Establishment Clause Limits on Government Support for Religion

Establishment Clause Limits on Government

Support for Religion

August 22, 2025

Congressional Research Service

https://crsreports.congress.gov

R48645

SUMMARY

Establishment Clause Limits on Government

Support for Religion

The first two provisions of the U.S. Constitution’s First Amendment, known as the Religion

Clauses, state that “Congress shall make no law respecting an establishment of religion or

prohibiting the free exercise thereof.” Together, the Establishment and Free Exercise Clauses

require the government to be neutral toward religion, neither providing impermissible support nor

demonstrating impermissible hostility.

R48645

August 22, 2025

Valerie C. Brannon

Legislative Attorney

The Supreme Court acknowledged in Walz v. Tax Commission, 397 U.S. 664, 668 (1970), that the Religion Clauses “are not

the most precisely drawn portions of the Constitution.” Over the years, the Supreme Court has employed a variety of

different tests to analyze whether government support for religion violates the Establishment Clause. Among other methods

of analysis, the Court employed the three-part Lemon test for several decades. The Lemon test asked courts to ensure that the

challenged government action has a secular purpose, that its primary effect neither advances nor inhibits religion, and that it

does not foster excessive government entanglement with religion.

In 2022, however, the Supreme Court announced that it had “long ago abandoned Lemon.” Kennedy v. Bremerton School

District, 597 U.S. 507, 534 (2022). Instead, the Court instructed lower courts to interpret the Establishment Clause by

“reference to historical practices and understandings.” Id. at 535. While Kennedy announced this new “historical practices

and understandings test,” it did not overrule prior decisions that applied the Lemon test or other types of analyses. Lower

courts are still bound to follow the outcomes of any Supreme Court precedent that has direct application, even if those cases

applied the Lemon test.

The Kennedy decision did not instruct courts on how to conduct an analysis by reference to historical practices and

understandings, but older Supreme Court precedent contains some guidance. For example, the Court looked to Founding-era

understandings and practices to invalidate state laws requiring public officials to declare a belief in the existence of God, as

well as a state law requiring a specified prayer to be recited at the beginning of a school day. The Court also looked to

evidence of long-standing historical practices to uphold prayers at the beginning of legislative sessions. In Kennedy itself,

however, the Court did not look to specific historical practices similar to the challenged government action. Instead, it

referenced its own prior precedent evaluating whether similar practices were coercive. Kennedy thus suggests some prior

Supreme Court precedent that did not look to historical practices or understandings may remain good law.

After Kennedy, lower courts have taken different approaches to evaluating Establishment Clause claims. Many courts have

attempted to apply Kennedy’s instruction to look to historical practices and understandings, seeking evidence of practices

from the Founding era or later that are in line with the challenged government action. At the same time, lower courts have

continued to apply controlling Supreme Court precedent that did not employ a historical analysis. For instance, one federal

appeals court held that it was bound to follow a Supreme Court case that had applied the Lemon test because the facts were

materially identical to the government action it was considering. Other courts have continued to follow the reasoning and

analysis of Supreme Court cases that evaluated the presence of coercion or entanglement.

This unsettled state of jurisprudence can make it difficult to assess how courts might review existing federal laws or any new

laws that arguably support religion. Courts might look to historical practices and understandings, but there are open questions

regarding how to conduct this inquiry. Further, courts may still have to assess whether existing Supreme Court precedent has

direct application to the facts before them, requiring them to rule a certain way regardless of historical practices or

understandings.

Congressional Research Service

Establishment Clause Limits on Government Support for Religion

Contents

Overview of Establishment Clause Tests ........................................................................................ 2

Kennedy’s Adoption of Historical Practices and Understandings Test ............................................ 5

Post-Kennedy Establishment Clause Analysis ................................................................................. 7

Historical Practices and Understandings Test ........................................................................... 9

Other Establishment Clause Tests ........................................................................................... 12

Coercion ............................................................................................................................ 12

Precedent Applying Lemon ............................................................................................... 13

Entanglement .................................................................................................................... 14

Evaluating Government Support for Religion ............................................................................... 15

Contacts

Author Information........................................................................................................................ 18

Congressional Research Service

Establishment Clause Limits on Government Support for Religion

he First Amendment to the U.S. Constitution provides that “Congress shall make no law

respecting an establishment of religion, or prohibiting the free exercise thereof.”1 This

relatively sparse language supplies the framework for a more complicated jurisprudence.2

While the First Amendment’s text refers to “Congress,” it also restricts action by the executive

and judicial branches as well as the states.3 The Establishment Clause bars the government from

providing certain types of support for religion,4 while the Free Exercise Clause prohibits

government hostility to religion.5 Together, the Religion Clauses require the government to be

neutral toward religion.6

T

Government support for religious institutions can implicate both Religion Clauses, as the

Supreme Court explained in the foundational 1947 case Everson v. Board of Education.7 In

Everson, a state reimbursed parents for bus fare to send their children to school, including

parochial schools.8 The Court held that the program did not violate the Establishment Clause even

though it used “tax-raised funds” to help some children “get to church schools.”9 The Court

observed a tension between the two Religion Clauses: the Establishment Clause prohibited the

state from using public funds to support “an institution which teaches the tenets and faith of any

church,” while the Free Exercise Clause prohibited the state from excluding individuals from

“public welfare legislation” because of their faith.10 Balancing these two principles, the Court said

that in preventing “state-established churches,” it did not want to “inadvertently prohibit [the

state] from extending its general state law benefits to all its citizens without regard to their

religious belief.”11 Everson provided a blueprint for much of the Court’s subsequent jurisprudence

on public support for religion.

In the past few decades, Supreme Court opinions have trended toward a view of the Religion

Clauses that allows more government support for religion. For example, the Supreme Court has

“abandoned” two doctrines that arguably had limited the government’s ability to support

religion.12 Further, starting in 2017, the Court issued a series of opinions holding that the

government violated the Free Exercise Clause by excluding religious entities from public

1 U.S. CONST. amend. I. See generally Cong. Rsch. Serv., Overview of the Religion Clauses (Establishment and Free

Exercise Clauses), CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/essay/amdt1-2-1/

ALDE_00013267/ (last visited Aug. 22, 2025).

2 See Walz v. Tax Comm’n, 397 U.S. 664, 668 (1970) (saying “the purpose [of the Religion Clauses] was to state an

objective not to write a statute”).

3 See, e.g., Trump v. Hawaii, 585 U.S. 667, 699–710 (2018) (considering and ultimately rejecting on the merits an

Establishment Clause challenge to a presidential proclamation); Presbyterian Church in the U.S. v. Mary Elizabeth

Blue Hull Mem’l Presbyterian Church, 393 U.S. 440, 445, 447 (1969) (“[T]he First Amendment severely circumscribes

the role that civil courts may play in resolving church property disputes.”); Everson v. Bd. of Educ., 330 U.S. 1, 8

(1947) (saying the Religion Clauses were “made applicable to the states” by the Fourteenth Amendment).

4 E.g., Everson, 330 U.S. at 15–16.

5 E.g., Masterpiece Cakeshop, Ltd. v. Colo. C.R. Comm’n, 584 U.S. 617, 639–40 (2018).

6 E.g., Sch. Dist. of Abington Twp. v. Schempp, 374 U.S. 203, 222 (1963).

7 Cong. Rsch. Serv., Relationship Between the Establishment and Free Exercise Clauses, CONSTITUTION ANNOTATED,

https://constitution.congress.gov/browse/essay/amdt1-5/ALDE_00000039/ (last visited Aug. 22, 2025).

8 Everson, 330 U.S. at 3.

9 Id. at 16–17.

10 Id. at 16.

11 Id.

12 Agostini v. Felton, 521 U.S. 203, 223–30 (1997); Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 534 (2022). As

discussed below, the doctrine deemed abandoned in Kennedy was a general test the Court had applied both to approve

and invalidate various government actions. The Kennedy opinion, however, expressed concern with the way the test

had been applied to suppress private religious activity. See 597 U.S. at 534–35.

Congressional Research Service

1

Establishment Clause Limits on Government Support for Religion

benefits—in other words, the decisions required the government to provide benefits to religious

entities in certain instances.13 In 2025, the Supreme Court held that a state violated the

Establishment Clause by excluding specific religious organizations from a tax exemption.14

Both Religion Clauses are relevant to evaluate government support for religion. Shifts in the

Supreme Court’s Establishment Clause jurisprudence have further complicated the matter, raising

questions about when the government can support—or decline to support—religion. This report

first describes the various tests the Supreme Court has historically used to analyze Establishment

Clause challenges. It then discusses Kennedy v. Bremerton School District, a 2022 case

instructing “that the Establishment Clause must be interpreted by ‘reference to historical practices

and understandings.’”15 The report reviews cases interpreting and applying Kennedy, exploring

open questions in Establishment Clause jurisprudence that may be of interest to Congress. While

the Supreme Court has abandoned certain approaches to interpreting the Establishment Clause, it

has not officially overruled some of the cases that applied those doctrines. The Establishment

Clause continues to prevent the government from providing direct financial aid or certain other

types of support to religious beliefs or activity.16 Nonetheless, in light of open jurisprudential

questions, lower courts have taken different approaches to reviewing Establishment Clause

challenges, potentially making it more difficult to predict how courts will rule.

Overview of Establishment Clause Tests

Over the years, the Supreme Court has evaluated various types of government support for religion

under the Establishment Clause, including financial aid, displays that include religious imagery,

and laws requiring participation in religious practices.17 As a historical matter, the Court has used

various analyses to determine whether the government has violated the Establishment Clause in

providing these and other kinds of support.18 The Court has acknowledged that the Religion

Clauses “are not the most precisely drawn portions of the Constitution.”19 While serving on the

Court, Justices O’Connor and Breyer both argued that there could be no single Establishment

Clause test, as different types of cases call for different analyses.20

13 Trinity Lutheran Church of Columbia, Inc. v. Comer, 582 U.S. 449, 454 (2017) (excluding “churches and other

religious organizations from receiving grants under [a Missouri] playground resurfacing program”); Espinoza v. Mont.

Dep’t of Revenue, 591 U.S. 464, 467–68 (2020) (excluding religious schools from a Montana tax credit program

benefiting parents of private school students); Carson v. Makin, 596 U.S. 767, 772–73 (2022) (excluding sectarian

schools from a Maine program providing tuition assistance to a private school of the parent’s choice).

14 Cath. Charities Bureau, Inc. v. Wisc. Lab. & Indus. Rev. Comm’n, 605 U.S. 238, 241–42 (2025). Specifically, as

discussed infra “Post-Kennedy Establishment Clause Analysis,” the Court held that the state tax exemption created a

denominational preference that triggered strict scrutiny.

15 Kennedy, 597 U.S. at 535 (quoting Town of Greece v. Galloway, 572 U.S. 565, 576 (2014)).

16 Cong. Rsch. Serv., Overview of Financial Assistance to Religion, CONSTITUTION ANNOTATED,

https://constitution.congress.gov/browse/essay/amdt1-3-4-1/ALDE_00013074/ (last visited Aug. 22, 2025); Cong.

Rsch. Serv., Overview of Non-Financial Assistance to Religion, CONSTITUTION ANNOTATED,

https://constitution.congress.gov/browse/essay/amdt1-3-5-1/ALDE_00013080/ (last visited Aug. 22, 2025).

17 See id.

18 See, e.g., Am. Legion v. Am. Humanist Ass’n, 588 U.S. 29, 48–50 (2019) (plurality opinion).

19 Walz v. Tax Comm’n, 397 U.S. 664, 668 (1970).

20 Elk Grove Unified Sch. Dist. v. Newdow, 542 U.S. 1, 33 (2004) (O’Connor, J., concurring); Am. Legion, 588 U.S. at

66–67 (2019) (Breyer, J., concurring). Both Justices sometimes influenced Establishment Clause jurisprudence by

providing the necessary votes to reach a judgment in cases where no single opinion commanded a majority of the

Court. E.g., Van Orden v. Perry, 545 U.S. 677, 698 (2005) (Breyer, J., concurring); Mitchell v. Helms, 530 U.S. 793,

836–37 (2000) (O’Connor, J., concurring).

Congressional Research Service

2

Establishment Clause Limits on Government Support for Religion

As a general matter, the Supreme Court has long recognized that the Establishment Clause’s

prohibition on laws “respecting an establishment of religion” not only prevents the government

from establishing an official religion, such as the Church of England, but also bars other types of

support “respecting” an establishment.21 The Supreme Court has said that “for the men who wrote

the Religion Clauses of the First Amendment the ‘establishment’ of a religion connoted

sponsorship, financial support, and active involvement of the sovereign in religious activity.”22 In

Everson, rather than outlining a general test, the Supreme Court identified specific activities that

would violate the Establishment Clause:

The “establishment of religion” clause of the First Amendment means at least this: Neither

a state nor the Federal Government can set up a church. Neither can pass laws which aid

one religion, aid all religions, or prefer one religion over another. Neither can force nor

influence a person to go to or to remain away from church against his will or force him to

profess a belief or disbelief in any religion. No person can be punished for entertaining or

professing religious beliefs or disbeliefs, for church attendance or non-attendance. No tax

in any amount, large or small, can be levied to support any religious activities or

institutions, whatever they may be called, or whatever form they may adopt to teach or

practice religion. Neither a state nor the Federal Government can, openly or secretly,

participate in the affairs of any religious organizations or groups and vice versa. 23

Subsequent cases outlined more general principles to guide Establishment Clause analysis. In the

1970 case Walz v. Tax Commission, the Supreme Court upheld a property tax exemption for

religious organizations after concluding that “the legislative purpose of tax exemption is not

aimed at establishing, sponsoring, or supporting religion” and that “the end result—the effect—is

not an excessive government entanglement with religion.”24 As part of its analysis of the law’s

effect, the Court emphasized the strong historical basis for such exemptions.25 In the Court’s

view, while the “unbroken practice” alone could not establish a right to the exemption, the

practice also could not be “lightly cast aside” given that “two centuries of uninterrupted freedom

from taxation has [not] given the remotest sign of leading to an established church or religion.”26

Accordingly, Walz looked to the law’s purpose and effect, judging effect in part by reference to

government entanglement and history.

A year later, in Lemon v. Kurtzman, the Supreme Court outlined an Establishment Clause test that

it said combined “the cumulative criteria developed by the Court.”27 To comply with the

Establishment Clause, the Court said, “first, the statute must have a secular legislative purpose;

second, its principal or primary effect must be one that neither advances nor inhibits religion ... ;

finally, the statute must not foster ‘an excessive government entanglement with religion.’”28 This

three-prong Lemon test looking to purpose, effect, and entanglement provided the primary (but

not sole) Establishment Clause test for decades following its announcement.29

21 U.S. CONST. amend. I. See generally Cong. Rsch. Serv., General Principle of Government Neutrality to Religion,

CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/essay/amdt1-3-1/ALDE_00013071/ (last visited

Aug. 22, 2025).

22 Walz, 397 U.S. at 668.

23 Everson v. Bd. of Educ., 330 U.S. 1, 15–16 (1947).

24 Walz, 397 U.S. at 674.

25 Id. at 676.

26 Id. at 676–78.

27 Lemon v. Kurtzman, 403 U.S. 602, 612 (1971), abrogation recognized by, Groff v. DeJoy, 600 U.S. 447, 460 (2023).

28 Id. at 612–13 (quoting Walz, 397 U.S. at 674).

29 E.g., Zelman v. Simmons-Harris, 536 U.S. 639, 668 (2002) (O’Connor, J., concurring) (describing Lemon as “[a]

(continued...)

Congressional Research Service

3

Establishment Clause Limits on Government Support for Religion

The Supreme Court articulated additional doctrines governing the Lemon test’s application in

subsequent cases. For example, to evaluate a program’s effect, the Court sometimes looked to

whether a reasonable observer would view the government as endorsing religion.30 Further, in

opinions that were later overruled, the Supreme Court invalidated support to religious schools

because it said the schools’ pervasively sectarian character would lead to impermissible

entanglement.31

In addition, the Supreme Court used Establishment Clause analyses that did not look to the three

Lemon factors. In a 1982 case, for instance, the Court explained that if a law contains a

“denominational preference”—that is, an express preference for one religious denomination over

another—the Court must “apply strict scrutiny in adjudging its constitutionality.”32 Under this

strict scrutiny test, the Supreme Court said a law would be unconstitutional unless it was “closely

fitted” to advancing “a compelling governmental interest.”33 The Supreme Court has applied a

strict scrutiny test in other constitutional contexts as one of the so-called “tiers of scrutiny.”34 In

the realm of the Establishment Clause, however, the Court has not generally employed these

familiar levels of scrutiny.35 Strict scrutiny apparently applies only to denominational preferences.

As another example of a non-Lemon analysis, the Supreme Court looked for impermissible

coercion in cases evaluating government support for prayer and religious teaching.36 In a 1948

case, the Supreme Court held that it was unconstitutional to allow private religious teachers to

teach religion in public schools, in part because the state was using its “compulsory public school

machinery” to “provide pupils” for religious classes.37 A 1992 case emphasized that there were

“heightened concerns” with “subtle coercion” in “elementary and secondary public schools.”38

Accordingly, the Supreme Court has held that schools violate the Establishment Clause by

sponsoring prayer at voluntary activities if the circumstances indirectly coerce students to

central tool” in Establishment Clause analysis); Doe v. Elmbrook Sch. Dist., 687 F.3d 840, 849 (7 th Cir. 2012) (saying

that Lemon is the prevailing test for Establishment Clause claims). For an in-depth discussion of how the Supreme

Court applied the Lemon test to evaluate financial assistance, see CRS Report R46517, Evaluating Federal Financial

Assistance Under the Constitution’s Religion Clauses, by Valerie C. Brannon (2020).

30 See Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 534, 535 n.4 (2022) (explaining this “variation” on Lemon);

Cong. Rsch. Serv., Endorsement Variation on Lemon, CONSTITUTION ANNOTATED, https://constitution.congress.gov/

browse/essay/amdt1-3-6-6/ALDE_00013089 (last visited Aug. 22, 2025).

31 See, e.g., Aguilar v. Felton, 473 U.S. 402, 412 (1985), overruled by, Agostini v. Felton, 521 U.S. 203 (1997);

Wolman v. Walter, 433 U.S. 229, 250–51 (1977), partially overruled by, Mitchell v. Helms, 530 U.S. 793 (2000). See

also Cong. Rsch. Serv., Lemon’s Effect Prong and Pervasively Sectarian Institutions, CONSTITUTION ANNOTATED,

https://constitution.congress.gov/browse/essay/amdt1-3-6-4/ALDE_00013086 (last visited Aug. 22, 2025).

32 Larson v. Valente, 456 U.S. 228, 246 (1982). In Larson, the Supreme Court invalidated a state law imposing

registration and reporting requirements on charitable organizations. Id. at 255. The law excluded religious

organizations that received more than half of their contributions from members. Id. at 230. The Court explained that

this provision “makes explicit and deliberate distinctions between different religious organizations.” Id. at 246 n.23.

33 Id. at 247.

34 See, e.g., CRS Report R48448, Gender and School Sports: Federal Action and Legal Challenges to State Laws, by

Madeline W. Donley and Jared P. Cole (2025); CRS Report R47986, Freedom of Speech: An Overview, by Victoria L.

Killion (2024).

35 See, e.g., Richard H. Fallon, Jr., Tiers for the Establishment Clause, 166 U. PA. L. REV. 59, 60–61 (2017).

36 In addition to the coercion approach, the Court has also applied the Lemon test to evaluate religious activities in

public schools. See, e.g., Edwards v. Aguillard, 482 U.S. 578, 580–81, 583 n.4 (1987) (ruling unconstitutional a state

law requiring the teaching of “creation science” in certain circumstances, and saying that a “historical approach” to

analyzing the Clause would not be “useful in determining the proper roles of church and state in public schools, since

free public education was virtually nonexistent at the time the Constitution was adopted”).

37 Illinois ex rel. McCollum v. Bd. of Educ., 333 U.S. 203, 212 (1948).

38 Lee v. Weisman, 505 U.S. 577, 592 (1992).

Congressional Research Service

4

Establishment Clause Limits on Government Support for Religion

participate in worship.39 In contrast, the Court has concluded that prayer before the meetings of

legislative bodies is not impermissibly coercive where adults understand they can come and go

freely and their choice to remain in the room or leave will not necessarily be understood as either

endorsing or disrespecting the prayer.40

The Supreme Court has also looked to historical practice to evaluate whether government

activities violate the Establishment Clause.41 In the early 1960s, the Supreme Court reviewed

colonial history in opinions holding that the Establishment Clause outlawed state laws requiring

public officials to declare a belief in the existence of God,42 as well as a state law requiring a

specified prayer to be recited at the beginning of a school day.43 In both cases, the Court said that

although there was historical precedent for the practices, there was also growing opposition in the

colonies and the Founding era demonstrating that the First Amendment outlawed the practices.44

Decades later, the Supreme Court used a historical analysis to reject Establishment Clause

challenges to state legislature and town board practices of opening their meetings with prayer.45 In

the Court’s view, “[t]he opening of sessions of legislative and other deliberative public bodies

with prayer is deeply embedded in the history and tradition of this country.”46 The Court ruled

that so long as a legislative prayer practice is consistent with historical practices and not

otherwise coercive, it will not violate the Establishment Clause.47

In summary, in the years since deciding Everson, the Supreme Court has applied a variety of tests

to evaluate whether the government violated the Establishment Clause, including Lemon’s threeprong test looking to purpose, effect, and entanglement; strict scrutiny analysis; an assessment of

whether the government action was coercive; and an evaluation of historical practices.

Kennedy’s Adoption of Historical Practices and

Understandings Test

In recent cases, the Supreme Court has said that courts should evaluate all Establishment Clause

claims by looking to historical practices and understandings. In 2022’s Kennedy v. Bremerton

School District, a majority of the Supreme Court said it had “long ago abandoned Lemon and its

39 Id. at 593 (holding “the school district’s supervision and control of a high school graduation ceremony places public

pressure, as well as peer pressure, on attending students to stand as a group or, at least, maintain respectful silence

during the invocation and benediction”); Santa Fe Indep. Sch. Dist. v. Doe, 530 U.S. 290, 312 (2000) (holding a prayer

before a football game had coercive effect, where the school was involved in the selection and delivery of the prayer).

40 Town of Greece v. Galloway, 572 U.S. 565, 590 (2014).

41 Prior to its decision in Lemon, the Court’s analyses sometimes looked to history to ensure laws had a secular purpose

or effect. See McGowan v. Maryland, 366 U.S. 420, 452 (1961); Walz v. Tax Comm’n, 397 U.S. 664, 676 (1970). In

McGowan, the Court rejected an Establishment Clause challenge to a state law prohibiting commercial activities on

Sunday. 366 U.S at 433–36. The Court concluded that while initially, English and colonial laws requiring Sunday

closures were intended to aid “the established church,” over time, the statutes became more secular both in text and in

justification. Id. at 433, 444, 453. Walz is discussed supra in the text accompanying notes 24 to 26. See also Locke v.

Davey, 540 U.S. 712, 722–23 (2004) (looking to evidence from “around the time of the founding” to support a state’s

interest in not funding degrees for devotional theology).

42 Torcaso v. Watkins, 367 U.S. 488, 496 (1961).

43 Engel v. Vitale, 370 U.S. 421, 424 (1962).

44 Torcaso, 367 U.S. at 490–92; Engel, 370 U.S. at 425–30.

45 Marsh v. Chambers, 463 U.S. 783, 795 (1983); Town of Greece, 572 U.S. at 569–70.

46 Marsh, 463 U.S. at 786.

47 Town of Greece, 572 U.S. at 577, 587–89.

Congressional Research Service

5

Establishment Clause Limits on Government Support for Religion

endorsement test offshoot.”48 Instead, the Court said “the Establishment Clause must be

interpreted by ‘reference to historical practices and understandings.’”49 Kennedy suggested this

analysis should focus on history and original meaning, referring to the understanding of the

Founding Fathers.50 At the same time, Kennedy’s own analysis only briefly referred to historical

understandings, and primarily focused on whether the facts demonstrated coercion.51

In Kennedy, the Supreme Court considered whether a school would have violated the

Establishment Clause by allowing a football coach to pray at the fifty-yard line immediately after

games.52 In evaluating the school’s decision to stop the coach’s prayer practice, the opinion

analyzed whether the coach had impermissibly coerced students into praying.53 Prior cases

evaluating prayer in schools also had looked for coercion, and in Kennedy the Court explained

that this inquiry was consistent “with a historically sensitive understanding of the Establishment

Clause.”54 In the Court’s view, “coercion” such as forcing citizens to engage in religious exercise

“was among the foremost hallmarks of religious establishments the framers sought to prohibit

when they adopted the First Amendment.”55

Unlike some of the earlier prayer cases discussed above, the Kennedy opinion did not review

Founding-era practices.56 Instead, to determine whether the football coach’s prayer was coercive,

the Court looked at the response of students and parents to his actions.57 The Court said there was

no record evidence that students felt pressured to participate apart from “hearsay.”58

Distinguishing prior cases that involved impermissible school-sponsored prayer, the Court said

the coach’s prayers “were not publicly broadcast or recited to a captive audience,” there was “no

formal school program accommodating the religious activity at issue,” and “[s]tudents were not

required or expected to participate.”59

Kennedy’s announcement that the Supreme Court had “abandoned” Lemon’s three-prong test

raised a number of additional questions about Establishment Clause jurisprudence in its wake.60

First, Kennedy did not overrule prior Establishment Clause cases that had applied the Lemon

48 Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 534 (2022). The last time a majority opinion had applied the Lemon

test to resolve an Establishment Clause case was in McCreary County v. ACLU, 545 U.S. 844, 861 (2005). A plurality

of the Court applied the endorsement test in Salazar v. Buono, 559 U.S. 700, 720–21 (2010) (plurality opinion). In

Espinoza v. Montana Department of Revenue, 591 U.S. 464, 474 (2020), and Carson v. Makin, 596 U.S. 767, 781

(2022), the Court rejected Establishment Clause challenges by citing to Zelman v. Simmons-Harris, 536 U.S. 639

(2002). The Zelman majority did not explicitly cite Lemon but nonetheless asked whether the challenged law had the

“‘purpose’ or ‘effect’ of advancing or inhibiting religion.” 536 U.S. at 648–49 (quoting Agostini v. Felton, 521 U.S.

203, 22–23 (1997)).

49 Kennedy, 597 U.S. at 535 (quoting Town of Greece, 572 U.S. at 576).

50 Id. at 536.

51 Id. at 537–40.

52 Id. at 512–14.

53

Id. at 537.

54 Id.

55 Id.

56 Id. at 537–42. See also, e.g., Randy E. Barnett & Lawrence B. Solum, Originalism After Dobbs, Bruen, and

Kennedy: The Role of History and Tradition, 118 NW. U. L. REV. 433, 474 (2023) (expressing the view that “Justice

Gorsuch’s opinion is not analytically precise about the roles that history and tradition play in the Court’s reasoning”).

57 Kennedy, 597 U.S. at 538–42.

58 Id. at 539.

59 Id. at 541–42.

60 Id. at 534.

Congressional Research Service

6

Establishment Clause Limits on Government Support for Religion

test.61 Instead, the Kennedy opinion itself looked to Supreme Court precedent on school coercion,

rather than a review of Founding-era practices, to decide the case.62 Further, as litigants have

noted before the Supreme Court,63 the Court evaluated Establishment Clause claims by looking to

a government action’s purpose, effect, and potential for entanglement in cases predating Lemon

itself.64 It therefore could be unclear whether the statement in Kennedy that “this Court long ago

abandoned Lemon and its endorsement test offshoot”65 also undermines cases predating Lemon

that applied the same factors.66

Post-Kennedy Establishment Clause Analysis

Since Kennedy, the Supreme Court has issued one opinion resolving a case on Establishment

Clause grounds: Catholic Charities Bureau, Inc. v. Wisconsin Labor & Industry Review

Commission.67 Despite Kennedy’s instruction to interpret the Establishment Clause by reference

to historical practices and understandings, Catholic Charities Bureau did not apply a historical

analysis.68 The case involved a religious exemption from a state unemployment insurance

program.69 To determine whether religious organizations were exempt from paying

unemployment taxes, the state looked to whether they engaged in religious activities such as

proselytizing or limiting their services to members of their religion.70 In a unanimous decision,

the Supreme Court said the law created “a denominational preference by explicitly differentiating

between religions based on theological practices”: “namely, whether to proselytize or serve only

co-religionists.”71 In line with its prior jurisprudence addressing express denominational

preferences, the Court held that the law was subject to strict scrutiny, meaning the state had to

show that it was “closely fitted” to “a compelling governmental interest.”72 The Court further

61 See id.

62 See id. at 536–42. See also, e.g., Jusino v. Fed’n of Cath. Tchrs., Inc., 54 F.4th 95, 102 (2d Cir. 2022) (concluding

“regardless of whether Kennedy actively overruled Lemon or simply recognized that Lemon was already a dead letter,

one thing it indisputably did not do was overrule” a separate case the court believed remained binding).

63 E.g., Transcript of Oral Argument at 111, Cath. Charities Bureau, Inc. v. Wisc. Lab. & Indus. Rev. Comm’n, 605

U.S. 238 (2025) (No. 24-154) (claiming that “entanglement predated Lemon”); Brief in Opposition at 21, City of

Pensacola v. Kondrat’yev, 139 S. Ct. 2772 (2019) (mem.) (No. 18-351) (claiming that “the secular purpose requirement

exists independent of ... and long predates Lemon”).

64 E.g., Walz v. Tax Comm’n, 397 U.S. 664, 674 (1970); see generally Cong. Rsch. Serv., Purpose and Effect Test

Before Lemon, CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/essay/amdt1-3-5-3/

ALDE_00013082/ (last visited Aug. 22, 2025).

65 Kennedy, 597 U.S. at 534.

66 Cf., e.g., Roake v. Brumley, 141 F.4th 614, 652 (5th Cir. 2025) (Dennis, J., concurring) (agreeing with “scholarship”

arguing that “Kennedy repudiated only the endorsement test ... and left intact the broader framework of Establishment

Clause doctrine: the requirement of a secular legislative purpose, the prohibition on policies whose primary effect

advances religion, and the concern about excessive entanglement between church and state”).

67 Cath. Charities Bureau, 605 U.S. 238. The Court heard oral arguments in another case involving the Establishment

Clause but ultimately did not issue an opinion. Okla. Statewide Charter Sch. Bd. v. Drummond ex rel. Oklahoma, 605

U.S. 165 (2025) (affirming the judgment by an equally divided Court). For more on Drummond, see CRS Legal

Sidebar LSB11342, Religious Charter Schools Remain Unlawful in Oklahoma After Supreme Court Judgment, by

Valerie C. Brannon and Whitney K. Novak (2025).

68 Cath. Charities Bureau, 605 U.S. 238.

69 Id. at 242.

70 Id. at 249–50.

71 Id. at 250.

72 Id. at 252 (quoting Larson v. Valente, 456 U.S. 228, 246–47 (1982)).

Congressional Research Service

7

Establishment Clause Limits on Government Support for Religion

concluded the state had failed to satisfy its burden.73 The majority opinion did not cite to Kennedy

or the historical practices and understandings test, even when it discussed an argument that

involved one of the Lemon factors.74 The state had claimed that its tax exemption would allow it

to avoid entangling questions about religious doctrine.75 The Court rejected this argument—not

on the basis that entanglement is no longer a valid factor for states to consider, but on the basis

that the law was not closely tailored to this entanglement interest.76

Thus, while the Supreme Court announced in Kennedy that courts should evaluate Establishment

Clause claims by looking to historical practices and understandings, neither Kennedy nor Catholic

Charities Bureau provided significant guidance on how to conduct such an inquiry. Instead, both

cases cited older precedent looking for coercion or a denominational preference. This is

consistent with the idea that Kennedy “abandoned” Lemon’s three-prong test but did not overrule

earlier cases.77 These cases might also be read to suggest that Kennedy’s historical test replaces

“Lemon and the endorsement test,” but not the other types of Establishment Clause inquiries the

Court previously employed.78

Lower courts remain bound by Supreme Court rulings that have “direct application” to a case,

even if the analysis has been “rejected in some other line of decisions.”79 Only the Supreme Court

can overrule its own cases.80 Following Kennedy, then, lower courts remain bound to follow the

rulings of any cases with similar fact patterns that have direct application. Kennedy and Catholic

Charities Bureau further suggest that courts might continue to follow not only the outcomes but

also the reasoning of some cases that did not primarily conduct a historical analysis.

Federal and state court opinions following Kennedy show that courts have applied the historical

test to evaluate Establishment Clause claims. In certain cases, however, courts have followed

Supreme Court precedent the courts believed remained binding, even if that precedent followed

an analysis using some of the Lemon factors. The overview of post-Kennedy jurisprudence below

focuses primarily on appellate court rulings, though select trial court decisions are discussed as

well.

73

Id. at 254.

74 See id. at 253–54.

75 Id.

76 Id.

77 Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 534 (2022).

78 Id. at 535; cf. id. at 536 (“An analysis focused on original meaning and history, this Court has stressed, has long

represented the rule rather than some ‘exception’ within the ‘Court’s Establishment Clause jurisprudence.’” (quoting

Town of Greece v. Galloway, 572 U.S. 565, 575 (2014))).

79 Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484 (1989).

80 Id.

Congressional Research Service

8

Establishment Clause Limits on Government Support for Religion

Historical Practices and Understandings Test

In light of Supreme Court guidance in Kennedy, lower courts have begun to grapple with how to

apply a historical approach to the Establishment Clause. One Fourth Circuit81 opinion identified a

number of open questions about how to evaluate Establishment Clause violations under

Kennedy’s historical test:

What kinds of evidence are relevant? ... What kinds of evidence are the most useful? ...

Which periods of history are relevant—the era of the Bill of Rights, 1791, or the era of the

incorporation of the Bill of Rights, 1868—and which period is most important?82

In its own pre-Kennedy Establishment Clause rulings evaluating historical practices, the Supreme

Court looked to history from the colonial era and shortly thereafter83 and searched for evidence of

longstanding, unbroken practices.84 The federal appeals courts that have applied Kennedy’s

historical test have seemed to focus first on Founding-era practices or understandings.85 Beyond

that, one court said that it would “look next to the ‘uninterrupted practice’ of a law in our nation’s

traditions”86 to determine whether the practice “withstood the critical scrutiny of time and

political change.”87

As the Supreme Court observed in 1987, it may be unclear how to use a historical test to analyze

practices that have arisen in modern contexts.88 In that opinion, the Court said that a “historical

approach” to analyzing the Establishment Clause would not be “useful in determining the proper

roles of church and state in public schools, since free public education was virtually nonexistent

at the time the Constitution was adopted.”89 Responding to this Supreme Court passage, the Third

Circuit stated in Hilsenrath ex rel. C.H. v. School District of Chathams that “[h]istorical tradition

can be established by analogical reasoning.”90 This may raise further questions about what

historical practices are sufficiently analogous to the challenged action.

Hilsenrath seemed to take a somewhat generalized approach to looking for analogous historical

practices.91 The court was evaluating whether instructional videos about Islam shared in a World

81 For ease of reading, references in this report to a particular circuit (e.g., the Fourth Circuit) refer to the U.S. Court of

Appeals for that particular circuit (e.g., the U.S. Court of Appeals for the Fourth Circuit).

82 Firewalker-Fields v. Lee, 58 F.4th 104, 122 n.8 (4th Cir. 2023). In addition, different Founders and different

jurisdictions had different views on religious establishments. See generally, e.g., Carl H. Esbeck, Uses and Abuses of

Textualism and Originalism in Establishment Clause Interpretation, 2011 UTAH L. REV. 489 (2011) (discussing

disagreements over the original meaning of the Establishment Clause); Cong. Rsch. Serv., Introduction to the

Historical Background on the Religion Clauses, CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/

essay/amdt1-2-2-1/ALDE_00013268 (last visited Aug. 22, 2025).

83 See Engel v. Vitale, 370 U.S. 421, 425–30 (1962); Torcaso v. Watkins, 367 U.S. 488, 490–92 (1961). In both of

these cases, as well as in McGowan v. Maryland, 366 U.S. 420, 431–40 (1961), the Court concluded that the historical

evidence showed shifts in practices and understandings that were critical to its analysis.

84 Walz v. Tax Comm’n, 397 U.S. 664, 676–78 (1970); Town of Greece v. Galloway, 572 U.S. 565, 576–77 (2014).

85 See Hilsenrath ex rel. C.H. v. Sch. Dist. of Chathams, 136 F.4th 484, 491 (3d Cir. 2025); Hunter v. U.S. Dep’t of

Educ., 115 F.4th 955, 965 (9th Cir. 2024); Freedom from Religion Found., Inc. v. Mack, 49 F.4th 941, 951–54 (5th Cir.

2022).

86 Hunter, 115 F.4th at 965–66 (quoting Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 536 (2022)). The quotation

comes from an explanatory parenthetical in Kennedy.

87 Id. (quoting Town of Greece, 572 U.S. at 577).

88 Edwards v. Aguillard, 482 U.S. 578, 583 n.4 (1987).

89 Id.

90 Hilsenrath, 136 F.4th at 491 (quoting Range v. Att’y Gen., 124 F.4 th 218, 228 (3d Cir. 2024) (en banc)) (alteration in

original).

91 Id.

Congressional Research Service

9

Establishment Clause Limits on Government Support for Religion

Cultures and Geography class violated the Establishment Clause.92 In applying Kennedy’s

“historical practices and understandings” test, the Third Circuit relied on a concurring opinion

from Justice Gorsuch.93 According to Justice Gorsuch, the “telling traits” of government

established churches include (1) “control over the doctrine and personnel”; (2) “mandated

attendance”; (3) punishment of dissenters for their religious exercise; (4) restriction of “political

participation by dissenters”; (5) financial support for the established church; and (6) using the

established church to carry out civil functions.94 Under the Third Circuit’s approach, the plaintiff

would have to prove that the curriculum resembled one of these “hallmarks” of an established

church.95 The plaintiffs asserted the presence of “coercion” (hallmark 2) and “non-neutrality”

(hallmark 5).96 The Third Circuit concluded that the school’s curriculum was not akin to

mandating attendance because the videos were not coercive or proselytizing.97 The court further

held that the school was not effectively providing financial support for a preferred religion

because the record did not “show favoritism” for Islam over other world religions.98

The Fifth Circuit disagreed with this approach in Roake v. Brumley, a case involving a Louisiana

law requiring public schools to display the Ten Commandments.99 That court said the Kennedy

majority did not adopt these “historical hallmarks of religious establishments” “as the exclusive

Establishment Clause test.”100 Instead, the Fifth Circuit asked whether “the permanent posting of

the Ten Commandments in public school classrooms fits within, or is consistent with, a broader

tradition of using the Ten Commandments in public education.”101 Given the case’s procedural

posture, the court accepted the plaintiff’s allegations that there was “no longstanding tradition of

permanently displaying the Ten Commandments in public[] school classrooms.”102

The Fifth Circuit again looked for more specific types of historical practices in Freedom from

Religion Foundation, Inc. v. Mack.103 To consider whether a judge could open his court sessions

with prayer, the Fifth Circuit discussed historical evidence of judicial practices.104 The court

looked to practices “around the Founding” and “the time of incorporation,” but it also reviewed

various practices from the 1830s and as late as 1996.105 Ultimately, the Fifth Circuit concluded

that although there was no evidence of prayer before daily court sessions, the scattered evidence

92 Id. at 486.

93 Id. at 491.

94 Id. (quoting Shurtleff v. City of Boston, 596 U.S. 243, 286 (2022) (Gorsuch, J., concurring)). Justice Gorsuch drew

these elements of a religious establishment from a law review article that reviewed “the period between initial

settlement and ultimate disestablishment” to summarize the “ad hoc and unsystematic” laws constituting American

religious establishments. Michael W. McConnell, Establishment and Disestablishment at the Founding, Part I:

Establishment of Religion, 44 WM. & MARY L. REV. 2105, 2131 (2003).

95 Hilsenrath, 136 F.4th at 491.

96 Id. at 491–92.

97 Id. at 493. This section of the opinion also looked to the Supreme Court’s precedent on school prayer. Id.

98 Id.

99 Roake v. Brumley, 141 F.4th 614, 645 (5th Cir. 2025).

100 Id. at 645–46.

101 Id. at 646.

102 Id. (alteration in original). The plaintiffs specifically asserted there was no evidence supporting the display of the

Ten Commandments in public schools, discounting the government’s evidence that early religious schools displayed

the Ten Commandments. Roake v. Brumley, 756 F. Supp. 3d 93, 173–74 (M.D. La. 2024). The plaintiffs also asserted

that Founding-era understandings did not support an official government version of the Ten Commandments. Id. at

172–73.

103 Freedom from Religion Found., Inc. v. Mack, 49 F.4th 941, 951–54 (5th Cir. 2022).

104 Id. at 951–54.

105 Id.

Congressional Research Service

10

Establishment Clause Limits on Government Support for Religion

of prayer before federal court-term openings and other judicial references to “God” were

sufficiently analogous to provide a historical basis supporting the challenged practice.106

To evaluate the constitutionality of a religious exemption from a federal nondiscrimination law,

the Ninth Circuit similarly reviewed the history of tax exemptions for religious institutions in

Hunter v. U.S. Department of Education.107 In the Ninth Circuit’s view, tax exemptions were the

closest analogy to the nondiscrimination exemption, “[g]iven the dearth of historical

equivalents.”108 Although the laws had differences, the court nonetheless concluded that “the

history of tax exemptions near the time of the Founding suggests that the statutory exemptions

that operate as a subsidy to religious institutions do not violate the Establishment Clause

according to its original meaning.”109 Additionally, the court reviewed more modern attempts to

accommodate religion in various federal and state programs, saying these provisions “evince a

continuous, century-long practice of governmental accommodations for religion.”110 The Ninth

Circuit acknowledged that the Supreme Court had used the Lemon test to approve of many of

these modern accommodations but said the decisions nonetheless provided evidence of historical

practices and understandings.111

Although Kennedy abandoned “Lemon and the endorsement test,”112 one trial court nonetheless

concluded that under the historical practices and understandings test, the Establishment Clause

“serves as a bulwark against government sponsorship or endorsement of religion.”113 The court

concluded a city had violated the Establishment Clause by “conceiving, organizing, and

implementing” a prayer vigil, impermissibly sponsoring “a religious event.”114 The court said the

city’s involvement with the prayer vigil was “dissimilar” to prior historical examples such as

legislative prayer or national days of prayer.115 Further, the court seemingly relied not just on

historical practices but also the Founders’ understandings of religious establishment.116 While

framing its analysis as an application of the historical practices and understandings test, the trial

court also cited decisions that analyzed the Lemon factors.117

In some contrast, a South Carolina trial court refused to consider prior Supreme Court cases based

on the Lemon test.118 In that South Carolina case, a foster-care agency that contracted with the

state to provide child placement services would work only with Christian foster parents who

affirmed a doctrinal statement of faith.119 A potential parent who refused to affirm these beliefs

argued this requirement violated the Establishment Clause in three ways: by (1) coercing

religious belief and exercise; (2) delegating government power to religious entities; and

106 Id. at 955–57.

107 Hunter v. U.S. Dep’t of Educ., 115 F.4 th 955, 966 (9th Cir. 2024).

108 Id. at 965.

109 Id.

110 Id. at 966. These provisions included another exemption from nondiscrimination law but also exemptions from the

draft or Social Security taxes, as well as prisoner accommodations. Id.

111 Id.

112 Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 535 (2022).

113 Rojas v. City of Ocala, 739 F. Supp. 3d 1068, 1090 (M.D. Fla. 2024).

114 Id. at 1088.

115 Id. at 1083.

116 Id. at 1084.

117 Id. at 1089.

118 Maddonna v. U.S. Dep’t of Health & Hum. Servs., 696 F. Supp. 3d 176, 191 (D.S.C. 2023).

119 Id. at 180–81.

Congressional Research Service

11

Establishment Clause Limits on Government Support for Religion

(3) demonstrating governmental favoritism for the agency’s religion.120 The trial court rejected

these arguments.121 It said the plaintiff failed to show the existence of a “historically disfavored

establishmentarian practice,” such as proving that religious agencies had a monopoly over

providing this civil service.122 To the extent the plaintiff’s arguments relied on Supreme Court

cases that applied the Lemon test, the trial court refused to consider those claims.123

Other Establishment Clause Tests

In other Establishment Clause cases decided post-Kennedy, courts have followed controlling

Supreme Court cases that were not decided using a historical practices or understandings test—

and in some cases, courts have applied factors from the Lemon test to evaluate whether a

government action violates the Establishment Clause.

Coercion

In line with Kennedy itself, some courts have continued to evaluate whether a challenged

government action is coercive. One trial court concluded that “a historical analysis [was] not

necessary” to review the constitutionality of a school program educating students on

transcendental meditation.124 The court said that “Kennedy ... did not overrule prior decisions”

using a coercion analysis to evaluate prayer in public schools, so it applied the reasoning from

those cases.125 The Second Circuit also applied Kennedy’s coercion analysis to reject a

photographer’s argument that New York would violate the Establishment Clause if it applied its

nondiscrimination laws in a way that required her to provide services to same-sex couples.126 The

appeals court concluded that the laws would “only require [her] to provide her wedding

photography services.”127 The laws would not require “her active religious participation in the

weddings that she photographs,” particularly given that “mere presence does not equate to

coerced participation in any religious activity.”128

The Supreme Court of Oklahoma discussed cases involving coerced prayer in public schools

when it held that the state could not approve a religious charter school.129 Specifically, the court

concluded “the creation of a religious public school” would violate the Establishment Clause

because such a school would “requir[e] or expect[] students to participate in religious

activities.”130 This ruling was appealed to the U.S. Supreme Court, which heard oral arguments in

120 Id. at 188.

121 Id.

122 Id. at 190–91 (quoting Firewalker-Fields v. Lee, 58 F.4th 104, 122 n.7 (4th Cir. 2023)).

123 Id. at 190–92.

124 Williams v. Bd. of Educ. of City of Chicago, 673 F. Supp. 3d 910, 921 (N.D. Ill. 2023).

125 Id. at 921–22. Specifically, the trial court denied both the school’s and the plaintiff’s motions for summary

judgment, saying a reasonable jury could find in favor of either party. Id. at 923. See also Lozano v. Collier, 98 F.4th

614, 627–28 (5th Cir. 2024) (remanding case to the trial court for further consideration after concluding a plaintiff had

presented evidence of potential coercion requiring inmates to take Christian-based classes).

126 Emilee Carpenter, LLC v. James, 107 F.4 th 92, 96, 111 (2d Cir. 2024).

127 Id.

128 Id. at 112.

129 Drummond ex rel. State v. Okla. Statewide Virtual Charter Sch. Bd., 558 P.3d 1, 14 (Okla. 2024).

130 Id. at 13.

Congressional Research Service

12

Establishment Clause Limits on Government Support for Religion

April 2025.131 The Supreme Court ultimately did not issue an opinion in the case, instead

affirming the judgment below by an equally divided court.132

In Freedom from Religion Foundation, the Fifth Circuit primarily reviewed historical practices to

determine whether courtroom prayer violated the Establishment Clause.133 The court also asked,

however, whether the judge’s prayer practice was coercive, looking to prior Supreme Court

precedent.134 The Fifth Circuit acknowledged disagreement among various opinions on how to

determine whether government action is unduly coercive, but it ultimately looked for “objective

evidence that a person has been treated differently from others.”135 In the court’s view, there was

no such evidence that any of the plaintiffs were subject to “disfavor” because of their failure to

participate in the prayer.136 Their “subjective perception” that the judge “disliked them” was

insufficient, according to the court.137

Precedent Applying Lemon

As mentioned above, one South Carolina trial court refused to rely on pre-Kennedy Supreme

Court cases that were based on the Lemon test.138 In contrast, in Roake, the Fifth Circuit

concluded it was bound by Stone v. Graham—a Supreme Court opinion that applied the Lemon

test—because Stone was not expressly overruled by Kennedy.139 Specifically, Roake held that a

Louisiana law requiring public schools to display the Ten Commandments was “materially

identical” to a Kentucky law that Stone ruled unconstitutional.140 The Fifth Circuit looked not

only to the text of the laws but also to legislators’ statements, determining that, as in Stone, those

statements revealed the Louisiana law did not have any valid secular purpose.141

The Oklahoma Supreme Court ruling on religious charter schools mentioned above appears to

have obliquely referenced cases that applied the Lemon test to prohibit direct government funding

for religious activity.142 In addition to its holding on coercion, the court emphasized that

131 See CRS Legal Sidebar LSB11342, Religious Charter Schools Remain Unlawful in Oklahoma After Supreme Court

Judgment, by Valerie C. Brannon and Whitney K. Novak (2025).

132 Okla. Statewide Virtual Charter Sch. Bd. v. Drummond, 605 U.S. 165 (2025).

133 Freedom from Religion Found., Inc. v. Mack, 49 F.4 th 941, 950 (5th Cir. 2022).

134 Id. at 958.

135 Id. at 958–59.

136 Id. at 960–61.

137 Id. at 961.

138 Maddonna v. U.S. Dep’t of Health & Hum. Servs., 696 F. Supp. 3d 176, 191 (D.S.C. 2023).

139 Roake v. Brumley, 141 F.4th 614, 642 (5th Cir. 2025). Accord Stinson v. Fayatteville Sch. Dist. No. 1, No. 5:25-CV5127, 2025 WL 2231053, at *11 (W.D. Ark. Aug. 4, 2025); Nathan v. Alamo Heights Indep. Sch. Dist., No. 5:25-cv00756, 2025 U.S. Dist. LEXIS 162056, at *70 (W.D. Tex. Aug. 20, 2025).

140 Roake, 141 F.4th at 643.

141 Id. at 643–45. The court explained that it did not “undertake this analysis to revive Lemon, but only for the limited

purpose of deciding whether Stone’s facts and reasoning control.” Id. at 643 n.21. Further, as discussed supra in the

text accompanying notes 99 to 102, the Fifth Circuit also applied Kennedy’s historical practices and understandings test

as an alternative analysis. Id. at 645 (“[E]ven if Stone were overturned tomorrow, H.B. 71 violates the Establishment

Clause under Kennedy.”).

142 Drummond ex rel. State v. Okla. Statewide Virtual Charter Sch. Bd., 558 P.3d 1, 14 (Okla. 2024); see also, e.g.,

Comm. for Pub. Educ. & Religious Liberty v. Nyquist, 413 U.S. 756, 780 (1973) (“In the absence of an effective means

of guaranteeing that the state aid derived from public funds will be used exclusively for secular, neutral, and

nonideological purposes, it is clear from our cases that direct aid in whatever form is invalid.”).

Congressional Research Service

13

Establishment Clause Limits on Government Support for Religion

approving a religious charter school would impermissibly “permit state spending in direct support

of the religious curriculum and activities.”143

Entanglement

Finally, some courts evaluating Establishment Clause claims have continued to look for

impermissible entanglement between religion and government, even though this was one of the

Lemon factors.144 The question of entanglement came up in two cases argued during the Supreme

Court’s October 2024 term.145 For instance, in the Catholic Charities Bureau case described

above, the state court disagreed that its interpretation of a state tax exemption law would require

an intrusive and entangling inquiry into religious organizations.146 Ultimately, the Supreme Court

rulings in both cases did not address the entanglement issue.147 Nonetheless, at oral argument in

the cases, Justices asked whether various outcomes would create excessive entanglement.148

Justice Thomas was the only member of the Court to ask during Catholic Charities Bureau

arguments whether entanglement was a permissible inquiry as a “standalone consideration,” or

whether instead it was part of the “hopefully defunct” Lemon test.149 Responding to his question,

the attorney for the state cited Walz, a case decided before Lemon, as the source for the

entanglement factor.150

In contrast to the state court’s ruling in Catholic Charities Bureau, some courts have concluded

post-Kennedy that state actions did create an unconstitutional entanglement.151 For instance, in

Does 1–11 v. Board of Regents of the University of Colorado, the Tenth Circuit held that a

143 Drummond, 558 P.3d at 14.

144 For a discussion of some of the cases analyzing the entanglement factor, see Cong. Rsch. Serv., Lemon’s

Entanglement Prong, CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/essay/amdt1-3-6-5/

ALDE_00013087/ (last visited Aug. 22, 2025). After Kennedy, one trial court rejected a plaintiff’s entanglement claim

on the basis that it rested “on an outdated and thus inapplicable test.” Gen. Conf. of Seventh-Day Adventists v. Horton,

No. 24-2866-TDC, 2025 WL 1703806, at *8 (D. Md. June 18, 2025), appeal docketed, No. 25-1735 (4th Cir. July 1,

2025).

145 See CRS Legal Sidebar LSB11283, Supreme Court Hears Argument on Religious Tax Exemptions, by Valerie C.

Brannon (2025); CRS Legal Sidebar LSB11342, Religious Charter Schools Remain Unlawful in Oklahoma After

Supreme Court Judgment, by Valerie C. Brannon and Whitney K. Novak (2025).

146 Cath. Charities Bureau, Inc. v. Lab. & Indus. Rev. Comm’n, 3 N.W.3d 666, 686–87 (Wis. 2024), rev’d and

remanded, 605 U.S. 238 (2025). Specifically, the Wisconsin Supreme Court said that an inquiry into whether an

organization both had a religious motivation and engaged in religious activities “requires minimal judicial inquiry into

religion.” Id. at 687. The inquiry was neutral and secular: courts would examine only whether the activities are

religious, not (for example) whether they “are ‘Catholic’ enough.” Id.

147 Cath. Charities Bureau, 605 U.S. 238; Okla. Statewide Virtual Charter Sch. Bd. v. Drummond, 605 U.S. 165

(2025).

148 E.g., Transcript of Oral Argument at 70, 104, Cath. Charities Bureau, 605 U.S. 238 (No. 24-154); Transcript of Oral

Argument at 9, Drummond, 605 U.S. 165 (No. 24-394).

149 Transcript of Oral Argument at 111, Cath. Charities Bureau, 605 U.S. 238 (No. 24-154).

150 Id.; Walz v. Tax Comm’n, 397 U.S. 664, 674 (1970) (stating that courts must determine “that the legislative purpose

of tax exemption is not aimed at establishing, sponsoring, or supporting religion” and that “the end result—the effect—

is not an excessive government entanglement with religion”).

151 E.g., Does 1–11 v. Bd. of Regents of Univ. of Colo., 100 F.4 th 1251, 1270 (10th Cir. 2024); Tilsen v. Benson, 299

A.3d 1096, 786 (Conn. 2023). In Tilsen, the Connecticut Supreme Court evaluated entanglement in the context of a

doctrine involving both Religion Clauses—the religious autonomy doctrine—which the court said had not been

disturbed by Kennedy. 299 A.3d at 774 n.8. The Alabama Supreme Court applied the same doctrine in a 2024 case,

although one concurring judge raised concerns about whether the majority opinion’s analysis was “similar to the Lemon

test.” Ex parte Ala.-W. Fla. Conf. of United Methodist Church, Inc., 401 So. 3d 1123, 1133 (Ala. 2024); id. at 1140

(Parker, C.J., concurring in part and concurring in the result).

Congressional Research Service

14

Establishment Clause Limits on Government Support for Religion

university’s COVID-19 vaccine policy violated the Religion Clauses.152 The university only

granted a religious exemption if a person cited to “the official doctrine of an organized religion”

that was “opposed to all immunizations,” requiring applicants to explain in detail why their

sincerely held beliefs prevented them from getting vaccines.153 The university rejected some

applicants’ explanations of their own beliefs—for example, rejecting a Roman Catholic’s

application after concluding that it was morally acceptable for Catholics to take the COVID-19

vaccines.154 The court concluded this policy required an “intrusive inquiry into the validity of the

[plaintiffs’] religious beliefs,” leading to “the sort of religious entanglement the Establishment

Clause proscribes.”155 In the opinion, the Tenth Circuit did not address Kennedy or the fact that

the entanglement analysis was part of the Lemon test.156

Evaluating Government Support for Religion

In Kennedy, the Supreme Court announced it had “long ago abandoned Lemon and its

endorsement test offshoot.”157 Nonetheless, the Court did not overrule any of its prior cases that

evaluated Establishment Clause challenges by looking to the government’s purpose, effect, or the

possibility of creating impermissible entanglement.158 While lower courts have attempted to

follow Kennedy’s command to interpret the Establishment Clause by reference to historical

practices and understandings,159 they have largely still considered themselves bound to follow

existing Establishment Clause precedent that was not overruled by Kennedy.160 Further, some

opinions suggest not only that courts might remain bound by the factual outcomes of directly

applicable decisions (e.g., holding a requirement to display the Ten Commandments to be

unconstitutional if it matches a law previously ruled unconstitutional161), but also that courts

might apply the reasoning of cases that did not apply a historical analysis (e.g., holding any law to

be unconstitutional if it creates an impermissible entanglement with religion162).

This lack of clarity about the state of Establishment Clause jurisprudence can make it difficult to

predict how courts will evaluate laws that provide or limit support for religious activities. PreKennedy cases that have not been overruled may continue to bind lower courts, though there

could be questions about whether the facts of those cases are sufficiently similar to control the

outcome in any new disputes.

The principles governing direct financial aid to religion provide one example. In the 1973 case

Committee for Public Education & Religious Liberty v. Nyquist, the Supreme Court ruled

unconstitutional a state program that provided grants to private schools to maintain and repair

152 Does 1–11, 100 F.4th at 1256–57.

153 Id. at 1257.

154 Id. at 1271.

155 Id.

156 Does 1–11’s entanglement analysis cited another Tenth Circuit case, Colorado Christian University v. Weaver, 534

F.3d 1245 (2008). Colorado Christian University said “[t]he anti-entanglement rule originated in the context of

education, changing with re-interpretations of the famous doctrine of” Lemon. 534 F.3d at 1261. That case’s analysis

cited Supreme Court cases applying the Lemon test. Id. at 1261–66.

157 Kennedy v. Bremerton Sch. Dist., 597 U.S. 507, 534 (2022).

158 See id.

159 Id. at 535; see supra “Historical Practices and Understandings Test.”

160 See supra “Post-Kennedy Establishment Clause Analysis.”

161 Roake v. Brumley, 141 F.4th 614, 642–43 (5th Cir. 2025).

162 Does 1–11 v. Bd. of Regents of Univ. of Colo., 100 F.4 th 1251, 1271 (10th Cir. 2024).

Congressional Research Service

15

Establishment Clause Limits on Government Support for Religion

their facilities.163 The state thus provided direct financial support to religious schools eligible for

these grants.164 To evaluate the program, the Supreme Court said the state law must satisfy the

three Lemon factors—but in setting out these factors, it cited cases predating Lemon.165 The Court

held that the grant program was unconstitutional, observing that it authorized “direct payments to

nonpublic schools, virtually all of which are Roman Catholic schools,” without restricting

payments “to the upkeep of facilities used exclusively for secular purposes.”166 The program

therefore had “a primary effect that advances religion in that it subsidizes directly the religious

activities of sectarian elementary and secondary schools.”167

Nyquist and other cases suggest that a court would likely hold a law to be unconstitutional if it

directly funds religious activities.168 Kennedy did not expressly overrule Nyquist or the pre-Lemon

cases cited in Nyquist, and courts likely remain bound by the holding that the government cannot

directly fund religious activities—or, at least, cannot create a program that mirrors the specific

grants in Nyquist.169 Different courts, however, might take different views of the scope of this

precedent. Some courts might look to the facts of Nyquist to see if the case is directly applicable

by factual analogy.170 For instance, the program in Nyquist only offered grants to certain private

schools, virtually all of which were one religion.171 If a program is more widely available to both

religious and nonreligious schools, a litigant might argue it is distinguishable. A court might then

decline to follow Nyquist and evaluate instead what types of direct financial support are

permissible based on historical practices and understandings.172 It is also possible that a court

could continue to evaluate financial assistance by looking to the program’s purpose, effect, or

potential for entanglement, potentially by citing pre-Lemon cases considering such factors.173

Adding to the complications, the specifics of what Nyquist and similar cases require from

governments were debated even before Kennedy. In Nyquist, the Supreme Court emphasized that

the grants did not contain any provisions expressly restricting their use to secular purposes.174 In a

later case, however, the Supreme Court rejected an Establishment Clause challenge to a federal

program funding adolescent health services even though the program did not expressly prevent

163 Comm. for Pub. Educ. v. Nyquist, 413 U.S. 756, 762–64 (1973). The state also created a tuition reimbursement and

tax relief program for parents. Id. at 764–67. These indirect aid provisions are discussed in Cong. Rsch. Serv., Zelman

and Indirect Assistance to Religion, CONSTITUTION ANNOTATED, https://constitution.congress.gov/browse/essay/amdt13-4-5/ALDE_00013078/ (last visited Aug. 22, 2025).

164 See Nyquist, 413 U.S. at 767–68.

165 Id. at 773.

166 Id. at 774.

167 Id.

168 Accord Mitchell v. Helms, 530 U.S. 793, 818–19 (2000) (plurality opinion); id. at 840 (O’Connor, J., concurring).

169 Cf., e.g., Carson v. Makin, 596 U.S. 767, 810 (2022) (Sotomayor, J., dissenting) (stating that although the majority

opinion allows states to give money to parents through a voucher program, the state could not “contract directly with

schools that teach religion”).

170 Cf., e.g., Roake v. Brumley, 141 F.4th 614, 642–43 (5th Cir. 2025) (comparing the facts of the case before the court

to the facts of a Supreme Court case).

171 See Mitchell, 530 U.S. at 819 n.8 (noting “serious concerns about whether the payments [in Nyquist] were truly

neutral”).

172 Cf., e.g., Freedom from Religion Found., Inc. v. Mack, 49 F.4th 941, 951–54 (5th Cir. 2022) (evaluating historical

evidence of “public, government-sponsored prayer”).

173 Cf., e.g., Does 1–11 v. Bd. of Regents of Univ. of Colo., 100 F.4 th 1251, 1270 (10th Cir. 2024) (concluding a policy

created an excessive entanglement and was therefore unconstitutional).

174 Comm. for Pub. Educ. v. Nyquist, 413 U.S. 756, 774, 777–78 (1973). See also, e.g., Tilton v. Richardson, 403 U.S.

672, 683 (1971) (plurality opinion) (ruling unconstitutional a federal provision that would have allowed federally

funded facilities to revert to religious purposes after twenty years); accord id. at 692 (Douglas, J., dissenting).

Congressional Research Service

16

Establishment Clause Limits on Government Support for Religion

grants from being used for religious purposes.175 While adhering to its prior cases that had

invalidated programs lacking such restrictions, the Court held that this particular program did not

present a risk of violating the Establishment Clause.176 Specifically, the Court said that unlike in

prior cases, the federal health grant program allowed the agency to monitor funds to ensure the

statute did not have the primary effect of advancing religion.177 Further, because the religious

organizations receiving grants might not be “pervasively sectarian” in the same sense as religious

schools, the Court concluded this “less intensive monitoring” did not risk excessive entanglement

with religion.178 Accordingly, while Nyquist has not been overruled, aspects of it have arguably

been weakened by this later case.

A number of existing federal statutes expressly prohibit federal funds from being used for

religious worship or instruction179 or maintaining buildings in which religious instruction or

worship occurs.180 Today, one preliminary question in assessing these restrictions might be

whether they impermissibly discriminate against religious entities, violating the Free Exercise

Clause.181 The restrictions might only be constitutionally permissible if they are required by the

Establishment Clause.182 Under Lemon, whether these restrictions were required by the

Establishment Clause might have depended on the nature of the program, as the Supreme Court’s

analysis of purpose, effect, and entanglement led to different outcomes in different contexts.183

After Kennedy, however, a court might not only have to assess whether the facts of a particular

program are similar to one previously considered by the Supreme Court, but also resolve what

test it should use to evaluate the permissibility of direct funding.

Nyquist provides one example of existing jurisprudence that may be subject to additional

questions in light of recent Supreme Court cases. As the cases discussed illustrate, similar

questions may arise in Establishment Clause challenges to other types of government support for

religion. Ongoing litigation in lower courts may give the Supreme Court additional opportunities

to clarify the doctrine in the future.

175 Bowen v. Kendrick, 487 U.S. 589, 593, 614 (1988).

176 Id. at 614–17.

177 Id. at 615.

178 Id. at 616–17. While some Supreme Court cases expressed particular concern about aid being provided to

“pervasively sectarian” institutions like religious schools, the Court later retreated from the “presumption” that neutral

nonfinancial aid would be used to inculcate religion in such schools. Agostini v. Felton, 521 U.S. 203, 224–26 (1997).

See also Mitchell v. Helms, 530 U.S. 793, 826 (2000) (plurality opinion) (saying “the Court should regret” the period

where it looked to whether a school was “pervasively sectarian”).

179 E.g., 20 U.S.C. § 1011k(c); 34 U.S.C. § 12161(b)(B)(iv); 42 U.S.C. § 290kk-2; id. § 9920(c). Cf., e.g., 25 U.S.C.

§ 278a (“Funds appropriated ... to the Secretary of the Interior for the education of Indian children shall not be used for

the education of such children in elementary and secondary education programs in sectarian schools.”).

180 E.g., 20 U.S.C. § 1087-53(b)(1)(C); id. § 10004(c)(3); 25 U.S.C. § 1813(e); 29 U.S.C. § 3248.

181 See supra note 13.

182 See, e.g., Carson v. Makin, 596 U.S. 767, 780–81 (2022) (holding that a state law excluding religious schools from a

voucher program could not satisfy strict scrutiny where the law did not “offend the Establishment Clause”).

183 See generally Cong. Rsch. Serv., Application of the Lemon Test, CONSTITUTION ANNOTATED,

https://constitution.congress.gov/browse/essay/amdt1-3-4-4/ALDE_00013077/ (last visited Aug. 22, 2025).

Congressional Research Service

17

Establishment Clause Limits on Government Support for Religion

Author Information

Valerie C. Brannon

Legislative Attorney

Disclaimer

This document was prepared by the Congressional Research Service (CRS). CRS serves as nonpartisan

shared staff to congressional committees and Members of Congress. It operates solely at the behest of and

under the direction of Congress. Information in a CRS Report should not be relied upon for purposes other

than public understanding of information that has been provided by CRS to Members of Congress in

connection with CRS’s institutional role. CRS Reports, as a work of the United States Government, are not

subject to copyright protection in the United States. Any CRS Report may be reproduced and distributed in

its entirety without permission from CRS. However, as a CRS Report may include copyrighted images or

material from a third party, you may need to obtain the permission of the copyright holder if you wish to

copy or otherwise use copyrighted material.

Congressional Research Service

R48645 · VERSION 1 · NEW

18