The Yale Law Journal


RFRA and First Amendment Freedom of Expression

16 Mar 2016
Robert Post

I have very little expertise in the Religious Freedom Restoration Act (RFRA)1 or in the underlying constitutional law of freedom of religion that RFRA seeks to codify. I therefore venture into the debate surrounding Douglas NeJaime and Reva B. Siegel’s Conscience Wars: Complicity-Based Conscience Claims in Religion and Politics2 with some diffidence, and primarily to respond to Douglas Laycock’s specific argument that NeJaime and Siegel advance a position that is inconsistent with the First Amendment rights of freedom of expression of religious objectors.3

I. the controversy

NeJaime and Siegel’s article is an effort to understand how best to interpret and apply RFRA, which provides:

(a) In general. Government shall not substantially burden a person’s exercise of religion even if the burden results from a rule of general applicability, except as provided in subsection (b) of this section.

(b) Exception. Government may substantially burden a person’s exercise of religion only if it demonstrates that application of the burden to the person—

(1) is in furtherance of a compelling governmental interest; and

(2) is the least restrictive means of furthering that compelling governmental interest.4

NeJaime and Siegel argue that in certain circumstances granting RFRA exemptions from rules of general applicability based on objectors’ “complicity-based conscience claims” can pose special dangers of imposing “material and dignitary harms” on third parties.5 NeJaime and Siegel contend that this point is relevant to determining whether a rule of general applicability, when applied to a particular objector, serves a compelling governmental interest and is the least restrictive means of furthering that interest.

NeJaime and Siegel consider complicity-based claims in the context of antidiscrimination law. Because the government enacts such laws to protect individual and social interests that have both material and expressive dimensions, NeJaime and Siegel contend that

[i]f religious accommodation (1) would inflict material or dignitary harm on those the statute is designed to protect or (2) would produce effects and meanings that undermine the government’s society-wide objectives, this impact is evidence that unimpaired enforcement of the law is the least restrictive means of furthering the government’s interest.6

It is to this formulation that Laycock strenuously and primarily objects.

Laycock believes that NeJaime and Siegel’s conclusion is blatantly inconsistent with First Amendment rights of freedom of expression. At times he asserts that NeJaime and Siegel would deny religious accommodations because a claimant engages in political speech.7 But this is not what NeJaime and Siegel argue. They instead seek only to implement the literal provisions of RFRA; they ask whether granting a RFRA accommodation will in any particular case prevent a generally applicable statute from fulfilling the compelling governmental interests that underwrite it. In the context of antidiscrimination law, these interests can include values like equal citizenship and the prevention of dignitary harm.

Laycock seems to believe that protecting these governmental interests are inconsistent with First Amendment values. He asserts that if “preventing . . . ‘meanings’ and ‘values’” communicated by conscientious objection “is a compelling government interest,” then conscientious objectors would in effect “lose their right to conscientious objection” whenever they engaged “in a political argument.”8 His rhetoric is on this point is unsparing:

This is indefensible. Religious conservatives are absolutely entitled to argue for their views on the regulation of sex, however mistaken some of those views may be. And their exercise of that right is not a ground for forfeiting other rights they may have, including their right to religious exemptions.

. . . .

And if these acts of conscience are . . . a form of political speech, as NeJaime and Siegel argue, then the acts of conscience are doubly protected.

. . . The government cannot justify restrictions on discriminatory expressive conduct with the goal of “produc[ing] a society free of the corresponding biases.” That would be “a decidedly fatal objective.”9

Laycock’s argument turns on how First Amendment rights of freedom of expression intersect with claim for religious exemptions under RFRA. His argument raises complex and difficult issues about how communicative rights should be conceptualized in the context of RFRA. In analyzing his argument, I use the term “First Amendment rights” to refer exclusively to First Amendment rights of freedom of expression.

II. coverage versus protection

At the outset, we should distinguish First Amendment coverage from First Amendment protection.10First Amendment coverage exists whenever the constitutionality of a government action must be determined by the distinctive doctrinal tests of First Amendment jurisprudence. First Amendment protection, by contrast, refers to whether these doctrinal tests permit or invalidate that government action. It is quite common for a statute to trigger First Amendment coverage and yet to survive constitutional scrutiny.11 Most forms of government action, however, do not raise questions of First Amendment validity, and consequently their constitutionality is not determined by the application of doctrines specific to First Amendment jurisprudence. With regard to such actions, there is no First Amendment coverage.

The question of First Amendment coverage is theoretically fundamental: it forces us to decide whether we face an issue that should trigger distinct First Amendment tests and doctrines. First Amendment coverage can be triggered in either of two ways: by the nature of the behavior that government seeks to regulate or by the purpose that government regulation seeks to advance.12 I shall discuss each of these in order.

A. The Behavior Regulated by Government

Government regulation typically triggers First Amendment coverage if government seeks to control a speaker’s ability to participate in a medium for the communication of ideas.13 But government regulation typically does not trigger First Amendment coverage merely because someone engages in behavior that happens to be expressive.14

So, for example, if persons are prosecuted for defacing public property by graffiti, the proscribed behavior does not trigger First Amendment coverage even though it is expressive. If terrorists use their crimes to communicate political messages, they do not thereby acquire the right to a First Amendment defense when they are prosecuted. If a manufacturer sells a product with inadequate warnings, it cannot cite First Amendment precedents to defend against tort liability, even though the adequacy of its warnings may depend upon their expressive content. Nor can a landlord rely upon the First Amendment to defend against a prosecution for violating a rent control ordinance, even though the landlord’s infraction may turn on meanings contained in a lease. In each of these examples, First Amendment coverage is not triggered by expressive conduct.

NeJaime and Siegel orient their article around the recent case of Burwell v. Hobby Lobby Stores, Inc.15 In Hobby Lobby,the owners of a closely held corporation challenged Department of Health and Human Services (HHS) regulations requiring employers to offer employees health-insurance coverage for certain methods of contraception. The owners of the Hobby Lobby Stores, David and Barbara Green and their children, objected to covering “four FDA-approved contraceptives that may operate after the fertilization of an egg,”16 because they regarded the provision of such insurance as approving and enabling abortion.17

It is plain that HHS could enforce its regulations against the Green family without triggering First Amendment coverage. The Greens would have no plausible First Amendment defense against government sanctions for refusing to offer the mandated insurance coverage. It is irrelevant that the Greens regard the insurance as expressing approval for abortion, just as it is irrelevant to the prosecution of a terrorist that he believes that his crime expresses opposition to the American government. From the First Amendment perspective, the HHS regulations simply regulate conduct.

I expect that this same doctrinal conclusion would apply to most of the circumstances that are at issue in the debate between Laycock and NeJaime and Siegel. Requiring that bakers or landlords not discriminate against patrons is paradigmatically categorized as a simple regulation of conduct. It does not matter whether the owner of an inn regards leasing rooms as expressive; First Amendment coverage is not triggered by the application of a general rule requiring landlords to lease rooms without discrimination.18 First Amendment doctrine does not conceive these kinds of conduct regulations as preventing or inhibiting expression, political or otherwise.

Those subject to antidiscrimination laws are of course always free to make whatever political arguments they wish. This freedom is securely guaranteed by the First Amendment. But they are not free to engage in conduct the law otherwise prohibits. This also holds true for the Greens, who are free to engage in political advocacy whether or not the HHS regulations are enforced. From the perspective of the First Amendment, the HHS regulations have nothing to do with political advocacy. That is why the Greens would not have a First Amendment defense were they to refuse to offer the mandated insurance coverage.

One can make this point more concretely. Imagine a case in which government seeks to regulate participation in a medium for the communication of ideas. It prohibits publishing magazines without a license. Imagine that a person seeks to publish a religious magazine that is essential to the practice of her religion. She could challenge the proposed government ordinance under either the First Amendment or RFRA. Because there are many advantages to claiming constitutional rather than statutory rights, we can expect her to invoke RFRA only if her First Amendment claims are weak. In the hypothetical case about a religious magazine, it would be astonishing if a person even bothered to invoke RFRA; because the government is regulating an expressive medium, there is an underlying First Amendment freedom at play. First Amendment coverage is obviously triggered. Conversely, in cases like Hobby Lobby, RFRA alone is involved precisely because there is no underlying First Amendment right to prevent the federal government from requiring employers to insure their employees.

It is therefore difficult to make sense of Laycock’s rhetoric. He is of course completely right to assert that “[r]eligious conservatives are constitutionally entitled to argue for their views on the regulation of sex.”19 But that assertion does not seem to have anything to do with the question before us, which is whether RFRA should be interpreted to apply to conduct that is otherwise without First Amendment coverage. The HHS regulations do not prevent anyone from arguing about anything. The RFRA question arises only becausethe conduct at stake is precisely not “a form of political speech.”20 If it were a form of political speech, there would be a far more direct way to provide legal protection than RFRA.

Laycock’s invocation of Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc.,21 is thus mysterious. Hurley involved the application of a Massachusetts antidiscrimination statute to a parade, which the Court regarded (and has traditionally regarded) as a “medium for the communication of ideas,”22 or, as Justice Souter put it, “a form of expression, not just motion.”23 First Amendment coverage is typically triggered whenever the state seeks to regulate media for the communication of ideas.24

Hurley explicitly contrasts statutes that regulate media for the communication of ideas with antidiscrimination laws that regulate only conduct. Hurley affirms the constitutionality of antidiscrimination laws that prohibit “the act of discriminating against individuals in the provision of publicly available goods, privileges, and services on the proscribed grounds.”25 Such laws are “well within the State’s usual power to enact when a legislature has reason to believe that a given group is the target of discrimination, and they do not, as a general matter, violate the First or Fourteenth Amendments.”26 The applications of RFRA that NeJaime and Siegel consider, and that Laycock discusses,27 are analogous. They involve the regulation of ordinary forms of conduct, rather than of media for the communication of ideas.

Laycock’s appeal to Hurley seems decidedly misplaced, as a fortiori does his appeal to the unconstitutional conditions doctrine.28 NeJaime and Siegel’s interpretation of RFRA does not require the waiver of any First Amendment right.

B. The Purpose of Government Regulation

We might rehabilitate Laycock’s First Amendment argument if we remember that First Amendment coverage can be triggered when the state acts for constitutionally suspicious reasons, regardless of whether it seeks to regulate media for the communication of ideas.29 Thus the famous R.A.V. case held that a statute that proscribes only fighting words, which are without First Amendment coverage, may nevertheless be vulnerable to constitutional challenge if its purpose is to exercise “hostility—or favoritism—towards the underlying message expressed.”30 Laycock appears to interpret NeJaime and Siegel as reading into RFRA just such hostility toward the beliefs of religious persons. Laycock may believe that NeJaime and Siegel run afoul of the First Amendment because they seek to apply RFRA in a way that advances constitutionally improper purposes.

Determining which government purposes trigger First Amendment coverage often raises profound and complex questions. But this is not the case with respect to NeJaime and Siegel’s argument. Their thesis, in a nutshell, amounts to nothing more than a restatement of RFRA’s explicit provisions, which direct courts not to grant religious exemptions from generally applicable statutes insofar as such statutes serve compelling interests and insofar as granting exemptions would undermine those interests. They believe that RFRA should be interpreted in light of the interests served by the general statute to which RFRA applies.

In NeJaime and Siegel’s argument, therefore, as in RFRA itself, there is a constitutional identity between the purposes served by RFRA and the purposes served by the statutes of general applicability to which RFRA applies. It follows that if the interests served by a generally applicable statute do not trigger First Amendment coverage, interpreting RFRA to advance those same interests should also not trigger First Amendment coverage. Someone who wishes to argue that interpreting RFRA to serve the interests of a generally applicable statute should trigger First Amendment coverage must be committed to the proposition that First Amendment coverage should also triggered by the application of the statute itself.

This symmetry has important implications for RFRA in the context of the antidiscrimination laws that lie at the heart of NeJaime and Siegel’s article. A fundamental purpose of antidiscrimination law is to prevent “the deprivation of personal dignity that surely accompanies denials of equal access to public establishments.”31 Because the law commonly conceptualizes the dignity of persons as dependent upon how they are regarded by others,32 legal efforts to uphold dignity typically have the purpose and effect of regulating conduct that transmits messages of disrespect. That is why antidiscrimination law characteristically prohibits conduct that creates social meanings associated with the stigmatization or stereotyping of protected groups.

Consider, for example, how the Court understands the purpose of a Minnesota antidiscrimination statute:

By prohibiting gender discrimination in places of public accommodation, the Minnesota Act protects the State’s citizenry from a number of serious social and personal harms. In the context of reviewing state actions under the Equal Protection Clause, this Court has frequently noted that discrimination based on archaic and overbroad assumptions about the relative needs and capacities of the sexes forces individuals to labor under stereotypical notions that often bear no relationship to their actual abilities. It thereby both deprives persons of their individual dignityand denies society the benefits of wide participation in political, economic, and cultural life. These concerns are strongly implicated with respect to gender discrimination in the allocation of publicly available goods and services.33

This point can be generalized. Whenever law seeks to alter the social status of a group, it will regulate conduct that transmits meanings inconsistent with the desired status. Consider, for example, Chief Justice Rehnquist’s opinion for the Court in Nevada Department of Human Resources v. Hibbs, which upheld the Family and Medical Leave Act (FMLA).34 The FMLA was not on its face an antidiscrimination law, yet its purpose was to alter the status of women. The Chief Justice applauded the efforts of Congress to transform

“[t]he prevailing ideology about women’s roles [that] has in turn justified discrimination against women when they are mothers or mothers-to-be.” Stereotypes about women’s domestic roles are reinforced by parallel stereotypes presuming a lack of domestic responsibilities for men. Because employers continued to regard the family as the woman’s domain, they often denied men similar accommodations or discouraged them from taking leave. These mutually reinforcing stereotypes created a self-fulfilling cycle of discrimination that forced women to continue to assume the role of primary family caregiver, and fostered employers’ stereotypical views about women’s commitment to work and their value as employees. Those perceptions, in turn, Congress reasoned, lead to subtle discrimination that may be difficult to detect on a case-by-case basis. We believe that Congress’ chosen remedy, the family-care leave provision of the FMLA, is “congruent and proportional to the targeted violation.”35

The regulation of conduct to prevent the transmission of social meaning is so pervasive that the Court has interpreted even the Equal Protection Clause itself as protecting “the dignity and worth” of individuals,36 and hence, as the Court explained in JEB v. Alabama ex rel. T.B.,37 as regulating behavior conveying “discriminatory and stereotypical presumptions that reflect and reinforce patterns of historical discrimination”:

Striking individual jurors on the assumption that they hold particular views simply because of their gender is “practically a brand upon them, affixed by the law, an assertion of their inferiority.” It denigrates the dignity of the excluded juror, and, for a woman, reinvokes a history of exclusion from political participation. The message it sends to all those in the courtroom, and all those who may later learn of the discriminatory act, is that certain individuals, for no reason other than gender, are presumed unqualified by state actors to decide important questions upon which reasonable persons could disagree.38

NeJaime and Siegel believe that in applying RFRA to laws that seek to alter the social status of groups, it is relevant whether granting a religious exemption would “inflict material or dignitary harm on those the statute is designed to protect or . . . would produce effects and meanings that undermine the government’s society-wide objectives.”39 This is merely to argue that RFRA should be interpreted to allow exemptions from laws only if such exemptions do not defeat the purpose of such laws. It is simply to restate the explicit terms of RFRA.40 It follows that if Laycock believes that interpreting RFRA in this way triggers First Amendment coverage, then he must also believe that First Amendment scrutiny is triggered by statutes that seek to transform the social status of groups.

This is a highly ambitious but most implausible claim. It would suggest that our entire tradition of antidiscrimination law is suspect under the First Amendment. Sometimes Laycock seems to embrace this wide-ranging claim, as when he states that “The Court has already closed the door on ‘social meaning’ as a compelling government interest.”41 In the context of antidiscrimination law, Laycock’s assertion is plainly and manifestly false, especially so when his authority for this startling proposition is Hurley. And yet at other times Laycock seems to reject such a radical position, as when he argues “I do not claim that civil-rights laws generally violate the Free Speech Clause, as Robert Post seems to think. It is NeJaime and Siegel, not me, who say that these claims of conscientious objection are a form of political speech in support of the conservative religious position on the underlying issues.”42

The logical difficulties that beset Laycock’s argument suggest how cautious we must be in interpreting a case like R.A.V. Almost all human behavior is saturated with meaning, and for that reason law frequently seeks to regulate conduct precisely because of its meaning.43 Criminal law illustrates the point.44 Antidiscrimination law is not exceptional in this regard. If we take a decision like R.A.V.too literally, and if we seek to strike down laws because they seek to repress certain (harmful) messages communicated by acts that lie outside media for the communication of ideas, we endanger large swaths of our existing legal system.


The great tradition of American First Amendment jurisprudence did not arise from striking down laws that sought to regulate conduct because of its social meaning. It instead focused on laws that sought to regulate participation in media for the communication of ideas. Without denying the validity of a case like R.A.V.,45 there is a deep puzzle about how exactly to generalize its holding.46 I have elsewhere advocated that we should do so through forms of reflective equilibrium,47 by inquiring carefully into the consequences of our formulation of First Amendment principles and their compatibility with fundamental constitutional and social commitments. It should count powerfully and persuasively against any abstract or rigid formulation of First Amendment principles that it would invalidate our entire tradition of antidiscrimination law.

I acknowledge that such modesty is in tension with the First Amendment literalism that seems recently to have gripped those with a strong deregulatory agenda.48 But I would be very surprised to learn that Laycock believes antidiscrimination law should be constitutionally invalidated because it seeks to maintain human dignity by regulating conduct that promulgates harmful meanings stigmatizing to protected groups. Were Laycock to embrace such an extreme position, he would need to offer a great deal more justification than what is presently contained in his Response. Without it, he is not yet entitled to the indignant tone in which he couches his First Amendment contentions.

How broadly or narrowly we should apply RFRA is a different question on which I take no position. I am concerned only with Laycock’s argument that NeJaime and Siegel’s thesis is unconstitutional under the First Amendment. And the precise answer to Laycock’s argument is that NeJaime and Siegel’s thesis is neither more nor less unconstitutional than the antidiscrimination statutes to which RFRA is being applied. If the interests served by antidiscrimination law do not trigger First Amendment coverage, then neither should interpreting RFRA to effectuate these interests. Conversely, if Laycock truly considers NeJaime and Siegel’s thesis to be “indefensible,” then so must he consider the many antidiscrimination statutes that are now subject to RFRA’s knife.

This same logic applies to the HHS mandate at issue in Hobby Lobby. The mandate serves a number of possible interests. If the primary interest is to protect the health of women, then it is difficult to understand how any First Amendment issue can arise from interpreting RFRA to advance this interest. RFRA states that it should not be interpreted to authorize exemptions that undermine the achievement of a compelling purpose like women’s health.

If the HHS mandate is instead understood to serve the purpose of altering “stereotypical views” about women’s reproductive functions, the mandate must be conceptualized as transformative in the same way that Hibbs interpreted the FMLA to be. If Laycock’s real argument is that interpreting RFRA to advance such a transformative social purpose ought to trigger First Amendment scrutiny, then his true disagreement lies with Hibbs and its approval of statutes like the FMLA. His true disagreement lies with a main purpose of our traditional antidiscrimination law.

Robert Post is Dean and Sol & Lillian Goldman Professor of Law at Yale Law School.

Preferred Citation: Robert Post, RFRA and First Amendment Freedom of Expression, 125 Yale L.J. F. 387 (2015),