Are Organizations’ Religious Exemptions Democratically Defensible?

Many democratic societies are pluralistic: people from different cultural, ethnic, and religious backgrounds live together, with different plans and values, and they disagree strongly about the permissibility of particular practices. Yet coordination and cooperation require that all citizens are united under one set of laws. Sometimes, this tension between pluralism and unity produces a religiously grounded exemption: there is a generally applicable law, but some are granted an exemption from that law because of religious conviction. 

Thus, the United Kingdom’s Highway Code requires that “On all journeys, the rider and pillion passenger on a motorcycle, scooter or moped MUST wear a protective helmet.” Yet, “This does not apply to a follower of the Sikh religion while wearing a turban.”1 In other cases, the exemption is granted for religious reasons, but the exempt party is not an adherent of the religion: in the Australian state of Victoria, local councils have successfully applied for exemptions from antidiscrimination legislation so they can run women-only swimming classes targeted at Muslim women.2 Here, the exempt parties are the councils, yet the exemption is justified with reference to the religion of individuals (swimming pool users).

In the 1990s, there was heated philosophical debate over such exemptions. Some viewed them as the proper response to individuals’ autonomy or need for recognition.3 Others argued that exemptions are unnecessary if we have robust freedom of association or that the values underlying the general laws are sufficient to reject exemptions (and if the values are not sufficient for this, then the general law should be scrapped altogether, rather than exempting some from it).4

All this concerns individuals’ religious claims. But recently, organizations’ religions have loomed large in pluralistic democracies. In 2014, Ashers Bakery in Northern Ireland refused to bake a cake with the slogan “Support Gay Marriage” because the slogan was “inconsistent” with the company’s religious beliefs. The customer sued the company for discriminating against his sexual orientation and political beliefs. In October 2018, the Supreme Court ruled in favor of the bakery, stating that service providers may refuse to endorse messages they profoundly disagree with.5

A legislative example comes from Australia, where the Sex Discrimination Act allows an “educational institution that is conducted in accordance with the doctrines, tenets, beliefs or teachings of a particular religion or creed” to “discriminate against another person on the ground of the other person’s sex, sexual orientation, gender identity, marital or relationship status or pregnancy” if that person is a potential staff member, contract worker, or student.6 Thus, religious educational institutions may refuse to accept gay or trans people as staff or students, though such refusal would be unlawfully discriminatory if enacted by a nonreligious educational institution. Thus, religious educational institutions are exempt from generally applicable antidiscrimination laws.

This essay examines justifications for exemptions that protect the religious convictions of organizations, including schools, hospitals, businesses, charities, churches, and others. My aim is not to justify or reject particular exemptions, such as those described above. My aim is more fundamental. I ask whether organizations’ religious convictions can give rise to claims at all, even before those claims have been weighed against individuals’ competing claims. I argue that exemptions should almost always be judged with reference to the religious convictions of individuals, not organizations. I reach this conclusion by examining four arguments for organizations’ religious exemptions, only one of which succeeds, and then only rarely.

To start, what are organizations? They are a type of collective agent. A collective agent is constituted by agents that are united under a group-level, rationally operated, distinct decision-making procedure. In general, a collective agent might be large or small, formal or informal, short-lived or long-lasting, and so on, including families, sports teams, reading groups, and many more. Organizations, though, are specific: they have “(a) criteria to establish their boundaries and to distinguish their members from non-members, (b) principles of sovereignty concerning who is in charge and (c) chains of command delineating responsibilities within the organization.”7

Collective agents–including organizations–can form irreducibly group-level religious convictions. To see this, consider that a “decision-making procedure” takes in reasons, beliefs, and preferences, and processes them to produce decisions. Organizations’ procedures include voting, committees, meetings, and so on, but their procedures are often informal and tacit, with the organization’s true beliefs and preferences revealed by the on-the-ground behavior of members (when acting within and because of their role), rather than by the “official party line.” Whether formal or informal, an organization’s procedure is “distinct” in that 1) the reasons it takes in tend to differ in kind from the reasons any of its members take in when deciding for themselves (consider: votes, proposals, and so on); and 2) its method for processing those reasons is different from the method of any one member when deciding for herself. For example, an organization might take the meeting contributions of members and process these using conversation-based consensus, thereby using a distinctive set of inputs and procedures to arrive at organizational beliefs. Members are unlikely to use these inputs, processed in this way, when settling the beliefs they hold themselves. If a procedure is “rationally operated,” it is operated with the aim of ensuring that current decisions follow from current beliefs and preferences, and that current beliefs and preferences are consistent with past beliefs, preferences, and decisions, plus any new evidence that has arisen since those were formed.8

The rational operation of a distinct procedure can mean a collective’s current beliefs are determined by its past beliefs, rather than by members’ current beliefs. For example, if a school has a long-standing practice of focusing on Christianity when teaching religion, then it might be rational for the school to continue this practice (maintain this preference), even if some, many, most, or even all current teachers and managers would prefer the school teach all religions equally. This possibility of departure is crucial, since–as I will explain–it allows a collective to have a religious conviction that no member has. 

With this characterization of organizations in hand, how might we justify their religious exemptions? A first strategy emphasizes that organizations are intimately related to members. That intimacy inheres in at least two strands. First, organizations largely supervene on members: many ways of changing organizations require changing the members. For example, one natural way to alter an organization’s convictions is for enough members to alter their inputs in the decision-making procedure. Second, organizations’ actions are largely constituted by members’ actions: an organization usually cannot implement a policy, sign a contract, and so on, without members’ actions.

Given this intimate connection, perhaps organizations’ religious exemptions are justified via the religious convictions of members. That would be convenient, since we have well-established theories justifying religious exemptions for individuals. Perhaps the religious convictions of bakery owners generate a claim of the bakery itself. Perhaps the religious convictions of schools’ managers justify a claim of the school itself.

To assess this, we must justify individuals’ religious exemptions, returning to the 1990s debate. Philosopher Will Kymlicka has focused on “societal cultures” rather than religions, but his points can be extended to religions. For Kymlicka, a societal culture is “a culture which provides its members with meaningful ways of life across the full range of human activities including social, educational, religious, recreational, and economic life, encompassing both public and private spheres.”9 Kymlicka’s crucial premise is that “freedom involves making choices among various options, and our societal culture not only provides these options, but also makes them meaningful to us.”10 Kymlicka insists people do not need “freedom to go beyond one’s language and identity, but rather the freedom to move around within one’s societal culture.”11 Plausibly, this role–of providing options, making options meaningful, and allowing us to choose among them–extends to religious tenets, practices, and communities, rather than being restricted to societal cultures.

Kymlicka argues that we need exemptions in order to preserve societal cultures, which in turn are needed because of their value for individual autonomy, understood as the capacity to make choices from among meaningful options. By “meaningful” options, I take Kymlicka to mean options for which there are self-identity connotations to choosing one way or another; an option is meaningful if it reflects some core feature of a person’s identity. Kymlicka’s argument resonates with philosopher Joseph Raz’s autonomy-based conception of well-being, according to which “a person’s well-being depends to a large extent on success in socially defined and determined pursuits and activities. . . . [People’s] comprehensive goals are inevitably based on socially existing forms.”12 That is, our well-being depends upon our ability to select from among options that are already well-established within our society or, more important for present purposes, our religion.

A different argument for individuals’ exemptions draws on philosopher Michael Sandel’s idea that humans’ constitutive ends define our personal identity, such that we are “thick with particular traits.”13 These ends and traits are not chosen, as the autonomy argument asserts. Rather, one’s religion (and culture more broadly) is “an attachment they discover, not merely an attribute but a constituent of their identity.”14 Similarly, philosopher Robert Audi endorses “a protection of identity principle: The deeper a set of commitments is in a person, and the closer it comes to determining that person’s sense of identity, the stronger the case for protecting the expression of those commitments.”15 Audi points out that “as a matter of historical fact and perhaps of human psychology as well, religious commitments tend to be important for people in both ways: in depth and in determining their sense of identity.”16 Thus we have the identity-based argument for claims to religious exemptions: our religion is constitutive and/or determining of our (sense of) identity; our (sense of) identity should be respected and protected; therefore, our religion should be respected and protected, which will sometimes require that we are exempt from generally applicable laws.

How might humans’ autonomy-based or identity-based claims transfer to organizations? The idea is this: When Ashers Bakery endorses a message, this implies that (some of) its members endorse that message. But the option not to endorse that message is crucial for members’ autonomy or identity. So, for members’ autonomy or identity to be respected, the bakery must be granted a claim to resist endorsing the message. The action transfers down (from organization to member); so the claim not to perform that action transfers up (from member to organization).

The problem is that the action does not transfer down. So there is no reason for the claim to transfer up. Ashers Bakery endorsing a message does not imply that any individual member endorses the message. Even if it is true that–to respect and protect individuals’ autonomy or identity–individuals should be free not to endorse messages they disagree with, this individual freedom is not infringed upon when an organization of which they are a member endorses a message. The transferring-up strategy commits the fallacy of assuming that when a whole has some property, some constituent part of the whole also has that property. If a wall is eight feet tall, that does not imply that any brick constituting the wall is eight feet tall. Likewise, when a bakery endorses a message, this does not imply that any member endorses the message. 

Nonetheless, sometimes some, most, or even all organization members will feel (or be interpreted as) tainted by the behaviors of their organization. A school’s hiring a gay teacher does not imply that any member hires the gay teacher. But the school’s hiring might cause individuals on the hiring committee to do things inconsistent with their autonomy or identity. If so, do members’ claims transfer up to the organization, despite the action not transferring down? 

No. Members claims might be real, in such cases. But members’ claims do not generate a claim of the organization itself. To be clear: members’ claims need to be balanced against the claim of the potential new hire, before an all-things-considered judgment is made. If the former claims outweigh the latter, then members are permitted not to be involved in the organization’s action. If there is no other way for the organization to perform the action, then the organization is permitted not to perform the action. But this does not mean that the organization has a claim. Instead, it is akin to the Australian city councils being granted exemptions to run women-only swimming classes. There, it was not that Muslim women’s rights were transferred up to the city council, such that we were respecting the council’s claim and right to have its religious convictions respected. Instead, granting the council an exemption was a means of respecting the women’s rights. 

Similarly, sometimes an organization’s action would have detrimental effects on members’ autonomy or identity. The members may have a claim not to be involved in that action. But these are members’ claims, not the organization’s claims. This is important for two reasons: 1) such member claims will likely change as the composition of the organization changes–present members’ autonomy and identity do not say anything about future members’ autonomy and identity, so the organization’s exemption should not be projected into the future; and 2) if we view the organization’s exemption as grounded in a claim of the organization rather than of the member(s), then we may be misled into thinking the claim is unduly weighty (because organizations are large, powerful, and subsume many members). When we view the claim as held by the relevant member(s), it will be easier to give it proper weight balanced against the competing claims of other individuals (such as potential new staff of the school). 

Additionally, there are practical upshots to viewing the claim as held by members rather than by the organization. If members make a claim based on being tainted by the organization’s action, then the first response should be to find other members who do not mind such “taint.” The first response should not be to grant the organization (as a whole) the permission not to perform the action. Furthermore, members’ claims must be treated on a case-by-case basis: in an instance in which all members refuse to be involved in the organization’s action, this might (pending consideration of competing claims) justify allowing the organization not to perform that action in that instance. But it would not justify a general and ongoing exemption from the organization performing actions of that type.

In sum, we must not confuse an organization’s claims with its members’ claims. The latter do not give rise to the former, even if the latter can justify organizational noncompliance with laws in some instances. To believe otherwise is to neglect the ontological distinctness of the organization and its members.

A second argument suggests organizations have their own claims to autonomy and/or identity-protection. Take a university with a religious character. The interests of the university are not merely a product of the interests of its members; its interests may run counter to their interests. So perhaps it has its own right to autonomy or identity-protection.

Take autonomy first. The idea is that one’s religion provides one with options, and choosing from among those options is highly valuable: “the sort of freedom . . . they [that is, people] most value, and can make most use of, is freedom . . . within their own societal culture.”17 This argument is grounded in the liberal conception of the self: the self is a fundamentally free being. In philosopher John Rawls’s words, “the self is prior to the ends which are affirmed by it,” such that individuals “do not think of themselves as inevitably bound to, or as identical with, the pursuit of any particular complex of fundamental interests that they may have at any given time.”18 Instead, they choose from among the options their societal culture gives them.

This conception of the self is not applicable to organizations. Organizations cannot “step back” from their goals like individuals can. To see this, imagine what it would take for a university to reflect upon whether to pursue the goals of teaching and research. Its decision-making procedure is set up such that these goals are built in. The university qua university cannot consider neglecting these goals. A university can decide among some options: it might decide to invest in humanities rather than sciences, for example. But it does so against the background of fundamental preexisting commitments, not from the position of being “prior to” the ends it affirms. The autonomy-based argument is inapplicable.

The identity-based argument was grounded in the proposition that our sense of self and our life’s meaning would be lost if we could not act in ways that express the central aspects of that identity. However, organizations do not have a sense of self or life’s meaning, as individuals do. Such senses require phenomenal consciousness: a subjective experience, an inner world, a creature with the sense. Organizations lack phenomenal consciousness.19 So it is false that the organization’s sense of self and its life’s meaning would be lost, were it not permitted certain practices. I have argued that organizations have beliefs, including beliefs that are so unshakeable they amount to convictions. But a sense of self or sense of meaning is a qualitatively different thing from beliefs, however unshakeable.

In this way, organizations do not fit within either the autonomy-based or identity-based defenses of religious exemptions. Organizations are mainly constituted by persons, but they must not be equated with persons. Not all agents are persons.

A third argument observes that we engage with organizations through what philosopher Daniel Dennett has called “the intentional stance”: we take a stance toward organizations that imputes to them beliefs, preferences, intentions, and actions.20 One of the main reasons we do this is that organizations “perform in a certain way”: specifically, they give explanations of their actions.21 Political theorists Christian List and Philip Pettit wrote: “Let the agent be a Martian, or a robot, or a chimp that has been trained or engineered to a higher level of performance. If it proves capable of engaging us on the basis of commonly recognized obligations . . . we have every reason to incorporate it in the community of persons.”22 Organizations can offer accounts of their actions, in which those accounts acknowledge their obligations to others. They are therefore part of our accountability-community.

Philosopher Leonie Smith has used this reasoning to argue for organizations’ rights.23 Crucially, Smith’s argument does not rely on substantive normative commitments. In pluralistic democracies, citizens disagree about such commitments. Thus, a justification of organizations’ rights that relied on such commitments would enjoy scant support in pluralistic democracies. Instead, Smith has argued for granting organizations only those rights that are reasonable preconditions for them to offer accounts of their actions. If some rights are reasonable preconditions for such account-giving, and if we have good reasons to bolster organizations’ account-giving abilities, then we have good reasons to grant organizations some rights.

Indeed, we do have good reason to bolster organizations’ account-giving abilities: such abilities allow us to demand explanations of their failures, to blame them when they do wrong, and to bestow obligations on them. These are valuable social-political practices. The question becomes: which rights must organizations enjoy if they are to perform in this way? Smith suggests that they need “the right to free speech, to free association, and to be able to enter into legal contracts, among others.”24 Yet she suggests that, for example, “the right to a family private life” might not be necessary for organizations, even if this is needed “in order to be human.”25 And closer to our purposes, she asserts that a profit-driven organization “may not have a right to religious belief as it does not need this to perform . . . in the particular social sphere within which it is capable of participating, and in which it is structured to participate.”26

Smith is tentative in her endorsement of some rights and her rejection of others. To build more certainty, we should consider what it takes to give an account of one’s actions. In a pluralistic democratic society, I suggest, an organization’s public explanations of its actions should refer only to public reasons, where a public reason is, roughly, a reason that all sensible and informed citizens recognize as a reason.27 For example, if an organization refuses to do business with a gay person “because our holy book says homosexuality is wrong,” then it has given a nonpublic reason for its action. By contrast, if it refuses to do business with someone “because that person broke a contract with us in the past,” then this reason is public: it is a reason all sensible and informed people would take to be a reason.

Of course, the line between a public and a nonpublic reason is vague and contestable. But reasons that refer to substantive religious doctrines are clearly nonpublic. If an organization owes society-at-large an account of its actions, then it is not helpful if the organization appeals to a religious doctrine that other members of the society do not endorse. Such an explanation is not intelligible to all reasonable and informed members of society, so it is not the kind of explanation that we should use organizations’ rights to facilitate. Religiously grounded exemptions to generally applicable laws protect actions that are, in this way, not publicly justifiable. By contrast, generally applicable laws are publicly justifiable. So claims to such exemptions from generally applicable laws cannot be justified with reference to organizations’ need to perform as accountable members of the moral community: such exemptions do not bolster their ability to give public justifications of their actions.

There is a fourth and final strategy. It starts from the fact that organizations are set up for a particular purpose, to be pursued in a particular way. We saw this when discussing the second strategy. There, I noted that the autonomy-based defense of religious exemptions is inapplicable to organizations, because organizations lack the relevant autonomy. A university, for example, cannot consider giving up the goals of teaching and research. Those goals are fundamental to its decision-making. More generally, an organization cannot decide to perform an action if its decision-making procedures, and fundamental goals, render it unable to decide to perform that action. 

Building on this, I suggest we conceive of religiously grounded exemptions as liberty-rights, rather than claim-rights: religiously grounded exemptions amount to the lifting of a legal duty to perform some action (the action of abiding by the generally applicable law), rather than amounting to the presence of a legal duty (held by an entity other than the right-bearer) to respect the content of the right.28 Most members of society have a duty to abide by the generally applicable law. Any entity that has an exemption lacks that duty. When exemptions are thus framed as absences of duties, it is easy to see how they might be justified. Simply, a duty to perform an action implies that the duty-bearing entity has the ability to perform that action: “ought” implies “can.” By contraposition, if an entity lacks the ability, then it lacks the duty. Thus, if an organization’s fundamental goals or decision-making procedures render it unable to abide by a generally applicable law, then it cannot have a duty to abide by that law. Thus, it must be granted a liberty-right (an absence of a duty) regarding that law: an exemption from the duty to abide by it.

The question is under what conditions an organization’s procedures and goals render it constitutionally unable to abide by a law. When assessing this, we should not simply take organizations at their word. After all, a school with a religious character might suddenly find itself able to abide by antidiscrimination laws if its funding becomes conditional on its doing so.29 In this way, organizations might misunderstand their own constitutional inabilities. 

This suggests a test for organizational abilities: would the organization abide by the general law if it were given an incentive for doing so? If yes, then we should reject any assertion that it is constitutionally incapable of abiding. This follows political theorist Zofia Stemplowska’s account of feasibility, according to which “motivational failure is an instance of mere unwillingness when there exists a conceivable incentive that would bring the agent’s motivational state in line with what is needed to perform the action in question.”30 By contrast, if there is no incentive that could induce an organization to abide by the generally applicable law, then we should take seriously its claim to be unable to abide.

Morally speaking, it is important that the incentives are not threats.31 To ensure this, the offered incentive must not infringe upon the organization’s rights (here referring to rights other than the right to religious exemptions). I assume these other rights are antecedently given, for example, via Smith’s strategy discussed earlier. Thus, I assume organizations do have some rights, including claim-rights and liberty-rights. My argument takes no stand on how these nonreligious rights are justified or what their content is. The argument so far has concerned rights to religious exemptions only. When deciding whether an organization has the specific liberty-right to a religious exemption, we should offer the organization an incentive that does not infringe its rights that are not religious exemptions.

This introduces a temporal dimension to organizations’ religious exemptions. After all, an organization may be unable now to abide by some law, while being able now to take steps to make itself able at a later time. That is, it might have the “diachronic ability” to abide by the law, while lacking the “synchronic ability.”32 Regarding antidiscrimination laws, for example, one might think of Christian churches’ shifting perceptions of women and LGBT+ people: while it might be plausible now for an educational institution with a religious character to claim that it is constitutionally incapable of making the decision to employ a trans person, any such claim will become less plausible as more churches slowly liberalize their attitudes toward homosexuality. What’s more, such changes often happen in an unofficial way: not through decrees of leaders, but through changing practices and norms among the foot soldiers of the organization, as I mentioned when characterizing organizations’ agency. If an organization can render itself able to abide by some law, then its exemption might legitimately be temporally constrained. Such organizations might be required to review their approach to the generally applicable law, with the exemption in turn being reviewed every five or ten years. This prevents “perverse incentives” whereby organizations are given license to avoid the law by constituting themselves unable to abide by the law.

Another constraint on this strategy derives from individuals’ moral duties. As emphasized above, an organization’s procedures and fundamental goals are conceptually–and often substantively–different from members’ procedures and goals. If a collective’s duty is ruled out due to its constitutional constraints, then members may have moral obligations to act upon the collective from the outside with the aim of revising the constitution. By from the outside I mean acting beyond what is mandated by their role within the organization. Of course, members might also have moral obligations to act within their role to change the constitution. But such internal actions are best construed as constituting actions of the organization itself, and therefore conceived of as the exercise of the organization’s diachronic ability to abide by the law. By contrast, actions from the outside may become morally necessary when the organization is both synchronically and diachronically unable to abide by the law. Neither the internal nor the external actions of members are likely to be strictly enforceable by law, due to their demandingness and potential infringement of individuals’ basic liberties. But, notably, the nonenforceability of such obligations does not derive from the organization’s claim to have its religious convictions respected. And if members face moral-political pressure to fulfill such obligations, then the organization may well find itself able to abide by the law after all, thus dissolving its liberty-right not to abide.

This fourth strategy might appear overly permissive, insofar as its rationale extends beyond religious organizations. For example, can a white supremacist organization assert its inability to abide by antiracism laws because its constitution is racist? I make two points in response. First, I have sought to find a plausible justification for existing laws that provide religiously grounded exemptions to organizations. If that justification extends beyond religious organizations to other (more sinister) organizations, this does not show that the law should be changed to allow exemptions to the latter organizations. Second and more important, even if the fourth strategy does apply beyond religious organizations, some procedures and fundamental goals are beyond the democratic pale. Plausibly, religiously grounded exemptions apply only to those that are within the pale. The pale might be set in various ways, such as with reference to a harm principle or to basic liberal rights. But it will rule out certain organizations as impermissible, even before those organizations’ exemptions can arise as a political question.

Where does this leave us? Consider again the Australian law: religious educational institutions may discriminate against potential staff members, contract workers, or students on the basis of sex, sexual orientation, gender identity, marital or relationship status, or pregnancy. This is not justified by an organization having a claim of its own that is transferred up from the claims of members (the first strategy). Nor should we view the exemption as protecting the autonomy or identity of the organization itself (the second strategy). Neither is the exemption necessary for the accountability of the organization (the third strategy). Perhaps members have claims not to be involved in the hiring or teaching of people, because of those people’s sex, sexual orientation, gender identity, marital or relationship status, or pregnancy. This essay has not sought to assess that idea. By looking directly to that possibility, we avoid giving members’ claims more weight than they deserve, by imbuing them with the size, power, and longevity of the organizational entity. When members’ claims are balanced against those of potential staff members, contract workers, or students, the latter may well win. But this is a matter of balancing individuals’ claims: it is not a matter of a claim held by the organization itself.

That said, there may be some cases in which religiously grounded exemptions are justified with reference to the organization itself. These cases fall under the fourth strategy, in which an organization’s procedures or foundational goals prevent it from being able to abide by the generally applicable law, thus preventing it from having a duty to so abide. To test whether this strategy can legitimately be taken by Australia’s religious educational institutions, I proposed an incentive test: would sticks and/or carrots suffice to induce compliance with nondiscrimination laws? Even when the answer is no, such that the fourth strategy can be taken, that strategy is unlikely to last: organizations will often have the long-term (if not short-term) ability to abide by the general law, and members will often have a moral duty to bring such an ability into existence if it does not yet exist. The result is that religious exemptions for organizations should be few and far between.

author’s note

For helpful feedback, I thank all participants at the “Religion and Democracy” seminar funded and hosted by the Australian Catholic University, and particularly Clive Cookson for his commentary at that seminar. I also thank participants at the 2019 Australasian Workshop in Moral Philosophy (Australian National University) and the 2019 Australasian Political Theory and Philosophy Conference (University of Canberra).

Endnotes

• 1United Kingdom Government, Department for Transport, “The Highway Code: Rules for Motorcyclists,” updated August 20, 2019.
• 2John Masanauskas, “Muslim Women Thrive with Pool Man Ban,” news.com.au, July 23, 2012.
• 3Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights (Oxford: Clarendon Press, 1995); and Charles Taylor, “The Politics of Recognition,” in Multiculturalism: Examining the Politics of Recognition, ed. Amy Gutmann (Princeton, N.J.: Princeton University Press, 1994).
• 4Chandran Kukathas, The Liberal Archipelago (Oxford: Oxford University Press, 2003); and Brian Barry, Culture and Equality: An Egalitarian Critique of Multiculturalism (Cambridge, Mass.: Harvard University Press, 2001), esp. 39, 167.
• 5Lee (Respondent) v. Ashers Baking Company Ltd and Others (Appellants) (Northern Ireland)[2018] UKSC 49.
• 6Australian Legal Information Institute, “Sex Discrimination Act 1984–Sect 38: Educational Institutions Established for Religious Purposes.”
• 7Geoffrey M. Hodgson, “Institutions and Individuals: Interaction and Evolution,” Organization Studies 28 (1) (2007): 97–116.
• 8Stephanie Collins, Group Duties: Their Existence and Their Implications for Individuals (Oxford: Oxford University Press, 2019).
• 9Kymlicka, Multicultural Citizenship, 76.
• 10Ibid., 83.
• 11Ibid., 90.
• 12Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1988), 309.
• 13Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982), 150–165, 100.
• 14Ibid., 226. See also Alastair MacIntyre, After Virtue (Notre Dame, Ind.: University of Notre Dame Press, 1981).
• 15Robert Audi, “Church-State Separation, Healthcare Policy, and Religious Liberty,” Journal of Practical Ethics 2 (1) (2014): 1–23, 5.
• 16Ibid., 5.
• 17Kymlicka, Multicultural Citizenship, 93.
• 18John Rawls, A Theory of Justice, rev. ed. (Cambridge, Mass.: Belknap, 1999), 491, 131.
• 19Eric Schwitzgebel has argued that organizations are phenomenally conscious; see Eric Schwitzgebel, “If Materialism Is True, the United States is Probably Conscious,” Philosophical Studies 172 (2015): 1697–1721. But Christian List provides a compelling rejoinder; see Christian List, “What Is It like to Be a Group Agent?” Nous 52 (2) (2018): 295–319.
• 20Daniel Dennett, The Intentional Stance (Cambridge, Mass.: The MIT Press, 1996).
• 21Christian List and Philip Pettit, Group Agency: The Possibility, Design, and Status of Corporate Agents (Oxford: Oxford University Press, 2011), 171.
• 22Ibid., 174.
• 23Leonie Smith, “The Curious Case of Ronald McDonald’s Claim to Rights: An Ontological Account of Differences in Group and Individual Person Rights,” Journal of Social Ontology 4 (1) (2018): 1–28.
• 24Ibid., 17.
• 25Ibid., 24.
• 26Ibid., 26.
• 27This term was made most famous by Rawls, though he claims the duty to give justifications in terms of public reasons applies only in the public political forum of courts and legislatures, not in the “background culture” of a society. John Rawls, Political Liberalism (New York: Columbia University Press, 1996), esp. 220.
• 28That distinction was conceptualized in Wesley Hohfeld, Fundamental Legal Conceptions (New Haven, Conn.: Yale University Press, 1919).
• 29I thank Jeanette Kennett for this point.
• 30Zofia Stemplowska, “Feasibility: Individual and Collective,” Social Philosophy and Policy 33 (1–2) (2016): 273–291, 280.
• 31I thank Robert Audi, Colleen Murphy, and Paul Weithman for pressing this.
• 32This terminology comes from Mark Jensen, “The Limits of Practical Possibility,” Journal of Political Philosophy 17 (2) (2009): 168–184.