The Uncertain Good of Overruling Smith


Gary J. Simson

Thirty years ago, when the Supreme Court decided Employment Division v. Smith,1494 U.S. 872 (1990). I never imagined that I’d ever respond with anything but enthusiasm to the news that the Court would be deciding a case squarely raising the question whether Smith should be overruled. In early November, however, the Court will be hearing oral argument in just such a case, Fulton v. City of Philadelphia,2Cert. granted 140 S. Ct. 1104 (Feb. 24, 2020). and I look forward to the Court’s decision in the case with trepidation at best.

At first glance, Justice Scalia’s opinion in 1990 for a 5-4 majority in Smith may not seem all that revolutionary. In holding that the Free Exercise Clause generally doesn’t afford religious claimants a vehicle for securing court-ordered exemptions from generally applicable laws that, though religion-neutral on their face, substantially burden the claimants’ religious liberty, Justice Scalia went to considerable lengths to make it appear as if the Court were saying nothing particularly new. He maintained that the Court had long been interpreting the Clause as not authorizing court-ordered exemptions, and he characterized cases seemingly to the contrary as exceptions of one or another sort. In terms of candor and persuasive argument, Justice Scalia’s opinion for the Court clearly left a great deal to be desired. The bottom line, though, is that he managed to cobble together a majority for a new conception of the Clause – one plainly less protective of religious liberty than the one in place since the early 1960s, if not longer.

To say that the new conception of the Clause met with popular disapproval would be a gross understatement. It sparked strong opposition across the political spectrum, and in 1993 a virtually unanimous Congress enacted the Religious Freedom Restoration Act (RFRA)342 U.S.C. §§ 2000bb to 2000bb-4. for the express purpose of undoing what the Court in Smith had done. Several years later, the Court held that, as applied to state and local government, RFRA was beyond Congress’s constitutional authority to enact, but it left RFRA intact as to federal governmental actions.4City of Boerne v. Flores, 521 U.S. 507 (1997).

As the Court this Term in the Fulton case ponders the question of whether to overrule Smith, RFRA’s importance extends beyond its continued applicability in the federal realm. In particular, it stands as a codification of the conception of religious liberty that mustered such widespread support among Congress and the populace as a replacement for the conception that the 5-4 majority in Smith put in place. What was the conception of religious liberty that Congress, in enacting RFRA, codified into law? By all indications, it was the longstanding conception that Smith summarily banished with one blow.

To say that the new conception of the [Free Exercise] Clause met with popular disapproval would be a gross understatement.

As the title, Religious Freedom Restoration Act, itself affirms, RFRA was very deliberately designed to restore the conception of religious liberty that prevailed prior to Smith. RFRA explicitly identifies Smith as the culprit that “virtually eliminated” the “workable test” that had been “set forth in prior Federal court rulings” for “striking sensible balances” between “religious liberty,” on the one hand, and “governmental interests,” on the other. Lest anyone wonder which prior rulings Congress had in mind, RFRA explicitly named two leading Supreme Court free exercise decisions of the era upended by Smith Sherbert v. Verner5374 U.S. 398 (1963) in 1963 and Wisconsin v. Yoder6406 U.S. 205 (1972). in 1972. Drawing directly from language in the majority opinions of the two cases, RFRA provided that religious claimants must be granted an exemption from a generally applicable law if they can prove that the law “substantially burden[s]” their religious liberty and the government can’t prove that requiring them to comply with the law is the “least restrictive means” of serving a “compelling governmental interest.”

Although the religious claimants prevailed in Sherbert and Yoder and were granted the court-ordered exemptions they sought from generally applicable laws, the Congress that in 1993 enacted RFRA to “restore” the “workable test” of the Sherbert-Yoder era couldn’t have helped but be aware that success was hardly the typical fate of free exercise claimants who sought court-ordered exemptions during that era. The won-lost record in the Supreme Court of free exercise claimants seeking court-ordered exemptions was unenviable, to say the least: five victories and fourteen defeats. Furthermore, it appears that such claimants’ success rate in the federal appeals courts was even worse. According to a much-cited 1992 study in the Virginia Law Review,7James E. Ryan, Note, Smith and the Religious Freedom Restoration Act: An Iconoclastic Assessment, 78 Virginia Law Review 1407 (1992). in the ten years before Smith, there were ninety-seven reported federal appellate decisions on free exercise exemption claims, and religious claimants emerged victorious in only twelve.

But if Congress indeed understood in enacting RFRA that it was reinstating a test for exemptions that only a relatively small proportion of free exercise claimants would be able to meet, why bother? Most obviously, some successful free exercise claimants are better than none at all. Beyond that, however, I suggest that Congress in enacting RFRA, like the Supreme Court in Sherbert, Yoder, and other cases of that era, saw free exercise exemptions as playing a minor, but vital, role in the legislative process: a safety-valve role of remedying a certain kind of occasional malfunction in the legislative process that is probably inevitable in our state and federal systems of majoritarian-minded lawmaking bodies.

In essence, when legislators adopt a generally applicable law, they are unlikely to overlook or fail to consider seriously the law’s impact on adherents of widely shared religions and the benefits of creating a legislative exception to relieve such adherents of a substantial burden that the law would otherwise place on their ability to practice their religion. Very simply, legislators are inattentive or insensitive to the needs of adherents of widely shared religions at their peril. Such behavior is apt to come back and bite those legislators at the polls. In contrast, when adherents of religions shared by only a small minority of the population are adversely affected by a generally applicable law, the political process doesn’t provide the same kind of check. Legislators are much more likely to be unresponsive to those constituents’ needs because they are much more likely not to recognize that the law will even have an adverse effect or that the adverse effect is as serious as it in fact is.

The Sherbert-Yoder test struck what Congress in RFRA saw as a “sensible balance” between religious liberty and governmental interests because it provided a judicial safety valve that operated at the margins. The safety-valve conception of court-ordered exemptions reflected in the test was fully consistent with the notion that legislatures typically should be trusted to weigh the costs and benefits fairly in deciding whether to accommodate religious exercise by making an exception to a generally applicable law. The generally applicable laws at issue in Sherbert and Yoder – a state unemployment compensation law and a state compulsory education law, respectively – were typical of the sort of laws challenged in the free exercise cases of that era: laws adopted with unquestionably important and honorable general welfare objectives in mind. Unlike, for example, laws explicitly singling out racial minorities for disadvantage, the laws challenged in Sherbert and Yoder were fundamentally sound laws that sparked no suspicion that, whatever the lawmakers may have claimed, their objectives in fact were invidious ones. For that reason, it was entirely fitting that the Supreme Court and lower courts applied the least-restrictive-means requirement of the Sherbert-Yoder test with significantly greater deference than they customarily afforded the state when applying the comparable requirement in equal protection challenges to racial classifications.

In short, I regard the Sherbert-Yoder test, as applied by the Supreme Court and lower courts in the pre-Smith era, as striking the “sensible balance” that Congress in RFRA characterized it as striking. That’s not to say that, if I were on the Court during the pre-Smith era, I would have decided the various free exercise exemption cases that came before it no differently than the Court decided them. In particular, I would have been slow to join the Court in insisting that the Orthodox Jewish merchants in Braunfeld v. Brown8366 U.S. 599 (1961). abide by Maryland’s Sunday Closing Laws even though doing so might well drive them out of business. Nonetheless, I don’t see Braunfeld or any of the Court’s other free exercise exemption cases of that era as cases that couldn’t reasonably be decided as they were.

When Justice Scalia in Smith essentially jettisoned the safety valve provided by the prior case law, he offered an alternative conception of religious liberty – one that relegated its realization, at least insofar as exemptions from generally applicable laws are concerned, to majority will as reflected in the legislative process:

[A] society that believes in the negative protection accorded to religious belief can be expected to be solicitous of that value in its legislation. . . . But to say that a nondiscriminatory religious practice exemption is permitted, or even that it is desirable, is not to say that it is constitutionally required, and that the appropriate occasions for its creation can be discerned by the courts. It may fairly be said that leaving accommodation to the political process will place at a relative disadvantage those religious practices that are not widely engaged in; but that unavoidable consequence of democratic government must be preferred to a system in which each conscience is a law unto itself. . . .9Smith, 494 U.S. at 890.

Justice Scalia in 2013 / Wikimedia

Unlike most of Justice Scalia’s majority opinion, the above portion can’t be faulted for lack of candor. In it, he expressly identified two key ingredients of the Smith Court’s conception of religious liberty: First, legislatures, not courts, are the appropriate forums to decide whether to create religious exemptions to generally applicable laws. Second, the fact that legislatures are not likely to grant exemptions to adherents of minority religions is simply a natural consequence of living in a democratic society that selects its legislators by majority vote. In addition, Justice Scalia at least implicitly appeared to suggest a third ingredient: A legislative decision to grant a religious exemption is purely a matter of legislative discretion, not a decision bounded by the Free Exercise Clause’s companion provision – the Establishment Clause. 

Those ingredients of the Smith Court’s conception of religious liberty obviously differ sharply from those that constituted the pre-Smith era’s conception. Furthermore, as I’ll discuss below, they also differ significantly from those that make up yet another conception of religious liberty – the conception that the Justices currently most eager to revisit Smith appear to hold.

I’m not certain when a third conception of religious liberty first began to gather momentum on the Court, but I’d date its emergence as a force no later than the Supreme Court’s widely discussed 2014 decision in Burwell v. Hobby Lobby Stores, Inc.10573 U.S. 682 (2014). Hobby Lobby and two other closely held for-profit corporations sued under RFRA for an exemption from federal regulations promulgated pursuant to the Affordable Care Act. The regulations require a broad range of employers to provide their employees with health care coverage that includes certain contraceptive methods that work in a way that, in the three corporate plaintiffs’ view, amounts to inducing abortion. The plaintiffs framed their complaint in terminology mirroring RFRA and the Sherbert-Yoder test on which RFRA was modeled. Specifically, they maintained that they adhere to Christian beliefs that prohibit abortion as sinful, that requiring them to provide health care coverage that their employees may use for abortion-like contraceptive methods substantially burdens their religious liberty, and that the court should exempt them from having to comply with the relevant regulations because requiring their compliance is not the least restrictive means that the federal government could use to achieve its objectives.

By a 5-4 vote, the Court in Hobby Lobby held that RFRA required granting the plaintiffs the requested exemption. As the four dissenters argued at length, the Court’s reasoning in granting the exemption is simply incomprehensible in terms of the requirements for a court-ordered exemption set forth in RFRA if, as Congress by all indications intended, those requirements are understood to mean what they meant during the era that Smith abruptly brought to a close. Given current space limitations for this essay, I won’t attempt to reiterate those arguments here. Instead, suffice it to say that the Court’s reasoning was unpersuasive in two basic respects. First, the burden on the plaintiffs’ religious liberty that the Hobby Lobby Court characterized as “substantial” fell significantly short of the formidable burden standard reflected in the pre-Smith case law. Second, the least-restrictive-means showing that the Hobby Lobby Court found inadequate to justify the burden was obviously adequate under the interpretation of least-restrictive-means in the pre-Smith case law. In fact, it was so obviously adequate that the Court in Hobby Lobby felt obliged to concede that its finding of inadequacy was predicated on an understanding of RFRA as codifying a more exemption-friendly test than the test the Court had applied in the pre-Smith era.

Although the Court’s opinion in Hobby Lobby was an interpretation of RFRA and not the Free Exercise Clause itself, the Court’s “immoderate reading of RFRA,” as the dissenters put it, very reasonably may be understood to signal that the Court was moving toward an interpretation of the Free Exercise Clause significantly at odds with Smith. In terms of the key ingredients, described above, of the Smith Court’s conception of religious liberty, the Court’s opinion in Hobby Lobby most obviously is in tension with the Smith Court’s view that legislatures, not courts, are the appropriate forums to decide whether to create religious exemptions to generally applicable laws. True, RFRA – the basis for the exemption granted to Hobby Lobby and the other plaintiffs – was the work of a legislature, not a court. However, RFRA’s delegation to courts of the authority to carve out exemptions when its terms are met plainly vested in courts the ultimate responsibility to decide when an exemption is or isn’t warranted, and one would have expected a Court skeptical of courts’ competence to carve out exemptions to interpret narrowly, not expansively, RFRA’s license to courts to carve out such exemptions.

Unlike the Smith Court, the Court in Hobby Lobby gave no indication that it was content to leave religious groups seeking an exemption from a generally applicable law to fend for themselves in the political process. The Court’s aggressive reading of RFRA seemed to announce loud and clear that religious groups in search of an exemption have an ally in the Court. Furthermore, that aggressive reading of RFRA in the case at hand also seemed to affirm that the Justices in the majority in Hobby Lobby, unlike those in the majority in Sherbert, Yoder, and other cases of that era, were not operating under any sort of safety-valve conception of the courts’ role in protecting religious exercise. The corporate plaintiffs in Hobby Lobby were anything but the kind of political outsiders featured in the pre-Smith era cases, and their religious opposition to abortion and abortifacients was hardly a tenet shared by only a small religious minority and apt to go unnoticed or unappreciated by legislative majorities.

The only ingredient of the Smith Court’s conception of religious liberty that the Hobby Lobby Court seemed to embrace was an attitude toward the Establishment Clause as placing little, if any, constraint on the creation of religious exemptions to generally applicable laws. The dissenters in Hobby Lobby maintained that, in taking a strong pro-exemption reading of RFRA, the Court had “ventured into a minefield” of Establishment Clause problems. If those in the majority were at all worried about any such problems, however, it wasn’t apparent from anything said in the opinion of the Court.

Another indication of the Court’s movement toward a new conception of religious liberty came in 2019 in the exceptionally unlikely form of an opinion concurring in the Court’s denial in Kennedy v. Bremerton School District11139 S. Ct. 634 (2019). For a fuller statement of the facts than in the concurring opinion, see Kennedy v. Bremerton School Dist., 869 F.3d 813 (9th Cir. 2017). of a petition to review a federal appellate court’s rejection of a free speech claim. Not surprisingly, in light of the thousands of cases that the Justices are petitioned to review each year, the Justices deny the overwhelming majority of those petitions and don’t attempt to provide an opinion explaining to each unsuccessful petitioner the reason for their decision. On rare occasions, a Justice who votes to review a case that doesn’t get the four votes needed to secure review feels strongly enough about an issue in the case to take the time to write a dissent from the denial of review. Even rarer is for a Justice who agrees with the Court’s denial of review to take the time to write an opinion concurring in that denial. For a Justice writing such a concurrence to use that opportunity, as Justice Alito did in Kennedy, to write about an issue not even raised in the petition for review is beyond rare. That Justice obviously has something that he or she is bursting to say. In Kennedy that something was a not very subtle statement that the time to revisit Smith was long overdue, and that statement was made by a Justice speaking not only for himself, but for three of his colleagues – Justices Gorsuch, Kavanaugh, and Thomas – as well.

The facts of Kennedy warrant brief recitation because they provide important insight into the nature of the free exercise exemption test that Justices Alito, Gorsuch, Kavanaugh, and Thomas appear to be contemplating putting in place if and when Smith is overruled. The petitioner, Joseph Kennedy, was fired from his position as football coach at a public high school, and the firing was upheld by a federal district court and court of appeals. In his petition to the Supreme Court, Kennedy maintained that the firing had violated his free speech rights, because it was predicated on his refusal to desist from a practice that the school superintendent had told him would require his dismissal: kneeling and praying at midfield immediately after each of his team’s games. Although Kennedy didn’t ask his players to join him at midfield, many of them commonly did, as did some members of the opposing team and some of the fans in attendance. The superintendent had told Kennedy that his practice raised serious Establishment Clause problems and that school policies required all staff to avoid acting in their employment in ways apt to encourage or discourage student engagement in religious activities.

Justice Alito devoted almost all of his opinion to Kennedy’s free speech claim and emphasized that in his view the claim couldn’t be meaningfully reviewed until some factual matters in the case had been clarified in the lower courts. Then, after noting that “the petition now before us is based solely on the Free Speech Clause,” he called attention in the final two paragraphs of his opinion to the “live” free exercise claim that Kennedy could still elect to pursue. He suggested that Kennedy may have decided to “rely primarily” on his free speech claim “due to certain decisions of this Court,” and he then named Smith as the likely reason for Kennedy’s reticence about pressing his free exercise rights. He described Smith in unmistakably critical terms as a case in which the Court “drastically cut back on the protection provided by the Free Exercise Clause.” Finally, lest anyone somehow fail to grasp the reason for his detour into free exercise law, he concluded by noting that there was nothing more to be said at the moment about Smith because “we have not been asked to revisit” Smith in the case at hand.

In and of itself, Justice Alito’s and his three colleagues’ readiness to revisit Smith doesn’t tell us a lot about the conception of religious liberty that they hope to put in place, but their encouraging Kennedy to pursue his free exercise claim does. Under the Court’s decision in 2000 in Santa Fe Independent School District v. Doe,12530 U.S. 290 (2000). Kennedy’s prayer practice almost certainly violated the Establishment Clause. In Doe, the Court held that the Establishment Clause prohibited a public high school from having an elected student representative deliver a prayer over the public address system at the start of each of the school’s home football games. Granted, students watching Kennedy’s team play or playing for him could avoid hearing or participating in Kennedy’s midfield prayer, whereas those attending the games in Doe couldn’t help but hear the prayer, like it or not. However, Kennedy’s prayer practice – an end-of-game midfield prayer by a prominent paid employee of the school joined by various students and in full view of many others – was at least as likely to send a message of school endorsement of prayer as the student-led prayer in Doe. If Justices Alito, Gorsuch, Kavanaugh, and Thomas were looking for a fact setting that would facilitate their making clear that their conception of religious liberty revitalizes the Free Exercise Clause at the expense of Establishment Clause values, what better way to make their point than slam Smith in the context of the Kennedy facts and treat Doe and the Establishment Clause as if they didn’t exist?

The Kennedy facts also provided an ideal opportunity for the four Justices to make clear how very expansive a view they are prepared to take of the demands of free exercise. If, as they appeared to be suggesting, Kennedy had a very credible free exercise claim if Smith were out of the way, the magnitude of the burden on religious exercise needed to justify a court’s carving out a free exercise exemption is remarkably low. According to Kennedy, he felt called at the end of every game to “take a knee at the 50-yard line and offer a brief, quiet prayer of thanksgiving for player safety, sportsmanship, and spirited competition.” To use the Court’s language in the Sherbert-Yoder test and Congress’s in RFRA, did the school “substantially burden” Kennedy’s religious exercise by telling him that, if he wished to offer that prayer, he had to do so at some time and place other than in the middle of the football field right after the game? The four Justices implicitly seemed to be answering “yes.” In light of the ease with which Kennedy could have offered his prayer after the game in a more private venue, such as his office or car or home, the burden on his free exercise qualifies as “substantial” only if “substantial” is understood far more broadly than it ever was understood in the Court’s free exercise cases prior to Smith.

On rare occasions, one or more Justices who vote to review a case that doesn’t get the four votes needed to secure review feel strongly enough about an issue in the case to take the time to write a dissent. 

Lastly, the facts in Kennedy also offered a perfect backdrop to clarify that the emerging conception of religious liberty is one having nothing in particular to do with ensuring protection for religious beliefs apt to get lost in a majoritarian political process. Engaging in prayer to express gratitude to a Higher Power for the good things in our lives is hardly some sort of obscure religious practice unfamiliar to, or unappreciated by, majority-minded legislatures. For confirmation, look no further than the legislative prayer practices that the Court approved over Establishment Clause objections in 1983 in Marsh v. Chambers13463 U.S. 783 (1983). and in 2014 in Town of Greece v. Galloway.14572 U.S. 565 (2014).

As my final piece of evidence of the new conception of religious liberty that I anticipate the Justices will put in place if they overrule Smith, I call attention to the Court’s rulings in two free exercise cases not involving exemption claims, Trinity Lutheran Church v. Comer15137 S. Ct. 2012 (2017). in 2017 and Espinoza v. Montana Department of Revenue16140 S. Ct. 2246 (2020). in 2020. Trinity Lutheran raised the question whether a state, though not compelled by the Establishment Clause to deny certain funding to religious institutions, was barred by the Free Exercise Clause from providing such funding only to secular institutions. Missouri had denied a church a grant for playground resurfacing while making available such grants to secular nonprofits. Espinoza raised the question whether a state, though not compelled by the Establishment Clause to deny certain funding to private individuals for use for religious purposes, was barred by the Free Exercise Clause from providing such funding only to private actors for secular uses. Montana had denied tuition assistance to parents sending their children to private sectarian schools while providing such assistance to parents sending their children to private nonsectarian schools. In both cases, the states maintained that their state constitutions required them to take a more separationist approach to funding than the Supreme Court has interpreted the Establishment Clause to require.

With Chief Justice Roberts writing for the Court in both cases, the Court held that the Free Exercise Clause required the states to disregard their state constitutional limitations. Instead, they must provide funding without regard to whether the would-be recipient is a religious or secular institution or whether the would-be recipient plans to use the funds to pay for sectarian or nonsectarian schooling. For the states to do otherwise, the Chief Justice explained, would constitute discrimination against religion.

Trinity Lutheran and Espinoza raise difficult issues on which reasonable people may disagree. For present purposes, however, I call attention to the cases only to highlight the strong interpretation of free exercise that they reflect and the extent to which that interpretation coincides with the new conception of religious liberty reflected in the Court’s opinion in Hobby Lobby and in the concurring opinion in Kennedy. Like those two opinions, the Court’s opinions in Trinity Lutheran and Espinoza vigorously enforce Free Exercise values and marginalize values that the Court for many years regarded as central to the Establishment Clause.

After reading the Court’s opinions in Trinity Lutheran and Espinoza, it is instructive to go back and read Justice Black’s majority opinion in Everson v. Board of Education,17330 U.S. 1 (1947). the 1947 decision in which the Court first held that the Establishment Clause applies not only to federal governmental actions but state and local ones as well. It’s striking today to realize that when Justice Black stated that the Establishment Clause “means” that “[n]o 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,” he was announcing a principle that all nine Justices in Everson accepted as true. By a 5-4 vote, the Court in Everson held that that principle wasn’t violated by a state law authorizing localities to cover the cost of children’s transportation to parochial schools. However, the Justices weren’t divided about the validity and importance of the principle itself. Both the majority and dissent hailed its origins in Madison’s famous Memorial and Remonstrance and in the renowned Virginia Bill for Religious Liberty that Jefferson had such a large part in drafting. To borrow language that both the majority and dissent quoted from the Virginia Bill, all nine Justices saw it as a fundamental violation of religious conscience for government to force anyone to “furnish contributions of money for the propagation of opinions which he disbelieves.”

Similarly, looking back from Trinity Lutheran and Espinoza, it’s eye-opening to read Chief Justice Burger’s once-influential explanation in 1971 of the Establishment Clause’s foundational principles. Writing for all but one member of the Court in Lemon v. Kurtzman,18403 U.S. 602 (1971). the Chief Justice underlined that “[i]n the absence of precisely stated constitutional prohibitions, we must draw lines with reference to the three main evils against which the Establishment Clause was intended to afford protection,” and one of the three evils that he named was “financial support” of religion.

For all practical purposes, the Court expunged financial support of religion from that list of Establishment Clause evils when it held in 2002 in Zelman v. Simmons-Harris19536 U.S. 639 (2002). that the Clause was no obstacle to a state’s providing tuition assistance to enable children to attend sectarian schools. In Espinoza the Court didn’t quite complete a 180 degree turn when it held that the Free Exercise Clause precluded Montana from subsidizing private-school tuition only at nonsectarian private schools. The Court is yet to hold that a state is discriminating against religion in violation of free exercise when it provides children with a free public education in nonsectarian schools but fails to fund children’s education at sectarian schools. However, there can be no question that the opinions of the Court in Trinity Lutheran and Espinoza, like the Court’s opinion in Hobby Lobby and the concurring opinion in Kennedy, go a long way toward a new conception of religious liberty that bears almost no resemblance to the one that Smith displaced and relatively little resemblance to the one that Smith put in place.

In short, Smith was wrongly decided, but at present, the Court appears primed to compound that error. To overrule Smith but put in its place a lenient standard for free exercise exemptions shortchanges Establishment Clause values.20See generally Gary J. Simson, Permissible Accommodation or Impermissible Endorsement? A Proposed Approach to Religious Exemptions and the Establishment Clause, 106 Kentucky Law Journal. 535 (2018). It also threatens to hobble the government’s ability to pursue effectively a host of important general welfare objectives. Much as Smith should be revisited, even more important is the Court’s first rethinking the path that it has been on. ♦


Gary J. Simson is the Macon Chair in Law and former Law Dean and Senior Vice Provost at Mercer University, as well as Professor Emeritus of Law at Cornell University. He is the author of numerous articles on the Establishment and Free Exercise Clauses and other constitutional law topics.