As Jules Coleman would say: This is a very, very rough (but potentially insightful) draft. Please do not quote or cite without permission.

Accountability and Political Authority SCOTT HERSHOVITZ* No account of authority has been more influential among legal philosophers than Joseph Raz’s service conception. With the service conception, Raz aims to answer this question: “How can it ever be that one has a duty to subject one’s will and judgment to those of another?”1 The inquiry is motivated in part by concern that authority is incompatible with reason and autonomy. In an early essay, Raz summarized several “paradoxes of authority” that once had currency in the philosophical literature: To be subjected to authority, it is argued, is incompatible with reason, for reason requires that one should always act on the balance of reasons of which one is aware. It is of the nature of authority that it requires submission even when one thinks that what is required is against reason. Therefore, submission to authority is irrational. Similarly, the principle of autonomy entails action on one’s own judgment on all moral questions. Since authority sometimes requires action against one’s own judgment, it requires abandoning one’s moral autonomy.2

Philosophers no longer talk much about the paradoxes of authority because Raz refuted them.

* 1

Assistant Professor, University of Michigan Law School. J. Raz, The Problem of Authority: Revisiting the Service Conception, 90 MINN. L. REV. 1003, 1012 (2006). 2 J. RAZ, THE AUTHORITY OF LAW, ESSAYS ON LAW AND MORALITY 3 (1979).

Accountability and Political Authority The key to Raz’s refutation of the paradoxes lies in his theory of legitimacy, which is captured in the normal justification thesis: The normal and primary way to establish that a person should be acknowledged to have authority over another person involves showing that the alleged subject is likely better to [conform] with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding, and tries to follow them, than if he tries to follow the reasons which apply to him directly.3

That is a mouthful, but the idea is simple: If a person will do better (with respect to reason, not interests) by following an authority than by working out what to do on his own, the authority is legitimate for him, and he is obligated to do what the authority requires. An attractive feature of Raz’s analysis is its generality. Raz presents the normal justification thesis as a test for the legitimacy of political authority and parental authority. He would apply it to matters medicinal (a doctor’s authority over a patient) and musical (a conductor’s authority over the members of his orchestra). And he argues that it is equally suited to explaining the authority one person has over another, as it is to accounting for the authority the state has over the masses. Not long ago, many might have balked at the idea that the authority of a parent, a conductor, and a state rest on similar foundations. But though we would tell different stories about why parents, conductors, and states satisfy the normal justification thesis (if 3

J. RAZ, ETHICS IN THE PUBLIC DOMAIN: ESSAYS IN THE MORALITY OF LAW POLITICS 214 (1994). In Raz’s original formulation of the normal justification thesis, “comply” appeared where I have substituted “conform.” For technical reasons, Raz now uses “conform.” See Raz, supra note 1, at 1014. For the reasons underlying the switch, see S. Hershovitz, Legitimacy, Democracy, and Razian Authority, 9 LEGAL THEORY 201, 206 (2003).

AND

2

Accountability and Political Authority they do), it is now commonplace among legal philosophers that their authority is grounded in the service they provide their charges.4 Recently, however, several scholars have raised doubts that Raz’s account applies as widely as he claims. In an earlier essay, I argued that the normal justification thesis is not an adequate test for the legitimacy of political authority.5 Our discourse about political legitimacy places great emphasis on procedural matters, on the way that government makes decisions. Yet the normal justification thesis does not appear, at least on first glance, to take into account of whether, for example, government operates transparently or affords adequate opportunity for those affected by its actions to influence outcomes. A political authority which satisfies the normal justification thesis may be illegitimate, I claimed, if it makes decisions in ways that are disrespectful of the dignity of its citizens as potential agents in their own self-government. And I asserted that the reverse is true as well: Democratic authorities that fail Raz’s test may nevertheless be legitimate because sometimes making decisions together through fair procedures is more important than reaching the right result.6 To different extents, and in different ways, Samantha Besson,7 Scott Shapiro,8 and Jeremy Waldron9 have also called into question 4

J. WALDRON, LAW AND DISAGREEMENT 84 (1999) (“The conception of authority standardly accepted among legal philosophers at the moment is that of Joseph Raz.”). 5 Hershovitz, supra note 3. 6 Id., at 216-219. 7 S. Besson, Democracy, Law and Authority, 2 JOURNAL OF MORAL PHILOSOPHY 89 (2005). 8 S. Shapiro, Authority, in THE OXFORD HANDBOOK OF JURISPRUDENCE AND PHILOSOPHY OF LAW 431-39 (J. Coleman and S. Shapiro, eds. 2002). 9 J. Waldron, Authority for Officials, in RIGHTS, CULTURE, AND THE LAW: THEMES FROM THE LEGAL AND POLITICAL PHILOSOPHY OF JOSEPH RAZ 6366 (L.H. Meyer, et al. eds. 2003).

3

Accountability and Political Authority the adequacy of the normal justification thesis as a test for the legitimacy of political authority. Those who have argued that the normal justification thesis fails to capture the conditions under which political authorities are legitimate have tended to suggest that there is something distinctive about political authorities, specifically democratic ones, which renders Raz’s test invalid as applied to them. My argument had this flavor, but I have come to think it a mistake. That is, I have come to regard the failure of the normal justification thesis to account for the legitimacy of democratic political authorities is a consequence of a defect more general. An authority is justified, on Raz’s picture, solely by reference to its ability to help subjects conform better to the reasons they have. But there is good reason to think that a successful explanation of the legitimacy of any authority — parental, political, or otherwise — must appeal to the nature of the relationship between the authority and its subjects, not simply to the quality of the directives the authority issues. The centrality of the relationship between an authority and its subjects has been obscured by the restricted view philosophers commonly have as to what authority consists in. Raz follows a long line of thinkers who hold that authority is comprised of a “right to rule, where that is understood as correlated with an obligation to obey on the part of those subject to the authority.”10 And Raz, like many others, approaches authority from the point of view of those subject to authoritative demands. That is why he asks the question, “[h]ow can it ever be that one has a duty to subject one’s will and judgment to those of another?”11 And it is why he “assume[s] that all other problems regarding authority are 10 11

J. RAZ, THE MORALITY OF FREEDOM 23 (1986). Raz, supra note 1, at 1012. See also RAZ, supra note 3, at 356 (“What can justify holding some people to be duty-bound to obey others? Under what conditions can some have a right to rule others?”).

4

Accountability and Political Authority subsumed under [that question].”12 The problem is that authorities do more than issue orders intended to bind people to a course of action. They also hold people accountable for their behavior. Sometimes an authority holds people accountable for violations of duties it has imposed. But in other cases subjects are held accountable for running afoul of duties they have independently of any action the authority has taken. This is most obviously true of children, whose parents may hold them accountable for wrongdoing whether or not it runs afoul of a parentally-imposed duty. The normative power to hold individuals accountable is, thus, separable from the normative power to issue binding orders. But if possessing practical authority entails having both powers, as I shall argue it does, a complete account of the legitimacy of any authority must include an explanation of how the authority acquires the moral standing to hold others accountable, as well as an explanation of how it imposes genuine obligations on others. The standing to hold others accountable for wrongdoing, I shall argue, is not typically acquired on the basis of one’s competence to do so. Rather, such standing is normally grounded in a relationship between the authority and its subjects that licenses the particular ways in which the authority seeks to call its subject to account. The relationship which confers a power to hold others accountable will, of course, take different forms depending on whether the authority in question is a parent, an orchestral conductor, or a state. This is in part because different sorts of authorities hold their subjects accountable in different ways, but also because authorities operate in diverse domains. If I am right about all this, we can deepen our understanding of why the normal justification thesis is inadequate as a test for the legitimacy of political authorities. The answer is not that 12

Raz, supra note 1, at 1003.

5

Accountability and Political Authority there is something special about democratic authorities that requires us to judge them differently than we would other practical authorities. We do not simply give democracies extra credit because we like the way they make decisions. Rather, our intuition that democratic procedures have the potential to confer legitimacy on a political authority stems from the fact that the procedures through which a state makes decisions are partly constitutive of the relationship between the state and its citizens. Democratic procedures, on this view, play a role in an explanation of the legitimacy of a state because they may confer the moral standing necessary to allow the state to hold its citizens accountable. Part I of this paper briefly canvasses the reasons for thinking that the normal justification thesis is an inadequate test for the legitimacy of political authority. Part II argues that an authority must possess the standing to hold its subjects accountable, and that such standing is typically grounded in a relationship between the authority and its subjects. Part III connects the positive view developed in Part II to the criticisms advanced in Part I, in an effort to show that the procedures a state uses to make decisions are relevant to assessing its legitimacy because they have the potential to confer the moral standing necessary to allow the state to hold its citizens accountable. I. Democracy and the Normal Justification Thesis “It is . . . no accident,” Raz recently explained, that his “account of authority makes no special reference to democratic authority.”13 Indeed, Raz cautions us that it is “of vital importance that we . . . not fall prey to the current, and much abused, democratic rhetoric, and maintain a clear-sighted and

13

Raz, supra note 1, at 1031.

6

Accountability and Political Authority critical perspective on the nature of democratic institutions.”14 Democracies, in Raz’s view, have a leg up on other political arrangements only to the extent they produce better results. The legitimacy of a democratic state, on his analysis, depends solely on whether its directives help its subjects conform better to reason’s requirements. That is, of course, precisely the same test Raz would apply to a political authority of any sort. Raz’s skeptical attitude toward democracy is further evident in encouragement that we “acknowledge the possibility that what pass for democratic regimes could completely lack legitimacy.”15 Raz is right to caution us against getting swept up in fashionable rhetoric and warn that many political systems which popularly pass for democracy lack legitimacy. Nevertheless, Raz’s claim that the normal justification thesis captures the conditions under which a state is legitimate sits uncomfortably with the fact that, as Scott Shapiro has observed, the legitimacy of a political authority “is generally not judged exclusively, or maybe even primarily, by its output, but rather by its input.”16 In assessing whether a political authority is legitimate, we do not simply consider the obligations it imposes on its citizens. Rather, we critique structures of government based on considerations such the degree to which people have access to the political process and efficacy within it. We think, to pick a small but important example, that decision-making behind closed-doors is inappropriate even if the decisions reached are good ones. We do not prize transparency only because we think that open decision-making leads to better outcomes. On some occasions, in fact, we recognize that a better decision might be reached if secrecy prevented interest groups from applying 14 15 16

Ibid. Ibid. Shapiro, supra note 8, at 434.

7

Accountability and Political Authority pressure. Nevertheless, many think it disrespectful of the dignity of citizens as agents in their own self-government to make important decisions affecting their lives without allowing them to witness and contribute to the debate shaping the outcome, even if only indirectly through public comment and protest. In this respect, and in many others, our discourse about the legitimacy of political authority reflects the fact that we think that a democratic political authority may be legitimate even if another political arrangement would produce better decisions. This is not the place to explore whether that is true, nor is it the place to consider which political systems that popularly pass for democracy actually deserve the title. The point here is simply that if democracy is uniquely well-suited to promote values such as autonomy, equality, and justice, as many argue it is, we have strong reason to prefer it even if decisions taken by other The normal means might be substantively superior.17 justification thesis appears to overlook the role that democratic procedures may play in legitimizing a legal regime, and this has led several scholars to conclude that it is not an adequate test for the legitimacy of political authority. Responding to this criticism, Raz has recently suggested that the “suppleness of the service conception” has been “underestimated.”18 Democratic governments, he claims, may “enjoy some qualified or limited authority . . . because of their ability to give expression to people’s standing as free autonomous agents, or [because of] whatever other values they serve.”19 What exactly does Raz mean, and how are we to square this new claim with the traditional understanding of the normal justification thesis? The answer lies in two examples 17 18 19

See Hershovitz, supra note 3, at 213-216. Raz, supra note 1, at 1030. Raz, supra note 1, at 1031 n. 20.

8

Accountability and Political Authority Raz offers. “Some people believe,” he says, “that one has a duty to obey anyone who is elected by a majority. . . . If that is so it simply shows that the conditions of the service conception are met regarding anyone who is so elected.”20 Likewise, if members of a certain group have a duty arising from historical circumstances to obey a particular institution, Raz claims, the normal justification thesis is satisfied because “[b]y obeying that . . . institution, one is discharging that duty.”21 We can generalize the point behind Raz’s examples as follows: If one has a duty to obey an authority, the authority satisfies the normal justification thesis because obeying it helps one to conform better to the reasons one has (which, ex hypothesi, mandate obedience) than one otherwise would. This argument blunts the criticism that the normal justification thesis cannot account for the possibility that a democracy might be legitimate even if its decisions are substantively inferior to decisions taken by other means. (It does not answer the criticism that political authorities which pass the test of the normal justification thesis may nevertheless be illegitimate in light of how they make decisions.22) If it could be shown that people are obligated to obey the dictates of a democratic political authority irrespective of the quality of its decisionmaking (or, more likely, so long as its directives do not fall below some minimal threshold of acceptability), the normal justification thesis would in fact be satisfied on the view that Raz now advocates. The problem with Raz’s new approach is that it renders the normal justification thesis an empty formalism. To decide whether the normal justification thesis is satisfied, it is no longer enough to determine whether compliance with 20 21 22

Id., at 1031. Id., at 1030. Arguably, Raz can address this criticism through his independence condition . . .

9

Accountability and Political Authority authoritative directives will allow a subject to conform better to the reasons she has on the matter the directive regards. We can no longer decide whether, for example, financial regulations are authoritative by asking if the banks subject to them would conform better to what reason requires of them with respect to their financial activities by following the regulations than they would by following their own lights. We must also consider, in every case, whether a further criterion of legitimacy (not specified by the normal justification thesis) gives rise to a duty for the subject to comply, irrespective of the instrumental value of authority’s directives. The upshot is that, on Raz’s new approach, the normal justification thesis ceases to be a competitor with other theories of legitimacy because it subsumes any theory which on its face conflicts with it but turns out to be true.23 To my thinking, Raz has made a mistake in suggesting that the normal justification thesis is satisfied whenever an individual has an independent duty to follow an authority’s orders. To be sure, his argument has the potential to save the normal justification thesis from the criticism that it fails to account for the possibility that a democratic political authority may be legitimate even with respect to citizens who would do better making decisions for themselves. However, the new approach drains the normal justification thesis of much of its theoretical appeal, as it no longer provides a guide to determining when, in fact, people are under a duty to obey an authority. Regardless of whether one accepts Raz’s revisionary take on the normal justification thesis, however, two further features of his account call into question whether he has successfully captured the conditions of political legitimacy. The first is, 23

For a more detailed version of this objection to Raz’s new view, see Hershovitz, supra note 3, at 219-220.

10

Accountability and Political Authority once again, democracy specific. As noted above, Raz takes the central question about authority to be some form of the following: “What can justify holding some people duty-bound to obey others?”24 Many, however, understand democracy to involve self-government, casting doubt on the distinction between rulers and subjects presupposed by Raz’s analysis. As I have elsewhere argued, democracies may have different criteria of legitimacy than other sorts of governments because of the fact that, unique among political arrangements, democracy does not involve the rule of one group of people over another.25 Even if the normal justification thesis answers the question when some people are duty bound to obey others, it may not answer the question when people, acting in concert with others, successfully bind themselves. The final oddity in Raz’s approach is that, under the normal justification thesis, the legitimacy of political authority is both personal and subject-matter specific. I may be bound by legal regulations detailing the appropriate use of pharmaceuticals, because, lacking medical knowledge, I will no doubt do better following Food and Drug Administration rules than by trying to figure out what drugs to take on my own. A pharmacist or a medical doctor, however, may do at least as well or better by following her own lights, in which case the same regulations would not be legitimate for her. Some types of law might be legitimate for almost everyone (e.g., traffic laws which solve coordination problems), while other kinds of law will be legitimate for almost no one. Strangely, many criminal laws fall into the later category, as most set the floor of acceptable behavior, rather than providing a guide to acting well. Trying to follow legal proscriptions against murder is probably not a way of improving upon my ability to avoid unjustified killings. 24 25

RAZ, supra note 3, at 356. See Hershovitz, supra note 3, at 209-212.

11

Accountability and Political Authority To be sure, laws against murder are binding for some people on Raz’s analysis, but it is odd that for me and many others, they are not. Raz argues at length that “flexibility in determining the scope of authority” is a virtue of his account.26 Whereas most discussions of authority ask whether a government has all the authority it claims or none, Raz says his account is more discriminating in that it allows the possibility that the government has some of the authority it claims but not all. I might agree with Raz that this is a virtue of his account if his claim was merely that a government might be “legitimate with respect to some areas of regulation but not others.”27 But the notion that a democratic government might be legitimate for me and not my neighbor flies in the face of many arguments that are commonly made in support of democracy, arguments that place emphasis on the need of people living in the same community to make decisions together through fair procedures.28 II. Accountability and Authority Let us turn our attention from political to parental authority. As I noted in the Introduction, parents may hold their children accountable for wrongdoing, whether or not it violates a parentally-imposed duty. A child caught punching his younger brother cannot avoid punishment by arguing that, though his action may have been wrong, his parents had not issued an order forbidding it. Defiance of an order is not a necessary predicate to parental punishment. 26 27 28

RAZ, supra note 10, at 73. Shapiro, supra note 8, at 403. For an argument of this sort, see J. WALDRON, LAW AND DISAGREEMENT 99-101 (1999).

12

Accountability and Political Authority The scope of a parent’s power to hold her own child accountable for wrongdoing is almost unparalleled. Teachers may be next in line after parents, but their power to hold children accountable covers a narrower range of behavior (e.g., a teacher may not punish a child for failing to respect her grandmother), and a teacher has more limited tools at her disposal (e.g., a teacher may not order a child to perform chores at the teacher’s house). As a general rule, the more attenuated a relationship between an adult and a child becomes, the more limited the adult’s power to hold the child accountable for wrongdoing. We cannot explain the scope of an adult’s power to hold a child accountable by appealing to the adult’s competence in such matters. I may be a wonderful parent, skilled in the art of disciplining children, but that does not license me to reproach every misbehaving child I happen upon. If I reprimand someone else’s child, the parent may object: “Who are you, and why are you speaking to my child that way?” “I’m his teacher” would be a good answer in some situations, were it true, but “I have better parenting skills than you” is not sufficient even if accurate. Competence does not confer standing to discipline a child.29 How are we to explain the power parents have to hold their children accountable? The power must be grounded in the parent-child relationship; that much is implicit in the mother’s question, “Why are you speaking to my child that way?” I do not know precisely what features of the relationship give rise to the power, but I suspect that an important part of the answer is the special responsibility that a parent has for the moral development and well-being of her children. Indeed, it is difficult to imagine that a parent could discharge this 29

Though, as Britney Spears has learned, one can lose the standing to discipline a child on the basis of incompetence.

13

Accountability and Political Authority responsibility without the power to hold a child accountable for wrongdoing. Notably, the other adults we can most readily imagine having standing to discipline a child (e.g., teachers and babysitters) also have some responsibility for the child’s moral development and well-being, though more limited. Thus far, I have been careful to refer to a parent’s “power” to hold her child accountable, rather than her “authority” to do so, because I do not want to beg the question whether this power is partly constitutive of parental authority. I shall argue below that it is, but first I want to turn to a situation that does not involve authority at all. Spouses have the power to hold each other accountable for wrongdoing. My wife, for example, has standing to criticize me for failing to do my share of the work around the house, and she may demand an explanation if I miss an important family event. In each case, I violate a duty that I plausibly owe to her, and when a duty is owed to someone else, the person to whom the duty is owed commonly has standing to hold the other person accountable for violation of it. The scope of my wife’s power to hold me accountable is, however, more limited than the analogous power in the context of a parent-child relationship. My wife can demand that I explain myself, and perhaps even that I make repair, but she cannot punish me. Nor can she hold me accountable for all of my wrongdoing; some things I do are her business, others are not. For example, if my wife hears me being snide to my friend, she does not have standing to demand that I apologize. My duty to be kind to my friend is a duty to my friend, not to my wife. Importantly, there are tools my wife can use to expand her power to hold me accountable: She can make requests and elicit promises. To build the last example, my wife might say, “Your friend seems emotionally fragile right now. Please, if only for me, don’t tease him.” If I do not do as my wife requests, then she may hold me accountable, perhaps demanding an explanation or apology. I can avoid responsibility for my 14

Accountability and Political Authority failure to act on the request if I can show that the request did not tip the balance of reason.30 But, assuming that my wife had standing to make the request (there are some things even a spouse cannot ask of her partner), she has created a relationship of accountability where none existed before. She could do the same by eliciting from me a promise to her not to mistreat my friend. My wife’s arsenal for expanding the scope of her standing to hold me accountable is missing one tool she might wish she had: Except in rare circumstances, spouses cannot issue orders to one another. This is a consequence of the fact that, in our society, spouses are supposed to relate to one another as equals. In hierarchical relationships, however, an order accomplishes much the same as a request or a promise: If validly issued, an order creates a relationship of accountability. Suppose that a employer orders her employee to prepare a report. If the employee fails to do so, the boss has standing to demand an explanation, and perhaps even to punish the employee. Of course it is possible the employee had sufficient reason to prepare the report even before she was instructed to do so. If that is the case, the employer would have had standing to complain even if she had never ordered the report’s production. Employers, like parents, have standing to hold employees accountable for wrongdoing, even when it does not violate an express order.31 But failure to follow a validly issued order is a separate wrong, and the employee’s disobedience may (depending on particulars) license a range of responses that would not be justified simply on the basis of substandard performance. Because this is so, issuing an order is often a sensible thing for a boss to do even when an employee already 30 31

Requests, unlike orders, do not create exclusionary reasons. An employer’s power, however, is narrower than a parent’s, as it is limited by relevance to the workplace.

15

Accountability and Political Authority knows what she ought to do. Much the same is true in the parent-child case, where, as we saw before, there is a background norm that allows a parent to hold a child accountable for all wrongdoing. When a parent issues an order proscribing something that is wrong (e.g., “do not read your sister’s diary”), and a child defies it, the parent may hold the child accountable for two wrongs: first, that the child misbehaved; and second, that the child disobeyed an order. Let us now consider the relationship between authority and the normative power to hold a person accountable. The spousal case shows that authority over another person is not is a precondition of the power to hold that person accountable for wrongdoing. There are, however, strong reasons to conclude the opposite, that one cannot posses authority over another person unless one has the power to hold her accountable. In a recent paper, Stephen Darwall adopts one of Raz’s examples to make the point.32 Raz posits that “John is an authority on Chinese cooking,” in which case it follows that if one “want[s] nothing but to prepare the best Chinese meal . . . then [one] should just follow John’s instructions.”33 Darwall asks us stipulate that a person, call her Sara, “has no reason to do anything other than prepare the best Chinese meal.”34 If that is so, on Raz’s analysis, John’s instructions create protected reasons for Sara.35 Sara will do better if she follows John’s instructions than if she decides how to prepare the meal on her own. Thus, she has a reason to do as John says and a reason not to act for countervailing reasons. But does that mean that John

32

S. Darwall, Authority and Second-Personal Reasons for Action 22-23, unpublished manuscript, on file with author. 33 RAZ, supra note 10, at 64. 34 Darwall, supra note 32, at 22. 35 Darwall suggests that this may not be true, but let us grant (as Darwall does) that it is. See id., at 22 n. 22.

16

Accountability and Political Authority has practical authority over Sara? Darwall argues that he does not: Of course, [Sara] would be foolish not to follow [John’s] instructions. But if [she] didn’t it is difficult to see why John would have any standing to complain or otherwise hold [Sara] to account. Raz says that those with practical authority “have the right to replace people’s own judgment on the merits of the case.” But what right could John have in such a case? . . . It follows from our stipulations that John’s instructions are not mere advice; they provide preemptive reasons. But it is hard to see how that gives John any right to [Sara’s] compliance with his directives or [Sara] any obligation to comply with them.36

What is missing from the story about John and Sara is a relationship between the two that confers on John the standing to make demands of Sara. If we add the additional supposition that John is the executive chef and Sara a line cook in a Chinese restaurant, then it is manifest that John has a right to Sara’s compliance and the standing to call her to account when she ignores his instruction. In that case, John’s instructions do not simply create protected reasons for Sara; they give her an obligation to him to do as he directs. A full explanation of why an executive chef has the power to hold a line cook accountable is beyond the scope of this paper, but the rough outlines are clear enough. An executive chef’s standing is likely grounded in the fact that members of a kitchen staff are committed to a joint endeavor for which he has ultimate responsibility. Somebody else in the kitchen (the sous chef, perhaps) may have greater expertise or be better suited to coordinate the activities of the various cooks, but the standing to make demands on others attaches to the responsibility vested in the executive chef. Whatever the particulars of the story, however, for our purposes it is enough to see that the nature of 36

Darwall, supra note 32, at 22-23.

17

Accountability and Political Authority the relationship between an executive chef and a line cook can empower someone in John’s position to make demands on someone in Sara’s position that competence alone would not license. Darwall’s argument suggests that Raz is wrong to locate authority in the power to create protected reasons. That power is indeed an element of authority, as Raz has long argued. But unless the power to create protected reasons is coupled with the power to hold others accountable for failing to act as those reasons require, one does not have authority in the sense that Raz has long sought to explain: “a right to rule, where that is understood as correlated with an obligation to obey on the part of those subject to the authority.”37 The normal justification thesis appears to tell us whether someone would make a good authority, not whether someone possesses it. To know whether one person has authority over another, it seems we must take stock of the relationship between them. Once again, Raz has recently addressed the criticism that the normal justification thesis “describes the conditions under which an authority is a good-enough authority . . . but . . . does not explain what it is to be an authority.”38 He confines his response to considering the case of political authorities, and he asserts that the service conception can indeed distinguish between possessing the qualification to be an authority and actually being one. The role of political authorities, Raz says, “requires them to be successful in coordinating the conduct of large numbers of people.”39 Because this is so, only a de facto authority can possess legitimate political authority.40 Moreover, Raz claims, “governments of the kind we are familiar with can 37 38 39 40

RAZ, supra note 10, at 23. Raz, supra note 1, at 1032. Id., at 1036. A de facto authority is one which succeeds in imposing its will on others. De facto authorities may or may not be legitimate.

18

Accountability and Political Authority only succeed in meeting the conditions of legitimacy (according to the service conception) if they have the authority to use and are successful in the use of force against those who flout certain of their directives.”41 Raz does not tell us what it means to “have the authority to use . . . force,” so it is unclear whether Raz is arguing that a political authority only satisfies the the normal justification thesis if it has standing to hold its subjects accountable in the ways we have been discussing. (There is a difference between having standing to make a demand on others and being justified in coercing them). As we shall see in the next Section, there is reason to doubt that this is what Raz has in mind. But, in any event, it bears note that the argument Raz makes about political authority is not generalizable to other cases of practical authority. Political authorities must possess de facto authority to solve coordination problems, but many practical authorities do not aim to solve coordination problems. Parents of an only child, for example, do not need to solve coordination problems among their subjects, nor do doctors generally need to solve coordination problems among their patients. Parents and doctors are authorities for their children and patients whether or not their charges are in the habit of complying with their directives. Thus, Raz’s strategy for answering the criticism that he confuses the qualification to be an authority with the possession of it only gets him so far. Perhaps it follows from the normal justification thesis that a political authority must have the power to hold individuals accountable for wrongdoing to be legitimate (though Raz does not show that it does), but it does not follow that this is so for all practical authorities. Yet, if Darwall’s argument is right, as I think it is, then no authority can claim the right to rule (or a corresponding obligation to obey), unless it has such a power. 41

Raz, supra note 1, at 1036.

19

Accountability and Political Authority III. Accountability and Political Authority In his recent essay, Raz bracketed the question under what conditions a government might justifiably deploy force against those who flout legal directives. In The Morality of Freedom, however, he addressed the question head on. There Raz asserts that “remedies for breaches of the law . . . can be morally justified even when applied to those who are not subject to the authority of the government and its laws.”42 Raz illustrates the point as follows: Consider for example the law of defamation. Assuming that it is what it should be, it does no more than incorporate into law a moral right existing independent of law. The duty to compensate the defamed person is itself a moral duty. Enforcing such a duty against a person who refuses to pay damages is morally justified because it implements the moral rights of the defamed. One need not invoke the authority of the law over the defamer to justify such action. The law may not have authority over him. It makes no difference. The importance of the law in such matters is creating a centre of power which makes it possible to enforce moral duties.43

Raz’s view appears to be that the state is morally justified in enforcing a duty if the state has the power to do so and the duty is genuine. It matters not, according to Raz, whether the state has authority over those against whom it is enforcing the obligation. There is considerable tension between Raz’s take on state standing to enforce moral obligations and the lessons we learned in the last Section. Suppose that Mike is both wise and strong, such that he typically knows what people’s moral duties are and is capable of holding people accountable for failure to discharge them. Do those two facts alone confer on Mike the 42 43

RAZ, supra note 10, at 103. Ibid.

20

Accountability and Political Authority normative power to hold others accountable for violation of their moral duties? If the others we have in mind are children, the answer is “no.” As we saw before, competence does not confer standing to discipline a child. The people who have standing to enforce a child’s moral obligations (parents, teachers, etc.) acquire that standing in virtue of a relationship to the child, which gives the adult in question special responsibility for the child’s moral development. Nor does it appear that Mike’s wisdom and strength would confer on him the power to hold other adults accountable either. As we saw in the spousal case and the Chinese cooking case, such standing is justified by reference to a relationship which licenses one person to make demands on another. What we know about Mike suggests that he may do a good job of enforcing people’s moral obligations, not that he has the standing to do so. The same is true of the state in Raz’s defamation example. The fact that the law has created a “centre of power” through which the moral rights of the defamed can be enforced implies only that the state may be competent to hold the tortfeasor to account, not that the state has standing to do so. The state’s standing to punish someone who harms another’s reputation (or to demand that he pay compensation to his victim) is licensed, if at all, by a relationship between the state and the defamer that empowers the state to call him to account for his wrongdoing. In fact, we must appeal to such a relationship to explain why it is the state and not Mike, or someone else similarly well-suited to the task, that has the normative power to provide a remedy for the breach of the moral obligation not to defame. The legitimacy of the state’s authority is as much on the line when it seeks to hold people to account for their misbehavior as when it issues an order intended to bind people to a course of action. In each case, one subject to the exercise of the state’s power can sensibly ask, “What business does the state have making demands on me?” An adequate answer to that question calls for more than a demonstration that one would do well to 21

Accountability and Political Authority follow the state’s orders or to comply with its enforcement efforts. As Darwall explains, the fact that it would be foolish not to recognize someone as an authority does not by itself establish that the person possesses authority. Now that we can see clearly that a full justification of the state’s authority requires an explanation of why the state has the standing to hold its subject to account for wrongdoing, we can deepen our understanding of why the normal justification thesis fails as a test for the legitimacy of political authority. The chief criticism of the normal justification was that it assesses the legitimacy of a political authority exclusively in terms of its output rather than its input. That led many to argue that Raz’s account overlooks the possibility that well-constructed democratic procedures may legitimize a legal regime, even with respect to subjects that may not do better complying with the law than they would following their own lights. This argument can now be reframed: The means by which a political authority makes decisions are, in part, constitutive of the relationship between the authority and the people it seeks to govern. A state in which citizens contribute to the development of the law may have a stronger claim to the standing to call its citizens to account than a state that excludes them from the political process. So too for governments which that open decisionmaking processes or in other ways manifest respect for citizens as potential agents in their own self-government. This is not the place to argue that this is so. My goal here is not to establish that democracy in general, let alone any particular democratic procedures, confers on states the normative power to make demands on citizens. The point is simply that the arguments in favor of the legitimacy of democratic political authorities that appear out-of-place in Raz’s picture are the sort of arguments that might justify a political authority if we think legitimacy primarily a question of the relationship between a state and its citizens.

22

Accountability and Political Authority With our new perspective, we can also explain more clearly why it is that the fact that legitimacy is personal and subjectmatter specific seems such an odd feature of Raz’s analysis. If the state’s authority to make demands on its citizens depends more on the nature of relationship between them than it does on the state’s claim to expertise, and the relevant relationship is constituted by the method by which the state takes decisions, the state has no weaker claim on a doctor’s obedience to pharmaceutical regulations than it does to anyone else’s. As Waldron puts it, a decision taken through democratic means “stands for the time being in the name of the whole community . . and is entitled to whatever respect that communitarian status confers on it, without regard to — indeed bracketing away from — the substantive merits of its content.”44 A doctor might do better on her own than she would adhering to democratically enacted law, but being a doctor does not entitle one to special status in a democracy. In a democracy, we are bound together, if at all, in virtue of our shared circumstances and the need to resolve disputes together through fair procedures. Finally, with our new view in mind, we can also sketch an account of how it is that the criminal law could be legitimate, notwithstanding the fact that, for most of us, the directives that comprise it do not satisfy the normal justification thesis. The criminal law proscribes many actions that the vast majority of us know are wrong and would not undertake even absent the criminal law’s prohibition. Since we will do at least as well or better following our own lights, much of the criminal law appears, on Raz’s analysis, to be illegitimate and hence not binding. But as we saw above, it can be sensible to issue an order to subjects who already know what their responsibilities are. The reason is that orders, like requests and promises, can create accountability relationships. We don’t have murder 44

WALDRON, supra note 28, at 101.

23

Accountability and Political Authority laws, one might argue, to provide epistemic guidance as to what counts as an unjustified killing. Rather, legal proscriptions on murder are adopted primarily to confer on the state standing to hold individuals accountable for actions that are wrong quite apart from what the law says. One possible reason for this is that, unlike parents, a state may not be justified in holding subjects accountable unless it has given them advance notice of the circumstances in which it will deploy coercive force. IV. Conclusion The aim of this paper was to deepen our understanding of the reasons that the normal justification test fails to account for the legitimacy of political authority. If the argument in Section II is right, however, the normal justification thesis is not an adequate test for legitimacy of any authority. To be an authority, one must posses the normative power to hold subjects accountable. The standing to do so is typically not acquired on the basis of competence. Rather, it is justified by reference to a relationship between the authority and the subject which licenses the authority to call the subject to account. In all cases, then, we must take stock of the relationship between an authority and its subjects in assessing its legitimacy. Democracy is not a special case. It just so happens that, for political authorities, making decisions through democratic means may be a way of acquiring the normative standing to make demands on subjects.

24