Introduction
The Supreme Court’s recently developed religious antidiscrimination principle asserts that when the government provides a generally available benefit, it may not bar religious entities from that benefit because of their religious identity.
In the public-education context, the doctrine has, over the last few years, altered longstanding barriers between church and state that have historically served to ensure that public schools do not become embroiled in the religious divisions of the day.
St. Isidore v. Drummond,
decided in May 2025, tested religious antidiscrimination’s outer bounds. A private, Catholic, virtual school, St. Isidore sought to become an Oklahoma charter school, contending that it is religious discrimination to exclude religious schools while contracting with secular charter schools.
But charter schools in Oklahoma are strictly public: They are approved by a state board, subject to stringent curricular requirements, obligated to follow state antidiscrimination law, and funded by public tax dollars.
When the Oklahoma state charter board approved St. Isidore to become a public school in violation of state law, the Oklahoma Attorney General brought suit in state court.
St. Isidore responded with a First Amendment religious discrimination claim.
After the Oklahoma Supreme Court held that, under state law, charter schools are public and therefore required to be nonsectarian,
St. Isidore brought its Free Exercise claim to the United States Supreme Court.
Following a string of successful cases beginning with Trinity Lutheran Church of Columbia, Inc. v. Comer,
many observers decried the lack of a limiting principle on the Court’s religious antidiscrimination principle, and St. Isidore was widely expected to extend the Court’s previous doctrine.
Instead, the Court punted. In a 4-4 per curiam decision without an opinion, the Court affirmed the Oklahoma Supreme Court’s judgment, and St. Isidore remains barred from becoming America’s first religious public school.
What explains this outcome? In the absence of a written decision, the Court’s religious antidiscrimination principle remains opaque, with public schools persisting as an apparent exception to the existing doctrine. Almost certainly, new plaintiffs will bring similar claims, this time hoping to avoid the recusal of a member of the Court’s conservative supermajority.
And the question will be posed again: What’s so different about public schools?
This Comment seeks to provide one possible answer and, at the same time, develop a coherent limiting principle for future cases. Looking to Professor Reva Siegel’s antibalkanization theory,
this Comment argues that the concerns that have motivated the Court to curb race-based affirmative action apply with equal or greater force in the context of religious antidiscrimination in public schools. Part I examines the free exercise doctrine the Court has developed in the Trinity Lutheran line of cases and concludes that the doctrine presently lacks any coherent limitation that can explain the outcome in St. Isidore. Part II identifies and distills antibalkanization principles as limitations on the Equal Protection Clause’s antidiscrimination principle in the Court’s racial equality cases. Part III develops the normative and historical justification for applying those principles in the Free Exercise context.
I. The First Amendment’s Antidiscrimination Principle
This Part describes the current status of the Court’s religious antidiscrimination principle: If states provide a generally available benefit, they may not exclude religious entities “solely on account of religious identity.”
It synthesizes the Trinity Lutheran line of cases, identifying two possible exceptions to the principle’s broad scope: (1) the Locke v. Davey
exception and (2) the public school exception. It then demonstrates that the Court has yet to provide a coherent doctrinal justification for either, ultimately concluding that the outcome in St. Isidore is doctrinally tenuous.
A. The Trinity Lutheran Trilogy
In Trinity Lutheran, Chief Justice John Roberts, writing for a 7-2 majority, struck down a Missouri constitutional provision barring aid to religious entities.
The Court found Missouri engaged in impermissible religious discrimination by “expressly requir[ing] Trinity Lutheran to renounce its religious character in order to participate in an otherwise generally available public benefit program.”
Because Missouri “impose[d] a penalty on the free exercise of religion,” its program could “be justified only by a state interest ‘of the highest order.’”
Despite stipulating that there would be no Establishment Clause violation if the state had provided playground funding to Trinity Lutheran, the only interest Missouri asserted was a desire to avoid Establishment Clause concerns.
Chief Justice Roberts had no difficulty dismissing this interest: The state’s policy of “expressly denying a qualified religious entity a public benefit solely because of its religious character” simply went “too far.”
At first blush, the decision was a limited one. In footnote three, Chief Justice Roberts cabined the decision to “express discrimination based on religious identity with respect to playground resurfacing,” reserving decision on “religious uses of funding or other forms of discrimination.”
Justice Neil Gorsuch, on the other hand, was keen to make clear the breadth of Trinity Lutheran’s holding, joining the opinion but for the footnote because “the general principles here do not permit discrimination against religious exercise—whether on the playground or anywhere else.”
Three years later in Espinoza v. Montana Department of Revenue,
Justice Gorsuch’s concurrence proved prescient. In a 5-4 decision drawing three concurrences and three dissents, the Court again struck down a state constitutional no-aid provision.
Montana had adopted a school-choice program that provided a tax credit to individuals who donated to a fund that supported families sending their children to private schools, but it barred any use of the tax-credit-based funding by religious schools.
Citing Trinity Lutheran for the proposition that the Free Exercise Clause “‘protects religious observers against unequal treatment’ and against ‘laws that impose special disabilities on the basis of religious status,’” the Court held that the Montana policy discriminated against the religious plaintiffs.
Because Montana required that a school “divorce itself from any religious control or affiliation” to receive aid, the state had engaged in “status-based discrimination . . . subject to ‘the strictest scrutiny.’”
Montana trotted out the same anti-establishment interest that Missouri attempted in Trinity, and the Espinoza Court batted it down just as swiftly.
The Court gave similarly short shrift to Montana’s interest in protecting religious freedom by ensuring that taxpayers did not fund views that they opposed, asserting that the program was not liberty-enhancing because it burdened both religious organizations and religious persons.
Most significantly, the Court rejected Montana’s asserted interest in safeguarding public school funding. Montana’s program was “fatally underinclusive” because it had itself established a system that diverted public funding to private schools; the state could not then claim an interest in preserving that same funding for traditional public schools.
As Chief Justice Roberts put it, “[O]nce a State decides to [subsidize private education], it cannot disqualify some private schools solely because they are religious.”
Just two Terms later, the Court doubled down in Carson ex rel. O.C. v. Makin.
Maine authorized a program of tuition assistance for rural families that excluded religious schools.
Unsurprisingly, the Court found that the Maine program fell directly under Espinoza.
Like Montana, Maine had “‘disqualif[ied] some private schools’ from funding ‘solely because they are religious,’” so strict scrutiny applied.
But rather than asserting general anti-establishment interests as in the previous cases, Maine argued that it sought to provide the “rough equivalent” of a public education, a distinction sufficient for the First Circuit to distinguish Espinoza.
Not so for Chief Justice Roberts. Maine’s interest in providing the “rough equivalent” of public education really meant an interest in depriving religious private schools from funding.
Echoing Espinoza, Chief Justice Roberts concluded that Maine’s mistake was providing funding for private schools in the first place: If Maine wanted to keep public funding out of religious schools, it had to keep public funding out of all private schools.
B. The Illimitable Antidiscrimination Principle
The thrust of these opinions is that a state unconstitutionally discriminates against religion by making a public benefit generally available and then withholding it from religious entities. The Court has nodded to two possible limiting exceptions, but, so far, no asserted government interest has satisfied “the strictest scrutiny.”
-
The Locke Exception. — In Trinity Lutheran and each of its progeny, the state defendants analogized to Locke v. Davey, which upheld a scholarship program that prohibited using state funds to pursue a degree in theology.
While Locke itself stated its anti-establishment holding in rather broad terms,
the Court has progressively chipped away at it. By Carson, the Court, drawing on Justice Gorsuch’s concurrence in Trinity Lutheran, had distinguished Locke as standing for a “history and tradition” exception to First Amendment antidiscrimination.
Espinoza makes clear this is a narrow exception. In Espinoza, the Court explained that no “historic and substantial” tradition could support Montana’s interest in precluding religious schools from receiving aid because “[i]n the founding era and the early 19th century” some state governments funded denominational schools.
Never mind that more than thirty states had adopted no-aid provisions by the second half of the nineteenth century: “Such a development, of course, cannot by itselfestablish an early American tradition.”
Further, Chief Justice Roberts concluded that the no-aid provisions “hardly evince[d] a tradition that should inform our understanding of the Free Exercise Clause” because they were tinged with anti-Catholic animus.
This normative framing allows the Justices to pick and choose what history they are willing to recognize as “historic and substantial.” The Locke exception is thus highly manipulable and evidently limited only to those exceptions that have an unbroken chain of tradition from the Founding to today.
-
The Public School Exception. — The public school exception may have more legs. The plaintiffs in Carson explicitly acknowledged the exception at oral argument, conceding they could not argue for “a constitutional right to a publicly funded religious education” because “Espinoza said point blank a state need not subsidize private education.”
Still, Espinoza provided no rationale for why states that do not want to fund religious schools can avoid the antidiscrimination principle by refusing to fund private schools.
It could be that public education is a different sort of public benefit and therefore does not trigger the antidiscrimination principle in the same way that a reimbursement, tax credit, or scholarship does. Or it could be that states have a sufficient interest in maintaining a system of public schools to overcome strict scrutiny, but only if the state does not undercut that interest by funding secular private schools.
Neither rationale fits neatly with the First Amendment antidiscrimination principle. The Court has evinced no desire to draw lines between different state benefits that might trigger strict scrutiny. Justices Gorsuch and Thomas specifically disavowed such an approach in their Trinity Lutheran concurrence.
Moreover, funding traditional public schools to indirectly benefit families can plausibly be described as a public benefit comparable to funding private school attendance. Maine ran into precisely this problem in Carson, failing to adequately define a “‘rough equivalent’ of a public education.”
Without a clear articulation of some meaningful difference between public and private education, public schooling looks like any other public benefit subject to the antidiscrimination principle.
Applying the public school exception at the compelling interest stage of strict scrutiny might be more appealing to the Court’s conservatives. Doing so would not risk diluting strict scrutiny’s application to other public benefits, avoiding the line-drawing problem. Still, state defendants asserting public-education interests run into Carson’s underinclusivity problem. Many states provide funding to private and religious schools for things like busing or technology.
Even assuming a state system gave no tuition aid to private schools, providing these services would likely undercut the financial rationale in the face of the “strictest scrutiny” standard.
C. St. Isidore: A Brief Respite From the Inevitable Tide
Whatever the doctrinal explanation, Chief Justice Roberts has been emphatic that a state “may provide a strictly secular education in its public schools.”
And yet in St. Isidore v. Drummond, the Court granted certiorari to consider “[w]hether a state violates the Free Exercise Clause by excluding privately run religious schools from the state’s charter school program solely because the schools are religious.”
Tellingly, the question presented failed to note that this is a public charter school program—Oklahoma’s charter schools are fully integrated into the public school system and subject to virtually identical legal requirements.
Moreover, the facts of St. Isidore were different from the Trinity Lutheran line in at least one key respect: Oklahoma had not already opened its public school funding to private schools like the Carson and Espinoza states had.
The case thus directly questioned the vitality of the public school exception.
The Justices arrived at a dead-even split, by default upholding the safe harbor established for states that do not fund private schools.
Without a vote count or an opinion, it is difficult to say what exactly happened in St. Isidore. Given the Court’s treatment of the Locke exception in the past, it’s unlikely to have played a role. That leaves only the public school exception as a possible ground for the decision. Chief Justice Roberts, having created the public school exception,
is the Justice most likely to have joined the liberal wing of the Court, and his questions at oral argument point to a concern with maintaining a safe harbor for state systems that preserve funding exclusively for public schools.
Without a decision, however, the public school exception remains unmoored from any compelling doctrinal explanation, making it almost certain the Court will face this question again. To survive, the exception will require a stronger theoretical basis.
II. Racial Equality, Antibalkanization, and the Roberts Court
Looking to Equal Protection Clause doctrine may help states define a sufficiently compelling interest to justify certain religious classifications. Over the last three decades, the Court’s affirmative action decisions have shown a concern with the social cohesion and racial antagonism that informs the sorts of interests that justify racial classifications or, at the least, race consciousness.
Dubbed “antibalkanization” by Professor Reva Siegel, this approach to Equal Protection Clause jurisprudence has led the court to strike a balance between “assur[ing] members of underrepresented groups that they have an opportunity to participate, while doing so in ways designed to reassure majority groups that their participation is not thereby unjustly constrained.”
Identified by Siegel as an alternative to the prevailing anticlassification and antisubordination approaches to the Equal Protection Clause, antibalkanization “holds that the goals of avoiding racial conflict and eradicating racial inequality are not only compatible, but that achieving the goal of eradicating racial inequality essentially requires avoiding or minimizing race-based social conflict where possible.”
Antibalkanization sees the Court’s “overarching role” as “that of drawing the complex doctrinal ‘boundary lines’ that allow the government to address the ongoing problem of racial inequality while preventing the government from creating counterproductive racial conflict and tension while doing so.”
As a result, antibalkanization most often calls for an antidiscrimination principle that seeks to avoid entrenchment and division along suspect class lines.
Similar entrenchment concerns arise in the religious context, and antibalkanization could provide a coherent limitation on First Amendment antidiscrimination doctrine. This Part distills the core values of antibalkanization and argues that those values continue to motivate the Court.
Beginning with his concurrence in Parents Involved v. Seattle Community Schools, Justice Anthony Kennedy incorporated the antibalkanization perspective as an essential aspect of the Court’s racial equality cases.
In an opinion written by Chief Justice Roberts, the Court struck down two K–12 programs that used race-conscious means to achieve diverse student bodies.
Justice Kennedy’s concurrence criticized the Court for being “inconsistent in both its approach and its implications with the history, meaning, and reach of the Equal Protection Clause.”
Justice Kennedy agreed that the programs were a facial classification subject to strict scrutiny and that the school districts had not provided sufficient evidence of narrow tailoring.
But the majority’s “all-too-unyielding insistence that race cannot be a factor” went too far.
Justice Kennedy recognized that barring all government consideration of race would entrench racial divisions and bar government from “recogniz[ing] and confront[ing] the flaws and injustices that remain” in American society.
Eschewing the majority’s rigid anticlassification approach, Kennedy asserted that school districts could use general, race-conscious means to achieve race-conscious ends.
Recognizing the reality of latent racial disparities in American society meant that state and local education agencies could “consider the racial makeup of schools and . . . adopt general policies to encourage a diverse student body, one aspect of which is its racial composition.”
So long as programs did not “reduce children to racial chits,” “avoiding racial isolation” was a compelling interest.
Justice Kennedy struck a similar tone, joined by the Court’s conservative wing, in Ricci v. DeStefano, a case involving a city employer that had repealed its own qualification test on the basis that it had produced a disparate racial impact.
For Justice Kennedy, it was no problem that the employer had undertaken a race-conscious effort in designing a qualification test that would “ensure broad racial participation.”
The statutory violation occurred only after “the raw racial results [of the test] became the predominant rationale for the City’s refusal to certify the results.”
Designing a program with racial effects in mind ensured racial inclusion; discarding results because of their racial effects risked engendering racial animus among applicants.
In 2015, Justice Kennedy again incorporated the antibalkanization perspective in a 7-2 majority opinion, this time joined by the Court’s liberal wing.
In Texas Department of Housing & Community Affairs v. Inclusive Communities Project, Inc. (ICP), the majority acknowledged the ongoing racial stratification that existed in housing and confirmed the ongoing viability of disparate claims under the Fair Housing Act (FHA) as a way of addressing it.
Citing Parents Involved, the majority confirmed that although “automatic or pervasive injection of race into public and private transactions covered by the FHA has special dangers, it is also true that race may be considered in certain circumstances and in a proper fashion.”
Speaking in antibalkanization terms, the majority warned against “reduc[ing] homeowners to nothing more than their race” while confirming that “foster[ing] diversity and combat[ting] racial isolation” were valid government interests.
Although Justice Kennedy has retired, the doctrines laid down in Ricci and ICP remain good law, and, in that sense, his antibalkanization reasoning remains persuasive.
More importantly, language in Students for Fair Admissions v. President & Fellows of Harvard College (SFFA) may indicate that antibalkanization continues to motivate some members of the Court, namely Chief Justice Roberts.
Indeed, much of the opinion’s qualms with the universities’ affirmative action policies sound in the same racial essentialism that concerned Justice Kennedy.
For Chief Justice Roberts, assigning preferences on the basis of “race qua race” reduces individuals to a racial category, mirroring Justice Kennedy’s concern in Parents Involved.
In a tone akin to Justice Kennedy’s, Chief Justice Roberts emphatically asserted that universities must treat applicants as individuals.
And Chief Justice Roberts’s problem with the University’s stated interest in diversity was not so much that diversity is not a valid governmental interest but that the racial categories used to achieve that interest devolved into racial essentialism.
What Chief Justice Roberts declined to say is important, too. The majority opinion in SFFA did not explicitly overrule Grutter v. Bollinger or Fisher v. University of Texas, the cases that had previously upheld but narrowed race-conscious higher education admissions.
Indeed, the opinion does not go so far as to say that no compelling interest could justify an affirmative action program.
Were the decision motivated solely by anticlassification theory, its holding could have been much broader, in line with the concurrences.
Instead, Chief Justice Roberts is careful to note that “nothing in this opinion should be construed as prohibiting universities from considering an applicant’s discussion of how race affected his or her life.”
This caveat fits within the antibalkanization framework, implicitly recognizing that students grow up in a racialized society and therefore will have racialized experiences that universities may properly credit in the admissions process.
Combined with the retention of diversity as a potentially compelling interest, this strikes some balance—however skewed toward anticlassificationist sentiment—which allows universities to achieve a certain racial integration while avoiding the Court’s perceived constitutional sin of racial essentialism.
In any case, the point is not so much that antibalkanization is a concrete set of doctrines that has survived SFFA. Rather, the values of antibalkanization, as evidenced by Chief Justice Roberts’s discomfort with his colleagues’ fulsome attack on affirmative action,
could plausibly endure as a motivating factor in antidiscrimination law. Just as these values have led the Court to place limitations on racial antidiscrimination doctrine, they may also motivate limitations in the religious discrimination context.
III. Antibalkanization’s Resonance in St. Isidore
This Part explains how antibalkanization values resonate in the free exercise context. It argues that these values support the outcome in St. Isidore and the vitality of the public school exception in future cases. St. Isidore asked whether states must include private religious schools in their system of public education.
Antibalkanization values counsel a resounding no.
Antibalkanization supports an approach to antidiscrimination that treats outright suspect classifications with strict scrutiny but allows compelling interests that can be justified on social cohesion grounds.
This logic extends with equal force to the religious context. Religion, like race, has been the locus of significant historic discrimination,
and therefore direct religious classifications merit the highest scrutiny.
Yet, also like race, overt government entanglement with religion risks entrenching existing divisions along religious lines. That high risk supports the public school exception articulated in Espinoza and Carson.
As a general matter, concerns over religion’s divisiveness have a long historical pedigree, and the Founders, while seeking to ensure religious freedom, were deeply concerned that religion would tear the fledgling nation apart.
James Madison warned that the American public’s “zeal for different opinions” would “divide[] mankind into parties, inflame[] them with mutual animosity, and render[] them much more disposed to vex and oppress each other.”
Thomas Jefferson espoused deep concerns with compelling public funding of religious endeavors, calling it “sinful and tyrannical.”
These fears are the same that motivate antibalkanization: Too great an entanglement with religion would “destroy [the] moderation and harmony” among America’s “several sects.”
In the public school context, this risk is at its highest because schooling involves the inculcation of moral and civic virtues.
If states are required to fund private religious schools as they do public schools, taxpayers are compelled to fund training children in ideas that they may vehemently oppose. Moreover, taxpayers are forced to subsidize religious flight from traditional public schools.
As it stands today, parents are disincentivized from placing their children in religious private schools because doing so means they miss out on the subsidy provided by state-funded public education.
Requiring states to fund religious schools removes this incentive structure and would lead many parents to enroll their children in schools aligned with their religious preferences. The result is a public school system stratified along religious lines, with traditional public schools drained of their religious diversity and a cadre of denominational schools that fail to expose their students to alternative viewpoints.
Secular education became a lodestar of American public schooling precisely to mitigate these divisive effects. From the beginning, Thomas Jefferson “opposed sectarian instruction at any level of public education.”
That is why Jefferson campaigned to establish a nonsectarian system of public schooling in Virginia from before the Founding until his death in 1826.
Following Jefferson’s lead, the vast majority of states, from 1818 to 1971, adopted some limitation on allowing public funds to go to religious education.
By the time the Fourteenth Amendment was adopted, extending the Free Exercise Clause’s protections against the states, twenty-three of the thirty-seven states had adopted some form of prohibition on funding specifically in religious schools.
After the Fourteenth Amendment was adopted, Congress itself required nonsectarian public education as a condition of statehood for twelve states.
The Court has routinely credited that history and the risks of religious anomie in its Establishment Clause jurisprudence.
As Justice Stephen Breyer put it, “The Religion Clauses were written in part to help avoid that disunion.”
“[S]ecular public school,” as Justice Frankfurter opined four decades earlier, was designed to bolster the clauses as “the means of reconciling freedom in general with religious freedom.”
The schools were to be free from religion, not out of religious animus, but because of “the need of a democratic society to educate its children, insofar as the State undertook to do so, in an atmosphere free from pressures in a realm in which pressures are most resisted and where conflicts are most easily and most bitterly engendered.”
As Breyer recognized, with “well over 100 different religious groups” in the United States today, the “risk of religiously based strife, conflict, and social division” has only grown.
Religious entities’ favored status in general antidiscrimination law exacerbates social division even further. Title VII already allows religious entities to discriminate in hiring based on religion,
and the Court’s recent broadening of the “ministerial exception” suggests that religious entities may freely discriminate along suspect class lines in the hiring and firing of instructors teaching secular topics.
In the religious education context, the teacher workforce may thus reflect religious schools’ preferences and prejudices, allowing religious schools to limit along suspect-class lines both the beliefs and the types of people to which students are exposed.
That portends suspect-class isolation in the very institutions meant to develop citizens equipped to participate in a robust, pluralistic society.
If significant social cohesion concerns are at play in the religious schooling context, how might antibalkanization be applied in St. Isidore? At the outset, antibalkanization could still vindicate the history of religious discrimination by calling for the application of strict scrutiny on the basis that the state relied on a religious classification, much as it did in Parents Involved.
At the same time, treating social cohesion in public schooling as a compelling interest provides a normatively coherent and historically sound basis for upholding the public school exception already articulated in Carson and Espinoza. This avoids the line-drawing problem emphasized in Justice Gorsuch’s Trinity Lutheran concurrence because Justices motivated by antibalkanization “differentiat[e] between government policies that entrench and repair . . . inequality” and “assess[] the constitutionality of government action by asking about the kind of polity it creates.”
Antibalkanization therefore looks to the nature of the public benefit only in so far as restricting access to that benefit affects the polity.
Whether Oklahoma’s charter schools are funded directly or through contract, funding religious schools risks the divisive results described above: taxpayers forced to fund a balkanized school system that limits the ideas and individuals to which impressionable students are exposed.
Avoiding a public school system divided along religious lines is thus precisely the sort of compelling government interest that the antibalkanization view credits. And just as antibalkanization “permit[s] and sometimes encourage[s] government to act in ways that promote racial integration,” so too can it permit government to act to ensure religious integration.
Vindicating this interest through an antibalkanization frame would also preserve the holdings of the Trinity Lutheran line and can explain why the asserted interests in those cases failed. Trinity Lutheran could not have been justified by a compelling interest in social cohesion because there is minimal risk of exacerbating religious divisions by subsidizing playgrounds used by religious organizations.
Carson and Espinoza could have asserted those interests, and perhaps implicitly did, but, as Chief Justice Roberts pointed out, their programs were fatally “underinclusive.”
This is because social cohesion interests apply with less force when the state has already accepted some amount of balkanization by subsidizing private education.
Oklahoma did not face this problem in St. Isidore because its charter program is part and parcel of its public school system.
Oklahoma approves its charter schools’ curriculum, requires that the schools are open to all students, and audits the public charter school funding just as it does for traditional public schools.
Oklahoma’s program could thus survive narrow tailoring because it requires charter schools to retain the features of public schools which ensure a socially cohesive environment.
Conclusion
As Justice Kennedy put it in Parents Involved, the “[n]ation has a moral and ethical obligation to fulfill its historic commitment to creating an integrated society that ensures equal opportunity for all of its children.”
If the Founders and nearly two hundred years of American history are to be believed, a school system divided along religious lines can only produce social division and anomie. In St. Isidore—and the cases that will inevitably bubble up in its wake—that is a result that antibalkanization cannot tolerate.