Navigation
Newsletters Index (06:2)
apaOnline Home Page

APA Newsletters

Spring 2007
Volume 06, Number 2

Newsletter on Philosophy and Law

Articles

Two Problems of Political Authority

Stephen Perry
University of Pennsylvania Law School

Legitimate political authority, it is generally agreed, involves a right to rule. A political regime has a right to rule only if it has, among other things, a fairly extensive power to change its subjects’ normative situation by, for example, imposing obligations or conferring rights on them. No doubt a right to rule involves much more than this. Perhaps it involves not just a power to impose obligations but also a right to compel compliance with those obligations by means of force or the threat of force. For present purposes, however, it will be sufficient to focus on the point that legitimate political authority requires possession of extensive normative powers. Joseph Raz calls political regimes which have effective control over a population and which also claim legitimate authority for themselves, meaning they claim extensive powers to change the normative situation of their subjects, de facto authorities. Raz argues, in my view rightly, that a political regime can only have law and a legal system if it is a de facto authority in this sense. For present purposes it will be convenient to focus on the case of de facto authorities, which we can think of loosely as the governments of states with legal systems. Because such governments can, in the exercise of the extensive authority that they claim for themselves,1 affect their subjects’ lives in very significant ways and often against their will, the authority they claim to possess is moral in nature.2 Although legal systems claim to be able to change the normative situation of their subjects in almost any conceivable way—for example, by granting permissions, conferring rights, or creating powers—I will focus on the most fundamental power that legal systems claims for themselves, which is the power to impose obligations. Let me refer to the product of the exercise of a claimed power to impose obligations as a directive. If a directive was issued by an organ of a government that not only claims but also possesses legitimate authority, then those persons who fall within the scope of the directive have an obligation to obey it. Because legitimate authority is moral authority, this obligation is a moral obligation.
The notion of political authority gives rise to a number of different philosophical problems, but two are particularly salient. The first concerns whether or not and under what circumstances political authority can ever be morally legitimate. The concern, in other words, is with the question of whether or not and to what extent the claim of de facto authorities to be legitimate can ever be a justified one. Let me call this the problem of justification. Since the claim to possess legitimate political authority is moral in nature, the problem of justification is a moral problem, and attempts to address it must have recourse to substantive moral argument. For example, the traditional Thomist response to the problem of justification is the thesis that, in John Finnis’s words, "the sheer fact that virtually everybody will acquiesce in somebody’s say-so is the presumptively necessary and defeasibly sufficient condition for the normative judgment that that person has (i.e., is justified in exercising) authority in that community."3 As Finnis recognizes, this is a substantive moral claim and must be defended as such. The second problem of political authority concerns the question of whether or not an individual can ever be justified in subjecting his or her will to that of another person. Let me call this the subjection problem. This, too, is a moral problem, at least insofar as the issue is taken to be one of maintaining moral autonomy4 and making moral decisions in accordance with one’s own independent judgment.

The two problems of political authority that I have identified are clearly related, since a response to the justification problem that purports to show that political authority can under some circumstances be legitimate must show that, at least under those circumstances, it is justified to subject one’s will to that of another. But the two problems are nonetheless distinct. A demonstration that one is justified, in appropriate circumstances, in subjecting one’s will to that of another is a necessary condition of the legitimacy of political authority, but there is no reason to think that it is, in general, a sufficient condition. Beyond that, the problem of subjection arises in many contexts besides that of political authority. According to one understanding, the problem of subjection arises whenever one is faced with a demand or request by another person that one behave in a certain way. Not only are demands that one do such-and-such often made in contexts besides that of political authority, they are not always even intended to be exercises of a normative power; sometimes a demand is just a demand. According to another, still broader understanding of the subjection problem, it arises whenever one is faced with the possibility of relinquishing control over some aspect of one’s life by acting not according to one’s own judgment about what one ought to do in a given type of circumstance, but, rather, according to someone else’s view. On this broader understanding, the subjection problem arises, for example, for a person who more or less automatically relies on the advice of a friend, or on the general recommendations of a self-help book, about what to do in a particular kind of situation, rather than on his or her own judgment.

Raz’s well-known service conception of authority appears to have been offered, at least in its early formulations, as a response to both the justification problem and the subjection problem. In this article, I will argue that while the service conception does indeed offer an important insight about when one is justified in subjecting one’s will to that of another, and to that extent addresses an important dimension of the justification problem, it nonetheless does not succeed as a general account of the conditions that a de facto political authority must meet in order to possess the legitimate moral authority that it claims for itself. The difficulty, in brief, is that even if the service conception correctly identifies (many of) the situations in which one is justified in subjecting one’s will to that of another, it does so in a way that does not depend on whether or not the other person has exercised or purported to exercise a normative power to impose a binding obligation. But the possession of a moral power to change the normative situation of another is, by Raz’s own lights, the heart of legitimate authority. If the service conception does not directly address the question of whether or not the possession of such a power is justified, then it has not truly come to grips with the problem of justifying authority.

The heart of the service conception is the normal justification thesis (the NJT), which in Raz’s early formulations asserted that "the normal way to establish that one person has authority over another person," and hence the normal way to show that the latter "should acknowledge the authoritative force of [the former’s] directives," is to show that the second person will in general do better in complying with the reasons that apply to him "if he accepts [the first person’s] directives as authoritatively binding and tries to follow them, rather than trying to follow the reasons that apply to him directly."5 Notice that the NJT is explicitly offered here as a full-fledged response to the justification problem: To "establish" that one person has authority over another person can only mean to offer a substantive justification for the conclusion—explicitly acknowledged by Raz to be moral in nature—that the one has legitimate authority over the other.6 Of course, Raz also offers the NJT as a response to the subjection problem.7 Roughly speaking, the general idea is that, if one will better comply with right reason in a specified type of circumstance by allowing oneself to be guided by the judgment of another rather than by trying to act according acording to one’s own judgment about what ought to be done, then one is justified in subjecting one’s will to that of the other. For the moment, however, my concern is with the NJT understood as a full-fledged response to the justification problem.

The difficulty that the NJT faces in justifying legitimate authority can be brought out by considering one of the two main types of case to which Raz says that the NJT has application, namely, cases in which the person issuing the directive has greater expertise than the person to whom the directive is issued. (The other type of case, which involves situations in which the person issuing directives is in a position to achieve valuable coordination among the activities of several people, will be discussed later.) To avoid unnecessary complications, I will focus on a case in which the "background reasons" that apply to the person to whom directives have been issued are both moral in character and constitute categorical reasons, where a categorical reason is one that applies to the person independently of his or her particular goals and aspirations. Suppose, for example, that the issue is how safely to transport a certain dangerous substance, where transporting the substance poses potentially serious risks of harm to other persons. The relevant background reasons that apply to transporters of the substance are reasons to avoid creating such risks for others, and are therefore categorical in nature. Suppose that a given governmental agency in fact knows much more about the transportation of this substance than I do, and that I will therefore do much better in complying with the background reasons of safety that apply to me if I obey the agency’s directives than if I try to decide for myself how the substance ought to be transported. The general idea of the NJT is that this fact gives me reason to comply with the directives the agency has issued, and that since it would be "double-counting" if I were also to try to take account of the background reasons directly, I should treat the directives as "preempting" those background reasons.8 The fact that the background reasons are preempted is supposed to explain the mandatory nature of my reason to follow the directives.9 Furthermore, by preempting the background reasons, the directives are supposed to replace those reasons for me. Since the background reasons are categorical in nature, it is very plausible to think that a reason that replaces them is likewise categorical. If whenever I engage in the activity of transporting the dangerous substance I have a reason to follow the agency’s directives that is both mandatory and categorical, that would appear to be sufficient to establish that I have a moral obligation to follow the directives. If I have a moral obligation to comply with the directives, that would appear to be sufficient to establish that the agency has the legitimate authority to issue the directives, which is simply to say that it possesses the normative power that it claims for itself and that it purports to be exercising in issuing the directives.

Let me call the argument of the preceding paragraph the justificatory argument. The difficulty with the argument, which comes in the last step, is that it proves too much.10 Suppose that everything about the example remains the same, with the sole exception that the governmental agency issues advisory recommendations about how the dangerous substance ought to be transported, rather than directives that are meant to be obligatory. In other words, the agency does not purport to exercise a power to impose an obligation on me; it simply offers me advice, and then leaves it up to me to make a decision about what precautions to take if I decide to transport the substance in question. To ensure that no other aspect of the hypothetical is modified, we have to assume that the background reasons of safety, at least insofar as they apply to me, are not affected by the possibility that fewer persons overall will comply with the agency’s views about how to transport the substance if those views are issued not as obligatory directives but simply as advisory recommendations. Let me therefore assume that the case is one in which my reason to follow the agency’s directives depends solely on its expertise.11 The assumption, then, is that I will do better in complying with the safety reasons that apply to me in transporting the dangerous substance if I follow the agency’s views about what ought to be done, regardless of how many other people do likewise. On that assumption, the justificatory argument appears to go through, right up to the last step. Given that the background reasons are categorical, and given that I will do better in complying with those reasons if I treat the agency’s recommendations as preemptive and hence as mandatory, it is difficult to see how to avoid the conclusion that I have a moral obligation to follow the agency’s views even though they are only issued as recommendations and not as directives. In other words, I have a moral obligation regardless of whether the agency purported to exercise a normative power to impose an obligation on me. But if I have the obligation regardless of whether or not the agency claimed to possess and purported to exercise a power to obligate me, it is difficult to see how the argument can justify the conclusion that the agency possesses such a power.

Perhaps it might be suggested that even though I have the obligation in both cases, that is not a reason to deny that, if the agency does happen to claim and to purport to exercise a power to impose obligations, the justificatory argument is sufficient to justify the conclusion that it possesses this power it claims for itself. But this suggestion would lead to some strange results. Suppose that no governmental agency exists to regulate the transportation of dangerous substances, but that I have a friend who has exactly the same level of expertise that we have been attributing to the agency. My friend gives me advice about how to transport the substance that is identical in content to the directives or recommendations that we were supposing would have been issued by the governmental agency. So long as I have reason to know that I will do better in conforming with the background categorical reasons that apply to me by following my friend’s advice than if I were to try to follow my own judgment, it is once again difficult to avoid the conclusion that the justificatory argument establishes that I have a moral obligation to follow my friend’s advice. But now suppose that, instead of simply advising me, my friend claims to exercise a power to obligate me to do as she says. In accordance with this claimed power she does not simply advise me but commands me, or issues a directive to me, to do such-and-such if and when I transport the dangerous substance. Does the justificatory argument establish that she does, in fact, possess the moral power to obligate me that she claims for herself? It seems very odd that the mere possession of a certain kind of expertise, coupled with nothing more than a claim that one possesses a normative power to obligate others, should be sufficient to establish that one does, in fact, possess such a power. Raz himself draws attention to the similarity, from the point of view of the NJT, between obligatory directives and advice when he writes that, in pure expertise cases, "the law is like a knowledgeable friend."12 But this similarity, far from lending support to the justificatory argument, actually undermines it. If, in cases of pure expertise, the justificatory argument is sufficient to justify the existence of a moral obligation to follow the views of another person about what to do, it does so whether the other person simply offers advice or claims to issue binding directives. And if the suggestion is made that whenever the justificatory argument justifies an obligation on the part of B to do as A says then it also justifies any claim that A might happen to make to have the power to obligate B, then we will be led to find normative powers in some rather odd places.
To come at the difficulty from a slightly different direction, consider what it means to say that one person possesses a normative power over another. Raz defends a very plausible view that takes something like the following form: One person A has a power to effect a certain kind of change in the normative situation of another person B if there is sufficient reason for regarding actions which A takes with the intention of effecting a change of the relevant kind as in fact effecting such a change, where the justification for so regarding A’s actions is the desirability of enabling A to make this kind of normative change by means of this kind of act.13 The difficulty that I am arguing the justificatory argument faces is that, in trying to justify the conclusion that one person possesses authority over another, it makes no essential reference to the desirability of the first person being able intentionally to change the normative situation of the other. The NJT looks out at the world, so to speak, from the perspective of an individual who is seeking assistance wherever he can find it in helping him to conform to right reason. From this point of view, leaving aside for the moment problems of coordinating activity, there is no particular reason to distinguish advice from directives claiming to be authoritative. For that matter, it does not particularly matter whether the "advice" comes from a person or from a black box, which, for all I know, might not even be subject to the control of another person. So long as I know or can readily come to know, on whatever basis, that if I follow the "recommendations" that appear on the screen of the black box then I will do better in complying with the background reasons that apply to me, the justificatory argument seems to go through. Notice that it is not sufficient to distinguish advisory recommendations from obligatory directives to say that, in the former case, the recommendations give me reasons for belief but not reasons for action, or that the advisor is a mere theoretical authority for me but not a practical authority. In the case of the governmental agency, for example, the justificatory argument appears to go through whether the agency issues advisory recommendations or obligatory directives. Furthermore, in the case of both recommendations and directives, the argument only goes through if I have a basis for knowing, as a general matter, that I will do better by complying with the agency’s views than if I try to act on my own judgment.14 It may or may not be true that, on any particular occasion when I transport the dangerous substance, I have a reason to believe that I will do better on that occasion by following the agency’s views, and this is so whether those views take the form of advisory recommendations or authoritative directives. What matters in both cases is, to repeat, that I must have good reason to believe that I will in general better comply with the reasons of safety that apply to me if I follow the views of the agency.

In light of the above considerations, it is noteworthy that, in his most recent work on authority, it is no longer clear that Raz means to offer the service conception as a full-fledged response to what I have been calling the justification problem, i.e., the problem of how to justify legitimate political authority. In a recent article, Raz states that the service conception is driven by a theoretical problem and by a moral problem.15 The theoretical problem concerns the issue of "how to understand the [normative] standing of an authoritative directive." The moral problem is, "How can it ever be that one has a duty to subject one’s will and judgment to those of another?" The moral problem is, clearly, just another name for the subjection problem that I identified earlier. The theoretical problem, however, is not the same as the justification problem. In response to the theoretical problem, Raz observes that "[a] person can have authority over another only if there are sufficient reasons for the latter to be subject to duties at the say-so of the former." He adds that this observation does not tell us when anyone has authority over another or even that anyone ever can have such authority. It only states what has to be the case for one person to have authority over another, and "[t]hat is all that one can ask of a general account of authority." The theoretical problem thus apparently raises no more than a bare conceptual issue, which is answered, in essence, by pointing out that one person can have authority over another only if there are sufficient reasons to justify the possession by the former of a normative power to impose obligations on the latter. Not only is the service conception not explicitly advanced as a general response to the justification problem, the justification problem is not even formulated as a distinct problem in its own right. The task of determining who has authority over whom and with regard to what "is a matter of evaluating individual cases."

Raz then goes on, in the same article, to offer the service conception as a specific response to the moral problem, which, as already noted, is what I have been calling the subjection problem. It is worth quoting Raz’s restatement of the service conception in full:

The suggestion of the service conception is that the moral question is answered when two conditions are met, and regarding matters with respect to which they are met: First, that the subject would better conform to reasons that apply to him anyway (that is, to reasons other than the directives of the authority) if he intends to be guided by the authority’s directives than if he does not (I will refer to it as the normal justification thesis or condition). Second, that the matters regarding which the first condition is met are such that with respect to them it is better to conform to reason than to decide for onself, unaided by authority (I will refer to it as the independence condition).16

Raz anticipates the objection that the two conditions do not solve the moral problem because the second condition merely restates it. He responds that the independence condition "merely frames the question," and that part of the answer to the moral problem is to be found in the first condition, namely the NJT.17 Raz argues that "the key to the justification of authority" is that, far from hindering our rational capacity, i.e., our capacity to guide our conduct in accordance with our own judgment, authority actually facilitates this capacity. We value our rational capacity not just because we value the freedom to use it, but also because "its purpose…by its very nature, [is] to secure conformity with right reason." When the conditions of the NJT are met, we are better able to achieve this purpose by acting in accordance with the relevant authority’s directives than if we try to exercise our rational capacity directly. In such cases, the importance of conforming our actions to right reason outweighs the importance of self-reliance and acting on our own independent judgment. Raz goes on to point out that, because we are hardwired to respond instinctively to certain dangers, sometimes we do better in achieving the purpose of our rational capacity by acting on our emotions. More generally, authority is just one technique among others that is capable of helping us to achieve this purpose: The primary value of our general ability to act by our own judgment can also be met by "making vows, taking advice, binding oneself to others long before the time for action with a promise to act in certain ways, or relying on technical devices to ‘take decisions for us’, as when setting alarm clocks, speed limiters, etc." To understand authority properly we must therefore see it "not [as]a denial of people’s capacity for rational action, but rather [as] simply one device, one method, through the use of which people can achieve the goal (telos) of their capacity for rational action, albeit not through its direct use."
This discussion of our rational capacity is illuminating and insightful. It shows that the two conditions of the service conception, namely, the NJT and the independence condition, together offer a response to a very broad understanding of the subjection problem. In doing so it makes clear why we are justified, in appropriate circumstances, in subjecting our will not just to certain demands or requests by other persons that we behave in a certain way, but also to advice that others give us and even to mechanical devices such as speed limiters. (Recall the earlier discussion of a black box that dispenses "advice," but which may not be under the control of, or even have ever been programmed by, a human being.) However, in drawing attention to the fact that authority is just one device or technique among others that can enable us better to comply with right reason, this response to the subjection problem simply underscores my earlier point that the NJT makes no essential reference to the claim of de facto authorities to be exercising a moral power to impose obligations when they issue their directives. As I noted earlier, the NJT looks out at the world from the perspective of an individual who is seeking assistance wherever he can find it in helping him to conform to right reason. So long as I have reason to know that I will in general do better in complying with the reasons that apply to me in a given type of case by following the views of another person rather than by acting on my own judgment, it does not matter whether those views are offered in the form of advice or in the form of directives. Thus, although it may well be a necessary condition of one person’s possessing legitimate authority over another that the NJT or some similar condition be met, the service conception cannot, certainly as a general matter, offer a full-fledged response to the problem of justifying legitimate authority. The service conception is not in the end best viewed as a conception of authority as such but, rather, as a general response to a very broad understanding of the subjection problem.

In light of the preceding discussion, it is misleading to say, as Raz does in his recent article, that "[t]he function of authorities is to improve our conformity with…background reasons by making us try to follow their instructions rather than the background reasons."18 In the case of political authority, the function of governments is not to improve our conformity with right reason for its own sake but, rather, to accomplish important moral goals that governments are uniquely suited, or at least particularly well suited, to achieve on behalf of their subjects by means of the normative instrument of a power to impose obligations. Background reasons will figure in the justification of the power, but that is not the same as saying that the power is justified because it improves conformity with background reason. For example, Raz has always regarded appeal to the ability of governments to achieve coordination, understood in a broad sense and not just as a solution to a narrowly-conceived, Lewis-style coordination problem, as one of the two main ways that the service conception can justify political authority.19 (The other main way, which I have already discussed, is grounded in the possession of greater expertise.) Although any response to the justification problem that is based on securing coordination must contend with a number of difficult questions,20 it nonetheless seems clear that, one way or another, the ability of governments to coordinate complex activity will inevitably play a prominent role in the justification of most political authorities, to the extent that they can be justified. It also seems clear that any justification of political authority that rests on the ability of governments to coordinate activity must at some point appeal to an interest that people generally have in the successful achievement of coordination. There is no harm in saying that, because people have such an interest, they have a background reason "to wish for a convention,"21 as Raz has put it, or something along those lines.

We nonetheless overlook an important dimension of what is valuable about coordination if we limit ourselves to saying that the power to issue coordination-securing directives is justified because the possession and exercise of this power will enable people better to conform their own actions with the requirements of right reason. Even if "a reason to wish for" coordination is sometimes a background reason for me to act in such a way that my own activities are coordinated with those of others, I have an interest in the existence of general social coordination, meaning coordination in many different aspects of public life, which goes far beyond ensuring that my own actions conform to this kind of background reason. I will benefit, for example, in countless different ways because other people’s activities have successfully been coordinated so as to ensure that the highways are adequately maintained and the air traffic control system operates efficiently. The achievement of coordination is a valuable moral goal in its own right, quite apart from the extent to which particular individuals might in their own actions better conform with reasons that apply to them. The fact that many people will benefit from a set of directives that successfully coordinates the air traffic control system will inevitably figure in any argument that is sufficient to establish that the government is justified in issuing these directives, and this is so even though the directives do not directly engage the background reasons for action of many of the people who benefit in this way.

As Raz has observed, his general view of authority entails the substantive moral thesis that all political reasons are subordinated to ordinary individual morality.22 But even if this thesis is true, which is not a self-evident matter, it is important to make clear that the relevant background reasons of individual morality are often very general and only indirectly related to the actions the government is empowered to take. As Raz says, the fact that everyone has reason to improve their own economic situation does not mean that everyone has a reason to raise taxes: "[T]hose helping us may have good grounds for pursuing the goals set by reasons that apply to us in ways that are not open to us," and, indeed, "they may be assigned the task of helping us precisely because of that."23 While this is certainly correct, it nonetheless seems strained to say that the government’s role in such a case is to help every individual to act in ways that better conform with the reason they have to improve their own economic situation, as opposed to saying that the government’s role is, quite simply, to achieve the moral goal of improving everyone’s economic situation.

To see the general point more clearly, consider the following example. Suppose that a society has good reasons of security to have a standing army of a certain size, but that it would be wasteful of resources to have an army of any larger size. Assume that the government is justified in achieving this goal by conscripting the required number of soldiers from the pool of citizens who are generally fit to serve in the army, and that it is also justified in determining who is to be drafted by instituting a fair lottery. It is no doubt true, in such a case, that each citizen has a background moral duty to contribute where necessary to maintaining the security of the society, so that it is true of each citizen who is drafted that she is complying with a background reason that applies to her. But does it follow that she is better complying with that reason by obeying the directive to serve in the army? After all, she only has the duty to serve because her name happened to come up in the lottery. Similarly, is it sensible to say that the function the government is fulfilling in instituting the draft-by-lottery is to enable each citizen who has been drafted to conform better with a background reason that applies to her? If that were the government’s function, why is it only helping Susan by drafting her, and not helping John by drafting him as well? Are we to say instead that the government is somehow serving the goal of helping each citizen who might have been drafted to better conform with a background reason? Surely we make the best sense of such a case by saying, quite simply, that the government’s function is to take steps to achieve the general moral goal of providing for the society’s security by means of the exercise of a power to impose duties on (some) citizens. No doubt there is a background reason at work in any argument that might plausibly be offered to show that the government does, in fact, possess this power, but we nonetheless mis-describe the situation if we say that the possession of the power is justified because its exercise ensures that those whose normative situation is affected will better comply with that background reason than would otherwise be the case. That is why I said earlier that it may well be a necessary condition of a government’s possessing legitimate authority over any particular person in regard to any particular matter that the NJT, or some condition similar to the NJT, be satisfied. What is necessary is that the justification of the power invoke some background reason that applies to any person whose normative situation is affected by its exercise. But it does not follow that the exercise of the power enables any such person to better comply with that reason. Still less does it follow that this was the purpose of exercising the power.
There is another, related difficulty to which the service conception gives rise, insofar as it holds that the function of authority is, quite simply, to improve conformity with right reason. Raz has observed, in connection with earlier formulations of the service conception, that the NJT "invites a piecemeal approach to the question of the authority of governments, which yields the conclusion that the extent of governmental authority varies from individual to individual, and is more limited than the authority most governments claim for themselves in the case of most people."24 Thus if, to revert to an earlier example, I happen to know much more about how safely to transport a dangerous substance than does the governmental agency that has been charged with regulating such matters, the piecemeal approach suggests that the agency does not have the authority over me that it claims and that I am not morally bound by its directives. If we say that the function of authority is, quite simply, to enable me better to conform with background reasons, then it becomes difficult to make sense of the fact that legal systems claim virtually unlimited power to regulate any activity or aspect of life,25 and that they do not acknowledge any exceptions to that authority that are not explicitly recognized by the law itself. No doubt Raz is correct to say that it is not a condition of adequacy of an explanation of the concept of authority that those who have authority accept, even implicitly, that the explanation is correct.26 Even so, one would nonetheless expect any adequate explanation to make sense of, in the minimal sense of being consistent with, the behavior of those who claim authority. If the function of authority really were to enable persons better to conform with right reason, then one would expect there to be many cases in which the claim of authority could be justified by appeal to pure expertise. One would nonetheless also expect in such cases that the law would act, as Raz has put it, "like a knowledgeable friend,"27 acquiescing gracefully when a subject either knows more than it does or succeeds in finding some third party who knows more.28 Regarded as a general response to the subjection problem, the NJT counsels me to seek assistance in conforming to right reason wherever I can find it, so that if I have good grounds for believing that I will generally do better in safely transporting a dangerous substance by following the law of California than by following the law of Pennsylvania, then I should follow the law of California. But the law of Pennsylvania does not see the matter this way. It insists that, insofar as I fall within its jurisdiction, I have an obligation to follow its directives except to the extent that the law itself makes room for some exception. Far from encouraging me to seek assistance in conforming to right reason wherever I might find it, the law tells me that I am obligated to follow its directives, and that is the end of the matter. This suggests that there is an almost unavoidable tension in regarding the NJT as a full-fledged response to both the subjection problem and the justification problem.

Insofar as he has, at least in the past, treated the service conception as a full-fledged response to the justification problem, Raz has tended to view the analysis of authority as beginning with the one-to-one relation between an authority and a single person subject to the authority, so that authority over a group of persons is to be accounted for by reference to authority relations between individuals.29 This aspect of the service conception goes hand-in-hand with the piecemeal approach to political authority, which suggests, in turn, that the paradigm of an authority relation is a case of pure expertise. If, however, we adopt the view that the principal function of political authority is not to enable individuals taken one by one to better conform with right reason but that it has the function, rather, of accomplishing important moral goals by means of a power to impose obligations, then we are likely to see the paradigmatic instances of authority as those in which the government’s actions can be understood, to borrow a term from Finnis, as intended to advance the common good. As Finnis recognizes, large-scale coordination of complex activity is perhaps the best example of how the common good can be advanced.30 Once we reject the idea that the function of political authority is to enable individuals to better conform with right reason one by one, then we are also naturally led to the view that an exercise of political authority can probably only rarely be justified by appealing solely to greater expertise. One way or another, the justification of a power to issue directives about, say, the proper manner of transporting a dangerous substance will almost always have some coordinating aspect, such as the value of ensuring that persons generally, and not just those who are engaged in the activity of transporting the substance, are able to form reliable expectations about how this activity will be carried on.31 If we understand such cases as almost always resting on more than just a claim of greater expertise, then the task of justifying political authority becomes a much less formidable task than it would otherwise be, and the authority of governments is far less likely to be piecemeal. By the same token, however, we do not succeed in justifying political authority simply by showing that the persons over whom authority is claimed will better comply with a background reason that applies to them than if they were to try to act according to their own judgment. To say this much is to say no more than that those persons may well be justified in subjecting their will to that of the authority. But solving the subjection problem does not mean that we have solved the justification problem. The NJT, and the service conception generally, must be understood as contributing primarily to the solution of the subjection problem. Although a response to the subjection problem will almost invariably figure in any attempt to justify a claim of legitimate political authority, we have no reason to assume that, as a general matter, this will be sufficient by itself to establish that a given claim to authority can be justified.

Endnotes

1. Raz argues that because legal systems, unlike other normative systems such as commercial companies and sports organizations, acknowledge no limitations on the spheres of behavior which they have the power to regulate, their claim of authority is "comprehensive." Joseph Raz. The Authority of Law (Oxford: Clarendon Press, 1979), 116-17. Elsewhere he makes a similar point by observing that even in states in which the power of the legislature to legislate is constitutionally limited, "the law" nonetheless claims unlimited authority for itself "because the law provides ways of changing the law and of adopting any law whatsoever…." Joseph Raz. The Morality of Freedom (Oxford: Clarendon Press, 1986), 76-77.

2. See, e.g., Joseph Raz, "Hart on Moral Rights and Legal Duties," Oxford Journal of Legal Studies 4 (1984): 123-31.

3. John Finnis. Natural Law and Natural Rights (Oxford: Clarendon Press), 250.

4. Robert Paul Wolff. In Defense of Anarchism (New York: Harper & Row, 1970), 1-19.

5. Raz, The Morality of Freedom, 53.

6. Cf. ibid., 63: "The service conception is a normative doctrine about the conditions under which authority is legitimate and the manner in which authorities should conduct themselves."

7. Cf. ibid., 38-42, where Raz discusses the problem of "surrendering one’s judgment."

8. Ibid., 41-42, 57-59. It is worth noting that, in Raz’s early formulations of the service conception, binding directives were apparently supposed to exclude acting on all the background reasons (or dependent reasons, as Raz sometimes calls them) because that would be double-counting. See, e.g., ibid., 42: "[R]easons that could have been relied upon to justify [an arbitrator’s] decision before his decision cannot be relied upon once the decision is given." However, in a recent article, Raz appears to have weakened the double-counting prohibition to some extent. Binding authoritative directives are now said to exclude acting only on those background reasons which the lawmaker was meant to consider and which conflict with the directive. Raz observes, quite sensibly, that the preemptive or exclusionary nature of a binding directive does not exclude relying on reasons for behaving in the same way as the directive requires, but only those reasons "on the losing side of the argument." "The Problem of Authority: Revisiting the Service Conception," Minnesota Law Review 90 (2006): 1003-44, 1022. Sensible as this is, it clearly permits double-counting, since there is nothing to prevent me from acting on both the directive and the winning background reasons. Furthermore, since it is difficult to see how I can act on the winning reasons without knowing and taking account of the fact that they outweigh the losing ones, it seems to follow that, even though I cannot act on the losing reasons alone, I can nonetheless act on the totality or balance of background reasons, and, indeed, I must do so whenever I act on the basis of the winning reasons. In light of these considerations, it is no longer clear exactly what, on Raz’s general account of practical reasoning, a preemptive or exclusionary reason is supposed to exclude.

9. Raz, The Morality of Freedom, 60.

10. Cf. Stephen Perry, "Law and Obligation," The American Journal of Jurisprudence 50 (2005): 263-95, 280-81.

11. This means, among other things, that we are not dealing with a certain kind of case that Raz has discussed elsewhere, in which it would be futile for me to follow a given standard of conduct unless many other people also follow it. See Raz, The Authority of Law, 247-48. Raz gives the following example. If a sufficiently large number of people refrain from polluting a river, then the river will be clean. Any given person has a reason not to pollute only if a sufficiently large number of others do likewise, since otherwise the action of a single individual in refraining from polluting will be futile. Notice that this is a case where any given individual’s reasons are affected by the existence or non-existence of a general social practice, but where the practice does not amount to a Lewis-style solution to a coordination problem.

12. Joseph Raz, Ethics in the Public Domain (Oxford: Clarendon Press, 1994), 332.

13. Raz, The Authority of Law, 18.

14. Raz explicitly acknowledges the epistemic basis of his account of legitimate authority when he writes that "[i]f one cannot have trustworthy beliefs that a certain body meets the conditions of legitimacy [i.e., the NJT and an "independence condition" which holds that it more important, in the relevant type of case, to conform to reason than to decide for oneself], then one’s belief in its authority is haphazard and cannot…be trusted. Therefore, to fulfill its function, the legitimacy of an authority must be knowable to its subjects." Raz, "The Problem of Authority: Revisiting the Service Conception," 1025. It is worth noting in this regard that Raz further acknowledges that, in the case of both authority and advice, it is ultimately up to each individual to decide for himself whether or not the conditions of legitimacy are met: "[I]n following authority, just as in following advice…one’s ultimate self-reliance is preserved, for it is one’s own judgment which directs one to recognize the authority of another, just as it directs one too keep one’s promises [and] follow advice…." Ibid., 1018.

15. Ibid., 1012-13.

16. Ibid., 1014.

17. Ibid., 1017-18.

18. Ibid., 1019 (emphasis added).

19. See, e.g., Raz, The Morality of Freedom, 49-50.

20. Cf. Leslie Green, The Authority of the State (Oxford: Clarendon Press, 1988), 89-121.

21. Raz, The Morality of Freedom, 50.

22. Ibid., 72.

23. Raz, "The Problem of Authority: Revisiting the Service Conception," 1030.

24. Raz, The Morality of Freedom, 80.

25. See note 1 above.

26. Raz, "The Problem of Authority: Revisiting the Service Conception," 1006.

27. Raz, Ethics in the Public Domain, 332.

28. Cf. Perry, "Law and Obligation," 283-84.

29. Raz, The Morality of Freedom, 71.

30. Finnis, Natural Law and Natural Rights, 153.

31. Raz acknowledges that expertise and coordination are, in the case of political authority, usually "inextricably mixed," and that because of the resulting importance of achieving coordination, generally speaking only de facto authorities, meaning bodies that are effectively able to ensure that people will do what those bodies say, can be legitimate. Raz, "The Problem of Authority: Revisiting the Service Conception," 1031, 1036.

Previous Article | Index | Next Article

APA Newsletters

Spring 2007 Volume 06, Number 2

Spring 2007
Volume 06, Number 2