In the United States, Canada, and Europe, courts generally approach the question of religious accommodations using a three-prong analysis. First, the sincerity of the claimant has to be established. Second, the substantiality of the burden or interference engendered by the law or regulation on the religious believer has to be assessed. And finally, there is often some kind of proportionality test to justify the burden with the government interest in order to ensure that the law or regulation is the least restrictive means of achieving or advancing that interest. The precise form and weight accorded to each prong varies across jurisdictions,1 and certainly, in the case of the European Court of Human Rights, or ECtHR, where the court generally accords a wide margin of appreciation with respect to actions undertaken by the national authorities of its member-states.2 Although there has been a considerable amount of scholarly literature on the propriety of religious exemptions writ large, as well as the difficulties associated with determining religious sincerity,3 there is less discussion on the other parts of this framework, much less on how courts should assess the substantiality of a burden.
A recent spate of both decided and pending cases in various jurisdictions however is beginning to place this question front and center. In some cases, the presence of a burden seems clear. For instance, in 2016, the ECtHR held the rejection of a permit for applicants to use an apartment as a place of worship to be interference as this rendered the members of the Jehovah's Witnesses religious community unable to practice their religion.4 But in some cases, it is not as clear. In Hobby Lobby v. Burwell, the U.S. Supreme Court deemed the contraceptive mandate under the regulations promulgated by the Health and Human Services department to be a substantial burden on the exercise of religion by the individual owners of the closely held corporation Hobby Lobby.5 In particular, the owners were concerned that providing full coverage would force them to facilitate wrongdoing by employees who might use the insurance to purchase forms of contraception, despite the charge of the critics that the claim seems too attenuated to merit protection. The court deferred to the claims of the petitioners in determining whether the burden was substantial enough to trigger the application of the Religious Freedom Restoration Act, or RFRA, and based its decision on the result of the proportionality analysis.
Part of the difficulty has to do with conceptualizing what is a burden. In theory, any government action that impacts religious exercise, whether directly or indirectly, is a burden. For example, a mandatory voting law would be objectionable and deemed burdensome to someone whose religious beliefs require abstention from politics. Accordingly, accommodations analysis mandates an assessment of whether the imposed burden is serious or substantial. Courts from different jurisdictions have similar approaches despite differences in nomenclature. Before the Supreme Court of Canada, it is required that the infringement on religious exercise be deemed “non-trivial.”6 To illustrate, a requirement that all driver's license photos show the entire face without any head covering is deemed a nontrivial infringement to a female member of a Hutterite community who believes that such practice is objectionable according to the dictates of her faith.7 In the jurisprudence of the ECtHR, implicit in its assessment that “interference” is present in the exercise of rights secured by Article 9 of the European Convention on Human Rights, is the requirement that it be not de minimis, though ECtHR case law on this tends to subject practices under scrutiny less in recognizing interference8 compared to its American counterpart. In a recent decision, the court's judgment did take into account the small amount of monetary impact on a Mormon church.9
Nonetheless, even with this qualifier, present judicial understandings of burden have been inadequate to capture a wide range of impact that government acts and regulations have on the individual or community practice of religion. An oft-cited example of this problem is the legal protection of indigenous sacred sites under the rubric of freedom of religion. In a famous decision, Lyng v. Northwest Indian Cemetery Protective Association, the U.S. Supreme Court held that the construction of a proposed road through a forest territory deemed sacred by Native Americans would not violate the First Amendment regardless of its effect on the religious practices of the respondents because it compels no behavior contrary to their belief.10 Another example involves complicity cases. Laws and professional codes mandating “effective referrals” by physicians who refused to participate in assisted suicide or abortion procedures have been recently challenged in Australia, Canada, and the United Kingdom. How immediate does the causal chain have to be in order for a burden to be considered one under the relevant legislation or constitutional provisions protecting freedom of religion? These kinds of claims do not readily fall under existing categories spelled out in case law.11 Accordingly, courts and many scholars readily come to the conclusion that there is no burden involved if there is no issue of direct choice or any form of coercion.
This article is set against the background of this ongoing discussion in courts and within the legal academy. I suggest that there is not a single conception, but a variety of burdens that courts have to take into account in order to provide some insights as to how substantiality can be assessed. I outline three possible conceptions of burden, namely (1) burden as coercion; (2) burden as impact; and (3) burden as ratification, and I evaluate each with respect to the central values underlying religious accommodations. Disaggregating burden in this way expands our frames of understanding and allows courts to make better sense of the claims before them. Significantly, it also allows courts to consider what Christopher McCrudden calls a “cognitively internal viewpoint,”12 that is, to understand religion and its significance from the perspective of religious believers, when weighing accommodation claims. This would inevitably require that any evaluation of burden would have to take into account the system of beliefs from which these perceived burdens emanate.13 Although that would understandably raise questions as to the propriety of courts evaluating religious or theological issues,14 it is probably unavoidable to a certain extent. To understand how a law impairs a practice presupposes an understanding of what is important and religiously significant about the practice. As the Supreme Court of Canada held in Bruker v. Marcovitz, 15 “to determine whether a particular claim to freedom of religion is entitled to protection, a court must take into account the particular religion, the particular religious right, and the particular personal and public consequences, including the religious consequences, of enforcing that right.” Indeed, in many cases, courts already do this with respect to the assessment of sincerity despite protestations to the contrary.16 In other words, not all legal questions involving religion are constitutionally suspect.17 It only means that courts should not evaluate religious decisions, that is, the resolution of religious or theological disputes, or the enforcement of religious obligations or the use of secular law to assure observance of religious practices.18
This article undertakes a holistic examination of various kinds of cases before three high courts—American, Canadian, and the European Court of Human Rights—in order to give us a sense of the type of claims found along this spectrum. Given the similarities found in the tests articulated by these courts, the suggestion in this article provides some much-needed guidance that could prove to be workable across these jurisdictions. To reiterate, acknowledging that religious believers confront different kinds of burdens does not relieve courts of the responsibility to assess whether these are indeed substantial enough to merit further balancing with legitimate government objectives by way of balancing or proportionality tests. It does however go a long way toward addressing the gaps that exist in current religion-related case law in the jurisdictions under consideration, the foremost of which is to take religion a lot more seriously than is currently done. To be clear, the argument to expand our understanding of what burden means is distinguishable and analytically separate from the ongoing lively scholarly debate on how to assess its substantiality without running afoul of constitutional limitations. In sharp contrast, the aim here is more modest. Before one can assess the gravity of a claim, the court has to find an a priori burden. Without the conceptual tools that allows courts a more nuanced understanding of burden, religious believers may find their claims dismissed without a just evaluation. They may or may not win their claims in court, but when they lose, it should not be because the court has found they were not deemed to be burdened in the first place.
THE ROLE OF BURDEN IN ACCOMMODATIONS ANALYSIS
Justifying Religious Accommodations
That a select number of individuals and institutions could be exempted from generally applicable laws or regulations on account of their religious beliefs has long been considered a self-evident feature of liberal, democratic societies. But religious accommodations have been met by a number of challenges in recent years, principally the charge that religion is no different from other deeply held moral commitments, and as a result, religion-based exemptions are normatively indefensible as a matter of law and policy.19 Skeptics view this as especially problematic because of the costs it imposes on third parties.20 Nonetheless, for a variety of reasons and underlying values, religion has remained a subject of distinctive treatment in almost all constitutional orders. Accordingly, stating these values is important because they inform each step of an accommodation analysis and they justify what some view as a departure from the rule of law. More importantly, these values help define the contours of religious liberty.
At the heart of the enterprise of religious exemptions lies the value of autonomy. It allows religious believers to hold and manifest their beliefs in ways contrary to what a generally applicable law or regulation would mandate. As a practical matter, it is also not in the interest of the state to put its citizens into a dilemma of having conflicting duties, much less to turn its citizens into martyrs by coercing them. Hence, Catholics could partake of sacramental wine during the Prohibition era and Catholic clergy are exempt from performing same-sex weddings. Part and parcel of the liberal paradigm is for the state to respect the right of people to self-determination and to treat all people equally. In religious exemption cases, religious believers are effectively placed at a disadvantage relative to those who do not subscribe to the same beliefs vis-à-vis what the law requires; hence, exemptions serve to level the playing field between religious and nonreligious persons in fulfilling their duties.21 This is what substantive neutrality, in Douglas Laycock's terms,22 looks like. For example, a Sikh person who must wear a turban at all times on religious grounds could conceivably ask for an exemption from mandatory motorcycle helmet laws, and indeed many jurisdictions such as some provinces in Canada and the United Kingdom provide for this exemption.23
Another justification that undergirds religious exemptions is civil peace. In Lemon v. Kurtzman, the U.S. Supreme Court singled out the divisive political potential of religion as a justification for the separation of church and state.24 This justification has a longer pedigree however. The canonical philosopher of liberalism John Locke attributed “all the bustles and wars, that have been in the Christian world, upon account of religion”25 to the lack of toleration of those who have different opinions. This pragmatic formulation aims to reduce the causes of public conflict and is intended for the benefit of both believers and nonbelievers living together.
A third, though less often employed, justification suggested by scholars is the protection and encouragement of the role of religious institutions and communities as intermediate institutions. On this view, religious accommodations are a good thing because we want to be able to encourage individuals not only to exercise their freedom individually but in concert with others. This mediating function of religious institutions has two dimensions. The first dimension is a negative one in the sense that it serves as a check on the totalizing power of the state. As Mark DeWolfe Howe, an American legal scholar, argued in 1953, “private groups within the community are likewise entitled to lead their own free lives and exercise within the area of their competence.”26 This justification underlies many claims for church or religious autonomy in modern case law involving religion. While constitutional courts are understandably reluctant to concede that churches are alternative sovereigns,27 there is ample evidence to show that some aspects of the jurisdictional approach—or at least a recognition of the collective nature of religious exercise—is making headway at least in Europe and Canada. In Schüth v. Germany, the ECtHR held that “the autonomous existence of religious communities is indispensable for pluralism in a democratic society and is thus an issue at the very heart of the protection which Article 9 affords.”28 Additionally, in a more positive vein, drawing from the tradition of civic republicanism, religious institutions are likewise seen as beneficial for cultivating public virtues. This dimension partly underlies the public funding of religious schools in Europe.
At this point, it should be noted that all these justifications could very well be claimed on behalf of a nonreligious viewpoint, and while that is true, it is liberty of religious conscience that is the subject of this discussion. Whether that is an anachronistic take or not, given the dramatic change in contemporary religious demographics, the morality of treating religion as special in this respect will not be addressed in this article. In any case, the ECtHR has adopted a more capacious view of beliefs and as such, Article 9 of the Convention covers ideas and philosophical convictions of all kinds contrary to the restrictive approach taken by national courts in the United States and Canada.
As mentioned earlier, these values inform all parts of the test most courts use to determine whether to grant an accommodation. These tests, while differing in structure, often require the sincerity of the believer or community, an infringement or interference by the state, and some kind of proportionality or balancing test depending on the jurisdiction concerned. As a reflection of the bifurcated structure of fundamental rights adjudication in general, this framework not only flows from the nonabsolute nature of these rights but also functions as a structure through which burdens of proof are distributed. The requirement of sincerity on the part of the believer respects the freedom of every individual to hold and adhere to certain religious beliefs, however unorthodox or unusual. This is one of the main reasons why courts have abandoned the practice of evaluating the centrality of beliefs in a religious tradition as well as adopting a subjective view of religious sincerity. Thus, the U.S. Supreme Court held that “courts should not undertake to dissect religious beliefs because the believer admits that he is struggling with this position or because his beliefs are not articulated with the clarity and precision.”29 Similarly, the Canadian high court also categorically stated that claimants do not need to prove the objective validity of their beliefs either by reference to experts or other members of a particular religion.30 Justice Frank Iacobucci, writing for the majority, held that a practice is protected “irrespective of whether [it] is required by official religious dogma or is in conformity with the position of religious officials.”31 In Europe, the ECtHR held that an individual's rejection of technologically derived markers such as a tax identification number on religious grounds is protected by Article 9 of the Convention even though this was in sharp contrast to the position expressed by the central governing body of the Russian Orthodox Church to which the individual belonged.32 It is thus sufficient that the individual holds an honest belief. The same rationale operates with regard to churches and religious institutions. Churches are exempt from paying taxes; they have the freedom to hire and remove their own employees without being subject to antidiscrimination laws; and they are exempted from acts contrary to their beliefs, such as performing same-sex weddings under marriage equality laws in place in many jurisdictions.33
The justification stage, on the other hand, reflects the notion that no right, including freedom of religion, is absolute, and which is especially true in a pluralistic society. This stage, whether done by way of European-style proportionality tests or an American balancing one,34 represents the right of the state to uphold the public interest by prescribing laws or enacting regulations for their benefit. There are two usual components in this test. In the language of the RFRA, first, there has to be a compelling government interest that is sought to be achieved by the government act. And second, the act should be the least restrictive means of advancing that government interest.35 In Canada, all the rights enumerated under its Charter of Rights and Freedoms are subject to a proportionality test, namely that these limits should be prescribed by law and as demonstrably justified in a free and democratic society.36 It closely tracks the language of Article 9 of the European Convention on Human Rights which mandates that any limitations “should be prescribed by law and are necessary in a democratic society in the interests of public safety, for the protection of public order, health, or morals, or for the protection of the rights and freedoms of others.” This stage focuses on and safeguards the substance as well as the means by which the state protects the general welfare.
The Evolution of Burden
Between the sincerity requirement and the justification stage stands burden. Burden—substantial burden in the United States, nontrivial infringement in Canada and interference under the ECHR—occupies a peculiar place in accommodations analysis. The requirement takes into account the gravity of the impact of the challenged measure on the religious individual or institution and qualifies the substance of what the believer is being sincere about. But at the same time, it also acts as a gatekeeper doctrine—a legal threshold—in that it compels courts to determine whether the challenged act is significant enough to be balanced against an equally or even greater compelling government interest. The inclusion of the burden requirement thus distinguishes meritorious from nonmeritorious claims such that if a claim is deemed nonsubstantial, however sincere, then there is no need for further justification of the law or policy. In other words, it is a doubled-edged requirement that serves the interest of both the individual asking for an accommodation and the government tasked with advancing its own objectives.
Part of the reason why burden has only recently been the subject of extended scholarly discussion is that it does not have a ready-made place in a jurisprudential world where the state can absolutely regulate external manifestations of religion. As early as 1879, the U.S. Supreme Court adopted the “belief-action distinction” in its decision in Reynolds v. United States, which upheld the conviction of George Reynolds for the commission of the federal crime of polygamy. According to legal scholar Ira Lupu, “Reynolds “drained the free exercise clause of its primary constitutional function[,]”37 because the presence of a legitimate secular purpose has all but eliminated the need to provide any kind of religiously motivated exemptions. Indeed, the overly wide net cast by Reynolds rendered the constitutional guarantee of religious liberty a hollow promise. It should be noted that this decision was largely spurred by nineteenth century fears of the Mormon religion; however, this rule was reaffirmed in the 1940 case of Cantwell v. Connecticut, which incorporated the Free Exercise Clause against the states, and where the court held that: “the [First] Amendment embraces two concepts—freedom to believe and freedom to act. The first is absolute but, in the nature of things, the second cannot be.”38 Despite its erosion in later American free exercise case law, vestiges of this Reynolds/Cantwell line of reasoning remain today. The ECtHR views Article 9 of the Convention as offering absolute protection to one's inner religious freedom (forum internum) but not for the practice of one's religion or conviction (forum externum). In a 2001 decision involving conscientious objection over the selling of contraceptives, the court opined that Article 9 of the Convention “does not always guarantee the right to behave in public in a manner governed by that belief.”39 There was no balancing undertaken in the decision. The ruling simply hinged on the fact that the selling of contraceptives is legal in France, and that the claimants in this case cannot give precedence to their religious beliefs and impose them on others. In these foregoing cases, there is no need to consider burden as part of the legal analysis for as long as the state can advance a legitimate objective. From a rights-protective stance, it is easy to see the flaws in this framework. Religious liberty without the freedom to act on those beliefs is a contradiction. Neither the value of autonomy nor the positive role of religion in a liberal democracy is taken seriously in this equation. Moreover, it also relies on and assumes a distinction between action and belief that is structurally intelligible only within a Christian context.40
Lupu credits the emergence of the modern concept of free exercise burden in the twin cases of Bowen v. Roy and Lyng v. Northwest Cemetery, involving, not coincidentally, Native Americans.41 In Bowen, Stephen Roy objected to the requirement of acquiring a Social Security number for his daughter in order to receive welfare benefits as it would violate their family's religious beliefs. He posited that using a number to identify his daughter would “rob [her] spirit.”42 After the Pennsylvania Department of Public Welfare cut benefits, Roy filed suit. The Supreme Court responded by clarifying what is not a burden, namely, not all burdens on religion are unconstitutional and that the statutory requirement of a Social Security number is a neutral and generally applicable rule that was not meant to discriminate against religion.43 Roy and, later on, Lyng, were significant for equating burden with compulsion and noted that “the Free Exercise Clause does not afford an individual a right to dictate the conduct of the government's internal procedures.”44 In Lyng, the majority opinion wrote that “in neither case … would the affected individuals be coerced by the Government's action into violating their religious beliefs; nor would either governmental action penalize religious activity by denying any person an equal share of the rights, benefits and privileges enjoyed by other citizens.”45
To be sure, this was a significant upgrade compared to the nonexistent balance struck by the Reynolds/Cantwell line of cases, as it at least took into account the possibility that religion might be constitutionally burdened, even if only under a very narrow set of circumstances. Moreover, it also presented a set of guidelines as to how to reach the constitutional threshold of burden. In Sherbert v. Verner, the case that inaugurated the birth of the compelling interest test, the court recognized that the statutory denial of unemployment benefits due to Sherbert's dismissal from work on account of her religious beliefs, as imposing a constitutionally cognizable burden.46 It also gave a clue as to an incipient understanding of burden as coercion as the court pointed out that the denial of benefits essentially forced her to choose between following the dictates of her religion and forfeiting benefits.47 Putting a believer into this kind of dilemma essentially amounted to the same kind of burden as would a direct fine against religious worship. As such, because of the presence of this burden, the government then had to show a compelling, paramount interest in the enforcement of the statute. The burden requirement was not spelled out with clarity, and it would only emerge as a full-fledged threshold doctrine more than twenty years later in Roy and Lyng, but this was a considerable advance from the simplistic distinction made in Reynolds.
As these early American cases illustrate, the emergence of burden as a legal threshold largely served the interests of religious believers, or at least some of them. By providing for a place in the doctrinal framework where claims of gravity can be asserted, it gave claimants an opportunity to prove that their claims were worthy of constitutional protection and, under certain circumstances, override a compelling government interest. The legal threshold of burden renders religion and all its intricacies visible. It is no accident that the evolution of burden coincided with the rise to prominence of religious accommodations as part of, and indeed, according to some scholars, even required by a constitutional commitment to freedom of religion.48
In the European context, burden or interference followed a similar trajectory. Prior to its 2013 decision in Eweida v. United Kingdom,49 the ECtHR had seldom engaged with any indirect interference claims under Article 9. For instance, in a series of rulings involving “a right to resign,” the court held that the opportunity to seek employment elsewhere was an adequate form of protection in cases of a clash between workplace regulations and religious beliefs.50 In a non-employment-related case, the ECtHR ruled that a refusal to grant approval for a religious organization to conduct their own ritual slaughter was not interference within the scope of Article 9 on the grounds that there was another licensed slaughterer in the area and it did not make it impossible for ultra-Orthodox Jews to eat meat slaughtered in accordance with their religious prescriptions.51 Consequently, Eweida marked a sea change in ECtHR Article 9 case law as the court used a proportionality test that seriously takes into consideration the rights involved. With the benefit of learning from the American experience,52 the Supreme Court of Canada has always distinguished between trivial and nontrivial infringement. In its landmark decision of Amselem, the Supreme Court of Canada categorically stated that “unless the impugned provisions or standards infringe the claimant's rights in a manner that is more than trivial or insubstantial, the freedom of religion guaranteed by the Charter is not applicable.”53 What is striking is that, in sharp contrast to, or perhaps as a result of an engagement with early American and European case law, the Supreme Court of Canada seems to have recognized dimensions of burden beyond direct coercion at the outset when it adopted an effects-based approach toward the rights enumerated under the Charter, and hence, even the presence of a generally applicable law, as was the case in Bowen and Lyng, would not prevent consideration of the impact of the regulation on the exercise of religion.
The Gravity of Burden
The foregoing discussion shows the difficulty of defining the contours of burden as a legal threshold. It also says little about the further qualification regarding its gravity. In Alberta v. Hutterian Brethren of the Wilson Colony, the Supreme Court of Canada simply defined nontrivial or insubstantial interference (or infringement) as interference that does not threaten actual religious beliefs or conduct.54 It remains to be seen, however, what that standard looks like in practice. In many instances, it is easy to substitute a claimant's sincerity over his/her beliefs with the substantiality of the impact, that is, a person who is sincere about his or her belief is, by necessity, substantially burdened. This is what happened in Hobby Lobby when Justice Samuel Alito, writing the majority opinion, held that there was a substantial burden on the Greens because noncompliance with the mandate would amount to a fine of as much as $1.3 million per day.55 But this is a mistake. Sincerity is a factual question, while it must be emphasized that substantial burden is foremost a legal conclusion which courts could and should consider and decide.56 Conflating the two essentially leaves the determination of both sincerity and substantial burden up to the claimants and it puts courts in a bind. It shields claims from any judicial scrutiny as most claims under sincerity are protected by the religious question doctrine, which bars courts from adjudicating issues of theology or belief.57 If the claimants say, as they did in Hobby Lobby, that their souls would be eternally damned if they act pursuant to what the law requires, it would not be easy, or indeed even constitutional, for courts to say otherwise. The conflation thus renders the substantial burden prong superfluous. As such, it presents a rule-of-law problem. As Frederick Gedicks notes, courts have a duty to review claims of substantial burdens otherwise, it “invests believers with a presumptive entitlement to exemption from any federal law they feel inclined to challenge.”58
The other end of the pendulum can be equally problematic however. If substantial burden is often mistakenly equated with sincerity by courts, academic literature in the wake of the Hobby Lobby decision has conflated burden with its required gravity in an effort to provide guidance for future decisions. Gedicks, for example, argues that courts can resort to relevant secular doctrines such as causation and harm as defined under tort law to measure substantiality.59 Chad Flanders has argued for a focus on secular penalties as a proxy for gravity.60 In his view, once a claimant demonstrates the secular costs of a law would lead to an abandonment of a sincere religious practice, then the substantial burden requirement has been satisfied. Similarly, Michael Helfand suggests that courts should look at the substantiality of civil penalties triggered by religious exercise in order to determine whether there was a substantial burden or not.61 He also notably counters that Gedicks's secular principles proposal misses the entire object of RFRA, which is to take seriously into account the burdens on the exercise of religious freedom. A believer's experience of religious burden would, in many cases, be deemed insubstantial if evaluated against prevailing legal standards. In any case, an examination of causation would inevitably lead to a prohibited theological inquiry because it would involve the court assessing the substance of a religious claim. A good example here is the claim made by the Little Sisters of the Poor, an order of Roman Catholic nuns that operates nursing homes, that the health care mandate to cover contraceptives in their self-insured health plan goes against their religious beliefs.62 The nuns could exempt themselves if they notify the government in writing, but they likewise object to this requirement because that written notice would effectively implicate them in the use of contraceptives. In their view, participating in a chain of events that leads others to the use of contraceptives would make them complicit in what they deem to be sinful activity. Evaluating this causation claim means the court has to delve into the theology behind it, which would then implicate the court in a constitutionally forbidden enterprise.
All these are important and useful efforts to introduce some clarity to a debate that has recently acquired salience, but it is equally significant to note that they do not adequately address the basic, prerequisite question of what a burden is. This matters because, as stated at the outset, the burden prong of a religious accommodations regime is meant to serve the interests of both the believer as well as the government. While it is true not all burdens are equal—hence the substantial or nontrivial qualifier—courts have often gone the route of equating substantiality with burden and that should not be the case. Consider the following example: a person sincerely believes that the government cannot impose taxpayer numbers on him because this constitutes a mark of the antichrist and therefore contrary to his religious beliefs. Without any further act on the part of the government to condition his access to benefits or services on the use of this number, this can be considered a burden on religion and still not be substantial enough to merit justification. It is thus entirely conceivable for a court to do any of the following: (1) identify a burden, consider it substantial and then the claimant loses at the justification stage because the interest of the state is compelling and it was the least restrictive means to achieve that interest; (2) identify a burden, consider it nonsubstantial, and then forego the justification stage entirely, in which case the claimant also loses; or (3) identify a burden, consider it substantial, and then the claimant wins at the justification stage. It is important that burden is distinctively identified because it performs a procedural and substantive function. As the late Robert Cover notes in his classic article Nomos and Narrative, an invasion of the nomos of insular communities deserves to be treated with a commitment as fundamental as that of the majority community's.63 It is part and parcel of recognizing the value of pluralism in a society committed to liberal democratic ideals. Contrary to conventional wisdom, that recognition is done not at the level of respecting a believer's sincerity but at the level of acknowledging burden. To believe someone or some group is sincere is to acknowledge the autonomy of individuals and communities to choose their beliefs, whether religious or not. It refers simply to the honesty with which they hold such belief. To recognize burden however is to acknowledge the normative pull of those beliefs.64 Suppose, for example, a claimant is sincere about wearing a crucifix. Workplace regulations require that no visible religious symbol should be worn as part of the company aesthetic. If one is simply to rely on the sincerity prong of the test alone, it tells us very little about the significance of wearing a crucifix as part of one's spirituality and why one might be entitled to an accommodation. Burden thus speaks to the actual conflict that results from the application of the law or regulation to the believer's observance of his or her religion. Religious beliefs, after all, are not products of mere choice. Indeed, the Lyng decision met a huge amount of criticism precisely for its dismissive approach to the Native American way of life. To be sure, drawing lines between no-burden/burden and substantial/nonsubstantial burdens is a difficult exercise, but it is far from impossible; courts are engaged in line-drawing most of the time.
In addition and more significantly, these proposals also assume a singular conception of burden—one which is largely defined by coercion. It seems prudent then for courts to consider several conceptions of burden, especially as we broaden the milieu beyond the American context involving the First Amendment and RFRA, and more importantly, beyond mainstream religions. For instance, in the Canadian case of Ktunaxa First Nation v. British Columbia, the question is whether the claim that the construction of a ski resort, including permanent overnight accommodation, on a mountain deemed sacred by aboriginal peoples will cause the Great Spirit Bear to leave the area and thus render all their religious activities meaningless, is covered by the freedom of religion guarantee in the Charter of Rights and Freedoms. As in Lyng, the set of facts does not lend itself to easy analysis within the existing doctrinal framework. There is no coercion involved and it is easy to imagine that a court could be at a loss as to how to evaluate such claim using the “nontrivial infringement” criteria. Sure enough, the Supreme Court of Canada came up with an inscrutable answer by stating that the Charter “protects freedom of worship but does not protect the spiritual focal point of worship,”65 and therefore there was no infringement.
Thus, widening the frame of understanding as to encompass the varieties of burden that religious believers and communities routinely encounter could help courts in recalibrating the weight put on each prong of their respective religious accommodations regimes. In keeping with the place of burden in accommodation, a partially subjective view of burden mandates that the claimants should be able to assert what they consider to be a burden on their own religious beliefs but at the same time, it is also a duty of the court to determine if there has been some kind of pressure induced by the government.66 This solves the conundrum of courts not getting into the business of defining people's religious commitments for themselves and likewise illustrates the close link between sincerity and burden.
BURDEN AS COERCION
The paradigmatic example of burden is one of coercion. This is borne out by the history of religious liberty, which is replete with extreme examples of state-sponsored physical coercion involving persecution, torture, and even death for religious dissenters. Out of this long history have come philosophical tracts that have formed the conceptual underpinnings of how we understand religious freedom in the present. Today, there is still no shortage of these forms of persecution in many parts around the world.67 On the less extreme end of the spectrum, however, lie dilemmas that involve imprisonment,68 fines,69 denial of benefits,70 and even seizures of property,71 as equally unpalatable alternatives. More recently, this type of burden is also envisioned in most conscientious objection legislation involving health care.
In all these cases, the believer is put to a stark choice: either comply with the law or incur a cost. Commentators usually disaggregate cost into two categories, namely religious cost and secular cost. Religious cost pertains to the psychic damage—often a reflection of the degree of religious conviction—that the claimant will feel if he or she is forced to comply with the law or regulation that goes against the dictate of his or her faith. Examples include a Jehovah's Witness doing a flag salute, a Muslim individual who misses Friday prayers, an observant Jewish prisoner who is given a nonkosher meal, or a Hutterite who has a photograph taken for a driver's license. Religious costs or burdens are, for good reason, immune from secular evaluation as they mostly pertain to beliefs that are, by their nature, mystical and often not susceptible, in large part,72 to claims of rational evidence. Like claims made under the sincerity prong of religious accommodation frameworks, this is correctly shielded from judicial scrutiny by virtue of the religious question doctrine or its analog which disables courts from passing judgment on any issue that involves theology.73 For example, in Amselem, the claimant wanted to be able to erect a personal succah to celebrate the Jewish holiday of Sukkot contrary to the building code which only allowed for the construction of a communal one. The Supreme Court of Canada pointed out that the infringement on the right to freedom of religion was more than trivial because “the alternatives of either imposing on friends and family or celebrating in a communal succah … will subjectively lead to extreme distress and thus impermissibly detract from the joyous celebration of the holiday.”74 While the sentiment is commendable, this statement is rather problematic because whether or not this practice is indeed emotionally and psychologically onerous is not a judgment that a court should feel readily competent to make. In practice though, courts are, and rightfully so, generally deferential with respect to the asserted religious cost of a religiously motivated practice.
A secular cost, on the other hand, refers to the actual, objective effect on the claimant who is being essentially forced to choose. In Multani v. Commission Scolaire, for example, the Supreme Court of Canada recognized that the interference with Gurbaj Singh's freedom of religion is neither trivial nor insignificant as he is faced with the choice of leaving his kirpan (a kind of ceremonial dagger) at home and leaving the public school system.75 Singh eventually decided to leave his public school rather than give up carrying his kirpan on a daily basis. In Eweida, the ECtHR held that the employee is forced to choose between manifesting her belief by wearing a visible cross contrary to existing uniform regulations and losing her employment.76 In Hobby Lobby, the U.S. Supreme Court pointed out that the cost for the Greens in sticking to their religious beliefs was a fine of almost 1.3 million dollars a day. In Wisconsin v. Yoder,77 the secular cost was a combination of fine and imprisonment for violating the state's compulsory school attendance law in refusing to send their children to high school in keeping with the tenets of Old Order Amish communities. Of course, all religious obligations carry both secular and religious costs. Nevertheless, compared to religious costs, a court can objectively ascertain the extent of gravity based on this secular criterion.
BURDEN AS IMPACT
Another way of understanding burden is one of impact. This arose from and is largely based on Justice William Brennan's dissent in Lyng and is directed toward religious beliefs and practices that do not quite fit within a coercion-based framework, the principal example of which is the protection of sacred sites. This is certainly not only limited to indigenous peoples’ religious practices although principally applicable to them as most of their sacred lands are often public property. Churches and mosques are more than likely to be privately owned and can be protected under the rubric of other rights such as property rights.
Remember that in Lyng the Supreme Court acknowledged that the proposed road construction would virtually destroy the Indians’ ability to practice their religion but ruled anyway that believers do not have a veto over the government's decisions. The opinion further continued that “incidental effects of government programs which may make it more difficult to practice certain religions but which have no tendency to coerce individuals into acting contrary to their religious beliefs”78 cannot require the government to offer a compelling interest argument. In his scathing dissent, Justice Brennan forcefully argued against the court's emphasis on the form, rather than on the adverse effect, of government regulations. Note that the decision in Lyng came about notwithstanding the existence of the American Indian Religious Freedom Act,79 which in theory is intended to protect all aspects of Native American spirituality, including traditional religious rites and cultural practices. More recently, the Lyng reasoning was repeated in Navajo Nation v. US Forest Service,80 which involved a claim by Indian tribe members that the use of recycled wastewater to make artificial snow for skiing on the Snowbowl, an area that covers a part of the San Francisco Peaks located in northern Arizona, will spiritually contaminate the entire mountain and devalue their religious exercise. The Ninth Circuit dismissed the claim, and remarked that “there is nothing to distinguish the road-building project in Lyng from the use of recycled wastewater on the Peaks.”81 Accordingly, no burden in Lyng also meant no burden in Navajo Nation. The court noted that the “only effect” of the upgrades would be on the subjective and emotional religious experience of the Navajos.82 It characterized existing Supreme Court precedent as mandating that the diminishment of spiritual fulfillment could not be considered as a substantial burden on the free exercise of religion. This type of reasoning, as problematic in 2008 as it was twenty years prior, animates this present effort to clarify the variety of burdens that religious believers face for the sake of a more reasonable accommodations regime.
Moreover, the text of the Religious Land Use and Institutionalized Persons Act explicitly supports this distinct conception of burden. Consider its subsection, which defines its scope of application: (a) the burden is imposed in a program or activity that receives federal financial assistance, even if the burden results from a rule of general applicability, or (c) the substantial burden is imposed in the implementation of a land use regulation or system of land use regulations.83 The Act was drafted to address the plight of churches and other religious institutions and prevent discrimination against them by local governments in the exercise of their zoning power.84
Conceptualizing burden in this way is important as the number of religious claims emanating from indigenous communities everywhere increase in number.85 Unfortunately, Article 9 of the European Convention on Human Rights did not take indigenous concerns into account when initially drafted, nor does it do so as currently interpreted by the ECtHR.86 Similar to other jurisdictions, freedom of religion is largely interpreted as an individual right that protects personal beliefs and under certain circumstances, the right to manifest those beliefs in practice. A further obstacle consists of most courts’ reluctance to recognize collective rights. As political scientist Lori Beaman wrote, Native Americans are doubly disadvantaged because “the space they identify as sacred does not resonate with the religious views of the Christian mainstream, and the manner in which their claims are articulated fall outside of the acceptable framework for rights claims … .”87 In Canada, the early anticipation directed toward the Supreme Court of Canada's resolution of the Ktunaxa case is due in part to the lack of any applicable framework within which to evaluate claims relating to aboriginal spirituality. Indeed, the Court of Appeal decision did not even explicitly state whether the construction of the resort amounted to an infringement, and whether, in turn, such infringement or interference was trivial or not.88 The decision did mention that the construction implicates the vitality of the Ktunaxa spiritual community but it did so without making a pronouncement on its implications. After the Canadian Charter came into existence in 1982, there have been few cases which involved aboriginal claims of religious freedom, and in those instances, the high court resolved them by recourse to aboriginal-specific legal doctrines, such as treaty rights, rather than through a general rights framework.89 For example, in R v. Sioui, a Huron band of aboriginal persons were convicted by a court for cutting down trees, camping, and making fires in an unauthorized area of a park located outside the Indian reserve. The band claimed that these activities were in performance of ancestral customs and religious rites.90 The Supreme Court of Canada struck down their criminal convictions on the ground that these activities were covered by a treaty agreed to between the Hurons and Britain in 1760, which is in turn, protected by a statute. Whether or not aboriginal spirituality is better off protected by a regime distinct from that of a bill or charter of rights applicable to everyone else is a matter of debate. But that should be an available choice in the first place.
This conception can also be useful in non-Indigenous contexts, although it is and presumably will be invoked far less frequently. It can render comprehensible, for instance, state protection for the religious feelings of believers, which would also otherwise fall outside the conventional purview of burden as coercion. Provocative or gratuitous negative portrayals of religions are considered serious burdens on one's religious feelings and therefore could be legitimately addressed by government measures even if these would amount to a curtailment of freedom of expression. In Otto-Preminger-Institut v. Austria, the ECtHR defined freedom of religion to include, in extreme cases, protection from the provocative portrayals of objects of religious veneration, such as a film which portrays religious figures such as Jesus Christ and the Virgin Mary in a critical manner.91 The Court upheld the state measures of seizure and forfeiture of the film to be motivated by a legitimate aim and necessary in a democracy society. Consider also the scenario in the case of Braunfeld v. Brown,92 in which Orthodox Jewish merchants who ran small retail businesses challenged a state Sunday closing law as a violation of their free exercise. Because of their faith, these merchants must close shop on both Saturday and Sunday which they argued would lead to their financial ruin. Braunfeld was not decided using the contemporary framework provided in RFRA and accordingly, the court did not have to go through the tripartite test set in the statute, but Chief Justice Warren, writing for the plurality, acknowledged that the statute imposed an indirect burden on the merchants but that the state had a legitimate secular objective in imposing a uniform day of rest.93
Finally, viewing burden through this lens can also make sense of the burdens imposed on the exercise of religion in the special context of prisons. In an older European Commission case of Galloway v. United Kingdom,94 a prisoner objected to the enforced exposure of his private parts in the context of a compulsory drug test as an “abomination, and strictly against the teachings of his church and the principles of his belief,” and claimed a violation of Article 9. The European Commission held up a familiar problematic and rather narrow view of interference and ruled that since the examination did not coerce him into changing his beliefs or inhibiting his free exercise of religion therefore there is no legally cognizable interference with his freedom of religion.95 The European Commission could have acknowledged the presence of an interference if only to recognize the importance of freedom of religion, even as it concludes that the interference is justified in the interest of public order.
BURDEN AS RATIFICATION
A third understanding of burden is one of ratification.96 In this conception, the act of ratifying what one considers to be an immoral or sinful act is deemed a burden that should merit legal cognizance. This burden is exemplified by the variety of religious and moral objections to abortion, same-sex marriage, and assisted suicide. For instance, one of the underlying issues behind the critique that the U.S. Supreme Court was overly deferential to what the Hobby Lobby plaintiffs asserted as a substantial burden is the notion that complicity-based claims cannot be considered burdens for its attenuated character.97 That is, as the dissenting opinion notes, there is no direct through line between the provision of the health plan by the business owners and the commission of the wrong, as it is up to the individual employees to avail themselves of the contraception coverage. A similar objection met the claims for exemptions by the University of Notre Dame and the religious order Little Sisters of the Poor as when the university argued that mailing a certification form giving notice to the health care providers that they are exempt would “trigger” the coverage of contraception and make the university an accomplice in the wrongful conduct,98 and when the religious sisters likewise argued that filling out a form in which they register their religious objections and which would prompt a third-party to provide coverage instead, would do the same.99 Caroline Mala Corbin argues for example that the belief of these nonprofit organizations that they are complicit in the sin of contraception rests on the mistaken assumption that it is their written refusal that triggered the provision of contraception.100 In Canada, a similar situation arose recently in the aftermath of the Supreme Court of Canada's decriminalization of assisted dying when a group of Christian medical professionals challenged the policy of the Ontario College of Physicians which requires doctors to perform an “effective referral” in cases of conscientious objections.101 These professionals argued that the mandatory referral policy made it a moral equivalent of providing the procedure themselves.102 This situation can be distinguished from cases of direct conscientious objection where the burden involved is properly considered to be one of coercion, not ratification. Viewed from another angle, these pertain to what scholars Douglas NeJaime and Reva Siegel term as “complicity-based claims,” which is deemed distinct by virtue of their unique capacity to harm other citizens.103
One common charge against these claims pertains to their purported falsity. Filling out a form in which the religious group registers their objections would actually not trigger contraception. If this was an analysis being done under the sincerity prong of the accommodations analysis, this would not pose any difficulty. Courts have generally adopted a hands-off approach when it comes to the truth of any religiously grounded claims. Because this question is asked pursuant to a court's determination of whether a legally cognizable burden exists, the notion that it would bring about the evil that the members of the concerned group wanted to avoid is deemed to be absurd. As Helfand remarked, what makes these claims absurd is not necessarily their falsity but that it would be harder to sustain the link between the sincerity in his belief that he or she is being substantially burdened.104 Burden, as earlier argued, is a mix of both law and fact. But the requirement that there should be a directly traceable showing of causation between the belief and the conduct being objected to unfairly disregards the psychic cost for religious believers, and operates as an indirect way of passing and substituting judgment on theological questions which would have been otherwise invalid. Is there really no cognizable burden involved when a religious believer considers it immoral to participate in what he or she views as a sinful activity? Catholic doctrine, for instance, distinguishes between formal and material cooperation. Material cooperation occurs when one participates in some way in the wrong act even if one does not intend the outcome.105 There are degrees involved in this kind of cooperation, given the voluntariness and the proximity to the result, but it is quite clear that Catholic organizations and individuals are not permitted to engage in immediate material cooperation in actions that the church deems to be intrinsically immoral such as abortion or euthanasia.106 From a conscientious believer's point of view, these religious norms have the force of law even though in practice, responses are more complex and varied than just simple dissent or compliance.107 Thus the psychic cost involved in what the believer deems to be a sinful act despite the absence of direct causation is thus best explained as burden of ratification. The burden lies in the act of expressing sanction or approval of the end result, not necessarily that filling out the necessary paperwork would literally bring it about. As legal scholar Amy Sepinwall explains, what objectors in the abovementioned cases involving the University of Notre Dame and the Little Sisters of the Poor are essentially objecting to is the fact that “acceding to the government's accommodation process makes them cogs in what they deem to be an abominable wheel.”108 Filling out a form that explicitly states that “The organization or its plan must provide a copy of this certification … in order for the plan to be accommodated with respect to the contraceptive coverage requirement” consequently facilitates the sisters’ participation.109
To be clear, the foregoing categories are not necessarily stand-alone categories, nor are they exclusive. It is entirely plausible that a burden though primarily one typified by coercion, would necessarily involve some impact or ratification and vice versa. These categories are used mainly to introduce some analytical clarity in this muddled area.
How Substantial a Burden?
Thus far my argument turns the conventional wisdom on its head by suggesting that it is on the claimant to argue whether there is a burden in the first place (at least any of the three kinds proposed here), taking into account the centrality and importance of one's religious beliefs, before putting the onus on the court to ascertain its substantiality.110 The main import of this framework is that it disaggregates the substantial burden requirement into (1) the assertion of a burden and (2) the required gravity. To illustrate the argument, consider once more the Little Sisters’ argument that signing the disputed form, even if its immediate object was to provide notice of their religious objections and thus exempt them, effectively facilitates the provision of contraceptive coverage in contravention to their religious beliefs’ and therefore makes them complicit. As an appellate court correctly held, not all burdens amount to substantial burden.111 One possible analysis could recognize that there was indeed a burden (ratification) on the religious exercise of the Little Sisters but that it would not be substantial because their interpretation of what the federal law requires was wrong. This conclusion might not seem all that different from the Tenth Circuit's holding that the law did not require the court to defer to the erroneous view about the operation of the health care law and its regulations.112 But at least it recognizes the weight that this requirement places on the Little Sisters and is arguably a much better approach than just completely dismissing their claim.
To be sure, as other scholars have correctly argued, believers should not be entitled to complete deference on whether a burden is substantial or not.113 Articulating the varieties of burden affords courts some clarity in determining what is a burden without changing that partially subjective position. Thus, in the Ktunaxa case, the recognition that the construction of permanent human accommodation on a sacred mountain posed a legally cognizable infringement on the beliefs would still require the Supreme Court of Canada to evaluate whether such infringement is nontrivial. Acknowledging that there is a burden on the Little Sisters of the Poor or the University of Notre Dame when they refuse to sign the necessary paperwork to trigger the third-party process that provides contraceptive coverage for their employees does not prevent the U.S. Supreme Court from determining whether that is substantial enough in order to shift the burden to the government to justify its compelling interest.
Admittedly, the search for a limiting principle still remains. Obliging courts to evaluate the substantiality of a claimed burden not only serves rule-of-law concerns, but given the capacious view of burden I suggest here, it would also help in limiting the number of claims that can be put forward. There are plenty of good reasons why burden has been defined narrowly by courts. A major objection is the understandable concern that it might open the floodgates to all kinds of claims, putting a strain on judicial resources and subjecting all kinds of government action to a theoretical veto. A related objection is that, at least in the American context, a broad conception of burden would subject all government action to a strict scrutiny test, and thus dilute the gatekeeping force of the test. It bears emphasizing however that American-style strict scrutiny or European-type proportionality tests, only come after a court has determined that (1) there was a burden and (2) it was substantial. In other words, the argument is simply that burden comes in many shapes and sizes, and that courts have thus far failed to recognize them. That does not absolve the court from having to determine the gravity of the burden and consequently, weighing it against other equally important interests.
So how should a court assess substantiality? Fortunately, this is not an altogether uncharted territory. In a case involving parents’ objections to having their children participate in a mandatory Ethics and Religious Culture program as it infringed on their right to pass their faith on to their children, the Supreme Court of Canada articulated an additional requirement of proving interference from an objective standpoint, that is, “proving infringement requires an objective analysis of the rules, events or acts that interfere with the exercise of the freedom.”114 As with U.S. scholars currently grappling with and debating this issue, the court likewise held a similar concern then. To rule otherwise would allow persons to conclude themselves that their rights had been infringed and thus to supplant the courts in their role.115 This objective dimension of the evaluation of burden closely tracks Michael Helfand's suggestion to consider the civil penalties for noncompliance as well as that of Paul Billingham's suggestion that “courts should determine the extent of the burden on a religious practice by using an impartial theory of individual interests to determine the cost that the individual bears” upon compliance.116 The main difference between Helfand's account and the one described in this article is that Helfand combines substantiality with the presence of burden, which under some circumstances, leads to the unfortunate result of conflating money with principles.117 By contrast, under the account I proposed here, the evaluation of whether the monetary penalty reaches the substantial threshold would only come after the court has already determined the presence of burden. For example, in the Hobby Lobby case, the Supreme Court could recognize the burden suffered by the claimants as one that involves an unacceptable ratification of what the U.S. Health and Human Services mandate leads to, which in their view, is abortion, and at the same time assess whether the accompanying daily fine of two thousand dollars would be substantial enough to make that burden meet the essential threshold under RFRA. In the Canadian context, the Supreme Court of Canada could have conceivably recognized the burden on the Ktunaxa First Nation community as a result of the desecration of what they deem to be sacred ground, and given that the main source of the community spirituality will be completely eradicated, it could evaluate and conclude that this is clearly a nontrivial infringement. In both cases, it is up to the court to decide the gravity of the burden and weigh it against the countervailing interests of the state. How they arrive at that assessment, however, is currently the subject of a lively scholarly discussion in the United States and beyond the scope of this article.
Some questions that can be presumably raised include: as the US Supreme Court ultimately ruled, whether the contraceptive mandate was the least restrictive means of achieving the objective of advancing public health and particularly women's health care. In the Ktunaxa case, the question would be whether the countervailing interest of allowing commercial development on the mountain is more important. Taking burden seriously in the manner proposed here would do away with the “meaningful choice” standard that was invoked in the Supreme Court of Canada cases of Hutterian Brethren and R v. NS 118 (involving the question of whether a witness could be required to remove her niqab when testifying in court) in the same way that the Eweida case shifted the trajectory of ECtHR Article 9 case law by departing from the “right to resign” rule, that is, there is no interference when a person voluntarily accepts a position where limits are placed on the free exercise of religious beliefs or where an employee is free to leave his or her employment to continue following his or her religious observances. In all likelihood, courts would still consider these factors but at least they would do so in the context of balancing competing claims, and not use them as another basis to dismiss outright any claims of burden, which, in turn, would preclude any further justification on the part of the government.
Ultimately, it is up to the court to decide and weigh the importance of the interests involved, but it should do so with a proper recognition of those interests. Of course, there are always understandable constitutional concerns about the propriety of courts entering the thicket of religion-related matters119 but as mentioned earlier, a reasonable religious accommodations regime is not realistic without the courts’ involvement on each step of it. For instance, it would hardly be anathema for secular courts to take judicial notice of the fact that opposition to abortion is part of Catholic doctrine, or that various religions have rules on dietary restrictions. This type of inquiry is not the kind of theological inquiry envisioned by the so-called religious question doctrine to be avoided by courts. Indeed, there is an important difference between ruling on the validity of a claimant's beliefs or adjudicating intra-faith differences and that of examining whether the asserted belief or infringement has any relation to the creed or tradition to which one subscribes for the purpose of determining the presence of burden.120 That is clearly one that courts can and should evaluate in order to give life to the constitutional guarantee of freedom of religion.
Balancing constitutionally protected religious rights with legitimate policy objectives by way of religious accommodations is far from being a straightforward exercise.121 At the moment, courts have largely clung to restrictive interpretations of what constitutes a burden, which do not necessarily map onto the myriad claims that believers of all stripes inevitably encounter in their daily lives. This leads to unfortunate results and makes short shrift of the constitutional promise to celebrate pluralism and its guarantee to protect religious liberty. While there is much room for further analysis to clarify how the proposal in this article would work in practice given the complexity of these types of cases, a legal recognition of how believers experience religion and the effects of how attempts by the state to restrict these practices in the name of a higher goal would go a long way in crafting a fair and clear accommodations regime.