A Secular Test for a Secular Statute

For RFRA’s strict scrutiny to be triggered, first there has to be a “substantial burden” on religious exercise. The Court has not yet taken that test seriously, but there are straightforward ways to borrow from well-established First Amendment (and other) doctrine. This short piece also comments on treatments of the substantial burden question by Amy Sepinwall and Michael Helfand.

 The Religious Freedom Restoration Act (“RFRA”) provides that “[g]overnment shall not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability[.]”1 “Exercise of religion” includes “any exercise of religion, whether or not compelled by, or central to, a system of religious belief.”2 The government may prevail, though, by showing that applying the law to the claimant is the least restrictive means of furthering a compelling government interest.3

 The law does not say that the government may not apply a law whenever a claimant believes that the law substantially burdens her religious exercise. Some sort of objective test seems warranted by the statutory language. But this is not how Justice Alito approached the matter in Hobby Lobby.4 Instead, he made two other moves. First, Alito referred to the significant tax consequences to Hobby Lobby were it to disobey the health care coverage law, deeming the sums “surely substantial.”5 Second, he deferred to claimants’ belief that the law in question substantially burdened their religious exercise. (Actually, he referred several times to the law burdening or violating claimants’ religious beliefs,6 not exercise. More on that later).

 The issue in Hobby Lobby, and in Zubik,7 turns on purported facilitation of the immoral (on religious terms) act of another. Claimants believe, on religious grounds, that their acts (of providing group health insurance or of filling out a particular form) make them complicit in down-the-line immoral acts (abortions). Alito explained, correctly, that it is not for the government (agencies or courts) to determine the complex moral answers to questions of this sort.8 He then maintained (in a wildly unsupported leap) that concluding there was no substantial burden on claimants’ religious beliefs would “in effect tell the plaintiffs that their beliefs are flawed[,]”9 adding that “it is not for us to say that [plaintiffs’] religious beliefs are mistaken or insubstantial.”10 But, the Court could accept claimants’ testimony about how the acts in question would render them complicit in the immoral end, and still apply a secular test of substantiality to determine what sort of burdens the law covers. This would not tell plaintiffs their beliefs are flawed, mistaken, or insubstantial (or unreasonable11). The Court would not be adjudicating any of these things. Instead, it would be taking them at face value, understanding them, and then developing and applying a secular test for this secular statute.

 Sadly, in the Zubik oral argument, Justice Kennedy appears to have bought into Alito’s total deference notion of figuring out “substantial burden.” Here was one exchange:

JUSTICE KENNEDY: Well, do—do you question their belief that they’re complicit in the moral wrong?

GENERAL VERRILLI: No, we do not.

JUSTICE KENNEDY: Well, then—then it seems to me that that’s a substantial burden.12

Here was another:

JUSTICE KENNEDY: But you’re saying, don’t worry, religions, you’re not complicit. That’s what you’re saying? [Verrilli responds that “the judgment about complicity is up to [the claimants], but that there is an objective limit that RFRA recognizes on the scope of what is a cognizable burden[.]” Kennedy responds:] It seems to me then the analysis has to be whether or not there are less restrictive alternatives and if—is this the least restrictive alternative?13

 Justices Alito and Kennedy are smart guys. Regardless of what one thinks about their (judicial) politics, they are usually pretty good about tracing their arguments from points A to B. But there is no tracing here; there is no explanation of how imposing a secular substantial burden test would be questioning plaintiffs’ religious assertions, or of how accepting claimants’ religious beliefs about complicity ipso facto means there is a substantial burden per the secular RFRA statute. Arguments are available, though, for reaching the conclusion the Hobby Lobby Court reached on the substantial burden question. Amy Sepinwall maintains the law should accept a capacious understanding of how we are complicit in moral wrongs.14 Michael Helfand suggests that any purportedly objective test for substantial burden in the RFRA setting will involve the courts in determining religious questions, in violation of the Establishment Clause.15 (In its place, he proposes that courts examine the size of the penalty for noncompliance.)16 Neither of these arguments, however, is successful. Instead, as they usually do with statutory standards such as “substantial burden,” courts should develop, through the case-by-case interpretive process, a secular, objective, and limited understanding of which burdens from generally applicable law are legally substantial, and which are not. I will start with a description of how such a test might work, and then fend off Sepinwall’s and Helfand’s resistance.

 Law can burden religious exercise in various ways; I will describe three. The first is if law prohibits conduct that, for some, constitutes religious exercise. Smith17 is a good example of this—a law making various controlled substances illegal and a religion that includes ingestion of one of those drugs as part of its religious ritual. Although I have not thought through all possible cases here, there seems a good prima facie argument that any such law imposes a substantial burden on religious exercise, regardless of the size of the penalty. (More on that later when I discuss Helfand’s argument.)

 The second is if law establishes an administrative benefits scheme, and sometimes denies benefits for religious exercise, purportedly for secular reasons. The Sherbert18 line of cases is relevant here—employees refuse work for religious reasons and are denied unemployment benefits on the ground that they lacked good cause for so refusing. These cases are tricky for several reasons, the first being that the laws arguably are not of general applicability, but rather are best seen as authorizations for administrative officials to make case-by-case, fact-intensive decisions. That aside, the burden question is of a different sort because the laws are not directly burdening religious exercise. They are neither forbidding conduct the claimant says is part of her religion, nor requiring conduct the claimant says violates her religious tenets. I forego further discussion of this category here, pointing the reader to my prior discussion of it.19

 The third category was in play in Hobby Lobby and is in play in Zubik: A law of general applicability requires conduct that the claimant says violates her religious tenets.20 In both settings the law is general in two senses—it covers a variety of health-care options (the arguably abortifacient contraception is just a small part of such options), and it applies to all manner of employers (religious employers being a subset). And in both settings, the connection between the required act and the claimed immoral end is indirect—pay for group health insurance and some employees choose to use the arguably problematic contraception in question; or fill out the form requesting an opt-out, which allows the government to put the wheels in motion to offer an alternative method to providing the contraception. Without questioning claimants’ arguments that their participation in these ways renders them complicit in the resulting harms, can we evaluate the substantiality of the burdens? Yes, and we do so in quite a similar fashion in at least three other areas of First Amendment law.

(i) Government funds and religious schools.21 After years of complexity, the doctrine in this area now focuses on generality and indirectness. To take a key example: If the state provides tuition vouchers for education, and some are used at religious schools where religious as well as secular subjects are taught, even though the state is facilitating the religious education (and thus is arguably complicit in it), we do not attribute those ends to the state. Why? Because the program is general (it supports both secular and religious schools) and the money flows through the hands of the parents and then to the school (so the support is indirect). Despite the critiques of the dissenting Justices, this doctrine is well established.

 (ii) Public school space and religious speech.22 In several cases, the Court held that opening public school classroom space for religious extra-curricular activities is permissible. (In fact, it is required under the Free Speech Clause if the space is opened for secular activities.) Even though the school is thereby facilitating the religious activity and ends, and is thereby arguably complicit in such ends, we do not attribute those ends to the state. Why? Because the program is general (it supports both secular and religious speech) and individual students and groups (not state officials) are making the choices of which activities to hold and which expression to engage in.

 (iii) Compelled speech and public forums.23 In spite of strong doctrine protecting a right against compelled speech, the Court has created an exception: If we can see the law as requiring the speaker to host what is in essence a public forum—and thus it is not fostering any particular message but a wide assortment of messages—then although the host is facilitating the speech it does not want to host, and is arguably complicit in such expression, the Court does not see the situation as unconstitutional compelled speech. The forum is general (for an array of speakers) and the host’s connection to the speech is indirect (the speakers choose the expression).

 These doctrines track ordinary ways of thinking about responsibility, from tort and criminal law’s use of proximate cause analysis, to common moral understandings of responsibility to those with whom we have direct, chosen connections as stronger than to those with whom we have indirect, often unchosen connections.24 The doctrines, and the analysis from which they flow, can easily support the conclusion in both Hobby Lobby and Zubik that the acts required by law are not sufficiently connected to the claimed down-the-line harms to constitute substantial burdens on religious exercise.

 Sepinwall acknowledges that on standard legal understandings of complicity, Hobby Lobby should have lost on the substantial burden test.25 She claims, however, that such understandings are based in considerations of proximity that are given a prominent role because “we tend to feel more implicated in conduct to which we bear a closer causal relation[.]”26 But since this tracks a “subjective sense of complicity[,]”27 Sepinwall argues, that is not a good reason “to privilege the law’s conception of complicity over that of the religious objector when the religious objector happens to feel complicit in a greater range of conduct than the standard legal account contemplates.”28 In many cases, she says, we are “without the moral clarity or authority to challenge someone’s belief that the conduct legally required of him would make him complicit in what he perceives as a wrong.”29 She maintains that “many assertions of complicity appear far more compelling from a first person, rather than third person, perspective[,]”30 and that “all else equal, we should deem a complicity claim compelling if the objector deems it so[.]”31

 She gives several examples in which we should defer to a claimant’s view of complicity. Thus, we allow objecting doctors to opt out of performing abortions because (at least here) we protect people from participating in acts they deem immoral.32 We should not distinguish active euthanasia from cases in which doctors provide life-ending drugs to patients, because the “physician is not more morally implicated in the death when she administers a lethal injection that kills her patient than when she hands over the lethal pills and simply watches as the patient kills himself.”33 And a pharmacist who fills an abortifacient prescription is no more or less implicated in the outcome than the cashier who rings it up (assume both know what is being sold and for what reason).34

 I take Sepinwall to be arguing (i) that proximity in the law tracks a subjective sense of complicity, (ii) that we should defer35 to a claimant’s greater subjective sense of complicity than the law allows (in part because we already base proximity determinations in such a subjective sense, and in part from an argument for value pluralism), and (iii) that, in any event, the law’s reliance on the indirectness of a causal relationship does not accurately track moral responsibility. In response, in this brief space, I suggest that our legal conceptions of proximity are in accord with proper, objective moral understandings; and thus I challenge points (i) and (iii), and the part of point (ii) that argues for deference to more capacious senses of complicity, because our law is already tracking a subjective sense. Focusing on the three First Amendment areas I discussed above36 (one could introduce other legal doctrines, as well), our law has a deeply engrained understanding that concepts such as generality, directness, proximate cause, and the like, are appropriate ways of allocating responsibility, and of thinking about harm. If I provide a pot of money for a thousand people to use as they see fit, it is harder to associate me with what they do with the money than if I provide money just for one person. And if I set in motion a chain of events that is affected by a third party’s conscious, intentional, intervening act, it is harder to blame me for the outcome, even if I am an actual cause of the outcome. Do we think reasoning of this sort is based in a subjective sense of complicity? Or do we think it is, in fact, the case that we are less implicated when our connection to an outcome is more general, less direct, and/or broken up by the acts of others? I submit that the latter is the case, that the many legal areas in which generality and indirectness matter track standard, objective, moral reasoning about complicity and responsibility. Remember, we are not questioning religious claimants who say their faith holds a different view of complicity. We are not assessing that at all; we are just applying a secular test to them as well as to others.

 That leaves what I believe is an alternative argument for point (ii): that the law should defer to more capacious senses of complicity from an argument for moral pluralism. I agree that arguments for religious accommodation are well grounded in value pluralism,37 but there have to be limits, and the limits need not come only at the strict scrutiny phase of cases. RFRA states one such limit—the state action in question must place a “substantial burden” on a claimant’s religious exercise. This can be worked out through standard legal conceptions of generality, indirectness, and the like, and not by deferring to claimants’ more capacious understandings of how and when they are implicated in moral wrong.

 Now, to Helfand’s conclusion that applying the type of test I say we should apply would implicate the religious question doctrine, and thus violate the Establishment Clause. He writes: “Interrogating the religious substantiality of conduct on a theological metric runs afoul of core Establishment Clause prohibitions.”38 “Courts lack the tools to engage in line drawing when it comes to determining and calibrating the degree of theological impact a particular law imposes on religion.”39 Additionally, Helfand states:

[A] court cannot reject the religionist’s experience of a substantial burden simply because that experience would be insubstantial if evaluating against prevailing legal standards. To do so, notwithstanding the attempt to employ secular legal standards, would be to take the court’s understanding of religious obligations as relevant over and above the claimant’s understanding. And it is precisely that type of analysis that violates the strictures of the Establishment Clause.40

But courts would not be doing any of this under a generality/indirectness test. They are not evaluating the “religious” substantiality; they are not using a “theological” metric; they are not assessing the “theological” impact; and they are not displacing claimant’s “understanding of religious obligations.” Courts are listening to claimants and (sincerity inquiries aside) are accepting claimants’ testimony about how the law would burden their religious exercise. But the test that the court then applies is secular—it treats the facilitation and complicity claims of religious claimants as it would of secular claimants, accepting or rejecting them as substantial burdens based on a secular metric.41

 In her Hobby Lobby dissent, echoing Justice Stevens, Justice Ginsburg raises a different sort of Establishment Clause concern:

There is an overriding interest, I believe, in keeping the courts ‘out of the business of evaluating the relative merits of differing religi-ous claims,’ or the sincerity with which an asserted religious belief is held. Indeed, approving some religious claims while deeming others unworthy of accommodation could be ‘perceived as favoring one religion over another,’ the very ‘risk the Establishment Clause was designed to preclude.’42

If I am right that courts can sensibly develop and apply a secular substantial burden test (and can do so similarly when it comes to assessing compelling state interest and least restrictive means), then we should not be concerned, as Justices Stevens and Ginsburg have been, about perceived “favoring” of religions that win accommodations claims over those that do not. As Justice Brennan eloquently wrote in Goldman, a regime with no accommodations favors the majoritarian status quo, courts are capable of exercising reasoned, case-by-case judgment, and reasonable perception of such should appreciate that some claims are stronger than others on secular tests, not that courts are improperly favoring religions that secure accommodations over those that do not.43

 Helfand would not just defer to claimants on the substantial burden matter; rather, he would ask whether the penalty/fine/tax/sanction for law violation is substantial.44 I have one principal concern with this test. Helfand gives the hypothetical of a law banning circumcision of a minor child (assume it is a law of general applicability, not targeted or discriminatory), with a $1 fine.45 He says there is no substantial burden. But this seems like a Holmesian “bad man” approach to things. On this approach, prohibitory laws do not impose duties, they just set prices for certain conduct. This is not a very attractive way of seeing law that prohibits conduct, regardless of what one thinks about the related law and economics approach to contract or tort law (which I am not terribly fond of, either).46

 Finally, I want to raise a cautionary note. I mentioned earlier that Justice Alito referred to RFRA burdening or violating claimants’ religious belief, even though RFRA refers to burdens on religious exercise. Perhaps this does not matter. If free exercise of religion involves living in accord with one’s religious beliefs, then laws that burden or violate religious beliefs thereby burden religious exercise. As the Court put it in Lee, “[b]ecause the payment of the taxes or receipt of benefits violates Amish religious beliefs, compulsory participation in the social security system interferes with their free exercise rights.”47 And Justice Stevens wrote, in his separate opinion, “[t]he clash between appellee’s religious obligation and his civic obligation is irreconcilable. He must violate either an Amish belief or a federal statute.”48 So obeying the law may violate a religious belief, which means a burden on religious exercise. In the context of laws requiring conduct that the claimant says violate her religious beliefs, the potential scope of claims is vast, especially when we consider complicity or down-the-line harms, as opposed to direct harms. That Justice Alito mostly talks about burdening or violating religious beliefs, rather than religious exercise, in Part IV of his majority opinion (the part about substantial burden),49 is worth reflecting on. Why does he not just say that providing the group health insurance coverage would burden Hobby Lobby’s religious exercise? Is it because it is somewhat difficult to see how the act of paying for the insurance (or, in Zubik, of filling out the form) burdens religious exercise? Burdening or violating beliefs gets us to conscientious objection more generally; “exercise,” part of the First Amendment, focuses us more on practices, activities, and conduct that are part of religion. Cases involving prohibitions—such as Smith 50 and Goldman51—are easier to see as core, free exercise of religion cases. So, as we move forward in this world of claims based on laws requiring participation in secular federal programs, that are claimed to be inconsistent with a claimant’s religious beliefs, we should appreciate that to accept such as burdens on religious exercise takes some work, and opens the door to a wide assortment of claims. This is one among other reasons for thinking we have to develop an interpretation of “substantial burden” that is consistent with how we think about facilitation, complicity, and causation elsewhere in the web of secular law.

Leonard F. Manning Professor, Fordham Law School. He is the author of two books, Against Obligation and Understanding the 2000 Election, and many law review articles focusing on religion and speech. He clerked for Justice Paul Stevens of the United States Supreme Court. Professor Greene would like to thank Michael Helfand and Amy Sepinwall for helpful exchanges on all of this.

1. 42 U.S.C. § 2000bb-1(a) (2012). In City of Boerne v. Flores, the Court invalidated RFRA, as applied to state and local governments, as exceeding Congress’ powers under section five of the Fourteenth Amendment; but its application to the federal government remains intact. 521 U.S. 507 (1997).

2. 42 U.S.C. § 2000cc-5(7)(A) (2000); see id. § 2000bb-2(4).

3. See id. § 2000bb-1(b).

4. Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014).

5. Id. at 2776.

6. See id. at 2759, 2775, 2777, 2779, 2782.

7. Zubik v. Burwell, 578 U.S. __ (2016).

8. See Hobby Lobby, 134 S. Ct. at 2778.

9. Id.

10. Id. at 2779.

11. Id. at 2778.

12. Transcript of Oral Argument at 46, Zubik v. Burwell, 136 S. Ct. 891 (2016) (No. 14-1418).

13. Id. at 49–50.

14. See Amy J. Sepinwall, Conscience and Complicity: Assessing Pleas for Religious Exemptions in Hobby Lobby’s Wake, 82 U. Chi. L. Rev. 1897 (2015).

15. See Michael A. Helfand, Identifying Substantial Burdens, 2016 U. Ill. L. Rev. (forthcoming Aug. 2016), available at, at *3.

16. See id. at *3–4.

17. Employment Div. v. Smith, 494 U.S. 872 (1990).

18. See Abner S. Greene, Religious Freedom and (Other) Civil Liberties: Is There a Middle Ground?, 9 Harv. L. & Pol’y Rev. 161, 181–83 (2015) (discussing “the Free Exercise Clause unemployment compensation cases”).

19. See id.

20. See generally Zubik v. Burwell, 578 U.S. __ (2016); Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014).

21. See Greene, supra note 18, at 185–87.

22. See id. at 187–88.

23. See id. at 188–90.

24. For a similar approach borrowing from other areas of law, see Frederick Mark Gedicks, “Substantial” Burdens: How Courts May (and Why They Must) Judge Burdens on Religion Under RFRA, 84 Geo. Wash. L. Rev. (forthcoming 2017), available at, at *28–34.

25. See Sepinwall, supra note 14, at 1909, 1937–38, Part II.

26. Id. at 1909.

27. Id.

28. Id.

29. Id. at 1908.

30. Id. at 1956.

31. Id. at 1958.

32. See id. at 1948–51.

33. Id. at 1953.

34. See id. at 1955–56.

35. At least prima facie, before getting to compelling interest, least restrictive means, harm to third parties, etc.

36. See supra text accompanying notes 21–23.

37. See Abner S. Greene, Against Obligation: The Multiple Sources of Authority in a Liberal Democracy 4, 21–23, 139 (2012).

38. Helfand, supra note 15, at *18.

39. Id.

40. Id. at *20.

41. See Gedicks, supra note 24, at *18.

42. Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751, 2805 (2014) (Ginsburg, J., dissenting) (quoting United States v. Lee, 455 U.S. 252, 263 n.2 (1982) (Stevens, J., concurring in the judgment)).

43. See Goldman v. Weinberger, 475 U.S. 503, 519–22 (1986) (Brennan, J., dissenting).

44. See Helfand, supra note 15, at *24.

45. See id.

46. Fred Gedicks says, in response to my argument that a law prohibiting conduct that is part of religious exercise imposes a substantial burden even if the penalty is low, “[b]ut if a claimant may disobey a religiously burdensome law without legal repercussions, it’s hard to see any burden at all, let alone a ‘substantial’ one.” Gedicks, supra note 24, at *17 n.91. My response to Gedicks is the same as to Helfand: When the law says, “you may not do X,” that imposes a substantial burden, regardless of the existence or size of the legal penalty for noncompliance. That law imposes a duty, and you are supposed to heed and follow the law and comply with the duty.

47. 455 U.S. at 257.

48. Id. at 261 (Stevens, J., concurring in the judgment).

49. See Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751, 2777–79 (2014).

50. Employment Div. v. Smith, 494 U.S. 872 (1990).

51. Goldman v. Weinberger, 475 U.S. 503 (1986).

The full text of this Article is available to download as a PDF.