The Rise and Fall
of Church-State Separation
Photographs in the Carol M. Highsmith Archive, Library of Congress, Prints and Photographs Division.
This article is part of our “Kennedy, Carson, and Dobbs: Law and Religion in Pressing Supreme Court Cases” series.
If you’d like to check out other articles in this series, click here.
For the past 60 years, American education has been governed by a policy of “strict separation” of church and state that formally disallowed both the teaching of religion in public schools, and public funding for religious schools. Like all policies, strict separation was a political settlement that emerged thanks to a strategic campaign by separationist actors in the years immediately after World War II. Motivated by a vision of public institutions that protected religious rights by vigilantly guarding against religious favoritism, these separationists took advantage of the accessible and receptive nature of American courts in the postwar era to judicialize, and thereby nationalize, a policy on religious education that until that time had been fragmented across a patchwork of state and local jurisdictions.
This summer, the United States Supreme Court took a sledgehammer to what remained of strict separation. In Carson v. Makin, the Court ruled that a Maine program that paid for residents of rural communities without their own public school to send their kids to non-sectarian private schools fell afoul of the First Amendment’s Free Exercise Clause. Maine could not single out religious schools for exclusion from this benefit, the court reasoned, and as such any voucher program that made funds available to private schools in general needed to make them available to religious private schools. In other words, voucher or tuition benefit programs must now be made available to religious schools or fall afoul of the Constitution.
Then in Kennedy v. Bremerton School District, the Court held that a school district could not prevent a football coach from praying on the 50-yard line of a football field immediately following a football game. That coach was engaged in “private prayer” outside of his “official duties,” and was thus not bound by the constitutional restrictions that otherwise hold that teachers and coaches are presumed to be state actors while on the job.
Although the Court presented these decisions as narrowly tailored to the peculiarities of each case, their implications are much broader. They are consistent with a long-term trend in Supreme Court jurisprudence that reflects a profound shift in the politics of religious education. For at least the past thirty years, the Supreme Court has been progressively elevating the Free Exercise Clause of the First Amendment, which protects religious freedom, over the Establishment Clause, which prohibits government support of religion. These decisions definitively announce the supremacy of Free Exercise. They effectively open the doors to substantial funding of religious schools (particularly if structured “indirectly”), and to religious expression within the public schools (so long as it can plausibly be construed as “individual”). Strict separation has passed from the scene. The Establishment Clause that sustained it has been effectively marginalized. How did this happen? And what may happen next?
I. The Politics of Religious Education Policy
My first book, Secular Conversions, was a comparative history of religious education policy in the United States and Australia. Shortly after I started working on it, a distinguished senior scholar gave a lecture at Berkeley, where I was completing my Ph.D., and I was asked to escort them to a nearby bar for post-colloquium drinks. As we talked on our walk over, I described my project and my broader interest in the separation of church and state. “Interesting,” they said, in a way that suggested they did not find the topic particularly interesting. “But hasn’t it always been that way?”
Many Americans share my distinguished colleague’s misapprehension that strict separation has always been the law of the land. However, it is (was) a relatively recent phenomenon, dating only to a series of court decisions in the quarter-century after World War II. Prior to that time, the Supreme Court paid almost no attention to the First Amendment’s religion clauses. What existed instead was a widespread (but by no means universal) political settlement that understood “separation of church and state” to permit religion in the public schools while prohibiting the expenditure of public funds on religious schools. This settlement was consistent with the preferences of the Protestant Establishment that dominated American politics from the late nineteenth through the early twentieth centuries, who wanted their Bible taught in the public schools and wanted to prevent public funds from flowing to Catholic religious schools.
The emergence of strict separation was a political accomplishment that reflected the changing religious politics of the mid century. As the Protestant Establishment began to fracture, a new coalition of civil libertarians, Jews, and Protestants came together to advocate for a vision of church-state separation that was much more absolute. This coalition strategically targeted the courts, ultimately succeeding in engaging the Supreme Court and getting it to accept its interpretation of the Constitution’s Establishment Clause as prohibiting both religion in the public schools and public funding for religious schools. Because their opponents were not mobilized for legal battles, separationists had a decided advantage in the courts. They thus had a good 30-year run of continually expanding separationist jurisprudence, by carefully selecting cases and recruiting sympathetic plaintiffs to maximize their chance of creating lasting precedents. In so doing, they elevated the Establishment Clause to a preeminent position in American jurisprudence.
Since the 1970s, however, America’s religious politics has changed. The longstanding Protestant-Catholic divide that structured American politics has dissolved, replaced by an conservative-progressive divide that spans religious traditions. Today, a coalition of evangelical Protestants, conservative Catholics, and Orthodox Jews make common cause against strict separation, while mainline Protestants align loosely with liberal Jews, liberal Catholics, and the religiously unaffiliated. The conservative coalition is now fully organized and very active. This is particularly so in the legal arena, where the demise of strict separation is largely the handiwork of an organized campaign by a bevy of anti-separationist groups. Like their midcentury separationist predecessors, these groups have organized to promote religious speech, an expansion vision of religious freedom under the Free Exercise Clause, and public aid to religious groups. Also like their predecessors, they have carefully selected test cases and built their own set of precedents to bring us to this moment. By contrast, the separationist coalition has atrophied as its religious members, especially mainline Protestants, have declined steeply in both membership and in the resources they devote to the defense of separationism. Newer atheist groups have taken up the separationist cause, but often without the same strategic sensibility of either the early separationists or today’s religious conservatives.
The demise of strict separation and the subordination of the Establishment Clause did not just happen; like the rise of separation, its fall is the product of legal mobilization. Anti-separationism has been a prominent if often-overlooked theme of religious conservatives for decades, and of the contemporary Republican Party and the Christian nationalists who dominate it today. Their takeover of the courts has been a top priority for the past twenty years, and it has been accompanied by efforts to install sympathetic officials and policies at the local and state levels at the same time. That this Court, installed by the leaders of this political movement, would act to eviscerate the Establishment Clause, is not at all surprising; indeed, it is precisely what they have been promising to do for years.
The broader politics and institutional positioning of the conservative anti-separationist coalition made some version of this week’s rulings inevitable. There are some interesting, and very American, wrinkles in how the Court has decided to approach these questions, however. They have not demanded that states provide direct funds to religious schools or overturned precedents barring teachers from leading prayers directly. Instead, they have engaged in creative boundary-work around state action to create the illusion of private choice even in the most statelike activities.
II. Schrödinger’s State
One of the jurisprudential through-lines connecting these cases is the Court’s ability to make the state appear and disappear at will. Unlike the Free Exercise Clause, which is (for now) understood to protect individual rights, the Establishment Clause only prevents the state from taking actions “respecting an establishment of religion.” For an Establishment Clause claim to be maintained, one must demonstrate that “state action” — that is, an action by some arm of the state — is at issue. One important move in strict separationist jurisprudence was to recognize that public schools were, in fact, state actors subject to the restrictions of the Establishment Clause.
For those seeking closer ties between church and state, one way to minimize the power of the Establishment Clause is therefore to reduce or deny the presence of state action. In both Carson and Kennedy, this is exactly what the Court has done. Carson is only the latest in a line of cases that asserts that funding schemes that operate through the “private choice” of individuals cannot run afoul of the Establishment Clause. This line begins with Zelman v. Simmons-Harris (2002), which upheld voucher programs on the grounds that, once money had been disbursed to individuals, they were free to spend them on whatever kind of school they so chose. Similarly, in Carson, the Court has declared that “a neutral benefit program in which public funds flow to religious organizations through the independent choices of private benefit recipients does not offend the Establishment Clause” (10).
State boundary-work of this sort has a long history in the church-state arena. Some of the earliest Establishment Clause cases, such as Cochran (1930) and Everson (1947), upheld subsidies for library books and bus transportation on the “child benefit” theory, whereby state funds were understood to flow to individual children, and not to the religious schools they were attending. Later, federal aid to education only passed after a last-minute reworking of the Elementary and Secondary Education Act (1965) that allowed some assistance to Catholic Schools to flow via a scheme whereby public schools would “loan” instructional materials to Catholic schools while retaining their title.
For the past thirty years, vouchers, and other kinds of schemes that deliver public money to religious schools through the mechanism of individual choice, have become the preferred means of avoiding the restrictions of the Establishment Clause. In addition to voucher programs like the one at issue in Zelman, and tuition assistance schemes like the one at issue in Carson, the Court has also upheld tax scholarship foundations, whereby individuals receive a tax credit for donating to scholarship funds which may then issue scholarships to students to attend a religious school. In a related vein, charter schools, which are publicly chartered but privately operated, hold a particularly ambiguous status, with courts ruling that they can be simultaneously public and private depending on the question at hand. All of these mechanisms blur the boundary between public and private, state and society, in ways that allow for creative evasion of the Establishment Clause.
In the Kennedy decision, the Court similarly creatively redrew the boundaries of the state. The question was whether a public school employee, a football coach, was praying on his personal private time, or in his role as a state employee. The Court found that praying on the football field immediately after a football game was private conduct because it did not occur within “his duties as a coach” (3). As the dissenting justices observed, this was a fairly creative interpretation of what constituted his duties, which formally included supervising students “until the students were released to their parents or otherwise allowed to leave” (dissent, p. 3). Nevertheless, the Court equated public postgame prayer with students with “praying quietly over…lunch in the cafeteria” (18). What is notable here is that an acknowledged state actor was deemed not to be a state actor for Establishment Clause purposes at a particular time and place. In other words, state actors can go in and out of stateness over the course of the day, subject to the interpretation of the courts.
In sum, to evade the Establishment Clause and expand Free Exercise, the courts have created a framework wherein public institutions like charter schools may be considered part of the state for some purposes but not others, and where public officials may be considered part of the state for certain parts of their day but not others. Like Schrödinger’s cat, the institutions and actors in question are simultaneously part of the state and set apart from it. Only the courts, it seems, can tell us if the state is alive or dead.
III. The Specter of Religious Conflict
Justice Breyer’s dissent in Carson emphasized the importance of judiciously using church-state jurisprudence as a means of “avoiding religious strife” (dissent, p. 5). In so doing, he was extending a long history of concern with religious conflict that extends back at least to Madison’s Memorial and Remonstrance, and which was a major theme in important early separationist decisions like Everson and Schempp (1963). Popular commentary in the wake of the decisions has also alluded to the risk of conflict among America’s multitude of religious denominations in the context of a lowered wall. How concerned should we be about religious conflict in the schools, in light of these cases?
The Australian experience suggests that concerns over conflict around voucher programs are probably overblown. In Australia, the introduction of direct forms of assistance to religious schools — which are more generous than those yet considered by the Supreme Court — led even those denominations which had been opposed to the reintroduction of “state aid” to scramble to secure checks for their own schools. The bigger danger of the Carson decision, to me, seems to be not religious conflict but the threat it poses to the public schools. In its decision, the Court expanded a line of argumentation begun in Trinity Lutheran (2017) and Espinoza (2020) that declares that any program of grants or subsidies made generally available to nonstate actors must be made available to religious groups on the same terms. This clearly opens the door for more direct grant-based programs to private schools of all kinds, the majority of which would flow to religious schools. So long as they are made available to all private schools, they could not be denied to religious schools. It seems to me like it is only a matter of time before some enterprising legislature decides to enact such a program. But the risk here would be primarily to the integrity and mission of public education, which will almost certainly face lower enrollments and fewer resources as a result of such programs — as has happened in Australia since the introduction of state aid — rather than of an explosion of religious conflict over government funding.
There is much greater potential for conflict in the schools as the result of the Kennedy decision. The risk of an expansive Free Exercise jurisprudence is that, eventually, the protection of one individual’s religious freedom claims will come into conflict with those of others. We have seen this already in cases using the Free Exercise Clause to permit religious adoption providers to discriminate on the basis of their religious beliefs: although the case in question (Fulton v. Philadelphia, 2021) centered on whether religious providers were allowed to discriminate against LGBT parents, other Christian providers have used the case to defend their unwillingness to broker adoptions for Jewish parents. The religious freedom of the Christian organization, so understood, directly impinges upon the religious freedom of such parents. When it comes to schools, as the dissenting justices make clear, freedom for teachers to pray inevitably raises concerns about coercion and their students’ right to pray (or not) as they wish.
It is thus not hard to see how the Court has opened a Pandora’s box of potential free exercise claims springing from teachers and coaches praying in their “down time.” Will teachers recite the Lord’s Prayer at their desk “privately” at the start of a period after students have arrived but before “official” instruction begins? Can a guidance counselor set up a table in the cafeteria to pray “privately” in front of students during their lunch break? Will students feel coerced by such prayers and sue to protect their own religious rights? The Kennedy decision downplays the coercive potential of such actions, and the Court’s rulings to date suggest they would be more receptive to the claims of the teachers. But what would happen in a predominantly Christian community where a Muslim teacher prays publicly between periods or over her lunch break, for instance? Would the Court be as solicitous of her claims?
The Kennedy decision suggests that the answer here would likely be “no.” An important aspect of Gorsuch’s opinion is its call for a new set of tests to determine whether religious expression falls afoul of the Establishment Clause. Both the half-century-old Lemon test, where a law’s purposes, effects, and potential for entanglement are considered in determining questions of establishment, and more recent coercion tests, where a practice’s potential for coercion are centered, are to be abandoned. In their place is a new test that focuses on “historical practices and understandings” (23). This move elevates recent decisions such as Town of Greece (2014) and American Legion (2019) which upheld public prayers and religious symbols on public lands on the grounds that they constituted historical traditions that should be understood as forms of cultural heritage rather than as religion. There is clearly a tension — probably better understood as a maneuver — here, in that these earlier decisions tried to deny that the practices in question were actually religious, whereas the Kennedy case affirms their religious character but insinuates that their historical existence should be seen as justification for their permissibility.
It seems entirely likely to me that all manner of public religious expression will eventually be permitted by being grandfathered in as “traditional.” Notably, however, Christian expressions and symbols will have far greater access to these claims than others — which is likely part of its appeal to this Court. The “culturalization” of religion, whereby courts treat religious expressions as a “cultural tradition” that is part of a nation’s “heritage,” and thus beyond the scope of disestablishment provisions, is an increasingly common move across Western courts who seek to maintain the appearance of religious neutrality while finding ways to privilege dominant religious traditions. The American Supreme Court is a little behind the curve here, but catching up rapidly. Given the Court’s reading of the facts of the Kennedy case, it seems unlikely that future decisions will attend to the true historical complexity of America’s religious heritage, rather than adopting the “Christian nation” mythology so central to the modern Religious Right. Viewed through that lens, the claims of non-Christian — perhaps even non-conservative-Protestant — religious groups to religious rights are likely to be severely hampered. The Muslim teacher praying at lunch is likely to be seen as “innovating” in an unprotected way, whereas her Christian colleague’s prayer is protected as “traditional.”
IV. The Establishment Clause in Exile
For those who favor a stronger separation of church and state, the road back will be long — if it is even possible. It took 60 years for Christian conservatives to complete their takeover of the courts and the marginalization of its Establishment Clause jurisprudence. It will likely take at least as long, possibly longer, to fight back. The coalition dedicated to an expansive Free Exercise interpretation of the First Amendment is strong and has a robust legal infrastructure it can use to enforce its new jurisprudence. The political party that is its vehicle has a growing institutional advantage in the American political system and a worrisome willingness to constrain democratic participation. Meanwhile, it is unclear what the political coalition in favor of separationism looks like. Its mainline Protestant champions are in terminal decline, while the religious minorities and “religious nones” who might otherwise form part of the coalition are not yet fully mobilized around this issue. If court decisions reflect political battles as played out in particular institutional systems, a revitalized Establishment Clause faces extremely long odds.
Above all, separationists need to recognize that the courts are now hostile territory. Sixty years of separationism has led to a reliance on the courts to enforce separation at the expense of a broader vision of politics. If it were not clear before this week that this is no longer a tenable strategy, there should be no illusions about that now. The good news is that the courts are not the only place where political fights for disestablishment can play out. Strict separationists triumphed by shifting battles over religion and education to the federal courts in the mid-twentieth century because it was an available and welcoming institutional venue for pressing their claims. In light of recent cases, it is time for separationists to shift to other political venues.
This is not to say that the courts should be ignored altogether. Indeed, the triumph of religious conservatives occurred because they engaged in political struggles across multiple institutional fronts — from school boards to town councils to legislatures to regulatory agencies to courts. And while separationists should engage along all these fronts, they should also engage with the courts in a strategic fashion. The conservative legal movement did not develop its free speech and free exercise arguments immediately. Rather, it experimented with a variety of challenges to separationist jurisprudence and pursued the avenues that forced the courts to acknowledge inconsistencies and contradictions within that jurisprudence. Strict separationist jurisprudence reached its zenith in the early 1970s, but thereafter its limits and internal tensions became increasingly clear. It is not clear that we have yet reached the zenith of Free Exercise jurisprudence, but as I noted above, some clear tensions have already emerged around questions of coercion and the difficulty of reconciling competing religious freedom claims. Finding cases that magnify these concerns and bringing them before the courts seems like a natural first step to forcing the court to rein in its more expansive religious freedom arguments.
Sixty years of separationist jurisprudence has also, I think, led to a general acceptance of the idea that courts should be given broad deference on questions pertaining to religious rights. By contrast, and particularly in situations where religious freedom claims have been overextended, courts ought to be understood as but one important player within a larger ecosystem of venues where decisions about how to balance religious freedom claims versus other interests are hashed out. Separationists can and should seek out friendlier venues where they can push back against the courts by passing laws and policies that articulate the appropriate limits of free exercise. Laws and policies prohibiting compelled worship, for instance, would be a logical starting point.
Another straightforward response should be to make the state as unambiguous as possible. As I discuss above, both of these decisions rely on the ambiguities presented or produced in what counts as state action. Eliminating voucher programs, and passing laws and policies that assert that public funds may only go to public schools and that charter schools must be understood to be public institutions, for instance, would all make clearer what counts as the state and what does not — thereby making it more difficult to bypass the Establishment Clause via “individual choice” arguments. Failing that, legislatures could insist that any school that receives government funding, either directly or indirectly, must be open to all students, just as public schools are. In addition to reducing possibilities for religious discrimination, such laws would also reduce the potential for “creaming” that risk downgrading public schools into an educational institution of last resort. Similarly, states and school boards can and should clarify what counts as the duties of teachers, coaches, and administrators, what counts as “down time” within those duties, and where such “down time” may occur. The Court has currently given itself carte blanche to make these decisions, but clarifying these policies in writing has the potential to constrain it.
All of these strategies require cultivation and use of venues other than the courts. They reject the idea that courts alone get to determine what passes for policy, and assert that other institutions have a rightful say in making political decisions. Yet should the Court reject these relatively mild attempts to counter the framework it has asserted, it should be remembered that more aggressive forms of dejudicialization exist. The composition of the court can be altered through expansion or contraction. Its jurisdiction and even its power of judicial review can be curtailed. And political actors can always refuse to comply with the court’s ruling.
It seems unlikely to me that separationists can muster the political capital needed to do these things on their own. However, the breathtaking range of transformations that the Supreme Court has undertaken in its current term appears to have whetted an appetite for stronger forms of dejudicialization among a larger segment of the left. To the extent that stronger forms of dejudicialization become politically feasible, separationists could benefit by joining a coalition that seeks to make a hostile Supreme Court a less powerful player in the political ecosystem. There are serious risks to the rule of law that attend these kinds of dejudicialization, but the Court is already in the middle of a stark legitimacy crisis. What is clear is that, absent any action along these lines, the Establishment Clause has been sent into deep exile, and will be for some time. ♦
Damon Mayrl is Associate Professor of Sociology at Colby College, where he studies the intersection of religion, politics, law, and culture. He is the author of Secular Conversions: Political Institutions and Religious Education in the United States and Australia, 1800-2000 (Cambridge University Press, 2016). He can be found on Twitter @yrlsoc.
Mayrl, Damon. “The Rise and Fall of Church-State Separation.” Canopy Forum, October 13, 2022. https://canopyforum.org/2022/10/10/the-rise-and-fall-of-church-state-separation/