1. See, e.g., Corp. of Presiding Bishop of Church of Jesus Christ of Latter-day Saints v. Amos, 483 U.S. 327, 339 (1987) (holding that a custodial worker at a religiously affiliated gymnasium could be required to convert to Mormonism to keep his job); Edmondson v. Shearer Lumber Prods., 75 P.3d 733, 739 (Idaho 2003) (holding that an employer could lawfully fire an employee for opposing a national forest development project).

2. See 26 A.L.R.3d §1333 (2018) (organizing legal findings of insubordination under a heading entitled “Profane language or backtalk”).

3. Karl Marx, On the Jewish Question, in Early Writings 219 (Rodney Livingstone & Gregor Benton trans., 1975).

4. For arguments about the centrality of bargaining power to labor law theory and jurisprudence, see generally P. L. Davies & M. R. Freeland, Kahn-Freund's Labour and the Law 18 (3d ed. 1983); Collins, Hugh, Lester, Gillian & Mantouvalou, Virginia, Introduction: Does Labour Law Need Philosophical Foundations?, in Philosophical Foundations of Labour Law (Hugh Collins, Gillian Lester & Virginia Mantouvalou eds., 2018)CrossRefGoogle Scholar; Langille, Brian, Labour Law's Theory of Justice, in The Idea of Labour Law 101–19 (Guy Davidov & Brian Langille eds., 2011)CrossRefGoogle Scholar; Ayres, Ian & Schwab, Stewart, The Employment Contract, 8 Kan. J. L. & Pub. Pol. 71 (1999)Google Scholar. For critical views of bargaining power theories of labor and employment law, see, e.g., Collins, Hugh, Is the Contract of Employment Illiberal?, in Philosophical Foundations of Labour Law 52 (Hugh Collins, Gillian Lester & Virginia Mantouvalou eds., 2018)CrossRefGoogle Scholar; Langille, Brian, What Is Labour Law? Implications of the Capabilities Approach, in The Capabilities Approach to Labour Law 122 (Brian Langille ed., 2019)CrossRefGoogle Scholar.

5. Labor republicanism is an exception in its general focus on objectionable hierarchy in employment. See infra note 98. Marxist critiques of work are similarly exceptional, but the substantive question of what we may reasonably ask of one another at work would arise even if production were completely socialized. Nevertheless, several of the major themes in this article—of mutual recognition, of not subordinating our moral lives to our working lives—are Marx-inspired, though grounded in liberal philosophy.

6. The paid workplace is of course not the only site in which we perform morally and socially significant work. We also make social, emotional, cultural, and intellectual contributions through participation in the family, volunteer work, membership in churches, political participation, and through participation in voluntary associations such as clubs.

7. John Rawls, Justice as Fairness: A Restatement (Erin Kelly ed., 2001), at 4.

8. By “legal norm” I refer inclusively to both legal rules and legal principles that guide the development and application of legal rules. See, e.g., Bhasin v. Hrynew, 3 SCR 494, para. 64 (2014) (explaining that the principle of good faith in performance is an “organizing principle” in contract law and, as such, “is not a free-standing rule, but rather a standard that underpins and is manifested in more specific legal doctrines and may be given different weight in different situations”); R. M. Dworkin, Is Law a System of Rules?, in The Philosophy of Law 38 (R. M. Dworkin ed., 1977).

9. See, e.g., Hanoch Dagan & Avihay Dorfman, Justice in Private: Beyond the Rawlsian Framework, 37 Law & Phil. 171 (2018); Sophia Moreau, What Is Discrimination?, 38 Phil. & Pub. Aff. 143, 144–46 (2010).

10. By “ability” I mean a realized capacity.

11. In the overwhelming majority of U.S. jurisdictions, an essential part of what it is, legally, to be an employee is to labor under the control of an employer. See, e.g., 20 A.L.R. 684, §1. New Jersey and California are exceptions. See, e.g., Hargrove v. Sleepy's, 106 A.3d 449 (N.J. 2015) (adopting the “ABC test,” according to which an individual is presumed to be an employee unless all three of the following are met: “(1) the employer neither exercised control over the worker, nor had the ability to exercise control in terms of the completion of the work; (2) the services provided were either outside the usual course of business or performed outside of all the places of business of the enterprise; and (3) the individual has a profession that will plainly persist despite termination of the challenged relationship”); Dynamex Operations W. v. Superior Court, 416 P.3d 1 (Cal. 2018) (adopting the ABC test for state wage orders); California Assembly Bill (AB) 5 (effective Jan. 2020) (codifying and extending the ABC test to all California Labor Code wage and hour violations). The British and Canadian paradigms of employment are similarly hierarchical. See, e.g., Laws v. London Chronicle (Indicator Newspapers) Ltd., 1 WLR 698 (1959) (explaining that “willful disobedience of an order” violates an “essential” condition of the employment relationship, “namely, the condition that the servant must obey the proper orders of the master”); Stein v. British Columbia Housing Management Commission, 65 B.C.L.R. (2d) 181 (BC CA 1992) (“The employer is the boss and it is an essential implied term of every employment contract that . . . the employee must obey the orders given to him.”); Hugh Collins, Employment Law (2d ed. 2010), at 10 (“The paradigm of the employment contract thus contains an authority structure at its heart.”).

12. See John Rawls, A Theory of Justice (1971), at 19, 505–10.

13. See id. (explaining that people are moral equals “as moral persons, as creatures having a conception of their good and capable of a sense of justice”).

14. Seana Valentine Shiffrin, Speech Matters: On Lying, Morality, and the Law (2014), at 68–69.

15. See, e.g., Rawls, supra note 7, at 23–24.

16. See id.

17. As opposed to freedom of the will.

18. John Rawls, Political Liberalism (1996), at 293.

19. As Jonathan Quong and Seana Shiffrin have argued, paternalism is also a paradigmatic way we can fail to treat each other as moral equals. See, e.g., Jonathan Quong, Liberalism Without Perfection (2011), at 101–06; Seana Valentine Shiffrin, Paternalism, Unconscionability Doctrine, and Accommodation, 29 Phil. & Pub. Aff. 205, 218 (2000).

20. See Jeremy Waldron, Can There Be a Democratic Jurisprudence?, 58 Emory L.J. 675 (2008) (describing the “public character of the law” as including its ability to “deal with matters . . . in a way that can stand in the name of public”); Seana Valentine Shiffrin, Speaking Amongst Ourselves: Democracy and Law, in 37 The Tanner Lectures on Human Values 150–52 (Mark Matheson ed., 2018) (arguing that social affirmation of our moral equality is often part of what it is to treat one another as equal members of our society). For a skeptical view about the law's communicative potential, see, e.g., Barbara Baum Levenbook, The Meaning of a Precedent, 6 Legal Theory 185 (2000).

21. See, e.g., Quong, supra note 19, at 316 (“By offering each person certain fundamental rights and liberties, liberalism affirms citizens’ moral right to direct their own lives consistent with a similar right of others.”); Rawls, supra note 18, at Lecture VIII, §3.

22. While Dagan and Dorfman are skeptical of distributive justice having much, if anything, to say about how private parties ought to relate to one another, they nevertheless argue that liberal justice generates requirements of “relational justice” to realize substantive equality. See generally Dagan & Dorfman, supra note 9. Seana Shiffrin also contends that distributive justice does not settle all that we owe one another as equal moral agents in our legal arrangements. See, e.g., Seana Valentine Shiffrin, The Divergence of Contract and Promise, 120 Harv. L. Rev. 708, 716–17 (2007).

23. Private employees in the United States cannot call upon the protection of the First Amendment to engage in political expression without fear of employer retaliation. See, e.g., Edmondson v. Shearer Lumber Prods., 75 P.3d 733, 739 (Idaho 2003). Some local jurisdictions, such as Connecticut, have nevertheless extended free speech protections to private employees by statute. See Conn. Code §31–51q (West 2005) (granting private employees the same free speech rights as public employees under federal and state constitutional law and creating a private cause of action for damages for violations of those rights); infra note 81.

Even though the First Amendment is a source of protection for public employees’ political expression, public employees’ rights to engage in such expression are highly circumscribed. Public employee speech is often not protected if it is about work, especially if the speech is perceived as a response to an interpersonal “controversy with [the employee's] superiors.” Connick v. Myers, 461 U.S. 138, 147–48, 151–52 (1983). When public employees speak pursuant to their professional duties, they are likewise not protected from employer retaliation. See Garcetti v. Ceballos, 547 U.S. 410, 422 (2006) (holding that a district attorney “did not speak as a citizen by writing a memo that addressed the proper disposition of a pending criminal case” because he was speaking in his capacity as a “government employee”). For a defense of the limits of First Amendment protections for public employees, see Robert Post, Participatory Democracy and Free Speech, 97 Va. L. Rev. 477, 482–86 (2011) (arguing that public employers must be able to exercise substantial control over speech in and about the “managerial domain” in order to effectively implement public policy). For criticism of this jurisprudence, see, e.g., Shiffrin, supra note 14, at 208–10; Cynthia L. Estlund, Speech on Matters of Public Concern: The Perils of an Emerging First Amendment Category, 59 Geo. Wash. L. Rev. 1, 37–39 (1990).

24. See Rawls, supra note 12, at 224–27 (arguing that distributive justice requires not just the formal equality of political liberties, but also their equal worth); Samuel R. Bagenstos, Employment Law and Social Equality, 112 Mich. L. Rev. 225, 256 (2013) (arguing that employer control over employees’ off-duty political speech is morally objectionable because it results in “a skewed political discourse” where “employers’ voices are amplified and workers’ are squelched”); cf. Joseph Fishkin & William E. Forbath, The Anti-Oligarchy Constitution, 94 B.U.L. Rev. 669 (2014) (arguing that a commitment to constitutional democracy requires legislative action to prevent social and economic inequality from coalescing into oligarchy). There is no state action doctrine in liberalism. Securing social conditions of equal liberty also requires regulating the interactions of private individuals. See, e.g., Rawls, supra note 12, at 203 (“Not only must it be permissible for individuals to do or not to do something, but government and other persons must have a legal duty not to obstruct.”) (emphasis added).

25. For arguments for the agential importance of political liberties, see, e.g., Rawls, supra note 18, at 293–99; Frederick Schauer, Free Speech and Obedience to Law, 32 Const. Comment. 661 (2017) (“[F]reedom of political speech is normatively justifiable as a necessary component of a normatively justifiable form of governance in which citizens have substantial input by voting and otherwise into the decisions that will affect them and that will control their activities.”). It does not, however, follow that all legal norms permitting an employer to regulate employees’ moral and political expression are objectionable. See infra Sections II.B and III.C (discussing possible grounds for employers to dismiss employees for communicating inegalitarian beliefs).

26. See 26 A.L.R.3d §1333 (2018).

27. See, e.g., 19 Williston on Contracts §54:23 (4th ed. 2018); Collins, supra note 11, at 9–11.

28. The legal concept of “cause” is still relevant even when employment is at will, as it often is in the United States. For example, being terminated for cause often precludes an employee from securing unemployment benefits. See, e.g., In re Stergas, 673 N.Y.S.2d 223, 223 (1998) (finding that a video store employee's refusal to keep his hair shorter than two inches above his collar per company policy was cause for firing him and denying him unemployment benefits).

29. See 26 A.L.R.3d §1333 (2018).

30. See P. F. Strawson, Freedom and Resentment, in 48 Proceedings of the British Academy 1, 1–25 (1962) (arguing that reactive attitudes of, for instance, resentment are natural and morally healthy responses to ill will or the unjustified indifference of others, and are ways in which we interact with one another as morally responsible beings).

31. See Laws 649d-e, 863b, 869a, 934d.

32. Martha C. Nussbaum, Anger and Forgiveness: Resentment, Generosity, Justice (2016).

33. See R. Jay Wallace, Trust, Anger, Resentment, Forgiveness: On Blame and Its Reasons, 27 Eur. J. Phil. 537, 540–42 (2019).

34. On the moral power of apology, see generally Jeffery S. Helmreich, The Apologetic Stance, 43 Phil. & Pub. Aff. 75 (2015).

35. Even if the wrongdoer realizes that she wronged someone, refusing to share indignation with the wrongdoer can deprive her of essential information needed to take steps to repair the relationship. Anger is often an authentic indication of how seriously a person feels she has been wronged and can accordingly provide the wrongdoer with information needed to determine how best to respond. Perhaps a longer discussion is needed before a meaningful apology can be issued, or perhaps a promise to undertake some kind of personal change is needed. Even when a person's anger is not warranted, the communication of that anger can prompt discovery of points of moral disagreement upon which agreement urgently needs to be sought. Communication of indignation, when directed to a believed wrongdoer, can thereby support moral progress and lend stability to relationships going forward.

36. For a discussion of warrant to blame, see Pamela Hieronymi, The Force and Fairness of Blame, 18 Phil. Persps. 115 (2004).

37. 108 F.3d 789 (7th Cir. 1997).

38. Id. at 792–93.

39. Id. at 792–94.

40. Id. at 794.

41. Id. at 794, 799 (holding that Indiana Sugars had a good faith belief that McClendon was insubordinate and, thus, that firing McClendon was not a pretext for acting on discriminatory motives). It may be objected that I should not use cases of discrimination to illustrate a more general wrong of failing to treat workers as equal moral agents. Discrimination is, after all, surely a distinctive kind of wrong. Yet we can recognize that fact while also noticing that examples of discrimination are among our most compelling and carefully studied paradigms of objectionable hierarchy. It would consequently be a mistake to inquire into the nature of objectionable workplace hierarchy with a methodology that takes these paradigms off the table. The mistake appears all the more egregious when one considers whose experiences we would risk treating as paradigmatic were we to try to evaluate workplace hierarchy by analyzing away any potential discrimination: those of white, heterosexual men in the workforce. Cf. Kimberlé Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, U. Chi. Legal F. 139 (1989) (arguing that treating the experiences of white men as a theoretical baseline for equality encourages a discrimination analysis along single axes of race, gender, and the like that results in the theoretical erasure of Black women).

42. There is, of course, much more to be said about McClendon. In particular, the duty of obedience may have been deployed as a pretext for discrimination. For a discussion of such problematic uses of the duty of obedience, see generally Susan D. Carle, Angry Employees: Revisiting Insubordination in Title VII Cases, 10 Harv. L. & Pol'y Rev. 185 (2016).

43. Similar servility objections apply to a lack of legal protection from termination for communicating indignation at clientele. Consider, for example, Frank Rollins's experience as a Pullman Porter. At the end of the American Civil War, the Pullman Company recruited former slaves and, later on, Southern Black workers to serve its luxury sleeper cars. See National Public Radio, “Former Pullman Porter Subtly Confronted Racism” (May 8, 2009), https://www.npr.org/templates/story/story.php?storyId=103945861. Frank Rollins, who worked as a porter for Pullman in the 1940s, recalls that the company preferred “Southern boys,” believing that “they had a certain personality and certain demeanor that satisfied the Southern passengers better than the boys who came from Chicago.” Id. Rollins's job training specially addressed how to respond to discriminatory remarks by clientele. Id. Rollins recalls his trainer explaining,

Id. (quotation marks omitted). Similar to McClendon's experience, Rollins was asked to submit to subordinating relationships as a condition of his employment, and that subordination was accomplished in part through a requirement that Rollins refrain from communicating indignation to individuals in response to those individuals’ wrongful treatment of him. The moral significance of communication is underscored by the transformative character of Rollins's response. Unwilling “to be humiliated,” Rollins eventually started greeting new passenger cars with the following:

Id.

44. See Stephen Nayak-Young, Revising the Roles of Master and Servant: A Theory of Work Law, 17 U. Pa. J. Bus. L. 1223 (2015). Whereas Nayak-Young argues that status inequality in work law is distinct from contract law, I argue here that that status inequality is constituted in part by implied contractual duties of obedience and loyalty. See id. at 1238–51. Nayak-Young also seeks to justify status inequality in work law, whereas I argue that we should be skeptical of such a conceptualization of workplace relations. See id. at 1252–56. For a similarly critical stance, see generally Aditi Bagchi, Exit, Choice and Employee Loyalty, in Contract, Status, and Fiduciary Law 271–92 (Paul B. Miller & Andrew S. Gold eds., 2016).

45. Collins, supra note 11, at 34; see also Alan Fox, Beyond Contract: Work, Power and Trust Relations (1974), at ch. 4.

46. Collins, supra note 11, at 34.

47. Similar to Dagan and Dorfman, I argue here that employers sometimes need to absorb the costs of, and hence accommodate, their employees’ agential activity. Whereas Dagan and Dorfman focus on the needs to accommodate employees’ traits and choices to pursue various conceptions of the good, such as religion, I focus on the need to accommodate expressive activities that are partially constitutive of egalitarian communication. See Dagan & Dorfman, supra note 9, at 190–94. The principles of accommodation I present here are thus distinct from principles of accommodation in employment discrimination law. Further, while Dagan and Dorfman present their principles of accommodation as legal duties, the requirements of accommodation I argue for here are not themselves legal duties, but rather should be implemented as second-order legal principles to guide the development of legal doctrine. On the possibility of such second-order legal principles, see supra note 8.

48. Shiffrin, supra note 14, at 114.

49. See Restatement (Second) of Contracts §§237, 241 (Am. L. Inst. 1981).

50. See id. at §241. The doctrine of material breach resembles but should be distinguished from the Anglo-Canadian doctrine of repudiatory breach. See, e.g., Hong Kong Fir Shipping Co. v. Kawasaki Kisen Kaisha Ltd., 2 Q.B. 26 (1962) (holding that a breach is repudiatory and hence entitles the aggrieved party to bring the contract to an end when the breach deprives the party of “substantially the whole benefit” of the contract). Whereas repudiatory breach automatically gives rise to a right in the aggrieved party to terminate the contract, material breach requires that the aggrieved party give the party in breach a chance to “cure” their defective performance. See Restatement (Second) of Contracts, supra note 49, at §237; John D. McCamus, The Law of Contracts (2d ed. 2012), at 685–86.

51. See Restatement (Second) of Contracts, supra note 49, at §237.

52. Cf. infra text accompanying notes 86–92 (discussing efforts to contractually silence employees’ moral criticism through nondisclosure agreements and secret arbitration).

53. The irreparable breach rule is in some respects both broader and narrower than the warranted indignation approach. The former rule permits a broader range of expression because the mere fact that indignation is not warranted does not deprive it of protection. Rather, protection turns on the degree to which the communication (or course of communication) is injurious to ongoing trust and cooperation. The irreparable breach rule is also for this reason potentially narrower than the warranted indignation approach, for it is at least in theory possible that the communication of warranted indignation could be destructive of ongoing trust and cooperation. More often, it will likely be the case that when warranted indignation compromises ongoing trust and cooperation, it is not in the first instance because of the communication of that indignation, but rather because of the act by the employer that occasions the indignation. Although I cannot fully develop the point here, the irreparable breach rule points to a ground for employees to have a symmetrical right to end an employment relationship—and thus a claim for wrongful (constructive) dismissal—when a supervisor's communications irreparably damage ongoing trust and cooperation. Such an application of the irreparable breach rule thus resembles the U.K. implied duty of mutual trust and confidence. See, e.g., Bournemouth University v Buckland [2010] EWCA Civ 121 (upholding a tribunal's decision that a professor-complainant had been constructively dismissed when his employer accepted another professor's remarking of the complainant's exams and thereby “destroy[ed] the relationship of trust and confidence which is implicit in all employment contracts”).

54. See, e.g., McKinley v. BC Tel, 2001 SCC 38; Henry v. Foxco Ltd, 2004 NBCA 22; Donovan v. New Brunswick Publishing Co., 1996 CanLII 4832 (NB CA); Bennett v. Cunningham, 2006 CanLII 37516 (ON SC). The irreparable breach rule is also similar to the U.K. rule requiring a fundamental breach of contract to warrant summarily dismissing an employee without reasonable notice. See, e.g., Wilson v. Racher [1974] ICR 428 (CA) (holding that the owner of an estate wrongfully dismissed his head gardener when he summarily fired the gardener for telling him to “Go and shit yourself” in response to a barrage of abusive and unreasonable criticism, and noting that a “Czar-serf” relationship no longer animates service contracts); see also Collins, supra note 11, at 164–65.

55. Henry, 2004 NBCA 22, at paras. 1, 5.

56. Id. at paras. 39–41

57. Id. at para. 38.

58. Id. at para. 40.

59. Id. at para. 40.

60. Id. at paras. 40, 129.

61. Id. at para. 40.

62. Id. at para. 130.

63. Id. at para. 130.

64. Id. at para. 5. The Court of Appeal ultimately deferred to the trial court's finding that Henry's recounting of the facts were less credible than Foxco's. See id. at paras. 5, 26.

65. Harris v. Forklift Systems, Inc., 510 U.S. 17, 22 (1993).

66. Harris, 510 U.S. at 22.

67. See Shiffrin, supra note 14, at 9.

68. See id. It may be objected that the legal protection of indignant expression argued for in Section III is in tension with this point. How to communicate indignant expression in respectful ways is surely something we rely on one another for learning, and it may be argued that the status quo, which gives employers broad latitude to terminate employees for indignant expression, furthers this end. In response, first, it should be noted that the irreparable breach rule does not preclude an employer—or colleagues—from telling the inappropriately indignant employee that she is wrong, rude, or offensive. Second, the rule actually does serve to establish shared norms of respectful communication. The rule communicates that what makes indignant expression problematic in zones of public life is not a matter of the particular ethical views of the parties, but rather a matter of whether such communication compromises our ability to work together on terms of equality. In contrast, the status quo risks publicly communicating a false, problematic ground for norms of respectful communication: namely, that an employer is owed respect qua employer, and that that respect is not equally owed to the employee.

69. See supra note 35.

70. See Seana Valentine Shiffrin, What Is Really Wrong with Compelled Association?, 99 Nw. U. L. Rev. 839, 862, 866, 869–70 (2005).

71. Marion Crain, Managing Identity: Buying into the Brand at Work, 95 Iowa L. Rev. 1179, 1201 (2010) (quoting Pierre Berthon, Michael Ewing & Li Lian Han, Captivating Company: Dimensions of Attractiveness in Employer Branding, 24 Int'l J. Advert. 151, 153–54 (2005)).

72. See id. at 1220–32.

73. See id. at 1209–13.

74. Id. at 1212 (“‘The real secret to Southwest's marketing is its almost religious fervor to maintain and perpetuate the core values of the [corporate] culture.’ Southwest's philosophy is that employment at the airline is not a job, it's a ‘crusade.’”) (quoting Kevin Freiberg & Jackie Freiberg, Nuts! Southwest Airlines’ Crazy Recipe for Business and Personal Success (1996), at 10, 267).

75. In the United States, government employers generally may not take adverse employment actions against their employees on the basis of their political affiliation; such actions are presumptively unconstitutional under the First Amendment. See Rutan v. Republican Party of Ill., 497 U.S. 62 (1990). But that presumption does not apply where “party affiliation is an appropriate requirement for the effective performance of the public office involved,” such as for a presidential speechwriter. Branti v. Finkel, 445 U.S. 507, 518 (1980); see also Elrod v. Burns, 427 U.S. 347, 366 (1976).

76. Cf. Edmondson v. Shearer Lumber Prods., 75 P.3d 733, 739 (Idaho 2003).

77. See Jespersen v. Harrah's Operating Co., 444 F.3d 1104, 1110–11 (9th Cir. 2006) (holding that company policy requiring female casino bartenders to wear makeup did not constitute sex discrimination under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-2 (2012)).

78. Cf. Shiffrin, supra note 70, at 851–73.

79. Id. at 840; but see Boy Scouts of America v. Dale, 530 U.S. 640, 655–56 (2000) (suggesting that the wrong of compelled association has more to do with the ability of the association's group to have control over the message that it outwardly communicates).

80. Making workplaces sites for moral association may, in addition to the reforms discussed here, also require creating legal space for religious workplaces. I address the complexities of creating such legal space while also giving appropriate weight to workers’ rights against discrimination elsewhere, in Disentangling Religion and Public Reason: An Alternative to the Ministerial Exception, 106 Cornell L. Rev. (forthcoming).

81. To be sure, labor law can offer employees a protected opportunity to voice criticism of their wages, vacation time, and other terms and conditions of work. See National Labor Relations Act, 29 U.S.C. § 157 (2012). But that protection is limited by the legal presumption that employers are to have control over their broader aims and philosophy. See Good Samaritan Hosp., 265 N.L.R.B. 618, 626 (1982). Consequently, open criticism of or failure to comport with the aims and policies of one's job usually provides cause for termination. See, e.g., Five Star Transp., Inc., 349 N.L.R.B. 42, 44–45 (2007); Good Samaritan, 265 N.L.R.B. at 626.

82. See, e.g., Kent Greenfield, The Unjustified Absence of Federal Fraud Protection in the Labor Market, 107 Yale L.J. 715, 719–20, 735–38 (1997).

83. See Crain, supra note 71, at 1212. Workplace cultures can also be discriminatory in ways that are difficult to regulate through antidiscrimination law. See Devon W. Carbado & Mitu Gulati, Working Identity, 85 Cornell L. Rev. 1259 (2000).

84. I am indebted to an anonymous reviewer for this example.

85. Cf. Seana Valentine Shiffrin, Egalitarianism, Choice-Sensitivity, and Accommodation, in Reason and Value: Themes from the Work of Joseph Raz 270 (Philip Pettit et al. eds., 2004) (arguing that we may sometimes need to accommodate one another's morally flawed activity to preserve deliberative space to reflect on the values that directly bear on that activity).

86. See David A. Hoffman & Erik Lampmann, Hushing Contracts, 97 Wash. U. L. Rev. 165 (2019).

87. For descriptions of a variety of such contracts, see, e.g., id. at 166–67, 174–77.

88. See id. at 198.

89. Id. at 179.

90. See Erik Encarnacion, Discrimination, Mandatory Arbitration, and Courts, 108 Geo. L.J. 855 (2020).

91. Cf. Rawls, supra note 12, at 334 (explaining that individuals are under a duty of justice to uphold and maintain just institutions).

92. See supra Section II.B.

93. On the moral importance of shared free time, see generally Julie L. Rose, Free Time (2016).

94. See, e.g., Brian Langille, Human Development: A Way out of Labour Law's Fly Bottle, in Philosophical Foundations of Labour Law 87 (Hugh Collins, Gillian Lester & Virginia Mantouvalou eds., 2018). Langille also argues that work law should support the development of human capabilities more broadly. Examples of such capabilities include health, skills, and meaningful opportunities to participate in public and political life. See id. at 91–94. Social conditions for moral agency will thus often overlap with the kinds of capabilities Langille identifies, and so a moral agency approach will typically condemn the same forms of control as a capabilities approach, such as unfettered employer control over employees’ off-duty political activity or workplace policies that compromise employee health. A moral agency approach can complement Langille's theory of work law by supplying a standard for determining when employment unreasonably restricts the exercise of our capabilities.

95. Good Samaritan Hosp., 265 N.L.R.B. 618, 626 (1982).

96. For a skeptical view that exit rights are sufficient to protect the basic liberties of members of voluntary associations, see Les Green, Rights of Exit, 4 Legal Theory 165 (1998).

97. I thus agree with Hugh Collins's point that it is not submission but rather subordination that constitutes the “crucial subversion” of liberal values. Collins, supra note 4, at 63.

98. The liberal principle of treating workers as equal moral agents thus offers an alternative to labor republicanism, which tends to characterize objectionable employment relationships as ones in which employers can arbitrarily interfere in the lives of employees. See, e.g., Alex Gourevitch, From Slavery to the Cooperative Commonwealth (2015), at 11; Iñigo González-Ricoy, The Republican Case for Workplace Democracy, 40 Soc. Theory & Prac. 234 (2014). Of course, it may be that, for labor republicans, the only state of affairs in which employers do not arbitrarily interfere in the lives of employees is one that, among other things, treats people as equal moral agents. Even so, the emphasis and animating ideals remain different: whereas the republican ideal is one of independence, substantive liberal equality, as I have developed it here, is an ideal of cooperative interdependence.

Although a full critique of labor republicanism is not possible here, to highlight a further difference between a moral agency approach and labor republicanism, it may help to point out one kind of difficulty that tends to confront labor republicanism. In order to generate principles for evaluating when employer power is arbitrary, labor republicans often rely on an analogy to the state. For example, Elizabeth Anderson argues that employment relationships are forms of authoritarian governments because employers can exercise legal, economic, and other power over employees and those employees have little say in how that power is exercised. See Elizabeth Anderson, Private Government: How Employers Rule Our Lives (and Why We Don't Talk About It) (2017), at 37–74. Anderson then proposes drawing on our understanding of the just state to develop a “workplace constitution” to restrict employer power. See id. at 65–71; Elizabeth Anderson, Equality and Freedom in the Workplace: Recovering Republican Insights, 31 Soc. Phil. & Pol'y 48 (2015) (arguing that we should draw on a Lockean conception of justice to guide legal reform of employer control). While I am sympathetic, states are relevantly different from workplaces (as Anderson herself notes). See Anderson, supra, at 66–68. For example, it would seem to violate the freedom of speech for the state to ban all discriminatory speech, yet prohibiting a wide range of discriminatory speech in the paid workplace is surely a requirement of justice. Consequently, it is not clear how principles limiting state action are to guide principles limiting employer control over employee expression, and other like substantive dimensions of employment. We need some other ideal, apart from political justice, to develop moral standards for the workplace. For another skeptical view of the use of workplace-state analogies, see Collins, supra note 4, at 60.

99. Rawls, supra note 18, at 293.

100. I am skeptical that bargaining power could ever truly be equalized. Even in a world of widespread social equality, people would still have differences in emotional dispositions, personal needs, and the like that may give some of us greater ability to advance our ends in negotiations than others.

101. Given the existence of such predictable pressures on expression, even if it were possible to somehow get rid of the legal institution of paid work more broadly, and to simply produce through volunteer work, we would still have reason for employment law and hence for employment relationships. Even under such idealized and distant circumstances, some and likely many of us would still need to teach, farm, build shelter, care for one another, and the like. Working together makes possible ways of living that are not dominated by the pressure to survive. Cf. Joseph Raz, The Morality of Freedom (1986), at 374 (describing a fictional situation in which a person is “hounded” by a “fierce carnivorous animal” and “never has a chance to do or even to think of anything other than how to escape from the beast”). Further, as epistemically and morally limited beings, we depend on one another to acquire and develop knowledge, and to create a flourishing culture for realizing and experimenting with various conceptions of happiness and the good life. Producing for moral agency can thus require a lot of labor. If, under such idealized circumstances, we found that, to provide for our needs and interests as moral agents, we needed to work together regularly, for many hours, we would be justified, if not required, to publicly regulate how we work together to prevent the very kinds of predictable and objectionable control—over character, over egalitarian communication—discussed here from arising in our working lives.

102. See, e.g., Patricia Hill Collins, Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment (2d ed. 2000, 2014), at 48–64 (describing how Black women's work after the Civil War has repeatedly recreated relationships of “interpersonal domination” and domestic service reminiscent of slavery and American apartheid); see also Catharine A. MacKinnon, Butterfly Politics (2018), at 110–25 (describing the law's role in supporting and obscuring subordination in employment); Tommie Shelby, Dark Ghettos: Injustice, Dissent, and Reform (2016), at 199 (explaining that the work available to poor Black women is often “domestic service in the homes of affluent white families” that reinforces the “ideological image of the ‘mammy’ . . . used to justify the exploitation and subordination of black women under slavery”).

103. To be sure, the government could make public announcements through, say, its department of labor, to the effect that everyone is an equal in their workplace relationships. But without legal change on the ground, such messages risk being inauthentic and obfuscating.

104. This is in contrast to low-wage workers, workers in nonunionized workplaces, as well as women, Black, and Indigenous workers, and other workers of color or from historically marginalized and oppressed groups more broadly. Even when such individuals occupy prestigious jobs, they may still have reason to believe that because of racism and other forms of bias, their positions are less secure than those of their more socially privileged colleagues.

105. One might also object that the reforms argued for here risk having the perverse consequence of further marginalizing already marginalized and historically disempowered individuals. Giving employees greater protection from termination for their indignant and morally critical expression may lead employers to preempt possible conflicts by hiring on the basis of a variety of pernicious stereotypes, treating race, religion, or other protected statuses as proxies for antisociality.

In response, such employer backlash may simply be an unavoidable part of transitioning to a more egalitarian legal and employment culture. Cf. Bhabha, Faisal, “Islands of Empowerment”: Anti-Discrimination Law and the Question of Racial Emancipation, 31 Windsor Y. B. Access Just. 65 (2013)Google Scholar (“[R]ather than being cause for alarm and retreat of the rights agenda, what is characterized as ‘backlash’ may in fact be part of a longer process of attitude and norm shifting.”). As such, the risk of backlash may caution against an abrupt transition, or may signal that the transition needs to be undertaken in conjunction with other reforms, such as better enforcement of employment rights against indirect discrimination. But given the importance of the proposed reforms to our equal status, it may be that initial backlash is an is an acceptable cost of bringing about a more egalitarian scheme of labor and employment.

106. See, e.g., Gourevitch, Alex, The Limits of a Basic Income: Means and Ends of Workplace Democracy, 11 Basic Income Stud. 17 (2016)CrossRefGoogle Scholar; Landemore, Hélène & Ferreras, Isabelle, In Defense of Workplace Democracy: Towards a Justification of the Firm-State Analogy, 44 Pol. Theory 53 (2016)CrossRefGoogle Scholar; supra note 98.