Our thesis is that the degree of permissible interaction between government and religion and its policy implications is most strikingly documented in the educational arena. To support our claim, we make three arguments: (1) by the early 1970s, New York, Pennsylvania, Rhode Island and other states were experimenting with accommodationist policies in education that neither increased religious divisiveness nor denied religious freedom. These policy experiments contributed to the Supreme Court’s new understanding of religious accommodation under the First Amendment; (2) beginning in the late 1970s, a scholarly reappraisal of the establishment clause – based on a close reading of the debates over the First Amendment, in Congress and the states – was demonstrating that reasonable governmental accommodations, particularly in education, would not violate the establishment clause; and (3) that periodizing the establishment clause’s history, by focusing on state-level innovations and scholarly reinterpretations, enhances the understanding of the policy developments of the First Amendment, particularly in the educational field.
Although we highlight the constitutional viability of educational accommodation, others argue that church–state entanglements endanger foundational constitutional values. Isaac Kramnick and R. Laurence Moore believe that strict separation between church and state protects religious liberty, fosters pluralism, and limits government intrusion in the public sphere.Footnote 1 In particular, they emphasize Thomas Jefferson’s and James Madison’s support for disestablishment while downplaying federalism and Anti-Federalist influences on the First Amendment. Leonard Levy is also a separationist, but he places the separationist–accommodationist debate within democratic and federalist contexts, noting that not all government interactions with religion are unconstitutional.Footnote 2 Michael Malbin, however, sees in founding-era debates greater accommodation. He suggests that the original intent of the First Amendment would have allowed federal aid to religious schools, although he thinks that because of growing religious diversity, school prayer might be “inconsistent”Footnote 3 with the original intention.
The divide over separation and accommodation is certainly intractable, but it is not insuperable. In Everson v. Board of Education (1947),Footnote 4 the Supreme Court upheld a New Jersey law authorizing reimbursement for transportation costs to parents who sent their children to parochial schools. Seemingly accommodationist, a strong separationist line of thought permeates the decision. “The ‘establishment of religion’ clause of the First Amendment means at least this,” Justice Hugo Black wrote in Everson: “Neither a state nor the Federal Government can set up a church. Neither can pass laws which aid one religion, aid all religions, or prefer one religion over another.”Footnote 5 Everson created the paradigm for denying state aid to religious institutions, with ramifications to this day. But recent scholarship has challenged that idea. “[C]hurches and schools are not the same,”Footnote 6 Douglas Laycock writes, and the varied perspectives that constitute the accommodation-separation divide suggest room for state-level approaches that encourage religious pluralism—especially in education—without violating the establishment clause.Footnote 7 Accommodation does not involve compulsion or government endorsement of a belief system. Rather, it maintains a fair environment that respects the deeply held convictions of all citizens. In this policy domain, the core principles of American constitutionalism—federalism and the incremental nature of American political development as well as the various manifestations of the culture wars over religion and education that have punctuated American history—come into focus.Footnote 8
How do they come into focus? Supreme Court justices from the New Deal to the Burger Court failed to see them. In Wallace v. Jaffree (1985), the Burger Court held that Alabama violated the First Amendment’s establishment clause by authorizing “teachers to lead ‘willing students’ in a prescribed prayer to ‘Almighty God … the Creator and Supreme Judge of the world.’”Footnote 9 Yet prayer in schools existed before the Civil War and for one hundred years after the passage of the Fourteenth Amendment.Footnote 10 Since the founding of this country, both houses of Congress have begun their day with prayer and most states and some localities either have or had a paid chaplain or paid one on a per diem basis to lead their legislatures in prayer.Footnote 11
In his dissent in Wallace, Justice William Rehnquist conducted a historical analysis of the founders’ intentions that reshaped the Court’s post–World War II understanding of the establishment clause.Footnote 12 He contended that an approach emphasizing state autonomy within the federal system would give states more room to accommodate religious expression without breaching constitutional constraints. For Rehnquist, the establishment clause addresses a federalism problem because it is constitutionally difficult to assert a “personal right of nonestablishment.”Footnote 13 In our view, Rehnquist’s dissent breaks with Everson’s logic and ushers in a new era in church–state accommodation with respect not only to sectarian institutions but also within public institutions.Footnote 14 By highlighting the incremental nature of American public policy, long thought of as a central paradigm for innovation and political development,Footnote 15 Rehnquist’s dissent provided parties, legislatures, and courts with a historical and constitutional justification to continue experimenting with state aid to education.
Periodization and the Problem of History
Periodization’s purpose is to isolate substantive changes in policy history that develop over time. Periodization reduces the desire to see history as positivistic, emancipatory, and universalFootnote 16 and instead forces a greater emphasis on the role of social relations, institutions, and political interactions among state actors and interested parties in the development of ideas and policy. Periodization, Martin Sklar writes, “imposes a discipline upon inquiry that acts as a control against the presuppositions, the fashionable interpretations, as well as the appealingly irreverent ones.”Footnote 17 Byron Shafer argues that “Cutting political history into periods makes us distinguish between greater and lesser influences on the ongoing structure of politics. It is the stability of these influences, and then their change, that demarcates periods.”Footnote 18 Rather than emphasizing a seamless web of ideas and practices that extend throughout history, periodization creates a “dialogue”Footnote 19 between past and present—as Rehnquist’s dissent does—highlighting path dependencies in law and history.Footnote 20 Periodizing the establishment clause recovers the role of the states and of political culture as integral parts of the First Amendment’s meaning.
Our primary focus is the second era of the establishment clause. In this section, we offer a brief overview of how to periodize First Amendment’s establishment clause. The following section will explore alternative approaches to periodization and explain why we believe our framework best captures how Rehnquist’s dissent reshaped the Court’s establishment clause jurisprudence.
The first era extends from the founding to 1947. During this era, federalism was the dominant constitutional principle delineating the degree of interaction between church and state. Federalism is explicit in Article VI of the Constitution, which proscribes religious tests for federal office. Implicit in this formulation is that states were free to stipulate adherence to Protestant Christianity or to Christianity in general for office holding or have no qualifications at all. Some states had religious tests until 1961.Footnote 21 The states could also establish a religion and require a tax to support it, as Connecticut did until 1818, New Hampshire until 1819, and Massachusetts until 1833. The demise of state-supported religion is the result, Steven Green writes, of fierce partisan battles, the weakening of the Federalist Party in the Jeffersonian era, and increasing secularization.Footnote 22 The states could also provide aid to parochial schools (or to the parents of children in parochial schools), require prayers and Bible reading in public schools, and have Sabbath and blasphemy laws.Footnote 23 “Sabbatarianism and Sunday mail delivery punctuated the public discourse”Footnote 24 long after the Fourteenth Amendment was ratified.
The second era extends from 1947 to 1985. Until Everson, establishment meant “a formal, legal union of a single church or religion with government.”Footnote 25 To erect a high wall between church and state, Justice Black, the author of the Everson majority opinion, relied, in part, on President Jefferson’s 1802 letter to the Danbury Baptists, which has become the “authoritative declaration of the scope and effect of the amendment.”Footnote 26 In the letter, Jefferson wrote, “I contemplate with sovereign reverence that act of the whole American people which declared that their legislature should ‘make no law respecting an establishment of religion, or prohibiting the free exercise thereof,’ thus building a wall of separation between church and State.”Footnote 27
It is now clear that Black placed too much emphasis on this letter to determine the amendment’s meaning. First, the Danbury Baptists were Jefferson’s allies in seeking to disestablish the Anglican church in Virginia and elsewhere. Jefferson wasn’t saying anything the Baptists didn’t already believe. Moreover, as Vincent Phillip Muñoz has demonstrated, in a careful reading of Jefferson’s correspondence with the Danbury Baptists, Jefferson’s letter to the Danbury Baptists mostly recapitulates the Danbury Baptists’ original letter to Jefferson.Footnote 28 And second, as Donald Drakeman has shown, Jefferson’s letter had no effect on religion or politics until the late nineteenth century, when changing configurations of religious diversity became politically salient.Footnote 29
The shift toward a stricter separationist logic became particularly pronounced during the last quarter of the nineteenth century, when some states adopted “Blaine amendments” that prohibited aid to private (religious) schools.Footnote 30 Everson builds on that legacy and reconfigures establishment as a limit on state-level religious accommodation. Everson, therefore, marks the beginning of an era in which the Supreme Court will prohibit aid to parochial schools and further secularize public schools. After Everson, American religious history was to be understood in the service of liberal secularism.Footnote 31 We maintain, however, that liberal secularism is not the natural outcome of American constitutional history but has a political history that periodization reveals. Underneath the claim of liberal secularism was a lively debate in the states about religious accommodation to public and nonpublic schools.Footnote 32
Rather than focusing on the two staples of the Court’s separationist jurisprudence since Everson—Jefferson’s wall metaphor and Madison’s “Memorial and Remonstrance,” written in 1785 to protest against a provision in the Virginia constitution to pay “teachers of the Christian religion,”Footnote 33 neither of which, one by dint of timing, the other by institutional specificity, significantly shaped the public understanding of the First Amendment at its ratification and for the next three-quarters of a centuryFootnote 34—Rehnquist emphasized two points that the Everson Court overlooked, one general, the other more specific: (1) constitutional doctrine must be grounded in accurate historical understanding and (2) “the true meaning of the Establishment Clause can only be understood through its historical context.”Footnote 35
In returning to history, Rehnquist showed (1) that Jefferson was not a good representative of the sense of the amendment, as he was in France at the time the Bill of Rights were debated; (2) that Jefferson’s “letter to the Danbury Baptist Association was a short note of courtesy, written 14 years after the Amendments were passed by Congress”Footnote 36; (3) that some congressmen, for example, Representative Peter Sylvester of New York, disliked the initial language of the proposed amendment “because it might have a tendency ‘to abolish religion altogether,’”Footnote 37 resulting in at least 25 changes to Madison’s original languageFootnote 38; and (4) that in explaining the history of the religion clauses, the political, cultural, and religious practices of the American people that the Court overlooked in 1947 should matter more than relying on Jefferson’s 1802 letter and Madison’s 1785 memorandum.Footnote 39 In short, Rehnquist’s dissent returned jurisprudence to its origins as a historical endeavor, grounded in the practices of a specific people organized around state and local governance. “[S]tare decisis,” Rehnquist wrote, “may bind courts as to matters of law, but it cannot bind them as to matters of history.”Footnote 40
The idea that history is different from precedent was a novel one at the time Rehnquist dissented in Wallace v. Jaffree. Charles Miller’s The Supreme Court and the Uses of History (1969) shows that “the decline of the historical approach to constitutional law”Footnote 41 was symptomatic of the justices and of scholars in political science and history throughout the 1960s. To be sure, some justices in the second era engaged in historical investigations into religion and the founding.Footnote 42 For example, in Justice William Brennan’s concurrence in Abington v. Schempp (1963), which held the practice of school prayer unconstitutional, Brennan provided a lengthy analysis of church-state relations.Footnote 43 But his methodology was “law-office history,”Footnote 44 an arrangement of preconceived facts to further a legal point rather than an independent and critical investigation into a set of historical events. Brennan’s opinion did not stray far from recent precedent, limiting the historical value of his opinion.Footnote 45
Rehnquist’s more extensive historical analysis demonstrated that the Court’s strict separationist understanding was a historically contingent interpretation of the First Amendment, foisted on the justices by political and social interests operating within an increasingly secular milieu.Footnote 46 Furthermore, Rehnquist revealed that the First Amendment’s establishment clause had a history embedded within state religious practices going back to the founding and even earlier, which formed and informed the state policies that, for example, sent children to school to pray and learn, until the Supreme Court ruled otherwise. Because this aspect of American culture was so deeply ingrained within historical practices, Rehnquist held, it could not be brushed aside by the creation of a novel constitutional interpretation that upended the ways of a large number of people throughout the United States.Footnote 47
The third era begins in 1985, with Rehnquist’s critique of the Court’s historiography of disestablishment, and ends in 1997, when the Court, in Agostini v. Felton, overturned Aguilar v. Felton. Footnote 48 Agostini begins a new era because it is the first case in which the Court overruled an establishment clause precedent.
The fourth era extends from Agostini to the present, with the Court breaking away from Everson’s strict separationist logic and embracing Rehnquist’s understanding of the establishment clause in Wallace. Judicial reinterpretation, of course, does not follow a straight line. As late as 2000, Rehnquist could write that a decision that denied a student-led prayer at a high school football game “bristles with hostility to all things religious in public life.”Footnote 49 Nevertheless, the capstone of Rehnquist’s Wallace dissent is Kennedy v. Bremerton School District (2022). In that case, the Court rejected the Burger Court’s 1971 “Lemon test” as an “ahistorical, atextual” approach to discerning “Establishment Clause violations.” Although the Court did not formally overrule Lemon v. Kurtzman,Footnote 50 the Court replaced the test with an originalist reading of the establishment clause that relies on “reference to historical practices and understandings.”Footnote 51
Periodizing the development of the establishment clause reveals that government accommodation of religion has evolved over time. This evolution has been shaped by shifting notions of pluralism, changing judicial interpretations, persistent state-level innovation, and patterns of immigration. In the late nineteenth century, for instance, state legislatures could not have imagined lending audio-visual equipment to religious schools or using vouchers and tax credits to enhance students’ educational experiences. These post–World War II developments not only reflect state-level legislative innovation, they also challenge the separationist argument that church–state relations have followed a linear path from disestablishment to separationism to secularism.
The Periodization of the Religion Cases Explained
Rehnquist’s break with the Court’s religion cases since Everson was not possible without an informed understanding of changes occurring within the scholarly literature on religion and the founding era.Footnote 52 Attempts to reinterpret the establishment clause appeared as early as 1978, with Michael Malbin’s Religion and Politics: The Intentions of the Authors of the First Amendment. Malbin’s main insight, which Rehnquist takes up implicitly, is that once the establishment clause is understood to be dependent upon an understanding of federalism, then a “return to the original conception … would surely alter the basic view of the relationship between religion and government in a way that would affect future legislation and future court cases.”Footnote 53
In 1982, Robert Cord published an equally revisionist account of the establishment clause. And once again, “Justice Rehnquist’s dissent drew heavily on Cord’s history without citing it.”Footnote 54 Cord concluded that there was no “historical evidence that the First Amendment was intended to provide an absolute separation or independence of religion and the national state. The actions of the early Congresses and Presidents, in fact, suggest quite the opposite.”Footnote 55 Other scholars began reaching similar conclusions. By 1984, William Van Alystyne could write of “Mr. Jefferson’s Crumbling Wall.”Footnote 56
The justices from the New Deal to the Burger Court had access to the sources used by Malbin, Cord, and others, but rarely relied on them.Footnote 57 Yet the congressional and state debates make clear that the establishment clause was neither understood as mandating governmental neutrality between religion and irreligion nor as a bar to federal nondiscriminatory aid to religion. There is no historical foundation for the proposition that the founders intended to build the kind of “wall of separation” the Everson Court constitutionalized. To support this claim, we have periodized the Court’s establishment jurisprudence in light of revisionist insights and enduring state-level practices. Admittedly, other frameworks for understanding the Court’s engagement with policy history—ones that do not rely on periodization—also merit exploration. To clarify our understanding of periodization, we also address how others have periodized establishment cases.
Empirical or attitudinal approaches to judicial policy history emphasize correlations among justices’ policy preferences, their framing of legal issues, and judicial outcomes.Footnote 58 However, such approaches overlook less visible influences on policy change—such as scholarly reinterpretations and state-level political dynamics—that clearly shaped Justice Rehnquist’s Wallace dissent. In contrast, Cornell Clayton and J. Mitchell Pickerill offer a “political regimes” framework, which underscores the Court’s interdependence with Congress and the presidency, and its reliance on evolving “regime values.”Footnote 59 Clayton and Pickerill then situate federalism among a constellation of political considerations the Court confronted during the 1970s.Footnote 60 As Clayton and Pickerill write, when “regime values pass from one institutional context to another, they are … reformulated in ways that may assert an independent influence on other institutions.”Footnote 61 In short, different institutions, and actors within them, conceptualize federalism in distinct ways. This is a compelling argument, and we concur with their portrayal of federalism as an interdependent issue. The “political revival of federalism,” they argue, “was a result of the interaction of multiple forces in American society.”Footnote 62 Our divergence from their framework lies only in the observation that, like empirical approaches, Clayton and Pickerill’s model does not fully account for the specific factors that informed Rehnquist’s federalist reframing of the establishment clause debate in Wallace,Footnote 63 including emerging scholarly interpretations, state-level political developments, the historical trajectory of religious pluralism, and broader currents in political culture. In this context, national political actors—the parties, the president, and Congress—played a comparatively limited role in reshaping the discourse on religion and state aid to education.
Another noteworthy approach to judicial policy change emphasizes the emergence of public-interest law firms and intellectual networks—such as think tanks and other organized groups—that together form what is now known as the conservative legal movement. Three works in particular focus on religious and legal interests: Steven Brown’s Trumping Religion: The New Christian Right, the Free Speech Clause, and the Courts (2002),Footnote 64 Hans Hacker’s The Culture of Conservative Christian Litigation (2005),Footnote 65 and Daniel Bennett’s Defending the Faith: The Politics of the Christian Legal Movement (2017).Footnote 66 And another three ignore the religious Right’s influence and focus instead on secular and conservative interests. These books include Steven Teles’s The Rise of the Conservative Legal Movement: The Battle for Control of the Law (2008),Footnote 67 Ann Southworth’s Lawyers on the Right: Professionalizing the Conservative Coalition (2008),Footnote 68 and Amanda Hollis-Brusky’s Ideas with Consequences: The Federalist Society and the Conservative Counterrevolution (2015).Footnote 69
Each of the three books that examine religion survey the groups and ideas that constitute the “New Christian Right.” Brown finds that “[t]he sole purpose for which they were founded is litigation,”Footnote 70 a point with which Hacker and Bennett agree. By contrast, separationists such as Kramnick and Moore make much of the polarizing influence of religious movements, such as the Moral Majority and the Christian Coalition, on American politics during the 1980s and 1990s. While these interests clearly had political clout (elite access, name recognition, and a strong media presence, for example), Melissa Deckman’s study of school board battles between secularists and the Christian Right describes a decentralized and unorganized Christian movement during the Reagan and Bush presidencies that limited their political effectiveness, locally and nationally.Footnote 71
Consequently, as both Brown and Bennett demonstrate, Christian thinkers and donors came to reject direct political activity in favor of legal strategies. According to Brown, “the involvement of religious conservatives in the political arena had had a corruptive influence on the church while accomplishing relatively little in terms of social change.”Footnote 72 Clyde Wilcox and Carin Robinson similarly observe, “The Christian Right has paradoxically been the most successful social movement in influencing elections and party politics over the past century and the least successful in influencing policy and culture.”Footnote 73 This is so, according to Bennett, because “the Christian Right … came to recognize the value of legal advocacy in pursuing its multifaceted agenda. For social movements, legal advocacy affords opportunities not available in other venues.”Footnote 74
Steven Teles, whose book does not address the Christian conservative legal movement, comes to the same conclusion about secular and conservative political interests.Footnote 75 For Teles, during the 1990s, the conservative legal movement was shifting away from the “demand side” (courts) and beginning to focus on the “supply side” (litigants).Footnote 76 He argues that legal change is not solely the product of judicial decisions but also the result of strategic actions within and around legal and political institutions. Teles situates the emergence of the conservative legal movement within the broader context of the institutionalization of legal liberalism across law schools, government agencies, private philanthropies, and public-interest legal organizations. Accordingly, his analysis offers a comprehensive framework for understanding legal and political transformation, incorporating a range of factors such as ideological infrastructure, policy entrepreneurship, and the long-term strategies of actors operating within a regime characterized by gradual, deliberative change. To support this analysis, Teles draws on historical institutionalism, organizational theory, and the sociology of knowledge. He then explains how conservative elites orchestrated what Ann Southworth describes as the conservative legal movement’s “court-centered strategies.”Footnote 77
As Teles acknowledges, an approach to policy change involving interest groups and electoral changes that alter the makeup of the federal judiciary is no longer satisfactory because of the negligible influence that religious interests have had on the judiciary. He also points to the weakening of parties and the “‘thickening’ of the American political system” under the weight of the growth of the American state, “which has … weakened the mechanisms of disruptive, electorally inspired change.”Footnote 78 Accordingly, we draw on these insights to construct a nuanced interpretation of religious accommodation under the establishment clause that is both politically and jurisprudentially informed.
Finally, we address other periodizing methodologies. John Jeffries and James Ryan’s periodization of establishment cases extends from 1947 to 1996 “because that period covers the great majority of all Establishment Clause decisions and because it stops just short of the first clear signal of change in 1997,”Footnote 79 when the Court decided Agostini v. Felton.
According to Jeffries and Ryan, legal change can best be understood by examining “political contests among various interest groups, both religious and secular.”Footnote 80 We have already shown that recent scholarship casts doubt on the effectiveness of this strategy. At a methodological level, we disagree with this approach because it relies on an insufficiently political understanding of legal development and change. First, by ignoring the literature on pluralism and elite theories of power,Footnote 81 Jeffries and Ryan fail to explain how these theories might alter their more limited understanding of legal change. Second, they assume that the concept of the political is interest-group based only, thereby excluding philosophical principles, cultural factors, and historical experiences that have clearly influenced the Court’s religion cases over time. Consequently, their understanding of legal change is neither sufficiently political nor historical.Footnote 82
Mark David Hall has created a formidable historical database of the Court’s religion cases, although he refrains from explicitly referencing periodization. For Hall, the Court’s establishment jurisprudence extends from Everson to Marsh v. Chambers (1983) and then from Wallace v. Jaffree to Cutter v. Wilkinson (2005),Footnote 83 a case concerning the religious practices of prisoners.Footnote 84 Unlike Hall, we see Rehnquist’s dissent in Wallace as straddling two eras because it is historically more informed than Burger’s majority opinion in Marsh v. Chambers, in which the Court upheld Nebraska’s funding of a legislative chaplain, and Rehnquist’s majority opinion in Mueller v. Allen (1983), in which the Court upheld a Minnesota tax deduction that applied to parents who used the deduction for religious instruction.Footnote 85 Hall acknowledges this, writing that Rehnquist’s dissent drew from more historical sources than any other opinion in Supreme Court history. Rehnquist also relied on the most “diverse group of Founders than any other Religion Clause opinion ever written.”Footnote 86 Significantly, Hall writes that Rehnquist “appeal[ed] to Madison nine different times, but never to his Memorial and Remonstrance.”Footnote 87 Hall’s assessment, although not couched in explicit periodization, affirms our contention that Rehnquist’s Wallace dissent represents an end and a beginning in the Court’s establishment clause jurisprudence.
Steven Green argues that the period from 1940 to 1975 created a “Third Disestablishment” in American politics, “the first being the political disestablishment of the founding period and the second being the legal and institutional disestablishment of the nineteenth century.” According to Green, the third disestablishment involves “the replacement of Protestantism and the struggle between secularism and pluralism,”Footnote 88 resulting in a greater emphasis on secularization in the public sphere. First, we have already argued that the Court’s 1947 Everson decision marks the beginning of the second era. And second, as we will show, the cases the Court decided after 1975, such as Agostini v. Felton and other such cases, challenge the separationist notion of a linear development toward greater separation.
Having examined a variety of interpretations of legal change and establishing the grounds for periodizing the establishment clause with Rehnquist’s dissent at the heart of it, in the sections that follow, we will present a history of the establishment clause against the backdrop of Rehnquist’s revival of federalism and the states’ attempt to come out from underneath Everson’s separationist logic. We begin about midway through the first era because, by the 1870s, there were acute challenges to the meaning of the separation of church and state that had not existed before.
The First Era (1789–1947)
One major effect of Rehnquist’s dissent was to shine a light on the nineteenth-century’s “religious intolerance and conflict”Footnote 89 that informed the Court’s understanding of the religion clauses since Everson. By the last quarter of the nineteenth century, some states with significant Catholic populations began to supply parochial schools with extra funds and secular and dissenting Protestant groups pushed back. In 1876, the Republican party platform called for a constitutional amendment “prohibiting the use of public funds or property ‘for the benefit of any schools or institutions under sectarian control.’”Footnote 90 The next year, President U.S. Grant asked Congress to pass a constitutional amendment to that effect.Footnote 91 Just one week after President Grant’s address, Congressman James Blaine proposed an amendment to the Constitution, the so-called Blaine Amendment, which stated, in part,
[n]o State shall make any law respecting an establishment of religion or prohibiting the free exercise thereof; and no money raised by taxation in any State for the support of public schools, or derived from any public fund therefor, nor any public lands devoted thereto, shall ever be under the control of any religious sect.Footnote 92
The House passed the amendment, 180-7, but the Senate did not, falling “4 votes shy of the two-third necessary for passage.”Footnote 93 Succeeding Congresses, however, required newly admitted states to have Blaine amendments, and some existing states passed their own, setting “rigid standards of separation of church and state.”Footnote 94 By 1890, twenty-nine states had “prohibitions against the transfer of public funds”Footnote 95 to sectarian schools.
The proposed Blaine Amendment shows that the Constitution acknowledged the states’ authority to regulate religion and that the Fourteenth Amendment did not change this original interpretation.Footnote 96 In Roberts v. Bradfield (1898), the United States Court of Appeals for the District of Columbia Circuit made this point explicitly, noting that neither state Blaine laws nor the First Amendment prevented “the incorporation of churches or sectarian establishments,”Footnote 97 a position the Supreme Court upheld the following year. Indeed, a number of pre-incorporation decisions by the Supreme Court indicate that public welfare spending, at the federal and state levels, could be consonant with the establishment clause. Three cases form the unstated background of this era.
Since 1832, Congress had “incorporate[d] in the District of Columbia charitable, educational and religious societies.”Footnote 98 In Bradfield v. Roberts (1899), Joseph Bradfield challenged a congressional grant to Providence Hospital in Washington, DC, run by a Catholic order of nuns, as a violation of the establishment clause. Justice Rufus Peckham, writing for a unanimous Supreme Court, held that the hospital was for “the care of such sick and invalid persons as may place themselves under the treatment and care of the corporation” and that the hospital conducted itself as “a secular corporation.”Footnote 99 Later analysts will note that a hospital is distinct from an educational institution, limiting the ruling’s reach. But the Court in the pre-incorporation era did tackle religious-education issues, failing to find an establishment clause violation in the federal government’s funding of the Catholic Indian Bureau.Footnote 100
Nativism and anti-Catholic sentiment, key elements of religious conflict in the nineteenth century that persisted into the twentieth,Footnote 101 drew the Court’s attention when Oregon attempted to standardize education by requiring all children aged 8 to 16 to attend public schools. The intent of the statute, passed by the initiative and referendum process, was to assimilate all children in the state to a common American culture.Footnote 102
In Pierce v. Society of Sisters (1925), the Court held that the Oregon statute violated the Fourteenth Amendment’s due process clause.Footnote 103 Pierce established the right of nonpublic schools to exist, but it did not clarify the problem of public aid to sectarian schools. Shortly after Pierce, the Court addressed state aid to students in nonpublic schools in Cochran v. Louisiana State Board of Education (1930). The Louisiana Board of Education provided free textbooks to students across the state, public and nonpublic. A unanimous Court noted that the beneficiaries of the aid were not the schools but “the school children.” Finding that the aid was for the “common interest,” the Court noted that the state could not “segregate private schools, or their pupils, as its beneficiaries.”Footnote 104 In cases in which a state provides aid to sectarian schools or to parents of children attending such schools, Cochran suggests that justification might be found through the principles of federalism and the concept of child benefit. The child benefit theory “asserts that certain kinds of public aid, such as school busses and textbooks, are aid to the child rather than aid to the religious school which he attends.”Footnote 105 Although Everson partly adopted the child benefit theory, the separationist opinions by the other justices undermined the states’ support for the idea.Footnote 106
The Second Era (1947–1985)
Philip Hamburger situates the origins of mid-twentieth-century anti-Catholicism in the mid-nineteenth century reshaping of the American idea of the separation of church and state. According to Hamburger, once eighteenth-century thinkers established the separation of church and state as a “constitutional guarantee of religious freedom,” nineteenth-century state actors imposed “an aggressively Protestant ‘Americanism’ on an ‘un-American’ Catholic minority.”Footnote 107 Protestant elites gave new meaning to the separation of church and state to include “religious liberty, [the] Protestant religion, and American identity.”Footnote 108
In the 1940s, both New York and New Jersey, over considerable opposition, passed legislation to provide transportation that included students attending parochial schools. Arch Everson had been a member of the “New Jersey chapter of the Junior Order of United American Mechanics,” which sounds innocuous enough. But “[a]n old nativist order initiated the Everson case,” Hamburger writes,Footnote 109 setting forth a separationist paradigm for the next two generations of law and policy.Footnote 110 The order “had among its objects: to prevent ‘foreign competition’ from immigrants and ‘[t]o maintain the Public School System … and to prevent sectarian interference therewith.”Footnote 111 It was “an open secret,” Justice Clarence Thomas has written, “that ‘sectarian’ was code for ‘Catholic.’”Footnote 112
After World War II, American political development had entered a new era in which the states—even those with Blaine laws, like New York—and the federal government became more involved in education.Footnote 113 In 1944, Congress passed the GI Bill, which allowed veterans to use federal funds at sectarian colleges and universities.Footnote 114 In 1946, Congress passed the National School Lunch Act, which authorized states to provide lunch for poor children in both public and nonpublic schools.Footnote 115 Between 1945 and 1965, Congress passed a number of bills that sent money to the states for education, but none included support for sectarian schools. The Elementary and Secondary Education Act (ESEA) of 1965Footnote 116 changed that.
Title I provided “aid for the education of disadvantaged children residing in areas with concentrations of children from low-income families.”Footnote 117 Families that met the criteria for aid and who sent their children to religious schools were entitled to a portion of the money distributed. Nonpublic schools were also entitled to receive secular instructional materials and textbooks. But “[c]ontrol of federal funds, provision of services, and ownership of textbooks and other instructional materials were to be maintained by public school authorities.”Footnote 118
The issue of whether and how far to include nonpublic schools in granting aid seemed open in the postwar period because Cochran and Everson indicated that modest state expenditures for textbooks and transportation for students were constitutional. By 1966, seven states permitted “public transportation of parochial school children,” twelve required it, and nineteen prohibited it.Footnote 119 Various states and localities also provided funds for nativity scenes, reimbursed or provided tax credits to parents of nonpublic school children for the costs of “secular, nonideological textbooks, instructional materials and supplies,” and leased classroom space to parochial schools for secular subjects. Each of these actions was consistent with the pre-Everson understanding of the First Amendment and historical practices dating back decades—and yet, federal courts struck down these state laws on separationist grounds.Footnote 120 Demographic changes involving an increase in the school-age population, coupled with a decrease in the ability of religious schools “to rely on relatively low-cost religious staff as teachers,”Footnote 121 also contributed to state experimentation with child benefit policies at the height of the Court’s separationist logic.
Lending textbooks for secular subjects in religious schools was a natural next step in the states’ challenge to the Court’s Everson reasoning. In Board of Education v. Allen (1968), the Court considered whether a New York law that required “public school authorities to lend textbooks free of charge to all students in grades seven through 12,”Footnote 122 including religious school students, violated the establishment clause. The Court, relying on Pierce, Cochran, and Everson, saw no violation, writing that the “law merely makes available to all children the benefits of a general program to lend school books free of charge.”Footnote 123 Justice William Douglas dissented, conjuring a scenario in which sectarian groups would vie to gain control of state agencies which approve the textbooks “to disseminate ideas most congenial to their faith.”Footnote 124
At about the same time that New York provided indirect aid to religious schools, Rhode Island and Pennsylvania developed programs to partially reimburse parochial schools for the salaries of teachers who taught secular subjects. The programs excluded any payment for teaching religion. These programs resulted in the Earley v. DiCenso and Lemon v. Kurtzman rulings,Footnote 125 the latter creating the three-pronged “Lemon test” that has provided the backdrop for government aid to nonpublic schools for more than fifty years.
The Lemon test requires that a state statute (1) “must have a secular legislative purpose”; (2) “its principal or primary effect must be one that neither advances nor inhibits religion”; and (3) it “must not foster ‘an excessive government entanglement with religion.’” The Court in Lemon held that the “cumulative impact of the entire relationship arising under the statutes in each State involves excessive entanglement between government and religion.”Footnote 126
Despite the burden the test imposed on the states, some state legislatures came up with innovative proposals to provide aid to nonpublic schools. In 1970, New York amended its education and tax laws, establishing financial aid programs, using, in part, the child benefit concept.Footnote 127 Noting that many nonpublic schools needed “maintenance and repair,” the state provided per pupil grants not to exceed equivalent services in public schools. Additionally, low-income families would be eligible for a tuition reimbursement program for grade schools and high schools, as long as the grants did not exceed 50% of the actual tuition. The grants were modest, but the rationale was to help preserve the “stability of urban neighborhoods.”Footnote 128 For parents with income above the low-income threshold, New York allowed deductions from the adjusted gross income tax owed for the state income tax. Additionally, the New York legislature appropriated funds to reimburse nonpublic schools “for expenses of services and inspection in connection with administration, grading and the compliance of and reporting of the results of tests and examinations, maintenance of records of public enrollments … health records, records of personal qualifications,”Footnote 129 and other reports mandated by the state. These provisions are known as “mandated services.”
In two separate decisions decided on the same day in 1973, the Court struck down the New York programs. The Court in Levitt v. Committee for Public Education Footnote 130 thought there was a substantial risk that “examinations, prepared by teachers under the authority of religious institutions will be drafted with an eye, unconsciously or otherwise, to inculcate students with the religious precepts of the sponsoring church.”Footnote 131 And in Committee for Public Education and Religious Liberty v. Nyquist, Justice Lewis Powell called New York’s plan “ingenious … for channeling state aid to sectarian schools.”Footnote 132 Because, the Court held, the provision for the maintenance of the physical plant of the school could be used to maintain the school chapel and renovate classrooms used to teach religion, the law “subsidizes directly the religious activities of sectarian elementary and secondary schools.”Footnote 133 The tuition grants and deduction provisions, Powell maintained, failed the second prong of the Lemon test by not “guaranteeing that state aid derived from public funds will be used exclusively for secular, neutral, and nonideological purposes.”Footnote 134 Implicit in Powell’s opinion is the fungibility argument often advanced by separationists during this era: that money not spent for one purpose could be used for another—specifically, the promotion of sectarianism. Powell was particularly mindful of the Court’s role as the guardian against the “potential divisive political effect of an aid program.”Footnote 135
The idea that religious expression in public schools creates divisiveness within communities first appeared in McCollum v. Board of Education (1948), in which the Court ruled that Illinois’ voluntary religious instruction in public schools was unconstitutional.Footnote 136 The Lemon decision turned the concept of state aid to religious institutions as the incubator of divisiveness into a foundational principle of constitutional law.Footnote 137 In Meek v. Pittenger (1975), for example, the Court’s plurality held unconstitutional a Pennsylvania law that provided auxiliary services to nonpublic schools but upheld the lending of secular textbooks (but not “instructional materials,” such as films). Auxiliary services refer to “counseling, testing, and psychological services, speech and hearing therapy, teaching and related services for exceptional children, for remedial students, and for the educationally disadvantaged.”Footnote 138 Because the services were “provided only on the nonpublic school premises,”Footnote 139 it created “the danger that religious doctrine will become intertwined with secular instruction.”Footnote 140 Justice Brennan dissented from the part of the opinion upholding the lending of textbooks, writing that it had “the potential for political divisiveness.”Footnote 141
The challenge with asserting that auxiliary services, provisions for maintenance and repair, and lending nonreligious textbooks is inherently divisive is twofold: First, it invites the judiciary to render political judgments on the nature of the political process. The founders, well-versed in the tumultuous history of religious conflict, understood that because no consensus on religious belief could be achieved, such matters were best left to the states. This principle is particularly beneficial in an extended and commercial republic, structured, as Madison observed, to mitigate the influence of intense factions.Footnote 142 The Court’s role in reducing divisiveness should apply only when majorities are formed hastily and without deliberation.Footnote 143
Second, divisiveness is vague in form and grounded in subjective perception rather than consistent principle. The Supreme Court detected no signs of divisiveness when it revisited New York’s “mandated services” plan in Committee for Public Education v. Regan (1980).Footnote 144 The Court upheld the law because New York did not “reimburse nonpublic schools for the preparation, administration, or grading of teacher-prepared tests,” and it required the auditing of state funds, “thus ensuring that only the actual costs incurred in providing the covered secular services are reimbursed out of state funds.”Footnote 145 It is worth noting that New York’s aid program, in effect for four decades, has not ushered in discernible divisiveness in that state, whether religious or political.Footnote 146 Mandated services is a state experiment that has withstood the test of time and provides a not-inconsiderable amount of aid to educational institutions (approximately 8% of a typical religious school budget).Footnote 147
The End of Everson and Lemon?
Although earlier attempts at public–parochial cooperation had run afoul of nativism, the state Blaine amendments, and the Lemon test, by the 1970s, state policies to foster cooperation between religious schools and local public school boards were becoming more common.Footnote 148 In Wolman v. Walter (1977), a plurality of the justices held that lending secular textbooks and the use of standardized testing and scoring, diagnostic services, and therapeutic and remedial services on sectarian schools’ premises could pass constitutional muster under certain conditions. In Lemon, those participating in these programs were physically in the church-related school. In Wolman, however, Justice Harry Blackmun observed that “providing therapeutic and remedial services at a neutral site off the premises of the non-public schools will not have the impermissible effect of advancing religion.”Footnote 149
To be sure, the states’ attempt to limit Everson’s and Lemon’s reach was slow and difficult. In cases from Michigan and New York, the Court struck down state experiments with shared programs using state and federal funds, respectively. Grand Rapids School District v. Ball (1985) involved two programs, Shared Time and Community Education. The Shared Time program offered classes to nonpublic school children during the regular school day. The program was designed to supplement “core curriculum” courses required by the state and taught by full-time teachers employed by public schools. The Community Education program offered voluntary courses after the regular school day. The courses were enrichment to academic subjects, such as arts and crafts, home economics, drama, chess, and gymnastics. The state considered the teachers for this program part-time public school employees, who were, for the most part, employed in the nonpublic school. The rooms used for instruction were free of religious symbols.
Writing for the Court, and relying on Lemon and its progeny, Justice Brennan noted that the “government’s activities in this area can have a magnified impact on impressionable young minds, and the occasional rivalry of parallel public and private school systems offers an all-too-ready opportunity for divisive rifts along religious lines in the body politic.”Footnote 150 And in Aguilar v. Felton, decided the same day as Grand Rapids School District, the Court struck down New York City’s use of federal funds from Title I of the Elementary and Secondary Education Act to pay the salaries of public school employees who taught in parochial schools to meet the needs of educationally deprived children. These programs involved readings skills, remedial mathematics, English as a second language, and guidance services using teachers, guidance counselors, psychologists, and social workers. New York City even adopted a system of monitoring the potential of conveying religious content. Brennan, however, stuck to the Lemon ruling, writing that “the scope and duration of New York City’s Title I program would require a permanent and pervasive state presence in the sectarian schools receiving aid,”Footnote 151 leading to an increase in “political divisiveness.”Footnote 152 The Court’s opinion is remarkable for its insistence that the demarcation between religious and secular could be breached even where there was no evidence of this having occurred. The solution, often criticized as expensive and inconvenient, was to have the publicly funded remediation placed outside the physical premises of the parochial school, in vans parked outside.
Above all, these cases show the states’ commitment to policy innovation and religious accommodation in the strict separationist era, commitments that continued into the next decades, with greater success. In the early 1980s, some states adopted policies that were directed at providing general public benefits without specifically favoring religious institutions. For example, Minnesota allowed taxpayers to deduct up to $500 for grades K–6 and $700 for grades 7–12 for expenses related to “tuition, textbooks, and transportation.” The Court upheld the law in Mueller v. Allen (1983). Rehnquist, writing for the majority, noted that there are many tax deductions in the Minnesota statute that taxpayers may use, such as those for medical and charitable deductions, and that “the deduction is available for educational expenses incurred by all parents including those whose children attend public schools and those whose children attend nonsectarian private schools or sectarian private schools.”Footnote 153 Mueller is the Court’s last word on establishment until Wallace v. Jaffree. After Wallace, in cases such as Espinoza v. Montana Department of Revenue (2020), the Court will be more explicit that sectarian identity need not preclude an institution or, as in Mueller, individuals, from a general public benefit.Footnote 154
Conclusion
In this article, we have examined policy change through three primary dimensions. First, we adopted a periodized framework to more effectively trace the evolution of legal doctrine over time. Second, we analyzed how scholarly reinterpretations of the establishment clause prompted the Court, decades later, to revise its understanding of the establishment clause. And third, we emphasized state-level institutional, political, and cultural dynamics that influenced legal transformations in the 1990s and beyond.
We have, furthermore, argued that the First Amendment’s establishment clause, made applicable to the states via the Fourteenth Amendment’s due process clause, prohibits the government from promoting or affiliating with a specific religion but denied that it requires the exclusion of religion from the public sphere. Periodization of the establishment clause situates the clause’s meaning in its proper historical relationship to federalism, religious pluralism, and cultural practices. From the earliest days of the American regime until just after World War II, federalism, religious pluralism, and policy incrementalism set the stage for state-level innovation and school choice—policies that persisted even after the Court nationalized the establishment clause in Everson—relied too heavily on Jefferson’s letter and Madison’s “Memorial and Remonstrance” to limit accommodation, and crafted a strict-separationist doctrine amid concerns about religious divisiveness.
The Lemon decision, fearing the “divisive political potential”Footnote 155 of state aid to nonpublic schools, seemed to preclude additional experiments by states in the form of teachers’ salaries, school repairs, tuition assistance, and other types of auxiliary services, such as providing remedial education, reimbursing school trips, and offering loans of audio-visual equipment. But as we have shown, the states continued to challenge this ruling and the potential for divisiveness has always been overstated. Periodization also reveals that different modes of accommodation change over time. A recent crop of cases in this century—Zelman v. Simmons-Harris (2002),Footnote 156 Trinity Lutheran Church of Columbia v. Comer (2017),Footnote 157 and Espinoza v. Montana Department of Revenue (2020)—shows that states have rejected the divisiveness argument and established the principle that religious institutions cannot be excluded from general public welfare programs on the basis of religious identity. The passage of the Religious Freedom Restoration Act (1993),Footnote 158 which “prohibits the Federal Government from taking any action that substantially burdens the exercise of religion,”Footnote 159 coupled with the decision in City of Boerne v. Flores (1997),Footnote 160 returns to the states the question of the type of permissible state interaction with religion. The movement for religious charter schools, just getting off the ground this decade, is the latest effort in the innovations begun more than a century ago, rooted in the flexibility of the nation’s founding documents and institutions.Footnote 161
