Liberalism and the Distinctiveness of Religious Belief

35 Const. Comment. 207

Abner S. Greene[1]


Every person reading this Review will have pondered, and perhaps resolved, his or her religious identity. Some are devout, and their relationship with and faith in God—in a higher power, an extrahuman source of generative and normative authority—is of central importance to who they are as human beings. Others are still theists, of a sort, but their religiosity is more backgrounded in their everyday lives. And yet others are agnostic—open to theistic belief but not yet convinced—or atheist, and denying God’s existence. In the United States, today, these religious differences mostly do not lead to significant conflict. But because the devout, the mildly religious, the agnostic, and the atheist usually coexist in where they live and work, some conflict based in religious belief and difference is inevitable. And when the government is involved, such conflicts often take on a constitutional dimension. To what extent may, or must, the state acknowledge God’s existence and help theists in their quest to have their religious beliefs be central to their existence? To what extent may, or must, the state adjust its laws to help those for whom God’s mandates should take precedence? For both types of question—what we under our Constitution sort under the Establishment and Free Exercise Clauses of the First Amendment—how do we best preserve the religious liberty of those for whom God’s will competes with the state’s and those for whom this is not the case, and who might be harmed by state action aiding the devout?

When they arise, the conflicts are front-page news items. Consider these, all cases that ended up at the United States Supreme Court—Should the state be permitted to maintain a 32-foot-high Latin cross on a large pedestal at a public highway intersection?[2] Should the state be permitted to use taxpayer dollars to fund private secular but not religious schools?[3] Must the state exempt, from public accommodations anti-discrimination law, a devoutly Christian baker who won’t make a custom cake for a same-sex wedding celebration?[4] Must the state permit religious employers to opt out of providing the portion of group health insurance that includes contraception, because the employer deems the contraception an abortifacient?[5]

Constitutional and statutory jurisprudence in the United States has sought a middle ground between permitting the state to achieve secular, sometimes liberal (as in “left of center”) political ends, and ensuring robust religious liberty. In Liberalism’s Religion, a work of political theory with important intersections with U.S. law and religion jurisprudence, Cécile Laborde confronts the establishment and free exercise dilemmas that arise in a religiously heterogeneous liberal democracy. She offers a middle-ground solution, permitting a small swath for state acknowledgment of religion, and requiring a somewhat larger space for state exemptions for religion from generally applicable law. But she does this with barely a mention of God or theism. This is because, for Laborde, the solution to the dilemmas a religiously integrated liberal democracy face is, strikingly, to drop religion out of the picture. I exaggerate, but only a bit. Laborde’s thesis is that “we should disaggregate religion into a plurality of different interpretive dimensions” (p. 2); she claims that “religion is not uniquely special: whatever treatment it receives from the law, it receives in virtue of features that it shares with nonreligious beliefs, conceptions, and identities” (p. 3). She casts her lot with scholars such as Chris Eisgruber, Larry Sager, Micah Schwartzman, and Nelson Tebbe, describing her approach as “liberal egalitarianism”[6] (p. 4; hereinafter “LE”). Thus, on the nonestablishment side of things, the state should (for the most part) not endorse or establish religion, but “only because it does not establish or endorse any conception of the good in general” (p. 5). For example, Laborde contends, “[a] state that enforces a secular comprehensive conception of the good—Rawls’s favorite example was a philosophy of Kantian autonomy, but we could think of other comprehensive world-views, such as ecocentrism—would fall foul of liberal legitimacy on exactly the same ground as would a comprehensively religious state” (p. 145). And on the free exercise side of things, the state should (sometimes) protect religion, but “only as one of the ways in which citizens live a life they think good” (p. 5). For example, Laborde says the following situations that might give rise to exemption claims should at least get to a balancing of individual versus state interest (and not be excluded at the get-go), whether based in religion or other deep commitments: “[a] parent sincerely believes that strict discipline—including justly administered mild corporal punishment—serves the moral edification of her child”; and “[a] bakery owner . . . does not mind serving a gay customer but objects to writing a pro-same-sex marriage slogan on the cake that the customer wishes to purchase” (p. 211). In addition to casting her lot with the LEs, Laborde describes a separate group of scholars, such as Stanley Fish and Steven Smith, as “critical religion theorists” (p. 14), focusing on their claims that there is no stable, neutral governmental approach that does not involve the state’s staking its own positions on matters of religion.[7]

But Liberalism’s Religion devotes almost no space to an important third position, which contends that religious belief and practice are distinctive and deserving of distinctive legal treatment. This “religion as distinctive” camp—of which I am a member[8]—begins with recognizing that for many religious people, God exists and their faith in and relationship to God is front and center in their lives. For many religious people, belief in God and what follows from that is not comparable to anything, and cannot properly be disaggregated into just another set of beliefs and practices. For the most devout religious people, God’s being extrahuman is at the core of their lives and their devotion. How can law take seriously that beliefs and practices not based in commonly shared material-scientific fact animate the lives of many of our fellow citizens?[9]

One goal in this Review is to examine whether Laborde’s LE is a defensible approach to the role of religion in the liberal state. At the same time, I will examine how well her approach fits with the U.S. constitutional settlement regarding law and religion. Although her book is not meant to be primarily an analysis of U.S. law, Laborde relies on the work of important U.S. constitutional scholars, who are seeking both a political theoretic and constitutional answer to questions of nonestablishment and free exercise. I will claim that a different kind of egalitarianism is the better answer from both political theory and constitutional law: an egalitarianism that does not disaggregate religion and approach it similarly to other beliefs and practices, but rather that takes seriously religion’s distinctiveness—its basis in one’s relationship to God—and thus that treats religious people and institutions as fully equal participants in our liberal democratic project. I will also challenge a predicate to Laborde’s argument, that the sovereign state in a liberal democracy has legitimate authority to draw jurisdictional lines between church and state and thus justifiably resolve the difficult establishment and free exercise issues that arise. My objection to Laborde on this question of authority isn’t that the state has no role in drawing church-state lines, and isn’t that the state is never justified in so doing; rather, my position is that there is no general argument that backs the state in applying all laws to all persons and institutions. We should see questions of political obligation (is there a moral duty to obey the law?) and political legitimacy (is the state justified in demanding our legal obedience?) as correlative; we should appreciate that there is no valid, general, affirmative answer to these questions; and thus, the state must earn its stripes, as it were, law by law or case by case. This understanding of political obligation and legitimacy undergirds one way of arguing for exemptions.[10]

Part I of this Review will describe Laborde’s argument. Part II will offer two critiques of her position—that she is defending the wrong kind of egalitarianism for the religion and state settlement we have reached in the U.S. and that would be best for any liberal democracy; and that her “jurisdictional” argument backing the wholesale legitimacy of the liberal state’s role in policing the religion-state settlement cannot be properly sustained.

  1. Laborde’s Case

Part I of Liberalism’s Religion (“Analogizing Religion”) begins with a chapter containing LE’s response to three challenges offered by the critical religion theorists. These challenges are (a) semantic—that liberalism has not offered a “stable, universally valid empirical referent for the category of religion” (p. 18), (b) protestant—that “liberal law is based toward individualistic, belief-based religions” (p. 21), and (c) realist—that the liberal state’s treatment of religion is either “the naked exercise of arbitrary power” (p. 24) and/or that it establishes liberalism as a kind of religion. Laborde’s answer to the semantic challenge is centered in the anti-religious-distinctiveness position taken by the LEs: The state may sometimes protect religious belief and practice, but only because of qualities shared by secular belief and practice; therefore, the LE needn’t get embroiled in difficult questions of defining what is and isn’t “religion.” Her answer to the protestant challenge is to reject the critical religion theorists’ premise: In a properly functioning liberal democracy, majoritarian power and prejudice isn’t reducible to a bias favoring “belief-based, voluntarily chosen religious practices” (p. 33); rather, law protects a “right not to be coerced into changing or abandoning the beliefs or way of life that one in fact has,” including both those who “choose” and those who are “called” (p. 34). As for the realist challenge, Laborde makes the easy point that there are defensible and indefensible exercises of state power. Of greater importance and connection to her LE position, Laborde claims that LE “is not . . . grounded in any comprehensive metaphysical, ontological, or ethical doctrine” (p. 40). Thus, although there’s no such thing as the state’s governing in a value-free way, the critical religion theorists are wrong in thinking the liberal state is establishing an official religion of secularism (or anything else); rather, “liberalism is based on the idea that all individuals should enjoy as much freedom as is compatible with the freedoms of others” (p. 38). This follows from Laborde’s discussion of classic Millian harm theory, and is not meant to prefer the autonomous life over other ways of life.

The next two chapters of Part I describe possible approaches to what in constitutional law we would call Free Exercise Exemptions issues and Establishment Clause issues. Laborde sets forth approaches from different liberal theorists, with some support and some critique for each, all from her LE perspective (which she then cashes out with an affirmative case in Part II). Chapter 2 is about exemptions. Laborde begins with a description and critique of what she calls “dissolving religion” (p. 44), focusing on Ronald Dworkin’s take on exemptions in his final work, Religion Without God.[11] On the one hand, Dworkin claims freedom of religion is a general and not special right. This means we shouldn’t apply elevated scrutiny to protect it; rather, the state “must not appeal to the superiority of any way of life over another” (p. 45), religious or otherwise. In this way, freedom of religion is “an instantiation of a more general right of ethical independence” (p. 46), and is an example of appropriate limits on the state’s reasons for action. Carried to its extreme, this focus on the government’s reasons, and not on possible disparate impact, looks like Brian Barry’s no-holds-barred approach: If the state has defensible reasons for a law of general applicability, then there’s not a good case for exemptions.[12] But, as Laborde shows, Dworkin was in fact concerned with laws that disparately affect what a group might regard as “sacred” (p. 50)—although for Dworkin this may be theistic or not theistic. This expanded view of the sacred as religious leads to a potentially expansive view of exemptions and to a view that forbids laws from being justified on grounds that take positions on what counts as sacred. This in turn requires some understanding of religious distinctiveness, even if not limited to a God-centered understanding.

Laborde turns next to a view she dubs “mainstreaming religion” (p. 50), covering a position advanced by Eisgruber and Sager, which would protect religious interests comparably to similarly serious nonreligious interests, against intentional or unintentional state discrimination against vulnerable/minority groups.[13] Exemptions are sometimes warranted, but not because of anything distinctive about religious belief. In the U.S. constitutional law community, Eisgruber and Sager are leaders in the “religion as non-distinctive” school; they are LEs, so Laborde’s critiques are internal to LEs. Her main point is that Eisgruber and Sager “have not settled for a single criterion of comparability between religion and nonreligion and instead oscillate between normatively distinct criteria” (p. 54). One criterion is “vulnerability to discrimination” (p. 54), about which Laborde is mostly in agreement; the problem, though, is figuring out which nonreligious commitments are ethically salient in the way religious commitments are, so we can then analogize among vulnerable groups and take steps toward awarding exemptions. One comparator is “depth of commitments” (p. 55), and although Laborde has some sympathy with this, she suggests we might sometimes want to protect a deeply held religious interest that has no obvious nonreligious analogy. Finally, Laborde evaluates Eisgruber and Sager’s focus on “close association” (p. 58) to sometimes permit exemptions for groups; here, too, Laborde’s LE approach is mostly sympathetic (e.g., a nonreligious as well as religious group should be able to choose its leaders regardless of a law of general applicability that would cover the normal business world); one critique she offers is that religious institutions sometimes might not properly qualify as truly close associations (as, say, a family or small club would).

The third approach to exemptions that Laborde discusses (and that she calls “narrowing religion” (p. 61)) is offered by Jocelyn Maclure and Charles Taylor, in their book Secularism and Freedom of Conscience.[14] Another LE offering, this book’s case for exemptions would cover “moral beliefs [that] structure moral identity” (p. 62), religious or nonreligious. Although Laborde applauds Maclure and Taylor’s taking our ethical pluralism seriously, she finds their focus on “categoricity” (p. 63)—on whether the claim is based on an unalterable call of conscience—too narrow. We might find religious practices of significance to a claimant’s “moral integrity” (p. 66), without necessarily being duties of conscience.

Chapter 3 of Liberalism’s Religion turns to the “state neutrality puzzle” (p. 69), which in U.S. constitutional law is Establishment Clause territory. Laborde once again gives us theories from Dworkin and Eisgruber/Sager; the third effort here comes from Jonathan Quong. All, again, are mostly fellow travelers with Laborde; she is offering and tweaking versions of LE, setting up her own case. Laborde reminds us of Dworkin’s core argument for “ethical independence” (p. 70); the state should be neutral about matters of religion as it should be neutral about “the good in general” (p. 71). Dworkin is careful to limit this claim to the state’s not using reasons based in foundational views of the good life to regulate matters of personal ethics, such as religion, family, and sexuality. Thus, “ethical independence does not demand neutrality toward other kinds of goods—the good of culture, the arts, or the environment, for example—because they do not fall into [the] domain of personal ethics” (p. 76). But as in much of the other liberal theory Laborde critiques, there’s a boundary problem; here the difficulty is in drawing the line between when the state should not be involved because a matter is one of “personal” (p. 80) ethics and when the matter might affect social justice and possibly harm to others. (As always, the core Millian notion of liberty until harm is caused requires a theory of harm that may be sharply contested.) Laborde concludes that “Dworkinian liberalism relies on a more substantive view of the liberal good than he acknowledged” (p. 82).

On the nonestablishment side of things, Eisgruber and Sager’s concern with vulnerable persons or groups—defined by religion or otherwise—leads to a fairly capacious limit on state speech, which Laborde dubs “civic disparagement” (p. 82). Not only do they adopt a fairly mainstream view about the state’s not endorsing a favored religion; they also contend the state may not use its expressive powers in a way that would disparage persons based on race or other vulnerable/minority characteristics. Laborde is sympathetic to this “religion as not distinctive” position, for theory and law. She reminds us that it can sometimes be hard to tell, in the setting of public education (one arena for state expression), whether the state is teaching a controverted and potentially caste-creating view as true, versus teaching about different positions in a complex area. She also notes that merely because a matter is divisive or controversial doesn’t mean the state’s taking a position is improper—Eisgruber and Sager’s concern with which she agrees is about state endorsement of favored positions on religion, race, and the like, which doubles as state disparagement (sometimes clearly so, sometimes more implicitly so) on these same grounds.

The third scholar whose views Laborde discusses as relevant to establishment clause issues is Jonathan Quong, and his book Liberalism Without Perfection.[15] Quong’s limit on state justifications for lawmaking is based in what Laborde dubs a theory of “foundational disagreement” (p. 92), and it requires distinguishing disagreements about the good from disagreements about the just. For Quong, the former are foundational and appeal to “no premise that all citizens can reasonably expect to share” (p. 93). He contends, though, that disagreements about the just are not foundational, but rely on properly public debatable and resolvable positions. (Note that Quong’s is an anti-religious-distinctiveness position; although it might include a case against laws primarily backed by religious justification, his is a broader Rawlsian public reason position.) Laborde’s key intervention is to claim that the boundary between the realms of the just and the good is often complex, “often requires drawing on contested ideas about the good” (p. 104), and that “liberal neutrality lacks resources to solve the boundary problem” (p. 107).[16] One might think this is a victory for the critical religion theorists, but later in the book Laborde resists this, defending the state’s legitimate authority to determine where justice reigns and thus the state may and must appropriately rule.

In Part II, Laborde develops her case for “disaggregating religion,” at each turn reducing religion to aspects of religious belief and practice that have secular analogues. In so doing, Laborde shows that different dimensions of religion (and their secular analogues) map onto different dimensions of “minimal secularism” (p. 116) (for the Establishment Clause type issues) and of “justice” (p. 197) (for the Free Exercise Clause type issues). Chapter 4 disaggregates religion for nonestablishment issues; Laborde defends what she dubs “minimal secularism” (p. 113). Drawing in part on her Chapter 3 discussion of Dworkin, Eisgruber and Sager, and Quong (and with references to John Rawls at key points), Laborde develops three necessary principles of minimal liberal secularism. These principles are about accessible reasons backing law (“the justifiable state” (p. 117)), the state’s refraining from disparaging vulnerable groups (“the inclusive state” (p. 132)), and not grounding law in comprehensive ethics/views of the good (“the limited state” (p. 143)). For each principle, religion serves as just an example of how the state might err. So, for the justifiable state, following Rawls’ public reason theory, Laborde claims that “state-proffered reasons for laws must be articulated in a language that members of the public can understand and engage with” (p. 119). Accessible reasons are the “currency of democratic debate” (p. 122). Religious reasons are not always inaccessible in this way; and sometimes nonreligious reasons can flunk the inaccessibility test (Laborde cites personal experience as an example). Next, for the inclusive state, tracking Eisgruber and Sager, Laborde claims that “[w]hen a social identity is a marker of vulnerability and domination, it should not be symbolically endorsed and promoted by the state” (p. 137). Here, too, religion is just one example of how the state might fail to adhere to this principle; and just as not all religious reasons are inaccessible, so too not all state religious speech is impermissibly disparaging. Finally, for the limited state, Laborde relies on a Rawlsian principle (previewed in the earlier discussion of Quong’s work), claiming that “[w]hen a practice relates to comprehensive ethics [(otherwise described as “comprehensive conception[s] of the good” (p. 145))], it should not be coercively enforced on individuals” (p. 144). This may cover nonreligious as well as religious doctrines; and when either religion or nonreligion is used in a noncomprehensive way, it is permissible. So, to summarize what Laborde is up to in Chapter 4: The state must turn square corners on all three metrics—laws must be based in accessible reasons; the state should not speak in a way that disparages vulnerable groups; and laws may not be based in comprehensive views of the good. State action regarding religious or secular beliefs and practices is problematic if it flunks any of the three metrics; if it satisfies all three criteria, then it is not problematic.

In Chapter 5, Laborde again treats religion as non-distinctive in developing a theory of freedom of association (for groups). Actually that’s the second part of Chapter 5; the first part is an important and to some extent stand-alone argument—it’s Laborde’s predicate for the whole book, where she defends the legitimacy of the liberal democratic state’s authority to be deciding issues of law and religion in the first place. I’ll return to that after describing the freedom of association argument and the exemptions argument from Chapter 6. To gain presumptive freedom of association a group must be voluntary (members may leave “at no excessive cost” (p. 174)) and identificatory (“individuals join to pursue a conception of the good that is central to their identity and integrity” (p. 174)). With these preliminaries in place, Laborde claims that groups—religious or nonreligious—may have either “coherence interests” or “competence interests” (p. 175). The former tracks what in U.S. constitutional law is a right of expressive association—to form a group to advance particular messages or interests, and thus to exclude those who don’t share that common purpose. Laborde sets forth some conditions for groups to claim freedom of association in this way—their professed doctrine must clearly be restrictive regarding membership; there should be formal organization; and there should be individual freedom to enter and exit. When a group meets these criteria, “full exemptions from laws of general application” (p. 175) are warranted. The key issues here are selection and exclusion of leaders and members. Competence interests are a bit trickier to grasp. Laborde says they “refer to associations’ special expertise in the interpretation and application of [a group’s professed] standards, purposes, and commitments” (pp. 190–191). In such settings—again, religious or otherwise—courts must exhibit sufficient deference to the group’s expertise and ability to answer what we might think of as internal questions (e.g., a question of religious truth or a question of who deserves tenure or promotion).

Laborde bases her theory of non-religion-distinctive exemptions, in Chapter 6, on the notion of integrity. “Integrity is an ideal of congruence between one’s ethical commitments and one’s actions” (p. 203). It is about “the values of identity, autonomy, moral agency, and self-respect” (p. 204). Based on this conception, Laborde contends that the state has reason to exempt “integrity-protecting commitments” (p. 203) (IPCs), of which there are two varieties—“obligation IPCs” (p. 215), which need not be connected to a narrow view of conscience (they can be, e.g., based in “[c]ultural or communal practices” (p. 215)), and “identity IPCs,” which are “non-obligation-imposing commitments and practices that comprehensively regulate the lives of the claimant” (p. 216). Laborde then develops the following architecture for courts to consider exemptions claims: Judges may examine a claimant’s sincerity; the practice claimed for an exemption must be nontrivial, important; “morally abhorrent claims” (p. 207) should be rejected at the get-go; “morally ambivalent claims” (p. 209) can pass step one and get to a balancing test at step two.[17] In general, exemptions claims are tricky because they represent a conflict between the state’s view of the good and the claimant’s/group’s view. In a properly functioning liberal democracy, we might want people to take responsibility for their beliefs (and obey the law or risk the sanction), but “only if background circumstances are fair” (p. 220). Laborde then develops two methods for determining when such circumstances may be sufficiently unfair as to warrant an exemption—“disproportionate burden” (p. 221), which mostly covers obligation IPCs, and “majority bias” (p. 229), which mostly covers identity IPCs.

Judges applying Laborde’s disproportionate burden test would examine several factors: “directness” of the burden (p. 221), “severity” of the burden (p. 222), proportionality of the burden considering the aim of the law (p. 225), and whether the exemption can be awarded without excessive cost-shifting (p. 227). “[E]xemptions are compatible with justice if the balance of these four reasons renders the burden [on the claimant] disproportionate” (p. 228). The majority bias test focuses on situations in which the majority protects its own interests but fails to similarly protect minorities (of various sorts). A key concern is if such bias affects “[c]ore societal opportunities [such as] access to primary goods, work, and education” (p. 231).

Finally, I turn to Laborde’s treatment of “state sovereignty and religious institutionalism” (p. 161). This is a predicate for the rest of her LE argument, as it purports to ground the state’s role in resolving difficult questions of establishment and free exercise. Earlier in the book, Laborde suggests that one of the challenges to LE posed by the critical religion theorists is that of “jurisdictional boundary” (p. 5), and she agrees that “liberals must think harder about the ultimate sovereignty of the state and its legitimacy in enforcing specific terms of liberal justice” (p. 6). In Chapter 5, Laborde rejects this challenge with a ringing endorsement of the legitimacy of the liberal democratic state in using coercive power to resolve various boundary questions and then the difficult as-applied questions that arise in the day-to-day operations of the state. We need, says Laborde, a “final, ultimate source of sovereignty” (p. 161) to resolve various conflicts. The state has liberal legitimacy for this task if it “pursues a recognizably liberal conception of justice, and does so democratically” (p. 168). It is the “only institution with the legitimacy to [resolve matters on which there is reasonable disagreement], because it can reliably enforce a scheme of cooperation over time, and because it represents the interests of citizens as citizens” (p. 168). Thus, the state has the legitimate “competence to adjudicate jurisdictional boundary questions” (p. 162), including proper lines of church/state, religious/secular, political/personal, and public/private. Laborde’s is a familiar argument for the legitimacy of a liberal state’s coercive power, focusing on such a state’s democratic underpinning and the systemic need for settlement.

  1. Two Critiques of Laborde’s Case
  2. An Opposing View—Religion as Distinctive

Liberalism’s Religion focuses on the liberal egalitarian position—both Laborde’s and other scholars whose related views she connects to her own. She briefly describes a “separationist approach” (p. 29) that sees religion as special, but spends little time on it. In Part II.A.1, I will attempt to fill that gap, presenting a normative and doctrinal case for religion as distinctive in a liberal democratic state, specifically in the setting of the U.S. Constitution. In Part II.A.2, I will show where Laborde’s “disaggregating” approach fails to fit with the religion-as-distinctive approach, while agreeing with her at times that it is appropriate to fold aspects of religion into larger categories.

  1. The case for seeing religion as distinctive

At least for the U.S. Constitution, and more broadly for any liberal democratic state, we should see religious belief and practice as distinctive in certain ways. Most issues of religious freedom in the U.S.—under both religion clauses—are connected to theistic belief, to belief in an extrahuman source of normative authority (and sometimes generative power). Several important religion law scholars over the last generation have developed a Laborde-like LE approach, deflecting from the central role faith in God plays in the lives of devout religious people, instead folding religion into larger categories of belief and practice, thus backgrounding the fact that the First Amendment makes religion the subject of nonestablishment and free exercise norms. These approaches do not take seriously enough the primacy of God’s commands for devout religious people. Even if such people are not theocrats, and acknowledge that God’s commands will sometimes have to yield to those of the state, nonetheless they deem the call from an extrahuman source of normative authority to be distinctive. Belief in God, for devout religious people, is not comparable to anything.[18]

The liberal democratic state was born in part from a desire to have a secular rather than theistic grounding for government. Along with this came strong arguments for broad toleration of religious faiths. This is a commendable version of political pluralism, accepting the various sources of the good and authority to which many of our fellow citizens adhere. Although later in this Review I summarize a non-religion-distinctive version of political pluralism in a liberal democracy, there is a case for focusing specially on theistic belief. Reliance on an extrahuman source of normative authority is not like relying on intrahuman theories of the good; often, a sectarian theistic view will appeal and speak to, and only be accessible to, those who share the relevant religious faith. “What is inaccessible to nonbelievers . . . is . . . the relationship between the human believer and the extrahuman source of authority.”[19] There is, thus, a distinctive reason to keep the state out of advancing or endorsing sectarian theistic norms, in various aspects of statecraft (including, as I will canvass further below, in the lawmaking process, public schools, and state speech). There is similarly a distinctive reason to grant exemptions from laws of general applicability to our fellow citizens who pledge allegiance to such an extrahuman source of normative authority (with appropriate consideration for the good advanced by uniform application of such law).

A question for liberal political theory that overlaps with a question for domestic constitutional theory is what kind of legal nonestablishment and free exercise rules should apply if we are appropriately agnostic, as a matter of political/constitutional theory, on the question of whether God exists. If we take liberalism to be a departure from state-sponsored religion, and based in a broad (though not unlimited) toleration of various religious beliefs and practices, then political agnosticism on the God question is the right fit. Were we to bake either theism or atheism into constitutional and other legal rules, we would improperly take positions on a question that not only divides people in a deep way, but also that is arguably the key question on which the modern liberal state has developed to avoid taking a firm position. Furthermore, one of the beauties of the U.S. constitutional system—regarding both structure and rights—is that it is agnostic at its core, in the following sense:[20] In devising both separation of powers and federalism, the framers were primarily concerned with avoiding concentration of power in any one branch or person or seat of government. This is a kind of structural agnosticism, making it hard for any answer to be firmly deemed correct, at least not without clearing difficult hurdles and not without opportunity for challenge and revision. And our core political rights—of speech, press, assembly, petition, and voting—are best understood as offering citizens robust opportunities to challenge the status quo. My claim is that liberal theory and constitutional law back an open-ended political agnosticism on the question whether God exists. To that end, U.S. constitutional doctrine has often reached a workable resolution by appreciating the special role God plays in the lives of the devoutly religious, while keeping the state out of advancing specific religious positions.

Here’s how the religion as distinctive position has played out at the Supreme Court, from its apex with Justice Brennan as the key figure, to some decline since. Let’s start with the Establishment Clause, and examine four types of problem. First, consider the understanding that law should not be based in express, predominant, religious justification. This norm applies to legislators, not citizens. One way of understanding this is as a political process point—reliance on an extrahuman source of normative authority, in whom only some of our citizens (and legislators) believe, cuts out others from meaningful access to the basis of the resulting general, binding law.[21] The Court has invalidated state action on this “religious purpose” test in Epperson, Wallace, Edwards, Santa Fe, and McCreary.[22] Although some scholars and Justices have questioned it—primarily proposing an alternative test that would uphold any law with a plausible secular justification, regardless of an actual express, predominant, religious justification—this line of cases has held fast. They are a prime example of the religion as distinctive position, for there is no comparable norm against laws based in express, predominant, secular justification.

Second, in the public school system, the state may not teach religious doctrine, and its teachers and other officials may not lead students in religious prayer (even if students may formally opt out).[23] That religion is distinctive for both points should be apparent—the state teaches secular doctrine as true all the time (various views on U.S. history, science, etc.); and teachers and other officials may lead students in nonreligious recitation (even if students may formally opt out).[24]

Third, the cases involving public funding of private, religious schools are complex in terms of whether religion is treated as distinctive. The core line of Establishment Clause cases now permits state funding (through various mechanisms) that benefits private religious schooling, so long as such funding is general (it’s not just for such schools) and parents have a choice whether to send their kids to various schools, private religious just one option among them.[25] The Court has mostly backed away from an earlier view that we should see religion as distinctive and disallow funding that benefits private religious education. Note, though, that state funds may not be earmarked for teaching religion as such; that would violate the Establishment Clause just as teaching religious doctrine in public schools would. Also, doctrine still allows some “play in the joints,” i.e., still allows the state to sometimes treat religion as distinctive in not including religious schooling among other generally available funding, even if the Establishment Clause would permit such inclusion. The key holding is Locke v. Davey,[26] permitting Washington to forbid otherwise generally available funding for post-secondary education from being used for ministerial studies. Chief Justice Rehnquist’s opinion treats religion as distinctive, explaining how the framers had a particular concern with public funding for religious ministers, with no similar concern about public funding for other professions or studies. The Court distinguished Locke in Trinity Lutheran Church v. Comer,[27] holding that the Free Exercise Clause forbids excluding religious institutions from receiving public support for building children’s playgrounds; there was no good reason to treat the religious recipient as distinctively unworthy of such funds. In Espinoza v. Montana Department of Revenue, the Court extended Trinity Lutheran to public funding of private education, while reaffirming Locke as a narrow holding.[28] If the state funds private secular education generally, it may not refuse to similarly fund private religious education generally; in this setting, says the Court, there is no “play in the joints” for the state to account for disestablishment concerns if there is no actual Establishment Clause problem.

Fourth, government sponsorship of religious symbols is another area in which religion is distinctive. The Court has refused to adopt one of two possible black-letter rules—the state may never sponsor a religious symbol, or it always may do so (so long as the symbol doesn’t operate in coercive fashion, which will almost never be the case). Rather, the Court has examined nuanced contextual factors to determine whether the state appears to be endorsing a preferred religion through the symbol, or whether the religiosity of the symbol is better seen as backgrounded against a broader secular meaning. So if a crèche stands alone atop courthouse steps, it is unconstitutional; but if it is part of a broader holiday display, it’s okay.[29] If the Ten Commandments are displayed on a courthouse wall, that’s a problem; but if they’re part of a monument park on state grounds, that’s okay.[30] Even a 32-foot-high Latin cross on state land may be constitutional if it is best understood as honoring the World War I dead, and not as the state’s promoting Christianity.[31] None of this contextual examination is needed when the state erects a secular symbol, such as the U.S. flag.

We may deduce one more principle from these religion-as-distinctive Establishment Clause areas of law—the state may advance various notions of the good, even hotly contested ones, so long as they are secular and not expressly those of specific religions or religion in general.[32] There is an important alternative view here, based in a robust Rawlsian conception of public reason, that would bar the state from advancing any comprehensive notion of the good, religious or secular. I’ll return to that—and challenge it—in the next section, when I discuss how Laborde’s view fits with that broader, religion-as-not-distinctive position.

On the Free Exercise Clause side of the ledger, the easiest rule is that the state may not discriminate against particular religions. The Court was unanimous, for example, in spotting such discrimination against a small religion in Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah.[33] This norm is not distinctive to religion, in the following sense—our Equal Protection Clause jurisprudence would reach the same result, but it would also reach this result for targeted discrimination against a specific race, or gender, or sexual orientation. What counts as invalid state discrimination against religion generally has proven to be complex terrain. An easy case was McDaniel v. Paty,[34] invalidating a state constitutional provision forbidding religious clergy from running for certain state offices. Locke/Trinity Lutheran/Espinoza explore a harder aspect of this (denial of otherwise generally available funds to religious persons or institutions). Although I need to think about the issue in more detail, my current view is that while Trinity Lutheran was correctly decided, Espinoza is a mistake.[35] The Missouri program in Trinity Lutheran involved state funds for nonprofits to rubberize playground surfaces. This is a purely secular use, the Establishment Clause does not bar religious institutions from receiving such funds, and there is no good disestablishment value in denying such funds to religious institutions. Thus, we can properly see Missouri’s “no state aid to religious institutions” rule as improperly discriminatory in this setting. On the other hand, a state does have a valid disestablishment-based reason to deny public funding for religious education: even if including religious schools as ultimate beneficiaries of state funds does not violate the Establishment Clause, in large part because of the evenhandedness of such funding, a state may legitimately wish to advance secular education with its funds, in both public and private schools, leaving religious education to be funded privately.[36] Generally speaking, the Court allows more leeway for differential government funding (and certainly for government speech) than for differential government regulation, so long as there is a legitimate public policy reason to be funding program X and not program Y. I realize that everything in this analysis turns on what counts as a legitimate versus illegitimate public policy reason; the argument I would need to develop in greater detail is why a strict public-private line regarding funding of religious education is at least a plausible, valid, state interest, even if the state is not required to adhere to that line. In sum, the position I am sketching here is that public funding of education is an area in which a state should be able to treat religion as distinctive if it wishes, and that the Court has erred in not recognizing that.

The big action in Free Exercise Clause constitutional and statutory law has been the question of exemptions—when, if ever, should religious persons or institutions receive an exemption from an otherwise valid law of general applicability? The question can be stated in two steps: (i) Does the Free Exercise Clause ever require such exemptions (and if so, should courts award them as well as legislatures)? (ii) If the Free Exercise Clause does not require such exemptions (either at all or in particular cases), what is the appropriate legislative role in awarding them as a matter of policy (often called accommodations), and are there Establishment Clause limits on such legislative power? From 1963 to 1990, the Court had reached a kind of settlement on (i): The Free Exercise Clause requires an exemption from a law that places a substantial burden on religious practice, unless the government can satisfy strict scrutiny—i.e., that the law satisfies a compelling state interest and that denying an exemption is a narrowly tailored means of achieving that interest. This was a period during which the Court treated burdens on religious practice as distinctive. The Court ruled against the state in four unemployment insurance cases applying this test, where claimants had been denied such insurance after refusing work on religious grounds.[37] And the Court ruled against the state in one other case applying this test, Wisconsin v. Yoder, allowing the Amish to pull fourteen- and fifteen-year-old children out of any schooling, public or private, contra state law requiring otherwise.[38] Otherwise the Court ruled for the government, and perhaps it’s best to see the practice over these years as applying elevated but not really strict scrutiny; nonetheless, burdens on religious practice were treated with special judicial attention. But in 1990, the Court shifted gears and held in Employment Division v. Smith[39] that even when laws of general applicability place substantial burdens on religious practice, only rational basis scrutiny is required. The concerns were of two sorts—one, that it’s too hard for courts to balance state interest versus religious interest on a case-by-case basis; two, that requiring exemptions for religious claimants (even after a balancing test) would lead to a kind of anarchy. These concerns aren’t specific to religion (the Court wouldn’t have required exemptions for nonreligious persons either); but the Court’s refusal to see harm to religious practice as requiring special treatment was a significant step away from the religion-as-distinctive approach.

And thus, we turn to step (ii): In the aftermath of Smith, Congress and several state governments enacted laws requiring courts to award exemptions to religious claimants substantially burdened by laws of general applicability, unless the government can persuade the court that uniform application of the law is needed to achieve a compelling state interest. In City of Boerne v. Flores,[40] the Court declared the federal version of such a law—the Religious Freedom Restoration Act (“RFRA”)[41]—unconstitutional as applied to state and local governments, but RFRA remains good law restricting the federal government. The federal RFRA, and state versions, are religion-as-distinctive laws—under them, substantial burdens on religious practice deserve amelioration in a way unmatched by similar burdens on nonreligious practice. The Court has so far been receptive to this approach;[42] in particular, any Establishment Clause objection to such distinctive exemptions for religion has not proven problematic. Courts are still working through a complex issue—to what extent does the Establishment Clause require accounting for harm to third parties in balancing state interest against the interest of a religious exemption claimant (assuming the strict scrutiny inquiries about compelling state interest and narrow tailoring invite a kind of judicial balancing)?[43]

Not long after the Court decided Smith, I criticized that decision, trumpeting the religion-as-distinctive settlement that had been in place.[44] My Free Exercise Clause argument was of a political participation legitimacy sort—if we exclude religious claims as express, predominant, justification for law, then we owe something to religious people (one cannot be fully bound by law where the lawmaking process excludes some of what one believes to be true), and a scheme of exemptions (with proper balancing) follows.[45] A softer version of my “compensation” argument can be made—we have struck a good political balance of keeping religion out of politics and politics out of religion (or should be doing so, with various adjustments and balancings). A deep constitutional commitment to religious freedom requires both a governmental sector free from sectarian doctrinal religious claims and protection for people to live by their religious faiths to the greatest extent possible with the equal freedom of others. This is an argument for a distinctive role for religious exemptions, which doesn’t have a counterpart for nonreligious claims, because we permit nonreligious arguments to be the express, predominant, justification for law, we permit the state to lead students in secular recitation and teach core secular values in public schools, and we permit the state symbolically to advance the truth of nonreligious positions, such as by flying the American flag.

  1. How Laborde’s “disaggregating religion” approach fits and doesn’t fit with the religion as distinctive approach

Recall Laborde’s central LE claim: “religion is not uniquely special: whatever treatment it receives from the law, it receives in virtue of features that it shares with nonreligious beliefs, conceptions, and identities” (p. 3). Everything she says about what the state may not do (nonestablishment) and what it must do (free exercise exemptions) she says through the prism of her claim that liberalism’s religion is not about what makes religion distinctive, but rather about what makes religion like other things. In the preceding section, I argued from a normative and descriptive perspective that liberal (and constitutional) theory can and should treat religion as distinctive. I now set forth several ways in which Laborde’s religion-as-not-distinctive approach conflicts with what I have claimed is the better view, and a few ways in which Laborde and I share common ground.

First, regarding nonestablishment, Laborde’s three principles are that the state should base coercive laws on accessible reasons, that the state should not endorse social identities when the upshot would be to harm vulnerable minority groups, and that the state should not base coercive law on comprehensive ethics, i.e., notions of the good. She means these principles to include religion as an example of the problem, but, since she has disaggregated religion into (what she claims are) its features, her claims at each turn are broader. I share her concern about accessibility but would limit its application in a way she does not; I mostly agree with her about state nonendorsement of powerful over vulnerable groups; and I mostly disagree with the concern about basing law on comprehensive notions of the good. The normative and descriptive theory I sketched in the prior section requires a powerful but narrow set of limits on state action regarding nonestablishment.

Laborde bases her theory of accessible reasons backing coercive law on a principle of democratic political participation, and, I would add, democratic legitimacy (and Laborde calls this principle “the justifiable state” (p. 117)). In the liberal state, legislators are representatives of and fiduciaries for the sovereign citizens; the currency of legislative deliberation should be based in reasons and justifications that all citizens, as citizens (rather than as, say, members of the dominant religious group), can follow and appreciate. This is properly seen as a burden on lawmakers; citizens as such may debate each other based on any kind of reason. Basing coercive legislation on express, predominant reference to an extrahuman source of normative authority—the majority’s understanding of God’s will—renders the resulting legislation illegitimate as to those who don’t share the relevant religious faith, because they (through legislators who would be able to speak for them) lack access to the justificatory grounding for the law. Laborde refers to this use of Rawlsian public reason theory as a “thin epistemic filter” (p. 119), and that seems right. As part of her LE, disaggregating religion approach, Laborde claims that “religious views . . . are not uniquely special from the point of view of accessibility” (p. 125). Her examples are “personal testimony” and “personal experience” (pp. 125–126). But these are unlikely candidates for the type of reason/justification that could expressly and predominantly capture a legislative body and lead to coercive law. It seems like a way of saying “see, religion really isn’t distinctive, we can be LEs here!” It distracts from the core animating concern about inaccessible reasons backing legislative action—religious ones. Moreover, the disaggregation approach fails to grapple with what religious people hold nearest and dearest to their hearts, minds, and souls—their faith in and relationship to an extrahuman source of normative authority, i.e., to God. The point isn’t that religious reference is always obscure or not based in reason; this is a complex issue, because for some religious people, God’s-will-based arguments are no less based in reason than are other arguments. The point is that the source of authority is outside the state’s legal apparatus—the framers of the Constitution and laws or those documents—and instead based in the eternal omnipotent framer and draftsperson, as it were. There is a unique problem with basing law on that kind of referent, and acknowledging that uniqueness allows us to focus on distinctive ways in which the liberal state should treat religious belief—regarding both limitations on and exemptions from lawmaking.

Laborde’s second nonestablishment principle is about the “inclusive state” (p. 132); it rejects government speech that would exalt some social identities over others. Although this notion of nonendorsement is closely tied to the similar idea Justice O’Connor developed in Establishment Clause jurisprudence,[46] Laborde follows Eisgruber and Sager (and other scholars such as Nelson Tebbe),[47] in applying this norm to any social identity that “is a marker of vulnerability and domination,” which could include “race, . . . culture or ethnic identity” (p. 137). A cultural-specific, context-dependent, objective inquiry is needed to determine where and how state speech runs afoul of this norm. As a matter of proper liberal democratic theory—focusing on equal liberty—and of U.S. constitutional law, the extension of this norm beyond religion seems appropriate. Our Equal Protection Clause is properly understood as blocking symbolic state action that disparages persons on the basis of religion, race, or similar dominant-over-vulnerable social markers. Religion should not be seen as distinctive on this score, and we can ground the norm solely in the Equal Protection Clause. So, a state’s placing a crèche atop its courthouse steps, or placing a symbol of white nationalism in the same spot, are equally unconstitutional examples of government speech (although I don’t mean to claim they have equal moral status; the former is a symbol to be welcomed when placed by private persons or churches on their own property; the latter is a symbol to be condemned even when placed by private persons on private property, though, under U.S. free speech law, with some exceptions, it is permitted).

Laborde’s third nonestablishment principle—“the limited state” (p. 143)—forbids state coercive action based on a “comprehensive conception of the good,” religious or otherwise (p. 145). This Rawlsian idea, previewed in Laborde’s analysis of Quong’s work, highlights the libertarian aspect of liberalism, in the following sense: it “prioritizes individual rights of self-determination in ethical matters” (p. 144). As a threshold question of proper grounds for state action, this principle is overly restrictive. Say the majority in a legislature are vitalists, and believe that state action should be predicated on preserving human life above all else. Or say they are staunchly committed to environmental stewardship, and believe that all state action should start from that premise.[48] One can imagine many other examples, including majorities that are in the grips of a particular political and/or economic theorist (Karl Marx, Adam Smith, Milton Friedman, etc.). A state may have a comprehensive view about the good (Laborde uses ethics and the good interchangeably here) rather than a piecemeal view about the good, and, as a step-one, threshold, presumptive matter of proper grounding for state action, there is nothing about liberalism that prevents the state from acting according to one or another comprehensive view, such as the examples I have mentioned. Whether a proper constitution in a liberal democracy should contain rights that trump state action (and the content of such rights) is a separate, second-step question. Some state-backed comprehensive notions of the good may never infringe constitutional rights; some may do so all of the time; and some may do so on a case-by-case basis. Thus, my response to Laborde’s religion-is-not-distinctive approach here is twofold: First, even lumping religion with other comprehensive notions of the good, the “limited state” principle is too limited. Second, we should treat religion as distinctive regarding proper grounds for coercive law, but that requires cycling back to principle one, regarding the justifiable state and accessibility.

Regarding exemptions, Laborde and I share a commitment to a fairly robust scheme that would require multi-factor, case-by-case, balancing.[49] I have developed two separate arguments for such a scheme—one that is internal to U.S. constitutional religion clause jurisprudence; the other that is not distinctive to religion, but based in a different theoretical predicate for why robust exemptions are appropriate for a liberal democracy. Regarding the former, as I have explained above, because the state is precluded from basing law on express, predominant, religious justification, from teaching the truth of a religion in public school (or more generally), from leading public school students in religious prayer, and from using its speech power to endorse a favored religion, it is also proper to let religious people out from under the clutches of law that substantially burdens their religious practice, that puts them in a deep conflict between the state and God. This is a strong public-private line settlement, one that we have come close to achieving at times in our constitutional (and statutory) jurisprudence. The case for exemptions here is distinctive to religion. I turn to the case for exemptions more broadly in the next section.

  1. An Alternative Predicate for Exemptions

Although it appears in Chapter 5, Laborde’s case for the legitimacy of the liberal state’s coercive authority, for its justifiable “competence to adjudicate jurisdictional boundary questions” (p. 162), is a predicate for the book. She contends that we are faced with “circumstances of reasonable disagreement” (p. 168), and only the state has the legitimacy to impose and enforce solutions. Laborde’s argument in this section is telescoped, and she does not claim to be setting forth the full case. Furthermore, her argument for state sovereignty isn’t necessary for the central claim in the book, about why and how we should disaggregate religion, and how that should play out for nonestablishment and free exercise. Nonetheless, her case for the legitimacy of the sovereign liberal democratic state is part of her overall framework.

The case, as a wholesale matter, is faulty, and its flaws can point us to a broad argument for exemptions, beyond religion. Assuming the correlativity of arguments for political obligation and legitimacy,[50] note that Laborde does not rely on arguments from consent (express or implied) or duty of fair play; nor does she rely on arguments from a natural duty to obey just institutions or from associative obligation (what we owe each other as fellow citizens).[51] Her position, rather, is a blend of arguments from democratic political participation and the need for systemic settlement. The former may be a necessary condition for political obligation and legitimacy, but it is not a sufficient condition.[52] Whatever mechanism is established for majority rule, individuals don’t knowingly and voluntarily cede self-government (we are born into a society and rarely have the opportunity to swear an oath to abide by majoritarian outcomes, and even such an oath would be problematic if we lack sufficient alternatives). Minority groups (of all sorts) are often not properly heard, and their interests are insufficiently considered. Moreover, the problem of the dead hand is real—constitutions and laws purport to govern indefinitely, covering people who have not had an opportunity to participate in their making, and the realities of political economy make change cumbersome. Laborde’s argument, thus, necessarily turns on there being some minimal democratic bona fides, plus the need for a diverse, complex society to settle various matters on which people disagree.

Arguments for a moral duty to obey the state’s laws (and correlative arguments for the state’s claims of legitimate authority) are meant to be general, to obligate and justify all state claims of authority.[53] These claims may be subject to override in individual instances, but the question up front is whether the general case for obligation and legitimacy can be made. The type of general settlement argument Laborde offers is subject to several objections.[54] One: The argument suffers from some of the same flaws from which consent and democratic participation arguments suffer: Who among us has consented to the political system into which we are born and then live? If individual notions of right—maybe we could call this conscience—are part of what make each of us human and are not displaceable by superior numbers and concretization over time, then “settlement” is just a word for describing the ruling group’s assertion of authority. Two: The argument from settlement does not properly account for subgroups within a polity; their core contention is that the state’s ground of authority is not valid as a trumping source just because a majority of citizens have established it; other sources of authority may exist, and although the state may properly prevail in some instances, it can’t properly claim general valid authority. This is a version of a core argument for political pluralism. Three: Error costs exist in all directions. Although in many cases the state may make correct (or at least best) decisions to settle reasonable disagreement, in others, it may not; perhaps other persons, maybe through non-state groups, can reach correct (or at least better) answers to complex problems. The state’s general claim to authority is too broad. Four: To the extent settlement arguments rely on Burkean common law notions of true/best answers working themselves out over time, they are subject to rejoinders that powerful majority interests may control over time in a way that does not properly obligate those out of power (with a correlative hit to the state’s legitimacy claim) and that accretion over time may represent mistakes building on mistakes and becoming blind to error.

The general case for political obligation and legitimacy—whether from the virtues of settlement or on other grounds—cannot be sustained. Partly that is because of the flaws in the various grounds, and partly that is because of the strength of the case for robust political pluralism, i.e., for permeable sovereignty.[55] We should treat all comprehensive views of the good as presumptively on equal footing. The state’s claim to authority is no more justifiable than, say, a claim based in religious, tribal, or cultural norms. But the liberal democratic state has a monopoly on armies and police and prisons, and so it is subject to a distinctive justificatory burden. Since it cannot carry that burden as a general matter, it must instead justify each law and each application of law. Those who follow different sources of normative authority—religious people, but others, too—thus have two related grounds for complaint about the state’s claim of authority just because it is the state: the general case for political obligation and legitimacy fails, and pluralism rather than state-monism is the better approach to a world of plural claimed sources of normative authority. Exemptions can serve as a kind of “partial exit”;[56] although they don’t let people who follow non-state sources of authority live purely on their own, they do achieve some amelioration. Such a system requires balancing, but the burden is on the state and not on the individual or group claimant.

This case for exemptions is not distinctive to religion and is not meant as an interpretation of U.S. constitutional doctrine. As a matter of ideal political theory, it can complement the more localized case for religious exemptions based in the balance we have struck between a robust Establishment Clause and a Free Exercise Clause that takes seriously the hit to religious people not just from discrimination but also from substantial burdens from nondiscriminatory laws. (This is how that clause should work, does work as filtered through federal and state RFRAs, and could work were the Court to overrule Smith.) This more general case for exemptions, not distinctive to religion, overlaps partly with Laborde’s case. Both of us would apply a balancing test to alleviate some burdens on persons whose commitments conflict with state law. The main difference between our approaches is the grounding for the claim against the state, for the claim that even nondiscriminatory laws of general applicability must sometimes (as a matter of right) give way. Laborde maintains that “the state should generally refrain from infringing integrity-related liberties” (p. 203); this is meant to protect “the values of identity, autonomy, moral agency, and self-respect” (p. 204). My argument is focused instead on competing sources of normative authority from comprehensive views of the good. A constitution with a capacious view of individual rights might cover some of what Laborde wants to include in her argument for exemptions, but protect them from laws that intentionally or directly infringe the rights in question, rather than as incidental burdens from otherwise valid laws of general applicability. If we are focused, though, on exemptions (as Laborde is), then we do better by approaching the matter from principles of political pluralism, pitting competing comprehensive views of normative authority against each other.[57]


The dawn of the liberal democratic state marked a new-found tolerance for persons to practice religion in various forms. Yet, finding the appropriate sweet spot for religion’s role in the state and how state action may affect the lives of religious people continues to be elusive. Cécile Laborde’s ambitious book Liberalism’s Religion comes down firmly on the side of seeing religion as not distinctive, even in a liberal democracy: To the extent that nonestablishment and free exercise norms should prevail, they should prevail insofar as we can disaggregate religion into components that it shares with nonreligious belief and practice. In this Review, I have advanced a position on which Laborde spends little time in her book—religion is distinctive because, for religious people, God is at the center of their beliefs and practices, and there’s nothing else like it. In so doing, I have suggested that there are good reasons for liberal democracy generally, and the U.S. constitutional order specifically, to respond to this sociological fact with nonestablishment and free exercise norms that are distinctive to religious belief and practice. Liberalism’s religion need not be disaggregative; it can remain true to core liberal principles while taking seriously the role that God plays in the lives of the devout.



         [1].    Leonard F. Manning Professor of Law, Fordham Law School. I dedicate this Review in loving memory of my dear friend, John Nagle, from whom I learned so much about religious devotion. For helpful written feedback on this essay, I thank Aditi Bagchi, Jim Brudney, Clare Huntington, Cécile Laborde, Russ Pearce, Aaron Saiger, Micah Schwartzman, Olivier Sylvain, and Nelson Tebbe. And I am very grateful to my Fordham Law colleagues who participated in a Zoom virtual workshop on this Essay during the novel coronavirus crisis.

         [2].    Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067 (2019).

         [3].    Espinoza v. Montana Dep’t of Revenue, No. 18-1195, 2020 WL 3518364 (U.S. June 30, 2020).

         [4].    Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719 (2018).

         [5].    Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014).

         [6].    See Christopher L. Eisgruber & Lawrence G. Sager, Religious Freedom and the Constitution (2007); Nelson Tebbe, Religious Freedom in an Egalitarian Age (2017); Micah Schwartzman, What if Religion Is Not Special?, 79 U. Chi. L. Rev. 1351 (2012).

         [7].    See Stanley Fish, Mission Impossible: Settling the Just Bounds Between Church and State, 97 Colum. L. Rev. 2255 (1997); Steven D. Smith, Discourse in the Dusk: The Twilight of Religious Freedom?, 122 Harv. L. Rev. 1869 (2009) (book review).

         [8].    See Abner S. Greene, The Incommensurability of Religion, in Law and Religion: A Critical Anthology 226 (Stephen M. Feldman ed., 2000); Abner S. Greene, The Political Balance of the Religion Clauses, 102 Yale L.J. 1611 (1993) [hereinafter Greene, Political Balance]. For other writings supporting religion as distinctive, see Andrew Koppelman, Defending American Religious Neutrality (2013); Thomas C. Berg, Can Religious Liberty Be Protected as Equality?, 85 Tex. L. Rev. 1185, 1211–15 (2007) (book review); Douglas Laycock, Formal, Substantive, and Disaggregated Neutrality Toward Religion, 39 DePaul L. Rev. 993, 1001–06 (1990); Douglas Laycock, Religious Liberty as Liberty, 7 J. Contemp. Legal Issues 313, 316–19 (1996); Michael W. McConnell, The Problem of Singling Out Religion, 50 DePaul L. Rev. 1, 3 (2000); Mark Storslee, On Religion’s Specialness, 81 Rev. Pol. 656, 661 (2019) (responding to Laborde’s disaggregation argument) (“When it comes to religion, the whole is more than the sum of its parts. And indeed, that might just be what makes it special.”).

         [9].    See Abner S. Greene, Religion and Theistic Faith: On Koppelman, Leiter, Secular Purpose, and Accommodations, 49 Tulsa L. Rev. 441 (2013) [hereinafter Greene, Theistic Faith]. One of the issues debated in law and religion scholarship is how to define religion, and whether belief in God is necessary to such definition. I have focused on belief in God—on an extrahuman source of normative authority—as typical and at the core of the law and religion debates in the U.S. Sometimes theory should be able to focus on typical examples, rather than necessary features, of the matter under discussion. See Frederick Schauer, The Force of Law 35–41 (2015) (defending his focus on law’s coercive nature, even though aspects of law are not coercive).

       [10].    See Abner S. Greene, Against Obligation: The Multiple Sources of Authority in a Liberal Democracy (2012).

       [11].    Ronald Dworkin, Religion Without God (2013).

       [12].    See Brian Barry, Culture and Equality: An Egalitarian Critique of Multiculturalism (2001). For my response, see Abner S. Greene, Three Theories of Religious Equality . . . and of Exemptions, 87 Tex. L. Rev. 963 (2009).

       [13].    See Eisgruber & Sager, supra note 6.

       [14].    Jocelyn Maclure & Charles Taylor, Secularism and Freedom of Conscience (2011).

       [15].    Jonathan Quong, Liberalism Without Perfection (2011).

       [16].    For Quong’s response, see Jonathan Quong, On Laborde’s Liberalism, Crim. L. & Phil., May 2019 (“All areas of social life are justice-apt.”) (“[D]isagreements between reasonable citizens about the political status of certain creatures can be resolved by appeal to the justificatory framework shared by all such citizens.”).

       [17].    A few commentators on Laborde’s book have taken her to task for allowing too many claims into the “morally ambivalent” category and thus subject to a balancing test. See, e.g., Alan Patten, Religious Accommodation and Disproportionate Burden, Crim. L. & Phil., Jan. 2020 (“[A] person has a pro tanto claim on others only for a fair opportunity to pursue and fulfill her IPCs. There is no pro tanto claim to realize IPCs that either by their very nature are inconsistent with the fair claims of others, or that for contingent reasons (resource scarcity, etc.) are incompatible with the fair claims of others.”); Lori Watson, Integrity: An Individual or Social Virtue?, 81 Rev. Pol. 652, 654–55 (2019) (“[C]laims for refusing to recognize or act on the basis of laws that demand equal recognition for others will not ground an integrity objection.”). But as Laborde has said in response to some commentators, hers is a modest liberalism that is a “substantive commitment that takes seriously pluralism and the burdens of judgment.” Cécile Laborde, Reply: Disagreement, Equal Respect, and the Boundaries of Liberalism, 81 Rev. Pol. 665, 666 (2019). She adds (responding to Watson) that the disagreement between relational egalitarians and orthodox political liberals is a “reasonable disagreement about liberal justice,” and that sometimes granting exemptions is one way to recognize this disagreement. Id. at 667.

I would go further than Laborde and suggest that an exemptions regime should be and, in fact, always is, a balancing test. When we exclude certain claims supposedly up front, as “morally abhorrent,” we are implicitly doing a kind of balancing—it’s just that the case is so heavily tilted toward denying the exemption that it looks as though we’ve done something at step one without reaching a step two balancing test. Here, as in free speech law, it’s better not to bury any of the normative work we’re doing. Rather, in both settings, we should acknowledge a wide array of claims for religious freedom or freedom of speech, and then realize that some of these claims are quite weak, either on their own merits, or because of strong countervailing state interests, or both. For more on this in the free speech setting, see Abner S. Greene, “Not in My Name” Claims of Constitutional Right, 98 B.U. L. Rev. 1475, 1508–11 (2018) [hereinafter Greene, Constitutional Right].

       [18].    For my challenge to several scholars who critique a God-centered view of religious freedom, see Greene, Theistic Faith, supra note 9 (discussing works by Richard Dawkins, Daniel Dennett, Sam Harris, and Brian Leiter).

       [19].    Abner S. Greene, Is Religion Special? A Rejoinder to Scott Idleman, 1994 U. Ill. L. Rev. 535, 538 (1994). Laws based in natural and social science are not similarly inaccessible to nonbelievers. “[E]ven if science . . . is based—as religion is—in an important way on faith (nondeducible premises), the critical difference is that by its own terms, science points to the human and natural world for the source of value, whereas religion, by its own terms, points not only to the human and natural world, but also outward to an extrahuman realm.” Id. at 540.

       [20].    For elaboration on this “multiple repositories of power” theory, see Abner S. Greene, Checks and Balances in an Era of Presidential Lawmaking, 61 U. Chi. L. Rev. 123 (1994); Abner S. Greene, Civil Society and Multiple Repositories of Power, 75 Chi.-Kent L. Rev. 477 (2000); Abner S. Greene, Kiryas Joel and Two Mistakes About Equality, 96 Colum. L. Rev. 1 (1996) [hereinafter Greene, Kiryas Joel].

       [21].    This is not to deny the deep influence of religion and religious ideas on secular law. For a recent thoughtful exposition of this, see Jeremy Waldron, Religion’s Liberalism, Crim. L. & Phil. (forthcoming 2020) (explaining that religious ideas that “can be discerned in liberalism’s abstract philosophical positions” include the sacredness of the person, human dignity, free will, personal responsibility, equality, respect for ordinary people, rights, and social justice).

       [22].    McCreary Cty. v. Am. Civil Liberties Union, 545 U.S. 844 (2005); Santa Fe Indep. Sch. Dist. v. Doe, 530 U.S. 290 (2000); Edwards v. Aguillard, 482 U.S. 578 (1987); Wallace v. Jaffree, 472 U.S. 38 (1985); Epperson v. Arkansas, 393 U.S. 97 (1968).

       [23].    Epperson, 393 U.S. at 106–07, and Edwards, 482 U.S. at 589–94, are clear on the former; Abington Sch. Dist. v. Schempp, 374 U.S. 203 (1963), and Engel v. Vitale, 370 U.S. 421 (1962), on the latter.

       [24].    See Aaron Saiger, Deconstitutionalizing Dewey, 13 FIU L. Rev. 765, 767–77 (2019). For some doctrinal difficulties with this position, see Abner S. Greene, The Pledge of Allegiance Problem, 64 Fordham L. Rev. 451 (1995).

       [25].    See Zelman v. Simmons-Harris, 536 U.S. 639 (2002).

       [26].    540 U.S. 712, 719 (2004).

       [27].    137 S. Ct. 2012 (2017).

       [28].    Espinoza v. Montana Dep’t of Revenue, No. 18-1195, 2020 WL 3518364, at *11 (U.S. June 30, 2020) (“A State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious.”).

       [29].    Compare Allegheny v. Am. Civil Liberties Union, 492 U.S. 573 (1989), with Lynch v. Donnelly, 465 U.S. 668 (1984).

       [30].    Compare McCreary County v. Am. Civil Liberties Union, 545 U.S. 844 (2005), with Van Orden v. Perry, 545 U.S. 677 (2005).

       [31].    Am. Legion v. Am. Humanist Ass’n, 139 S. Ct. 2067 (2019).

       [32].    For my support for a broad view of secular state speech, see Abner S. Greene, Government Endorsement: A Reply to Nelson Tebbe’s Government Nonendorsement, 98 Minn. L. Rev. Headnotes 87 (2013); Abner S. Greene, Government of the Good, 53 Vand. L. Rev. 1 (2000); Abner S. Greene, Government Speech on Unsettled Issues, 69 Fordham L. Rev. 1667 (2001); Abner S. Greene, State Speech and Political Liberalism, 28 Const. Comm. 421 (2013).

       [33].    508 U.S. 520 (1993).

       [34].    435 U.S. 618 (1978).

       [35].    I put to one side the anti-Catholic discriminatory animus that might have been instrumental in producing the relevant “no religious aid” provisions of the Missouri and Montana constitutions in Trinity Lutheran and Espinoza. See Espinoza v. Montana Dep’t of Revenue, No. 18-1195, 2020 WL 3518364 (U.S. June 30, 2020) (Alito, J., concurring). Discriminatory animus should be relevant and sometimes dispositive in invalidating state action.

       [36].    I am focusing on how the relevant funds are used—for rubberizing playground surfaces versus for scholarship tuition for schooling. Chief Justice Roberts’ opinion for the Court in Espinoza proceeds differently—it says the relevant state constitutional provisions in Trinity Lutheran and Espinoza discriminate based on the religious status of the ultimate funds recipient. See id. at *6–7. Justice Gorsuch in concurrence and Justice Breyer in dissent find fault in Roberts’ analysis—both think it more accurate to say that what is at issue in these cases is how the funds are used.  See id. at *21–22 (Gorsuch, J., concurring); id. at *27–28 (Breyer, J., dissenting).

       [37].    See Frazee v. Ill. Dep’t of Emp’t Sec., 489 U.S. 829 (1989); Hobbie v. Unemployment Appeals Comm’n, 480 U.S. 136 (1987); Thomas v. Review Bd., 450 U.S. 707 (1981); Sherbert v. Verner, 374 U.S. 398 (1963). Some argue that these are not best seen as exemptions cases, but rather as holdings preventing states from discriminating against religious reasons for not working. See, e.g., Christopher L. Eisgruber & Lawrence G. Sager, The Vulnerability of Conscience: The Constitutional Basis for Protecting Religious Conduct, 61 U. Chi. L. Rev. 1245, 1277–82 (1994).

       [38].    406 U.S. 205 (1972).

       [39].    494 U.S. 872 (1990).

       [40].    521 U.S. 507 (1997).

       [41].    42 U.S.C.A. §§ 2000bb to 2000bb-4 (West).

       [42].    See Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014); Gonzales v. O Centro Espirita, 546 U.S. 418 (2006); see also Holt v. Hobbs, 574 U.S. 352 (2015); Cutter v. Wilkinson, 544 U.S. 709 (2005) (construing related statutes).

       [43].    See Tex. Monthly, Inc. v. Bullock, 489 U.S. 1, 15, 18 n.8 (1989) (plurality opinion).

       [44].    See Greene, Political Balance, supra note 8.

       [45].    For some of my responses to objections, see id. at 1636–39. Laborde objects to my political balance argument in two ways: She reiterates her LE, disaggregation approach, separating why we might sometimes protect religious practice from why we might sometimes exclude it from backing lawmaking; and she rejects “the view that people should be compensated for the illegitimacy of some of the arguments they might present when acting in official capacity” (p. 304 n.90). My baseline differs from hers—mine is that all citizens have an equal right (through their legislators) to speak their truth in the lawmaking process; a partial gag rule on such speech creates a legitimacy problem for the resulting law that purports to bind all, including those whose arguments were excluded from being the express, predominant, justification for law. That there might be a good reason for the partial gag rule (on which Laborde and I agree) does not make this legitimacy problem go away.

In other words, we should distinguish two types of argument we might exclude from being the express, predominant ground for lawmaking. We have reached a justifiable settlement to exclude certain such grounds—e.g., racism or sexism. This is a kind of viewpoint exclusion for lawmaking. When we exclude a sectarian theistic claim from grounding lawmaking it isn’t because we have determined the viewpoint contained therein is wrong or otherwise unjustifiable. Rather, the state should take no position on contested religious truth claims, and, similarly, such claims should not be the express, predominant ground for generally binding law—such reliance on an extrahuman source of normative authority improperly excludes nonbelievers from the lawmaking process. But this does not mean the religious viewpoint asserted for lawmaking is illegitimate in the way a racist or sexist viewpoint is.

       [46].    See, e.g., Lynch v. Donnelly, 465 U.S. 668, 688 (1984) (O’Connor, J., concurring).

       [47].    See Eisgruber & Sager, supra note 6; Nelson Tebbe, Government Nonendorsement, 98 Minn. L. Rev. 648, 658–65 (2013).

       [48].    See Greene, Constitutional Right, supra note 17, at 1520–21.

       [49].    A certain subset of exemptions problems poses a particularly difficult challenge for me, namely, those that pit certain equality claims—say, to reproductive freedom or to LGBTQ rights—against claims for religious exemption. Laborde acknowledges that in many of these types of cases, the religious person is not making a morally abhorrent claim, but rather a morally ambivalent one, and thus a balancing test is needed (pp. 225–27). How we should conduct such a balancing test is one of the issues in a work-in-progress of mine, tentatively called “The Dilemma of Liberal Pluralism.”

       [50].    See Greene, supra note 10, at 24–34. My arguments in this section are drawn from chapters 1 and 2 of Against Obligation.

       [51].    For the case that none of these arguments properly grounds political obligation or legitimacy, see id. at 35–45, 56–94.

       [52].    See id. at 45–56.

       [53].    See id. at 15–17.

       [54].    See id. at 94–113.

       [55].    See id. at 20–24. Jean Cohen refers to my work (and the work of others) as advancing a kind of jurisdictional political pluralism. See Jean L. Cohen, Sovereignty, the Corporate Religious, and Jurisdictional/Political Pluralism, in Religion in Liberal Political Philosophy 83 (Cécile Laborde & Aurélia Bardon eds., 2017); Jean L. Cohen, On Liberalism’s Religion, Critical Rev. Int’l Soc. & Pol. Phil., June 2018. She offers several critiques of this position, which I will address in future work. Regarding Laborde’s defense of the legitimacy of the sovereign democratic state, a view Cohen shares, see id. Cohen notes the seriousness of the challenge from jurisdictional political pluralists, and cautions that “we need more from Laborde on the jurisdictional question because I doubt it will go away and I fear that those who do not accept the core intuition behind the notion of liberal democratic state sovereignty (the primacy of civil law) are quite serious about their projects.” Id. at 15. Indeed.

       [56].    See Greene, supra note 10, at 114–15; Greene, Kiryas Joel, supra note 20, at 8–57.

        [57].    Additionally, the “values” Laborde seeks to protect will ring true to comprehensive liberals, but maybe not always or as much to persons who adhere to group values less focused on, say, autonomy. This is another reason to favor an exemptions approach based on competing sources of normative authority from comprehensive views of the good and not on “integrity-related liberties.”