5th Cir. en banc  ·  No. 25-50695  ·  Decided April 21, 2026  ·  9-8 to reverse

What Is Left of the Wall?

Reading Nathan v. Alamo Heights and the Establishment Clause After Kennedy v. Bremerton

Format: An essay for law students
Companion to: the full case brief and the slip opinion
Length: ~5,500 words / ~25-minute read

I · The Question

In April 2026, a 9-8 en banc Fifth Circuit upheld Texas Senate Bill 10, a statute that requires every public school classroom in Texas to display, in a conspicuous place, a 16-by-20-inch framed copy of the King James Version of the Ten Commandments. Nathan v. Alamo Heights Independent School District, No. 25-50695 (5th Cir. 2026) (en banc). The majority read the Supreme Court's 2022 decision in Kennedy v. Bremerton School District as having abrogated not only Lemon v. Kurtzman but also Stone v. Graham — the 1981 per curiam that had, for forty-five years, prohibited exactly this kind of classroom Decalogue display. The dissenters protested that an inferior federal court cannot assume silent overruling of a Supreme Court case directly on point.

Students reading Nathan should not treat it as an ordinary application of existing doctrine. It is a test case for a bigger proposition: that the Establishment Clause, post-Kennedy, forbids only the formal features of a founding-era established church and has little to say about the saturation of public space — including the public classroom — with majoritarian religious messaging.

This essay is meant to give you the background and analytical tools to read Nathan carefully. It proceeds in three parts. Part II reconstructs the doctrinal landscape in which Nathan was decided: the original synthesis represented by Everson, Lemon, and Stone, the school-prayer line that ran alongside it, and the coercion-and-history turn that culminated in Kennedy. Part III walks through the Nathan majority's six-hallmark test and the dissents' three main attacks on it. Part IV presses two hypotheticals — a classroom portrait of Jesus, and a classroom crucifix — against the majority's framework, to show where the majority's rule does and does not constrain future legislation. The essay closes with an assessment of whether it is fair to read Nathan as leaving the Establishment Clause with "zero wall of separation."

Your aim in reading Nathan should be neither to cheer the majority nor to lament the dissents. It should be to understand what the new test is, what it permits, what it still forbids, and — the hardest question — what its underlying premises commit its authors to.

II · Background: The Wall, the Test, and the Turn

A. The Wall and the Tiered Test

Modern Establishment Clause doctrine begins with Everson v. Board of Education, 330 U.S. 1 (1947), which incorporated the Establishment Clause against the states and revived Jefferson's famous metaphor of a "wall of separation between Church and State." Id. at 16. Everson declared that neither a state nor the federal government "can set up a church," "force" religious belief or disbelief, "punish" the entertaining or professing of religious beliefs, or "pass laws which aid one religion, aid all religions, or prefer one religion over another." Id. at 15. The metaphor was aspirational; the rule-statement was ambitious.

For the next two decades, the Court experimented with applications of that ambition, striking down state-composed school prayers in Engel v. Vitale, 370 U.S. 421 (1962), and state-mandated Bible reading and Lord's Prayer recitation in School District of Abington Township v. Schempp, 374 U.S. 203 (1963). Schempp in particular announced what would become the first half of the canonical test: a statute could not survive the Establishment Clause unless it had "a secular legislative purpose and a primary effect that neither advances nor inhibits religion." Id. at 222.

In Lemon v. Kurtzman, 403 U.S. 602 (1971), the Court fused the purpose and effect requirements together with a third prong — that the challenged law not foster "an excessive government entanglement with religion" — and articulated the three-part test that dominated Establishment Clause analysis for the next half-century. Id. at 612-13.

Applied to classroom religious displays, Lemon produced the outcome in Stone v. Graham, 449 U.S. 39 (1981) (per curiam). Kentucky had required a copy of the Ten Commandments, purchased with private funds, to be posted in every public school classroom, accompanied by a small context statement describing the Decalogue as "the fundamental legal code of Western Civilization and the Common Law of the United States." Id. at 39 n.1. The Court summarily reversed the Kentucky courts, holding that "the pre-eminent purpose for posting the Ten Commandments on schoolroom walls is plainly religious in nature," and that "no legislative recitation of a supposed secular purpose c[ould] blind [the Court] to that fact." Id. at 41. Forty-two words of Kentucky boilerplate could not save a display whose "effect" would be to "induce the schoolchildren to read, meditate upon, perhaps to venerate and obey, the Commandments." Id. at 42. That, the Court said, was "not a permissible state objective under the Establishment Clause." Id.

Stone was short — barely four pages of Reporter text — but it was a directly-on-point, seven-justice per curiam. For forty-five years, federal courts treated it as controlling whenever a legislature tried to mount the Decalogue in public-school classrooms.

B. The School-Prayer Line

Running alongside and partially overlapping Lemon, a distinct doctrinal line addressed religious exercise in public schools specifically. Engel and Schempp established that state-composed or state-prescribed prayer in the classroom was impermissible. Lee v. Weisman, 505 U.S. 577 (1992), extended the principle to clergy-led prayer at a public-school graduation. Santa Fe Independent School District v. Doe, 530 U.S. 290 (2000), extended it again to student-led, student-initiated prayer broadcast over a high-school football stadium's public-address system.

The central insight of this line was that coercion in the school context is different from coercion elsewhere. Students are a captive audience: they are subject to compulsory-attendance laws, they are "impressionable," and their "attendance is involuntary." Edwards v. Aguillard, 482 U.S. 578, 584 (1987). In such an environment, even "subtle and indirect" pressure — the pressure to stand respectfully, to refrain from walking out, to blend in with one's peers — can amount to the "establishment of a state religion by means of such pressure." Lee, 505 U.S. at 592-93.

The school-prayer cases thus did two kinds of work. Doctrinally, they elaborated a coercion-based Establishment Clause theory independent of Lemon's secular-purpose test. Rhetorically, they placed public schools at the heart of the Establishment Clause's concerns. Stone v. Graham is best understood as a hybrid case, deploying Lemon's secular-purpose language but sitting intellectually within the school-prayer tradition.

C. The Long Retreat from Lemon

Virtually from the moment Lemon was decided, it attracted criticism — from Justices, academics, and even some of its own authors. Over the next fifty years the Court narrowed, refined, and sometimes ignored Lemon. The endorsement test — Justice O'Connor's refinement of the primary-effect prong — became the Court's preferred framework in the 1980s and 1990s. See, e.g., Lynch v. Donnelly, 465 U.S. 668, 687-88 (1984) (O'Connor, J., concurring). By the 2000s, the Court would apply Lemon in some cases (McCreary County v. ACLU, 545 U.S. 844 (2005)), replace it in others (Van Orden v. Perry, 545 U.S. 677 (2005) (plurality)), and in still others openly criticize it. The story of the 2000s and 2010s in Establishment Clause doctrine is a story of slow erosion without clean overruling.

Two cases set the stage for the eventual break. In Town of Greece v. Galloway, 572 U.S. 565 (2014), the Court upheld a town's practice of opening its legislative sessions with explicitly sectarian prayers, relying heavily on historical practice since the founding and distinguishing the town-council setting from the school setting. And in American Legion v. American Humanist Ass'n, 588 U.S. 29 (2019), the Court upheld the constitutionality of a 32-foot Latin cross on public land in Bladensburg, Maryland, emphasizing that "longstanding monuments, symbols, and practices" carry a "strong presumption of constitutionality" and declining to apply Lemon's "grand unified theory" to the display. Id. at 54-58 (plurality).

American Legion did not formally overrule Lemon. But it featured a concurring opinion by Justice Gorsuch pushing back hard against "offended-observer" standing, and a plurality opinion that expressly refused to reaffirm the Lemon test. The wall was crumbling; a demolition order seemed imminent.

D. The Kennedy Turn

The demolition order came in Kennedy v. Bremerton School District, 597 U.S. 507 (2022). A high-school football coach, Joseph Kennedy, engaged in a public, midfield prayer after games; the school district disciplined him, citing Establishment Clause concerns. The Supreme Court, 6-3, sided with Coach Kennedy and announced two propositions that would reshape the doctrine.

First, and most importantly for our purposes, the Court explicitly declared that Lemon was dead. "[T]his Court long ago abandoned Lemon and its endorsement test offshoot." Id. at 534. That sentence — "long ago abandoned" — was the formal funeral.

Second, the Court announced that the Establishment Clause "must be interpreted by reference to historical practices and understandings." Id. at 535 (quoting Town of Greece, 572 U.S. at 576). The governing inquiry, Kennedy suggested, looks to history: what the Framers understood an "establishment of religion" to mean.

Left unresolved was the question that Nathan would squarely confront: did Kennedy abrogate only Lemon itself (and its endorsement-test offshoot), or did it sweep away the entire edifice of doctrine Lemon had built? In particular, did the secular-purpose prong survive in some form? Did Stone v. Graham — a case that quoted Lemon but whose reasoning rested largely on Schempp and Engel — survive?

E. Two Survivors the Court Has (So Far) Preserved

Two other doctrinal lines matter for Nathan. First, the prohibition on denominational preference announced in Larson v. Valente, 456 U.S. 228 (1982), has been expressly reaffirmed as recently as Catholic Charities Bureau v. Wisconsin Labor & Industry Review Commission, 605 U.S. 238 (2025). Larson held that "the clearest command of the Establishment Clause is that one religious denomination cannot be officially preferred over another." Id. at 244. When a statute or policy draws lines among religious traditions — say, by providing benefits to some and not others — Larson triggers strict scrutiny. Larson survives Kennedy untouched.

Second, the "historical hallmarks" framework elaborated in Justice Gorsuch's concurrence in Shurtleff v. City of Boston, 596 U.S. 243, 286-87 (2022) (Gorsuch, J., concurring in the judgment), has emerged as the most fully developed originalist Establishment Clause theory. Drawing heavily on the historical scholarship of Professor Michael McConnell, Justice Gorsuch identified six hallmarks of a founding-era religious establishment: (i) government control over the doctrine and personnel of the established church; (ii) mandated attendance in the established church and punishment for failing to participate; (iii) punishment of dissenting churches and individuals for their religious exercise; (iv) restriction of political participation by dissenters; (v) financial support for the established church, often in a way that preferred the established denomination over others; and (vi) use of the established church to carry out civil functions. When none of those features is present, Gorsuch argued, the challenged practice is not an establishment in the sense the Founders forbade.

Whether the six hallmarks are necessary features, sufficient features, or merely illustrative examples of what an establishment looks like is one of the central doctrinal questions Nathan presses.

F. A Final Piece: Mahmoud v. Taylor

Before we turn to Nathan, one more piece of the puzzle must be introduced. In Mahmoud v. Taylor, 606 U.S. 522 (2025), the Supreme Court held that a Maryland school district violated the Free Exercise rights of religious parents by refusing to allow them to opt their elementary-school children out of classroom instruction using "LGBTQ+-inclusive" storybooks. The Court framed the harm in terms borrowed from Wisconsin v. Yoder: "A government burdens the religious exercise of parents when it requires them to submit their children to instruction that poses 'a very real threat of undermining' the religious beliefs and practices that the parents wish to instill." Id. at 544 (quoting Wisconsin v. Yoder, 406 U.S. 205, 218 (1972)).

Mahmoud was a Free Exercise case, not an Establishment Clause case. But it established a principle — that the school classroom is a special site of parental religious-direction rights — that runs against the direction the Court took in Kennedy for Establishment Clause doctrine. Nathan's majority and dissents both grapple with Mahmoud, and reach opposite conclusions about what it requires.

III · Nathan v. Alamo Heights: The New Framework

A. The Statute

Texas S.B. 10 is simple on its face. Every public elementary and secondary school classroom must display a 16-by-20-inch durable poster containing the King James Version of the Ten Commandments. The text must be "legible to a person with average vision from anywhere in the classroom." Tex. Educ. Code § 1.0041(b)(1). The posters must be hung in a "conspicuous place." No context statement is permitted. No accompanying secular-legal monuments or historical artifacts are required.

On the day the statute took effect, a group of public-school students and their parents — Protestant, Catholic, Jewish, Muslim, Hindu, Baha'i, Unitarian Universalist, and nonreligious — sued, alleging violations of the Establishment Clause and Free Exercise Clause. The district court preliminarily enjoined enforcement. The Fifth Circuit took the case en banc and reversed.

B. The Majority's Six-Hallmark Test

Judge Duncan's majority opinion (for nine judges) made five moves.

First, the majority held that Lemon had been fully abrogated by Kennedy, including Lemon's secular-purpose prong.

Second, and more controversially, the majority held that Stone v. Graham fell along with Lemon. Because Stone had applied Lemon's secular-purpose test to hold the Kentucky Decalogue-display statute unconstitutional, and because Lemon's secular-purpose test was no longer good law, Stone was no longer good law.

Third, the majority adopted Justice Gorsuch's Shurtleff six-hallmark framework as the governing test for Establishment Clause claims: only those practices that bear one or more of the six hallmarks of a founding-era establishment violate the clause.

Fourth, the majority concluded that S.B. 10 bears none of the six hallmarks. A passive classroom display does not install a controlled state church or its personnel; it does not compel attendance at an official church or punish non-attendance; it does not punish dissenters for their religious exercise; it does not restrict dissenters' political participation; it does not direct financial support to any established denomination; and it does not use an established church to carry out civil functions. Accordingly, S.B. 10 is constitutional.

Fifth, the majority distinguished Mahmoud v. Taylor. Mahmoud concerned "compulsory curriculum" — books that students would be required to engage with as part of graded classroom instruction. Posters, the majority said, are not curriculum. Students are not quizzed on them, not required to recite them, not graded on them. Whatever parental right Mahmoud recognized, it does not extend to the mere presence of religious text on a classroom wall.

C. The Dissenters' Three Attacks

The dissents, led principally by Judge Ramirez and Judge Higginson, pressed three distinct lines of attack.

Stare decisis. The Supreme Court has told inferior federal courts, repeatedly, that they cannot assume silent overruling. Rodriguez de Quijas v. Shearson/American Express, Inc., 490 U.S. 477, 484 (1989). Kennedy never mentions Stone. The majority's declaration that Stone has been overruled is, Judge Haynes's one-paragraph dissent argues, simply beyond the Fifth Circuit's power. This is the narrowest dissenting position and the most jurisprudentially conservative.

Coercion within the school context. Even if we accept Kennedy's historical turn, Kennedy expressly left intact the school-prayer line. Engel, Schempp, Lee, and Santa Fe all remain good law. Those cases establish that coercion in the elementary and secondary school context is "subtle and indirect," Lee, 505 U.S. at 592-93, and that children in compulsory-attendance settings are especially susceptible to pressure to conform. The Ramirez and Southwick dissents argue that a mandatory Decalogue display in every classroom exerts at least as much coercive pressure on students as the football-game prayer in Santa Fe or the graduation prayer in Lee.

Denominational preference and parental rights. The Higginson dissent — joined by Richman, Graves, Douglas, and Ramirez — argues that Texas chose a specifically Protestant text (the King James Version's particular translation and numbering) and thereby preferred Protestantism over Catholicism, Judaism, and every other religious tradition. That runs headlong into Larson v. Valente and Catholic Charities Bureau, which the Nathan majority did not purport to overrule. Higginson also argues that Mahmoud's parental-rights principle applies with greater force to S.B. 10 than to the storybooks in Mahmoud itself, because parents cannot opt their children out of a classroom display the way they could request opt-outs from particular books.

IV · Stress-Testing the Majority's Framework

The majority's framework is at its most intellectually serious when stated abstractly. It is at its most vulnerable when one presses it with hypotheticals. Two are worth working through.

A. The Jesus-Poster Hypothetical

Suppose Texas followed S.B. 10 with S.B. 11, requiring every public-school classroom to display a 16-by-20-inch poster of Jesus Christ. Same dimensions. Same "conspicuous place" requirement. Same prohibition on context statements. Different content.

Run that statute through the six-hallmark test.

Every box that was unchecked for the Decalogue is unchecked for Jesus. Under the majority's own framework, S.B. 11 is just as constitutional as S.B. 10.

The only doctrinal hook the majority left itself is Larson v. Valente's prohibition on denominational preference, which the Nathan majority expressly preserved. On that hook, S.B. 11 should fall: Jesus is specifically central to Christianity in a way the Ten Commandments (on the majority's telling) are not. Jewish, Muslim, Hindu, Buddhist, Sikh, and nonreligious families would all recognize a Jesus poster as distinctively Christian imagery.

So the answer to the first half of your question is "yes, but — the majority's six-hallmark test alone does not distinguish the Jesus poster from the Decalogue, and only the surviving Larson doctrine provides doctrinal grounds for distinguishing them."

That answer is unsatisfying for a reason. Whether the Larson hook holds depends entirely on how broadly a court is willing to draw the category of "religious tradition" for purposes of denominational preference. The Nathan majority drew it very broadly — so broadly that "Judeo-Christian" embraced a specifically Protestant translation of the Decalogue, on the theory that the Decalogue's "cultural resonance" transcends sectarian boundaries. If the same court were asked to save a Jesus poster, it could try the same move — Jesus is a figure recognized (if differently understood) by Christians, Muslims, Baha'is, and secular historians alike; his cultural resonance extends beyond any single denomination. The move would be weaker for Jesus than for the Decalogue. But the fact that the same move is available suggests that the Larson hook is not a stable limit; it is a rhetorical choice.

B. The Crucifix Hypothetical

Consider a harder variant. Suppose a Texas school district — not the state legislature — adopts a policy of hanging a small wooden crucifix in every classroom, funded by voluntary parent contributions.

Now the six hallmarks look even thinner. The state is not spending a dime. The state is not mandating anything. A crucifix is less text-heavy than the Decalogue; it is a symbol rather than a recitation of divine commands. It does not "induce" students to "read" or "meditate upon" anything in the direct way Stone described.

And yet nearly every lawyer's intuition is that a mandatory classroom crucifix is more troubling than a mandatory classroom Decalogue. Why? Because a crucifix is a pure sectarian emblem. It does not claim "Judeo-Christian" ecumenical cover. It is the central symbol of one tradition. Larson should bite hard.

But notice what has happened. The Nathan majority's test is doing almost none of the work. All the work is being done by Larson, which the majority inherited rather than constructed. The majority's framework, on its own, would not obviously stop a crucifix. It would stop one only by piggybacking on a doctrine the majority did not articulate and could not refine without admitting that its own "cultural resonance" analysis of the Decalogue was doing double duty as denominational-preference analysis.

C. What the Framework Still Forbids

Law students should not overread the majority. There is a real limiting principle in the Nathan framework, even if it is narrower than the pre-Kennedy Wall. The framework still forbids:

What the framework permits, at a minimum, is the saturation of public-school classrooms with passive religious text and imagery drawn from what a court will describe as "broad cultural-religious traditions." That is a major expansion of permissible state action compared to the pre-Kennedy regime.

V · Is This "Zero Wall of Separation"?

Your instinct — that the majority is close to reading the Establishment Clause as prohibiting only a literal founding-era establishment — is well-grounded and widely shared among critics of the Kennedy-era doctrine. Professor Ira Lupu and Professor Robert Tuttle have argued in a 2025 essay that the six-hallmark framework effectively reduces the Establishment Clause to a prohibition on what King George III's Anglican establishment actually did: taxation for clergy, compelled attendance at services, civil disabilities for dissenters. See Lupu & Tuttle, The Ten Commandments in Louisiana Public Schools: A Study in the Survival of Establishment Norms, 100 Chi.-Kent L. Rev. 601 (2025). Anything short of that formal fusion of church and state becomes permissible.

But "zero wall" is not quite right, and it is useful to see why not.

First, the school-prayer cases survive formally. The majority in Nathan does not purport to overrule Lee or Santa Fe. A state cannot mandate a prayer, a recitation, or a formal religious exercise in public school. The dissenters argue, with considerable force, that the majority has narrowed "coercion" to the point of near-irrelevance — but the doctrinal line still exists.

Second, Larson v. Valente's denominational-preference prohibition survives, and the Nathan majority expressly preserved it. A state cannot officially prefer one religious sect over another. Whether this doctrine can be made to bite in practice — whether courts will be willing to look past "pan-Judeo-Christian" framing to recognize denominational preference in the choice of a specific translation — is an empirical question the next round of litigation will answer.

Third, the Free Exercise Clause, as elaborated in Mahmoud v. Taylor, protects parents' ability to direct their children's religious upbringing through public-school instruction. The Nathan majority distinguished Mahmoud on the "compulsory curriculum" ground, but a different Supreme Court might well conclude that a daily visible poster is more intrusive into parental religious-direction rights than an opt-outable storybook. The dissenters believed that argument should have won Nathan on Free Exercise grounds alone.

So: what has changed is not that every religious display is now permitted. What has changed is that the presumption has flipped. Under Lemon and the pre-Kennedy cases, a religious display in a classroom was presumptively unconstitutional unless the state could articulate and defend a plausible secular purpose and a non-religious primary effect. Under Kennedy and Nathan, a religious display is presumptively constitutional unless the challenger can show that it checks one of the six hallmark boxes or that it violates one of the surviving discrete doctrines (denominational preference, formal coercion of worship, Free Exercise interference with parental direction).

That is not zero wall. It is a much thinner wall — a wall that now forbids only the most formal, textbook-example features of an established church, and permits a degree of state religious messaging that would have been unthinkable under Stone or McCreary County.

Whether that thinner wall is the correct reading of the First Amendment is a question on which reasonable originalists, living constitutionalists, religious liberty advocates, and separationists will disagree. What is clear is that Nathan marks an important doctrinal waypoint: the first en banc circuit decision to rest the permissibility of classroom religious displays squarely on the historical-hallmarks framework and to declare that Stone v. Graham did not survive Kennedy. Whether the Supreme Court will endorse that reading — or temper it, either by reinstating a secular-purpose inquiry or by taking the denominational-preference inquiry more seriously than the Nathan majority did — is the question on which the next generation of Establishment Clause litigation will turn.

VI · What to Take Away

A careful reader of Nathan should hold four propositions in mind simultaneously.

First, the case is a faithful application of the logic of Kennedy v. Bremerton to a problem Kennedy did not address. The majority did not invent the six-hallmark framework; it borrowed it from Justice Gorsuch's Shurtleff concurrence and applied it seriously. If Kennedy is correctly decided, something like Nathan follows almost inexorably.

Second, the majority made one aggressive move — declaring Stone v. Graham abrogated — that goes beyond anything the Supreme Court itself has yet said. That move is jurisprudentially bold, and it is the cleanest vehicle for Supreme Court review. Whether the Court will bless it or rebuke it is genuinely uncertain.

Third, the test the majority announced is underdetermined in ways that will generate litigation for decades. The "cultural resonance" move the majority used to save the Decalogue under Larson is not a rule; it is a rhetorical posture. Different courts will draw the line between denominational preference and pan-traditional cultural reference in different places. The first test case — a crucifix, a Jesus poster, a Book of Mormon display — will expose how narrow the doctrine really is.

Fourth, the dissents preserved a coherent alternative vision of Establishment Clause doctrine. That vision — anchored in Stone, the school-prayer line, Mahmoud, and Larson — is not gone. It is the vision the Supreme Court may reach for if and when it confronts the Texas classrooms that now, pursuant to S.B. 10, display the King James Decalogue on every wall.

The wall is thinner than it was. It has not yet disappeared. Whether it will depends less on the logic of the six hallmarks than on how the next Supreme Court majority chooses to balance history, doctrine, and the particular constitutional place that public-school classrooms have occupied in American law since Everson v. Board of Education first declared that separation is "impregnable."

An essay for law students, April 2026.

← Read the full case brief Read the companion essay on vouchers →