36 Authority
(2016: 92): "When voluntary rule is achieved, there is at least some partial alignment between what an individual values and what goods are promoted by the political order to which he is subject." States, according to her, can be more or less legitimate, depending on the proportion of people that consent (2016: 87). Grossly uninformed consent does not count, though, because it does not achieve the said alignment between what an individual values and what the state actually does (2016: 85).
Now the first thing to note is that rights are not scalar, but binary; one either has a right or one lacks it. So if the ideal of a voluntary state is to explain the right to rule and hence political authority, there has to be a threshold of support at which the state gets the right to rule. But every such threshold would look arbitrary; why should a state that enjoys the consent of 67 percent of its subjects have the right to rule, while a state that enjoys the consent of 66 percent does not, for example? Second, treating consent as a mere ideal simply does not explain how some people can come to have the right to rule over non-consenting equals. Accordingly, some advocates of the ideal of a consensual state feel free to simply ascribe political obligations to those who do not consent (Walzer 1970), or regard them as "political children" that may be governed without consent (Tussman 1960: 36-7). I conclude that treating consent as an ideal does not meet the explanation condition.
Summary
Consent could well explain how new rights and hence political authority have come into being. Unfortunately, our states actually do not have everyone's consent (whether explicit or tacit). If consent were necessary for state authority, all states would therefore lack political authority. Neither is it possible to achieve consensual states via institutional reform. Mere hypothetical consent or normative consent, on the other hand, cannot explain political authority.
3
The Service Conception of Authority
The service conception of authority has been devised by Joseph Raz. Its basic idea is to conceive authorities as a service for those who submit to them. It is an intriguing theory because it promises to account for all sorts of authority, from the authority of experts to the authority of parents and state officials. The service conception encompasses a theory about how authority should be exercised, a theory about how authority-based reasons work, and a theory about the justification of authority.
How to exercise authority
I start with the first part of the theory, how authority should be exercised. Authority, according to the service conception, is to be understood as a valuable instrument for those subject to it. They do better by following the advice or command of the authority. According to the service conception, the role and primary function of authorities is "to serve the governed" (Raz 1986: 56). It is to serve them by helping them to believe or do what they have reason to believe or do. Accordingly, authority should be exercised in line with the reasons that
38 Authority
apply to the subject independently of the authoritative directives. As Raz puts it: "[A]ll authoritative directives should be based on reasons which already independently apply to the subjects of the directives and are relevant to their action in the circumstances covered by the directive" (1986: 47). This is what Raz calls the dependence thesis. It is a normative thesis about how authority should be exercised. Of course it is not a thesis about what (de facto) authorities actually do. It is about what they should do.
Obviously, the dependence thesis is fairly abstract. The reasons that apply to a subject (and that the authority should base its directives on) can be of very different kinds. Sometimes an authority helps the subject to better further his or her own interests or well-being; sometimes the authority helps the subjects to follow the moral reasons that apply.
The interesting point is that the dependence thesis is to hold for all different kinds of authority. All authorities are to "serve the governed." In that sense, the service conception scores very high on the target condition (see p. 15), since it promises a unified account of all different forms of authority. The dependence thesis is certainly highly plausible for theoretical authorities. Opera experts and physicians can give bad advice on purpose, but this obviously is not what they should do. They should indeed help people who need their authority. A physician, for example, should advise the patient to take a particular antibiotic just in the case when the patient has good reasons to do so, usually because taking the antibiotic promises to restore the patient's health. If asked which recording of Tosca one should buy, an expert in Puccini operas should give advice in light of what he thinks the best recording is. In other words, he should give advice in light of the reasons that apply to people who are interested in purchasing a good recording of Tosca (independently of the advice of the expert).
How about practical authorities? Should they also base their directives on reasons that apply to their subjects independently of the directives? In the case of parents, the answer is "yes." It is quite intuitive to think that parents' authoritative directives should be a service to their children. For example, when a father orders his daughter to do her homework, then this is because doing her homework will
The Service Conception of Authority 39
help her succeed at school, which again will help her in her adult life. It is also plausible to conceive the authority of school teachers as a service to the children. Indeed, while consent theory can be understood as a theory that takes bosses and employees as the paradigm of proper - because consensual - relations of authority, the service conception can be understood as taking parents and children (or maybe teachers and children) as the paradigm of proper - because serviceable - relations of authority.
But the dependence thesis looks less plausible when it comes to the authority of bosses and the authority of states and their officials. Here it is harder to see in what sense authoritative directives could be a service to the governed. To understand how these forms of authority can be conceived as a service, it is important to recall that an authority need not base its directives on reasons that further the subject's personal interests. Even supporters of workplace democracy may concede that some sort of bossism is usually unavoidable for leading a successful firm. If we assume that an employee has reason to contribute to the success of the firm, then a boss who gives directives with an eye to the success of the firm can be said to base her directives on reasons that independently apply to the employee. Insofar as a boss should also care about the well-being of the employees, she should of course also base her directives on considerations about the well-being of the employees. The same holds with reasons that relate to the social and ecological responsibility of the firm: When a boss gives directives with an eye to this responsibility, she looks at reasons that again apply to the employee independently of the directives.
The same holds for states and state officials. A judge in a civil law case provides a service to the parties that take their argument to the court. The judge should decide the case in light of the facts and so of the reasons that apply to the parties independently of his judgment. No matter how he decides, the arbitration service will serve the parties. One thing that can be learned from this case, therefore, is that the reasons that the authority takes into account (the merits of the case) need not be identical with the parties' reasons to submit to an authority and to follow the directive of the authority (their need for an arbiter) (Viehoff 2011: 257-8).
40 Authority
Of course, states do much more than providing civil courts. But other activities can be reconstructed as a service, too: When the state enacts a law against murder, then of course this law can be regarded as based on reasons that apply to everyone independently of the law, namely on moral reasons not to murder. And so on.
Is there anyone who thinks that authorities may disregard the reasons that apply to those who are subject to the authoritative directives? How could one oppose the dependence thesis} One objection - discussed by Raz himself (1986: 45-6) - is that sometimes there are no antecedent independent reasons that would apply to the subjects. Before tax laws are enacted, for example, citizens do not have any reasons to pay taxes. Does this speak against the dependence thesis: Well, of course one could say that even before tax laws are enacted subjects have reasons to support certain causes that are advanced by taxes, or that they have reasons to in general support reasonably just states. In that sense, one can always come up with relevant reasons that apply to the citizens before the authority's directive. The dependence thesis, then, looks pretty plausible, if not trivial.
How authority-based reasons work
Authority should be a service to the governed, according to Raz. To better understand the service conception we should now discuss how authority-based reasons are supposed to work compared to other reasons for action or belief. By authority-based reasons, I mean the reasons that are constituted by an authority's order, advice, or opinion. These reasons can be either reasons for belief or reasons for action. When an expert on opera offers her opinion on an opera performance, people have reason to believe that her judgment is true or at least illuminating. When a physician gives advice on what medicine to take, one has reason to do as he says.
How do these authority-based reasons work? If we focus on practical authorities, the first thing to note is that authority-based reasons do not merely provide information about the reasons that apply independently to the agent. According to
The Service Conception of Authority 41
Raz, an authority's directive makes a difference to the balance of reasons (1986: 29-30, 48-51, 67). An authority, just by say-so, can add something to the balance of reasons. Take traffic laws, for example, like a law to drive on the right-hand side of the street. That the state orders drivers to drive on the right is not informative about an alleged independent reason to drive on the right. It constitutes all by itself a reason to drive on the right. This shows that a practical authority's power to impose duties is not a mere side-effect power (see p. 11): When an authority orders something, it creates a content-independent new reason for action by mere say-so.
How does all this fit with the dependence thesis from above? Did not the dependence thesis say that authorities have to give directives with an eye to the reasons that independently apply to the subjects? How can directives be made in light of reasons that apply independently, but on the other hand do more than inform about these reasons? Well, in the case of traffic laws, for example, the state should indeed base its directives on reasons that apply independently on the subject; but, of course, people do not have a reason to drive on any particular side of the street without the directive of the state (or some other institution or convention). They have reason to drive in a well-coordinated manner, and this is what the state should look at when giving directives about how to drive. Or take the case of tax laws again. Here quite obviously the state does not inform the citizens about reasons to pay taxes that apply independently of the state's directives. There are no reasons to pay taxes without tax laws, since it is not even possible to pay taxes without tax laws and the relevant fiscal authorities. Nevertheless, the state should give tax laws with an eye to reasons that apply independently to the citizens, like reasons to support the state's functioning or to support particular tasks of the state.
There are two caveats, though. First, when a practical authority gives a directive that merely specifies a pre-existing duty, it does not seem to add much to the balance of reasons. When, for example, the state enacts a law against murder, then the balance of reasons has not changed much. In addition to the moral duty not to commit murder, people now also have a legal duty not to commit murder (Sherman 2010: 426-7). Nevertheless, one can insist that it does add
an additional reason and is not merely informing about the balance of reasons. Second, that authority-based reasons are more than informative only holds for practical authorities. Theoretical authorities, for example experts on nineteenth-century opera, indeed provide reasons that merely inform about the balance of reasons, but do not add anything to the balance of reasons (see Raz 1986: 29).
I asked how authority-based reasons work. The first point was that they do more than informing about the balance of reasons: They change the balance of reasons (although this is not true in the case of theoretical authorities). The second point is that authority-based reasons are pre-emptive reasons. Authority-based reasons do not merely change the balance of reasons by adding another reason to the balance of reasons; they take the place of other reasons that apply to the agent. An agent follows an authority when she takes the directives of the authority as binding, such that she ignores what she might personally think she should do. In that sense, authority-based reasons replace other reasons. Raz's pre-emption thesis thus says: " [T]he fact that an authority requires performance of an action is a reason for its performance which is not to be added to all other relevant reasons when assessing what to do, but should exclude and take the place of some of them" (1986: 46).
Let me illustrate the pre-emption thesis with different kinds of authorities, starting with theoretical authorities. When I learn something about opera from Carolyn Abbate's and Roget Parker's book, I should take that piece of information or judgment as a reason for belief rhat replaces reasons to believe otherwise that I might have entertained before (see Raz 2006: 1033). This, at least, is what 1 do when I take them as authoritative. Likewise, when I think that I just have a light cold and can easily spend the day working, but my doctor says that I have the flu and should stay in bed and sleep all day, then I treat her as an authority when 1 stay in bed and ignore my own assessment of the matter. Same with practical authorities: When a police officer says that I may not continue driving, this is not just one reason among others that co-determines what I should do. I can no longer make a decision based on what I think the overall balance of reasons is. If I take the officer's directive as authoritative, her order replaces other reasons that might apply.
The Service Conception of Authority 43
Relatcdly, the pre-emption thesis can explain in what sense laws and rules can be called "authoritative" (and hence talk of the "authority of law"), even though laws and rules obviously cannot have rights (only people can). Rules and laws can be regarded as authoritative when they should be taken as binding and hence as replacing the reasons that apply to the agents independently of the rules (Raz 1979: Ch. 2; 1986: 57-9).
Two clarifications are important. First, the pre-emption thesis does not imply that one has to follow an authority's directives come what may. It just explains what it means to exercise authority and to follow authority. There are two kinds of cases where an agent should not follow an authority. One is the case of an illegitimate de facto authority. People take it as authoritative, but it is not. The other is the case of a legitimate authority that errs on a particular point and gives a bad directive. Usually, one should follow the directives of legitimate authorities even when one thinks they are bad, because otherwise there would be no point in submitting to the authority in the first place (Raz 1986: 47, 60-1; 2006: 1022-3). But sometimes, when an otherwise legitimate authority gives morally deeply problematic orders, one ought not to take that directive as authoritative. Note that the pre-emption thesis does not say that authority-based reasons should exclude and take the place of all other reasons that apply to the agent, but only some of them. Some reasons that are still taken into account by the agent may speak in favor of not submitting to the authority.
Second, depending on what kind of authority is at stake, one may have or not have a duty to follow an authority's directives (Sherman 2010: 424-6). When introducing the pre-emption thesis, Raz has practical authorities in mind, authorities with "the power to require action" (1986: 38). When we consider theoretical authorities, i.e. authorities who have great knowledge in a certain realm, people usually do not have a duty to follow their advice. Correspondingly, such authorities are usually not in a position to give orders. A physician, for example, usually offers advice, and the patient does not have a duty to follow. To be sure, the patient should take the physician's advice as a pre-emptive reason for action that replaces other teasons. But only
44 Authority
The Service Conception of Authority 45
practical authorities have the power to impose duties on people; it is only here that pre-emptive reasons come in the form of duties.
An objection against the pre-emption thesis says that // it is essential for legitimate authorities to provide pre-emptive reasons, then there might not be any legitimate authorities, because it is either irrational or immoral or both to take directives as pre-emptive. Why is that? Well, maybe because as human beings we ought to make up our own minds and decide autonomously what we should do (see Wolff 1970; Hurd 1991). This does not mean that we should never take the advice of experts. But in the end, we should always consider the balance of reasons for ourselves. This objection, though, exaggerates the conflict between autonomy and authority. Take arbitrators. Arbitrators can only function as arbitrators if their clients take their directives as binding, and of course people need arbitrators on many occasions. This does not seem to undermine people's autonomy. Moreover, as explained before, the pre-emption thesis does not claim that we should always follow legitimate authorities, come what may. There are occasions when one should not follow them.
Still, the objection makes an important point. It often seems rational not to take an authority's directive ot advice as a pre-emptive reason (Himma 2007: 12J-8, 131-3). When you would like to buy a recording of Verdi's Falstaff, then you certainly do well to consider the advice of Carolyn Abbate and Roger l'arker; but you can well take their advice as one consideration among others, and in the end assess the balance of reasons for yourself. Now one may say that you do not treat them as an authority if you assess the balance of reasons for yourself, but this seems to be an artificially narrow picture of what authorities ate. The pre-emption thesis, then, does not seem to be true of all authority-based reasons. Indeed Raz seems to agree that one need not take experts as providing pre-emptive reasons (2010: 300-1). An authority provides pre-emptive reasons mainly if some kind of coordination or arbitration is needed. This obviously narrows the applicability of the pre-emption thesis considerably. Only some authorities provide pre-emptive reasons; many do not.
The justification of authority
So far I have laid out the service conception's view of how authority should be exercised (it should be based on reasons that apply to the subject independently of authoritative directives) and how authority-based reasons work (they are pre-emptive and more than informative, at least on some occasions). What is most interesting in the context of this book is how we can justify authority: What has to be the case for someone to have legitimate authority? The service conception provides an answer that is closely related to the pre-emption thesis and the dependence thesis.
Consent theory claimed that practical authorities need the consent of the governed in order to be legitimate. (Parents and school teachers were exceptions because children cannot give valid consent to authority.) Valid consent was conceived as a necessary and sufficient condition for the legitimacy of these types of authority. What does the service conception say? The answer comes in the form of Raz's normal justification thesis:
The normal way to establish that a person has authority over another person involves showing that the alleged subject is likely better to comply with reasons which already independently apply to him [...] if he accepts the directives of the alleged authority as authoritatively binding and tries to follow thern, rather than by trying to follow the reasons which apply to him directly. (1986: 53)
Legitimate authorities really are a service to their subjects (and do not merely pretend to be). If people do better by submitting to an authority than by trying to figure out what to do by themselves, then the authority is legitimate. The account specifies who is an authority, over whom he is an authority, and with regard to what things he is an authority.
The normal justification thesis is coined for practical authorities, like parents, bosses, teachets, state officials, and religious leaders: Authorities that have the power to impose duties on people. Theoretical authorities cannot impose duties on people. As Raz notes, these authorities are not even properly said to have authority "over us" (2006: 1034).
46 Authority
The Service Conception of Authority 47
Nevertheless, something similar to the normal justification thesis could also be formulated for theoretical authorities (Zagzebski 2012). Arguably, someone is a theoretical authority relative to some subject when submitting to his or her judgment helps that subject to believe what she should believe or do what she should do.
Objections against the normal justification thesis
There are two important objections against the normal justification thesis. The first objection applies to the normal justification thesis as a claim about political authority. This objection says that the thesis fails to take account of the procedural and in particular the democratic sources of legitimate political authority (Waldron 1999: Ch. 5; Hershovitz 2003; Christiano 2004: 279-80; Himma 2007: 142-4; Schmelzle 2015: 140-9). Because we disagree about what reasons there are in politics and about what good or just laws would be, democratic procedures are essential to political authority. Yet in reply, one can argue that Raz's service conception can well account for the procedural dimension of democracies, since providing fair arbitration procedures is precisely a service as envisaged by Raz's theory (Vichoff 2011). It is a different question whether democratic procedures can by themselves explain political authority. I will get back to this in Chapter 5.
A second and in my view stronger objection is that the normal justification thesis fails to specify sufficient conditions for the moral power to impose duties and the liberty-right to enforce them that is essential for all forms of practical authority including political authority (Perry 2005: 280-3; Himma 2007: 140-2; Darwall 2009; 2010; Quong 2011: Ch. 4). That I would do better by following the directives of my dentist does not imply that my dentist has the moral power to impose duties on me and the moral liberty to coercively enforce them. Jonathan Quong presents the case of a tour company that offers by far the best tours of Peru at a reasonable price (2011: 108-9). I might best follow the reasons that apply to me if I book a tour with them and
follow their directives. Yet it seems that the company does not have moral powers over me as long as I do not actually book a tour with them. In particular, the fact that I would do better if I went with them docs not give the tour company the moral power to impose duties on me, like a duty to pay them, for example. Since it lacks a power to impose duties, it also lacks the liberty to coercively enforce these duties, of course. Providing a good service, then, is not sufficient for having rights to impose and coercively en force duties.
A first reply is that the normal justification thesis can be used separately for different aspects of authority. It could show that someone has the power to impose duties, but not to coercively enforce these duties, for example. Raz writes:
That an authority is entitled to impose a duty to F does not entail that it is entitled to impose a sanction for failing to F, or a remedy should any right be violated thereby. One needs a separate argument for chat, and the argument - on my account - would be provided if [the normal justification thesis] would apply to those additional measures. (2010: 300)
Similarly, sometimes the normal justification thesis may only show that someone is a theoretical authority (i.e. an expert in some field or good at doing something), but does not have the powet to impose duties. So far, so good. But one can stipulate that I would indeed better comply with reasons that apply to me independently of authoritative directives if the tour company could coerce me to comply with its commands. Once they brought me to Peru, I would see that it was good that they forced me on to their tour. Yet the tour company still does not seem to have the right to coerce me on to their tour, and so the first answer does not help to avoid the counterintuitive result that the tour company does have that right.
A second reply is that the normal justification thesis is not even supposed to specify sufficient conditions for legitimate authority. There are further necessary conditions for legitimate authority that are not satisfied in the tour company example. One is that only entities that claim to have authority could actually have authority (Raz 2006: 1005). Since the tour company does not claim authority,
48 Authority
it cannot have legitimate authority. This is true, and yet it seems that the tour company would lack authority even if it did claim to have authority.
Another necessary condition has to do with the value of independent decision-making. Raz makes clear that "the desirability of people conducting their own life by their own lights" can sometimes be more important than receiving the service an alleged authority could provide (1986: 57). He later calls this the "independence condition" (2006: 1014). The tour company example could thus be a case in point, where the independence condition explains why the tour company has no legitimate authority even though I would do better by complying with its directives. This reply is good, as far as it goes, but unless we are told under what conditions people should be conducting their own life by their own lights, even though submitting to an authority would be a service to them, the normal justification thesis is rather uninformative. In the end, it comes down to this: "The normal way to establish that a person has authority over another person involves showing that the alleged subject is likely better to comply with reasons which already independently apply to him if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly, except when this does not establish that the person has authority." The service conception could at best be one part of a theory that explains how persons can come to have authority over others, specifying one necessary condition among others. (Of course one could also have doubts that being serviceable is even a necessary condition for having authority; maybe valid consent is sufficient.) By itself, the service conception does not establish that someone has authority. In that sense, then, the service conception does not satisfy the explanation condition.
But let us put this objection against the normal justification thesis to one side and assume that the thesis is correct and has explanatory power. There is a third worry, then, namely that the normal justification thesis cannot show that at least some actual states have political authority (see Raz 1986: 74-8). At best, states could have vatying degrees of authority over their citizens. What authority they have
The Service Conception of Authority 49
would depend on whether a particular citizen needs certain state activities as a service or not. States would have more authority over one person than over another, and might well have no authority over some persons at all. Our states, on the other hand, claim a quite general right to rule over all citizens and persons in their territory (this is what I called the holistic nature of political authority; see p. 7). In that sense, the service conception cannot establish political authority as it is usually understood, and so it fails to meet the success condition, too.
Summary
The service conception conceives authorities as a service for those who submit to them. It encompasses a theory about how authority should be exercised, a theory about how authority-based reasons work, and a theory about the justification of authority. According to the last of these, one has legitimate authority if those who are subject to one's authority thereby better comply with reasons which already independently apply to them. The problem is that this fails to specify sufficient conditions for someone's moral power to impose duties and the liberty-right to enforce them, and thus the service conception cannot explain political authority on its own. Moreover, even neglecting that problem, the service conception cannot account for the holistic nature of political authority.
Community and Authority
Community-based justifications of political authority are the most straightforward ones. According to such justifications, political communities can be conceived by analogy to families. Just like there are moral bonds within families, there also are moral bonds within political communities, and these bonds include relations of authority.
Moral bonds in families
So let us start with families. We all believe that we have moral bonds with other family members, most clearly with our parents, children, and siblings, but also with less close relatives. These bonds are not emotional bonds, although of course there usually are emotional bonds, too; they are moral bonds which constitute moral obligations and rights. For example, as sons and daughters, we have moral obligations to support our parents when they get old and need our help. Obviously, this is not to say that the only reason to support them is that we have such obligations; if things go well, we support them because we love them. But the point is that we also feel that this is what we morally ought to do, that we would be morally blameworthy if we did not support them. As parents, we have moral obligations to raise and
Community and Authority 51
educate our children appropriately. Again, this is not to say that there are no excellent other reasons to raise and educate our children appropriately. If things go well, we raise and educate them as well as we can because we love them. But, again, we also feel that this is what we morally ought to do, and that we would be morally culpable if we failed to do so. Pointing out moral obligations does not mean saying that people should be primarily motivated by such obligations. The same holds for our brothers and sisters, and - to a lesser extent - for grandparents, uncles and aunts, etc. There is a sense of connection that most people feel is natural among family members.
The idea that we have familial obligations "is one of the most salient moral beliefs we have" (Hardimon 1994: 342). If we have moral beliefs that we are very certain about and could not give up without giving up our sense of who we are, these certainly include beliefs about moral bonds within families. Indeed we are so certain about moral obligations in families that they do not seem to require any further explanation; they seem to be morally basic.
It should be noted that biology is not decisive for the issue of familial obligations. Marriage usually counts as a way to become a member of a new family - the family of one's spouse. Adoption is a way to become a member of a new family. Relatedly, the strength of familial moral bonds depends on many different factors, and biology is just one among others in this regard. Biology matters, to be sure. That someone is one's biological aunt or cousin arguably constitutes some moral bonds; but more important is whether someone takes a certain familial role, in particular the role of a father or mother, brother or sister, and this is not necessarily the biological father, mother, brother, or sister. Moreover, the strength of familial moral bonds depends on how the connection between two persons is in real life, on the actual relationship (see Seglow 2013: Ch. 2). If someone has a close relationship with one's aunt, then the moral bonds arguably grow stronger than they usually are between aunt and nephew or niece. Conversely, if personal relationships are disrupted, the moral bonds become weaker than usual. In extreme cases, the moral bonds even disappear. When children are seriously mistreated by their parents, for
52 Authority
Community and Authority S3
example, their obligations to support their parents arguably vanish.
The existence of familial moral bonds can be regarded as a moral puzzle, even though we are so certain that familial moral bonds exist. They can be regarded as a moral puzzle because they imply a partiality that seems to be inimical to any universalist picture of morality with its emphasis on impartiality. On a universalist view of morality, all moral duties and obligations are either universal - owed to all other persons - or voluntarily incurred by means of promises, contracts, and the like. Familial obligations seem alien to this view of morality. As Michael Hardimon puts it (before defending familial obligations), there is a "horror at the thought of being impressed - like the seamen of old - into social roles and burdened with their attendant obligations against our will" (1994: 347). Familial and other partial obligations smack of a pre-enlightened, tribal view of morality that is to be overcome.
Of course there have been attempts by advocates of universalist ethical theories to accommodate the moral bonds in families, friendships, and political communities (Railton 1984; Goodin 1985: Ch. 5; 1988; Wellman 2000; Blake 2001; Stilz 2009). Thus Peter Railton has argued that our partiality toward our loved ones is justifiable on a consequentialist basis (1984). A sophisticated consequentialist is committed to what Railton calls "objective consequentialism." She wants to live a life that is justifiable from an impartial consequentialist point of view, but she is not committed to "subjective consequentialism" and thus rejects impartial consequentialist reasoning in her everyday life. A sophisticated consequentialist cannot even bring herself to impartial consequentialist reasoning when her loved ones are involved. The clue is that she has an objectively consequentialist justification for this: The world is a better place, from an impartial point of view, when people take special care for their families and friends. A sophisticated consequentialist can thus have true friendships and is not alienated from her deepest personal concerns. Railton's theory has been challenged (Badhwar 1991), and indeed it seems that a sophisticated consequentialist still has to condemn her actions as morally wrong when they favor her loved ones and are suboptimal from an impartial
consequentialist perspective (in a particular situation). This is still counterintuitive. Some, therefore, argue that there are moral bonds that are partial all the way down (Miller 1995: Ch. 3; Mason 1997; Scheffler 1997; Keller 2013; Seglow 2013). For the purposes of this chapter, it is not terribly important whether we believe that all partial moral bonds can in the end be based on impartial moral principles, or whether we endorse the view that some partial moral bonds cannot be based on such principles. What matters is that there arc moral bonds among family members, and that we are so certain about this that we do not seem to need to find any deeper grounding or justification for them.
So let us assume that indeed there are moral bonds within families, moral bonds that need no deeper explanation and thus can serve as a starting point in ethical reasoning. What has all this to do with authority} Moral bonds within families do not only consist in moral obligations. Parents also have powers with regard to their children, and these powers constitute their parental authority. Parents have the power to make decisions about medical affairs concerning their children, for example (see Cherry 2010). Parents also have the power to impose duties on their children. When a mother tells her fourteen-year-old son to go pick up a parcel for her at the post office or to be home at 10pm, then he arguably has a duty to pick up the parcel or to be home at 10pm, just because his mother said so. Parental authority is as salient as moral obligations in families are. It is part and parcel of the moral landscape as we know it.
It should be noted that parental authority is not unlimited, of course (sec Gheaus 2017). A father cannot impose a duty on his daughter to assist him in a bank robbery or to make erotic photos. It is a power to impose duties within a certain range. It is no different from political authority in this regard.
Furthermore, parents cannot impose duties on children before they reach a certain age. Of course, a mother can tell her four-year-old daughter not to do something, but it would be odd to say that the daughter therefore has an obligation not to do it. In order to have obligations, one has to be able to understand what an obligation is. Dogs do not have obligations, even though one can give orders to dogs. It does not matter, for our purposes, at what age a kid is usually
54 Authority
able to understand the idea of an obligation. Teens, at least, usually seem to understand what obligations are.
As in the case of moral bonds in families in general, we may ask about the deeper basis for parental authority in particular. Of course the service conception could be cited as a foundation for parental authority: Parents are then said to have authority over their children because and as long as this is a service to the children. But, within the context of the present argument, we can stay agnostic about this. Community-based arguments are trying to convince us that we need no deeper explanation for parental authority. We just know that parents have authority over their children. We do not need to find a deeper justification for parental authority to be assured that parents have parental authority. We could test ethical theories by asking whether they can accommodate familial bonds including patental authority. If they cannot, we are willing to drop the theory, not our conviction that parents have authority over their children.
In a way, a community-based account of authority thus tries to circumvent my explanation condition (see p. 15). Let us accept - for the sake of the argument - that a community-based argument has a great deal of plausibility with regard to parental authority. The question is whether it is also plausible when it comes to political authority. A community-based argument for political authority claims that we can treat political authority as analogous to parental authority, and assume moral bonds among compatriots like we can assume moral bonds among family members. If the analogy holds, we do not need a deeper explanation for political authority.
Political communities as analogous to families
What speaks in favor of treating political communities as analogues to families (and hence to refrain from asking for a deeper explanation of the authority relations within them)? Before starting with some considerations in favor of the analogy, it should be noted that most advocates of community-based arguments have focused on political obligations and
Community and Authority 55
not directly on political authority. Because they take one's association with others in a political community as the key to political obligations, they often speak of "associative obligations." But of course political obligations and political authority are related. It is possible for people to have political obligations without a state that has the authority to impose duties on them, but from the point of view of community-based arguments this seems rather implausible. For that reason, I here take such arguments as arguments for both political obligations and political authority at the same time. I mostly speak of "moral bonds" because this expression can cover all kinds of moral relations, including both obligations and authority.
One thing that speaks in favor of the analogy between family and political communities is that many people seem to be as certain about moral bonds in political communities as they are about moral bonds in families. Just like everyone takes it for granted that parents have an obligation to care for their children and to properly educate them, and that children have an obligation to support their parents when they get old and need their help, most people take it for granted that citizens have an obligation to pay taxes, to obey the law, and to vote. The same holds for authority relations. Most people think that of course the state has the right to enact and coercivcly enforce laws and of course it has the power to thereby impose duties on citizens. There are not many people who question political authority, just like there are not many people who question parental authority. Some political philosophers have argued that the ideas of political obligations and political authority just belong to our concept of a political society (Macdonald 1941; McPherson 1967).
Second, many people indeed feel connected to their compatriors in a way similar to how they feel connected to members of their family (Horton 1992; Tamir 1993). People identify themselves as members of families and as members of nations and/or states. Some may be willing to deny that, but on reflection they often have to confess that they feel pride or shame for their family or their country (or for particular family members or particular compatriots). Being on vacation abroad, one can easily feel shame for the behavior of one's drunken compatriots, just like one can feel ashamed of the
56 Authority
behavior of one's parents or siblings. Watching the Olympics, one can feel pride in the achievements of athletes from one's country, just like one can be proud of the achievements of one's sister. As John Horton remarks, an American citizen who opposed the war in Iraq felt different than a French citizen who opposed the war in Iraq, simply because the United States were involved, but France was not (1992: 171). This connection to family and country can deeply affect one's life. Imagine being a grandniecc of Hermann Goring, for example.
When speaking of "compatriots," I deliberately leave open whether this means other members of one's nation -no matter if this nation has a state or not - or whether it means other members of one's state - no matter if this state is multinational or not. Catalans, for example, do not have a state, but arguably are a nation with members living in Spain and in France. Identification and feelings of connectedness sometimes are complicated and can apply to one's nation or state or both. This is no different in families. I usually speak of "political communities," a term that is open enough to allow both interpretations.
An interesting issue is how much weight wc should give to objective and subjective aspects in identification (Horton 1992: 183-8). When persons are identified as members of a family or a political community by others and self-identify as membets of this family or political community, things are easy. But what happens to moral bonds when people are identified as members of a family or political community by others, but do not self-identify as such members? Does it mean that the moral bonds of family or political community vanish? What, on the other hand, happens when people identify themselves as members of a family or political community, but are not accepted as members by the family ot political community (and are not identified as members by outsiders)? Arguably self-identification cannot be sufficient for the existence of moral bonds including authority relations: One must actually be a member of a political community or family, and one cannot all alone decide about membership. Membership depends on conventions that are - in some way - under the control of the group. Objective membership is thus at least a necessary condition
Community and Authority 57
for the existence of moral bonds. Could it also be sufficient? Arguably it cannot. The subjective aspect, i.e. self-identification, matters, too (see van der Vossen 2011: 490). No one should be stuck in a family or political community; it must be possible to leave one's family and one's political community. Yael Tamir strongly emphasizes this subjective aspect. She writes that associative obligations "must be based on some sense of belonging, on an active and conscious discovery of one's position, and on an affirmation of this position" (1993: 135). But self-identification alone cannot be sufficient for the existence of moral bonds. One also has to be acknowledged as a member. This is especially clear when the moral bonds come in the form of authority relations. Self-identifying as a father of a child is not sufficient for having parental authority, when the rest of the family including the child does not acknowledge my fatherhood. (Of course, things get more complicated again when the person in fact is the biological father.) Self-identifying as the king of Sweden certainly does not give me the political authority the king of Sweden has. Is self-identification necessary for moral bonds? Arguably not. A father can owe moral obligations to his son even though he does not self-identify as a father. It seems, then, that objective identification is necessary but not sufficient for moral bonds, and that subjective identification is neither necessary nor sufficient, but can matter under certain conditions.
If political authority can be tteated as analogous to parental authority, why not treat other forms of authority as analogous to parental authority as well? Why not bosses or religious authorities, teachers or even theoretical authorities? An answer to this question is necessary if the approach is to satisfy the target condition. One has to explain why the approach applies to some forms of authority, but not to others. Community-based arguments are limited to parental and political community because other forms of authority are not relevantly similar. Theoretical authorities have nothing to do with special relationships or community, and the authority of bosses is based on voluntary consent in a way that parental and political authority is not. Yet the authority of religious leaders might be relevantly similar to the authority of parents. Insofar as one is born in a religious group, and insofar as people identify as members of religious
58 Authority
groups, there might indeed be a case for a community-based argument for the authority of religious leaders. I cannot discuss this issue any further here. The same holds for the authority of school teachers. Insofar as children are expected to go to schools, the authority of their teachers might be understood as a community-based form of authority. The authority of other teachers - like an adult's piano teacher -is based on consent, though. In any case, I do not think that community-based arguments for authority have a problem with the target condition. They promise to encompass a wide range of authorities and are quite attractive for that reason.
Summing up, then, there are some good reasons to treat political communities as analogues to families. First, many people are just as certain about moral bonds and authority relations in political communities as they are about moral bonds and authority relations in families. Second, people identify and are identified with their country or nation, just as they identify and are identified with their families.
Objections and refinements
There are several worries or objections about community-based arguments for authority. The first of them applies to community-based arguments for authority in general; the other three apply to community-based arguments for political authority in particular. The first worry is that there obviously are deeply unjust or otherwise immoral communities, be it political communities or families (Simmons 1996a: 85; Dagger 2000: 110-12; see also Vernon 2007). Quite obviously, people do not have moral obligations within such communities and there are no authority relations within such communities. A mother who abuses her son loses her parental authority, and a brutal dictator loses his political authority. This can be taken to show that membership in families and political communities alone does not ground moral bonds including authority relations. Advocates of community-based arguments fot both parental and political authority need an answer to this worry.
Community and Authority 59
Tamir bites the bullet; she argues that there are moral bonds even in morally reprehensible groups like the Mafia; it is just that these moral obligations are easily overridden by other moral concerns (1993: 101-2). But arguably a father cannot impose a duty on his son to rob the next bank with him. There is not a duty that gets overridden, but no overriding going on at all (Mason 1997: 438; Dagger 2000: 111; van der Vossen 2011: 489).
Margaret Gilbert's solution is to "de-moralize" obligations (2013: Ch. 17). If obligations within immoral communities are not moral obligations, then their existence surely becomes more plausible. But, on the other hand, if obligations are not understood as moral obligations, they do not help us with our task of explaining legitimate authority relations and in particular legitimate political authority.
A better answer to the first worry is to simply claim thar cerrain moral criteria have to be met for there to be moral bonds within both families and political communities (Dworkin 1986; Horton 1992; Hardimon 1994; Mason 1997; Seglow 201.3; Lazar 2016). Some suggest that we look at the generic good that a community provides in order to determine the moral criteria a community has to meet to give rise to moral bonds. In the case of political communities, Horton argues that this good is the provision of peace, security, and law and order (1992: 162, 176-7). If that is so, the moral criteria a political community has to meet to give rise to moral bonds are rather modest (1992: 160, 163, 178). Ronald Dworkin argues that a "true community" (not only a true political community, but a true community of any kind) must meet a couple of moral criteria, and one of them is that a group's practices display an equal general concern for the well-being of all members (1986: 200-1). Moral bonds including relations of authority can only be found in such true communities. Dworkin's account has the reputation of setting quite ambitious criteria for a proper community, but it should be noted that the criteria do not require perfect justice. There can be legitimate authority and moral bonds in moderately unjust communities (1986: 203-5). Andrew Mason grounds political obligations in the intrinsic value of equal citizenship and the recognition it provides (1997: 439—45). When a polity does not realize
60 Authority
equal citizenship, then there are no moral bonds and no legitimate authority.
For our purposes, it does not matter what moral criteria one should adopt, in the end. In any case, introducing such moral criteria is not an ad hoc move. All theories of political authority have to invoke external moral criteria at some point, including consent theory (Horton 1992: 161; Hardimon 1994: 344): Consent theory has to specify under which circumstances consent ceases to be valid, and it has to discuss whether there are things that we cannot give valid consent to (slavety may be such a thing).
Some suspect that once moral criteria are introduced, membership alone is no longer sufficient for the existence of obligations and authority. Richard Dagger writes: "Something extra must be added - an appeal to justice or to the nature of true community - to supply what a straightforward appeal to membership lacks" (2000: 110; see Simmons 1996a: 87-90; Wellman 1997: 199). I do not see much of a problem here (see also Horton and Windeknecht 2015: 912-14). Indeed an advocate of a community-based argument for authority should say that membership in a community alone is not sufficient for having obligations and for authority relations. The community must also meet certain moral standards and, at least sometimes, one also has to self-identify as a member. This does not undermine a community-based argument for authority. The core of a community-based argument for authority is not the claim that membership in a community is sufficient for standing in authority relations to other members. The core of a community-based argument for authority is that we do not need any grounding of authority relations. A community-based argument tries to convince us that we can take it as an established, fixed point that certain kinds of communities come with moral bonds including relations of authority. That certain kinds of communities do not come with such moral bonds does not undermine this idea.
A second worry about a community-based argument for political authority is that there are important differences between families and political communities (Simmons 1996a: 91-2; Wellman 1997: 188-9; Dagger 2000: 107-8; Jeske 2001). One of the most obvious differences is that political communities are quite anonymous, while families are not.
Community and Authority 61
We do not know most of our fellow compatriots, but we do know our family, at least that part of our family where we feel that moral bonds exist. Another important difference is that there usually are emotional bonds between family members; in the case of political communities, people may be emotionally attached to their community in the abstract, but they are not emotionally connected with each member. Finally, the emotional bonds within families seem to be natural in a deep sense; at least the emotional bonds between parents and children and between siblings are built into our "biological blueprint," so to speak. Political communities, on the other hand, are cultural artifacts and so our attachment to them cannot be as deeply embedded in our human nature.
How should a defender of a community-based argument for political authority reply to this? First, she could try to explain how and why these differences are matters of degree: We do not know all family members we are said to have moral bonds with, we do not have strong emotional bonds with all family members (for example to an uncle one rarely meets or one simply does not like), and the tendency to get emotionally attached to groups has deep evolutionary roots, too (Haidt 2012). After all, there is something to Aristotle's famous dictum that "man is by nature a political animal" (2013: Bk 1). Second, the defender could deny the relevance of these (gradual) differences. Sure, there are these differences, but why should they demonstrate that political authority cannot be conceived analogously to parental authority, i.e. as a basic feature of the moral landscape without any need for a deeper explanation? Both replies are good, as far as they go, although one may indeed become a little less confident in the adequacy of the analogy. If you think that there are moral bonds within families because there are (in general) emotional ties and a natural affection among members, then you will become suspicious about moral bonds in political communities. But, as I said, a defender of a community-based argument for political authority and mora! bonds in political communities need not believe that there are moral bonds only if and because there are such natural emotional ties. The core conviction is that we need no deeper explanation about moral bonds in families and political communities because we are so certain about them.
62 Authority
Relatedly, it has been argued that a community-based argument for political authority gives a problematically paternalist picture of the state (Dagger 2000: 108). If state authority is taken as similar to parental authority, does this imply that the state should educate us like our parents did? Should the state treat its citizens like children? In reply, one could make the case that - in a sense - of course the state should be for the good of its citizen and in that sense parents may indeed be a proper model (Rousseau 1762: Bk 1 Ch. 2). Moreover, one could point out that this need not imply that the state should treat citizens like children. A proponent of a community-based argument for political authority need not deny that citizens should be conceived as autonomous beings and that this sets limits to what states may legitimately do. To draw an analogy between parental authority and political authority does not commit us to follow it through in every respect.
The third worry is that the community-based argument for political authority confuses felt obligations with real obligations (Simmons 1996a: 75, 83; see Dagger 2000: 108-10); likewise, it confuses most people's de facto acknowledgment of political authority with legitimate authority. Just citing people's conviction that states have political authority is simply not sufficient to show that they actually have political authority. In reply, one can insist that - just like in the case of parental authority - we should take people's intuitions at face value, at least as long as there is no reason for doubt. Ethics and political philosophy have to start somewhere, and where else should we start if not with our firmly held convictions.
But now the objector can respond that indeed there are good reasons to doubt that there are moral bonds and authority relations in political communities. Christopher Wellman describes a cosmopolitan who feels no special connection to his compatriots and cares for people far abroad as much as he does for people at home. Wellman suggests that this person looks not morally depraved, but indeed superior and admirable (1997: 184-5). This is different from the case of someone who does not feel connected to her mother and father or to her children. Maybe, then, the moral bonds in political communities including authority relations turn out to be a mere prejudice that should be overcome?
Community and Authority 63
A second reason to doubt that there are legitimate relations of authority in political communities is that they constitute relations of inequality (see pp. 13-14). Most people adhere to the powerful idea that all humans are equals; and even though the idea of equality can certainly be interpreted in different ways, political authority plainly introduces massive inequality in the moral rights people have. Some are said to have the moral power to impose duties on others and the liberty to enforce them. Some are said to have the right to rule. This is an inequality in the most straightforward sense. It is true that the right to rule basically resides in political institutions, but it is also true that in the end real persons have to fill the institutional roles and therefore have the right to rule. Note that the inequality between parents and children looks much less problematic, since children are not yet fully rational humans. They are unequal in the relevant sense. Thus political authority (and moral bonds in political community in general) looks morally problematic in a way that parental authority (and moral bonds in families) does not.
Moreover, we know about people's willingness to accept illegitimate de facto authorities and should therefore be reluctant to treat people's acceptance of state authority as clear evidence that the state actually has legitimate political authority (see Huemer 2013: Ch. 6). The Milgram experiments (see Milgram 1974) showed a surprising willingness in people to submit to immoral directives from authorities, and the so-called Stockholm syndrome describes how victims sometimes tend to build emotional ties with their captors or abusers.
Thus political authority is in need of a deeper explanation, while parental authority is not, or at least not to the same degree. The strategy of just relying on our conviction that there are moral bonds in political communities including relations of authority is in the end unconvincing; one cannot circumvent the explanation condition when it comes to political authority.
Now friends of a community-based argument for political authority could of course try to change their strategy and try to actually provide a deeper explanation for how moral bonds including relations of authority are brought about in communities. A tempting way is to push the
64 Authority
community-based argument in the direction of consent theory (Gilbert 2006; Horton 2012; Rcnzo 2012; see also Raz 1984: 353-4). When we self-identify with a community, is not our membership consensual, in a way? Couldn't this explain why we are morally connected to the community? Of course, membership is usually not consensual in the sense discussed in Chapter 2, but we might try to attenuate the concept of consent a little bit. Maybe it is not literally consent that is required to establish political authority and, more generally, authority in communities that are not strictly voluntary. Maybe something similar, but weaker, is sufficient, like citizens' acceptance of the state or their membership in a community (Horton 2012; Renzo 2012; see also Murphy 1999). Horton emphasizes that acceptance does not itself "ground" legitimacy, though. He writes:
I consent to, or more properly recognize or acknowledge, the stare as legitimate, because it meets the salient criteria of legitimacy that are practically operative. I do not acknowledge its legitimacy because 1 have consented to it [.,.]. The affirmation of legitimacy matters, but that affirmation is grounded in something other than that affirmation itself. (2012: 142; see Horton and Windeknecht 2015: 909)
Nevertheless, acceptance seems to be necessary for legitimate authority, according to him.
Acceptance certainly looks similar to consent, at first sight. When people accept the state, they have some kind of positive attitude toward the state; when they give consent to the state, they express that attitude. But consent and acceptance are very different in their moral powers (Wendt 2016b: 237—40). Consent can only do its job of creating rights because it is not a mere mental act or mental state, but some kind of public performance, observable and understandable by others. This is especially clear in the context of consent to sex. My mental acts cannot give some other person the liberty-right to have sexual intercourse with me. For that reason, acceptance is very different from consent. Acceptance is a mental state, not a performance or public act. And as such it cannot create new rights. But a theory of authority must explain how the state could come to have rights that the governed lack. Acceptance cannot help with that task; only proper consent can.
Community and Authority 65
Now one may reply that acceptance is not only a mental act, but also behaviorally manifest. One can observe that others accept the state, for example in how they talk about it. But either behaviorally manifest acceptance is said to amount to tacit consent, in which case we are back to consent theory, or behaviorally manifest acceptance does not amount to tacit consent, in which case we still lack an explanation of how this acceptance could give rise to new rights for some persons.
A fourth and final worry about community-based accounts is that they probably fail to meet the success condition. At least when subjective endorsement of membership plays an important role, not every citizen will incur political obligations and be subject to political authority. Some punks, anarchists, and stubborn cosmopolitans will not be members in the relevant sense. This may be not so problematic when we merely aim to explain political obligations. One can concede that the account cannot explain everyone's political obligations, when it succeeds in explaining the political obligations of most or many citizens. But for an account of political authority, the fact that some are not members in the relevant sense is problematic, since states as we know them claim to have authority over all citizens and people who are in their territory (that is what I called the holistic nature of political authority; see p. 7). Political authority is not individualized, such that a state could have authority over some citizens, but not over others. Hence political authority in the full sense does not exist in any actual and imaginable future state, when we subscribe to a community-based argument for political authority (see also van der Vossen 2011: 493-4).
Summary
If we take moral bonds within political communities as analogous to moral bonds within families, we can try to circumvent the explanation condition and simply treat political authority as a basic fact that is not in need of any deeper explanation. The problem is that the analogy between
66 Authority
families and political communities does not hold insofar as there are good reasons to doubt that there are moral bonds within political communities and that states have political authority. For that reason, we need some explanation for political authority and cannot circumvent the explanation condition. Moreover, the community-based account cannot account for the holistic nature of political authority.
r
Natural Duties and Authority
Some have tried to ground political authority in our natural duties. Natural duties are nothing obscure; they are simply moral duties we all have as humans, i.e. independently of our specific roles as mothers or teachers and independently of our promises and contracts. I will discuss three natural duty-based theories. The first works with a natural duty of justice. The second does so as well, but claims to be able to explain not only the authority of the state, but also the specific authority of democracy. The third works with a natural duty to rescue others in emergency situations.
The natural duty of justice
At first blush, one might think that the connection between authority and justice should be very simple. When states do a reasonably good job of securing and promoting justice, then they have political authority for that very reason. Justice is a big word, of course. I would like to stay agnostic on the details of the requirements of justice. Justice will probably require that certain basic liberal rights (the right to physical integrity, freedom of speech, religious freedom, and so on) are protected, that resources and opportunities are fairly distributed in the society, that criminals are caught and
68 Authority
punished, etc. Justice is often conceived as the main virtue of social institutions, and so one might well be inclined to think that states that help to secure or promote justice theteby acquire political authority. But this line of thought would be a bit too simple. There are other institutions beyond the state that also help to secure and promote justice, and we are not willing to grant them political authority. Oxfam or Amnesty International, for example, may advance the cause of justice, but they do not have the authority to impose and coercively enforce laws on us. In fact, they do not have the moral power to impose any duties on us. For that reason, the connection between justice and authority is a bit more complicated. The simple fact that an institution helps with the cause of justice is not sufficient to bring political authority into the world.
A promising way to find a bridge from a state's justice to its authority is to appeal to people's "natural duties of justice." As explained, natural duties are duties one has simply as a human being, i.e. independently of community- or consent-based obligations. An example of a natural duty is our duty not to kill others. Some think that there also is a natural duty of justice "to support and to further just institutions." According to John Rawls, our natural duty of justice has two parts: "[Fjirst, we are to comply with and to do our share in just institutions when they exist and apply to us; and second, we are to assist in the establishment of just arrangements when they do not exist, at least when this can be done with little cost to ourselves" (1971: 334, and see 115). What is of interest for us is the first part, the duty to comply with and do our share in existing just institutions that apply to us. That we have such a duty sounds quite plausible; and if we do, then this is an important firsr step in an argument for political authority: We have a natural duty to comply with and do our share in just institutions including just states. This arguably means, in the context of states, that we have a duty to obey the laws, pay taxes, maybe to serve in the military, etc.
That we have such a duty is not sufficient to show that political institutions have authority. But when we add the assumption that we do not have "a right not to be coerced to do what we have an obligation of justice to do" (Buchanan 2002: 703), then we already have an argument why just states at least have a liberty-right to enact and coercively
r
Natural Duties and Authority 69
enforce laws. Before we take the second step from here to the state's power to impose duties (and hence to political authority), I would like to add a couple of comments and discuss two objections to the argument so far.
Recall the target condition (p. 15): It is nice if an account of political authority can also explain other forms of authority. But if it cannot, it should be able to explain why. The natural duty of justice is only invoked to explain state authority. Why does it not apply to theoretical authorities, to parental authority, and to the authority of bosses, religious leaders, and teachers? One may argue that only states have the job of securing and promoting justice. Parents, for example, have the job of promoting the well-being of their children, first of all, but not of promoting justice. Yet all authorities can influence the justice of a society, and all authorities arguably owe certain duties of justice to their subjects. To meet the target condition, it would therefore be worthwhile to try to develop natural duty-based accounts of parental authority, the authority of bosses, the authority of religious leaders, and even theoretical authorities. Nevertheless, I will here focus on political authority only.
If the natural duty of justice account succeeds, it seems that it will probably have no problem in accounting for the holistic nature of political authority (see p. 7), and this certainly is a great attraction. Because we all have natural duties of justice, no matter if we like it or not, the account promises to explain the state's authority over all of us. There is no way for some people to escape state authority (as long as states are just, of course). The theory thus promises to meet the success condition.
A natural duty of justice-based account of political authority has roots in Immanuel Kant. As explained in Chapter 2, Kant argues that we have a natural duty to leave the state of nature in order to achieve equal freedom under the rule of law (1793: Sec. 2; 1797: Part I §42).
Two objections
There are two objections to the story so far. The first objection is that it is not of much interest whether we have a
70 Authority
duty to comply with and do our share in just states, since no actual states are truly just, and none will ever be. Whether wc have a duty to comply with just states is of purely academic interest, a special issue in highly unrealistic Utopian political philosophy (see Horton 1992: 103-5). Moreover, there is disagreement about what justice requires precisely; how could we ever know that we live in a just state?
This objection is important, but the easy answer is to concede its point and downgrade the natural duty of justice a little bit. Jonathan Quong argues convincingly that the natural duty to comply and do one's share in just institutions does not only apply to perfectly just institutions. Because there is reasonable disagreement about justice, the natural duty of justice already applies when institutions are "reasonably just" (2011: 132-5). There may still be some vagueness in the idea that institutions are to be "reasonably just," but this seems unavoidable. In any case, the natural duty of justice is supposed to apply under realistic, non-utopian conditions. In a similar spirit, Allen Buchanan proposes a natural duty of justice "to help ensure that all persons have access to institutions that protect their basic human rights." That duty applies as soon as states do a "credible job of protecting at least the most basic human rights of all those over whom it wields power" (2002: 703). Doing a credible job of protecting the most basic human rights is certainly not Utopian, and it is a fundamental advancement of justice when this job is done.
The second objection is that the natural duty of justice cannot explain why we are bound to particular states, i.e. why a citizen of Ghana is bound to the Ghanaian state, while a citizen of France is bound to the state of France (presupposing, of course, that both Ghana and France are reasonably just states). Why is a Mexican citizen supposed to pay taxes in Mexico and not elsewhere? Why should a citizen of Vietnam do his military service in Vietnam and not elsewhere? How, in general, can the natural duty of justice explain special ties with one's country? This, in short, is the particularity objection (Simmons 1979: 147-56; 2005: 162-79).
The particularity objection does not apply to consent-based or community-based accounts. When political authority is grounded in consent, it is clear why I am bound to a
Natural Duties and Authority 71
particular state - because I consented to that state's authority. When political authority is grounded in community, it is also clear why I am bound to a particular state - because I am a member in that political community. But a natural duty of justice-based account cannot explain why people are bound to particular states, since there are many reasonably just institutions in the world and I could do my share in all of them. When we start with a universal natural duty of justice, we cannot end up with particularized duties to particular states, just because the very point of natural duties is that they apply independently of special roles or acts of consent that could constitute particularist connections.
If a natural duty of justice-based account cannot answer the particularity objection, it will also have a hard time accounting for political authority (in a second step). If there are no particularized bonds between individuals and states, one can hardly explain why states have legitimate authority over their citizens and everyone in their territory, but not over other people.
Now one may think that there is an easy answer to the particularity objection. According to the natural duty of justice-based account, wc have moral bonds with the institutions that apply to us. That is how Rawls formulates the duty. But what does it mean that a political institution "applies" to someone? John Simmons presents the example of the (fictional) Institute for the Advancement of Philosophers in Montana. Let us assume that the Institute pursues a just cause and that it "applies" to me, since I am a philosopher. Do I have a duty to support the Institute in one way or another? It does not seem so: "People cannot simply force institutions on me, no matter how just, and force on me a moral bond to do my part in and comply with those institutions" (1979: 148). Of course, things change once 1 actively become a member, but then the reason for the motal bonds is not a natural duty of justice, but my consent. Mere "application" cannot be the point, says Simmons.
One reply by defenders of a natural duty of justice-based account is to simply drop the idea of particularized moral bonds between states and citizens and to endorse a globalist, transnational duty to support just institutions (see Rinderle 2005: 261-3). But since a state's political authority certainly
72 Authority
is not global and transnational, but specific to a state's citizens and territory, this reply will not help if we would like to ground political authority (in a second step).
A better answer to the particularity objection is to claim that some institutions - namely states - are necessary for establishing justice and, second, that they require territorial jurisdiction in order to be able to establish justice (Waldron 1993; Quong 2011: 129-30). To establish institutions that protect people's basic liberal rights and try to achieve a fair distribution of resources and opportunities among citizens arguably requires territorial states. The states we live in are the ones we have to comply with because they apply to us qua territorial states. In that they differ from Simmons's Institute for the Advancement of Philosophers in Montana. In the end, then, institutions can indeed "impose" themselves on us (Waldron 1993: 27-30; Christiano 2004: 281-3).
Some may still be skeptical about this answer to the particularity objection. First of all, anarchists could of course question the necessity of the territorial state to establish justice. This is less of a philosophical doubt, of course, but I will briefly deal with it in Chapter 7.
A second problem is the moral arbirrariness of territorial borders (Simmons 2005: 173^1). If the territoriality of states is to particularize our duties of justice, then the actual territorial borders of states should better not be morally arbitrary or - even worse - based on a history of conquest and war. But a defender of a natural duty of justice-based account may insist that the justice of territorial borders is beside the point. Territorial states are needed for establishing justice, and when they are (now) reasonably just, we have to support the states we live in, no matter whether their borders are morally arbitrary. They are now morally salient, and so we have to work with them.
Third, one might worry that the particularity objection has still not been fully answered. That states are necessary for establishing justice and require territorial jurisdiction in order to be able to establish justice may explain particularized bonds between states and everyone in their territory. But it cannot explain why there should be any bonds between states and citizens who live abroad or visit some other country for a limited period of time. Conversely, it cannot explain why
]
Natural Duties and Authority 73
tourists and foreign residents in a state's territory are not bound to that state in the same way that citizens are (Mason 1997: 437).
But maybe proponents of a natural duty-based account need not worry too much about answering that version of the particularity objection. Maybe they can simply endorse the view that there are no natural duty-based bonds between expats and their home states, and that citizens, tourists, and foreign residents all have the same natural duty-based bonds with the territorial state they happen to be in. This would obviously change the scope of the state's liberty-right to enact and enforce laws: States could no longet be said to have that liberty-right with regard to expat citizens, and tourists would be beyond the reach of their home states while abroad. It would also change the scope of the state's power to impose duties (if a natural duty-based account can explain that power in a second step): States could no longer be said to have that power with regard to citizens who are not in their home state's territory. This would mean rethinking the holistic nature of political authority. A state's authority would still be holistic with regard to its territory, but not with regard to its citizens. That may be a bullet one can bite. But maybe one need not even bite that bullet: To account for the special ties of citizenship, proponents of a natural duty-based account might try to combine it with a community-based account (see Mokrosinska 2012: Ch. 7; Schmelzle 2015: 119-23) (see also pp. 91,100).
From natural duties of justice to political authority
Let us assume that we have established that people have a natural duty of justice to comply with a particular (reasonably just) state and that reasonably just states have the liberty-right to enact and enforce laws for a certain territory. Some proponents of a natural duty of justice-based account, like Rawls and Buchanan, stop at this point. But others aim to take a second step and to base the state's authority on a natural duty of justice (Quong 2011: 108, 128). We thus
74 Authority
have to show how reasonably just states could acquire the moral powet to impose duties on citizens.
One option would be to follow Elizabeth Anscombe when she writes: "If something is necessary, if it is, for example, a necessary task in human life, then a right arises in those whose task it is, to have what belongs to the performance of the task" (1978:17; see Sartorius 1981; Copp 1999). Because states are necessary for establishing justice and because they need political authority to do so, they have political authority. But this is unconvincing (see also Simmons 2005: 127-42), since it is simply a refusal to provide an explanation for political authority (which is problematic, as pointed out in the last chapter). Something more needs to be said, and the most promising way is to draw a connection between authority and the natural duty of justice again.
Quong proposes a principle that applies not only to political authority, but also to "local" authorities like that of the flight attendant in Estlund's example from Chapter 2 (pp. 31-2). Quong writes:
One way to establish that a person has legitimate authority over another person involves showing that the alleged subject is likely better to fulfil the duties of justice he is under if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to directly fulfil the duties he is under himself. (2011: 128)
The principle obviously resembles Joseph Raz's normal justification thesis (see p. 45), but it works with natural duties of justice instead of reasons. Quong presents the following example (2011: 127): Two persons, A and B, arrive at the scene of an accident. A has medical expertise, B has not. Quong suggests that, under these circumstances, B has a moral duty to follow A's ditectives, because this is the best way for her to fulfil her moral duty to help the victims, which is a natural moral duty. A thus acquires authority over B because B has a natural duty to help the victims.
In the example, A indeed seems to have acquired a moral power to impose duties. But it is weaker than a typical moral power to impose duties (see also Wendt 2016a: 117-21). First of all, it is a power that is constrained in many ways. It only applies under the present exceptional circumstances
Natural Duties and Authority 75
for a short period of time and for a narrow set of permissible directives that all have to do with helping the victims of the accident. Some may want to call it mere "leadership," not authority (see p. 32). Second, it is merely a power to specify pre-existing duties. A power to specify a pre-existing duty is not exactly the same as a mere side-effect power (see pp. 10-11). It indeed is a power to provide content-independent reasons for action. But it is also not a power to impose duties in its fullest sense, because it is not a power to impose new duties, but only to specify a pre-existing duty.
Can we understand political authority by analogy with Quong's example? Of course one might worry that it cannot establish a proper power to impose duties, merely a power to specify a pre-existing duty. But maybe one should not worry too much about this. A moral power to specify a pre-existing duty of justice might be all we need.
It is a bigger problem that the authority (or leadership) of the doctor and the authority of the state are so different in their comprehensiveness. The accident example is so convincing because there is a clear task (one has to help the victims), clear expertise (the doctor knows how to help, I do not), and clear limits to authority (once the task is done, authority ends). All these factors are different in the case of political authority. There is no clearly specified task, no clear expertise, and no clear limits to authority. Political authority implies a very general power to impose duties on all matters, and it is hard to see how this can be understood as necessary for giving direction to a pre-existing duty of justice.
Another problem is that private companies or NGOs may also be in a great position to specify my natural duty of justice. Why do they not acquire the same authority as the state? It seems that a natural duty of justice-based account cannot convincingly explain the authority of states and thus does not meet the explanation condition.
Democracy and authority
I now come to the second natural duty-based account. It aims at explaining not only state authority, but also the
76 Authority
specific authority of democracy. Most people believe that democracies are morally superior to othet types of state. Some think so for instrumental and pragmatic reasons (recall Winston Churchill saying "democracy is the worst form of government except for all those other forms that have been tried from time to time"); others think that democracies are intrinsically valuable for reasons that have to do with self-rule, freedom, and/or equality. Some argue that there also is a pretty direct link between democracy and political authority. They are inspired by, among others, Jean-Jacques Rousseau. Rousseau provides a social contract argument for state authority and argues that the only legitimate contract is one that establishes a (direct) democtacy (1762: Bk. 1 Chs. 6-7, Bk. 2 Chs. 1-2).
Among contemporary philosophers, Thomas Christiano has provided one of the best-known arguments for why states have political authority and have to be democratic (2004; 2008). Christiano starts with a principle of justice and then goes on to argue that democracies are "uniquely suited" to advancing that principle (2004:269-77; 2008: Chs. 1-3). The principle of justice is called the principle of the public realization of equal advancement of interests. What mattets most, in our context, is that it takes justice to require publicity. Justice not only has to be done, it has to be seen to be done. The background for the need for publicity is our fallibility in thinking about justice and our disagreement about justice. Under these conditions, everyone has an interest in being able to see that she is treated as an equal and thus able to "feel at home." Christiano thinks that democracy is uniquely suited for advancing the principle of the public realization of equal advancement of interests, because it gives everyone an equal say in political decisions and in that sense respects everyone's judgment (see also Waldron 1999: 113-16). Of course, democracy is not the only institutional means for the principle of the equal advancement of interests; basic liberal rights like freedom of speech and freedom of association have the same source.
Now what has all this to do with authority? Christiano claims, first of all, that the state is necessary to establish justice (2004: 281-3). Like Kant, he believes that justice is impossible without states. For that reason, he endorses the principle that
Natural Duties and Authority 77
"if legislative institutions publicly realize justice, then they have legitimate legislative authority over those people within their jurisdiction" (2004: 285). If that principle is convincing, it is plausible that only democratic states can have political authority, because only democratic states publicly realize justice in themselves, i.e. not due to the justice of the outcomes they produce, but simply because they are democratic. It should be noted that the authority of democracy is limited by the substantive justice of the democratic decisions, according to Christiano: Democratic decisions must not publicly violate justice by infringing liberal rights (2004: 287-90; 2008: Ch. 7). But when we put this caveat to the side, then Christiano's basic argument is that (1) only democracies are just, (2) just states have political authority, and (3) hence only democracies have political authority.
Now of course one may well question Christiano's starting point, the principle of the public realization of equal advancement of interests, and one can also raise doubts whether that principle really requires democracy (Wall 2006; Huemer 2013: 68-77; Viehoff 2014: 348-51). But the key step in Christiano's argument for political authority is the claim that states which publicly realize justice have political authority simply because they are necessary for establishing justice. Is not political authority established by mere say-so here? Why exactly do states that publicly realize justice thereby gain political authority? As long as there is no answer to this, the explanation condition is not satisfied.
Yet Christiano indeed provides the missing link. It is, again, a natural duty of justice. He writes: "The state is engaged in an activity that is a morally necessary one in the sense that someone who fails to comply with the state's publicly promulgated rules is merely violating a duty of justice to his fellow citizens" (2004: 283). In general terms, everyone has a natural duty of justice "to treat other human beings as equals and this implies that each person must try to realize the equal advancement of the interests of other human beings" (2008: 249). Accordingly, citizens have a duty to comply with democratic states because this is required by their duty of justice.
Now we can see that Christiano's argument has a similar basis to Rawls's. As such, it faces the same challenges, too.
78 Authority
One is to answer the particularity objection. Christiano writes:
Notice that each has a duty to comply with their own democratic institutions since these institutions are necessary to treating their fellows publicly as equals. The duty to treat people as equals is not fully discharged by trying to support the construction of democracy in other parts of the world. If one only did this and failed to act in accordance with a reasonably well-constituted democratic order, then one would be treating one's fellows publicly as inferiors. And this would be a very weighty violation of equality. (2008: 250)
This may again not fully answer the particularity objection, since it is not sufficient to explain why one has special moral ties with one's home country even when living abroad (sec p. 73), but - again - this may be a bullet one can bite.
The other problem is to find a bridge from the natural duty of justice to the state's power to impose duties. As we saw in our discussion of Quong, this is rather difficult. Political authority looks too all-encompassing to be understood by analogy with the authority (or leadership) of a doctor at the scene of a car accident who specifies people's natural duty to help the victims. As long as there is no better explanation, the explanation condition is not met.
There are philosophers who have proposed alternative arguments for a close connection between equality and democratic authority or between equal freedom and democratic authority (Buchanan 2002; Marmor 2005; Stilz 2009; Kolodny 2014; Viehoff 2014). I cannot discuss their views here. I believe that the difficulty of bridging the gap between people's natural duties and the state's moral power to impose duties will affect all these attempts as well.
As a side-note: It seems to me that most discussions of "democratic legitimacy" (see e.g. Cohen 1989; Habermas 1992; Peter 2008) are not really about political authority, understood as the state's right to tule, but about how political decisions ought to be made. Other debates about democracy are simply about what is good and bad about democtacies and whether democracy has not only instrumental, but also intrinsic value (as those who see democracies as closely related to values like equality and freedom tend to think). I
Natural Duties and Authority 79
think that all these debates are situated at a lower level than debates about political authority. They presuppose that states can have political authority, under certain conditions, and ask what institutional form states should take (and why).
Likewise, to show that only democracies are publicly justifiable (Lefkowitz 2005; Estlund 2008) is not to show that democratic states have political authority. The public justifiability of democracy means, roughly, that everyone has sufficient reason to accept it, or at least that no one can reasonably rejecr it. Public justifiability can be translated into the language of hypothetical consent: If democracy is publicly justifiable, then everyone will agree to endorse democracy (under appropriate circumstances). I think that the public justifiability of democracy could at best show that democracies are superior to other state forms, and that we have good reasons to support democracies, but it cannot establish the political authority of democratic states, for reasons spelled out in Chapter 2.
Samaritanism and authority
A third natural duty-based argument has been devised by Christopher Wellman. Wellman does not invoke natural duties of justice, but what he calls "samaritán duties." Samaritan duties are basically natural duties to rescue others in emergency situations. For example, Wellman presents the case of Antonio picking up a hitchhiker who wants to go to a nearby town called Pleasantville. Once they get there, it turns out that Pleasantville is not pleasant at all, but a "contemporary Hobbcsian state of nature." In that scenario, Antonio quite plausibly has a moral duty to bring the hitchhiker to safety; he may not leave her in Pleasantville (1996: 214). More generally, says Wellman, we have a duty to save others from peril when we can do so without unreasonable costs for ourselves and our help is indeed necessary (1996: 215; 2001: 744). This duty should indeed be rather uncontroversial.
Now what he calls "state legitimacy" - the liberty-right of the state to enact and coercively enforce laws - can be understood as based on people's samaritán duties, says Wellman. How does that work? Well, the state is necessary to establish
80 Authority
Natural Duties and Authority 81
peace, security, and law and order; without the police, military, and courts, we would have civil war and instability, basically like the Hobbesian state of nature. Because this is so, we have a samaritán duty to save all of us from the perils of the Hobbesian state of nature by allowing the state to do its job of establishing peace, security, and law and order. This establishes the state's liberty-right to enact and enforce laws. Like Antonio is not at liberty to leave the hitchhiker in Pleasantville, we are not at liberty to be free from laws and state coercion, because the territorial state saves us from the Hobbesian state of nature (1996: 216-19, 223; 2001: 745-7; 2005: 23). With its focus on peace, Wellman's theory is the natural duty account that is closest to Thomas Hobbes.
So far we have seen a samaritán duty-based defense of the state's liberty-right to enact and enforce (certain kinds of) laws. But, says Wellman, we have not yet established a general duty to obey on the part of the citizens, simply because it is not the case that every citizen's compliance is necessary for the state to be able to deliver its benefits: "The plain truth is that any given citizen's behavior typically has no discernible effect on her state's capacity to perform its functions" (2001: 749). But Wellman goes on to argue that everyone has to do his or her fair share in the societal rescuing enterprise, and that this means that everyone has a general duty to obey the law (2001: 749-51; 2005: 32-3).
Wellman's theory aims to overcome the particularity objection in a distinctive way. Duties of rescue are individuated by the facts of the emergency situation. If I pass by a pond and see a drowning child, it is me and not someone else who has the duty to rescue that child. Likewise, my political community is the pond in which the other members and I are all about to drown, and so I and not foreigners have a duty to rescue the members of my community. More precisely, Wellman argues that it would be unfair to claim discretion in how to discharge one's samaritán duty; one has to comply with the laws of one's own state, since wide conformity with the laws of a state is necessary to solve the problem territorial states are to solve (2005: 37-45). That is why I am particularly bound to my state.
But there are several worries about his argument (besides the obvious anarchist objection that the state is not necessary
to achieve peace; see Chapter 7). First, like other natural duty-based accounts (see pp. 73, 78), Wellman's attempt to overcome the particularity objection does not fully solve the problem. It cannot explain why there should be bonds between a state and expat citizens who do not live in its territory (Renzo 2008). Wellman simply concedes this and bites the bullet (2005: 46-52).
Second, one may doubt that the Hobbesian state of nature is an actual danger lurking in the background of our daily lives (Knowles 2010: 164-5). Most people never think about it at all, and, even on reflection, it does not look like an immediate threat, an emergency, or something we have to be saved from here and now, at least in countries that are not actually facing a civil war.
Third, it can be quite costly to comply with the state. One may have to go into the military, one has to pay taxes, etc. Since Samaritan duties only apply when there are no "unreasonable costs" to bear, they can arguably not ground a very robust duty to obey (Klosko 2005: 93-4; Simmons 2005: 181-2). The obvious reply is that indeed compliance with the state may sometimes be costly, but not unreasonably costly, given the benefits the state provides (Wellman 2001: 746; 2005: 32). But, as a rejoinder, one can well insist that individual compliance is simply not necessary to achieve peace, security, and law and order, and in light of this fact the costs may indeed be "unreasonable." If that is right, the Samaritan argument is undermined.
Fourth, and most importantly, Wellman does not attempt to take the second step from people's natural duties and the state's liberty-right to enact and enforce laws to the state's power to impose duties (i.e. political authority). The theory thus does not fare better in explaining political authority than other natural duty-based theories, and for that reason it does not satisfy the explanation condition.
Summary
I have discussed three theories that try to ground political authority in our natural duties. The first two start with a
82 Authority
natural duty of justice; what is specific about the second is that it tries to establish the specific authority of democracy as well. The third starts not with a natural duty of justice, but with a natural duty of rescue. There are two main problems for all natural duty-based accounts. One is to answer the particularity objection; the other is to explain the step from people's natural duties to the state's power to impose duties on them.
Fair Cooperation and Authority
Most people agree that the state does not have the consent of the greatest part of the citizenty. But maybe there are other ways to voluntarily incur obligations and establish relations of authority. The most prominent candidate is fairness obligations. When people voluntarily accept benefits that are produced cooperatively, then they arguably have an obligation of fairness to do their part in the cooperative scheme that produces these benefits. If the state could be conceived as part of such a cooperative venture, maybe people have a fairness obligation to comply with its laws and do their share, and maybe one can also understand the state's authority as based on fairness considerations. Like consent theory, this account has roots in Plato's Crito.
The principle of fairness
Consider the inhabitants of a rural village who suffer from yearly droughts and understand that they would all be better off with a dam. Building the dam, though, requires contributions from everyone, in terms of both money and time. Now the fairness principle says that each of the farmers has an