Establishment Clause Inversion in the Bladensburg Cross Case

Richard C. Schragger Perre Bowen Professor of Law and Joseph C. Carter, Jr. Research Professor of Law, University of Virginia School of Law and Micah Schwartzman Joseph W. Dorn Research Professor of Law, University of Virginia School of Law

One of the most memorable exchanges in the recent history of U.S. Supreme Court oral arguments happened in Salazar v. Buono.[1] The case involved an Establishment Clause challenge to a Latin cross displayed on public land in the Mojave National Preserve. The cross was built in the 1930s as a memorial to soldiers who died in World War I.[2] During argument in the case, Peter Eliasberg, an attorney for the ACLU whose father and grandfather were Jewish war veterans,[3] said that he did not think the Latin cross represented soldiers who were not Christian. That statement provoked a sharp response from Justice Antonin Scalia, and the following exchange ensued:

JUSTICE SCALIA: The cross doesn’t honor non-Christians who fought in the war? Is that – is that –

ELIASBERG: I believe that’s actually correct.

JUSTICE SCALIA: Where does it say that?

ELIASBERG: It doesn’t say that, but a cross is the predominant symbol of Christianity and it signifies that Jesus is the son of God and died to redeem mankind for our sins, and I believe that's why the Jewish war veterans –

JUSTICE SCALIA: It’s erected as a war memorial. I assume it is erected in honor of all of the war dead. It’s the – the cross is the – is the most common symbol of – of – of the resting place of the dead, and it doesn’t seem to me – what would you have them erect? A cross – some conglomerate of a cross, a Star of David, and you know, a Moslem half moon and star?

ELIASBERG: Well, Justice Scalia, if I may go to your first point. The cross is the most common symbol of the resting place of Christians. I have been in Jewish cemeteries. There is never a cross on a tombstone of a Jew.


The laughter recorded in the official transcript of the oral argument was from the courtroom audience, which perhaps recognized both the obviousness of Mr. Eliasberg’s point and the absurdity of Justice Scalia’s claim to the contrary. Justice Scalia was apparently angered by this reaction[5] and continued the argument by saying: “I don't think you can leap from that to the conclusion that the only war dead that that cross honors are the Christian war dead. I think that's an outrageous conclusion.”[6]

A decade later, and a few years after Justice Scalia’s death, the Supreme Court has revisited the question of whether a Latin cross is predominantly the symbol of Christianity or whether it can also be used as a universal symbol to represent soldiers, including non-Christians, who died fighting for our country.[7] In American Legion v. American Humanist Ass’n,[8] the Court rejected an Establishment Clause challenge to a forty-foot tall Latin cross, publicly owned and maintained as a World War I memorial in Bladensburg, Maryland. In upholding the constitutionality of the Bladensburg Cross,[9] seven justices accepted Justice Scalia’s view—that it is outrageous to claim that the Latin cross fails to represent all the war dead—and made it the law of the land.[10] An argument that was sufficiently “off-the-wall”[11] to prompt ridicule in the courtroom is now a matter of precedent in the Supreme Court’s interpretation of the Establishment Clause.

In this article, we focus on how this constitutional change occurred and what it means for the Establishment Clause going forward. We argue that American Legion represents a significant development in the dismantling of Establishment Clause jurisprudence. The Court has been chipping away steadily on precedent that restricts the government from providing material and expressive support for religion. American Legion could be read narrowly as part of the ongoing demise of the separation of church and state, which has been underway since the Rehnquist Court. But in our view, the decision amounts to something more, namely, an inversion of disestablishment principles. The Court did not use the Establishment Clause as a shield to protect vulnerable minorities, but rather as a sword against them. In upholding the Bladensburg Cross, and in declaring that its removal would show hostility toward religion, the Court has entrenched the views of past religious majorities and opened the way toward Christian preferentialism. In the midst of rising ethno-religious nationalism in the United States, and around the world, American Legion participates in undermining constitutional safeguards against the integration of church and state.

I. The Demise of Disestablishment

We start by putting American Legion in the context of a decades-long trend of undoing the separationist interpretation of the Establishment Clause, which restricted various forms of government aid to religion.[12] In the areas of 1) public funding of religious institutions, 2) government religious speech, and 3) religious accommodations, the Court has taken an increasingly deferential view toward government acts and expression that support religious beliefs, practices, and institutions.

First, in terms of funding, the Rehnquist Court was already pushing the doctrine toward permitting government funding of religious schools and social service organizations from the mid-1980s through the early 2000s.[13] This line of decisions culminated in the 2002 decision Zelman v. Simmons-Harris,[14] which held that a voucher program that included private religious schools was constitutional so long as parents and children had a genuine choice of where to attend school.[15] Zelman was the last in a line of cases that ultimately rejected a no-aid principle that the Court had applied, however unevenly, since the 1950s.[16] The replacement for the no-aid principle was “neutral aid.”[17] That principle permitted government to fund religious entities, but did not require it. State constitutional provisions that had more restrictive no-aid requirements could still be interpreted to prevent the funding of religious schools.[18]

More recently, however, the Roberts Court has held that neutral aid is not only permissible but, in some cases, required. In Trinity Lutheran Church of Columbia v. Comer,[19] the Court ruled that a state government violated the Free Exercise Clause by excluding religious organizations from equal access to a public funding program.[20] Applying a non-discrimination principle that borrows from equal protection and free speech doctrine,[21] the Court concluded that the government cannot impose special disabilities on an organization because of its religious status.[22] This shift from a no-aid to a “compulsory-aid” principle represents a dramatic change in the interpretation of the Establishment Clause over the course of the last half century.[23]

Second, as the Court has retreated from Establishment Clause restrictions on funding, it has also withdrawn limits on government religious speech. While on the Court, Justice O’Connor had interpreted the Establishment Clause to forbid government speech that endorses a particular religious perspective.[24] Her test, which the Court adopted and applied over the span of two decades,[25] asked whether a government-sponsored religious display “sends a message to nonadherents that they are outsiders, not full members of the political community, and an accompanying message to adherents that they are insiders, favored members of the political community.”[26] In recent cases, however, the Court has refused to apply the endorsement test and is much more likely to permit specifically sectarian speech in public settings. In Town of Greece v. Galloway,[27] the Roberts Court upheld a town’s practice of commissioning ministers to offer prayers at the openings of town council meetings. The prayers offered were overwhelmingly Christian and sectarian in nature.[28]

A majority of the Court now appears to reject the endorsement test. Some justices have previously advocated for a coercion test, which would prevent the Establishment Clause from being applied to non-coercive symbolic expression.[29] Others have advocated an historical approach that permits a wide range of public religious endorsements, though not ones that would represent an “establishment” in eighteenth-century terms.[30] It is unclear which non-coercive practices would be impermissible if such a test were applied, though it seems many if not most would survive such an historical review.

Third, in addition to loosening restrictions on government funding and expression, the Roberts Court has read the Establishment Clause narrowly to permit a significant expansion of constitutional and statutory religious accommodations, including to large, for-profit corporations.[31] In Cutter v. Wilkinson, which involved a facial challenge to the Religious Land Use and Institutionalized Persons Act, the Court held that exemptions for religious entities “need not come packaged with benefits to secular entities.”[32] The Court also held that, in granting exemptions, the government cannot favor particular religious denominations over others, and it must give adequate consideration to burdens that such exemptions impose on third parties.[33] But these rules set outer limits on religious accommodations, and it is unclear whether the Supreme Court will apply them in a meaningful way.[34] The issue is pressing, as religious organizations, including for-profit corporations, seek expansive exemptions from federal and state civil rights laws.[35]

Looming over all these areas of doctrine is a general question about how the Supreme Court interprets the meaning of the Establishment Clause. Starting in the early 1970s, the Court consolidated its approach under the well-known Lemon test, which holds that government action must have a secular purpose, that it must not have a primary effect of advancing or inhibiting religion, and that it must not foster excessive entanglement between church and state.[36] Justice O’Connor’s endorsement test, mentioned above, was initially presented as a gloss on the purpose and effects inquiries under Lemon.[37] But over the years, many justices have expressed dissatisfaction and severe criticisms of these tests. In some cases, the Court has sharply limited application of these tests, and in others, it has altogether ignored them.[38]

The jettisoning of Establishment Clause doctrine is thus occurring both in the formulation of specific rules for regulating government funding, speech, and accommodations,[39] and at the level of providing a larger framework within which such rules are determined. The entire edifice of the Establishment Clause has been collapsing, with the result that principles of disestablishment are becoming increasingly irrelevant.[40] They are either ignored or replaced by principles of neutrality and non-discrimination, on the one hand, and demands for special solicitude and public recognition of religion, on the other.[41]

II. American Legion and Government Religious Speech

American Legion continues the process of narrowing Establishment Clause jurisprudence. There was never any doubt about the outcome of the case. Given recent changes in the Court’s personnel, including the appointments of Justices Gorsuch and Kavanaugh, the Court was expected to affirm the constitutionality of the Bladensburg Cross.[42] The question was how the justices would do it, and whether they would provide guidance with respect to larger questions about the meaning of the Establishment Clause. The answer turned out to be highly fractured. The justices produced seven opinions—eight if one counts Justice Alito as having written both a majority for the Court and, in parts, a plurality for himself and three other justices.[43] These opinions reveal a Court sharply divided about the meaning of the Establishment Clause. It is nevertheless possible to distinguish a few main strategies used to affirm the government’s religious expression.

The most minimal approach, adopted by the Court, holds that the Bladensburg Cross is constitutional because it has taken on secular meaning as a war memorial.[44] Writing for the majority, Justice Alito enumerates various problems with interpreting the meaning of longstanding monuments: first, it can be difficult to identify the purposes for which they were created;[45] second, even if those purposes can be determined and even if they were religious, they may be obscured or proliferated over time as the government acquires secular purposes for maintaining old monuments;[46] third, the messages conveyed by monuments can also change over time;[47] and fourth, removing religiously expressive monuments that have taken on historical significance may convey hostility toward religion.[48] These considerations lead the majority to distinguish between old religious monuments and news ones. The Court holds that “[t]he passage of time gives rise to a strong presumption of constitutionality.” It then applies that presumption, along with the enumerated considerations, to find that the Bladensburg Cross is constitutional.[49]

A plurality of four justices—Justice Alito, joined by Chief Justice Roberts and Justices Breyer and Kavanaugh—would abandon the application of the Lemon and endorsement tests in cases involving religious symbols.[50] Justice Alito claims to apply “a more modest approach that focuses on the particular issue at hand and looks to history for guidance.”[51] The benchmarks for this approach are the Court’s legislative prayer decisions,[52] which give the government wide latitude in promoting religious speech. The plurality claims that the practice of legislative prayer “stands out as an example of respect and tolerance for differing views, an honest endeavor to achieve inclusivity and nondiscrimination, and a recognition of the important role that religion plays in the lives of many Americans.”[53] And it would hold that any well-established government religious speech or practice that fits this description is constitutional.[54]

The plurality’s emphasis on historical practice is open to various interpretations. In his concurring opinion, Justice Kavanaugh reads the plurality as moving toward a more comprehensive rejection of Lemon and as accepting an approach based on “history and tradition.”[55] In an ambitious interpretation of the Court’s recent Establishment Clause jurisprudence—perhaps of the type that the plurality otherwise cautions against in its criticism of Lemon—Justice Kavanaugh formulates “an overarching set of principles” according to which:

If the challenged government practice is not coercive and if it (i) is rooted in history and tradition; or (ii) treats religious people, organizations, speech, or activity equally to comparable secular people, organizations, speech, or activity; or (iii) represents a permissible legislative accommodation or exemption from a generally applicable law, then there ordinarily is no Establishment Clause violation.[56]

This multi-part test is certain to be the subject of much discussion and, we expect, confusion. For example, to the extent the government has nearly unfettered discretion to engage in secular speech, it would seem that any comparable treatment of religious expression would fall within category (ii) of Justice Kavanaugh’s “safe harbor.”[57] Perhaps the constitutionality of all, or nearly all, non-coercive government religious expression is an intended consequence of this test.

Justices Breyer and Kagan are more circumspect about the plurality’s appeal to “history and tradition.” In his concurring opinion, Justice Breyer explains that, in his view, the Court did not adopt a “history and tradition test” that would allow for the construction of new religious monuments;[58] rather, the majority placed significant weight on the age and “historical context” of the Bladensburg Cross.[59] Old monuments are one thing, but new ones, even if they reflect some historical practice, are another matter. For her part, in addition to joining both the majority opinion and Justice Breyer’s concurrence, Justice Kagan also writes separately to explain that she refused to join parts of Justice Alito’s opinion because, in her view, the Lemon test’s “focus on purposes and effects is crucial in evaluating government action in this sphere—as this very suit shows.”[60]

For those counting, however, it is clear that the majority of justices—all but Justices Kagan, Ginsburg, and Sotomayor—are prepared to reject Lemon, at least with respect to cases involving religious symbols. Piling on to the extensive criticisms from the four-justice plurality, Justices Thomas and Gorsuch write separate opinions concurring in the judgment, both of which indicate their eagerness to overrule Lemon in its entirety.[61] In his opinion, Justice Thomas reiterates his opposition to the incorporation of the Establishment Clause against the states. He would hold that the Establishment Clause simply does not apply to actions by state and local governments that promote religion.[62] And even if it did apply, Justice Thomas would reject challenges to any noncoercive government speech, both because “[t]he sine qua non of an establishment of religion is actual legal coercion”[63] and because the idea that such speech must remain religiously neutral “is inconsistent with our Nation’s history and traditions.”[64]

Yet another strategy for rejecting Establishment Clause challenges would be to deny standing to those who observe the government’s expression. Justice Gorsuch elaborates this view in his concurrence, joined by Justice Thomas.[65] Thus, in addition to disincorporating the Establishment Clause, or applying it only to coercive acts and never to state-sponsored sectarian speech, Justice Gorsuch would also hold that no one has standing as an “offended observer” to object to even the most blatant government endorsements of religion.[66]

Against all this, Justice Ginsburg responds with a beleaguered but powerful dissent, joined only by Justice Sotomayor. Applying a principle of religious neutrality, and relying on the endorsement test to articulate that principle, Justice Ginsburg argues that the Bladensburg Cross is a clear violation of the Establishment Clause.[67] “As I see it,” she writes, “when a cross is displayed on public property, the government may be presumed to endorse its religious content.”[68] According to Justice Ginsburg, that presumption is not overcome by viewing the Cross as a war memorial, because the history of memorializing the war dead shows that “[t]he cross was never perceived as an appropriate headstone or memorial for Jewish soldiers and others who did not adhere to Christianity.”[69] For the dissenters, the meaning of the Bladensburg Cross has not changed over time. It was and remains an affirmation and endorsement of Christianity.[70]

III. Establishment Clause Inversion

How does the cross, the predominant and exclusive symbol of Christianity, become more entrenched as a civic marker rather than less as a matter of Establishment Clause doctrine? The American Legion majority concludes that ordering the removal of the Cross would not be “a neutral act but . . . the manifestation of a hostility to religion that has no place in our Establishment Clause traditions.”[71] In this way, the Court arrives at a doctrinal position in which maintaining the central symbol of Christianity as opposed to taking it down is necessary to vindicate Establishment Clause values. A remarkable inversion occurs in American Legion: The central religious symbol of the dominant, majority religion must remain in place because its continuous presence for so many years represents “a society in which people of all beliefs can live together harmoniously” and to remove it would suggest otherwise.[72]

This inversion occurs in two steps.[73] First, the Court has to empty the Bladensburg Cross of its religious significance and meaning. Second, it has to credit the religious and civic majority’s understanding of what removing the Cross means over and above religious and civic minorities’ understanding of what the existing Cross represents. In this way, the Establishment Clause becomes a shield for existing civic-religious practices, and perhaps for future ones as well.

A. The Cross Is Not Religious

The first step requires the Court to engage in the semiotics of the Latin cross, even as it insists that meaning is uncertain and unstable. The majority spends a great deal of time discussing both the intent of those who created the Bladensburg Cross and the contemporary meaning of the memorial. The cross, asserts the Court, came into “widespread use as a symbol of Christianity by the fourth century, and it retains that meaning today.”[74] But there are also instances in which the cross has attained a secular meaning and, indeed, “instances in which its message is now almost entirely secular.”[75] The Court’s investigation into the cross’s meaning in the first part of its opinion could fit comfortably into a Justice O’Connor-style endorsement analysis—with likely the same outcome.

Indeed, Justice Alito begins with a discussion of the “indisputably secular” use of the cross as a trademark by businesses and secular organizations.[76] He refers to the use of the cross by the International Committee of the Red Cross and to its appearance on the Swiss flag. While the cross was originally used by the Swiss for religious reasons, he notes, it is no longer so viewed. Thus, “an image that began as an expression of faith was transformed.”[77] This same transformation happened in Bladensburg after World War I—when “a plain Latin cross . . . also took on new meaning.”[78] To determine that meaning, Justice Alito discusses the use of crosses in military cemeteries, the association of the cross with World War I dead, and the use of the cross in contemporaneous literature, poetry, and art.[79] This deep dive into the cultural history of the World War I-era cross is not gratuitous; the entire point of the exercise is to show that the Bladensburg Cross was and is primarily a symbol of the war dead and not primarily of Christianity or Christian superiority.

This discussion matters, even as some of the justices insist that it does not. As we have seen, after establishing the secular meaning of the cross, the Court describes various difficulties with applying the Lemon and endorsement tests. There are the usual problems of determining the intentions of state actors, along with the proliferation of public purposes over time, and the different meanings that may come to be associated with government-sponsored religious symbols.[80] Famous cathedrals, religiously-infused place-names, and other religious expressions become embedded in the cultural landscape and “[f]amiliarity itself can become a reason for preservation.”[81]

The Court raises these as objections to a meaning-based judicial doctrine for all long-standing government religious symbols. But one can ask why these problems cannot be overcome. All the examples the Court offers are relatively uncontroversial; they sometimes read as a parade of horribles, but none of the examples are particularly difficult. The secular-purpose prong of Lemon and the endorsement test, for all their flaws, provided a standard by which judges could assess the meaning of government speech, some of which did and would violate the Establishment Clause.[82]

And in fact, the Court applies a meaning-based analysis in the very case in which it eschews it, beginning its opinion by draining the Bladensburg Cross of its primary religious meaning. On an endorsement analysis, the case could have ended there. If the Bladensburg Cross is a mostly secular symbol of the war dead, then it likely does not convey a message of outsider status and, on that view, would not violate the Establishment Clause.[83]

B. Switching Perspectives

The Court takes a second step, however, shifting the perspective from which the Cross should be assessed. Once the Cross is drained of its religious purpose and meaning, the Court turns to assessing the meaning of removing the Cross from the perspective of those who would favor keeping it. This is the moment of Establishment Clause inversion. In a striking passage, Justice Alito writes:

[W]hen time’s passage imbues a religiously expressive monument, symbol, or practice with this kind of familiarity and historical significance, removing it may no longer appear neutral, especially to the local community for which it has taken on particular meaning. A government that roams the land, tearing down monuments with religious symbolism and scrubbing away any reference to the divine will strike many as aggressively hostile to religion. Militantly secular regimes have carried out such projects in the past, and for those with a knowledge of history, the image of monuments being taken down will be evocative, disturbing, and divisive.[84]

This rhetoric, involving the government “roam[ing] the land,” “tearing down monuments,” and “scrubbing away” references to the divine, is provocative and purposefully so. The reference to “militantly secular regimes” is meant to put the challengers to public religious symbols on the defensive. This language turns a disagreement about public meaning into a conflict over ideologies, with “militants”—presumably radical secularists—representing those who would insist on erasing public references to the divine.

Who exactly would be disturbed by the potential removal of the Cross? The endorsement test has generally been applied from the perspective of the reasonable observer, with some debate about whether that perspective should be defined in terms of a representative member of the political community or by the views of religious (or non-religious) minorities.[85] In either case, the endorsement inquiry concerns the meaning of the contested government symbol itself—in American Legion, what message the Cross conveys to those who observe it. But in Justice Alito’s inverted Establishment Clause, the government act under review is not display of the Cross, but rather the possibility that the Court might order its removal. The majority emphasizes the message being sent to those who support the Cross. Instead of asking what message the government sends by displaying the Cross, the Court asks what message it would send if it held that the Cross is unconstitutional.

This inversion of the Establishment Clause in American Legion is inconsistent with a central purpose of the First Amendment, which is to protect religious minorities from being subject to majoritarian religious practices. Those practices, including symbolic speech, are not supposed to signal the lesser status of nonadherents. There are many reasons we might be concerned about government symbolic speech that has this social meaning. Such speech might cause the government or other citizens to treat nonadherents differently in providing access to material, economic, or social benefits.[86] Government speech that endorses a particular religion (or religion generally) might also induce a sense of alienation among nonadherents, or it might stigmatize them, causing psychic harm.[87] Government actions, including symbolic speech, that convey a message of exclusion might also be wrong because they violate a constitutional requirement of treating all citizens with equal concern and respect.[88]

These are constitutional harms that follow from the government’s endorsement of religious symbols. But we do not often ask about the harms that might result if the government never engages in, or is forbidden from engaging in, religious speech. And why would we? Presumably, the harm of removing the Cross would be experienced by those who favor the government’s display of it. But a central idea of disestablishment is that there should be no favored religion. Messages of insider status are constitutionally inappropriate. Of course, those who enjoy such a message will be harmed when the government withdraws that message or is disallowed from expressing it in the first place. But if those harms count, then what becomes of disestablishment? As opposed to a check on religious favoritism, the Establishment Clause becomes a way of entrenching that preference.

In assessing the expressive effects of removing the Cross, American Legion’s majority starts down this path. The Court could have held that the Bladensburg Cross had a legitimate secular purpose and that it was not an endorsement of religion. It also could have held that the Cross is a de minimis establishment; perhaps regretful, but not so egregious as to require constitutional invalidation. Instead, the Court ruled that Establishment Clause principles require that the Court not order its removal, or for that matter, not order the removal of any long-standing monuments. For those with a favorable view of both the Cross and the meanings that it conveys, “destroying or defacing the Cross that has stood undisturbed for nearly a century would not be neutral and would not further the ideals of respect and tolerance embodied in the First Amendment.”[89]

This statement echoes Justice Breyer’s concern with religious divisiveness, most fully articulated in Van Orden v. Perry, which upheld the display of a large Ten Commandments monument on state-owned property.[90] But the application here goes one step further. In Van Orden, Justice Breyer asserted that a central purpose of the Establishment Clause is to prevent religious conflict.[91] In assessing the constitutionality of the Ten Commandments display in that case, he balanced the harms of the religious display against the harms of an order to remove it. In light of the potential political backlash that might follow from removal, he concluded that such an order would do more harm than good. He made a political calculation.

Justice Alito’s argument in American Legion appears to be different. If the Establishment Clause requires religious neutrality and removing the Cross would not be neutral, then it is possible that Bladensburg would be barred under the Establishment Clause from removing the Cross, even absent a court order. Perhaps those who favor the display have some right against government interference with it, such that any deviation from the existing baseline would express hostility toward them. Therefore, existing religious monuments must remain. Another possibility is that neutrality requires equal expressive treatment in the public square. Cities and towns across the nation cannot remove those monuments that already exist. But must they also build religious monuments when they put up secular ones, and would failure to do so show hostility toward religion?

The Court has already moved sharply in the direction of equal treatment in the area of religious funding, holding that neutrality demands equal access for equivalent secular and religious entities.[92] That neutrality principle has also been applied to private religious speech in public forums.[93] Of course, the Court does not hold that the state must preserve the Cross in American Legion. In his concurring opinion, Justice Kavanaugh makes this point explicitly.[94] The majority’s reasoning, however, raises the question of whether a voluntary decision by the state to remove the Cross would convey hostility toward religion. The Court’s holding—that removing the Cross would violate religious neutrality—turns the Establishment Clause on its head, and its logic points toward an even more complete and radical inversion.

C. Neutrality as Hostility

What is clear from American Legion is that government messages matter. The government conveys a message when it puts up a cross. It also conveys a message when it takes one down. Presumably the neutral position is for the government not to say anything about religion. But that is not how the American Legion majority sees it. The government cannot be neutral. Even when it says nothing, it might be conveying a message. Does the government exhibit hostility to religion when it refuses to erect a cross or a Christmas tree or a creche, or stops doing so after many years? These are the types of questions about social meaning that the Court’s decision raises.

In American Legion, Justice Alito mistakes the concept of neutrality for hostility toward religion. This is an ongoing theme in the rhetoric and doctrine of Establishment Clause jurisprudence. The rhetoric originates in early Burger Court cases in which the Court upheld religious displays in part on grounds that an aggressive judicial stance could be interpreted as hostility to religion; the state should instead exhibit a supportive relationship to religion, without running afoul of the Establishment Clause.[95]

This trope has been common among religious conservatives, who have long argued that separationist judicial decisions evince hostility to religion, and that religious entities, practices, and messages should receive treatment at least as favorable as their secular analogues. Certainly, any lesser treatment, they argue, is both denigrating of religion and non-neutral.[96]

What concept of neutrality is doing the work here? Disestablishment imposes special disabilities on religion that do not apply to many secular beliefs and practices. Some have argued that a principle of disestablishment should extend further to cover certain secular doctrines,[97] though the Court continues to treat religion as special for purposes of government speech, at least currently. But if neutrality means that the government must, even in its own speech, give equally favorable treatment to religion, then we have come a long way from the idea of disestablishment.

Some of Justice Alito’s rhetoric sounds in the cultural and political register of victimization.[98] Consider the “War on Christmas.”[99] Instead of being understood as an effort to be inclusive, non-specific holiday greetings are targeted for their hostility to Christians. “Neutral” greetings are suddenly non-neutral. In other areas, too, religious conservatives have argued that laws requiring neutrality instead evince hostility toward or discriminate against them. Anti-discrimination laws that protect LGBT people, for example, are challenged as hostile to traditional believers.[100] Similarly, removing a cross in order to be more inclusive of non-Christians is reinterpreted as animosity toward Christians.

Perhaps this reflects a transitional moment. In a society dominated by a particular religious outlook, namely one in which Christians exercise cultural, political, and social power, Establishment Clause doctrine could reflect a clear sense of who constitutes the majority and who constitutes the minority and how the practices of the former might fuel alienation in the latter. In a society in which Christians are less dominant, or which is conceived of as dominated by a “religion” of secularism, the alienation might run in the other direction. From Justice Alito’s perspective, by imposing its will on the public square, a secular majority is signaling its disapproval of traditional believers, and of Christians, in particular. The majority becomes the minority. And neutrality becomes hostility.

The assumed dominance of a secularist majority seems overstated. Traditional believers, including conservative Christians, do not appear to be a particularly weak or insular minority, despite what some have asserted. The deference to long-standing religious symbols that the Court adopts in American Legion instead seems to be responsive to a political problem, and that should be cause for regret. Justice Breyer’s Van Orden concurrence is an acknowledgement of the Court’s limited capacity to act effectively as a counter-majoritarian force. A decision against exclusionary government speech risks a political backlash, increasing religious polarization, and consequently less toleration of minorities. On Justice Breyer’s account, the Establishment Clause needs to be read so as to manage religious strife.[101]

Turning this political concern into a substantive argument for neutrality, as Justice Alito does, is a mistake, however. The move to equate non-neutrality with hostility advances a radical vision of the Establishment Clause, in which anything less than favorable treatment of religion constitutes an attack by “militant” secularists. The Establishment Clause thereby becomes an instrument for requiring that the government support and endorse religious activities, practices, and symbolic speech, where before it had served as a limit on that support and endorsement.

IV. After Endorsement

In light of the Court’s criticism of Lemon and the endorsement test, and despite its engaging in a similar meaning-laden analysis, it is fair to ask what limits the Court would impose on the government’s religious speech after American Legion, and what criteria courts should use to set those limits.

One possibility is that there are no limits. If judicial constraints on government religious symbols and practices evince hostility toward religion, government religious speech may be protected by the Establishment Clause, not threatened by it. The majority could also adopt Justice Gorsuch’s or Justice Thomas’s respective approaches. Recall that Justice Gorsuch argues that “observer standing” under the Establishment Clause is an anomaly and that noncoercive government speech should not be vulnerable to constitutional challenge.[102] Justice Thomas argues that the Establishment Clause does not apply to the states and that, even if it were applicable, it would only prohibit coercive government acts, not symbolic ones.[103]

The Court does not seem inclined to embrace these views. Most of the justices appear to contemplate at least some circumstances in which noncoercive government religious speech would violate the Establishment Clause. What the Court does state, repeatedly, is that there is a presumption of constitutionality for “longstanding” monuments, and perhaps for other traditional state-sponsored religious practices as well.[104]

What constitutes “longstanding” will be an on-going issue, of course. “Longstanding” might have two components. It seems to refer to those religiously-infused government practices that have 1) shed any primary sectarian meaning and have become so embedded in the civic fabric that 2) striking them down would convey a message of hostility toward religion. The old/new distinction seems intended to prevent the erection of new monuments or the adoption of new religiously-infused practices. But it is not entirely clear if “longstanding” refers to actual existing monuments or practices, or to government monuments and practices that have some grounding in historical practice. If the government has historically invoked the divine at town council meetings or during Thanksgiving celebrations, may it continue to do so next year? Presumably. But can it also erect new monuments, as long as those monuments do not depart from norms that have been historically accepted?

Justices Breyer and Kagan signal that they would resist new construction.[105] Perhaps some other justices would as well. The majority seems nervous that its opinion could be read to authorize a spate of cross building across the country. Perhaps the Court has an image of crosses on top of courthouses, in courtrooms, or on classroom walls.[106] Former Alabama Supreme Court Chief Justice Roy Moore’s placement of a two-ton Ten Commandments monument in the Alabama state courthouse still seems to be constitutionally out-of-bounds, even for some of the conservative justices.[107]

Nevertheless, state and local legislators regularly propose new laws for the display of the Ten Commandments in schools, courtrooms, and on other government property.[108] Several states have proposed laws that would require government buildings to post displays stating, “In God We Trust.”[109] That state and local officials choose these particular symbols and messages is neither accidental nor surprising. Those pursuing a public religious agenda are aware of the Supreme Court’s decisions, and they know that certain symbols and messages may be constitutionally permissible, at least in certain contexts.[110] The Court is right to worry about new crosses. Political entrepreneurs are happy to push the boundaries of acceptable religious speech once the Court has given them an opening.

The old/new distinction is not going to end the religious culture wars, despite what may be an effort by the Court to withdraw from them, at least in the context of government religious speech. Justices Breyer and Kagan, who joined the conservatives in the majority, must know this. They both dissented in Town of Greece v. Galloway, which upheld a town council’s practice of commissioning mainly Christian prayers to open its meetings.[111] This prayer practice was new, though the practice of opening a legislative session with prayers is quite old, as the majority noted.[112]

Justice Kagan’s dissent in Town of Greece makes clear what is at stake in these cases of state-sponsored religious expression. Though she does not explicitly invoke the endorsement test, her argument sounds in the same expressive register.[113] The state must not convey a message of exclusion that denies equal citizenship to those of different faiths. “A Christian, a Jew, a Muslim (and so forth)—each stands in the same relationship with her country,” she writes, “[s]o that when each person performs the duties or seeks the benefits of citizenship, she does so not as an adherent to one or another religion, but simply as an American.”[114] Government religious speech, especially when overwhelmingly sectarian, undermines that relationship of political equality.[115]

The justices in Town of Greece disagreed about the meaning and effects of the prayer practices in that case. But whether the justices agree on the particular facts, this form of analysis—taking into account the social meaning of religious expression—seems unavoidable. If the Court is going to strike down some types of government religious speech, it cannot avoid analyzing the meaning of the speech and the message it conveys to the reasonable citizen.[116] Roy Moore’s Ten Commandments monument sends a message that the Alabama Supreme Court is aligned with a particular religious faith to the exclusion and subordination of religious and nonreligious minorities. Otherwise, why treat it differently from the Ten Commandments frieze in the U.S. Supreme Court? Some of the justices, perhaps a majority, seem to believe that new crosses convey different messages than old ones. Though the Court disavows inquiries into the meaning of government symbols, it is difficult to imagine how else the justices could make these distinctions.

V. Christian Preferentialism?

Of course, it is possible that the Court’s reticence to police the public square will result in the proliferation of government religious expression, presumably in ways that endorse Christianity or certain forms of Christianity. This outcome is not predicted by the conventional political economy story of the Establishment Clause, which asserts that as the country becomes more religiously diverse, state-sponsored religious symbols that appear to favor one religion over others become less acceptable. Consider John Jeffries’s and James Ryan’s A Political History of the Establishment Clause.[117] They cite increasing religious pluralism and the decline of Protestant-Catholic tensions for the shift from a no-aid to a neutral-aid principle. As long as the government is even-handed, it should be able to fund religious organizations as it funds secular ones. At the same time, Jeffries and Ryan argue that increasing pluralism coupled with rising religious non-affiliation means that a return to a more assertive public cultural Christianity is unlikely.[118] In other words, increasing pluralism tends to support government funding of religious institutions, but not government-sponsored religious speech. Ryan and Jeffries thus predicted that the corresponding aspects of Establishment Clause doctrine would eventually diverge.[119]

The terrain of the religious culture wars has shifted considerably in the last few decades, however. The pluralism narrative does not account for two phenomena that have characterized politics in the early twenty-first century: the robust alliance of religious conservatives against the secular state and the rise of ethno-religious populism on a global scale. These forces may push church-state relations in the U.S.—as they have elsewhere—in a reactionary direction, leading to more preferentialism in the public square rather than less.[120]

A. The Attack on Secularism and the Rise of Christian Nationalism

The Catholic-Protestant tensions that characterized the early-to-mid-twentieth century church-state settlement have been replaced by alignment of religious conservatives, on one side, and liberal or progressive religious minorities and nonbelievers, on the other.[121] Traditional believers—mainly evangelical Protestants and conservative Catholics—have found common cause in their opposition to the sexual revolution. The LGBT civil rights movement, in particular, has mobilized an inter-denominational conservative coalition that had already been unified by opposition to abortion and contraception.[122] The Supreme Court’s decision to constitutionalize same-sex marriage in Obergefell, the expansion of civil rights laws to protect LGBT persons, and federal regulations mandating contraceptive coverage under the Affordable Care Act—these developments have led many religious conservatives to believe that they are culturally and politically under siege.[123] On the religious right, there is a shared sense of victimization at the hands of a relentlessly aggressive secular culture, leading conservatives to position themselves as oppressed minorities.[124] And for some, the proper response to these perceived threats is a more forceful public assertion of traditional forms of Christianity.[125]

This religious polarization coincides with rising ethno-nationalism, both in the United States and abroad. President Trump’s anti-immigrant policies and rhetoric, which target Latinos, Muslims, and other minorities,[126] have much in common with the programs of other ethno-religious populist movements.[127] In response to Muslim immigration and the perceived threat of Islam, countries like Hungary and Poland are promoting the practices and institutions of so-called “illiberal democracy.”[128] There, opposition to Islam has become a defining characteristic of religious, ethnic, and political identity. Islamophobia has been accompanied by a resurgence of Christian nationalism. Religious conservatives, or at least a significant subset of them, have become chief defenders of nativist regimes.[129]

This latter impulse suggests how the pluralism thesis might be mistaken, and how increasing religious diversity in Western nations may lead to a resurgence of populist Christianity. The perceived Islamic (and non-white) threat has led to calls for banning religious sites, outlawing Islamic law, enforcing restrictions on religious dress, and requiring outward assertions of fealty to Christian symbols.[130] Instead of resulting in less public support for official displays of Christianity, existing forms of pluralism have led to a popular backlash in the form of a more full-throated defense of Christian (or “Judeo-Christian”[131]) values, including efforts to coerce compliance with traditional religious norms.

President Trump’s travel ban is an assertion of Christian nationalism, in parallel with the practices of right-wing parties throughout Europe to restrict or eliminate immigration from the Middle East and to shut the door to Muslim refugees and asylum seekers. His repeated statements denigrating Muslims and his stoking of anti-Islamic sentiments in the larger public play on the fears of white Christians for their safety and cultural security. Controlling entry is a way of short-circuiting pluralism and is part of a larger agenda of preserving an imagined racial, ethnic, and religious American identity.[132]

B. The Supreme Court’s Role

What role is the Court playing in this cultural and political environment? The Roberts Court is continuing a revolution that began with the Rehnquist Court. The narrowing of the Establishment Clause and the expansion of free exercise—at least as applied to traditional believers—has been a goal of conservatives for some time. These doctrinal changes seem in some ways disconnected from global conflicts over ethno-religious identity. And though the Supreme Court is often sharply divided on matters of religious freedom, a number of recent decisions have enjoyed solid majorities, including members of the Court’s liberal wing. Certainly, the Court is on a conservative trajectory, slowly dismantling Establishment Clause restrictions that have characterized the last half-century of church-state jurisprudence. Nevertheless, one could view these shifts in doctrine as a relatively minor departure, an appropriate corrective, or part of the inevitable swings that accompany changes in the Court’s personnel.

A more skeptical view—and one we share—is that the Court’s doctrine is paving the way for a certain kind of religious preferentialism. Before his death, Justice Scalia was explicit about this agenda. He believed that legislatures could adopt laws that reflect particular religious commitments and that the government could promote religious speech.[133] His preferred Establishment Clause doctrine rejected the principle of neutrality among religious denominations. He would have permitted the state to favor monotheistic religions over non-monotheistic ones, and Christian (or “Judeo-Christian”) public practices over others.[134] He rejected the Lemon test’s “secular purpose” requirement[135] and would have replaced the endorsement test with an interpretation of the Establishment Clause that prohibited only government coercion of religious practices and perhaps proselytization.[136]

Justice Scalia’s views have now moved much closer to the center of the Court, as we have seen. Though the Court is not yet willing to adopt an explicit preference for monotheism, it is willing to permit a larger range of preferential government religious practices. Sometimes the Court does so after paying lip service to the demands of nondiscrimination and denominational neutrality,[137] or by invoking private choice as a way to insulate the government from directly supporting particular religious groups.[138] In other cases, the Court has adopted more rigorous standing requirements that make it increasingly difficult to bring Establishment Clause challenges in the first place.[139]

While the Court’s conservative majority may not be actively promoting an ethno-religious vision, its decisions have opened the door for political constituencies pressing in that direction. The Court cannot by itself advance the cause of religious preferentialism. But its decisions can permit religiously-aligned political majorities to achieve aims consistent with such preferentialism, freed from the constraints of the Establishment Clause. As conservative political discourse moves away from pluralism and toward ethno-religious nationalism, the Court has signaled its retreat from requirements of religious neutrality, except insofar as the government appears hostile to religion.


When the courtroom audience in Salazar v. Buono laughed after Mr. Eliasberg said that there is a never a cross on a Jewish tombstone, it was an appropriate response. Both the laughter and the observation that provoked it reflected a common understanding about the meaning of the cross. The suggestion that a cross can be a universal symbol is both offensive—to many Christians and non-Christians—and, more importantly, a statement of civic exclusion. But that view—our view—is now a dissenting perspective. It is remarkable that within a few short years, the absurdity of treating the cross as a secular symbol has become Supreme Court precedent, entrenched by an inversion of the Establishment Clause that equates religious neutrality with hostility to religion.

If Justice Scalia had lived so long, he would have had the last laugh.

Schragger is the Perre Bowen Professor of Law, University of Virginia School of Law.

Schwartzman is the Hardy Cross Dillard Professor of Law and Martha Lubin Karsh and Bruce A. Karsh Bicentennial Professor of Law, University of Virginia School of Law.

[1] Salazar v. Buono, 559 U.S. 700 (2010).

[2] Id. at 706.

[3] Dahlia Lithwick, Cross-Eyed: The High Court Looks Again at Religious Symbols on Public Lands, Slate (Oct. 7, 2009, 7:14 PM); see also Bill Mears, Justices Weigh Constitutionality of War Memorial Cross, CNN (Oct. 8, 2009).

[4] Transcript of Oral Argument at 38–39, Buono, 559 U.S. 700 (No. 08–472).

[5] See Adam Liptak, Religion Largely Absent in Argument About Cross, N.Y. Times (Oct. 7, 2009).

[6] Transcript of Oral Argument at 40, Buono, 559 U.S. 700 (No. 08–472).

[7] The Court had to return to this question because Salazar v. Buono did not resolve it. The Court instead remanded for further proceedings on whether a federal statute transferring the Mojave Desert Cross onto private property violated the Establishment Clause. Buono, 559 U.S. at 722.

[8] Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067 (2019).

[9] We also follow the Court’s practice of referring to this particular monument as the “Cross.” Id. at 2074.

[10] Id. at 2089–90 (“That the cross originated as a Christian symbol and retains that meaning in many contexts does not change the fact that the symbol took on an added secular meaning when used in World War I memorials.”).

[11] Jack Balkin, From Off the Wall to On the Wall: How the Mandate Challenge Went Mainstream, Atlantic (June 4, 2012) (“Off-the-wall arguments are those most well-trained lawyers think are clearly wrong; on-the-wall arguments, by contrast, are arguments that are at least plausible, and therefore may become law . . . . The history of American constitutional development, in large part, has been the history of formerly crazy arguments moving from off the wall to on the wall, and then being adopted by courts. In the process, people who remember the days when these arguments were unthinkable gape in amazement; they can't believe what hit them.”).

[12] See Steven G. Gey, Life After the Establishment Clause, 110 W. Va. L. Rev. 1, 21–26 (2007); John C. Jeffries, Jr. & James E. Ryan, A Political History of the Establishment Clause, 100 Mich. L. Rev. 279, 280–82 (2001); Ira C. Lupu, The Lingering Death of Separationism, 62 Geo. Wash. L. Rev. 230, 237–49 (1994).

[13] See Mueller v. Allen, 463 U.S. 388 (1983); Witters v. Wash. Dep’t of Servs. for the Blind, 474 U.S. 481 (1986); Zobrest v. Catalina Foothills Sch. Dist., 509 U.S. 1 (1993); Agostini v. Felton, 521 U.S. 203 (1997); Mitchell v. Helms, 530 U.S. 793 (2000).

[14] Zelman v. Simmons-Harris, 536 U.S. 639 (2002).

[15] Id. at 653 (“[T]he program challenged here is a program of true private choice, consistent with Mueller, Witters, and Zobrest, and thus constitutional.”).

[16] See 2 Kent Greenawalt, Religion and the Constitution: Establishment and Fairness 405–14 (2008).

[17] Zelman, 536 U.S. at 652 (holding “that where a government aid program is neutral with respect to religion . . . the program is not readily subject to challenge under the Establishment Clause”); see also Douglas Laycock, Churches, Playgrounds, Government Dollars—and Schools?, 131 Harv. L. Rev. 133, 140 (2017) (“The tide turned in 1986, and for more than thirty years now, the Court has been moving toward the view that government funding of secular services, including education, can flow to religious providers so long as it is distributed in religiously neutral ways.”).

[18] See Nelson Tebbe, Excluding Religion, 156 U. Penn. L. Rev. 1263, 1285–86 (2008) (discussing exclusions of religious organizations from public funding programs, including vouchers).

[19] Trinity Lutheran Church of Columbia v. Comer, 137 S. Ct. 2012 (2017).

[20] Id. at 2024–25.

[21] See e.g., Good News Club v. Milford Cent. Sch., 533 U.S. 98 (2001); Rosenberger v. Rectors & Visitors of Univ. of Va., 515 U.S. 819 (1995); Lamb’s Chapel v. Ctr. Moriches Union Free Sch. Dist., 508 U.S. 384 (1993); Widmar v. Vincent, 454 U.S. 263 (1981); see also 2 Greenawalt, supra note 16, at 196–206 (discussing equal access cases).

[22] Trinity Lutheran, 137 S. Ct. at 2021–22. The Court left open whether the government can exclude religious organizations when public funds are used to support religious activities. See id. at 2024 n.3 (“This case involves express discrimination based on religious identity with respect to playground resurfacing. We do not address religious uses of funding or other forms of discrimination.”).

[23] See Ira C. Lupu & Robert W. Tuttle, Trinity Lutheran Church v. Comer: Paradigm Lost?, 1 Am. Const. Soc’y Sup. Ct. Rev. 131, 133 (2017) (“[Trinity Lutheran] represents a stunning and thoroughly unacknowledged move from the religion-distinctive principle of ‘no funding’ to one of nondiscrimination.”).

[24] Lynch v. Donnelly, 465 U.S. 668, 688 (1984) (O’Connor, J., concurring).

[25] See County of Allegheny v. ACLU, 492 U.S. 573, 593–94 (1989); Santa Fe Ind. Sch. Dist. v. Doe, 530 U.S. 290, 309–10 (2000); McCreary Cty. v. ACLU of Ky., 545 U.S. 844, 866 (2005).

[26] Lynch, 465 U.S. at 688.

[27] Town of Greece v. Galloway, 572 U.S. 565 (2014).

[28] Id. at 573; see also Caroline Corbin, Christian Legislative Prayers and Christian Nationalism, 76 Wash. & Lee L. Rev. 451, 455–56 (2019).

[29] See County of Allegheny, 492 U.S. at 661–63 (Kennedy, J., dissenting).

[30] See, e.g., Marsh v. Chambers, 463 U.S. 783, 790 (1984); McCreary Cty., 545 U.S. at 886–94 (Scalia, J., dissenting); Town of Greece, 572 U.S. at 603 (Alito, J., concurring).

[31] See Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014). For discussion of Hobby Lobby and its implications, see Chad Flanders, Micah Schwartzman & Zoë Robinson, Introduction to The Rise of Corporate Religious Liberty xiii (Micah Schwartzman et al. eds., 2016); Elizabeth Sepper, Free Exercise Lochnerism, 115 Colum. L. Rev. 1457 (2015); Douglas NeJaime & Reva B. Siegel, Conscience Wars: Complicity-Based Conscience Claims in Religion and Politics, 124 Yale L.J. 2516 (2015).

[32] Cutter v. Wilkinson, 544 U.S. 709, 724 (2005) (internal quotations omitted).

[33] Id. at 720 (2005). See Micah Schwartzman, Nelson Tebbe & Richard Schragger, The Costs of Conscience, 106 Ky. L. J. 781 (2018) (discussing limits on religious exemptions based on third-party harms).

[34] Some lower courts have rejected exemptions that impose significant burdens on third parties, especially in cases involving for-profit corporations. See, e.g., Pennsylvania v. Trump, 930 F.3d 543, 574 (3d Cir. 2019) (affirming preliminary injunction in part because the Trump administration’s contraception exemption would “impose an undue burden on nonbeneficiaries—the female employees who will lose coverage for contraceptive care”); California v. Health & Human Servs., 351 F. Supp. 3d 1267, 1295–96 (N.D. Cal. 2019) (relying on Cutter’s requirement to consider third-party harms in granting a preliminary injunction against religious exemption to the contraceptive mandate). Some members of the Supreme Court have expressed skepticism about limiting religious exemptions on the basis of third-party harms. See Hobby Lobby, 573 U.S. at 729 n.37 (Alito, J.). But cf. Nelson Tebbe, Micah Schwartzman & Richard Schragger, When Do Religious Accommodations Burden Others?, in The Conscience Wars: Rethinking the Balance Between Religion, Identity, and Equality 328, 340–45 (Susanna Mancini & Michel Rosenfeld eds., 2018) (discussing Justice Alito’s treatment of third-party harms in Hobby Lobby).

[35] See Leslie Kendrick & Micah Schwartzman, The Etiquette of Animus, 132 Harv. L. Rev. 133, 154–62 (2018) (discussing religious exemptions from civil rights laws after Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719 (2018)).

[36] Lemon v. Kurtzman, 403 U.S. 602, 612–13 (1971).

[37] Lynch v. Donnelly, 465 U.S. 668, 688 (1984) (O’Connor, J., concurring).

[38] See Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067, 2080 (collecting Establishment Clause cases in which the Court “declined to apply the [Lemon] test or simply ignored it”).

[39] The Court has also cut back significantly on taxpayer standing to raise Establishment Clause challenges. See Ira C. Lupu & Robert W. Tuttle, Ball on a Needle: Hein v. Freedom from Religion Foundation, Inc., and the Future of Establishment Clause Adjudication, 2008 B.Y.U. L. Rev. 115 (2008); William P. Marshall & Gene R. Nichol, Not a Winn-Win: Misconstruing Standing and the Establishment Clause, 2011 Sup. Ct. Rev. 215, 216 (2011).

[40] See Richard C. Schragger, The Relative Irrelevance of the Establishment Clause, 89 Tex. L. Rev. 583 (2011).

[41] There is a deep tension between the neutrality principles that the Court holds applicable in the funding and speech contexts and the special treatment it provides to religious entities in accommodation cases. See Richard Schragger & Micah Schwartzman, Religious Antiliberalism and the First Amendment, 103 Minn. L. Rev. (forthcoming 2019).

[42] See, e.g., Garrett Epps, Why Is this Cross-Shaped Memorial Constitutional?, Atlantic (Feb. 19, 2019) (“Barring the unlikely event—one hesitates to use the word miracle in this context—of a decision against the cross, the Court has a lot of paths through this litigation, and the question is how far the new majority wants to go.”).

[43] Am. Legion, 139 S. Ct. at 2068–74. In Parts I, II-B, II-C, III, and IV, Justice Alito wrote for a majority, including Chief Justice Roberts and Justices Breyer, Kagan, and Kavanaugh. But Justice Kagan refused to join Parts II-A and II-D, which represent a plurality of the remaining four justices.

[44] Id. at 2089.

[45] Id. at 2082.

[46] Id. at 2082–83.

[47] Id. at 2084–85.

[48] Id. at 2085.

[49] Id.

[50] Id. at 2087.

[51] Id.

[52] Marsh v. Chambers, 463 U.S. 783 (1983); Town of Greece v. Galloway, 572 U.S. 565 (2014).

[53] Am. Legion, 139 S. Ct. at 2089 (plurality opinion).

[54] Id.

[55] Id. at 2092 (Kavanaugh, J., concurring).

[56] Id. at 2093.

[57] Id. at 2093 n.*.

[58] Id. at 2091 (Breyer, J., concurring).

[59] Id.

[60] Id. at 2094 (Kagan, J., concurring in part).

[61] See id. at 2097–98 (Thomas, J., concurring in the judgment); id. at 2101 (Gorsuch, J., concurring in the judgment).

[62] Id. at 2095. See also Elk Grove Unified Sch. Dist. v. Newdow, 542 U.S. 1, 49–51 (2004) (Thomas, J., concurring in the judgment); Van Orden v. Perry, 545 U.S. 677, 692–93 (2005) (Thomas, J., concurring); Zelman v. Simmons-Harris, 536 U.S. 639, 677–80 (2002) (Thomas, J., concurring).

[63] Am. Legion, 139 S. Ct. at 2096 (internal quotation marks omitted) (Thomas, J., concurring in the judgment).

[64] Id.

[65] Id. at 2098–03 (Gorsuch, J., concurring in the judgment).

[66] Id. at 2098–01.

[67] Id. at 2113 (Ginsburg, J., dissenting).

[68] Id. at 2106.

[69] Id. at 2109.

[70] Id. at 2106–07.

[71] Id. at 2074 (internal quotation marks omitted) (majority opinion).

[72] Id.

[73] See Richard Schragger, Of Crosses and Confederate Monuments: Considering the Constitutional Limits on Majoritarian Control of the Public Square (unpublished manuscript) (on file with authors).

[74] Am. Legion, 139 S. Ct. at 2074.

[75] Id.

[76] Id. at 2075.

[77] Id.

[78] Id.

[79] Id. at 2075–77.

[80] Id. at 2082–85.

[81] Id. at 2084.

[82] For qualified defenses of Lemon and the endorsement test as reflecting sensible approaches to government religious speech, see 2 Greenawalt, supra note 16, at 158–81; see also Nelson Tebbe, Religious Freedom in an Egalitarian Age 98–112 (2017) (defending a principle of government nonendorsement).

[83] Of the opinions offered in American Legion, Justice Kagan’s seems nearest to this view. See Am. Legion, 139 S. Ct. at 2094 (Kagan, J., concurring in part).

[84] Id. at 2084–85 (majority opinion).

[85] See B. Jessie Hill, Anatomy of the Reasonable Observer, 79 Brook. L. Rev. 1407, 1416–17 (2014); Caroline Mala Corbin, Ceremonial Deism and the Reasonable Observer, 57 UCLA L. Rev. 1545, 1596–97 (2010).

[86] See Caroline Mala Corbin, Nonbelievers and Government Speech, 97 Iowa L. Rev. 347, 392–99 (2012).

[87] See Marshall & Nichol, Not a Winn-Win, supra note 39, at 232–46 (arguing that a purpose of the Establishment Clause is to protect citizens from psychic harms).

[88] See Christopher L. Eisgruber & Lawrence G. Sager, Religious Freedom and the Constitution 122–27 (2007); Nelson Tebbe, Government Nonendorsement, 98 Minn. L. Rev. 648 (2012); Michael C. Dorf, Same-Sex Marriage, Second-Class Citizenship, and Law’s Social Meanings, 97 Va. L. Rev. 1267 (2011); Deborah Hellman, The Expressive Dimension of Equal Protection, 85 Minn. L. Rev. 1 (2000).

[89] Am. Legion, 139 S. Ct. at 2090.

[90] Van Orden v. Perry, 545 U.S. 677, 698–705 (2005) (Breyer, J., concurring in the judgment).

[91] Id. at 698 (“[The Religion Clauses] seek to avoid that divisiveness based upon religion that promotes social conflict, sapping the strength of government and religion alike.”).

[92] See supra Part I.

[93] See, e.g., Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819 (1995).

[94] Am. Legion, 139 S. Ct. at 2094 (Kavanaugh, J., concurring) (“The Court’s ruling allows the State to maintain the cross on public land. The Court’s ruling does not require the State to maintain the cross on public land.”).

[95] See, e.g., Lynch v. Donnelly, 465 U.S. 668, 673 (1984) (“Nor does the Constitution require complete separation of church and state; it affirmatively mandates accommodation, not merely tolerance, of all religions, and forbids hostility toward any.”).

[96] See, e.g., McCreary Cty. v. ACLU of Ky., 545 U.S. 844, 900–01 (2005) (Scalia, J., dissenting); Michael W. McConnell, Religious Freedom at a Crossroads, 59 U. Chi. L. Rev. 115, 124–27 (1992); cf. Frank S. Ravitch, The Supreme Court’s Rhetorical Hostility: What Is “Hostile” to Religion Under the Establishment Clause, 2004 B.Y.U. L. Rev. 1031 (2004) (discussing uses of the concept of hostility in Establishment Clause jurisprudence).

[97] See Tebbe, Government Nonendorsement, supra note 88, at 657–92; Micah Schwartzman, What If Religion Is Not Special?, 79 U. Chi. L. Rev. 1351, 1421–25 (2012).

[98] Cf. Melissa Murray, Inverting Animus: Masterpiece Cakeshop and the New Minorities, 2018 Sup. Ct. Rev. 257, 282 (“The oppressed victim of discrimination is no longer the ‘discrete and insular minorities’ contemplated in Carolene Products, but rather religious objectors who were once trumpeted as a ‘moral majority,’ but now cloak themselves as ‘religious minorities’ in need of state protection.”).

[99] See, e.g., John Gibson, The War on Christmas: How the Liberal Plot to Ban the Sacred Christian Holiday Is Worse than You Thought (2005).

[100] See, e.g., Masterpiece Cakeshop v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719, 1735–40 (2018) (Gorsuch, J., concurring); Christian Legal Soc'y Chapter of the Univ. of Cal., Hastings Coll. of the Law v. Martinez, 561 U.S. 661, 725–27 (2010) (Alito, J., dissenting); Steven D. Smith, Pagans and Christians in the City: Culture Wars from the Tiber to the Potomac 341–43 (2018); Michael McConnell, Dressmakers, Bakers, and Equality of Rights, in William N. Eskridge, Jr. & Robin Fretwell Wilson, Religious Freedom, LGBT Rights, and the Prospects for Common Ground 381–82 (2019).

[101] Schragger, The Relative Irrelevance of the Establishment Clause, supra note 40, at 617–18.

[102] Am. Legion v. Am. Humanist Soc’y, 139 S. Ct. 2067, 2100–04 (2019) (Gorsuch, J., concurring).

[103] Id. at 2096–97 (Thomas, J., concurring).

[104] Id. at 2082, 2085, 2089 (plurality opinion).

[105] Id. at 2091 (Breyer, J., concurring) (“Nor do I understand the Court’s opinion today to adopt a ‘history and tradition test’ that would permit any newly constructed religious memorial on public land.”).

[106] Cf. Salazar v. Buono, 559 U.S. 700, 715 (2010) (“[T]he [Establishment] Clause forbids a city to permit the permanent erection of a large Latin cross on the roof of city hall . . . because such an obtrusive year-round religious display would place the government’s weight behind an obvious effort to proselytize on behalf of a particular religion.” (quoting County of Allegheny v. ACLU, 492 U.S. 573, 661 (1989)).

[107] See Frederick M. Gedicks & Roger Hendrix, Uncivil Religion: Judeo-Christianity and the Ten Commandments, 110 W. Va. L. Rev. 275, 289–90 (2007); Steven Lubet, The Ten Commandments in Alabama, 15 Const. Comment. 471 (1998).

[108] See, e.g., Bill Chappell, Arkansas Installs A New Ten Commandments Monument At Its Capitol, NPR (April 26, 2018, 1:19 PM); Julius L. Lasin, Alabama Amendment Allows Display of Ten Commandments, Montgomery Advertiser (Nov. 7, 2018, 1:02 AM).

[109] Moriah Balingit, Does ‘In God We Trust’ Belong in Schools? More and More States Say Yes., Wash. Post (Dec. 1, 2018).

[110] See Jeremy W. Peters, Alabama Revisits Ten Commandments, Hoping for Help From Kavanaugh, N.Y. Times (Sept. 25, 2018).

[111] Town of Greece v. Galloway, 572 U.S. 565 (2014).

[112] Id. at 575.

[113] See id. at 632–33, 637 (Kagan, J., dissenting).

[114] Id. at 615.

[115] Id. at 621 (“And so a civic function of some kind brings religious differences to the fore: That public proceeding becomes (whether intentionally or not) an instrument for dividing her from adherents to the community’s majority religion, and for altering the very nature of her relationship with her government.”).

[116] See Hill, Anatomy of the Reasonable Observer, supra note 85, at 1408 (arguing that even if the Court rejects the endorsement test, it must still determine the social meaning of government speech based on the perspective of a reasonable observer).

[117] Jeffries and Ryan, supra note 12.

[118] Id. at 366–68; see also Gey, supra note 12, at 42–47 (arguing that religious pluralism will limit the rise of Christian preferentialism).

[119] Jeffries and Ryan, supra note 12, at 368.

[120] See Schragger & Schwartzman, Religious Antiliberalism, supra note 41.

[121] See James Davidson Hunter, Culture Wars: The Struggle to Define America (1991); see also Noah Feldman, Divided by God (2006).

[122] See NeJaime & Siegel, Conscience Wars, supra note 31, at 2544–52.

[123] See Various, Against the Dead Consensus, First Things (March 21, 2019); Ross Douthat, What are Conservatives Actually Debating, N.Y. Times (June 4, 2019).

[124] See, e.g., Smith, Pagans and Christians in the City, supra note 100, at 360–61­; Adrian Vermeule, As Secular Liberalism Attacks the Church, Catholics Can’t Afford to Be Nostalgic, Catholic Herald (Jan. 5, 2018); but cf. Murray, Inverting Animus, supra note 98, at 282–83 (criticizing the inversion of antidiscrimination law that occurs when Christian conservatives are considered “a beleaguered minority religious sect subject to invidious discrimination”).

[125] See, e.g., R. R. Reno, Resurrecting the Idea of a Christian Society (2016); Sohrab Ahmari, Against David French-ism, First Things (May 29, 2019); see also Alexander Zaitchik, Is Josh Hawley For Real?, New Republic (July 25, 2019) (discussing the rise of “post-liberal” conservative populists).

[126] Cf. Corbin, Christian Legislative Prayers, supra note 28, at 467–70.

[127] See generally William A. Galston, Anti-Pluralism: The Populist Threat to Liberal Democracy (2018); Roger Eatwell & Matthew Goodwin, National Populism: The Revolt Against Liberal Democracy (2018).

[128] See Zack Beauchamp, It Happened There: How Democracy Died in Hungary, Vox (Sept. 13, 2018, 9:30 AM); Jan-Werner Müller, The Problem with Poland, N.Y. Rev. of Books (Feb. 11, 2016).

[129] See Andrew Whitehead et al., Making America Christian Again: Christian Nationalism and Voting for Donald Trump in the 2016 Presidential Election, 79 Soc. of Religion 147 (2018); Harry Bruinius, Why Evangelicals Are Trump’s Strongest Travel-Ban Supporters, Christian Sci. Monitor (Mar. 3, 2017); Robert Duncan, Limiting Muslim Immigration Is Patriotic, U.S. Cardinal Says, Nat’l Cath. Rptr. (May 20, 2019); Aleksandra Wróbel, Orbán Pledges to Keep Hungary Safe and Christian, Politico (May 7, 2018, 1:49 PM).

[130] See, e.g., Dan Bilefsky, Quebec Bans Religious Symbols in Some Public Sector Jobs, N.Y. Times (June 17, 2019); Ed Pilkington, Anti-Sharia Laws Proliferate as Trump Strikes Hostile Tone Toward Muslims, Guardian (Dec. 30, 2017, 11:35 AM); Nick Cumming-Bruce & Steven Erlanger, Swiss Ban Building of Minarets on Mosques, N.Y. Times (Nov. 29, 2009).

[131] We are skeptical of the term. See Richard Schragger & Micah Schwartzman, Jews, Not Pagans, 56 San Diego L. Rev. 497, 509 (2019); Anna Gryzmala-Busse, Once, the ‘Judeo-Christian Tradition’ United Americans. Now It Divides Them., Wash. Post (April 17, 2019); Kevin M. Schultz, Arguing in Bad Faith: The Curious Appropriation of “Judeo-Christian” Values on the Right, New Republic (May 9, 2019).

[132] See Corbin, Christian Legislative Prayers, supra note 28, at 472–73; Christopher Stroop, White Evangelicals Have Turned on Refugees, Foreign Pol’y (Oct. 29, 2018, 8:56 AM).

[133] See McCreary Cty. v. ACLU, 545 U.S. 844, 900 (2005) (“[I]n the context of public acknowledgments of God there are legitimate competing interests: On the one hand, the interest of that minority in not feeling ‘excluded’; but on the other, the interest of the overwhelming majority of religious believers in being able to give God thanks and supplication as a people, and with respect to our national endeavors. Our national tradition has resolved that conflict in favor of the majority.”).

[134] Id. at 893 (“If religion in the public forum had to be entirely nondenominational, there could be no religion in the public forum at all. . . . With respect to public acknowledgment of religious belief, it is entirely clear from our Nation’s historical practices that the Establishment Clause permits this disregard of polytheists and believers in unconcerned deities, just as it permits the disregard of devout atheists.”); see Thomas B. Colby, A Constitutional Hierarchy of Religions? Justice Scalia, the Ten Commandments, and the Future of the Establishment Clause, 100 Nw. U. L. Rev. 1097 (2006).

[135] McCreary Cty., 545 U.S. at 900–02.

[136] See id. at 908–9; Lee v. Weisman, 505 U.S. 577, 631, 640–41 (Scalia, J., dissenting).

[137] Town of Greece v. Galloway, 572 U.S. 565, 585–86 (“So long as the town maintains a policy of nondiscrimination, the Constitution does not require it to search beyond its borders for non-Christian prayer givers in an effort to achieve religious balancing.”).

[138] Zelman v. Simmons-Harris, 536 U.S. 639, 652 (2002).

[139] Arizona Christian Sch. Tuition Org. v. Winn, 563 U.S. 125 (2011); Hein v. Freedom from Religion Found., Inc., 551 U.S. 587 (2007).