Abstract
An authoritative directive, when it has legitimacy, is a reason to exclude from consideration some of the reasons to act and not to act in this way. One is obliged to obey, even when one disagrees with the directive. Therefore, authority demands deference regarding how one acts, although one is free to think what one likes about that action. How can deference of this kind be compatible with freedom and rationality? That is the so-called moral problem of authority. For some, authority has legitimacy in a deliberative democracy because deference to its rules better meets the demands of freedom and rationality. As I hope to show here, this line of thought does not give an adequate account of authority. If a directive is legitimate only if it has a certain content, as deliberative democrats are saying, one does not defer to but rather merely complies with such a directive.
1. Introduction
An authoritative directive is a reason to act in a particular way, a reason to tolerate the lives others choose to lead, to pay one’s taxes to help remove unjust barriers to opportunity, and so on. However, an authoritative law, rule, order or command is not simply a reason to act insofar as it out-weighs other, competing reasons (see Section 2). Rather, it is both a reason to act independent of the content of the action, and a reason to exclude from consideration the weight of some of the reasons for and against acting in this way. That is, an authoritative directive is a content-independent, exclusionary reason (Raz 2009 [1979]; Hart 1990 [1982]; Simmons 2016). One is obliged to obey, even when one disagrees with the directive. Therefore, authority demands deference regarding how one acts, although one is free to think what one likes about that action. And what is referred to as the moral problem of authority is the requirement to explain how authority, and the deference it demands, can be compatible with freedom and rationality (Wolff 1998 [1970]; Raz 2006).
In this paper, I examine whether deliberative democrats succeed in their efforts to resolve the moral problem of authority. I focus (see Section 3) on two lines of argument: that authority is legitimate when justified with democratic norms (see Bohman and Richardson 2009; Christiano 2004; Macedo 1995) and that authority is legitimate when it has emerged from a democratic procedure (see Chambers 2017; Cohen 1986; Estlund 1997; Gutmann and Thomspon 1990; Manin 1987; Shapiro 2000; Stilz 2017). As we shall see, the differences between these two lines of argument are not unimportant. Not only that, each category is capacious: after all, the first encompasses those who make a clear distinction between political reason and moral truth (see Rawls 2001 [1997]; Larmore 2003), along with those for whom democratic norms have some basis in moral truth (see Gutmann 1995; Gutmann and Thompson 1990); and the second category includes liberals (see Rawls 1993) as well as those for whom liberalism is insufficient and must be supplemented with or even replaced by a commitment to republican politics (see Habermas 1990, 1994; Pettit 2001). Nonetheless, for the purposes of this paper, it is what they have in common that is most important, namely, the assumption that deliberative democracy can and does resolve the moral problem of authority (see Cohen 2003 [1999]; Shapiro 2000; Christiano 2004; Estlund 2008; Stilz 2016). For that reason, they stand apart from those theorists of democracy who are authority sceptics: including those who accept that one can have an obligation to obey a rule one disagrees with but who insist that such a rule does not operate as a second-order, exclusionary reason (see Alexander 1990; Gans 1986; Moore 1989; Regan 1989; Warren 1996).
The version of deliberative democracy focused on here, therefore, is one that does set out to resolve the moral problem of authority. The argument, in brief, is that deference to authority in a deliberative democracy is justified for the reason that it better meets the demands of freedom and rationality (see Section 4). In this paper, I cannot hope to establish that deliberative democrats need an account of authority, that scepticism about legitimate authority is mistaken (see Fives 2022, 2023). I restrict myself to the much more modest goal of evaluating whether those deliberative democrats who attempt to resolve the moral problem of authority are successful in that endeavour. And what I try to show is that success evades them (see Section 5 and 6). This is because deliberative democracy does not adequately account for authority. If a directive is legitimate only if it has a certain content, as deliberative democrats are saying, one does not defer to but rather merely complies with such a directive, and one does so only if doing so better meets the demands of freedom and rationality. Deliberative democracy, therefore, fails to address the moral problem of authority. Regardless of its merits as a theory of the just polity, it cannot convince as a theory of authority.
2. Authority, scepticism and deliberative democracy
In this paper, my interest is in legitimate political authority. What kind of authority is this? The first thing to note is that the authority in question is practical rather than theoretical (see Raz 2009 [1979], 8; Simmons 2016, 14; Peter 2017; Christiano 2020). Of those who have theoretical authority on a particular subject, it can be said their authoritative utterances count as reasons for belief. This is the case if they are an authority on French wine, say and claim that 2007 was not a good year for Beaujolais. Practical authority, in contrast, is a reason for others to act in some way. For example, those who have practical authority over legislation governing the sale of alcohol claim the right to implement rules governing who is free to sell Beaujolais (or any other wine), at what times of the day, in what type of commercial outlet, and so on. Both types of authority involve deference, but with a difference. Theoretical authority calls for surrender of judgement in terms of what others should believe. In contrast, practical authority leaves one free to think what one likes about the rules brought in, but one is to defer in the sense that one is to act as the rules stipulate, and do so regardless of one’s views (Raz 1986, 39).
To understand the precise way in which authority demands deference it is necessary to consider the distinctive way in which an authoritative directive operates as a reason. For a start, a legitimate authoritative directive is a content-independent reason for action. Hence, if a military command has authority, one should obey for the reason that it is a command, and independent of the nature of the action commanded, including when the action is unjust (Hart 1990 [1982], 101). Similarly, if a law has legitimacy, one has a duty to obey simply because it is law, and again, even if it is unjust (Raz 2001, 9; Simmons 1996, 24). However, a legitimate authoritative directive is not only content-independent. It is also pre-emptive. It is an exclusionary 1 reason. It is both a first-order reason for action (a reason to φ) and ‘a second-order reason’ for not acting on (all or some) reasons to φ and to not φ (Raz 2009 [1979], 17, 18; see also Hart 1990 [1982], 101; Hart 2012 [1961], 81; Simmons 2016, 13). One should comply with a mere first-order reason (a request say, or an offer of advice) insofar as it out-weighs other competing considerations. An authoritative directive, in contrast, is a higher-level reason, a reason to exclude from consideration the weight of some but not necessarily all other reasons.
An authoritative directive is a content-independent, exclusionary reason. However, not all content-independent, exclusionary reasons are authoritative directives. For instance, if one promises to contribute to a charitable cause, and if the promise was freely made, then one should do so. One’s promise makes one obligated, independent of the content of the action and regardless of the weight of reasons for and against spending one’s money in this way (Hart 2012 [1961], 197; Warnock 1971, 114). Nonetheless, when one keeps this promise, one is not deferring to the promisee. 2 In contrast, if one is ordered to obey a rule, and if the person or office issuing the order has legitimate authority to do so, the rule is not only a content-independent, exclusionary reason; one is also to defer. Imagine the rules on taxation favour the least advantaged members of society, on the principle that an injustice is an inequality not to the benefit of the least well off (Rawls 1971, 15). Let us also say that, as a committed libertarian, one disagrees with the rule, maintaining that, as a net contributor to the system, it is an unjust violation of one’s liberty, and it treats one simply as a means (Nozick 1974, 172). One may well have voted against the new rule (or voted against the political party who proposed and enacted the rule). Authority does not presume consensus. Indeed, it presumes disagreement, and is a response to disagreement. Nonetheless, even if one has reasons of one’s own to disobey, one is to exclude (put to one side) those reasons, and act as required simply because that is what the rules demand.
Deference is demanded by practical authority, we have seen. To be more precise, it is demanded by legitimate, de jure, practical authority. Of course, this assumes there is such a thing as de jure authority. And, as mentioned above, some theorists of democracy are sceptics regarding legitimate authority. There are of course those anarchists who reject authority on the grounds it is simply incompatible with autonomy. This is the argument of a priori philosophical anarchism (Wolff 1998 [1970], 6). But there is a second group of authority sceptics who are even more relevant to our purposes here. Like the deliberative democrats discussed in this paper, they reject anarchism. They maintain that a rule (etc.) agreed to in a democracy is a justified reason for action, including for those who disagree with it (see Gans 1986, 388; Moore 1989, 895). However, as authority sceptics, they do not see the rule as an authoritative directive, and so it is not a second-order, exclusionary reason. Instead, it is a rule only for so long as it out-weighs other, competing reasons. It is, therefore, a quasi-rule or ‘indicator rule’. Perhaps at times one should treat a rule as if it had authority, but this is only because what it demands is of great significance, or weight, as might be the case with a tax law that benefits the least advantaged. In addition, it ceases to be a rule if it conflicts with and is defeated by other more weighty reasons, which might well be the case with a tax law that imposed too great a burden on individual liberty (Regan 1989, 1009; Alexander 1990, 6; Gans 1992, 20, 21). Hence, democracy does not require, or justify, deference. In fact, the legitimacy of democratic institutions rests on the informed beliefs of its citizens, beliefs shaped through democratic participation. 3
These non-anarchist authority sceptics accept the idea of a duty to obey a rule one disagrees with, but reject the idea that such a rule is a content-independent, exclusionary reason. That is one approach we see among some deliberative democrats. However, in this paper, I focus on those deliberative democrats who believe in legitimate authority. Their position is not just that one can be duty-bound to obey a rule one disagrees with, but also, the rule does operate as a content-independent, exclusionary reason. It is not just that one imposes upon one’s future self the obligation to obey the rules that one disagrees with (Shapiro 2000, 53). A law passed by a democratic regime also may be judged unjust ‘from an independent standpoint’, and nonetheless, it can be ‘a pre-emptive and content-independent reason’ simply because it has been agreed to by a ‘democratic assembly’ (Christiano 2004, 268). More precisely, deliberative democracy ‘yields moral reasons to obey the law and a moral permission to enforce it’, and it follows, the ‘authority and legitimacy’ of the laws of a deliberative democracy ‘often extends even to unjust laws’ (Estlund 2008, 19, 20). Admittedly, deliberative democrats of this sort do not agree regarding every aspect of an authoritative directive, and there is some debate about whether an authoritative directive is both an exclusionary reason as well as a content-independent reason. 4 However, they are as one in accepting that authority demands deference. That is why, even if one believes the tax laws are unjust because they treat one merely as a means, nevertheless one is expected to defer.
The dispute here is over the manner in which the rules agreed to in a deliberative democracy then operate as reasons. It is not to be confused with the separate issue of whether some decisions should be removed from the orbit of democratic decision-making, whether because of the complexity of issues, or the large numbers of citizens affected, and so on. That is, I do not focus here on the question of the proper scope of democratic decision-making, important though that question is (Goodin 2000, 82; Warren 1996, 46). Deliberative democrats do argue for an extension of democratic decision-making to ever more areas of life. That is true, but also, again, not my primary concern. My interest is with those deliberative democrats who go on to say that, when democratic decisions have been taken, and when rules are implemented based on those decisions, the rules then operate as authoritative directives.
3. Democratic norms and democratic procedures
Of course, it is only when authority is legitimate that it is justified in demanding deference. And in this section, I explore the two different versions of the argument for the justification of authority found in deliberative democracy. I begin with the claim that authority is legitimate when justified with democratic norms. Here what matters is the substance of the reasons offered in democratic deliberation. There are two versions of that argument.
According to the first of these, political legitimacy is a matter of political reason rather than moral truth. Hence, it is argued, a regime is legitimate only if it satisfies the requirement of ‘reciprocity’, which in turn demands the basic structure of a society be justified with ‘public’ reasons: reasons about what is ‘reasonable’. As a reasonable citizen, one does not insist on the whole truth of one’s comprehensive doctrine, whatever that might be, as this is something about which, under conditions guaranteeing basic freedoms, there will be reasonable disagreement (Larmore 2003, 368; Rawls 1993, 55). Rather, one appeals to a conception of justice that is, itself, justified with public reasons. What is vitally important is that each person is ‘publicly treated as an equal’ (Christiano 2004, 282). And one is publicly treated as an equal when basic rules are justified with a conception of justice one can reasonably be expected to endorse as a free and equal citizen (Rawls 2001 [1997], 576, 581). To return to the example of a taxation system designed to benefit the least well off, the libertarian may judge the system unjust, but nonetheless should concede that the taxes are reasonable. Insofar as the libertarian (like everyone else) is ‘publicly treated as an equal’, they are duty-bound to obey the resulting directives, even when they have reason to think the particular directives unjust. More generally, a regime can be reasonable, and therefore legitimate, without being fully just (Rawls 1993, 428, 429). Therefore, although norms of justice play an indispensable role here in the justification of authority, providing as they do the content of the public reasons offered in any such justification, it may be that any one authoritative directive will clash with the demands of justice.
The argument under consideration is that authority is legitimate when justified with public reasons. However, others worry that democratic norms cannot perform the function expected of them here if they are not themselves based on moral truth. For example, if the libertarian is to defer to redistributive tax laws they believe to be unjust, it is not enough that the laws be reasonable. The laws must have some basis in moral virtue. To that end, it must be that (inter alia) moral positions are espoused not so as to further one’s self-interest but simply because they are moral positions (Gutmann and Thompson 1990, 76). This follows because it is not agreement per se that is of value (Bohman and Richardson 2009, 260). The aim should be to minimize unnecessary or unjustified disagreement; to seek an ‘economy of moral disagreement’ by securing grounds on which all, morally, ‘ought’ to agree (Gutmann and Thompson 1990, 81–82). At issue here is whether democracy requires a basis in moral truth, rather than in political morality. And this dispute between these two versions of the argument from democratic norms is significant, for it concerns precisely which democratic norms are necessary for legitimacy. But what is equally important is the consensus also evident here. For there is agreement that what matters ultimately is the substance of democratic norms. This is what gives legitimacy to authority.
So much for the argument from democratic norms. The second line of argument is that authority is legitimate when it has emerged from a democratic procedure. In one sense, the distinction between these two lines of argument is artificial. Often they are combined together, as we see in Rawls' work, for example. Nonetheless, for the purposes of conceptual precision, it is helpful to deal with each one in turn. Indeed, the second line of argument does introduce something new. For, it states that it is not the content of the reasons offered that matters. Rather, the argument goes, the outcome of a democratic debate has legitimate authority because and insofar as the procedure itself is just. For example, it is just if it gives everyone an equal say (Shapiro 2000, 77). Some (though not all) proceduralists maintain that democratic procedures play this role because they also bring us closer to the correct conclusion, that is, what is morally right or true. In this vein, it is variously argued that deliberative procedures have ‘epistemic value’ (Estlund 1997, 260); they are based on ‘procedural epistemic norms’ (Peter 2021, 403); and that equal participation and epistemic value are ‘tightly linked’ (Chambers 2017, 266). Despite this appeal to epistemology, it must be remembered, this is an account of practical rather than theoretical authority. What is at issue is whether the duty to obey is justified. Indeed, notwithstanding the epistemic value of a democratic procedure, it is still possible to have good reasons to disagree with any resulting authoritative directive, including when it is unjust. Nonetheless, the argument goes, one should obey the directive arrived it in this way because the fairness of the procedure gives it its legitimacy (Cohen 1986, 37).
Some versions of the procedural approach are part of a liberal political vision. It is one that takes certain matters off the agenda of political debate so as to better protect them, as is the case with ‘basic’ or ‘fundamental’ rights and duties. And one objection to the liberal approach is that it does not give enough scope to deliberation itself in agreeing the basic rules for deliberative democracy. That explains why some move deliberative democracy closer to the republican tradition, where what is called for is an actual discourse ‘to settle conflicts of action by consensual means’ (Habermas 1990, 67; see also Pettit 2001, 89). For instance, Jürgen Habermas' ‘ideal speech situation’ requires, inter alia, that no speaker will be prevented by either internal or external coercion from exercising their rights to question or introduce any assertion, or to express their attitudes, desires and needs (1990, 89). The procedure has value in large part because it ‘allows the better argument to come into play’; the outcome is then ‘collectively binding’, even when one disagrees (Habermas 1990, 67, 1994). These ideas then feed into an account of legitimation of law through what Habermas calls communicative power (1996, 148). This moves us some way away from Rawls' liberalism. And, of course, the debate between liberals and their republican critics is important, but, again, it should not obscure from view the significance of what is also agreed upon here: namely, it is the procedure employed in deliberation that gives legitimacy to the directives then agreed upon in this way.
I have been discussing two different approaches among those deliberative democrats who believe in legitimate authority. There are arguments about democratic norms and also arguments about democratic procedures. A democratic system, the argument goes, is one where either the procedure by which decisions are reached, or the norms upon which decisions are based, is just, and in being just confers legitimacy on what is agreed. The crucial point is that, according to this group of deliberative democrats, one should defer to the directives of a democratic regime. One should obey the directives, even when one has good reasons to disagree with them, including when they are unjust. The directives of a deliberative democracy are authoritative. They are reasons to act independent of their content; and reasons to exclude from consideration some of the reasons not to act in this way.
4. Resolving the moral problem of authority
Of course, any deliberative democrat who maintains that authority can have legitimacy in this way must then address ‘the moral problem with authority’ (Raz 2006, 1014). Deference to authority seems to be incompatible with the demands of freedom and rationality; and that is why sceptics (both anarchist and non-anarchist) reject authority. For the deliberative democrats under consideration here, however, deference is only problematic in this way when authority lacks legitimacy. In those situations, one perhaps should comply with a directive, but never defer. This is the case when, for example, a tyrannical regime passes a law and there happens to be independent reasons to comply with it: for example, one should comply with a law against theft because theft is wrong, not because theft is illegal in this regime (Klosko 2011, 506). Similarly, in non-tyrannical regimes, one might be faced with laws that command what is just, but do not have legitimacy. This is the case, for example, when one’s regime introduces laws ‘to regulate the conditions of employment’, but the laws lack democratic justification because they were not arrived at in a way that extends ‘respect’ to people’s judgements and is ‘responsive’ to those who disagree with a proposal (Christiano 2004, 280). Again, the argument goes, one should comply with such a law, but not defer to it.
Why is it argued that one should never defer to the directives of an undemocratic regime? It is that its directives lack the moral qualities needed for legitimacy. Specifically, while one is duty-bound to obey the directives of a just regime, one is never required to obey those of an unjust regime. This is to ensure that legitimacy can never be conferred on what are called ‘deeply unjust’ regimes (Christiano 2004, 278, 279). And while one is obliged to obey any single unjust law passed by a legitimate regime, nonetheless, ‘laws cannot be too unjust if they are to be legitimate’ (Rawls 1993, 429). It seems that, for deliberative democracy, legitimacy is tightly bound up with justice. In consequence, we can ask, is there a danger of conflating the two? As some deliberative democrats are themselves aware, to do so would be unacceptable (see Peter 2017). Indeed, an undemocratic regime can implement just directives, as Thomas Christiano has shown with the example, above, of an illegitimate law justly regulating employment conditions. Nonetheless, the point Christiano makes regarding that very example is that the law lacks legitimacy because it arose from an unjust decision. He is assuming that legitimacy is not possible without what he calls respect and responsiveness: but that means legitimacy is not possible without justice. However, as we said, conflating justice and legitimacy is a problem. It keeps from view the fact that one can be obliged to obey a regime even though it is not morally justified, say because one has consented to its authority; whereas another regime may be just, and so deserve support in various ways, even though it does not have authority (Simmons 1979, 78, 184, 1999, 750; see Hart 1955, 184).
Whether or not deliberative democrats fail to properly distinguish between justice and legitimacy is a question I leave to one side. My concern here is a different one. I want to examine whether deliberative democrats can, by proceeding in this way, resolve the moral problem with authority. They maintain that those who defer to merely de-facto authority renege on their responsibilities as autonomous agents, whereas deference to authority in a deliberative democracy simply better meets the demands of freedom and rationality. Again, there are two versions of that argument. For proceduralists, a democratic procedure both respects each person’s autonomy to control their own life and also shows respect for the rationality of each citizen. As a result, democratic directives are ‘empowering’ and also it is ‘unreasonable’ to disobey them (Shapiro 2000, 77, 85). A similar argument is made by those who instead focus on democratic norms. Here, what matters is that a democracy ‘treats each publicly as an equal and respects each citizen’s judgment’ (Christiano 2004, 284, 285, 287).
Why is it thought that deliberative democracy has this quality? It is because a democratic decision does not simply record individual preferences. Rather, it shapes them in a way ‘consistent with autonomy’, because preferences shaped by public deliberation are ‘the result of the power of reason’ (Cohen 2003 [1999], 347, 350). One way this happens is by excluding those whose proposals are based on ideas that ‘disqualify’ the point of view, as is the case with racist commitments, for example (Estlund 2008, 17, 26). That is why, it is argued, deference in a deliberative democracy actually better meets the demands of freedom and rationality. What we see here is a very specific conception of freedom and of rationality. Hence, institutions have political authority insofar as they are just, which means they respect the autonomy of each, protecting the freedom of each to shape deliberation in ways that are compatible with the demands of justice. In addition, it is assumed that some of the reasons one happens to have are bad reasons, they should carry little or no weight in public deliberations, including reasons that reflect nothing more than ideology and bias (see Christiano 2004, 278; Habermas 1990, 90). Such reasons do not give legitimacy to authoritative directives. What matter, instead, are the reasons arrived at after the transformative, and liberating, process of democratic deliberation (Manin 1987, 364; Peter 2021, 403).
Deliberative democrats of this kind are hoping to provide a justification for authority. But it must be remembered that any directive implemented by a regime justified in this way must nonetheless operate in the distinctive fashion of an authoritative directive. Deliberative democrats are saying that an authoritative directive can claim legitimacy only if there are good reasons for its legitimacy, but an authoritative directive also must operate in a specific way, one that demands deference when one disagrees with it. As noted already, what we require here is an account of practical rather than theoretical authority. Practical authority demands deference regarding how one acts, leaving one free to disagree with the action. Therefore, in advancing this argument, deliberative democrats must ensure that their picture of practical authority is accurate: at the very least, it must involve deference. Is that the case?
5. Democratic norms and deference
To critically examine the deliberative democratic account of authority, I begin with the view that authority is legitimate when justified with democratic norms. One version of that argument, presented earlier, is that legitimate authority requires what Rawls calls reasonableness, which presupposes not only a reasonable conception of justice, but also commitment to treating others reasonably, that is, treating them as free and equal citizens. When those conditions are met, both ‘the constitutional structure itself’ as well as the ‘particular statutes and laws enacted in accordance with that structure’ will express respect for persons as free and equal citizens (Rawls 2001 [1997], 578–9). Such a polity has legitimate authority, and one is to defer to its directives: a particular statute or law ‘may not be thought the most reasonable, or the most appropriate, by each’, Rawls says, but if it is both reasonable and has majority support then ‘it is politically (morally) binding on him or her as a citizen and is accepted as such’ (ibid., p. 578). That is why the libertarian opponent of redistribution, mentioned earlier, is to act as obliged, independent of the content of the action and regardless of their reasons for not doing so.
Nonetheless, deference has its limits. Rawls will not countenance deference to unreasonable directives, the explanation being they do not have legitimacy. One example of an unreasonable directive, for Rawls, is a paternalistic infringement of liberty, where one’s freedom is restricted without one’s consent so as to promote one’s own good (Fives 2017, 2018). This is unacceptable if, as Rawls insists, liberty has lexical priority over other principles. Liberty can be restricted, he says, only ‘only for the sake of liberty itself, that is, only to insure that the same liberty or a different basic liberty is properly protected’ (Rawls 1971, 204). That is why redistributive taxation is reasonable insofar as it does not infringe basic civil and political freedoms, whereas paternalism is unreasonable precisely because it does violate those freedoms. Paternalism is only ever justified for those unable to make decisions for their own good, those whose freedom we are not required to protect (Rawls 1971, 249–50). If one is able to make decisions for one’s own good, paternalism is not justified, and one has no obligation to defer to paternalistic directives. And paternalism is not obligatory because, for Rawls, it is not reasonable (see Fives 2019).
Only reasonable reasons give legitimacy to authoritative directives, Rawls is saying. Others have criticized this stance, on the ground it places an arbitrary limit on what can count as a reason in democratic debate (Yack 2006, 417). In doing so, deliberative democracy fails by its own standards, namely, the requirement to put moral truth to one side, for these constraints will be acceptable only to adherents of some moral doctrines, for example, individualists or liberals of various kind (Dryzek and Niemeyer 2006, 636; Talisse 2006, 109). However, it is not the fact Rawls has a liberal or individualist conception of rationality that is my concern here (see Fives 2013). As we saw above (see Section 3), there are other versions of the argument from democratic norms, and one of them permits us to employ ideas of moral truth. The latter is, in one sense, a less restrictive version of the argument, but the underlying aim is the same. In each case, they hope to show is that authority has legitimacy insofar as it better meets the demands of freedom and rationality. It is this that is a cause for concern, regardless of the distinctive way in which any one theorist conceptualizes those terms. And the concern is that this line of argument does not account for the mode of operation of an authoritative directive.
The objective, if we recall, is to give an account of the deference involved in practical authority, where one is to act as obliged, but where one is free to think what one likes about the action. However, deliberative democrats inadvertently do away with deference, and hence also authority itself. The implication of deliberative democracy is that, when one is faced with a directive issued by the political community, one must examine its content before deciding whether or not to accede to it. As we saw, on Rawls' version of the argument, one knows a paternalistic rule is not obligatory for an autonomous person such as oneself because of the actions it entails and because of what one should think of the reasons for such a directive. That is, paternalism fails to show the respect owed to autonomous citizens; and free and equal citizens cannot be expected to accept paternalism as a reasonable policy proposal.
It is also helpful to remind ourselves of the difference between deliberative democracy and the non-anarchist critique of authority surveyed earlier. For the latter, one can be obliged to obey rules that one disagrees with (so-called ‘indicator rules’), but the obligation is justified by the weight of reasons. The argument from democratic norms does not go that far. Rawls, for instance, is not saying that one is to weigh the reasons for obedience against conflicting reasons, and decide on the basis of their respective weights. Nonetheless, what he is saying is that there is a certain category of reasons (and an associated category of actions) that are incompatible with authority, namely, those that are unreasonable. The implication is that, when confronted with a putative directive, one first examines its content, to determine whether it falls into the category of directives that can have legitimacy, and only obey those in that category. What Rawls and others are saying is that only reasonable directives can have legitimacy, and so, one must first examine the content of a directive (to determine what category it belongs to) before deciding whether or not to accede. That just means, on this argument one does not defer to legitimate directives.
Perhaps there is a satisfactory response to this critique. It could be argued, deliberative democrats are simply offering reasons to accept a regime as legitimate. And they are explicit in stating both that one is obliged to obey the directives of a legitimate regime, even those one thinks unjust, and that one is free to disagree, and one can continue to use one’s reason to come to such a conclusion. All that is at issue here is the legitimacy of the directives of a regime. Even when directives have legitimacy, one is free to form one’s own view on the morality of those directives. It could be argued, it is important to make that distinction, and even that one of the strengths of deliberative democracy is its insistence on such a distinction, and the space that creates for legitimate moral disagreement (Rawls 1988, 267). Nonetheless, it is also clear that deliberative democrats are not simply concerned with the question of the legal validity of directives (whether rules, laws, orders or commands), if this means or entails they have no interest in the moral duty to obey directives. While, for instance, law and morality are different, one can still have a moral duty to obey a specific law, and a moral duty to obey the law.
The focus of this paper is the moral duty to obey an authoritative directive simply because it is a directive, and regardless of the reasons not to do so, including moral reasons. And when deliberative democrats say a directive has legitimate authority in a democracy, they are saying one has a moral duty to obey. As Rawls puts it, legitimate laws are ‘politically (morally) binding’ on those who disagree. The focus of this paper is a narrow one admittedly, that is, whether deliberative democrats are able to make sense of the moral duty to obey authoritative directives. Of course, there are intriguing questions to ask about the differences between moral obligation simpliciter and, for instance, the moral obligation to obey a law, but I am not able to do justice to those questions here. My point is that although deliberative democrats promise to provide an account of the moral duty to obey legitimate directives, they fail to do so. As we have seen in this section, some deliberative democrats are arguing one should obey a directive only if it is based on certain democratic norms: whether respect, integrity, reasonableness and so on. And the problem with this approach is that, in order to know whether a directive is, for example, unreasonable, one must first examine its content, that is, the quality of the actions it demands and the reasons for its justification. The result is that one does not defer to such a directive.
6. Proceduralism and deference
Does the argument from democratic procedure fare any better in accounting for the mode of operation of an authoritative directive? The basic idea here is that a democratic procedure is more likely to lead to the correct decision; and this is because it is both a free and rational procedure. Of course, this is not an argument justifying theoretical authority. At issue here is not whether an authoritative utterance counts as a reason to believe something, but whether, instead, it is a reason for action. And the argument is that, given the nature of the procedure, it is rational for those who disagree with the outcome nonetheless to defer. The attraction of a procedural approach is that it seems able to uncouple the legitimacy of an authoritative directive from its substantive content. That is, a directive can have legitimacy because it has been agreed to through a procedure that confers authority, and that this is, on one reading, ‘owing to the epistemic value, however imperfect, of the democratic procedure’ (Estlund 1997, 267). That might suggest the procedural approach, when compared with the argument from democratic norms, is better placed to account for the deference at the heart of authority: one is to obey the directives that result from a democratic procedure and do so regardless of one’s views on the quality of the actions they call for or the reasons for their justification. Nonetheless, as I hope to show, the procedural approach in fact has the same shortcoming as the argument from democratic norms.
I am considering the contention that it is the epistemic value of democratic procedures that gives legitimacy to the rules created by following those procedures. To put the same point another way, certain responses to disagreement could never have legitimacy, and this is because they lack the requisite epistemic value. As we have seen, the deliberative democratic procedure does not just record preferences. It shapes them (Cohen 2003 [1999], 347, 350). They are shaped by a just process. For some, the aim is to fight against ‘epistemic injustice’, where, because of one’s prejudices, others suffer an ‘unjust deficit’ of either ‘credibility’ or ‘intelligibility’, and so do not get a ‘fair hearing’ in deliberative settings (Fricker 2013, 1324). Thus, Fabienne Peter’s ‘Epistemic Justice Norm of Political Deliberation’ states that one’s contribution, p, is ‘not valid’ if one’s ‘confidence’ in p is the ‘result of discounting p-related testimony or disagreements because of identity prejudice’, including prejudice related to race, gender, or class (2021, 404; see also Pettit 2001, 288). It follows, however, such a procedure rules out certain decisions in advance. In particular, as I will now try to show, it could never result in an obligation to tolerate those one is opposed to.
What John Horton terms the ‘traditional’ conception of toleration requires that one not intervene in a practice (etc.) that one opposes for good reasons (Horton 2011, 290). The point to emphasize is that this conception of toleration recognizes the good reasons one has for one’s opposition to the practice in question. Opposition may arise for many reasons, including disgust, dislike, or disapproval (moral or otherwise). The traditional conception of toleration requires that one not interfere with those one is opposed to, but it leaves one free to retain one’s opposition, for which one has good reasons, as we have seen.
What do procedural deliberative democrats have to say about toleration? It is highly significant that the only kind of toleration compatible with deliberative democracy is one that requires people overcome their opposition, for example through a reflexive or deliberative process (Bohman 2003; see also Galeotti 2015). This corresponds with the more general idea that legitimacy arises not from tracking the reasons people do have, but rather from the deliberative process through which people’s beliefs are shaped and transformed (Manin 1987, 364).
There will be variation here, between those demanding extensive revisions to citizens' beliefs and those who instead take a more ‘restrained’ approach, where citizens are required to revise only those beliefs that, for example, ‘directly impact others’ enjoyment of the opportunities of equal citizenship’ (Stilz 2017, 350). But regardless of whether it is restrained or extensive in the revisions demanded, the deliberative democratic procedure requires that one ‘appropriately adjust’ one’s ‘original confidence’ in the proposition when that confidence arises from one’s opposition to others. As a result, the traditional conception of toleration is, for the deliberative democrat, unjustifiable. It could never have authority because of the quality of the actions it requires and because of what should be thought about the reasons for this directive. That is, one’s un-adjusted opposition to others is invalid as a reason, and if one retains one’s opposition then one’s toleration is unacceptable because as an action it has a begrudging and judgemental quality (Ceva 2012).
Perhaps, it could be argued in response, the procedural approach is more accommodating and flexible than this critique would suggest. Indeed, many deliberative democrats explicitly state they are engaged in ‘ideal theory’ (Rawls 1971, 9), and that their preferred procedure is an ideal, for example, the ideal speech situation (Habermas 1990, 89; see Benhabib 1994, 27). It is a separate matter to begin mapping out, through non-ideal theory, how to approximate that ideal. Perhaps the traditional conception of toleration could be one of the non-ideal steps taken in pursuit of the ideal of deliberative democracy? That might be, but it is also clear that, for a deliberative democrat, the directives demanding toleration of this sort could never carry legitimate authority. That is, deliberative democrats might conclude that one should comply with the directives demanding toleration when this is the best of a series of bad options. But they would say the same thing about one’s reasons not to steal property in a tyrannical regime: i.e., one should comply with the directive, but it does not have legitimate authority.
But if only a certain range of directives can have legitimacy, for example, those justified with appropriately ‘adjusted’ beliefs, it follows there is no deference even to legitimate directives. How is one to determine whether a directive has legitimacy? The implication of the procedural version of deliberative democracy is that one does so by examining its content, and by ensuring that, for example, it is not judgemental or justified with un-adjusted reasons. As a result, in the procedural approach, one does not defer to directives: they do not have authority. Instead, before acting, one judges each directive based on the quality of the actions it commands and the reasons for its justification. Therefore, as was the case with the argument from democratic norms, the procedural version of deliberative democracy also does not account for the mode of operation of an authoritative directive. There is no deference here.
7. Conclusions
The moral problem of authority arises because authority demands deference. One is to obey an authoritative directive if it has legitimacy, even when one has good reasons not to act as directed, including moral reasons. And as we have seen, to understand the precise way in which authority demands deference it is necessary to consider the distinctive way in which an authoritative directive operates as a reason: it is a content-independent, exclusionary reason. The question then is, how can deference of this kind ever be justified? One argument, the argument focused on here, is that deliberative democracy, and only deliberative democracy, can justify authority. There are two versions of that argument, once focusing on democratic norms, the other on democratic procedures. What I have tried to show is that each argument ultimately fails, and they fail because they are based on the assumption that authority has legitimacy only if it better meets the demands of freedom and rationality. In each case, there is no actual deference to legitimate directives. Before deciding whether to accede to the directives of a deliberative democratic regime, one examines their content, for only directives with a certain content can have legitimacy.
It is important to stress, it is not the commitment to democracy that has created the problem. Rather, it is the idea that authority has legitimacy insofar as it better meets the demands of freedom and rationality. Not only that, arguably a democrat can reject this idea without falling into inconsistency. For they can evaluate the legitimacy of a polity, as well as the directives issued by that polity, by appealing to a variety of reasons; and the commitment to democratic norms and democratic procedures can figure in any such evaluation, but so too will other commitments and values. However, if they are to avoid the pitfalls of deliberative democracy, they must also accept that, if directives have legitimate authority, they operate as authoritative directives. One does not examine their content before deciding whether or not one should comply with them. Instead, one is required to obey because the directives have legitimacy, although of course one may have good reasons not to do so. That is one way in which to reconcile democracy with authority, but it means giving up on deliberative democracy.
