1. In this paper, I use the notions of obligation and duty interchangeably. Similarly, I also use the notions of directive and command interchangeably.

2. Lefkowitz, David, The Duty to Obey the Law, 1 Phil. Compass 571, 572–573 (2006)CrossRefGoogle Scholar (emphasis original).

3. The idea of content-independence was introduced in H. L. A. Hart, Essays on Bentham: Studies in Jurisprudence and Political Theory (1982), ch. 10; see also Joseph Raz, The Morality of Freedom (1986), ch. 2. As John Gardner has argued, the notion of content-independence is potentially misleading. A better label would be “merit-independence”: the obligatoriness of a certain action stems not from the merits of that action, but from the source of the requirement to perform it (in this case, the law). Gardner, John, Legal Positivism: 5½ Myths, 46 Am. J. Juris. 199, 209 (2001)CrossRefGoogle Scholar. In the main text, I will sometimes use the “merit” terminology. Note that, in the paper just cited, Gardner discusses merit-independence in relation to legal validity rather than moral obligatoriness. Still, that notion carries over to the domain of moral obligation.

4. For related discussions, which also highlight ambiguities in the idea of content-independence, see Sciaraffa, Stefan, On Content-Independent Reasons: It's Not in the Name, 28 Law & Phil. 233 (2009)CrossRefGoogle Scholar; Markwick, Paul, Law and Content-Independent Reasons, 20 Oxford J. Legal Stud. 579 (2000)CrossRefGoogle Scholar.

5. One may wonder whether content-independence is indeed a necessary feature of the obligation to obey the law “proper.” In line with much of the literature, the present paper is developed under the assumption that it is. Skeptics about content-independence as a feature of political obligation are thus unlikely to be moved by my arguments. For reviews of the literature on political obligation and authority that state the content-independence condition, see, e.g., Lefkowitz, supra note 2; Edmundson, William A., State of the Art: The Duty to Obey the Law, 10 Legal Theory 215 (2004)Google Scholar; Christiano, Thomas, Authority, in The Stanford Encyclopedia of Philosophy (Zalta, Edward N. ed., 2013)Google Scholar, http://plato.stanford.edu/archives/spr2013/entries/authority/. For a recent defense of content-independence, see Adams, N. P., In Defense of Content-Independence, 23 Legal Theory 143 (2017)CrossRefGoogle Scholar.

6. For some, the obligation to obey is owed to the authority (i.e., it is correlative to the authority's right), and the authority may then hold those who disobey to account for their failures. Here, I remain agnostic about this feature of authority. My arguments apply whether the obligation to obey is taken to be correlative to a right or not. For defenses of the rights-correlativity of authority-imposed duties, see Darwall, Stephen, Authority, Accountability, and Preemption, 2 Juris. 103 (2011)CrossRefGoogle Scholar; Hershovitz, Scott, The Authority of Law, in The Routledge Companion to Philosophy of Law (Marmor, Andrei ed., 2012), at 65–75Google Scholar.

7. On this, see Hershovitz, supra note 6, at 69.

8. Id. at 69. Hershovitz uses the language of “grounds” instead of “justification.” For the purposes of my discussion, these should be seen as equivalent. I employ the expression “justifying political obligation” meaning “explaining what grounds political obligation.”

9. Interpretation (1) is in line with how, e.g., Leslie Green understands the content-independence of authority-imposed obligations: as placing constraints on the reasons that should guide their addressees’ practical deliberations and actions. See Leslie Green, The Authority of the State (1990), at 41–42, 225. For critical discussion of Green's interpretation of content-independence, see Perry, Stephen, Political Authority and Political Obligation, in Oxford Studies in Philosophy of Law 2 (Green, Leslie & Leiter, Brian eds., 2013), at 13–15Google Scholar.

10. Hershovitz comes to the same conclusion, by pointing out that the law is often indifferent to why we act as it commands, so long as we act as commanded. See Hershovitz, supra note 6, at 67–68. For a similar argument, see also Perry, supra note 9, at 14. Cf. Raz, Joseph, Authority and Justification, 14 Phil. & Pub. Aff. 3 (1985)Google Scholar, at 7.

11. Cf. Raz, Joseph, The Obligation to Obey: Revision and Tradition, 1 Notre Dame J. L. Ethics & Pub. Pol'y 139, 140–141 (1985)Google Scholar.

12. This is in line with Stephen Perry's view that content-independence should be understood as a “constraint on the type of argument that can be offered to demonstrate that a directive is obligatory … [, and not as] a constraint on the practical reasoning of persons who are supposedly bound by the directive.” Perry, supra note 9, at 15.

13. Cf. Klosko, George, Are Political Obligations Content Independent?, 39 Pol. Theory 498 (2011)CrossRefGoogle Scholar. See also the critical discussion in Walton, Kevin, The Content-Independence of Political Obligations: A Response to Klosko, 42 Pol. Theory 218 (2014)CrossRefGoogle Scholar.

14. But cf. Adams, supra note 5, at 151.

15. Raz, Joseph, The Problem of Authority: Revisiting the Service Conception, 90 Minn. L. Rev. 1003, 1013 (2006)Google Scholar.

16. Robert Paul Wolff, In Defense of Anarchism (1970).

17. Frankfurt, Harry G., The Anarchism of Robert Paul Wolff, 1 Pol. Theory 405 (1973)CrossRefGoogle Scholar.

18. This line of reasoning shares some similarities with the discussion in Raz, Joseph, Legitimate Authority, in The Authority of Law (1979), at 3–27Google Scholar. Raz is committed to the view that a valid authoritative directive gives one second-order, exclusionary (or protected) reasons. See id. at 17. A second-order, exclusionary reason to ϕ is a reason to ϕ that excludes some other reasons as grounds for (not) ϕ-ing (i.e., it is “protected” by this exclusion). This being so, on Raz's view, a valid authoritative directive always affects the proper content of an agent's practical deliberations: excluded reasons should not serve as grounds for action. However, if other, nonexcluded reasons bear on an agent's decision, these may figure in the agent's practical deliberations. This is why Raz concludes that, in the face of valid authoritative directives, it remains true that one ought to “act on one's judgement on what ought to be done, all things considered,” but it is not true that, in those cases, one ought to act on one's judgment “on the balance of first-order reasons.” See id. at 27 (emphasis original). Some such first-order reasons will be excluded by authoritative directives. I thank Joseph Raz for discussion of these matters. Unlike Raz, I am not committed to the view that authoritative directives generate second-order, exclusionary reasons. Therefore, on the framework I have proposed, an authoritative directive can be treated as a first-order reason (or a pro tanto obligation) to be added to the relevant balance of considerations. Note, also, that conceiving of valid authoritative directives as giving rise to exclusionary reasons is arguably reminiscent of interpretation (1) of the obligation to obey the law, discussed and rejected in the previous section. One may thus wonder whether, on the “exclusionary reasons” framework, an authoritative command to ϕ gives us an obligation to [ϕ because it was commanded, and not because of excluded reasons (not) to ϕ]. For critical discussion, see Hershovitz, supra note 6, at 69.

19. Raz, supra note 15, at 1014. Raz adds the caveat that authoritative directives are valid only regarding those issues for which “it is better to conform to reason than to decide for oneself” (Id. at 1014). Since this caveat is not relevant for my analysis, I have omitted it from the main text. For an earlier statement of Raz's view, see Raz, supra note 10, at 19.

20. See Hershovitz, Scott, The Role of Authority, 11 Phil. Imprint 1, 4–5 (2011)Google Scholar.

21. Cf. supra note 18.

22. Raz, supra note 3, at 74.

23. I am grateful to an anonymous reviewer for raising this objection.

24. Raz shows awareness of this difficulty, but says relatively little about it. His solution—not further elaborated on—is that “a person or body has authority regarding any domain if that person or body meets the conditions [set out by the service conception] regarding that domain and there is no proper part of the domain regarding which the person or body can be known to fail the conditions.” Raz, supra note 15, at 1027 (emphasis added). This quotation is not fully transparent, but can be interpreted as suggesting that, for any domain in which the person or body exhibits P (on the whole), the person or body has authority except for those “proper parts” of the domain for which it is known to fail to exhibit P. Presumably, those “proper parts” will be identified by looking at specific directives and asking whether they are known to exhibit P : if the answer is “no,” the directives will be removed from the domain of authority. Two points in relation to this response are worth making. First, Raz's short passage suggests that valid authority rests not merely on directives exhibiting P, but on it being known that they exhibit P. This differs from Raz's official formulation of the service conception and creates ambiguities (which, however, I do not have the space to explore here). Second, this way of individuating domains of authority is susceptible to my original objection: a change in the content of a directive may make a difference to whether that directive is known to satisfy P, hence to whether it is authoritative. Thanks to Massimo Renzo and Daniel Viehoff for discussion.

25. Raz, supra note 18, at 25. But cf. Raz, supra note 3, at 62.

26. Note that this way of proceeding is in fact at odds with a suggestion Raz makes elsewhere, discussed in note 24, supra.

27. John Rawls, A Theory of Justice (1999), at 96. For other versions of the principle, see Hart, H. L. A., Are There Any Natural Rights?, 64 Phil. Rev. 175, 185 (1955)CrossRefGoogle Scholar; Klosko, George, Presumptive Benefit, Fairness, and Political Obligation, 16 Phil. & Pub. Aff. 241 (1987)Google Scholar; Arneson, Richard J., The Principle of Fairness and Free-Rider Problems, 92 Ethics 616 (1982)CrossRefGoogle Scholar. For a critique, see Robert Nozick, Anarchy, State, and Utopia (1974), at 90–95. In the main text, I focus on the mere receipt of benefits, and set aside the issue of whether benefits have been willfully accepted. For discussion, see A. John Simmons, Moral Principles and Political Obligations (1979), ch. 5.

28. Thanks to Massimo Renzo for pushing me on this.

29. Smith, M. B. E., Is There a Prima Facie Obligation to Obey the Law?, 82 Yale L.J. 950, 958 (1973)CrossRefGoogle Scholar.

30. Klosko, George, John Simmons on Political Obligation, 7 APA Newsletter Phil. & Law 1, 3 (2007)Google Scholar.

31. David Lefkowitz and Massimo Renzo have independently pointed out to me that fair-play theory may be interpreted as involving a duty to do one's share in sustaining the benefit of the rule of law, which implies a duty to do whatever the law demands, without using one's discretion in determining what counts as one's share. This interpretation appears to deliver content-independence but, I suggest, it is ad hoc and substantively unconvincing. Consider, for instance, agent A being faced with a red light on an empty road. The present interpretation of fair play would require A to obey the law and not cross: doing otherwise would be a violation of the rule of law. But it is mysterious why crossing would count as unfair in a case like this. By crossing, A would not take advantage of others’ compliance with the law: the road is empty. In fact, if everyone, in A's situation, were to cross the road, the benefits of traffic regulation would still be delivered. So we are not dealing with an instance of free-riding. Yet, on the “rule of law” interpretation of fair play, crossing the road would count as unfair. This suggests that the “rule of law” interpretation is ad hoc and substantively dubious: it stipulates that one's fair share in reciprocating benefits always amounts to what the law demands.

32. John Rawls, in A Theory of Justice, treats it as a positive duty, while Thomas Pogge, for example, sees it as a negative duty. Thomas Pogge, World Poverty and Human Rights: Cosmopolitan Responsibilities and Reforms (2008).

33. Wellman, Christopher Heath, Toward a Liberal Theory of Political Obligation, 111 Ethics 735 (2001)CrossRefGoogle Scholar.

34. Cf. Christopher Heath Wellman & A. John Simmons, Is There a Duty to Obey the Law? (2005), http://ebooks.cambridge.org/ref/id/CBO9780511809286, at 182.

35. Renzo, Massimo, State Legitimacy and Self-Defence, 30 Law & Phil. 575 (2011)CrossRefGoogle Scholar. Though Renzo's argument focuses on the state's justification of coercion, he is explicit that he also intends to provide a justification for the state's authority (at 577).

36. Rawls, supra note 27, at 99; Waldron, Jeremy, Special Ties and Natural Duties, 22 Phil. & Pub. Aff. 3 (1993)Google Scholar; Anna Stilz, Liberal Loyalty: Freedom, Obligation, and the State (2009).

37. Naticchia, Chris, The Anarchist Within: Natural Duty of Justice Accounts of Political Obligation, 7 APA Newsletter Phil. & Law 14, 16 (2007)Google Scholar.

38. Id. at 17 (emphasis original).

39. Cf. the reasons John Simmons offers to show that natural-duty theories fail to meet the “particularity requirement” of political obligation. That is, they are unable to vindicate an obligation to obey the law of one's own state. This is because whether obeying the law of one's own state best promotes justice in general is contingent on what the law of one's state is. Simmons, supra note 27, ch. 6.

40. Simmons himself—who is a critic of natural-duty theory—raises this as a possible response to Naticchia's objections. See Simmons, A. John, The Particularity Problem, 7 APA Newsletter Phil. & Law 19, 24 (2007)Google Scholar.

41. For a Kant-inspired, natural-duty argument along these lines, see, e.g., Stilz, supra note 36.

42. Here I only discuss associative views concerned with moral obligations. I set aside views according to which membership—including in morally impermissible associations (e.g., the Mafia)—generates sui generis, not strictly moral, obligations. See, e.g., Margaret Gilbert, A Theory of Political Obligation: Membership, Commitment, and the Bonds of Society (2006).

43. See, e.g., Ronald Dworkin, Law's Empire (1986); David Miller, On Nationality (1995); Stilz, Anna, Nations, States, and Territory, 121 Ethics 572 (2011)CrossRefGoogle Scholar, sec. 5; John Horton, Political Obligation (2d ed. 2010); for discussion, see van der Vossen, Bas, Associative Political Obligations, 6 Phil. Compass 477 (2011)Google Scholar.

44. I am abstracting away from the details of each associative view, and just focusing on their main structural features. This rather coarse-grained description of associativism suffices for my purposes.

45. I am grateful to an anonymous reviewer for drawing my attention to this.

46. John Locke, Second Treatise of Government (C. B. Macpherson ed., 1980/1690); Pitkin, Hanna, Obligation and Consent—I, 59 Am. Pol. Sci. Rev. 990 (1965)CrossRefGoogle Scholar; Simmons, A. John, Tacit Consent and Political Obligation, 5 Phil. & Pub. Aff. 274 (1976)Google Scholar.

47. Thomas Christiano, The Constitution of Equality: Democratic Authority and Its Limits (2008); Gerald F. Gaus, Justificatory Liberalism: An Essay on Epistemology and Political Theory (1996); Robert A. Dahl, Democracy and Its Critics (1989); Viehoff, Daniel, Democratic Equality and Political Authority, 42 Phil. & Pub. Aff. 337Google Scholar; cf. David Estlund, Democratic Authority: A Philosophical Framework (2008).

48. Democracy is also defended as authoritative by virtue of the quality of its outcomes. Since, however, this defense of the value of democracy is clearly content-dependent, I set it aside for present purposes.

49. Simmons, A. John, Democratic Authority and the Boundary Problem, 26 Ratio Juris 326, 340–343 (2013)CrossRefGoogle Scholar. Thanks to Anna Stilz for discussion.