Responsible Religious Freedom: Factual Scrutiny in Free Exercise Doctrine

Abstract

The state of the Free Exercise Clause in U.S. constitutional law is uncertain. With an opportunity in Fulton v. Philadelphia to clarify the vitality of the current standard from Employment Division v. Smith, the United States Supreme Court has declined to do so. The lasting impact of Smith has been to move away from directly requiring government justifications for infringing free exercise. Instead, courts now use neutrality and general applicability as heuristics for government justification. Yet, relying solely on neutrality and general applicability to proxy for government justification when infringing religious exercise distracts courts from conducting a fact-based inquiry. This article demonstrates how more scrutiny of the legislative facts in free exercise doctrine may serve as a viable alternative to Smith’s flawed approach for evaluating government justifications. The author first shows empirically how more factual scrutiny—directly requiring the government to justify its actions with evidence—can benefit government and religious claimants and then discusses the normative advantages of a fact-intensive approach to constitutional scrutiny. During a moment of sharp division over religious freedom and other competing rights, factual scrutiny can be a powerful tool for handling free exercise challenges and promoting responsible religious freedom.

Introduction

What is the law protecting the free exercise of religion? This question was recently raised in what many anticipated as a blockbuster case on religious freedom in the United States: Fulton v. Philadelphia. ¹ However, rather than rewriting or reaffirming the current standard for the Free Exercise Clause under Employment Division v. Smith, ² the court declined to provide clarification. Consequently, many observers were left with the question Justice Barrett rhetorically asked in her concurring opinion: “[W]hat should replace Smith?” ³

Surveying the current state of the law and the goals of religious freedom in the United States adds to the growing body of scholarly work addressing Justice Barrett’s question. ⁴ The most familiar understanding of free exercise jurisprudence since the early 1990s has been a bifurcated approach where courts defer to lawmakers absent evidence of religious discrimination. ⁵ This approach has been criticized for, among other things, its failure to adequately address the core of free exercise concerns: government justifications. ⁶ When state action infringes on constitutional rights, courts play a vital role in ensuring government provides evidence for its burden on religious freedom. But under the approach in Smith, only certain types of government action require justification. In Smith, the court explained that laws that were lacking neutrality or general applicability require heightened government justification for infringing on religious exercise.

The approach that the court took in Smith creates a troubling diversion. In contrast to strict scrutiny, which directly requires government justifications for infringing on religious exercise, Smith redirects the free exercise inquiry to the threshold questions of neutrality and general applicability. This is problematic because neutrality and general applicability are unreliable heuristics for evaluating government justifications; indeed, they often ignore the facts of a case with assumptions about the law. ⁷

More recent court decisions have produced particularly outsized attention toward the question of general applicability. Employing an approach known as the most-favored-nation theory, courts increasingly require heightened government justification when a law contains secular exemptions without comparable religious exemptions. As a result, many more free exercise claims have the potential to reach strict scrutiny. Some fear that this will lead to an unyielding flow of religious exemptions or excessive religious accommodation. But evidence suggests otherwise. Drawing on 750 state free exercise cases decided over a period of fifty years, the empirical evidence I offer below demonstrates more government scrutiny reduces disproportionate outcomes for marginalized religious groups while preventing unfettered religious exemptions from impeding effective governance. A qualitative analysis of four case studies in the federal courts further supports this finding. These findings suggest more scrutiny of state action that infringes on free exercise has the potential to improve outcomes while avoiding pitfalls.

As a whole, the analysis I offer demonstrates that the doctrinal free exercise test must evolve by focusing less on neutral or generally applicable laws and more on the factual disputes of each case. While neutrality or general applicability may sometimes function as an accurate proxy for justification, they cannot adequately replace the evidentiary burden on government required under the tailoring prong of strict scrutiny. Unfortunately, the current discourse too often leaves out serious discussion of government justification due to a fixation on the threshold questions under Smith. By contrast, an emphasis on factual scrutiny—that is, directly requiring government to justify its burden on religion—would place greater attention on the factual questions that underly a strict scrutiny tailoring inquiry.

To be clear, factual scrutiny is nothing new; indeed, it is the motivating feature of tailoring in U.S. constitutional scrutiny and the necessity prong of proportionality in foreign courts. ⁸ Therefore, adopting constitutional scrutiny entails greater factual scrutiny if courts require evidence from the state that its actions are necessary to achieve its stated goals. ⁹ As courts put greater focus on this underlying factual inquiry in constitutional law, state actors will be incentivized to empirically strengthen their justifications when brought into court by a religious claimant. The benefit of this judicially driven incentive on the state actors or government officials is a subsequent paradigm shift from individual rights to government responsibilities in religious freedom. Such a shift in religious freedom discourse is particularly desirable when such conflicts are too often conceived as clashes of rights without any discussion of government responsibility.

Government Justifications in Free Exercise Doctrine

Following incorporation of the First Amendment to the states, ¹⁰ the United States Supreme Court’s subsequent free exercise cases principally focused on a means/ends rationality. First, a “compelling state interest” is required to justify “any incidental burden on the free exercise of [individual’s] religion.” ¹¹ Even if the state demonstrates an interest of the “highest order,” it must also choose the “least restrictive means” for achieving its interest. ¹² This constitutional test—known as heightened scrutiny or strict scrutiny—is common across many areas of U.S. constitutional law. ¹³

The core of constitutional scrutiny is about government justifications. When state action infringes on a fundamental right like religious freedom, the government must give reasons to show why taking such action was justified. Under U.S. constitutional scrutiny, this inquiry into government justifications occurs as a tailoring analysis between state goals and the means chosen to accomplish them. Internationally, the same type of analysis occurs under the necessity and suitability prongs of proportionality. ¹⁴ While both U.S. and foreign approaches to judicial review of religious freedom rights also include other elements, the primary focus tends to center on the tailoring or necessity stage of the inquiry. ¹⁵ In U.S. free exercise cases, the tailoring requirement took the following form: when state action infringes on religious exercise, government must show that such action was narrowly tailored to achieve certain public aims that would be threatened if religion were accommodated.

This inquiry was the default doctrinal test for free exercise in the United States until 1990. Leading up to its decision in Employment Division v. Smith in 1990, ¹⁶ the Supreme Court’s focus switched from directly weighing the strength of government justifications to using a shortcut for evaluating the state’s reasons for infringing on religious exercise. ¹⁷ Under Smith, the Court declared neutrality and general applicability as the gatekeepers for conducting additional scrutiny of the government’s actions. ¹⁸ Consequently, all the facts of the case were now siloed into arguing about the form of the law rather than its factual application to the circumstances. Additionally, the burden fell on religious claimants to make this showing before government was required to justify its actions. Only in cases where religious discrimination could be demonstrated was government required to provide justification.

While neutrality and general applicability are nice heuristics, they are chronically overinclusive and underinclusive. Recent disputes over vaccine mandates and religious exemptions are illustrative. In many cases, COVID-19 vaccine mandates could be labeled as a generally applicable law and thus courts will grant deference to government judgment. ¹⁹ But the fact that a vaccine mandate is generally applicable tells us little about whether the government is justified in denying a religious exemption. Answering that question requires the government to bring evidence, not just rely on the face of the rule it has crafted. For example, where many religious claimants are requesting an exception in the same geographically concentrated area, government has good reason for denying an exception due to concerns about herd immunity. ²⁰ By contrast, if the religious claimants are geographically scattered in areas with high levels of vaccination rates, government reasons for denying the exception appear to be less compelling. ²¹ Either way, knowing that the law is generally applicable tells us little about the strength of government justifications. ²²

The fallibility of the neutral and general applicable law standard is also clear in the context of religious exemptions to drug laws. Like the drug law at issue in Smith, many drug laws fit comfortably within the definition of neutral and generally applicable laws. Yet using neutrality and general applicability is often ineffective at evidencing the strength of the state’s interest in drug regulation. For example, Native American tribes have frequently requested a religious exemption to use peyote in limited quantities solely during a weekly sacramental ceremony. ²³ In other cases, religious claimants have argued that their religion requires them to broadly use, possess, distribute, and sell marijuana or other drugs without constraint. ²⁴ Comparing these two types of claims for religious exemptions to drug use demonstrates how the underlying facts impact the strength of the government’s argument for regulation. Government clearly has a better case in denying a broad religious exemption for marijuana use compared to a much narrower religious exemption for peyote use. ²⁵ However, because both religious groups are claiming an exemption from neutral drug laws of general applicability, deciding these disputes under Smith will miss the factual context necessary to evaluate the strength of governmental interests.

To the detriment of religious claimants, constitutional free exercise cases ²⁶ since 1993 have essentially followed a bifurcated approach between Smith’s religious discrimination threshold heuristics for strict scrutiny and strict scrutiny proper. ²⁷ Importantly, even though Smith remains the default, recent Supreme Court decisions have expounded the meaning of religious discrimination from merely evidence of religious targeting to searching for evidence that religion has been disfavored in comparison to secular interests. ²⁸ As noted above, this has become known as the most-favored-nation theory of free exercise. ²⁹ The progression of this doctrinal development now requires heightened scrutiny in all cases where secular exemptions are written into law but a religious exemption is denied. ³⁰

The most-favored-nation theory has drawn significant scholarly attention as a dramatic judicial change in free exercise doctrine. ³¹ For some, the logic of the most-favored-nation theory could potentially subject every law to strict scrutiny and, therefore, it will either be an unsustainable approach or courts will fail to faithfully apply it. ³² Yet, a close look at the operation of the most-favored-nation theory reveals that the core failure lies in using neutrality and general applicability to serve as reliable indicators for government justifications. Simply expanding the scope of cases that qualify as burdening religion through a neutral law of general applicability is only helpful if careful consideration is given to government’s reasons for choosing to burden religion in a certain way.

Several recent cases demonstrate how the most-favored-nation approach tries to use general applicability to require greater government justification. The two cases most commentators point to as exemplifying the most-favored-nation theory are Tandon v. Newsom and Fulton v. Philadelphia. In Tandon, the court granted injunctive relief against a California law banning in-home religious gatherings of three or more households due to COVID-19. ³³ The majority opinion explained that strict scrutiny is triggered when the government treats “any comparable secular activity more favorably than religious exercise.” ³⁴ And the dissent echoes this sentiment: “The First Amendment requires that a State treat religious conduct as well as the State treats comparable secular conduct.” ³⁵ The actual disagreement between the opinions ran to the factual question of comparability between religious and secular restrictions; the fact that the threshold question of general applicability moved the case to strict scrutiny was unremarkable.

Similarly, Fulton also suggests that general applicability under most-favored-nation theory is doing little work in these cases. In Fulton, the court decided whether the Free Exercise Clause required Philadelphia to grant an exemption to Catholic Social Services to its nondiscrimination policy in foster care contracts. A unanimous court in Fulton agreed that Philadelphia’s contract with Catholic Social Services allowed for “a formal system of entirely discretionary exemptions” that rendered its nondiscrimination requirement not generally applicable. ³⁶Applying strict scrutiny, the court found that Philadelphia had failed to justify its asserted interests. Critically, the existence of contractual exceptions was not why Philadelphia lost the case; rather, it was the absence of justification for denying Catholic Social Services an exception where other exceptions were available. ³⁷

As demonstrated in these cases, the most-favored-nation interpretation is unremarkable and leaves unresolved the core free exercise question of government justifications. The practical effect of the most-favored-nation theory perpetuates general applicability as a heuristic for determining whether a law is narrowly tailored to a compelling interest. Both Tandon and Fulton support the idea that the threshold questions of neutrality and general applicability are tools for evaluating government justifications. Consequently, moving toward more strict scrutiny under a most-favored-nation theory still leaves the core problem in place because determining which exemptions are comparable draws attention to high-level theoretical comparisons and away from the facts of the case. ³⁸ As demonstrated by the court’s analysis in Tandon and Fulton, the core disputes are factual in nature. Indeed, this is one way to reconcile the result in Tandon (5-4) with the result in Fulton (9–0) even though both decisions composed the same justices operating under the same strict scrutiny standard. The key difference was the factual strength of the government’s reasons for burdening religious exercise. ³⁹

In sum, shortcutting to government justifications through Smith’s heuristics, courts fail to rigorously engage the facts and consequently legislatures have little incentive to produce them. Even with the most-favored-nation theory nudging more claims toward review under strict scrutiny, general applicability remains inadequate to proxy for the facts of each case. For this reason, a tilt toward more focus on facts in free exercise cases is needed.

Quantitative Evidence for Factual Scrutiny

Existing Literature

The urge to rely on heuristics under Smith is likely tied to anxieties about courts facing strict scrutiny in free exercise cases. As Justice Scalia explained in Smith’s majority opinion, the idea that courts would need to apply strict scrutiny any time a law burdened religious exercise seemed to “court anarchy.” ⁴⁰ Many have continued this line of thinking by overlooking strict scrutiny entirely when discussing free exercise doctrine. Scholars often discuss the threshold questions of neutrality and general applicability and assume that the government loses the case if they lose on the threshold questions. ⁴¹ While this assumption is widely held, it is not well grounded in the evidence on the application of strict scrutiny. Scrutiny of government action in free exercise does not “court anarchy.” In fact, more factual scrutiny remedies disparate treatment of religious claimants from marginalized groups without impeding effective governance. These conclusions are the product of both quantitative and qualitative evidence on outcomes in state and federal cases handling free exercise.

Testing free exercise theory through empirical methods is not new. Many scholars have tested for the effect of Smith using a variety of descriptive studies. Two of the earliest studies came from Amy Adamcyzk, John Wybraniec, and Roger Finke, ⁴² who relied on the Religious Freedom Reporter, which collected both state and federal data on religious freedom cases between January 1981 and May 2002. ⁴³ For the first study, published in 2001, Wybraniec and Finke examined more than four hundred state and federal free exercise cases from January 1981 to January 1997. ⁴⁴ Using a logistic regression model, they probed, inter alia, the relationship between a case’s citation to Smith and the ultimate outcome. ⁴⁵ They concluded that Smith was associated with a negative outcome for religious claimants even after controlling for the introduction of RFRA. ⁴⁶ In their 2004 study, Adamczyk, Wybraniec, and Finke drew upon the same dataset and concluded that Smith had an especially negative impact on religious minorities compared to that on mainline Protestant religious claims. ⁴⁷

More recently, scholars have utilized online legal research services and databases such as Westlaw and LoislawConnect to identify and code free exercise decisions. ⁴⁸ For example, in a 2017 study, Caleb Wolanek and Heidi Liu relied on 264 published federal cases between 1990 and 2015 from Westlaw. ⁴⁹ They focused on the success of free exercise claims when heightened scrutiny is applied, finding that religious minorities were much more likely to bring a free exercise claim and claims generally succeeded if they could demonstrate a substantial burden on religious exercise. ⁵⁰ In 2019, Robert Martin analyzed 453 published state appellate court opinions on free exercise from legal research database LoislawConnect. ⁵¹ Martin’s study uses a multivariate regression analysis to demonstrate the increase in favorable case outcomes for religious claimants when the deciding court applies heightened scrutiny to government action. ⁵²

The longest empirical study of religious liberty decisions covers three decades of lower federal court opinions. In a series of papers, Gregory Sisk and Michael Heise relied on Westlaw’s digested opinions to conduct three empirical studies covering 1986–1995, 1996–2005, and 2006–2015. ⁵³ The thrust of this research is to identify which variables—judicial ideology, religion of judge, religion of claimant, type of case—provide the most explanatory power of outcomes in religious liberty cases. Their research produces three key conclusions across all three decades. First, in earlier decades certain religious claimants were more likely to succeed on their claims in court, ⁵⁴ but this disparity has been reduced in more recent years.⁵⁵ Second, success of religious liberty claims in federal courts has stayed relatively stable over thirty years—“consistently falling within a two percentage point band.” ⁵⁶ Finally, Sisk and Heise suggest that each period of study demonstrates a robust connection between the type of free exercise case and the ultimate outcome. ⁵⁷ In other words, the case context does make a difference in these decisions.

Taken together, empirical studies on the free exercise doctrine have coalesced around some overarching themes. First, the religious affiliation of the claimant has traditionally impacted the odds of a favorable outcome ⁵⁸—supporting what other legal scholars have contended in theory ⁵⁹—though this trend may be changing. ⁶⁰ Additionally, the introduction of government deference in free exercise cases following Smith exacerbates these odds, generally favoring religious majority claims over those from marginalized faiths. ⁶¹ Similar empirical findings are also present in the context of federal courts applying RFRA. ⁶²

I build on previous empirical work studying free exercise jurisprudence in three ways. First, I expands the time horizon of cases studied, allowing for a broader evaluation of change in free exercise doctrine. Most case studies that have empirically analyzed free exercise doctrine in case law are limited to a timeframe of twenty years or less. By contrast, I look at fifty years (1972–2022) of case law. Previous studies have only implemented empirical analysis surrounding the time of the largest changes in free exercise doctrine such as 1990, when Smith was decided. ⁶³ While these changes are among the most significant, studies with limited time frames may fail to fully capture potential drifts away from the doctrine in practice. Moreover, even when studies do cover longer time horizons such as a recent review of Supreme Court decisions on religion cases since 1953, ⁶⁴ these tend to focus on a subset of the judiciary and are thus less likely to be representative of all cases. ⁶⁵

Second, I evaluate the usage of constitutional scrutiny in free exercise decisions. Since the doctrinal shift in Smith, the empirical work on free exercise has generally looked to find relationships between claimant or case characteristics and case outcomes. ⁶⁶ Even within the two studies that buck this trend and focus on the impact of doctrinal tests, this analysis provides a unique angle. Wolanek and Liu focus exclusively on strict scrutiny ⁶⁷ while Martin compares rational basis review to “balancing tests”—categorizing more than half of the tests used by courts as “other.” ⁶⁸ In contrast, I order the doctrinal test into two categories: government deference or government scrutiny. By doing so, I engage a core question of Smith and modern free exercise debates: Is heightened scrutiny of government interests a viable path for protecting free exercise? ⁶⁹

Finally, I contribute to existing empirical literature on free exercise by focusing on state court decisions. Besides Martin, all other relevant empirical studies have focused on federal court decisions in their analyses. ⁷⁰ But state courts also merit attention. State courts provide a heterogenous sample of free exercise cases involving deference and scrutiny of government interests. ⁷¹ As in many areas of law, state courts are serving as laboratories for protecting religious exercise. ⁷² Additionally, many of the most significant cases influencing free exercise doctrine originated in state courts including Smith, ⁷³ Wisconsin v. Yoder, ⁷⁴ Masterpiece Cakeshop v. Colorado Civil Rights Commission, ⁷⁵ and Espinoza v. Montana Department of Revenue. ⁷⁶ In sum, state courts are vital to a thorough understanding of free exercise doctrine.

Empirical Design and Analysis

To make best use of the important work that has already been done, I compiled a unique dataset of free exercise cases spanning fifty years of data. Beginning with the Supreme Court decision in Yoder and continuing through the end of 2022, this unique dataset provides every state court decision where a free exercise issue was adjudicated. ⁷⁷ In compiling this dataset, I excluded unreported opinions due to variance in availability across jurisdiction and their lack of precedential value. ⁷⁸ Each case was hand-coded for three different values. First, the religion of the claimant involved in the dispute. Of the 819 total free exercise cases, 182 did not specify a religion at all. In the remaining 637 cases, 194 were a general Christian category, 93 were Protestant, and 61 were Catholic. The other 289 cases varied across many religious groups that, for the purpose of this analysis, will be defined as minority religions. ⁷⁹

The second data point collected from the case was the type of test used to adjudicate. Coders were instructed to distinguish between cases that used some form of scrutiny that required government justification (such as Yoder’s scrutiny test) versus those that deferred to government and inquired into the interests of the claimant (such as Smith’s neutral and general applicability standard). When more than one test was used, the highest form of scrutiny was coded following the approach other scholars have taken. ⁸⁰ Of the 819 cases in the final dataset, 44 percent used some form of heightened scrutiny on the government.

The third category recorded whether the religious claimant was successful on any point of the free exercise claim. A summary of the collected data is provided in table 1.

Table 1. Summary Statistics

Because this dataset spans fifty years from 1972 to 2022, it allows for eighteen years of case data prior to Smith and thirty-two years of data after. Figure 1 presents descriptive statistics on comparing number of successful claims between majority and minority religious groups before and after Smith. While both groups see a decline in successful claims following Smith, the decline is disproportionately larger for minority religions (46 percent decrease) than majority religions (31 percent decrease) and is statistically significant. ⁸¹

Figure 1. Successful Claims Before and After Smith.

Notably, religious minorities generally see better case outcomes than do religious majorities. This is at least in part because their claims are seemingly more likely to obtain strict scrutiny review as suggested by figure 2. Similar to win rates, the incidence of heightened scrutiny decline is greater among minority religions after Smith (42 percent decline) than it is for majority faiths (40 percent decline). The decline in both groups is statistically significant. ⁸²

Figure 2. Government Scrutiny Before and After Smith.

Finally, in figure 3, results for cases between different levels of scrutiny are provided. While arguing under a deferential standard certainly makes success unlikely for a religious claimant (8 percent win rate), it is also true that arguing under heightened scrutiny is not a sure victory (36 percent win rate). As other scholars have demonstrated more generally, requiring government to justify alleged intrusion on constitutional rights will not result in endless exemptions and render laws unworkable. ⁸³

Figure 3. Claimant Win Rate Based on Level of Scrutiny.

These descriptive results suggest that following Smith’s turn to greater government deference in free exercise cases, religious claimants generally won fewer challenges in state court with minority believers facing steeper declines than claimants of majority faiths. While these results are consistent with the literature, they expand the scope to a broader time frame and suggest that effects on minorities are persistent. Moreover, these results also demonstrate how constitutional scrutiny proves to be a workable standard in free exercise jurisprudence. As Heise and Sisk suggest in their most recent empirical study on free exercise, recent Supreme Court decisions placing greater scrutiny on government acts burdening religion “may be bringing greater quality in the process” for religious claimants. ⁸⁴ Although religious claimants generally fare much better under judicial scrutiny of government justifications, it is also true that government has successfully produced compelling evidentiary justifications against about two-thirds of free exercise claims in state court.

Importantly, as with all empirical work, the benefit of providing precise results is that it also allows for precise critiques. For example, the way of defining and then coding for the use of heightened scrutiny versus government deference is somewhat indistinct. Although textual hooks were used for defining one over the other (such as rational basis, compelling interest) the endeavor to neatly categorize remained imperfect. The same is true for defining religious minorities and majorities. Moreover, a threshold decision was made to exclude cases that did not raise a free exercise issue. In making this judgement call, some cases on the margins could have gone either way. Although a sample size of nearly eight hundred cases allows for a margin of error, and external coders were used to check the original coder, bias in the results ultimately remains a possibility. ⁸⁵ Additionally, there remain both observable and unobservable underlying differences between cases decided before and after Smith. ⁸⁶ Therefore, the full effect of case outcome differences may be attributable to other factors besides (or in addition to) the doctrinal change in 1990. Even so, the results remain consistent with the hypothesis that the doctrinal test is contributing to the observed change in outcomes.

Finally, there are also questions about the external validity of this study when the sample focused on state rather than federal cases. This is somewhat addressed by the fact that many if not most state courts explicitly relied upon the Supreme Court’s free exercise approach both before and after Smith. ⁸⁷ Moreover, it is helpful to look at the state and lower court context where so often constitutional law discourse suffers from selection bias in a handful of Supreme Court cases. ⁸⁸ However, to the extent analogies between state and federal courts remain unconvincing, qualitative analysis of four federal cases provides evidence for factual scrutiny.

Qualitative Evidence for Factual Scrutiny

The holding of Smith justifies incidental violations of religious freedom when oppressed religious practices are overlooked in lawmaking so long as the laws are both “neutral” and “generally applicable.” ⁸⁹ As suggested empirically above, stepping toward Smith and away from requiring constitutional scrutiny has proved both inequitable and unnecessary. This qualitative section highlights modern examples of how Smith’s diversion from government justification impacts marginalized religious groups. The four case studies demonstrate how more factual scrutiny in these cases offered a plausible and even preferable alternative. I draw upon pairs of cases involving Native American, Muslim, Buddhist, Sikh, and Hmong beliefs—religious groups that are generally considered minority religions in the United States. These cases—the experience of groups on the religious margins of the United States—model how factual scrutiny can ease the inequities displayed in the results above. Moreover, these cases also demonstrate how factual scrutiny returns focus to government justifications in the doctrine.

Perhaps the most widespread marginalized religious community in the United States is that of Native Americans. Since the creation of the First Amendment, Native American religious traditions have struggled to find purchase among the majority of the U.S. populace. To those versed in the rites and practices of the Judeo-Christian heritage that undergirds U.S. civil religion, Native American religion can seem fluid and difficult to classify. Native American religion and culture may appear indistinguishable at times. This can lead some outside the belief system to question the sincerity or depth of these religious practices, allowing them to be written off as cultural or even cultish actions that fail to fall within the generally accepted definition of religion. Even at best, when these Native American beliefs are nominally categorized as religion, it is unlikely that legislatures will make efforts to accommodate those beliefs in the law. ⁹⁰ This makes Smith’s neutrality requirement especially challenging since a Native American claimant bears the burden to show state action both recognizes and discriminates against Native American beliefs.

Two cases involving high school graduation ceremonies illustrate how Smith has made protecting Native American religious freedom especially difficult. During the nineteenth century, many Native American children were removed to boarding schools in the United States. The schools were designed to assimilate Native children into Western society, and their practices included forcing the children “to disavow their religious beliefs.” ⁹¹ With this history in mind, in modern times, graduation from high school or college has taken on new significance for Native Americans as an opportunity to honor their ancestors and honor their religious beliefs. Native students will frequently include a blessed eagle feather—a sacred object in many Native traditions—in their graduation caps to mark the religious significance of this ceremony. ⁹²

Unfortunately, school policies often prohibit students from wearing decorations on their graduation caps, including eagle feathers. In Griffith v. Caney Valley Public Schools, Hayden Griffiths—a member of the Delaware Tribe—was denied the opportunity to wear an eagle feather blessed for her high school graduation to celebrate “her passage into adulthood.” ⁹³ By failing to wear the feather because of the school’s no-cap-decoration policy, Griffiths “disrespected the feather, the tribal elder who gifted it to her, and God.” ⁹⁴ Under review in federal district court in Oklahoma, the court relied on Smith to hold no free exercise violation had occurred. Because the no-cap-decoration policy applies to all types of decorations for reasons unrelated to religion, the court found the policy to be both neutral and generally applicable. ⁹⁵ Consequently, the school was not required to justify its decision that prevented Griffiths from honoring her religious tradition.

How would this conflict have turned out differently under more factual scrutiny of the school district? A case in the Ninth Circuit involved a similar dispute between the Dysart School District in Arizona and Larissa Waln, a member of the Sisseton Wahpeton Oyate tribe. ⁹⁶ When Waln requested to wear an eagle feather on her graduation cap, the school district explained that its no-decoration policy permitted no exceptions. However, while Waln was excluded from her high school graduation for wearing the eagle feather, another student in the district was allowed to participate with a cancer awareness sticker on their cap in a different graduation ceremony. In the Ninth Circuit, this asymmetry took on constitutional significance. Because Waln carried her burden to demonstrate an instance of exception, the policy was not generally applicable, and the burden shifted to the district. In response, the district offered a compelling interest in complying with the Establishment Clause that justified their policy in denying Waln an exception. ⁹⁷ The Ninth Circuit was unconvinced. Ruling in favor of Waln, the court held that accommodating Waln’s religious beliefs would not threaten and perhaps even advance the objective of the Establishment Clause to “learn[] how to tolerate diverse expressive activities.” ⁹⁸

Side by side, these two cases highlight how factual scrutiny improves law and policy for government while strengthening religious freedom protections. In Griffiths, the lack of evidentiary burden on the government meant that the school district did not need to really address Griffith’s religious freedom concerns. In fact, the structure of Smith encouraged the school district to ignore the claims of the Delaware Indian tribe. By considering religious accommodation, a court may consider this practice part of “a mechanism for individualized exemptions.” ⁹⁹ Similarly, the more that school administrators interacted with students or parents about the school policy, the more likely evidence could come forth demonstrating religious hostility. ¹⁰⁰ Thus, when the school district opens itself up to considering accommodation of religious beliefs in its policies, Smith opens them up to new opportunities for liability. To its credit, Caney Valley did try to provide other options; yet none of those options addressed Griffith’s religious concerns probably because there was no strong incentive to do so. ¹⁰¹

By contrast, the same year that Griffiths was litigating her case against Caney Valley, school officials in North Dakota had enacted a ban on “personal additions” to graduation attire. ¹⁰² However, after “very informative” conversations with Native American parents, the school reversed course and allowed eagle feathers on graduation caps. ¹⁰³ Likewise, in Waln’s case, additional scrutiny prompted the school to consider the reasoning more carefully behind its policy. Indeed, by requiring justification, the Ninth Circuit’s decision actually helped the school district better achieve its asserted interest in harmonizing its policy with the pluralistic spirit of the Establishment Clause. ¹⁰⁴ Yet without the court’s close review of the school’s policy, the resulting policy would have likely been less helpful to the school and the students.

Similar to these graduation cases, the impact of factual scrutiny is also clear in two cases about grooming policies in prison. In the first case, Muslims, Rastafarians, and Native Americans challenged a South Carolina prison grooming policy requiring all prisoners “to keep their hair short and faces shorn.” ¹⁰⁵ For Muslims, keeping a beard is tied to a hadith or saying of the prophet Mohammad that many see as a religious obligation. ¹⁰⁶ Similarly, Rastafarians and Native Americans each abstain from cutting their hair according to religious beliefs. In its decision, the Fourth Circuit found no violation of the prisoner’s free exercise rights. The court found the prison’s grooming policy fulfilled Smith’s neutrality and general applicability requirements since the policy was not “enacted to burden anyone’s free exercise rights.” ¹⁰⁷ Instead, the court found that the policy “was implemented to help eliminate contraband, reduce gang activity, identify inmates, and maintain order in South Carolina’s prisons.” ¹⁰⁸ However, whether the policy advanced these interests went unaddressed.

Contrary to the Fourth Circuit, the Eleventh Circuit took a more critical approach to prison grooming policies in Georgia. The case was brought by Muslim inmate Lester Smith after Georgia prison officials denied him the right to grow an untrimmed beard in accordance with his Muslim beliefs. ¹⁰⁹ According to Georgia prison policy, beards may be grown up to only half an inch. ¹¹⁰ Noting the substantial burden this policy placed on Smith’s beliefs, the Eleventh Circuit proceeded to strict scrutiny of Georgia’s policy under the Religious Land Use and Institutionalized Persons Act. ¹¹¹ The court’s evaluation demonstrated that the application of the grooming policy was the least restrictive means to accomplish the prison’s interests in safety and security. Based on witness testimony and facility officer evidence, untrimmed beards allowed inmates to hide contraband and evade detection by shaving when trying to escape. ¹¹² Moreover, Smith in particular demonstrated an extensive record of disciplinary prison action that made him more likely to pose a security threat. ¹¹³ For these reasons, the court upheld the prison policy based on the facts underlying the policy’s rationale.

Importantly, both prison beard cases ultimately reached the same outcome of denying relief to the religious claimants. However, the Eleventh Circuit case was much more useful because it required the prison officials to proffer evidence and carefully consider its own policy implications for religious freedom. The reason that Georgia was ready to bring evidence in Smith’s case was due to a prior decision from the U.S. Supreme Court that required factual scrutiny of an Arkansas prison beard policy. ¹¹⁴ Following that decision, Georgia adjusted its grooming policy from prohibiting all beards to allowing half-inch beards. ¹¹⁵ This adjustment better fit Georgia’s interests in safety and security by only imposing necessary restrictions on religious freedom. In the South Carolina case, the state lacked incentive to carefully consider the religious freedom implications of its beard policies in prison. Instead, the rule under Smith encouraged avoiding religious discrimination without any concern for how effectively the resulting rule would achieve prison security interests or inmate religious freedom.

Examples from other faith groups further illuminate this point in the context of zoning decisions. After finding the traffic flow and other property specifications of a Buddhist temple failed to meet town code, the Supreme Court of Connecticut dismissed an appeal from the Cambodian Buddhist Society to build their house of worship. ¹¹⁶ In doing so, the court’s opinion was based largely on facts and information about a different Buddhist temple in Massachusetts. ¹¹⁷ In part, the Massachusetts town decision drew inferences about the noise level and traffic congestion based on the number of participants in Buddhist festivals. ¹¹⁸ While Cambodian Buddhist festivals can raise volume and visitors at temples, these events are similar to what many Americans do to celebrate the New Year. ¹¹⁹ Further, the vast majority of festivals performed involve chanting, prayers, and rituals that do not reach outside the walls of the temple. ¹²⁰ In justifying its conclusion, the Connecticut Supreme Court explained the town’s rejection of the Buddhist temple was “motivated not by religious bigotry but by neutral considerations.” ¹²¹ Religious interests were more easily set aside when a government body could show it was acting in a neutral or generally applicable manner toward practices it found largely unfamiliar. ¹²² Any careful inquiry into the state’s interests in constructing certain inferences about noise level and traffic was absent from the case.

In contrast to the decision in the Cambodian Buddhist case, Sikhs in Yuba City, California, successfully argued a violation of their religious freedom under strict scrutiny. The case involved Sikhs who desired to build a temple on land zoned as “agricultural.” ¹²³ The city denied their application due to a concern of “leapfrog development” even though other churches had been approved on similar parcels. ¹²⁴ Furthermore, because noise and traffic concerns would likely prevent Sikhs from building their temple on land in more urban areas, the denial appeared to be unredeemable. ¹²⁵ Consequently, the Ninth Circuit held that placing Sikhs in such a jam as to preclude any possibility of constructing a house of worship required a compelling justification from the city. For its part, the city failed to submit a compelling interest sufficient to counterbalance the Sikhs’ claims and thus infringed on the Sikhs’ religious exercise. ¹²⁶ Importantly, factual scrutiny here provided an important shift away from the Sikh’s needing to show discrimination on the part of the state. Instead of asking the religious claimant to explain the purpose behind the zoning board’s actions like in the Cambodian Buddhist case, the court reasonably turned its attention to the government decisionmaker to ask for its rationale.

A final pair of examples motivated by Hmong beliefs about autopsies contrasts the deficiency of neutrality and general applicability of laws and the virtues of factual scrutiny. Hmong religious beliefs include views on the sacred nature of the body which prohibits removing organs or performing autopsies after death. ¹²⁷ Neng Yang, the son of Hmong immigrants from Laos, tragically died in Rhode Island days after suffering a seizure. Because the specific nature of his death was unclear, the medical examiner’s office for Rhode Island was called upon to perform an autopsy according to state law. The autopsy was then performed without notifying the parents and in direct violation of their religious beliefs. The district court ultimately concluded that even though the state law “did profoundly impair the Yang’s religious freedom,” the fact that it was both “facially neutral” and “generally applicable” prevented the law’s harm to religious freedom from rising “to a constitutional level.” ¹²⁸As in the cases already mentioned, this interpretation places less weight on religious interests under the Constitution when the laws that burden free exercise are neutral and generally applicable.

Contrary to the decision in Yang’s case, a factual scrutiny approach was used to decide whether an autopsy performed on an executed inmate was violative of his religious freedom rights. ¹²⁹ Under Tennessee state law, “no governmental agency could burden the exercise of religion without requisite justification.” ¹³⁰ In its ruling, the Tennessee Court of Appeals found the state had demonstrated a compelling interest in performing an autopsy following an execution. As the state had explained, the autopsy was needed to ensure that the execution did not violate the Eighth Amendments charge against cruel and unusual punishment. ¹³¹ However, the court of appeals also concluded that the state failed to demonstrate that the least restrictive means was used to achieve that interest. ¹³² Because the circumstances under which the execution occurred were innocuous and the facts showed the execution was “without incident”, the court concluded the state could have employed other means besides autopsy to determine the execution did not violate the Eighth Amendment. ¹³³ This case demonstrates that when the state is required to articulate its reasons behind law and policy, there is a better fit between the means and ends of government action in addition to greater protection of religious exercise. Moreover, although Mr. Johnson’s beliefs would not be vindicated in his own case or that of his surviving spouse, the state would be cognizant of its responsibility to the religious freedom of the deceased going forward. ¹³⁴

In explaining these contemporary examples, it is nothing new to introduce criticism of the Supreme Court’s neutral and generally applicable test under Smith. While many scholars have presented differing legal reasons that this case is so problematic, ¹³⁵ these examples present the lack of factual scrutiny as a deficiency of Smith. Without adequate attention to facts in the courts, state actors fail to carefully consider the religious freedom implications of their decisions. Instead, government actors have incentive to actively avoid accommodating religion even when they can do so at a low cost. Moreover, when courts fail to require government responsibility for burdening religious exercise, the focus shifts to religious claimants trying to justify their constitutional rights. As a result, religious freedom rights become more about excluding others from infringing on your constitutional space rather than obligations of state and other actors to pursue compelling objectives in a responsible way.

The quantitative and qualitative findings above suggest there is a better way. As demonstrated through the foregoing evidence, factual scrutiny benefits free exercise doctrine in at least two empirical ways. First, more factual scrutiny reduces disparities between marginalized religious groups and religious majorities when compared to Smith. Second, factual scrutiny addresses this disparity while still ensuring effective governance is possible. Even when greater factual scrutiny is applied, these findings suggest that government often has strong reasons for limiting religious exercise under certain circumstances. Moreover, the number of cases where government lacks such reasons do not appear unmanageable or unwieldy.

While other qualitative examples could be cited and additional quantitative evidence is needed, the empirical findings above represent the opportunities for factual scrutiny to benefit free exercise jurisprudence and redirect the religious freedom narrative. Importantly, factual scrutiny does not have to work by giving the religious claimant a win in every case. Instead, by shifting toward direct evaluation of government justifications, courts no longer need to rely on neutral laws of general applicability as the sole indicators of a strong government interest. Building on these empirical findings presented above, the following section outlines the normative arguments for factual scrutiny while responding to potential concerns.

Normative Support and Potential Concerns for Factual Scrutiny

Factual scrutiny can bring normative advantages to applying strict scrutiny as a replacement to Smith. In contemplating the next doctrinal move in free exercise after Smith, a fact-intensive approach to constitutional scrutiny is normatively advantageous. It is normatively advantageous for three reasons: (1) factual scrutiny is consistent with history and precedent of the Free Exercise Clause; (2) factual scrutiny plays to the comparative institutional design advantages of courts and legislatures, and this is supported by the fact that courts regularly and ably evaluate evidence on government action under RFRA and RLUIPA; and (3) factual scrutiny encourages a narrative of responsibilities to counterbalance the heavy focus on competing rights in constitutional law discourse. Even accounting for potential concerns about factual scrutiny, the arguments in favor of factual scrutiny outweigh those in favor of Smith.

For courts and many observers searching for a new test to replace or modify Smith, strict scrutiny is perhaps the default option. ¹³⁶ Due to its use in other federal constitutional contexts in addition to many state-level free exercise cases, returning to strict scrutiny is a natural doctrinal step that the Supreme Court could make. However, even if the court returns to strict scrutiny in name, the way strict scrutiny is applied is perhaps the more critical question for free exercise doctrine. As expressed by Justin Collings and Stephanie Barclay, the name or label of the doctrinal framework itself is “less important … than whether the legal test requires the government to justify the burden it places on religious rights.” ¹³⁷ Therefore, the ultimate form of constitutional scrutiny that replaces Smith is not important. What is important is that applying strict scrutiny as a fact-intensive inquiry rather than a balancing test allows many of the benefits of strict scrutiny to be enjoyed with greater consistency and constraint on the judiciary.

Just as Smith distracted from government justifications for burdening religious exercise, it is easy for strict scrutiny to do the same without sufficient attention to facts. A common way this occurs is through viewing strict scrutiny as a balancing test where government interests are weighed against religious rights. At least two problems arise from this approach. First, courts are often ill-equipped to evaluate burdens on religious exercise. Even assuming courts could reliably distinguish substantial and incidental burdens on religious exercise, it is hard to imagine how this blunt evaluation could then be effectively compared to a governmental interest. Ultimately, courts will find themselves balancing incommensurate interests even assuming they can make a clean evaluation of both sides independently. ¹³⁸ In reality, courts are best positioned to apply law to the facts—evaluating evidence rather than balancing interests.

Understanding constitutional scrutiny as a fact-intensive inquiry into government justifications pulls courts away from balancing incommensurate interests. When viewing scrutiny analysis as an evidentiary burden on government, courts can simply judge whether government has demonstrated that the infringing on religious exercise is necessary to the aims of the state. ¹³⁹ The tradeoff, of course, is that much of the evidence required to show a state action is necessary will be empirically intensive. ¹⁴⁰ In fact, many questions will require counterfactual reasoning to determine how state action that infringes on religious exercise would compare to the same action that does not. But courts will not be the ones required to produce this evidence: that burden falls both to the state and the claimant seeking to counter the government’s claims. ¹⁴¹

Properly focused on directly evaluating government justifications through facts, scrutiny analysis is a normatively attractive alternative to Smith. First, factual scrutiny is supported in practice from the history and precedent of the Free Exercise Clause. Present within legal reasoning from Blackstone to the United States’ founding era is the idea that facts guide judges in interpreting the law rather than just pure text and logic. ¹⁴² In particular, factual inquiry was historically integral to conducting a scrutiny analysis “in a variety of natural rights contexts.” ¹⁴³ For example, the very first reported Supreme Court free exercise decision involved clear reference to legislative facts to justify denial of a religious exemption. ¹⁴⁴ In Permoli v. Municipality No. 1 of City of New Orleans, the Supreme Court determined that the city of New Orleans could forbid open-casket Catholic funerals during an episode of yellow fever in the city. ¹⁴⁵ In its reasoning, the court noted how facts demonstrating the “law of necessity” for enacting the public health ordinance “gives warrant” for disallowing open-casket funerals that would otherwise “impair [the ordinance’s] efficacy.” ¹⁴⁶ Moreover, this is supported by the Supreme Court’s more recent precedent under the Free Exercise Clause. Modern cases demonstrate how free exercise precedent supports a “precise analysis” ¹⁴⁷ informed by the facts that allow the court to “take relevant differences into account.” ¹⁴⁸

A second benefit from factual scrutiny in free exercise is the way it supports the institutional roles of courts and legislatures in free exercise conflicts. If courts require lawmakers to show that their law must be free from a religious accommodation to accomplish its purpose, lawmakers will be incentivized to search for that evidence. As Paul Yowell has explained, legislatures in the United States are well equipped to perform this function due to institutional resources such as legislative research services that courts often lack. ¹⁴⁹ Importantly, this approach also acts as a constraint on courts. Whereas Smith or balancing tests invite greater judicial discretion regarding “comparable” secular exemptions or the size of burden on a religious claimant, ¹⁵⁰ the question under factual scrutiny boils down to an evaluation of the necessity of fit between the law and its goals. Courts conducting such an analysis will largely be confined to the facts of the case and briefings of the parties before them. ¹⁵¹

Moreover, there is already good precedent supporting judicial capability for evaluating government justifications in the free exercise context. Since the enactment of RFRA in 1993 and RLUIPA in 2000, federal and state courts have regularly engaged in a close analysis of government interests burdening religious exercise. ¹⁵² Moreover, many state governments have shown increasing approval for courts to continue undertaking such scrutiny in free exercise. Currently, twenty-eight states have a RFRA on the books, with seven of those (25 percent) adopting their state-level RFRA since 2020. ¹⁵³ This judicial experience with RFRA and RLUIPA over the past several decades bolsters confidence in the judiciary to ably conduct scrutiny of government justifications.

Finally, and perhaps most importantly, focusing on government justifications through factual scrutiny shifts discourse on religious freedom from rights to responsibilities. Under the current paradigm, religious freedom challenges are framed as a right to be claimed and argued against government power. Both Smith and balancing tests encourage religious claimants to view themselves as rightsholders trying to establish entitlement. But under a factual scrutiny approach, this paradigm shifts. With a focus on government’s evidentiary burden to provide justifications, the right to religious freedom is secured by holding government accountable to its constitutional responsibilities. ¹⁵⁴ In a polarized environment where free exercise is often at the center of fights over competing rights, the reciprocal responsibilities that accompany those rights are often overlooked. Consequently, more cases are decided as one-sided conflicts in the courts and less effort is made to find better tailored resolutions in the legislature. ¹⁵⁵ But focusing on the facts of each dispute “concretize the interests on both sides, reduce the level of abstraction, and open up possibilities for compromise”—a benefit especially needed in the current constitutional environment. ¹⁵⁶

Even with the normative benefits outlined above, many may still be wary of factual scrutiny in the free exercise doctrine for several reasons. For example, in her concurring opinion in Fulton, Justice Barrett was openly skeptical “about swapping Smith’s categorical antidiscrimination approach for an equally categorical strict scrutiny regime.” ¹⁵⁷ She wondered how heightened scrutiny through proportionality or something similar will distinguish individuals from groups, and how a new standard will weigh indirect burdens against direct burdens of the religious claimants. ¹⁵⁸

These questions from Justice Barrett’s opinion highlight the problematic approach that Smith has introduced into the current Free Exercise Clause paradigm. Strict scrutiny need not be categorical in nature if the core question to be answered is whether government has shown the necessity of infringing on religious exercise to achieve some compelling interest. Indeed, the “more nuanced” approach that Justice Barrett seems to be searching for lies in making the test more responsive to the facts of the case. ¹⁵⁹ This is essentially the suggestion of factual scrutiny in free exercise: Let the government’s evidence drive the court’s inquiry in each case.

Justice Barrett’s Fulton concurrence also points out that a focus on government justifications still brings challenges for courts such as deciding what interests are “compelling” “substantial” or “important.” ¹⁶⁰ However, under a narrow tailoring analysis, courts do not determine these interests in the abstract. Instead, legislatures are required to assert these interests before the court and connect them to law enacted to accomplish those interests—a much narrower and more precise inquiry. ¹⁶¹ As other scholars have shown, the development of proportionality in foreign constitutional rights cases has proved it to be a valuable tool for apex courts to tackle challenging legal issues, suggesting scrutiny could function in a similar way. ¹⁶²

Beyond Justice Barrett, the academy has presented its own objections to constitutional scrutiny. Following free exercise doctrine, many scholars are resistant to the idea of defaulting to heightened scrutiny because, they assert, such a requirement is more likely to produce “third-party harms.” ¹⁶³ Importantly, however, requiring more justification from government with constitutional scrutiny of facts gives greater opportunity to point out such third-party harms as a part of demonstrating the necessity and suitability of a chosen approach. When relying on government justifications rather than weighing religious burdens or balancing, legislatures are incentivized to bring third-party harm interests to bear when they otherwise might go overlooked. ¹⁶⁴ For example, in both Hobby Lobby and Fulton, little time was spent by the state trying to argue the relationship between the government’s interest in contraception access or placement of foster children and how granting a religious exemption would threaten that interest. ¹⁶⁵ In each case, this resulted in rulings against the government. Knowing how a religious exemption might interact with these interests is a critical piece of evidence that government should provide to justify burdening religious exercise. Under a factual scrutiny approach, government will give greater consideration to this evidence and how it will impact all of society—religious or not.

Conclusion

In looking to the future of free exercise doctrine, I urge factual scrutiny—directly requiring government justifications when religious exercise is burdened. Whereas government action disproportionately impacts marginalized religious groups; conversely, when greater factual scrutiny is used in judicial decisions, law can better protect all religious groups without impeding effective governance. Moreover, adopting a fact-centered approach to constitutional scrutiny also provides three advantages over a strict scrutiny balancing test: (1) this approach coheres with the history and precedent of the Free Exercise Clause, (2) this approach supports the institutional roles of courts and legislatures, and (3) this approach moves constitutional discourse away from a singular focus on rights to a complementary recognition of responsibilities.

The approach I offer is one among many already proffered by First Amendment scholars to uphold, modify, or reject current free exercise doctrine. ¹⁶⁶ However, the foregoing evidence and arguments suggests that any serious doctrinal test of free exercise must hold government accountable to its constitutional responsibility. With greater emphasis on government responsibilities to rights holders through factual scrutiny, responsible protection of religious freedom can model to U.S. society the benefits of choosing cooperation over entitlement and pluralism over conformity.