A Little-Noted Puzzle in Religion Law, post-Bremerton
With Kennedy v. Bremerton School District, the Supreme Court finally drove the last nail in the coffin of the Lemon test for Establishment Clause violations (as Justice Scalia said it should do years ago). Over several decades, the Lemon test’s importance had dwindled to the point of having been de facto reversed. But the Court’s hesitance to explicitly overturn that precedent had allowed lower courts to continue drawing on it, as recently as in a case that reached the Court this term, Shurtleff v. Boston (which asked whether displaying a “Christian flag” on public grounds in Boston violated the Establishment Clause).
Justice Gorsuch suggests in his opinion for the Court that Bremerton’s reversal of Lemon should come as no surprise, since the Court “long ago abandoned Lemon and its offshoot[s].” But what Establishment Clause principles was Lemon originally trying to serve? And did those overarching principles have a complement in the Court’s free exercise jurisprudence of the same period? After all, as Gorsuch’s opinion emphasized, a “natural reading” of the First Amendment’s clauses is that they “have ‘complementary’ purposes, not warring ones where one Clause is always sure to prevail[.]” One aspect of the two Clauses’ parallel roles has received little attention: how they bear on a challenged law’s incidental effects on religion—that is, on the ways a law can either benefit or burden religion as a side effect of pursuing some other, religion-neutral goal.
Lemon and related cases’ test for establishments, like the Court’s free exercise cases from the same period, cared about a challenged law’s unintended impact on religion, not just its intent. Now Bremerton has killed Lemon, with its Establishment Clause concern about laws incidentally advancing religion, just as the 1990 Smith case made unintended burdens on religion irrelevant in free exercise cases. But has the Court really stopped caring about incidental effects on religion in both areas? And if not, what new doctrines should it use to scrutinize that impact in the wake of Bremerton and Smith?
I. Attention to Neutral Laws’ Effects on Religion under Lemon and Pre-Smith
Per Lemon, the Establishment Clause required that a law have a “secular legislative purpose,” its “primary effect” be “one that neither advances nor inhibits religion,” and must avoid causing “‘excessive government entanglement with religion.’” The line drawn between the first two prongs, a law’s “purpose” and “effects,” found a close analogue in the Court’s free exercise cases of the same period. The point of that distinction in Lemon, like the Court’s decisions in free exercise cases like Sherbert and Thomas, was to clarify that even when a law doesn’t intentionally, i.e., purposefully, endorse or forbid religious practice, the law might still violate the First Amendment if it has the effect of doing either of those things.
The Court struck down under Lemon, for example, many forms of public support for religious schools, not only when a law seemed purposefully designed to benefit those institutions, but even when the benefits were arguably incidental to another religion-neutral goal. Wolman ruled that states couldn’t provide busing for parochial school field trips, even as part of a nondiscriminatory transportation policy, because the state supervision of teachers required by that provision would be too entangling. Meek upheld in nonpublic schools the provision of some materials and “auxiliary services” (like counseling and speech therapy) that every child was entitled to under neutral education laws, but rejected the lending of any school equipment, like film projectors, “which, from its nature, can be diverted to religious purposes.” (As Chief Justice Rehnquist once put the oddity of this result, “[a] State may lend textbooks on American colonial history, but it may not lend a film on George Washington, or a film projector to show it in history class.”)
Similarly, in free exercise decisions prior to Smith, the Court had held that even laws with a neutral and generally applicable goal—e.g., employment or education laws—can violate the First Amendment if they incidentally burden religious practice in certain ways. A Seventh Day Adventist such as Adele Sherbert, the Court held in an opinion by Justice Brennan, couldn’t be denied unemployment compensation when she was fired for her inability to work on Saturdays, her Sabbath day. A Jehovah’s Witness working in a factory was entitled to the same compensation when he was fired for not contributing to arms productions. Amish parents couldn’t be penalized for ceasing their children’s public education after the eighth grade.
Then in Smith, in 1990, the Court officially reversed course: Justice Scalia, writing for the majority, held that the First Amendment’s protection did not require religious exemptions from laws that were otherwise neutral in purpose and applicable to everyone. The decision turned in part on some of the justices’ view that courts are not well-equipped to weigh and compare burdens on religion against whatever state interests a law might be trying to serve. Yet attention to those kinds of burdens did not disappear: bipartisan efforts to restore the Court’s prior approach of considering them, at least in statutory law, brought about federal and state versions of the Religious Freedom Restoration Act in the 1990s. And today, not all the conservative justices continue to agree with Scalia’s interpretation. In last year’s Fulton decision, Justice Alito put forward a thorough and compelling case for reversing Smith.
II. How Effects Still Count in Our Religion Clauses Jurisprudence
Though Smith remains on the books for now and Lemon has been officially reversed, the Court’s reasoning about incidental effects has played an important role in its recent free exercise and establishment cases. We can start with the cases involving public funding for religious schools—like Trinity Lutheran, Espinoza, and this term Carson. In these decisions, state refusals to give that benefit have been characterized as unconstitutional forms of discrimination against religion. As Chief Justice Roberts explained in Carson, the Court has “repeatedly held that a State violates the Free Exercise Clause when it excludes religious observers from otherwise available public benefits.” And while Roberts focuses on the discriminatory intent of Maine’s policy, which even Smith would regard as unconstitutional, he cites, for support, pre-Smith cases like Lyng, Sherbert, and Thomas, which concerned incidental burdens on religion.
The opinion explicitly points, for example, to Lyng’s holding that the “Free Exercise Clause of the First Amendment protects against ‘indirect coercions or penalties on the free exercise of religion, not just outright prohibitions.’” The law at issue in Carson, like those in Trinity Lutheran and Espinoza, violated that rule by putting religious schools to a difficult choice: between retaining their religious mission, on the one hand, and qualifying for a public benefit on the other. That is a point about the law’s effect, rather than its intent.
This reasoning is spelled out quite explicitly in Trinity Lutheran, and is in turn traced there to McDaniel v. Paty (another pre-Smith precedent, holding that under the Free Exercise Clause clergy members could not be prohibited from serving as political delegates). Take this passage from the majority opinion, also authored by the Chief Justice:
. . . the Department’s policy puts Trinity Lutheran to a choice: It may participate in an otherwise available benefit program or remain a religious institution. Of course, Trinity Lutheran is free to continue operating as a church . . . . But that freedom comes at the cost of automatic and absolute exclusion from the benefits of a public program for which the Center is otherwise fully qualified.”
So in short, Carson confirms that concern about effects is still operative in the Court’s thinking about free exercise, in spite of Smith’s retention (for now).
Likewise under the Establishment Clause: while Lemon has been increasingly ignored, its attention to incidental effects on religion has continued to play a role in the Court’s emerging “history and tradition” approach to interpreting the Clause—as recently as in American Legion v. American Humanist Association, the case that many had expected would reverse Lemon. In that decision, Justice Alito’s opinion for the majority argued that the Lemon test would not help resolve the question at hand: whether a WWII memorial cross could continue to stand in a Maryland traffic circle. That was because Lemon’s “secular purpose” inquiry didn’t have a clear answer in the case: the original reasons for the display, Alito argued, were hard to discern and had evolved over time to include secular ones (such as reverence for the soldiers’ sacrifice). Here’s a fair synopsis of the Court’s reasoning:
The cross is undoubtedly a Christian symbol, but that fact should not blind us to everything else that the Bladensburg Cross has come to represent. For some, that monument is a symbolic resting place for ancestors who never returned home. For others, it is a place for the community to gather and honor all veterans and their sacrifices for our Nation. For others still, it is a historical landmark. For many of these people, destroying or defacing the Cross that stood undisturbed for nearly a century would not be neutral and would not further the ideals of respect and tolerance embodied in the First Amendment. For all these reasons, the Cross does not offend the Constitution.
This passage is an argument about the memorial’s effects: the multiple meanings that observers could draw from the memorial—many of which may not have been intended by those who first erected it—are themselves treated as a reason to preserve a public display of religion. They are also meant to answer Justice Ginsburg’s concern, in dissent, about the memorial’s effects on those who feel excluded by its Christian associations.
So, to summarize: under Smith, and after Lemon’s reversal this term, incidental effects on religion are officially no longer relevant under the First Amendment’s religion clauses. But a concern about those effects has still crept into cases of both kinds. The Court either needs to jettison that concern or—if pre-Smith precedents and the Lemon test are not to be the standards—give an account of exactly when and how a law’s impact on religion matters. There is an opportunity for clarification here, which would find a perfect occasion in any future cases that ask the Court (as Fulton did) to consider overturning Smith. And by answering the question more fully, the Court could further spell out the complementarity of our religion clauses, the point Bremerton so strongly insisted upon this term. It’s exciting to think of the coherence that attention to this problem would help restore in our constitutional law.
III. When Law’s Effects on Religion Should Matter
Any effort to clarify the weight and relevance of incidental effects on religion would need to begin from a deeper explanation of what general principle on that point would make the religion clauses truly complementary.
Some members of the Court may hesitate to propose a “grand unified theory” of the religion clauses like the one offered in Lemon and just rejected in Bremerton. After all, Bremerton cast its more historical approach to interpreting the Establishment Clause as an express contrast to what it called Lemon’s overly “ambitious,” “abstract,” and “ahistorical” method. But even if one agrees with that characterization, it is hard to see how the Court can develop a coherent approach to handling incidental effects on religion, without some general theorizing about the broader relationship between the clauses. A few further questions to get the gears turning:
Is the First Amendment best understood to require minimal incidental burdens for religion, and maximum incidental benefits? That is the picture we get from the scholarship of Michael McConnell, for example, who has argued both that Smith is incompatible with an originalist understanding of the First Amendment, and that the Establishment Clause was originally understood to rule out only a very narrow range of impermissible supports for religion (e.g., the creation of a national church, or taxation for the purpose of paying religious ministers). But other constitutional scholars like Phillip Muñoz have proposed a different originalist account, one on which courts don’t need to limit incidental benefits or burdens for religion. For Muñoz, early American debates about religious freedom were concerned with preventing a national establishment of religion, and many thought that public support for religion in individual states was permissible. And the framers understood the Constitution to require religious exemptions from neutral laws only in a very limited number of cases (e.g., military draft exemptions for Quakers).
We can also sketch out some more targeted questions for free exercise and establishment cases yet to come, starting with the possibility of Smith’s reversal. Should the Court continue to hold, as Justice Scalia’s opinion did, that judicial weighing of burdens on religion is beyond the scope of the First Amendment as it was originally understood? From one perspective, such judicial parsing is an approach to free exercise that the framers never envisioned our Court would take. But from another, incidental burdens on religion are a frequent and inescapable fact of our current culture, in which ample space for religious freedom is increasingly challenged. Today, Smith’s interpretation of the First Amendment seems to have trouble fitting an American milieu in which purportedly neutral anti-discrimination laws informed by progressive tenets are burdening activities inspired by religious (or any) dissent.
If the Court does decide to overturn Smith, it will have to take up questions about incidental burdens on free exercise more directly. How best do we determine, for example, what counts as a “substantial” burden triggering strict scrutiny? Justice Barrett’s concurrence in Fulton pinpointed that question as foremost among those Smith’s reversal would ignite. (Sherif Girgis and I have both written papers on the topic.) Is there a “history and tradition” approach to substantial burden analysis still to be developed?
Analogously, even after reversing Lemon, it might make sense for the Court to build some of that case’s sensitivity to incidental support for religion into its new approach to establishment. This is arguably what Chief Justice Rehnquist began to do in opinions like Zelman—cases that drew lines, for example, between public funds channeled to religion through individual choices (e.g., of parents in selecting schools for their children) and those distributed by the government directly. Through recent cases like Trinity Lutheran and American Legion, the Court has suggested that more direct or visible forms of support for religion are compatible with the Establishment Clause. But what distinctions might remain to be drawn among those? Are there public displays of religion that remain unconstitutional because they indirectly endorse specific creeds in ways the framers would probably have rejected? Are some kinds of subsidy for religious exercise unconstitutional because in their effects they approximate too closely establishments that our framers thought the First Amendment prohibited?
Although Lemon’s reversal will strike many as long overdue, the Court’s official rejection of it may have bigger implications than we realize. A frequent complaint of judges and justices has been the seeming irreconcilability of our religious liberty precedents (e.g., on public funding for religion). The Court’s decisiveness on Lemon this term, and its signals that Smith’s reversal might be down the road, could be just the impetus toward coherence needed by our judiciary and those scholars whose work is devoted to guiding it.
* Postdoctoral Fellow, Ethics & Public Policy Center. Ph.D., Princeton, 2020.
 Kennedy v. Bremerton Sch. Dist., 142 S. Ct. 2407 (2022).
 Lemon v. Kurtzman, 403 U.S. 602 (1971).
 See Lamb’s Chapel v. Ctr. Moriches Union Free Sch. Dist., 508 U.S. 384, 397–401 (1993) (Scalia, J., concurring) (calling for reversal of Lemon).
 See Bremerton, 142 S. Ct. at 2427 (declaring Lemon “long ago abandoned”).
 See id.
 Shurtleff v. City of Bos., 142 S. Ct. 1583 (2022).
 Bremerton, 142 S. Ct. at 2427.
 Id. at 2426 (citation omitted).
 Emp. Div., Dep’t Hum. Res. of Or. v. Smith, 494 U.S. 872 (1990).
 See id. at 885.
 Cf. Fulton v. City of Phila., 141 S. Ct. 1868, 1883–84 (Barrett, J., concurring) (asking “what should replace Smith?” and listing “a number of issues to work through if Smith were overruled”).
 Lemon, 403 U.S. at 612–13 (citation omitted).
 Sherbert v. Verner, 374 U.S. 398 (1963).
 Thomas v. Rev. Bd. of the Ind. Emp. Sec. Div., 450 U.S. 707 (1981).
 See Wolman v. Walter, 433 U.S. 229, 254 (1977).
 See Meek v. Pittenger, 421 U.S. 349, 357 (1975).
 See Wallace v. Jaffree, 472 U.S. 38, 110–11 (1985) (Rehnquist, J., dissenting).
 See Sherbert, 374 U.S. 398.
 See Thomas, 450 U.S. 707.
 See Wis. v. Yoder, 406 U.S. 205 (1972).
 See Smith, 494 U.S. at 882.
 See id. at 886–87.
 See, e.g., 42 U.S.C. §2000bb (1993).
 See Fulton, 1883–1926 (Alito, J., concurring).
 Trinity Lutheran Church of Columbia, Inc., v. Comer, 137 S. Ct. 2012 (2017).
 Espinoza v. Mont. Dep’t of Revenue, 140 S. Ct. 2246 (2020).
 Carson v. Makin, 142 S. Ct. 1987 (2022).
 Id. at 1996.
 Lyng v. Nw. Indian Cemetery Protective Ass’n, 485 U.S. 439 (1988).
 Id. at 450.
 McDaniel v. Paty, 435 U.S. 618 (1978).
 Trinity Lutheran, 137 S. Ct. at 2021–22.
 Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067 (2019).
 See id. at 2082–85.
 Id. at 2090.
 See id. at 2107 (Ginsburg, J., dissenting) (“The principal symbol of Christianity around the world should not loom over public thoroughfares, suggesting official recognition of that religion’s paramountcy.”).
 See Bremerton, 142 S. Ct. at 2426.
 Id. at 2427.
 See Michael McConnell, Free Exercise Revisionism and the Smith Decision, 57 U. Chi. L. Rev. 1109 (1990).
 Brief for the Becket Fund for Religious Liberty as Amicus Curiae Supporting Petitioners at 13–15, American Legion v. American Humanist Assn., 139 S. Ct. 2067.
 See Vincent Phillip Muñoz, The Original Meaning of the Establishment Clause and the Impossibility of Its Incorporation, 8 U. Pa. J. Const. L. 585, 605–23 (2006).
 See Vincent Phillip Muñoz, Two Concepts of Religious Liberty: The Natural Rights and Moral Autonomy Approaches to the Free Exercise of Religion, 110 Am. Pol. Sci. Rev. 369, 374 (2016); see also Muñoz, The Original Meaning of the Free Exercise Clause: The Evidence from the First Congress, 31 Harv. J. L. & Pub. Pol’y 1083 (2008).
 See Fulton, 141 S. Ct. at 1883–84 (Barrett, J., concurring) (listing questions likely to arise).
 See Sherif Girgis, Defining “Substantial Burdens” on Religion and Other Liberties, 108 Va. L. Rev. (forthcoming 2022); Gabrielle M. Girgis, What is a “Substantial Burden” on Religion under RFRA and the First Amendment?, 97 Wash. U. L. Rev. 1755 (2020).
 See Zelman v. Simmons-Harris, 536 U.S. 639, 648–53 (2002).Read More »