411 Id.

412 Id.

413 Town of Greece v. Galloway, 572 U.S. 565, 615 (2014) (Kagan, J., dissenting).

414 Bostock v. Clayton Cnty., 140 S. Ct. 1731, 1754 (2020).

415 W.Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 638 (1943).

416 Trump v. Hawaii, 138 S. Ct. 2392, 2441 n.6 (2018) (Sotomayor, J., dissenting).

417 Sunstein, supra note 174.

418 Barnette, 319 U.S. at 638. 

419 Emp. Div. v. Smith, 494 U.S. 872, 890 (1990).

420 And even in Trump, there was broad agreement that rational basis is extremely deferential review not applicable to ordinary Establishment Clause claims.  See Trump, 138 S. Ct. at 2418–20; id. at 2441 (Sotomayor, J., dissenting).

421 McConnell, Origins and Historical Understanding, supra note 18, at 1516 (stating “[t]he Madisonian perspective points toward pluralism” as the animating principle of the Free Exercise Clause).

422 See Fulton v. City of Phila., 141 S. Ct. 1868, 1883 (2021) (Barrett, J., concurring).

423 Smith, 494 U.S. at 881.

424 Douglas Laycock & Steven T. Collis, Generally Applicable Law and the Free Exercise of Religion, 95 NEB. L. REV. 1, 11 (2016) (noting a circuit split and wide variance in the lower courts over how to determine general applicability under Smith).

425 Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2074 (2019).