* University Professor, Columbia Law School.
At the end of June 2014, the Supreme Court decided one of the most publicized controversies of decades. In a decision covering two cases, widely referred to as Hobby Lobby, 1 Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751 (2014). the Court held that closely held for-profit corporations, based on their owners’ religious convictions, have a right under the Religious Freedom Restoration Act (RFRA) 2 42 U.S.C. §§ 2000bb to 2000bb-4 (2012). to decline to provide employees with insurance that covers contraceptive devices that may prevent a fertilized egg “from developing any further by inhibiting its attachment to the uterus.” 3 Hobby Lobby, 134 S. Ct. at 2762–63. Relatively recent evidence indicates that one of the four devices, Plan B, may not have this effect. See Robin Fretwell Wilson, The Calculus of Accommodation: Contraception, Abortion, Same-Sex Marriage, and Other Clashes Between Religion and the State, 53 B.C. L. Rev. 1417, 1456–57 (2012) (noting “while Plan B appears to have no effect after fertilization, . . . post-fertilization effects cannot be definitively ruled out”).
The result has been widely approved by those who favor an extensive scope for religious liberty and strongly criticized by others who worried that it will undermine basic concerns about equality for women and nondiscrimination based on sexual orientation. 4 Among those who believe the decision was sound is Paul Horwitz, The Hobby Lobby Moment, 128 Harv. L. Rev. 154, 157 n.16 (2014) (“In short, I think the Court was right in Hobby Lobby.”). Among the critics is Ira C. Lupu, Hobby Lobby and the Dubious Enterprise of Religious Exemptions, 38 Harv. J.L. & Gender 35, 35 (2015) (asserting “grounds for deep skepticism of any sweeping regime of religious exemptions”). Among other effects, it has sharply intensified both efforts to enact new state laws like RFRA and the opposition to such efforts.
The focus of this Essay is not upon whether the result itself is on balance justified, which I urge is indeed a close question. Rather, it describes and criticizes the basic approach to interpretation and the standards adopted by the Court’s majority. It examines in successive Parts the Court’s treatment of the relevant statutory criteria, offering some critiques that, as construed, their effectiveness is sharply curtailed. But the overarching theme is that these standards should not have been addressed wholly independently of each other, but with consideration of how together they can accomplish what the statute aims to do.
Of course, a realist perspective suggests that what is articulated in a majority opinion does not necessarily accurately describe how the Supreme Court Justices have reached a conclusion; but as outsiders, that is what we have to assess matters, at least until decades later when internal documents may be made public. 5 For a treatment on this uncertainty about opinions, written to satisfy Justices who may have different perspectives and to put determinations as more obvious than they really are, see generally Kent Greenawalt, Statutory and Common Law Interpretation (2013) [hereinafter Greenawalt, Statutory]; Kent Greenawalt, Interpreting the Constitution (2015) [hereinafter Greenawalt, Interpreting]. Moreover, if one believes it is desirable for opinions generally to be candid about how judges reach conclusions, that also provides an important reason to focus on what opinions say.
RFRA constitutes a very unusual statutory provision. In order to reject the Supreme Court’s curtailment, in Employment Division v. Smith, 6 494 U.S. 872 (1990). of the extensive free exercise protection provided by rulings in the preceding three decades, Congress explicitly adopted what it conceived to be the appropriate constitutional standard. 7 See 42 U.S.C. § 2000bb(a)(5), (b)(1) (adopting compelling interest standard as test applicable to free exercise claims); S. Rep. No. 103-111, at 2–3 (1993) (same). Although the Court subsequently ruled that the law did not validly apply to states, the essential standard remains as enacted for federal law. 8 Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 439 (2006) (holding RFRA valid as applied to federal law); City of Boerne v. Flores, 521 U.S. 507, 534–35 (1997) (holding RFRA invalid with regard to state law). Given all this, how to construe and apply this law’s language involves a rather strange combination of statutory and constitutional interpretation. 9 For a lengthier discussion of these subjects, see generally Greenawalt, Statutory, supra note 5; Greenawalt, Interpreting, supra note 5.
The overall themes of both this Essay and those volumes is that basic approaches to interpretations are not simple nor should they be, despite the assertions in many opinions of Supreme Court Justices and some scholarly writing that suggest otherwise. 10 See supra note 5 (citing author’s previous work on uncertainty of opinions). The nature of particular provisions and the subjects they cover are elements of context that need to be taken into account. Often, components that may be articulated as distinct should be related to each other in their application. 11 A simple example of this is “probable cause” that someone has committed a crime, needed for an arrest; “probable” here depends not only on the degree of likelihood of criminal behavior but also on the gravity of a crime and the likelihood that the individual who may have committed it will be able to escape altogether if not stopped now. See Greenawalt, Interpreting, supra note 5, at 315–19 (exploring Court’s treatment of probable cause and highlighting problems implicating “extent to which judges should employ general categorical standards or undertake particular evaluations”). With many standards of application that courts may or may not choose to adopt, a crucial consideration can be their administrability. How easily can officials, lower court judges, and perhaps juries, apply those standards in a variety of instances that will arise?
The Essay’s fundamental contentions are that Justice Alito’s opinion in Hobby Lobby adopts an approach that is excessively formalistic, that it treats as separate certain elements of the statute that should be seen as interrelated, and that it takes inadequate account of concerns about administrability.
This Essay begins with a detached description of what the case involved and what the various opinions of the Justices actually assert. This descriptions is largely aimed to provide readers who are not already closely familiar with those opinions with a basis to assess the fairness of the main critique offered here.
The two families controlling the companies involved in Hobby Lobby possess religious convictions that once an egg is fertilized, it is a human life that deserves protection and therefore, when emergency contraceptives operate after fertilization, they constitute a kind of abortion. 12 Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751, 2766 (2014). Based on their convictions against providing insurance that assists this way of causing wrongful death, the company owners sought an exemption from the requirement of regulations issued under the Patient Protection and Affordable Care Act of 2010 (ACA), and as part of the 2011 Women’s Preventive Services Guidelines, which required employee insurance to cover such contraception, as well as all the other sixteen forms of contraception approved by the Food and Drug Administration (FDA). 13 Id. at 2765. The Supreme Court held that RFRA afforded the companies their claimed right. 14 Id. at 2785.
To reach that result, the Justices had to resolve at least three of four issues set by RFRA favorably for the companies: (1) They had to decide that the statutory protection includes closely held for-profit corporations, and does so even if the law from which the exemption is sought is designed to protect their workers, many of whom will not share the owners’ religious convictions; (2) the Justices also had to find the “substantial burden” on religious exercise, needed to qualify for a RFRA exemption; and (3) they further had to determine either that the government lacked a compelling interest behind its regulation or (4) had failed to employ an available “less restrictive means” to achieve its objective. 15 On the available means, the Justices focused on the Department of Health and Human Services’ delayed accommodation that already specified that religious nonprofit organizations did not need to provide such insurance directly—although insurance companies, either providing group insurance or engaged by third-party administrators for organizations giving self-insurance, did have to finance use of these contraceptives. Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751, 2781–82 (2014). That established a less restrictive means that could be extended to closely held for-profit corporations. Id.
Justice Alito wrote the majority opinion and Justice Kennedy added a relatively brief concurrence. Justice Alito’s opinion is an interesting mixture of rather broad propositions underlying a holding that is quite narrow. Especially since Justice Kennedy’s concurrence indicates that he might not accept giving an expansive scope of protection to for-profit corporations, one cannot easily foresee actual future outcomes, even prior to a change in the Court’s composition. What well may be even more important, and equally hard to predict, is the range of claims companies will make for exemptions, and how all this will affect political controversies about the appropriateness of special exemptions for those with religious convictions. The country has moved from a broad acceptance of religious exemptions to a perception by many that they tend to conflict with the rights of women and of gay individuals and couples; and recent proposals to enact state RFRAs have been claimed to reflect a hostility to equal treatment for same-sex marriage. 16 See Michael Paulson & Fernanda Santos, Religious Right in Arizona Cheers Bill Allowing Businesses to Refuse to Serve Gays, N.Y. Times (Feb. 21, 2014), http://www.nytimes.com/2014/02/22/us/religious-right-in-arizona-cheers-bill-allowing-businesses-to-refuse-to-serve-gays.html?_r=0 (on file with the Columbia Law Review) (discussing controversy around Arizona legislators’ attempt to enact state RFRA); see also Joshua Dorner, Ctr. for Am. Progress, Religious Liberty for Some or Religious Liberty for All? 6–7 (2013), https://cdn.americanprogress.org/wp-content/uploads/2013/12/ReligiousLiberty.pdf [http://perma.cc/9Lm3-8VF4] (discussing severe consequences of religious exemptions for women, same-sex individuals and couples, and families generally). But see Ryan T. Anderson & Leslie Ford, Protecting Religious Liberty in the State Marriage Debate, Backgrounder (Heritage Found., Washington, D.C.), Apr. 10, 2014, at 1–2, http://thf_media.s3.amazonaws.com/2014/pdf/BG2891.pdf (on file with the Columbia Law Review) (disputing idea that religious exemptions infringe on “sexual freedoms”). By extending the range of exemptions in a decidedly controversial setting, Hobby Lobby has intensified resistance to religious exemptions more broadly. 17 Paul Horwitz has offered an account of the breakdown of a relative consensus on religious liberty and exemptions in The Hobby Lobby Moment. Horwitz, supra note 4, at 166–84. Elizabeth Sepper has responded that the lack of consensus is specifically about corporate exemptions, particularly those that impose significant costs on third parties. Elizabeth Sepper, Reports of Accommodation’s Death Have Been Greatly Exaggerated, 128 Harv. L. Rev. Forum 24, 25–28 (2014), http://harvardlawreview.org/2014/11/reports-of-accommodations-death-have-been-greatly-exaggerated/ [http://perma.cc/9NSN-WUHZ].
Justice Alito’s opinion provides a striking example of a particular approach to statutory interpretation. Although quite acceptable in many straightforward cases, that approach is genuinely misguided in this particular setting. The approach is predominantly formalistic, treating each of the four issues separately as if they lack intrinsic connection. When resolving each issue in order, Justice Alito disregarded many considerations that could bear on overall application. A consequence is that RFRA is read to provide a clear answer that it does not actually contain. The opinion pays virtually no attention to whether its announced legal standards may entail sacrifices of workers’ legitimate interests and produce serious problems of administrability, leading to the granting of various claimed concessions that are really beyond what the statute covers.
Among the particular analytic problems with the majority opinion is its carving out of the crucial tax case of United States v. Lee. 18 455 U.S. 252 (1982). As explained later, the paying of taxes for a particular purpose to which one strongly objects bears much more resemblance to insurance coverage that may be used in part for behavior one regards as immoral than Justice Alito acknowledges.
As already noted, the particular controversy Hobby Lobby presented could fairly have been resolved either way. Taken alone, this isolated granting of an exemption seems acceptable, but the Court’s decision raises deep problems about general propositions and future applications. Although the dissenters may have been correct that RFRA should not have applied here, what this Essay urges is that the legitimate aspects that needed to be taken into account are considerably more complex and interrelated than the majority opinion indicates. Although readers can never be sure how far the somewhat artificial, legalistic reasoning in opinions represents the bases on which Justices actually reach their conclusions, 19 On this question, it may well be relevant that all of the five Justices in the majority, except Justice Kennedy, embraced Justice Scalia’s explicitly formalistic, text-centered approach in NLRB v. Canning, issued just days before Hobby Lobby. NLRB v. Noel Canning, 134 S. Ct. 2550 (2014). Justice Scalia’s concurring opinion in Canning strongly rejects Justice Breyer’s emphasis, in writing for the majority, on what will work effectively, and on historical practice, in determining whether a President can make recess appointments during intrasession recesses and for vacancies that occur prior to the recess but continue into it. Id. at 2592–618 (Scalia, J., concurring in the judgment). Because all the Justices agreed that appointments cannot be made during very brief intra-sessions recesses of the kind the actual case involved, Justice Scalia’s opinion, in sharp disagreement with Justice Breyer, was one concurring in the judgment. Id. at 2592 (Scalia, J., concurring in the judgment). identifying crucial flaws remains important. This Essay emphasizes what should have counted but was not reflected in Justice Alito’s opinion.
To be clear, most of this Essay’s criticisms do not cover typical issues of statutory interpretation, in which judges apply specific detailed language that clearly does or does not reach particular circumstances, thereby reflecting both what legislators intended and what knowledgeable readers would understand. Being vague and open-ended and built on prior free exercise law, RFRA is different. It is a kind of framework statute, requiring more complex reasoning and assessments of competing considerations. One cannot genuinely discern a clear original understanding of the statute’s application in the novel setting of this case. When determining outcomes and standards in such situations, judges should be guided substantially by what will work effectively, given the underlying values of the statute before them. That RFRA was designed to reinstate a prior constitutional approach provides a special reason for interpretation here to resemble that of the Constitution’s broad provisions.
This Essay first addresses some major concerns about the scope of RFRA. It then turns to application of each of the law’s standards. Part II examines RFRA’s application to closely held corporations. Part III addresses the question of whether requiring insurance coverage for employees to use the four contraceptive devices constituted a “substantial burden” on the owners’ exercise of religion. Because the Court found a substantial burden, Part IV analyzes Justice Alito’s consideration of whether the government has a “compelling interest” in requiring the contested insurance coverage. Finally, Part V examines the issue of whether denial of an exemption was the least restrictive means. For all these Parts, it is critical to keep in mind the central claim that each of the law’s standards should not have been treated in isolation.
I. Issues About the Scope of RFRA
The initial question in Hobby Lobby involved the scope of RFRA and whether it extended to closely held for-profit organizations. The law provides that the “[g]overnment shall not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability.” 20 42 U.S.C. § 2000bb-1 (2012). Strongly supported by a wide range of religious and civil liberties organizations, and adopted by a Senate vote of 97-3 and a House voice vote reflecting virtually no opposition, the law’s stated objective, as already noted, was to reestablish the free exercise law that Employment Division v. Smith had rejected. 21 Id. § 2000bb(a)(4)–(5), (b); see also Peter Steinfels, Clinton Signs Law Protecting Religious Practices, N.Y. Times (Nov. 17, 1993), http://www.nytimes.com/1993/11/17/us/clinton-signs-law-protecting-religious-practices.html (on file with the Columbia Law Review) (describing both “unusual coalition of liberal, conservative and religious groups that had pressed for the new law” and bill’s approval process). Contrary to earlier decisions, Smith had concluded that no valid constitutional claim can be made by those whose religious exercise is impaired by a general law not itself directed against religion. 22 See Emp’t Div., Dep’t of Human Res. of Oregon v. Smith, 494 U.S. 872, 881–82 (1990) (declining to extend constitutional protections to religiously motivated action in circumstances that do not involve “Free Exercise Clause in conjunction with other constitutional protections”). The Court thus rejected the free exercise claim of two members of the Native American Church who had used peyote as the central part of their worship services and were then fired for doing so. 23 Id. at 890. The two Native American Church members had been refused unemployment compensation because their firing was for violating a criminal statute not directed at religious practice. 24 Id. at 874. By enacting RFRA, Congress aimed to reinstate the First Amendment law that preceded Smith, which would have supported the claim in Smith unless the government had a “compelling interest” that could not be satisfied by a less restrictive means. According to the Senate Report, RFRA’s purpose was “only to overturn the Supreme Court’s decision in Smith.” 25 S. Rep. No. 103-111, at 2–3 (1993). Provisions of the statute itself say that its objective was to “restore the compelling interest test as set forth in Sherbert v. Verner and Wisconsin v. Yoder,” 26 42 U.S.C. § 2000bb(b)(1). and they characterize the “compelling interest test as set forth in prior Federal court rulings . . . [as] a workable test for striking sensible balances between religious liberty and competing prior governmental interests.” 27 Id. § 2000bb(a)(5).
In grasping the core grounds for both Employment Division v. Smith and Congress’s RFRA response, one question is critical: Whenever claims for exemptions are based on the magnitude and intensity of religious convictions, or conscience more generally, how can administrators and judges discern whether someone really meets the standards? The easiest circumstances are when no one has an incentive to make a claim insincerely. Also simple are claims by persons who are clearly aligned with a religious denomination that has settled standards, such as an Orthodox Jewish prisoner who refuses to eat pork. Harder situations arise when other people have unrelated reasons to avoid legal obligations and those making claims are not relying on standard church doctrines, as was true for many who claimed conscientious objections to military service. In such circumstances, the government may then attempt a serious inquiry into convictions or grant claims generally. Either approach allows some claimants who do not genuinely meet the basic standards to succeed. Smith reflected the view that these matters were too complicated to generate a constitutional right to exemption; RFRA was based on the contrary conclusion, at least for the kinds of situations courts had dealt with prior to Smith. What RFRA by itself does not tell us is exactly when, in light of these administrative difficulties, the value of conceding adherence to religious convictions is insufficiently great to warrant including entities as legitimate claimants. Nor does the statute make clear whether, when not excluding particular claimants, these difficulties should yield to judges applying “substantial burden,” “compelling interest,” and “least restrictive means” in a way that is deferential to government decisions, in order to avoid excessively broad coverage and serious problems of administrability.
As noted in the Introduction, after RFRA was enacted, the Supreme Court held it was invalid as applied to states, because Congress could not override the Court’s determination of the reach of constitutional provisions, 28 City of Boerne v. Flores, 521 U.S. 507, 536 (1997). although it was within the regulatory power of Congress to limit the effective coverage of federal laws. 29 Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 439 (2006). Consequently, many states have since adopted their own versions of RFRA. 30 See generally Christopher C. Lund, Religious Liberty after Gonzales: A Look at State RFRAs, 55 S.D. L. Rev. 466 (2010) (providing overview of state RFRAs).
An important feature of “compelling interest” in this context, although not explicitly acknowledged by the Supreme Court, needs to be recognized. 31 Although not actually acknowledging what follows, the Court has also not denied it. The “compelling interest” test was developed mainly in respect to laws that interfered with freedom of expression and disadvantaged racial minorities. In those contexts, only an extremely powerful interest allows a government restriction to survive; the use of the test has been close to a determination of invalidation. 32 Among the accounts of this use of the text are chapters 9 (Freedom of Speech and the Press) and 12 (Equal Protection) in Greenawalt, Interpreting, supra note 5, at 195–241, 336–68. How the test works for free exercise claims is treated in id. at 263–76 and in much more detail in 1 Kent Greenawalt, Religion and the Constitution: Free Exercise and Fairness, 201–32 (2006) [hereinafter Greenawalt, Religion and Constitution]. Despite some rhetorical formulations, such a high bar has not typically been employed for religious exemption claims. When people assert a constitutional right to an exemption from an otherwise valid statute, they seek special treatment not afforded to others; courts generally have not insisted on an overpowering government interest to reject the claim—a genuine substantial interest suffices. This was effectively sufficient in United States v. Lee, the 1982 case in which the Supreme Court rejected the claim of an Amish employer not to pay the Social Security Tax, and it was also reflected in a wide variety of lower court decisions dealing with religious claims. 33 United States v. Lee, 455 U.S. 252, 261 (1982) (“Congress and the courts have been sensitive to the needs flowing from the Free Exercise Clause, but every person cannot be shielded from all the burdens incident to exercising every aspect of the right to practice religious beliefs.”). This conclusion is supported with reference to a variety of cases in Greenawalt, Religion and Constitution, supra note 32, at 216–28 (examining cases in which courts were called on to balance interferences with religious exercise against degrees of government need, ranging from prisoners’ appearance to bankruptcy). Two other cases in which the Supreme Court rejected religious claims were Bob Jones University v. United States, 461 U.S. 574, 602–05 (1983), and Goldman v. Weinberger, 475 U.S. 503, 509–10 (1986). In the first of these, the Court did not allow a religious university that engaged in a form of racial discrimination to maintain its tax-exempt status as a charity. Bob Jones Univ., 461 U.S. at 605. Of course, the effort to combat racial discrimination could well count as compelling even in the strongest sense. In the second case, the Court held that a military rule against wearing headgear could apply even to an Orthodox Jewish psychologist at a mental health clinic. Goldman, 475 U.S. at 509–10. The military interest here seems far from compelling, but for this decision—made four years before Smith—most Justices were simply disinclined to sustain a religious claim to violate a general regulation. Since RFRA explicitly adopted the approach taken prior to Employment Division v. Smith and refers to the federal courts’ test as “striking sensible balances,” the statute did not seek to greatly increase the rigor of what the government needed to show. 34 42 U.S.C. § 2000bb(a)(5) (2012). Thus, its use of the “compelling interest” standard needs to be understood as a kind of intermediate scrutiny, more rigorous than “rational basis” but less than the demanding test used to invalidate laws effecting racial discrimination or interfering with core forms of protected speech.
In seeking to discern the scope of any statutory language, one may focus on legislative intent or reader understanding or both. In contrast to some decades ago, when large attention was given to legislative history as showing enactor intent, the Supreme Court’s main focus now is often on how a reader would understand statutory language. 35 See Greenawalt, Statutory, supra note 5, at 49–59 (describing shift in statutory interpretation from focus on legislative intent to textualism, especially since Justice Scalia joined Court). Despite huge complexities in exactly what “readers” should count, and the clear fact that with directive language, people understand what is communicated in terms of what they perceive writers or speakers as trying to convey, 36 There are, of course, various assertions about why no real intent of a legislative body is discernible from committee reports and statements in sessions. See supra note 5 and accompanying text (citing author’s work describing realist perspective’s uncertainty about opinions). most of the majority joining Alito’s opinion placed primary weight on readers, and not on enactors. 37 Cf. supra note 19 (discussing Justice Scalia’s opinion in NLRB v. Noel Canning, 134 S. Ct. 2550 (2014)). That is the gist of this opinion as well, although it does not explicitly discount the actual intent of members of Congress as irrelevant, and, given what it asserts, the distinction is not central to the decision.
This general account of the terms of RFRA, why it was enacted, and how both of these relate to earlier free exercise constitutional law, provide important bases for a fairly flexible interpretation in light of overarching values.
II. Application to Closely Held For-Profit Corporations
For Justice Alito, the primary coverage question is whether for-profit corporations count as “persons” within the statute. He relied partly on the broad precept that within the law, generally corporations count as persons, and partly on the Dictionary Act which provides that for acts of Congress, “unless the context indicates otherwise . . . ‘person’ . . . include[s] corporations, companies, associations, firms, partnerships, societies, and joint stock companies, as well as individuals.” 38 Burwell v. Hobby Lobby Stores, Inc., 134 S. Ct. 2751, 2768 (2014) (internal quotation marks omitted) (quoting Dictionary Act, 1 U.S.C. § 1 (2012)). Rejecting the counterargument that the Supreme Court had never sustained the free exercise right of a for-profit corporation prior to Smith, Justice Alito relied partly on Braunfeld v. Brown, 39 366 U.S. 599 (1961). although the Orthodox Jewish company that then sought an exception from a Sunday closing law was actually not incorporated. The Court’s rejection of that claim on the merits without expressing doubt about standing is taken as favoring the companies in Hobby Lobby. 40 Hobby Lobby, 134 S. Ct. at 2769–70 (“In Braunfeld, we entertained the free-exercise claims of individuals who were attempting to make a profit as retail merchants, and the Court never even hinted that this objective precluded their claims.” (citation omitted)).
To meet the contention that “context” in the Dictionary Act would here suggest an exception, Justice Alito responded both convincingly and unconvincingly. He is right that reading RFRA simply to preclude recovery by any entity that happens to be different from those winning cases prior to Smith would be inappropriate. When courts develop a settled interpretative approach to a broad constitutional or statutory provision, that does not foreclose appropriate coverage for a novel situation that differs from those previously resolved. The basic standards that have been articulated may be inconclusive or even point in favor of a new application. Justice Alito also said, unpersuasively, that the text of RFRA does not indicate a tie to “pre-Smith interpretation.” 41 Id. at 2772. In fact, although we do not know what the law entails for questions of coverage not clearly settled by pre-Smith cases, RFRA is clearly designed to reinstitute the interpretation that preceded Smith. 42 See City of Boerne v. Flores, 521 U.S. 507, 515–16 (1997) (discussing intent of RFRA).
Justice Alito further supported his position about the text’s coverage by noting one pre-Smith case 43 Hobby Lobby, 134 S. Ct. at 2772–73 (“The plurality opinion for [Gallagher] rejected the First Amendment claim on the merits based on the reasoning in Braunfeld, and reserved decision on the question whether the corporation had ‘standing’ to raise the claim.” (citing Gallagher v. Crown Kosher Super Market of Mass., Inc., 366 U.S. 617, 631 (1961))). in which a majority of Justices implicitly accepted the standing of a for-profit corporation. 44 Of course, if RFRA did not apply to corporations like Hobby Lobby, they would lack standing to raise a claim under it. The actual strength of this example is hardly clear. Three dissenters did vote that free exercise rights had actually been infringed, but a plurality of four specifically reserved the question of standing while rejecting the claim on the merits; two others rejecting the claim did not specify a reservation about standing. 45 Gallagher, 366 U.S. 617 (1961). For the separate opinion of Justice Frankfurter, released on the same day as Gallagher and joined by Justice Harlan, see McGowan v. Maryland, 366 U.S. 420, 459 (1961). For the dissenting opinion of Justice Brennan, released on the same day as Gallagher and joined by Justice Stewart, see Braunfeld, 366 U.S. at 610–16 (Brennan, J., dissenting). For Justice Douglas’s dissenting opinion, see McGowan, 366 U.S. at 561 (Douglas, J., dissenting).
Another argument Justice Alito made against limited coverage is the expansion that the Religious Land Use and Institutionalized Persons Act (RLUIPA), adopted seven years after RFRA, made to the definition of the “exercise of religion” in RFRA itself. 46 Hobby Lobby, 134 S. Ct. at 2761–62. The exercise of religion was made not to depend on “whether or not” it was “compelled by, or central to, a system of religious belief” and the measure was to be “construed in favor of a broad protection of religious exercise.” 42 U.S.C. §§ 2000cc-3(g), 2000cc-5(7)(A) (2012). Contrary to the dissent’s assumption that this change did not concern who could bring claims but merely made clear that claims might succeed without satisfying a centrality standard, 47 See Hobby Lobby, 134 S. Ct. at 2792 (Ginsburg, J., dissenting) (“RLUIPA’s alteration clarifies that courts should not question the centrality of a particular religious exercise.”). Alito contended that the amendment also supports an inclusive notion of “persons.” 48 He further rejected Justice Ginsburg’s assertion that the 2012 rejection by Congress of an amendment to broadly exempt claims of conscience from the Women’s Health Amendment, which had expanded the requirements of the Affordable Care Act, showed an intention to restrict what entities could bring claims. Id. at 2775 n.30 (majority opinion).
Alito further argued that, given the government’s concession that RFRA does apply to nonprofit organizations, it would not make sense to take “persons” as including some kinds of companies but not those incorporated “for-profit.” 49 Id. at 2768–69. He pointed out that many states now explicitly recognize hybrid corporate forms, according to which “for-profit” corporations may have other objectives apart from making money, some of which may actually reduce the likely amount of profit. 50 Id. at 2770–71. (“While it is certainly true that a central objective of for-profit corporations is to make money, modern corporate law does not require for-profit corporations to pursue profit at the expense of everything else, and many do not do so.”). Justice Alito’s reference to so-called “benefit” corporations in one respect actually cuts against his position. While many states have authorized corporations that may pursue objectives in addition to making profits, these corporations must contain a provision in their articles of incorporation stating that they are a benefit corporation. Plainly the corporations at issue in the case did not choose to be benefit corporations, which calls into question whether the inference Justice Alito wants to draw actually applies. Of course, the corporations may have been created prior to the enactment of the state statutes; the statutes do allow an existing corporation to convert to a benefit corporation, but that amendment requires at least a two-thirds vote, and a ninety percent vote in Delaware. See Brett H. McDonnell, Committing to Doing Good and Doing Well: Fiduciary Duty in Benefit Corporations 14 (Univ. of Minn. Law Sch., Minn. Legal Studies Research Paper No. 14-21, 2014), http://ssrn.com/abstract=2423346 (on file with the Columbia Law Review) (detailing state-specific legal rules and procedures regarding conversion to benefit corporation). At least in many circumstances, the failure of a corporation to amend its charter in this fashion could support an inference that a significant number of its shareholders did not want the corporation to combine profit and religion, risking a dilution of one or both. Responding to Justice Ginsburg’s argument that it does not make sense to think of “for-profit” corporations as actually exercising religion, 51 Hobby Lobby, 134 S. Ct. at 2793–94 (Ginsburg, J., dissenting) (“[T]he exercise of religion is characteristic of natural persons, not artificial legal entities.”). he urged that those who control closed corporations are human beings who are genuinely engaged in exercises of religion. 52 Id. at 2768 (majority opinion) (“An established body of law specifies the rights and obligations of the people . . . who are associated with a corporation in one way or another. When rights . . . are extended to corporations, the purpose is to protect the rights of these people.”). Justice Alito suggested that the corporation’s position is sheltered by the genuine religious beliefs of its controlling shareholders. But under standard corporate law principles, a controlling shareholder cannot cause the corporation to take action that benefits the controlling shareholder at the expense of the minority shareholders. If, because of his religious beliefs, a controlling shareholder is causing the corporation to decline to provide coverage in a way that could hurt the company economically, that could be seen as violating his fiduciary duty to the minority, absent a showing that the minority unanimously agrees with him (or the corporation is a benefit corporation and has declared its religious commitment, see supra note 50). To be sure, the controlling shareholder could take the position that not covering these contraceptive techniques is good business, perhaps because it will attract social conservatives to buy lots of things from the corporation, a position that if artfully presented would protect the controlling shareholder from liability under corporate law. However, establishing that position would require a factual showing that the shareholder in fact believed that the strategy would be profitable, a position that could undercut a claim that his religious beliefs are genuine.
Although Justice Alito wrote as if the answer to the coverage question is patently clear, in truth, if one asks what those voting for RFRA had in mind, the almost certain answer is that they did not think about this particular question. They were concerned with the Court’s denials of claims in very different contexts. 53 However, in an amicus brief, Douglas Laycock urged that a few years after RFRA was passed, debates over a nonadopted religious exemption bill indicated that RFRA was seen to cover for-profit corporations. Brief of Christian Legal Society, et al. as Amici Curiae Supporting Hobby Lobby and Conestoga Wood, et al. at 10–34, Hobby Lobby, 134 S. Ct. 2751 (No. 13-354), 2014 WL 411294. Professor Laycock’s basic position about coverage is challenged in James M. Oleske, Jr., Obamacare, RFRA, and the Perils of Legislative History, 67 Vand. L. Rev. En Banc 77, 82–87 (2014) http://www.vanderbiltlawreview.org/content/articles/2014/03/Oleske_Perils-of-Legislative-History.pdf [http://perma.cc/BN5D-UQ4Z] (arguing nonadopted religious exemption bill’s legislative history does not support coverage for for-profit corporations); defended in Douglas Laycock, Imaginary Contradictions: A Reply to Professor Oleske, 67 Vand. L. Rev. En Banc 89 (2014) http://www.vanderbiltlawreview.org/content/articles/2014/03/Laycock_Response.pdf [http://perma.cc/67PV-BYQV] (responding to Professor Oleske’s challenge to coverage for for-profit corporations); and challenged again in James M. Oleske, Jr., The Public Meaning of RFRA Versus Legislators’ Understanding of RLPA: A Response to Professor Laycock, 67 Vand. L. Rev. En Banc 125 (2014) , http://www.vanderbiltlawreview.org/content/articles/2014/05/Oleske-Response.pdf [http://perma.cc/RX9W-XGKC] (defending original challenge to Laycock’s position that RFRA covered for-profit corporations). And if one turns to reader understanding, one would have to imagine an incredibly sophisticated reader to reach the Alito resolution. An ordinary citizen simply reading RFRA could hardly jump to the Court’s understanding. Since few people think of business corporations as persons exercising religion, the relevant reader for Justice Alito would need first to understand the broader sense of “persons” within many legal circumstances, as well as the existence and content of the Dictionary Act. She would then need to overcome doubts raised by the special force of RFRA as a whole, and whether its inclusion of “a person’s exercise of religion” helps to create the “context” for an exception under the Dictionary Act. In other words, “the reader” would have to be exceptionally well informed or an imagined hypothetical expert to support Alito’s treatment of the statute’s language.
Addressing these questions carefully, one must acknowledge that neither legislative intent nor the understanding of even a very informed reader yields a decisive answer, both because prior law did not focus on nor resolve this particular issue, and because the “context” exception makes application of the Dictionary Act far from straightforward.
Given this conclusion, the issue of coverage really should come down to whether, given the objectives of RFRA and concerns about administrability, an extension to closely held for-profit corporations is sound. Aspects of that far from simple question could include whether the statutory language allows any distinction (1) between closely held corporations and other corporations, (2) between for-profit corporations and other organizations, and (3) between claims that concern government rules that are not aimed to protect existing or potential employees and those definitely designed to provide such protection. 54 Of course, laws protecting employees can play a role even when they do not present the fundamental issue at stake. Thus, Employment Division v. Smith concerned unemployment compensation, but the basic controversy was whether ingesting peyote as the center of a worship service could be criminal. 494 U.S. 872, 874 (1990). The actual text of RFRA does not contain any of these distinctions, but if “context” and likely difficulties of administration are considered, one or more of these ways of drawing a line could make sense.
On the issue of coverage for closely held companies, it is interesting how the Alito opinion itself treats likely claims by publicly held corporations such as IBM and General Electric. Continually emphasizing that the litigants involved here are closely held corporations in which particular families have complete control and possess religious convictions opposed to the statute’s mandate, Justice Alito said it “seems unlikely” that publicly traded corporations will bring RFRA claims, given the diversity of views of stockholders. He concluded that “we have no occasion in these cases to consider RFRA’s applicability to such companies.” 55 Hobby Lobby, 134 S. Ct. at 2774. However, imagine a company that lies somewhere between closely held corporations and those in which widely diverse shareholders hold a majority of stocks. One family, or a group of close religious associates could hold onto a bare majority of shares, making only a minority actually available for purchase by those with divergent views. The main stockholders could certainly feel that a requirement violates their religious convictions, and they might well be willing to reduce slightly the value of their stocks by not complying. For an ordinary for-profit corporation, minority shareholders may have a legal claim if the monetary value of their shares is sacrificed for other values. 56 See James D. Cox & Thomas Lee Hazen, Business Organizations Law § 14.15, at 437–38 (3d ed. 2011) (detailing various state decisions recognizing fiduciary duty of “utmost good faith and care” to minority shareholders in closely held corporations). However, that would not be true if those in control were adhering to one of the purposes of an established hybrid corporation. It might also not be true for an ordinary corporation if an initial contract with those buying stocks made clear other objectives besides profits. Given Justice Alito’s textual approach to “person,” it is very hard to see how RFRA will simultaneously include closely held for-profit corporations and exclude all publicly traded corporations, and religious exercise claims may well be made by those who control certain varieties of the latter.
The problem of different views among stockholders can, of course, arise even in closely held corporations when family members have competing views. In such circumstances, administrators and courts will need to decide who can speak for the corporation on the issue of religious conviction. This Part has indicated why it may not be at all simple to draw a line between closely held and other for-profit corporations and why the statute’s reaching of the former is far from obvious. This reality should have underlain an approach to this issue that took account of how well such applications would work in general in furthering the purposes of RFRA.
III. Substantial Burden
Having resolved that “person” within RFRA includes closely held for-profit corporations, Justice Alito turned to whether requiring insurance coverage for employees to use the four contraceptive devices constituted a “substantial burden” on the owners’ exercise of religion. 57 Hobby Lobby, 134 S. Ct. at 2775. None of the Justices doubted that the owners’ sincere religious convictions conceived these contraceptives as sometimes taking innocent life. Given that actual sincerity was not in question, how was one to determine if the burden was “substantial”? On this the Alito and Ginsburg opinions sharply disagreed; neither explored some nuances that could have made a difference in this respect. Among other things, Justice Alito discounted the conceivable relevance of United States v. Lee. 58 455 U.S. 252 (1982). In that predecessor to Smith, the Court, as Alito noted, did assume a substantial burden on religious exercise, relying instead on the absence of a less restrictive means to sustain the application of the Social Security tax law as it applied to an Amish employer. 59 Hobby Lobby, 134 S. Ct. at 2770. But one might see the Lee case, and other possible claims for tax exemptions, as actually supporting Justice Ginsburg’s dissenting argument that when a connection to a practice to which one objects becomes too attenuated, that should not count as a substantial burden. 60 Id. at 2797–99 (Ginsburg, J., dissenting). As developed below, the payment of taxes may be seen both as a connection that is not sufficiently direct and as somewhat similar to insurance coverage. 61 See infra notes 76–81 and accompanying text (presenting payment of taxes as analogy).
In discerning a substantial burden in Hobby Lobby, Justice Alito relied essentially on two factors—the owners’ sincerity that providing the insurance would seriously violate their religious beliefs and the powerful adverse economic consequences that a violation would impose on them. 62 134 S. Ct. at 2775. More specifically, if companies failed to include the required items in their insurance, they would be taxed $100 per day for each person; if they dropped insurance coverage altogether, they would be penalized $2000 per employee per year, as long as at least one employee was eligible for a subsidy on a government-run exchange. 63 The exact working of penalties is fairly complicated. See The Premium Tax Credit, IRS, http://www.irs.gov/Affordable-Care-Act/Individuals-and-Families/The-Premium-Tax-Credit [http://perma.cc/9L7W-XW8N] (last updated Sept. 29, 2015) (providing details of premium tax credit). 42 U.S.C. § 18071 (2012) details cost-sharing calculations and 26 U.S.C. § 36B(c)(2)(C) (2012) outlines what qualifies as an employer plan exempting the employee from eligibility for the premium tax credit. See Health Reform for Small Business: The Affordable Care Act Increases Choice and Saving Money for Small Businesses, The White House, http://www.whitehouse.gov/files/documents/health_reform_for_small_businesses.pdf [http://perma.cc/6U7Z-H7ZY] (last visited Oct. 1, 2015) (explaining new Employer Shared Responsibility provisions); Questions and Answers on Employer Shared Responsibility Provisions Under the Affordable Care Act, IRS, http://www.irs.gov/uac/Newsroom/Questions-and-Answers-on-Employer-Shared-Responsibility-Provisions-Under-the-Affordable-Care-Act [http://perma.cc/6L3J-94NJ] (last updated May 20, 2015) (same). In answer to the assertion by amici supporting the government that the penalty for dropping coverage completely would not be more expensive than providing it, 64 For a prominent statement of this claim, see Marty Lederman, Hobby Lobby Part III—There Is No “Employer Mandate”, Balkinization (Dec. 16, 2013), http://balkin.blogspot.com/2013/12/hobby-lobby-part-iiitheres-no-employer.html [http://perma.cc/S6RH-9RGE] (last updated Dec. 18, 2013). Justice Alito noted initially that the Court should not reach an empirical claim raised for the first time at this stage of the litigation, but he proceeded to announce that the argument was “unpersuasive.” 65 Hobby Lobby, 134 S. Ct. at 2776. His reason was that if a company failed to provide valuable insurance, it would need to raise wages to compensate, a raise for which, in contrast to insurance benefits, employees would have to pay taxes. 66 Id. at 2776–77. Although the exact economic consequence of the $2000 penalty could depend on the size of the company and whether at least one employee was eligible for a government subsidy, Justice Alito’s conclusion, that the overall economic consequences would generally be negative for companies that paid the penalty and compensated their workers for failing to provide insurance, makes sense.
A more fundamental problem concerning the practical consequences of a violation is why that should matter for whether the requirement itself poses a substantial burden. If compliance bothered someone only a little, his doing so would not involve a substantial burden no matter how severe were the penalties for violations. In past cases, the Court has assessed burdens in terms of the basic requirements, not the penalties. Justice Alito clearly assumed that the litigating companies would have accepted one of the prescribed penalties rather than actually providing the insurance. That willingness to suffer considerably helped demonstrate the strength and intensity of their convictions. This assumption raises a central issue about administrability and even people’s self-conceptions.
Justice Ginsburg argued that the Court’s “decision elides entirely the distinction between the sincerity of a challenger’s religious belief and the substantiality of the burden placed on the challenger.” 67 Id. at 2799 (Ginsburg, J., dissenting). For her, the latter is a legal matter, based on a more public appraisal of impairment. She referred to Bowen v. Roy, a case in which a father’s claim against the government using his child’s Social Security number was rejected because that use did not interfere with his own religious practices. 68 476 U.S. 693, 701–04 (1986). Without claiming that providing insurance coverage is exactly analogous, Justice Ginsburg concluded that the “requirement is too attenuated to rank as substantial.” 69 Hobby Lobby, 134 S. Ct. at 2799 (Ginsburg, J., dissenting). A company financing worker insurance does not decide whether a woman will use one of these contraceptives; that is left to her and her doctor, and the insurance company then pays for whichever product she chooses. The employer’s connection to actual use is simply too remote to amount to a substantial burden.
At one fundamental level, Justice Alito has the better of this disagreement. If a company owner believes, based on religious conviction, that use of a particular contraceptive device sometimes amounts to an abortion that constitutes the killing of innocent human life, he may conclude that he should not have to provide the device directly, any more than he should have actually to perform an act with that likely consequence. He may further think that if he provides money to someone knowing she will use it for that purpose, his involvement is still too great. And he may even believe that providing insurance coverage that some women will use in that way still keeps him so involved that he would rather suffer serious penalties. If he is convinced he would somehow be involved in taking innocent life, does that not constitute a substantial burden on his religious exercise, determined by his religious convictions?
What the Alito approach does not adequately take into account is the problem of administrability, partly illustrated by the comparison with paying taxes. On the general question of administrability, a helpful analogy of a different kind concerns the Roman Catholic belief that communion involves an actual transformation; wine, which Jesus used during the Last Supper in his reference to his body and blood, is regarded as a central element. Were a bar on the use of alcohol to make no exception whatsoever for Catholic Mass, it would be a substantial burden. Many Protestants believe communion is essentially symbolic, that it may be desirable to use wine, or to leave that choice to parishioners, but valid communion can take place with grape juice. Were a law to forbid any use of wine, and were this law actually enforced for the small amounts given at communion, these Protestants would use grape juice rather than suffer the penalties for illegal use or refrain from having communion altogether. We might well conclude that taking all this into account, an absolute, enforced bar on using wine would certainly impose a substantial burden on Catholic religious exercise, just as the bar on using peyote, upheld in Employment Division v. Smith, must be understood as such a burden on the religious exercise of members of the Native American Church. 70 494 U.S. 872, 903 (1990) (conceding “[t]here is no dispute that Oregon’s criminal prohibition of peyote places a severe burden on the ability of respondents to freely exercise their religion”). By contrast, the burden on the described Protestants would be less than substantial. But once an exemption was clearly established, how might the Protestants respond? If they thought it would be beneficial to use wine for all or some participants, would they not be tempted to claim that since communion is a central part of some of their worship services, the use of wine is very important? And on reflecting about whether they actually warrant an exemption, might they not persuade themselves that wine really is highly important and genuinely come to believe that the law really does impose a substantial burden? 71 During the Vietnam War, which many draftees perceived as unjustified, some applied for conscientious objector exemptions, which require such an objection to participating in “war in any form,” and many did actually arrive at this belief, although they would almost certainly not have in other circumstances, such as World War II. See David Malament, Selective Conscientious Objection and the Gillette Decision, in War and Moral Responsibility 159, 160 (Marshall Cohen et al. eds., 1974) (discussing universal versus selective conscientious objection to war in context of Vietnam War). Given these kinds of possibilities, how are outsider officials in a position to determine if a conviction is honest and really amounts to a substantial burden?
These concerns provide significant support for the argument that, given the need for exemptions that are administrable, the substantiality of a burden should be based partly on whether the connection between the actions of a claimant and the practices to which he objects is not too remote from a more general perspective. “Remoteness” here would need to rest on a determination that takes account of practical difficulties and more general public perceptions.
In supporting his conclusion about burden, Justice Alito unpersuasively relied upon Thomas v. Review Board of the Indiana Employment Security Division as “nearly identical.” 72 Hobby Lobby, 134 S. Ct. at 2778 (citing Thomas v. Review Bd. of the Ind. Emp’t Sec. Div., 450 U.S. 707 (1981)). In that case, the Court accepted a person’s refusal to participate in making turrets for tanks, even though he had previously helped manufacture steel that was used for weapons. 73 Thomas, 450 U.S. at 715. Rejecting the state court’s odd conclusion that Thomas’s choice was not really religious, Chief Justice Burger’s opinion did emphasize the importance of an individual’s own beliefs, but what the state court challenged was Thomas’s consistency and his fit with what other Jehovah’s Witnesses believed, not the connection of his work to a genuine objection to war. 74 See id. at 714–16 (discussing state court’s reasoning). In fact, part of the controversy was whether he could sensibly take a different perspective when his involvement became somewhat more direct, not more attenuated. 75 See id. at 715 (discussing lower court’s treatment of issue of directness of Thomas’s involvement in different scenarios). For these reasons, the Thomas decision does not actually provide powerful support for the Court’s resolution in Hobby Lobby.
This brings us to the relation to tax laws. Although recognizing that religious exercise was burdened in United States v. Lee, 76 See Hobby Lobby, 134 S. Ct. at 2770 (describing Lee’s holding as “compulsory participation in the social security system interferes with [Amish employers’] free exercise rights” (internal quotation marks omitted) (quoting United States v. Lee, 455 U.S. 252, 257 (1982))). Justice Alito decisively put aside compliance with tax laws as outside the realm of what RFRA covers. 77 See id. at 2784 (“Lee was a free-exercise, not a RFRA, case, but if the issue in Lee were analyzed under the RFRA framework, the fundamental point would be that there simply is no less restrictive alternative to the categorical requirement to pay taxes.”). It is interesting, if peripheral in this respect, that Chief Justice Roberts, who joined the Alito opinion, wrote for the majority in National Federation of Independent Business v. Sebelius that the requirement that individuals obtain health care insurance was justified under Congress’s power to tax. 132 S. Ct. 2566, 2594 (2012). By paying general taxes, individuals and companies provide indirect support for all sorts of activities, some of which the persons paying may find deeply objectionable on religious grounds. Whether one sees the reason as the absence of a substantial burden or the existence of a compelling interest with no less restrictive means, the Supreme Court has made clear that neither the First Amendment nor RFRA requires exemptions from tax payments. 78 See Hobby Lobby, 134 S. Ct. at 2784 (“Because of the enormous variety of government expenditures funded by tax dollars, allowing taxpayers to withhold a portion of their tax obligations on religious grounds would lead to chaos.”); Lee, 455 U.S. at 260 (“Because the broad public interest in maintaining a sound tax system is of such a high order, religious belief in conflict with the payment of taxes affords no basis for resisting the tax.”). Although a legislative exemption strategy could wisely allow those with strong objections to instead pay a higher amount of taxes which would then be used only for purposes they accept, 79 For a defense of this approach, see, e.g., Kent Greenawalt, Refusals of Conscience: What Are They and When Should They Be Accommodated?, 9 Ave Maria L. Rev. 47, 61 (2010) (suggesting government could “[l]et those opposed in conscience to paying certain taxes pay the amount owed plus an extra amount to some other valuable endeavor”). such a strategy has rarely been used and is definitely not required by any general law.
How different is insurance coverage from tax payments, especially if one considers the actual tax case in which a company’s claim was rejected? United States v. Lee involved a Social Security tax, not a general income tax. The Amish employer had only Amish workers, and the Amish did not believe in accepting Social Security from the government. 80 See Lee, 455 U.S. at 254–55 (noting appellee’s religious opposition to paying social security taxes). Were all the workers and their families to remain Amish, they would never take advantage of any Social Security benefits to which they were entitled. Given the employer’s religious conviction that acceptance of such benefits was wrong, and that his payment of the tax was to provide specifically for his workers being able to receive those benefits, his objection to payment does not seem very different from that involved in Hobby Lobby. It is true that taxes go into a general pool, whereas insurance coverage is for specific groups, but how much difference does that make for whether religious exercise is substantially burdened, or whether a less restrictive means exists? Very little, in fact. In brief, if one sees the tax cases as possibly about what should not amount to a “substantial burden,” independent of the individual religious convictions of particular claimants, adopting the Ginsburg approach to assess attenuation would be a genuine option.
Yet another option would be an intermediate approach. Judges might rely primarily on individual convictions when these are fairly obvious and do not extend too far in stretching connections between the degree of involvement and the practices to which the claimants object. However, if either individual convictions are very hard to determine or those of particular claimants extend beyond certain reasonable limits concerning connections, a more general sense of substantial burden would come into play. 81 For a discussion of this troubling issue in respect to what claims of conscience should be recognized, see Kent Greenawalt, Religious Toleration and Claims of Conscience, 28 J.L. & Pol. 91, 105–07 (2013) (“Requiring some closeness of connection to the act to which one objects can be an indirect way of assuring an employee’s basic sincerity and that his moral objection really rises to the intensity of conscience.”).
This Part has offered two basic criticisms of Justice Alito’s treatment of “substantial burden.” The opinion both fails to consider how its approach will work for a range of RFRA claims and effectively eliminates the burden requirement as a genuine limit on claims that can be made.
IV. Compelling Interest
Having found a substantial burden, Justice Alito turned to whether the government’s interest in requiring the contested insurance coverage was compelling. The opinion, after expressing some doubt, ends up with, “We find it unnecessary to adjudicate this issue,” instead assuming that the government’s interest in “guaranteeing cost-free access to the four challenged contraceptive methods” meets that standard. 82 Hobby Lobby, 134 S. Ct. at 2780. The opinion does counter any casting of the crucial interest in very broad terms such as “public health,” rather insisting that the “Government . . . demonstrate that the compelling interest test is satisfied through application of the challenged law ‘to the person’—the particular claimant.” 83 Id. at 2779 (internal quotation marks omitted) (quoting Gonzales v. O Centro Espirita Beneficente Uniao do Vegetal, 546 U.S. 418, 430–31 (2006)). This language taken from RFRA itself and from Gonzales v. O Centro, 84 546 U.S. at 430–31. which applied RFRA to the federal government, cannot be taken quite literally. The government will often lack a compelling interest in enforcing a restriction against one particular person or entity, so long as it reaches everyone else. Certainly that would have been true in United States v. Lee—Social Security taxes would hardly have been touched if Lee alone had failed to pay. One needs to understand this standard as at least positing that a single granted exception is not required if it will encourage multiple claims by others, thus undermining the enforcement of tax or other laws. More significantly, the government has a compelling interest in applying the law to the particular “person,” if non-enforcement against all others who are not distinguishable under RFRA would sacrifice that interest.
Justice Alito’s doubt about whether a compelling interest existed for the litigated cases relies on provisions allowing companies with fewer than fifty employees not to provide any health insurance and allowing companies with grandfathered plans to continue to provide some health insurance without meeting the act’s requirements. 85 See Hobby Lobby, 134 S. Ct. at 2780 (“As we have noted, many employees—those covered by grandfathered plans and those who work for employers with fewer than 50 employees—may have no contraceptive coverage without cost sharing at all.”). The opinion itself, though explaining the two bases for noncoverage, does not clearly indicate the difference between exactly what options are available to the two categories of companies. However, Justice Alito noted that the phasing out of grandfathered plans is not legally required. See id. at 2764 n.10 (“[T]here is no legal requirement that grandfathered plans ever be phased out.”). Justice Ginsburg’s counterargument is that specified exceptions do not eliminate a compelling interest and that the grandfathered exception was designed to give companies time to bring things up to date, not to be long-lasting in effect. 86 See id. at 2801 (Ginsburg, J., dissenting) (“Once specified changes are made, grandfathered status ceases.”). After noting his uncertainty, Justice Alito refrained from resolution about a compelling interest. 87 See id. at 2780 (majority opinion) (declining to adjudicate issue).
Justice Ginsburg argued powerfully that contraceptives covered by the debated insurance, notably intrauterine devices, are especially critical to protect women’s health and freedom of choice. 88 See id. at 2799–801 (Ginsburg, J., dissenting) (“[T]he corporations exclude intrauterine devices (IUDs), devices significantly more effective, and significantly more expensive than other contraceptive methods.”). Justice Kennedy, concurring, wrote that the “Department of Health and Human Services (HHS) makes the case that the mandate serves” a compelling interest, and goes on to “confirm” the “premise” of the Court’s opinion that such an interest is served. 89 Id. at 2785–86 (Kennedy, J., concurring). Since Kennedy’s language is about as close as he could get to stating explicitly that the compelling interest standard was satisfied, it appears that that at least five Justices definitely believed that it was, whatever were the doubts of Justice Alito and the remaining three Justices joining his opinion. The combination of the opinions of Justices Alito and Kennedy reveal that the presence or absence of a compelling interest was not central to the decision. In some other circumstances, that could matter for a RFRA claim. This Essay supports the view that that should be evaluated partly in terms of the strength of the competing religious claim and how workable its recognition will be.
V. Least Restrictive Means
Having assumed, without deciding, the existence of a compelling interest, Justice Alito turned to whether denial of an exemption was the least restrictive means. In general terms, Alito referred to least-restrictive means standard as “exceptionally demanding.” 90 Id. at 2780 (majority opinion). This phrasing lies in interesting contrast with the Court’s characterization of the compelling interest test in Cutter v. Wilkinson, 91 See 544 U.S. 709, 722 (2005) (“[A]n accommodation must be measured so that it does not override other significant interests.”). which applied RLUIPA to claims of federal prisoners, a case Justice Alito actually used to indicate that Congress was not bothered by difficulties of testing sincerity. 92 See Hobby Lobby, 134 S. Ct. at 2761–62 (citing Cutter in comparing scope of RFRA with that of RLUIPA). In Cutter, the Court stressed that context matters for the compelling interest test and that “due deference” was to be given to prison officials about maintaining order and safety. 93 544 U.S. at 723 (internal quotation marks omitted). Given that prison cases often concern whether prison officials will sacrifice an important interest if they accommodate a claim, “due deference” here obviously reaches what means will work, and is very far from any “exceptionally demanding” legal test.
As with other aspects, Justice Alito’s opinion throws out a rather broad possibility before resolving the case on a much narrower ground. The broad option is that the government could itself pay for the contraceptives without sacrificing a great deal financially. 94 See Hobby Lobby, 134 S. Ct. at 2780–81 (“The most straightforward way of doing this would be for the Government to assume the cost of providing the four contraceptives at issue to any omen who are unable to obtain them under their health-insurance policies due to their employers’ religious objections.”). The narrower ground relies on the fact that the Obama Administration had already granted an exemption for nonprofit religious organizations 95 See id. at 2781–82 (“HHS has already established an accommodation for nonprofit organizations with religious objections.”). —whatever their form of providing employee insurance, they need not pay for insurance for contraceptive devices whose use counters their religious convictions.
If one of these organizations has group health insurance, the company providing that insurance must pay for use of these contraceptive devices by those who are insured. The assumption has been that this will not increase the cost for the insuring company itself, even if it does not charge the religious organization more money to cover the contraceptives, because use of these contraceptives precludes future, more costly medical treatments for the unintended pregnancies and birth expenses of uncovered women who would not themselves pay for expensive intrauterine devices.
When it comes to self-insured organizations such as Notre Dame University, however, things work differently. 96 See id. at 2763 n.8 (discussing requirements in cases of self-insured religious organizations entitled to accommodation). A third-party administrator acquires the contraceptive insurance to which the organization objects from a company that administers Federal Facilitated Exchange Insurance. Having granted this insurance, that company is allowed to pay the government a reduced amount for the privilege of its participation in the federally facilitated exchange.
If this plan for independent coverage, in its two variations, will work for religious nonprofit organizations, presumably it can also do so for closely held for-profit corporations. Thus, it provides an effective less restrictive alternative to satisfy the government’s compelling interest (at least so long as the corporation’s owners do not believe that even the requirement that they cooperate by registering their objection in a certain form violates their religious convictions). 97 This related issue was sharply raised by the Court’s order, a few days after Hobby Lobby, not to require Wheaton College, a religious school, to submit the required exemption form until the merits of its claim against doing so are resolved. See Wheaton Coll. v. Burwell, 134 S. Ct. 2806, 2807 (2014) (declining to condition injunction on applicant’s use of specific form). Justice Sotomayor, joined by Justices Ginsburg and Kagan, strongly dissented from this ruling. Id at 2810 (Sotomayor, J., dissenting). As Justice Ginsburg noted in her dissent in Hobby Lobby, 134 S. Ct. at 2803 (Ginsburg, J., dissenting), it was not yet clear if the parties in Hobby Lobby would accept filing the required documents. Of course, if the particular form of the document is somehow objectionable, it may be that a less offensive form could satisfy the least restrictive means test.
In rejecting Justice Alito’s conclusion about a less restrictive alternative, Justice Ginsburg did not really show why the resolution in the narrow context of the case could not satisfy the government’s interest in insured use of the contraceptives. 98 She did indicate that women will have to take steps to learn about and sign up for coverage, but it is unclear just how this works and whether it is a real impediment. See Hobby Lobby, 134 S. Ct. at 2802 (Ginsburg, J., dissenting) (“Impeding women’s receipt of benefits ‘by requiring them to take steps to learn about, and to sign up for, a new [government funded and administered] health benefit’ was scarcely what Congress contemplated.”). In respect to administrability, the government has been assuming that religious organizations will not put forward a dishonest claim, instead accepting the mere submission of the proper notice as a basis for the exemption. Although one can imagine occasional incentives to advance insincere claims in this context, such as satisfying customers with a religious outlook one does not happen to share, any concern about false claims is not major.
The Court’s particular conclusion about a less restrictive means, persuasive though it may be in this specific circumstance, should lead to reflection on the question of government payments and costs, and on whether taxes are really so different in principle. Just how often an insurance company will actually save money by providing a benefit that the enterprises purchasing insurance for their workers have not included is far from clear. 99 See, e.g., Daniel Liebman, Demand Elasticity of Contraceptives and Their Cost-Effectiveness, a Follow-Up, Incidental Economist (July 11, 2014, 10:00 am), http://theincidentaleconomist.com/wordpress/demand-elasticity-of-contraceptives-and-their-cost-effectiveness-a-follow-up/ [http://perma.cc/E5L9-J22L] (noting contraceptive costs depend on type of method); Daniel Liebman, Does Contraceptive Coverage Pay for Itself? A Review of the Evidence., Incidental Economist (July 9, 2014, 6:55 am), http://theincidentaleconomist.com/wordpress/does-contraceptive-coverage-pay-for-itself-a-review-of-the-evidence/ [http://perma.cc/4KCM-CKSS] (noting contraceptive coverage might not be cost-saving from insurer’s perspective since many women will pay out of pocket for contraceptives). If providing an extra benefit, on balance, costs the insurance company money, it may need to raise its overall rates slightly or receive a governmental grant. If the companies raise rates for their customers generally to meet a RFRA need, that itself seems very close to the government imposing a modest tax to satisfy relevant religious convictions. At the same time, the companies with the religious objections may end up paying slightly less for their own worker insurance because they are not covering certain practices. 100 In respect to the organizations with group health insurance, it is unclear now whether they pay (1) less because of the devices not covered, (2) more because of increased risks of pregnancy, (3) the same, because these risks are reduced by the insurance company, or (4) the same, because there are set scales not responsive to such subtle variations in coverage. The Guttmacher Institute, in a 2003 report, claimed that “not covering contraceptives in employee health plans would cost employers 15–17% more than providing such coverage.” Cynthia Dailard, The Cost of Contraceptive Insurance Coverage, Guttmacher Report on Pub. Pol’y, Mar. 2003, at 12, 13. In other words, a cost would be shifted from those with the religious conviction to a more general public.
In fact, something very close to that actually takes place with the present scheme for self-insured religious organizations. Whatever the overall consequences of the provision of Federally Facilitated Exchange Insurance for workers using these contraceptive devices and their self-insuring employers, the independent company providing the contraceptive insurance does not directly save money, except by paying less to the federal government under its contract allowing it the special status for federally facilitated insurance. Although the government does not directly pay for the insurance itself, it does receive less money because the company provides it.
Given all this, perhaps it is not surprising that the Alito opinion flirts with the idea that government payment may itself amount to a relevant less restrictive alternative. 101 See Hobby Lobby, 134 S. Ct. at 2781–82 (“HHS itself has demonstrated that it has at its disposal an approach that is less restrictive than requiring employers to fund contraceptive methods that violate their religious beliefs.”). However, if this is so, and even if one focuses not on government payment but its receiving less money under a contract, why is taxation different? In United States v. Lee, 102 455 U.S. 252 (1982). had the Amish employer not had to pay Social Security taxes, the government might have had to pick up a modest bill down the road if some of the workers or their family members left the Amish denomination and were willing to accept government support. If the “less restrictive alternative” may sometimes be government payment, it is hard to see why either government payment to counter the negative consequences of a refusal to pay insurance, or a reduction in government income because a separate private company provides the insurance, is very different from government funding to compensate for a refusal to pay a tax that is imposed for a specific purpose, such as Social Security. Since Lee was decided prior to Smith and RFRA and Lee raised relatively few objections, we must assume that readers and enactors did not take the language of RFRA to indicate a contrary result, a point all the Hobby Lobby opinions take for granted. 103 Justice Ginsburg’s dissent notes that passages in Lee suggested that companies engaged in commercial activity should not be able to superimpose their religious views on statutory schemes or impose them on their employees. Hobby Lobby, 134 S. Ct. at 2803–04 (Ginsburg, J., dissenting).
The foregoing analysis indicates why what should have been crucial in Hobby Lobby was how its distinctive facts related to broader classes of situations. If the particular “less restrictive” means available here is not typically feasible more generally, perhaps it makes more sense simply to say that when what is involved are benefits for workers that cost money, for-profit companies, closely held or not, must comply with the law, unless the legislature chooses to grant a particular exemption. 104 Often, statutory provisions, such as Title VII, that allow some religious organizations to use religious criteria for employment are not taken to include typical for-profit corporations. See, e.g., Corp. of Presiding Bishop v. Amos, 483 U.S. 327, 340–46 (1987) (Brennan, J., concurring in the judgment) (noting why nonprofit organizations may be exempt); see also Fike v. United Methodist Children’s Home, Inc., 547 F. Supp. 286, 290 (E.D. Va. 1982), aff’d, 709 F.2d 284 (4th Cir. 1983) (holding United Methodist Children’s Home is “Methodist only in name” and not eligible for exemption). In other words, the general language of RFRA and the Free Exercise Clause would not then be taken to provide protection. One might reasonably say that if those in charge of a company choose incorporation—which not only entails nondiscrimination in hiring on religious grounds but also means that the individuals in control are not personally liable if the company suffers a financial disaster—it is appropriate to conclude that the owners sacrifice the ability to make RFRA claims for religious objections to requirements designed to protect their workers, many or most of whom will not share their religious outlooks.
On top of this concern is the genuine worry that once exemptions are granted, individuals and companies with what are actually weaker convictions may well be inclined to claim the same special treatment. There are thus strong practical reasons in support of concluding that RFRA does not protect for-profit corporations, closely held or not, from requirements designed to preclude disadvantages to present employees and discriminations of various sorts including hiring. 105 Interestingly, the result in Lee could be defended as helping to counter hiring discrimination in favor of Amish workers, since other workers would want and need Social Security protection.
This Part has urged strongly that what counts as a “less restrictive means” should not be limited to an isolated circumstance but include consideration of how that will work more generally. The more important point, however, is the general theme of this Essay: namely that this standard and the others need to be considered as related to each other, and assessed in terms of what will serve the values reflected in the statute.
If the more comprehensive and context-oriented analysis offered in this Essay had been employed, rather than application of single standards, such as the meaning of “a person’s exercise of religion,” the substantiality of a particular burden, or the presence of an unusual “less restrictive means,” the conclusion of coverage or noncoverage could have been reached only on the basis of a range of considerations. By contrast, the Alito opinion treats each standard as a distinct legal box not connected to the others in play. The difference between formalist treatment of individual segments and a more context-oriented approach to what Justices must decide represents a crucial variation in approach to open-ended statutes like RFRA.
This Essay does not address statutes with rather specific provisions that are clear in their implications for circumstances. There, the job of judges is to apply the law whether or not they agree with the legislative policy. But here, matters are much more complex. If one believes that taking everything relevant into account, RFRA coverage should not have been granted, one could think the best result would have been nonapplication of “a person’s exercise,” the absence of substantial burden, or a more relaxed less restrictive means approach that treated this like a tax case. Or one could rely on each of these grounds (as in Justice Ginsburg’s dissent). But whatever the stated textual basis, the true ground of resolution would rely on multiple factors that could bear on the proper coverage of all these three criteria of RFRA. As emphasized, that approach has special force for this particular law, which was designed to reintroduce a preexisting constitutional standard that was itself definitely being applied in a manner that was responsive to administrability in particular kinds of circumstances. 106 This assertion is supported by the range of cases addressed in Greenawalt, Religion and Constitution, supra note 32, at 216–28 (examining cases in which courts were called on to balance interferences with religious exercise against degrees of government need, ranging from prisoners’ appearance to bankruptcy); and using that kind of approach for many constitutional issues, including religious ones, is defended in a forthcoming book on constitutional interpretation, Greenawalt, Interpreting, supra note 5 at 3–104, 242–81 (defending general approach involving multiple interpretive criteria and applying approach to religion clause cases).
In the context of the Hobby Lobby issues and others in which administrability is a crucial factor, it would not be desirable for the Supreme Court, having arrived at a balanced assessment of the kinds of circumstances in which a statute should apply, to also set flexible legal standards that are open-ended for every application. Rather, the Court should use general considerations to set more specific criteria for who can make claims and in what circumstances. Such an analysis might well lead to the general conclusion that for-profit companies should not get religious exemptions under broadly worded statutes such as RFRA or when such exemptions will interfere with rights afforded to others or with serious concerns about practical or symbolic discrimination.
Although this Essay suggests such a conclusion, that is, of course, not its main theme. The main claim is that formalistic, section-by-section, reading is not really appropriate for this kind of statute. Instead, the Supreme Court should consider sections as related to each other and decide in terms of what will both serve the law’s objectives and be genuinely administrable.