Why This Supreme Court Should Overrule Employment Division v. Smith

Raphael A. Friedman

Photo of the Village Church of Lincolnshire by Andrew Seaman on Unsplash.

For several decades, parties asserting religious liberty claims under the Free Exercise Clause saw little success at the Supreme Court. Over the past few years, however, religious groups have fared better. Recently, they prevailed in cases challenging various restrictions imposed on houses of worship because of the COVID-19 pandemic. And in June, the Court ruled in a (shockingly) unanimous decision that a Philadelphia anti-discrimination law could not require a Catholic social services agency to work with same-sex couples wishing to become foster parents. 

In that case — Fulton v. Philadelphia — the Court granted certiorari, at least in large part, to decide whether to overrule Employment Division v. Smith. The Catholic agency prevailed, but the Smith decision survived, with only three Justices ready to use Fulton as an opportunity to overrule the controversial 1990 decision.

This article analyzes one of the COVID cases in which the free exercise claim succeeded, and argues that the Court’s opinion was strained because of a disingenuous adherence to precedent. It is argued that the Court’s opinion would be more persuasive had the Justices done what they really wanted: overturn Employment Division v. Smith. It is further argued that Smith was wrongly decided and should have been overturned long ago, but that doing so now is especially pertinent. Instead of explicitly rejecting Smith, the Court’s latest free exercise decisions tiptoed around it, leaving lower courts and state legislatures puzzled as to the scope of the Free Exercise Clause’s protection for religious practice.

COVID-19 Restrictions on Houses of Worship: Roman Catholic Diocese

In November 2020, the Court granted injunctive relief in Roman Catholic Diocese of Brooklyn v. Cuomo. In separate cases, the Roman Catholic Diocese of Brooklyn and Agudath Israel of America, a national Orthodox Jewish organization, sued then New York Governor, Andrew Cuomo, claiming an executive order he signed in October 2020 infringed on their constitutional right to freely practice religion. The order placed caps on the number of occupants allowed to attend services in houses of worship in locations with significant upticks in new coronavirus cases. Per the restrictions, attendance at houses of worship in designated “red zones” was capped at 10 persons and at 25 in “orange zones.” The law did not distinguish between buildings of different sizes.

Photo by Gabriella Clare Marino on Unsplash.

In a 5-4 per curiam opinion, the Court explained that the order treated houses of worship in red zones more harshly than other indoor secular venues deemed by the state to be “essential.” Those businesses deemed essential included acupuncture facilities and liquor stores. In orange zones, even “non-essential” businesses were allowed to admit as many people as they wished. While acknowledging New York’s strong interest in stemming the spread of COVID-19, the Court determined that the order was both unnecessary and overly harsh: unnecessary, because there was no evidence that the churches or synagogues contributed to spreading the virus; and overly harsh because the law could have been more flexible while still achieving its objective. For example, one of the Dioceses’ churches can seat over 1,000 people, and yet the law did not allow more than 10 people at a time — the same limit on a church that seats just 50 people.

In her dissent, Justice Sotomayor rebuked the Court for treating dissimilar things as equal. Health experts have identified large indoor gatherings where people speak or sing for extended periods as especially dangerous. None of this applies to liquor stores. In fact, Sotomayor argued, the law actually favored houses of worship over categorically similar venues such as sporting arenas, concert halls, and movie theaters, all of which needed to remain completely shuttered.

Did the Court Adhere to Precedent?

Some commentators have criticized the Supreme Court for flouting controlling precedent in its Roman Catholic Diocese decision. That precedent is the Court’s 1990 landmark decision in Employment Division, Dept. of Human Resources of Oregon v. Smith. The criticism is not without merit.

The Court does not seem to genuinely follow precedent, but then again, neither did Smith

Prior to 1990 — and prior to Smith — the Court’s free exercise jurisprudence was guided by the so-called “Sherbert test.” Under Sherbert v. Verner and its progeny, the Free Exercise Clause “require[es] the government to justify any substantial burden on religiously motivated conduct by a compelling state interest and by means narrowly tailored to achieve that interest” (Smith at 894). This rule created a two-step process for courts to follow when considering free exercise claims. First, parties had to show that a law substantially interfered with their religious practice. If they were successful, the government would then need to justify the law by proving the law furthered a “compelling state interest” and was “narrowly tailored” to best accommodate the religious practice (a legal test referred to as “strict scrutiny”). That all changed after the landmark ruling in Employment Division v. Smith

As Justice O’Connor repeatedly argued in her free exercise opinions, “Smith was wrongly decided” and was “supported neither by precedent nor…by history.” She is not alone.

In Smith, Justice Scalia, writing for the majority, declared: “The right of free exercise does not relieve an individual of the obligation to comply with a valid and neutral law of general applicability on the ground that the law proscribes (or prescribes) conduct that his religion prescribes (or proscribes)”. The government no longer needed to justify a law that is “neutral” and “generally applicable,” no matter how severely the law might inhibit religious practice. The Court insisted it was not diverting from precedent, claiming that “[w]e have never held that an individual’s religious beliefs excuse him from compliance with an otherwise valid law prohibiting conduct that the State is free to regulate.” But, as Justice O’Connor repeatedly argued in her free exercise opinions, “Smith was wrongly decided” and was “supported neither by precedent nor…by history.” She is not alone. As Justice Gorsuch noted in his concurrence in Fulton “[n]o fewer than ten Justices — including six sitting Justices — have questioned [Smith’s] fidelity to the Constitution.”

Indeed, some of the Smith majority opinion’s feigned adherence to precedent borders on the absurd. For example, Justice Scalia claimed that Wisconsin v. Yoder — a 1972 landmark decision in which the Court clearly held that a law which was both generally applicable and neutral violated the Free Exercise Clause as applied to members of the Old Order Amish — was limited to “hybrid right cases.” Yoder involved a successful challenge to Wisconsin’s compulsory high school-attendance law, brought by Amish Wisconsinites. They argued that sending their children to conventional secular high schools would greatly interfere with the fabric of their religious life. The Court agreed that “formal high school education beyond the eighth grade is contrary to Amish beliefs”. In its reasoning, the Court stated: “There are areas of conduct protected by the Free Exercise Clause of the First Amendment and thus beyond the power of the State to control, even under regulations of general applicability” (emphasis added). Unsurprisingly, the Court cited Sherbert v. Vernon to support that notion. According to Justice Scalia, however, the appellants in Yoder won only because they challenged the law as conflicting with two constitutional liberties — religious freedom and the freedom of parents to “establish a home and bring up children”, a right announced by the Court in 1923. Per that bizarre reading, the Sherbert standard was used in Yoder because it presented a “hybrid situation”. But that’s nonsense. Yoder never used the term “hybrid,” and the above-cited language makes it clear that the Free Exercise Clause protects religious liberties even when the law is generally applicable and neutral. The “hybrid right” doctrine was, in the words of one federal circuit court, “illogical.” Why would two unsuccessful constitutional arguments make for a successful one? Consistent with that sentiment, many courts have not taken the “hybrid rights” concept seriously.1The federal courts of appeals have split on how to interpret the “hybrid rights” doctrine. For a summary of the various approaches see Note, The Best of a Bad Lot: Compromise & Hybrid Religious Exemptions, 123 HARV. L. REVIEW 1494, 1498 (2010) (“ The courts have developed a variety of approaches in the two decades since Smith, each of them a somewhat different attempt to reconcile the irreconcilable by squaring Smith with the cases that came before it.”); Ryan S. Rummage, Comment, 64 EMORY L.J. 1175, 1184 (2015) (“Three interpretations of the hybrid rights doctrine have emerged- … and United States Circuit Courts of Appeals are split on which of these approaches should rule.”)

Besides the denunciation of Smith within the legal community, the decision raised great concern among politicians and the public. Congress passed the Religious Freedom Restoration Act of 1993 (RFRA) in direct response to the Smith decision. The bill passed the House in an anonymous vote (almost unimaginable in today’s political climate) and passed the Senate in an astonishing 97-3 vote. RFRA basically reinstated the Sherbert rule for deciding religious freedom cases. However, in City of Boerne v. Flores, the Supreme Court struck down RFRA as applied to the states. Thus, for religious claims challenging most state laws, Smith remains the controlling standard of review under the federal Constitution.

Since Smith, establishing that a law violates the Free Exercise Clause is nearly impossible, as a law that is “neutral” and “generally applicable” is virtually unchallengeable. Until a few years ago, the sole case in which a religious party managed to show that a law was not neutral in its treatment of religion was Church of the Lukumi Babalu Aye v. City of Hialeah. In an unanimous decision, the Court held that a city ordinance which prohibited ritual animal sacrifice was impermissibly hostile toward (the Santeria) religion. However, as Justice Blackman noted in a concurring opinion, that case “was an easy one.” The subject of that case was an ordinance that was neither neutral nor generally applicable. Rather, it was clear that the city through “careful drafting ensured that, although Santeria sacrifice is prohibited, killings that are no more necessary or humane in almost all other circumstances are unpunished.” Lukumi made clear that even after Smith, laws which are not neutral and generally applicable must meet the heightened standard articulated in Sherbert. Because the ordinance targeted the Santeria religion, and because the city failed to show both a compelling state interest and that the laws were narrowly tailored, the ordinance was struck down. 

Ostensibly, the decision in Roman Catholic Diocese was justified by its reliance on Lukumi. The Court wrote that Cuomo’s “regulations cannot be viewed as neutral because they single out houses of worship for especially harsh treatment.” However, the Court’s exacting treatment of the New York regulations does not follow easily from Lukumi. Whether one agrees with the majority or the dissent regarding the necessity of the harsh restriction on houses of worship, it is hard to argue that the laws were really not “neutral” and “generally applicable.” The majority’s arguments are unconvincing if Smith is still truly the law of the land. 

Smith holds that even unnecessary burdens on religious practice are permissible so long as the law does not target religion.

First, the Court’s argument that the restrictions should have been correlated to the size of the buildings (which normatively is a very compelling argument) has nothing to do with the neutrality/general applicability inquiry. The fact that the state’s interest in public health (which is undeniably “compelling”) could have been achieved through less restrictive means is only relevant if the laws were not neutral. That’s because Smith holds that even unnecessary burdens on religious practice are permissible so long as the law does not target religion. Second, the Court’s comparison between the restrictions on houses of worship and other venues, which at least addresses the general-applicability inquiry, is far afield from Lukumi. There, an accounting of the city’s ordinances on animal cruelty clearly demonstrated discriminatory treatment of the Santeria’s animal sacrifice, in contrast to other animal slaughter. In Roman Catholic Dioceses, reasonable minds can disagree on whether New York treated houses of worship more harshly than secular venues (majority and concurring opinions), or if New York treated them more favorably (the dissent). However, the majority’s conclusion that “the challenged restrictions are not ‘neutral’ and of ‘general applicability’” is a bold extension of the Lukumi rule. Even the majority should admit that it’s conceivable that New York’s treatment of houses of worship was based on a scientific belief that more restrictive regulations were necessary as compared to stores (even non-essential stores). Considering that Lukumi is the only precedent the Court cites for its decision, its legal reasoning stands on weak footing.  

Not so under the pre-Smith test articulated in Sherbert. Under the Sherbert framework, the opinion would have made far more sense. Per Sherbert, even laws of general applicability must satisfy strict scrutiny. Therefore, because the claim that the Governor’s regulations placed a substantial burden on religious practice was uncontroverted, the Court would have utilized strict scrutiny-review. The state would have needed to show that it had a “compelling interest” and used means “narrowly tailored” to achieve its goal. The majority’s arguments would have been: (1) the state could have accommodated religion while still protecting public health, by tethering the extent of the restrictions to the size of the building; (2) the state’s willingness to exempt other venues (so called “essential businesses”) from harsh restrictions also indicates that the regulations on houses of worship were unnecessary; and (3) there was insufficient evidence that there was a compelling interest in the first instance, because “there [was] no evidence that the applicants have contributed to the spread of COVID–19.”

However, Sherbert is allegedly not the law — Smith is. Therefore, the Court needed to be creative in finding that the New York regulations were not “neutral,” in order to justify using a strict standard of review. This approach is circuitous and disingenuous; overturning Smith outright would be better for everyone.

Professor Laycock’s Reading of Smith

Leading free exercise scholar Professor Douglas Laycock has long advanced a much more measured interpretation of Smith, per which the Roman Catholic Diocese decision would be far more sound. Professor Laycock argues that Smith never meant that every law free from blatant religious discrimination must receive deferential treatment as a law of “generally applicability.” Rather, whenever a law features secular-exemptions that undermine the object of the law, the government must explain why similar exceptions cannot be made for religious groups. However, even Professor Laycock acknowledges that the “rhetoric” of Smith implies a “general rule of no [religious] exemptions.” By and large Smith has not been understood to require exemptions for religious reasons whenever similar exemptions are granted for secular causes.  

Interestingly, as Professor Zalman Rothschild noted in different words, Justice Kavanaugh seemed to embrace Laycock’s version of Smith in his solo concurrence in Roman Catholic Diocese without so much as mentioning that such a reading of Smith was, at the very least, controversial. Kavanaugh wrote that whenever “a State creates a favored class” it “must justify why houses of worship are excluded from that favored class.” Moreover, in April 2021, a majority of the Court cited its Roman Catholic Diocese decision for the proposition that “government regulations are not neutral and generally applicable … whenever they treat any comparable secular activity more favorably than religious exercise.” Most remarkably, the majority in Fulton, which included the three liberal Justices on the Court, suggested the same (in what appears to be dicta). In explaining why Philadelphia may not insist the Catholic agency service same-sex couples, the Court found it significant that the city allowed similar exemptions to the anti-discrimination law for secular reasons. The Court wrote, also without any fanfare, that a law “lacks general applicability if it prohibits religious conduct while permitting secular conduct that undermines the government’s asserted interests in a similar way.” (For an excellent explanation of the difference between the type of “general applicability” in this part of Fulton and the “individualized exception” rule also discussed in Fulton, see Professor Patrick Hornbeck’s September 16, 2021, article on Canopy.) 

In the next section I argue that the Court should overrule Smith, but a less radical approach may be for the Court to expressly adopt Justice Kavanaugh’s formulation of Smith. Indeed, Professor Laycock has advocated for his reading of Smith, as a means for the Court to restore religious protections to the pre-Smith era, without actually overruling it.

Why This Court Should Overturn Smith

To be clear, when I say that overruling Smith will be better for everyone, I do not mean to suggest that those who oppose broader protections for religious freedom out of concern for countervailing interests (for example, an interest in giving states broad discretion to enact public health measures) should advocate for overturning Smith. Of course not. Even people like Professor Leah Litman who believe that “the Court is no longer really applying Smith” would rather keep Smith on the books at least formally, so that whenever the makeup of the Court shifts, religious freedom jurisprudence might return to the (real) Smith framework.  

Instead, what I argue is that the Justices should overturn Smith for three reasons. First, Smith was wrong the day it was decided. Second, since it seems that the Court (at least a majority of the Court) is not really adhering to Smith, it should be honest and announce what it is really doing: overturning Smith in favor of the Sherbert-test. This honesty is especially necessary now, when the integrity of the Court as an independent institution has increasingly been called into question. Third, the Court’s approach in Roman Catholic Diocese will only leave lower courts and state legislatures confused. 

An explanation of these three reasons follows.

Smith Was Wrong and Stare Decisis Is No Obstacle To Overturning It

As many Justices, legal scholars, practitioners, and politicians have argued, Smith was wrongly decided. Smith disregarded the history of the Free Exercise Clause. Per Smith, the protection the Framers envisioned when they drafted the First Amendment’s Free Exercise Clause was merely that the government may not enact laws which openly discriminate against religion. But that notion is hard to believe in light of the fact that the Framers placed protecting religious liberty at the top of the Bill of Rights. Plus, decades of Supreme Court jurisprudence prior to 1990 interpreted the Free Exercise Clause far more broadly. As explored briefly above, the Court in Smith simply ignored precedent.

Stare Decisis is not an obstacle in the way of giving Smith the death knell. As Justice Souter argued in his concurring opinion in Lukumi, “Smith may be examined consistently with principles of stare decisis.” One petitioner succinctly explained why in a petition for writ of certiorari filed at the Supreme Court (the petition was denied in June): “The factors this Court has recently identified as central to stare decisis all point toward overruling Smith here: Smith was poorly reasoned, inconsistent with other areas of First Amendment law, has been undermined by later legal developments, and has generated no reliance worth this Court’s protection.”

Disingenuous Adherence To Smith Will Harm the Integrity of the Supreme Court

While Professor Litman’s characterization of the Court’s most recent free exercise decisions as “no longer really applying Smith” is a bit of an overstatement, she certainly has a point. At the very least, it does seem the Court is doing its best to expand the protection the Free Exercise Clause grants beyond the narrow scope that has prevailed at the courts since Smith. Even if the Supreme Court is not openly flouting Smith, it is certainly interpreting Smith more broadly than in the previous three decades. 

Admonitions of the Court for ignoring Smith highlight another reason this Court should be inclined to overrule it. The integrity of the Court as an independent institution, isolated from political and public influence, has been repeatedly attacked in recent years. Many scholars and journalists have posited that Chief Justice John Roberts keeps public perception of the Court’s independence in the forefront of his mind (and decisions). 

Overruling Smith would undoubtedly create huge backlash from those who fear that heightened protection for religious liberty will endanger other liberties, such as reproductive rights, LGBTQ+ rights, and separation of church and state-related rights. However, by continuing to decide high stakes cases with a feigned adherence to precedent, the Court’s integrity will only suffer. The case for rejecting Smith is strong, and even more liberal-leaning Justices and scholars should be able to admit that. By showing its true colors and overturning Smith outright, the Court will be more honest with the American people. That’s good in its own right (honesty is a nice virtue), and it will be beneficial for the Court as an institution.

The Roman Catholic Diocese-Approach Will Confuse Courts and States’ Legislatures

Supreme Court commentators Dahlia Lithwick and Mark Joseph Stern wrote a damning opinion piece in response to the Supreme Court’s February 26, 2021 order that also struck down COVID-19 restrictions on free exercise grounds. Lithwick and Stern cautioned: “[p]ublic health officials and lower court judges alike are inevitably chilled by the prospect of future rulings casting aside their best efforts to balance pandemic health regulations against constitutional freedoms. Their impulse will be to stop trying because they will reasonably conclude that they no longer have a fair shot at persuading SCOTUS [to uphold their regulations and laws]”. While the prospect of states giving up on enacting public health restrictions during a pandemic due to a handful of cases at the Supreme Court is imaginary, their astute observation is important. 

The Sherbert test, while not always easy to apply, was at least clear and coherent. The test tells state legislatures to make sure their laws do not substantially burden religious practice, and if they do, to be sure they have a compelling explanation for it. Namely, a law which substantially burdens religion must be narrowly tailored to further the state’s compelling interest while still trying to accommodate religious practice. The Smith test is even clearer. It tells states to be sure their laws are generally applicable and neutral. So long as they are, the laws will withstand free exercise challenges. 

However, the Court’s rationale in Roman Catholic Diocese is very unclear. The result will be confusion at the lower courts as to what test they are to apply in deciding free exercise cases. The opinion also obfuscates guidelines for state legislatures and governors to adhere to when enacting law. Everyone will be better off with this Court saying what it means and doing what it really wants to.

The government would still be free to place restrictions on religious liberties so long as the governmental interest is important enough and the law is appropriately tailored to accommodate religion.

The Fears of Losing Smith Are Unwarranted

In addition to the benefits of overruling Smith, it should be noted that the doomsday predictions sometimes paraded about by proponents of keeping the decision, are grossly exaggerated. Imagery of scores of religious exemptions and a country constricted by the beliefs of the faithful few is pretty unrealistic. Even under the Sherbert framework, religious parties lost at the Court about half of the time. Moreover, returning to Sherbert would not mean that religious people will be given carte blanche to make their own rules. The government would still be free to place restrictions on religious liberties so long as the governmental interest is important enough and the law is appropriately tailored to accommodate religion. Professor Hornbeck cautioned in a piece on Canopy that returning to a Sherbert-like framework would result “in courts granting more exceptions to public health regulations.” But that’s only because, as Hornbeck acknowledged, “[m]any religious litigants would [be] able to make quite persuasive rejoinders” to states’ arguments that immutable health regulations were necessary. Arguments like the ones made in Roman Catholic Diocese for tethering capacity restrictions at houses of worship to the size of the building, for example. Well, if the arguments that may succeed under a Sherbert framework are “quite persuasive,” one may wonder why the prospect is so terrifying.

Fulton v. Philadelphia

Fulton involved the city of Philadelphia’s refusal to renew its contract with Catholic Social Services (CSS), a privately owned agency, unless the agency began certifying married same-sex couples as suitable foster parents. The agency said that doing so would be an “endorsement” of same-sex marriage and would violate their religious beliefs. The agency and three foster parents sued the city under the Free Exercise Clause and asked the Court to “revisit” the Smith decision. All nine justices ruled for the agency but the majority did so without overruling Smith

Despite Smith’s permissiveness toward “neutral laws of general applicability” which burden religion, the decision contained an important carve out. Where a law contains “a mechanism for individualized exemptions,” it is per se not “generally applicable.” Therefore, in such cases the government must offer a compelling reason why it cannot grant a religious exemption. Because Philadelphia’s contract allowed a commissioner to grant exemptions from the anti-discrimination requirements in “his sole discretion,” the city had to offer a compelling reason for why it could not exempt CSS. The Court held the city failed to do so. Therefore, according to the Court the city must lose even under Smith, leaving “no occasion to revisit that decision.”

In two concurring opinions, three of the Justices expressed their deep frustration with the majority for not revisiting Smith. Justice Alito penned a whopping 77-page opinion that explained why he would overrule Smith and bemoaned the fact that after “receiving more than 2,500 pages of briefing,” the Court avoided the question. In a separate concurrence, Justice Gorsuch went further, accusing the majority of taking a “circuitous path” to decide in favor of CSS in order to “steer the Court around the controversial subject matter” which Fulton presented.2Justice Gorsuch’s “judicial-realism” style interpretation of what motivated the Court in Fulton to avoid revisiting Smith is plausible, but considering Fulton in conjunction with another recent religion decision, one can’t help but conclude that Gorsuch’s assessment is correct. In what is now a famous case about a Catholic baker who refused to bake a cake for a same-sex wedding, the Court issued a strained and narrow decision when the case presented a perfect opportunity for the Court to decide whether public accommodation laws must exempt religious people from complying when doing so would violate their religious beliefs. Justice Gorsuch noted that the “Court granted certiorari in this case to resolve [Smith’s] fate” and “not a single Justice has lifted a pen to defend the decision.” He asked what the Court was “waiting for.”

(Justice Alito briefly mentioned Roman Catholic Diocese in his opinion. Amongst his myriad critiques of Smith, Alito argued that Smith did not create the kind of bright-line rule which Justice Scalia envisioned when he wrote the decision. Alito pointed to the disagreement between the majority and the dissent in Roman Catholic Diocese as to the fairest point of comparison for houses of worship as an example of the difficulty in determining whether a law is “generally applicable.” Justice Alito did not go so far as to accuse Roman Catholic Diocese of flouting Smith — unsurprising considering he joined that opinion — but he too recognized part of the problems with the decision, discussed above.)


The Court’s recent free exercise decisions do not rigorously apply the controlling precedent, Employment Division v. Smith. The Court’s displeasure with the Smith framework for adjudicating free exercise cases is well-founded. This article argues that rather than muddling the waters of free exercise jurisprudence, the Court should explicitly overturn Smith. Stare decisis does not pose an obstacle to doing so because Smith was wrongly decided. This bold action will actually enhance the integrity of the Court and will provide clear guidance to lower courts adjudicating these cases and to state legislatures trying to draft law which is constitutional. Fulton may not have been the right occasion to overrule Smith but considering this Court’s interest in religion cases, the next opportunity to revisit Smith is probably not long coming. When that opportunity arises, the Court should finally overrule Employment Division v. Smith.♦

Raphael A. Friedman is an associate at a New York City litigation boutique and a graduate of New York University School of Law who has written on the Free Exercise Clause, juvenile sentencing, and the intersection of tort and administrative law.

Recommended Citation

Friedman, Raphael A. “Why This Supreme Court Should Overrule Employment Division v. Smith.” Canopy Forum, January 24, 2022. https://canopyforum.org/2022/01/24/why-this-supreme-court-should-overrule-employment-division-v-smith/.