Concept
Critiques of Current Doctrine
Establishment
Critiques of Current Doctrine
Chief Justice Burger’s dissent in Wallace v. Jaffree shows how sharply the Justices have disagreed about the doctrine in this chapter. In his view the Court was striking down a statute that did nothing more than permit a moment of silence: the statute required no one to pray, prescribed no prayer, and merely allowed students who wished to pray to do so.1 Other Justices have attacked the framework itself. They have strongly criticized the Lemon and endorsement tests, and in various settings the Court has declined to apply them, at least as such, often emphasizing history and tradition instead.2 Justice Powell, concurring in Wallace, saw the danger in that trend: continued criticism of Lemon, he warned, could encourage other courts to feel free to decide Establishment Clause cases on an ad hoc basis.3
The Court’s own practice has been uneven. A plurality in Van Orden v. Perry found Lemon not useful for a passive monument on the Texas Capitol grounds and relied instead on the nature of the monument and the Nation’s history. It called the monument a far more passive use of the text than the classroom posting in Stone, where the text confronted elementary school students every day, and it noted that Texas had treated its Capitol monuments as representing strands in the State’s political and legal history.4 In McCreary County, by contrast, a 5–4 Court, per Justice Souter, applied Lemon and held that the displays violated the Clause, although the counties defended them as a secular effort to educate the citizenry on the foundations of American law and government. The secular purpose prong, the Court said, serves the important function of mandating government neutrality toward religion.5
Whether Lemon still governs depends on the date. From 1971 until 2022 the Establishment Clause test derived from Lemon v. Kurtzman, although the Court did not always use it, and a large body of cases grew up under it.6 In Kennedy v. Bremerton School District (2022), Justice Gorsuch’s opinion for the Court formally overruled Lemon and, rather than creating a new legal standard, endorsed a historical approach.7 It is unclear what that means for the many cases decided under Lemon, so the objections below remain live disagreements.8
Beneath the complaints about particular tests lies a disagreement about what the Clause is for. On the separationist view, religion and government exist in profound tension: separation protects the church from worldly corruption and protects secular interests from being co-opted by ecclesiastical influences.9 Its proponents argue that the Founders established a secular state walled off from religion, and they emphasize the writings and beliefs of Jefferson and Madison; the Court first hinted at the view in Reynolds v. United States (1878).10 On the non-preferentialist view, government may aid religion so long as it prefers no one religion over the others, because religions should multiply and prosper and compete as factions so that none dominates.11 A second divide concerns method. History and tradition are increasingly used, especially by accommodationists, to validate measures that might clearly be viewed as religious if government embraced them today for the first time, and it is often hard to predict the outcome of a present-day case.12
Three objections carry those disagreements into the doctrine: that the Clause forbids only preference among sects, that long-standing practice should control, and that the Clause may not bind the states at all. Each section below states the objection, gives the answer of the Justices who defend the settled law, and says what the dispute would change on an exam.
Nonpreferentialist Reading
Justice Rehnquist’s dissent in Wallace v. Jaffree is the leading statement of the nonpreferentialist reading. He urged a return to the language and intent of the drafters. The evil to be aimed at, he said, appears to have been the establishment of a national church, and perhaps the preference of one religious sect over another, but the Clause was “definitely not concerned about whether the Government might aid all religions evenhandedly.” The First Amendment, he concluded, did not command neutrality between religion and irreligion, much less create a wall of separation, and he called for the repudiation of the Lemon test.13 On his account the Clause stops the Federal Government from preferring one denomination or sect over others, and through incorporation it stops the States from establishing a religion or discriminating between sects. It does not stop Congress or the States from pursuing legitimate secular ends through nondiscriminatory sectarian means.14 Government may therefore promote religion generally, discriminating against nonreligion and supporting religious institutions, so long as it creates no preference for a particular religion.15
The reading has never commanded a majority in a modern Supreme Court opinion.16 The rule the Court has applied instead comes from Everson v. Board of Education, which applied the Clause to the States in 1947.17 Everson declared that neither a state nor the Federal Government can set up a church, and that neither “can pass laws which aid one religion, aid all religions, or prefer one religion over another.”18
Rehnquist’s case is mostly historical. He said that a historical statement on which the Court’s earlier opinions relied is “demonstrably incorrect as a matter of history,” and that repeating it in later opinions gives it no more authority than it has as a matter of fact, because “stare decisis may bind courts as to matters of law, but it cannot bind them as to matters of history.”19 He offered the First Congress’s own conduct as evidence of what the Religion Clauses meant. On the day after the House voted to adopt the form of the Religion Clauses that was ultimately proposed and ratified, Representative Elias Boudinot proposed that the President be asked to issue a Thanksgiving Day Proclamation. Boudinot said he could not think of letting the session pass without offering all citizens the chance to join in returning thanks to Almighty God.20 Rehnquist also relied on Justice Story’s reading that the amendment’s object was “to exclude all rivalry among Christian sects, and to prevent any national ecclesiastical establishment,” and so to cut off the means of religious persecution.21 From this evidence he concluded that the Clause had acquired a well-accepted meaning: it forbade the establishment of a national religion and forbade preference among sects.22
From the history Rehnquist drew a verdict on the doctrine built on the contrary reading. The Court’s recent opinions, many of them “hopelessly divided pluralities,” have conceded “with embarrassing candor” that the wall of separation is merely a “blurred, indistinct, and variable barrier” that can only be “dimly perceived.”23 The tests fared no better. The purpose prong, he argued, condemns nothing so long as the legislature utters a secular purpose and says nothing about aiding religion, so a statute’s constitutionality may turn on what the legislators put into the legislative history and what they leave out.24 McCreary County answers part of that charge. The purpose inquiry cannot be trumped by transparent efforts to circumvent the secular purpose requirement, and it must take into account all probative evidence of purpose, including the original purpose. A moment-of-silence law amended by adding a phrase such as “or meditation” might therefore still fail.25
The entanglement prong, Rehnquist said, creates a paradox with no solution in school aid cases once it is divorced from the logic of Walz. Aid to parochial schools must be closely watched lest it be put to sectarian use, yet the close supervision itself creates entanglement.26 The Court had tried to calibrate that problem by setting. It treated aid to church-related colleges more lightly because college students are less impressionable and less susceptible to religious indoctrination, and because college courses by their nature limit the opportunities for sectarian influence.27 Where religious indoctrination is not a substantial purpose of the institution, the risk that aid will support religious activities falls, intensive government surveillance becomes less necessary, and the resulting entanglements lessen.28 The Court later folded entanglement into the inquiry into a statute’s effect, while reiterating that entanglement that is “excessive” still violates the Clause.29
Rehnquist added that the results of these tests violate the historically sound principle that the Clause does not forbid governments to provide general welfare programs under which benefits are distributed to private individuals, even though many of them may elect to use the benefits in ways that “aid” religious instruction or worship.30 That principle resembles the private-choice reasoning taught earlier in this chapter. Defenders of the settled law answer that the doctrine can be refined without being abandoned. A concurring opinion in the Cleveland voucher case described the Court’s requirement that state aid flowing to religious organizations through beneficiaries must do so only at the beneficiaries’ direction as a refinement of Lemon that “surely does not betray Everson.”31
The argument is older than Rehnquist. In McCollum v. Board of Education, litigants argued that the First Amendment was historically intended to forbid only government preference of one religion over another, not impartial assistance to all religions. They also asked the Court to distinguish or overrule Everson’s holding that the Fourteenth Amendment applies the Clause to the States. After full consideration, the Court was unable to accept either contention.32 Justice Reed, dissenting, agreed that no governmental entity can “set up a church” and that none can “aid” all or any religions or prefer one “over another.” But he read “aid” narrowly, as “purposeful assistance directly to the church itself or to some religious group or organization doing religious work of such a character that it may fairly be said to be performing ecclesiastical functions,” and he read “prefer” to require an advantage given to one over another.33 He also distrusted the metaphor at the center of the Court’s account. The difference between the generality of Jefferson’s statements on separation of church and state and the specificity of the conclusions drawn on education is considerable, and “a rule of law should not be drawn from a figure of speech.”34
The two readings produce different results only where government supports religion generally without ranking sects.
Justice Souter’s reply, in his concurrence in Lee v. Weisman, begins with the text. The House conferees persuaded the Senate to accept the final wording: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof.” Unlike the earliest House drafts or the final Senate proposal, that wording is not limited to laws respecting an establishment of “a religion,” “a national religion,” “one religious sect,” or specific “articles of faith.” The Framers, Souter concluded, repeatedly considered and deliberately rejected such narrow language and extended their prohibition to state support for “religion” in general.35 He quoted a prominent commentator on what the narrow reading would require: confining the Clause to a prohibition on preferential aid “requires a premise that the Framers were extraordinarily bad drafters—that they believed one thing but adopted language that said something substantially different, and that they did so after repeatedly attending to the choice of language.”36
Madison’s own words point the same way. In his Memorial and Remonstrance he asked, “Who does not see that the same authority which can establish Christianity, in exclusion of all other Religions, may establish with the same ease any particular sect of Christians, in exclusion of all other Sects?”37 Later writers have drawn conflicting lessons from this history, but the current consensus seems to be that the record does not support the most limited view, which would allow government support for religion so long as no specific religion is preferred.38
Souter’s second reply does not depend on winning the historical argument. Nonpreferentialism requires some line between “sectarian” practices and those ecumenical enough to pass. By requiring that inquiry, nonpreferentialists invite courts to engage in comparative theology, a subject Souter could hardly imagine less amenable to the competence of the federal judiciary.39 The rabbi’s graduation prayer in Lee shows the difficulty. Even if its thought is generic enough for most believers, it embodies a theistic premise, and many Americans who consider themselves religious are not theistic. A nonpreferentialist who would condemn the Anglican liturgy at a public school graduation would still have to explain why the government’s preference for theistic over nontheistic religion is constitutional.40 Rehnquist’s attack on the doctrine as historically ungrounded leads directly to the next objection, that long-standing practice should decide Establishment Clause cases.
History and Tradition
The second objection concerns method. Justice Scalia’s dissent in Lee v. Weisman charged that the Court had replaced Lemon with a “psycho-coercion test” that has “no roots whatever in our people’s historic practice” and is “infinitely expandable as the reasons for psychotherapy itself.”43 His alternative looks to what coercion meant in history. The coercion that was a hallmark of historical establishments was coercion of religious orthodoxy and of financial support by force of law and threat of penalty. Attendance at the state church was typically required, only clergy of the official church could lawfully perform sacraments, and dissenters, if tolerated, faced an array of civil disabilities.44 On that definition a court asks whether the law backs a religious demand with a penalty. The Lee majority’s concern with the risk of indirect and subtle coercion asks a less bounded question, which is the point of Scalia’s objection.45
Dissenters had argued from history before. Justice Reed, dissenting in McCollum, said that for him “the history of past practices is determinative of the meaning of a constitutional clause, not a decorous introduction to the study of its text.”46 He pointed to federal practice. Each House of Congress has a chaplain who daily invokes divine blessings and guidance, and the armed forces have commissioned chaplains from early days and conduct public services on government property. Eligible veterans may receive training at government expense for the ministry in denominational schools, and the schools of the District of Columbia open with a Bible reading without note or comment.47 The Court has approved similar practices in dicta, even though by every indication they advance and endorse religion deliberately. They include presidential Thanksgiving proclamations with religious references since the time of George Washington; the Supreme Court’s own opening cry, “God Save the United States and this Honorable Court”; the national motto, “In God We Trust,” and its use on currency; and the words “one nation under God” in the Pledge of Allegiance since 1954.48
Opinions that accept the method still draw limits. One accepts that the Court properly looked to history in upholding legislative prayer, property tax exemptions for houses of worship, and Sunday closing laws, and it quotes Justice Holmes: “[i]f a thing has been practised for two hundred years by common consent, it will need a strong case for the Fourteenth Amendment to affect it.” The same opinion distinguishes Presidential Proclamations from school prayer. Proclamations are received in a noncoercive setting and are directed primarily at adults, who are not readily susceptible to unwilling religious indoctrination. Government-sponsored exercises directed at impressionable children who are required to attend school are much more likely to produce coerced religious beliefs.49
Justice Souter’s concurrence in Lee replies that early practice is relevant but not determinative. The practice was itself divided. President Jefferson steadfastly refused to issue Thanksgiving proclamations of any kind, in part because he thought they violated the Religion Clauses,50 and James Madison refused to call for days of thanksgiving and prayer during his first three years in office, though amid the political turmoil of the War of 1812 he did so on four separate occasions.51 The First Congress hired chaplains, and Presidents Washington and Adams marked days of public thanksgiving and prayer. But in the face of the separationist dissent those practices prove, at best, that the Framers did not share a common understanding of the Establishment Clause and, at worst, that they, like other politicians, could raise constitutional ideals one day and turn their backs on them the next.52 Souter tested the argument on a different clause. Ten years after proposing the First Amendment, Congress passed the Alien and Sedition Acts, and if the early Congress’s political actions were determinative, not merely relevant, evidence of constitutional meaning, the Court would have to gut current First Amendment doctrine to make room for political censorship.53
Recent decisions give history a larger role and show where tradition has done its work. In Town of Greece, the town neither reviewed the prayers in advance nor guided their tone or content, believing that any control would infringe the free exercise and speech rights of the ministers. Guest clergy composed their own devotions, and the prayers often sounded both civic and religious themes.54 Although the government sponsored the prayers, the private prayer-givers determined their content, a factor that implicates the value of religious free speech and the First Amendment’s aversion to censorship.55 School-sponsored prayer stays outside any tradition-based exception. The risk of indirect and subtle coercion discussed in Lee is reason enough to reject such an exception for school-sponsored nonsectarian prayer in the public school setting, even if the prayer is formally voluntary. The Pledge of Allegiance differs, as Chief Justice Rehnquist and Justice O’Connor emphasized in Elk Grove Unified School District v. Newdow, because it is not a prayer or religious exercise.56
Kennedy v. Bremerton made history the Court’s stated method. Justice Gorsuch’s opinion formally overruled Lemon and, rather than creating a new legal standard, endorsed a historical approach. The case began when a football coach was disciplined for leading his team out onto the field after a game for prayer; the players were not required to participate, and some chose not to.57 Joseph Kennedy had begun coaching at Bremerton High School in 2008 after nearly two decades in the Marine Corps, and he had made it a practice to give “thanks through prayer on the playing field” at the end of each game.58 Apart from its rejection of outright coercion to engage in religious belief, Kennedy leaves great uncertainty. It is unclear whether states can institute prayers in public schools, or even whether it would be unconstitutional for a state to declare a particular creed its official religion so long as it took no coercive measures to enforce that status. Presumably some inquiry into the original understanding would be necessary to decide those points.59 The same attention to history appears in the monument cases, from Van Orden to American Legion v. American Humanist Association.
An exam answer on a practice defended by its pedigree should therefore sort the facts into three questions: how old and widely accepted the practice is, who controls its content, and whether it reaches schoolchildren who must attend. The answer should also attribute the broader claims to Scalia’s dissent in Lee and not to a holding of the Court. Those claims are that a test with no roots in historic practice is illegitimate and that coercion means force of law and threat of penalty.
Incorporation Doubts
The third objection goes beneath the tests and asks whether the Establishment Clause should restrain the states at all. Its textual form appears in Justice Thomas’s concurrence in Zelman v. Simmons-Harris. The Clause says that “Congress shall make no law respecting an establishment of religion,” and on its face that provision places no limit on the States with regard to religion. In Thomas’s account the Clause originally protected the States, and by extension their citizens, from the imposition of an established religion by the Federal Government, so whether and how it should constrain state action under the Fourteenth Amendment is “a more difficult question.”62 He made the same point in Van Orden v. Perry, where he read the history and text of the Clause as resisting incorporation. On that view the Clause was simply a federalist limitation on the national government.63
The historical support for the federalism reading comes from Madison, who believed that even under the original Constitution there was “not a shadow of right in the general government to intermeddle with religion.” The framers and ratifiers agreed on the impropriety of national involvement but wished to make the point explicit. Directing themselves to Congress, they acted to limit national power and to preserve the power of the states to address religion and religious liberty as the states saw fit.64 Defenders of incorporation have two answers. First, critics of incorporation note that some states did establish religion by favoring particular denominations, but by the time of the Fourteenth Amendment those establishments had been eliminated.65 Opinions discussing that state-level history cite state constitutions of the 1870s, among them those of Illinois, Pennsylvania, Nebraska, and California.66 Second, the Establishment Clause protects individual liberty too. Its incorporation is more controversial than that of the Free Exercise Clause because the latter clearly safeguards individual liberty while the former seems directed at the government, but the Court has explained that the Establishment Clause also protects liberty.67
The coercion theory fits that second answer. Its principle is that “government may not coerce anyone to support or participate in any religion or its exercise.” That principle helps explain why the Clause applies to the states, because like other parts of the First Amendment the Clause is incorporated into the concept of “liberty” under the Fourteenth Amendment. Incorporation is harder to explain if the Clause is seen as pertaining primarily to the secular nature of government rather than to individual freedom, although the coercion approach, too, has significant problems.68
Thomas’s Zelman concurrence accepts that incorporation is about liberty and turns the point against the Court’s practice. Rights incorporated against the States through the Fourteenth Amendment, he wrote, should advance, not constrain, individual liberty.69 It follows, in his view, that while the Federal Government may “make no law respecting an establishment of religion,” the States may pass laws that include or touch on religious matters so long as the laws do not impede free exercise rights or any other individual religious liberty interest. A federal court asks which particular religious liberty right the State is said to have invaded, and so balances the demands of the Fourteenth Amendment against the federalism prerogatives of the States.70 In Zelman the point decided the argument. Whatever the textual and historical merits of incorporating the Establishment Clause, Thomas could accept that the Fourteenth Amendment protects religious liberty rights, but he could not accept using it to oppose neutral programs of school choice. There would be “a tragic irony in converting the Fourteenth Amendment’s guarantee of individual liberty into a prohibition on the exercise of educational choice.”71
Three positions are in play, and an exam answer should say which one it applies:
- Settled doctrine holds that the Establishment Clause is incorporated into the Due Process Clause of the Fourteenth Amendment and applies to the states as well as the federal government.72
- The Zelman concurrence suggests that state action may be judged on different terms, tested against the particular individual religious liberty right invoked, and that the Clause should not be used to oppose neutral school-choice programs.73
- The Van Orden concurrence maintains that the text and history of the Clause resist incorporation altogether, because the Clause was a federalist limit on Congress.74
The Court’s own position is that incorporation is settled. Cantwell v. Connecticut suggested that the Establishment Clause was likewise incorporated, and the Court so ruled a few years later, in 1947, in Everson v. Board of Education.75 In McCollum, the challengers asked the Court to distinguish or overrule Everson’s holding that the Fourteenth Amendment makes the Clause a prohibition against the States, and the Court, after full consideration, declined.76 The practice before it, the Court said, “falls squarely under the ban of the First Amendment (made applicable to the States by the Fourteenth) as we interpreted it in Everson.”77 More recently, a district judge surprisingly attempted to revise history to rule that the Clause should not have been applied to the states. The Supreme Court promptly entered a stay and without qualification reaffirmed its long-standing and consistent precedent incorporating the Clause.78
An exam answer on a state or local practice should therefore test it under the Clause as the doctrine does, as a limit that binds states as it binds Congress. The federalism reading belongs in the answer only as an argument from separate opinions, attributed to Thomas and not to the Court, with the note that his Zelman concurrence would use it to reject Establishment Clause challenges to neutral school-choice programs.