Political Obligation and the Need for Justice

Abstract This paper examines the claim that justice is necessary for a moral obligation to obey the law. By reflecting on the meaning of obedience, it identifies one version of the claim that must be right and another that must be wrong. It then focuses on the argument for a moral obligation to obey the law that most obviously includes the claim: John Rawls’s argument from the natural duty of justice. More specifically, it focuses on the degree of justice that is needed for this duty to ground a moral obligation to obey the law.

also sufficient. More specifically, I focus on the degree of justice that is needed for the natural duty of justice to ground a moral obligation to obey the law. I regard this issue of degree as the most pressing for Rawls's argument, more pressing than a well-known issue that John Simmons raises, yet discussion of it, including by Rawls, is lacking. I explain why Simmons's issue is less problematic than is generally supposed-in fact, it is not at all problematic-and offer a response, which might be classified as Humean and perhaps somewhat Rawlsian, 4 to the neglected issue of degree. 5 I suggest that the degree of justice that is needed is however much in the circumstances is required to indicate that justice is more likely to be promoted by maintaining current institutions than replacing them with others or none. In Section 3, I briefly conclude.

Justice and Obedience
Since the political obligation that is under examination here requires obedience, the claim that justice is necessary for it must be right in at least one sense. The claim must be wrong in at least one sense, too. In this section, I identify both senses by reflecting on the concept of obedience.
My reflection starts with the observation that obedience is an act that is consistent with a legal (or, indeed, other) norm that states a requirement. 6 But not all such acts are obedience. One might unintentionally conform to a legal norm. That is, one might act consistently with a legal norm without intending to act consistently with it. To obey a legal norm, however, one must intend-and not just happen-to act consistently with it. Only conformity that is intentional can be obedience. 7 Yet this does not mean that intentional conformity equates to obedience, since obedience is intentional conformity of a certain kind, a kind that entails the wrongness of one version of the claim that justice is necessary for a moral obligation to obey the law.
Before identifying that wrong version of the claim, I identify one that must be right. Its rightness follows from the intentionality of obedience. Consider what conditions must be satisfied for conformity to a legal norm to be intentional. Here is one: knowledge of the norm. No one can obey a legal norm of which they are unaware. In typing these words, I act consistently with many legal norms. But I lack knowledge of all of the legal norms with which my typing is consistent. 4. The pedigree of the response is not my primary concern, and I make no more than passing comments on whether it is Rawlsian. 5. One can agree with my comments on the issue of degree even if one disputes its significance because one disagrees with my comments on the issue that Simmons raises. 6. Norms that confer powers cannot be obeyed. See  Such ignorance precludes intentional conformity and, therefore, obedience. 9 Of course, the law cannot be known unless it is knowable. People can obey a legal norm only if, as Joseph Raz says, "they can find out what it is and act on it." 10 To know what it is, it must have certain properties. These properties are set out by Lon L. Fuller in his account of law's 'internal' morality. This morality, without which, in Fuller's opinion, a legal system cannot exist, comprises eight principles of legality. These principles state that officials should (i) enact general rules; (ii) promulgate enacted rules; (iii) enact rules prior to the occurrence of the situation to which they apply; (iv) enact rules that are clear; (v) not enact rules that contradict one another; (vi) not enact rules that demand the impossible; (vii) not change enacted rules too often; and (viii) behave in accordance with enacted rules. 11 The law cannot be obeyed to the extent that officials violate the first seven principles, says Fuller, and obedience to the law insofar as officials violate the eighth is pointless. 12 All but the sixth of the first seven principles state conditions that must be satisfied for a legal norm to be knowable. Knowledge of a legal norm is possible only if officials enact a general rule that is promulgated, non-retroactive, clear, consistent with other norms, and changed infrequently. Hence, if the claim that justice is necessary for a moral obligation to obey the law means that the law must be just in the sense that officials comply with these principles of legality, the claim must be right, since the law cannot be known, which means that it cannot be obeyed, without such compliance by officials. This sense of justice is no doubt limited. In fact, Fuller himself acknowledges that official compliance with the principles of legality is insufficient for justice. 13 For him, as for Rawls, it is merely necessary. 14 But the claim that justice is required for a moral obligation to obey the law might mean only that justice of a certain kind or a certain degree is required. In making the claim, then, someone might (though I am not sure that any theorist actually does) have in mind no more than the formal sort of justice that is involved in official compliance with the principles of legality. If that is what someone has in mind, the intentionality of obedience entails the rightness of the claim.
If they believe that a moral obligation to obey the law depends on each legal norm being just in a more substantive way, however, their claim cannot be right. It cannot be right because it is incompatible with the meaning of obedience. To be obedience, conformity to a legal norm must not only be intentional but also have a certain motive. More specifically, it must be motivated by the fact that the legal norm requires it. 15 An actor who is so motivated might be said to treat the norm as a 'content-independent' reason for action. 16 Although others use the term differently, I regard as content-independent an actor's reason for doing as the law requires if that reason is not contingent on any sort of appraisal (by the actor or by someone else on whom the actor relies) 17 of what the norm requires, independently of the fact that the norm requires it. 18 For instance, I do not act as a legal norm requires for a content-independent reason if I pick up and return a wallet that I saw fall from someone else's pocket due to my belief in the injustice, independent of the illegality, of taking another person's property without their consent. My appraisal of what the norm requires, independently of the fact that the norm requires it, means that I do not obey the law in this instance. Yet the need for such appraisal is implied by the claim that a moral obligation to obey a legal norm depends on the substantive justice of the norm. From the incompatibility of the claim with the content-independence of obedience the wrongness of the claim follows.
The claim that a moral obligation to obey a legal norm is subject to the substantive justice of the norm is not uncommon. Christopher Heath Wellman, for example, makes it when he states: "The distinction between just and unjust laws is an important one, and there is nothing about affirming a citizen's duty to obey the former that prohibits one from denying a citizen's obligation to comply with the latter." 19 I regard the discriminating approach that Wellman endorses as incompatible with the content-independent quality of obedience. But is my assertion of incompatibility too hasty? Perhaps those who claim that a moral obligation to obey a legal norm depends on the substantive justice of the norm, far from rejecting content-independence, merely identify the limits within which it operates. As Dudley Knowles says, "the thesis of content-independence should be understood to operate only within an accepted domain." 20 However, to work out whether a legal norm is within that domain and should be treated as a content-independent reason for action requires appraisal of what the norm requires, independently of the fact that the norm requires it, which is the very thing that content-independence precludes. Even if what the norm requires, independently of the fact that the norm requires it, is obviously just or unjust and can be appraised summarily, appraisal of it cannot be avoided.
One might have qualms about a conception of obedience from which the possibility of a moral obligation to obey an unjust law follows, since one might doubt that such an obligation can exist. Yet one should not forget that, if it exists, it need not always (though, presumably, it must sometimes) defeat all other considerations. 21 A moral obligation to obey the law does not mean that one ought on every occasion to do what the law requires because the law requires it. Perhaps, in a particular situation, what the law requires is so unjust, independently of the fact that the law requires it, that one ought, all things considered, not to obey the law, notwithstanding a moral obligation to do so. This reminder might dispel one's concerns. Alternatively, one might think that the potential for a content-dependent reason against doing what the law requires to defeat a content-independent reason for doing what the law requires undermines the content-independence of the latter reason. That is, one might doubt the possibility of doing what a legal norm requires for a content-independent reason if one ought not to act for such a reason without an appraisal of what the norm requires, independently of the fact that the norm requires it, just in case there is a content-dependent reason against doing what the norm requires. Yet the paradox to which my conception of obedience seems to lead is not genuine. One's appraisal of what a legal norm requires, independently of the fact that the norm requires it, before acting as the norm requires because the norm requires it, does not mean that one's motive is both content-dependent and content-independent, which would be paradoxical. Rather, one obeys the norm despite one's appraisal of what the norm requires, independently of the fact that the norm requires it, which cannot be taken to mean that one obeys the norm because of one's appraisal of what the norm requires, independently of the fact that the norm requires it, yet this is just how one's obedience to the norm must be taken for the paradox to be more than apparent. Even if my conception of obedience is not paradoxical in this way, perhaps it is paradoxical in another. According to Scott Hershovitz, the 'standard view' of obedience, which is my view, leads to 'the paradox of the just law.' 22 This paradox states that just laws seem to deserve obedience, but no morally decent person would obey a just law. Rather than doing what a just law requires because the law requires it, such a person would do what the law requires because what the law requires is just, independently of the fact that the law requires it. The paradox can be avoided, says Hershovitz, by rejecting the standard view and "identify[ing] obedience with conformity-with simply doing as the law requires whatever one's reasons." 23 But the paradox can also be avoided by distinguishing, again, between a moral obligation to obey the law and what, all things considered, one ought to do on this particular occasion, even if not on others. From the fact that no morally decent person would obey a just law, Hershovitz infers the absence of a moral obligation to obey the law. 24 But this absence does not follow. The existence of a moral obligation to obey the law is consistent with the fact that a morally decent person would not, all things considered, do what a just law requires because the law requires it. The response that just laws deserve, all things considered, need not be obedience.
In any event, Hershovitz concedes that the paradox, since it does not involve contradiction, is not a sufficient reason for rejecting the standard view and must be supplemented by a reason for preferring his obedience-as-mere-conformity view. This reason, he says, is that his view fits better than the standard view with the concept of authority. He claims that a necessary property of every authority is that it has a right to its subjects' obedience. He also claims that authorities are usually indifferent to their subjects' reasons for doing what they require. He then points out that these two claims, on the standard view of obedience, lead to the "bizarre" conclusion that "many exercises of authority carry a demand which the authority itself has not seen fit to make." 25 On his view of obedience, however, what authorities necessarily demand is no more than what they generally demand. He thinks that this is a compelling reason for rejecting the standard view in favour of his alternative. Yet he immediately recognises another way of aligning the concepts of obedience and authority: rather than rejecting the standard view of the former, the latter might be understood to include a right to no more than subjects' conformity. In recognising this possibility, he forsakes his case against the standard view. Only by engaging with the issue of which concept-obedience or authority-should be rethought can he resume it. But he refuses "to get bogged down with that question," 26 even though one might infer from the fact that the

200
Walton meaning of obedience, of which there is a 'standard view,' is less contested than the meaning of authority that the former should not depend on the latter. 27 Another critic of the standard view is George Klosko. 28 He argues that a moral obligation to obey the law need not be content-independent and advocates a "loose" understanding of obedience, according to which one obeys the law by "act[ing] as the law says one should, without reference to whether in doing so one is acceding to the will of the lawmaking authority." 29 Despite Klosko's characterisation of his critique, however, its real target is not the condition of contentindependence, to which he remains committed, as evinced by his remarks on what he calls 'formal' reasons. 30 Klosko objects to the condition of content-independence only when it is linked to another. This further condition states that one cannot obey a legal norm unless the reason for which one does what the norm requires is not only content-independent but also has a certain force. More specifically, the reason must be 'peremptory' in the sense that it blocks consideration of every other reason. 31 To obey, in other words, one must do what the norm requires because and only because-for no reason whatsoever except thatthe norm requires it. Only Robert Paul Wolff, on at least one reading of his much-disputed analysis, seems to appreciate the significance of this additional condition. 32 He realizes that, if a law precludes all action-oriented reflection by those to whom it applies-if the lawmaker's will supplants theirs-no moral obligation to obey that law can exist, since this obligation is among the further reasons whose consideration the law does not permit. A moral obligation to obey the law is a reason to do what the law requires for the reason that the law requires it. 33 If the latter reason is peremptory, the former is impossible. Klosko does not reject the condition of peremptoriness due to its incompatibility with a moral obligation to obey the law. He rejects it for other (arguably less compelling) reasons. In criticising it, however, he criticises a view of obedience to which hardly anyone subscribes. None of the many theorists for whom a moral obligation to obey the law is at least possible can endorse this view, even if they appear to do so. 34 Meanwhile, the more popular view that I favour, when disconnected from this implausible one, is unaffected by Klosko's critique. 35 I end this section by considering briefly two more concerns that one might have about my view that obedience must be content-independent (but not peremptory). First, one might worry about its compatibility with my approval of the claim that the formal sort of justice that is involved in official compliance with the principles of legality is necessary for a moral obligation to obey the law. One might think that this kind of justice depends no less than a more substantive kind on the content of the law. Yet, while the knowability of a legal norm surely depends on the norm's content, 36 it does not depend on appraisal of what the norm requires, independently of the fact that the norm requires it. Hence, it is consistent with my understanding of content-independence, for which, I suppose, the label 'merit-independence' might be more appropriate. 37 Second, one might worry that the content-independent nature of obedience is bound to scupper all of the many arguments for a moral obligation to obey the law that require, not the legal norm to which obedience is contemplated, but the system to which the norm belongs to be just. 38 However, the claim that a moral obligation to obey the law depends on the justice of the system is entirely consistent with my conception of obedience, which precludes only appraisal of what a particular norm requires, independently of the fact that the norm requires it. Obedience, on my understanding, entails neither the wrongness nor, indeed, the rightness of the claim that a legal system must be just. Whether it is wrong or right hinges, instead, on each argument for a moral obligation to obey the law. The need for a legal system to be just might be a feature of some arguments for a moral obligation to obey the law but not of others. I cannot examine every relevant argument here. I can, though, consider the argument that most obviously includes the claim that the justice of a legal system is needed. I do so in the next section, having identified in this one a sense in which the claim that justice is necessary for a moral obligation to obey the law must be right and another in which it must be wrong.

Rawls's Argument
An argument for a moral obligation to obey the law might state that justice is necessary but not all that is necessary. For instance, following John Locke, 35. Valentini provides a useful reminder of the need to distinguish between content-independence and peremptoriness (which she calls 'judgment-independence' 202 Walton one might claim that, in addition to justice, consent is also needed. 39 Then again, an argument might state that justice is the sole necessity. John Rawls makes such an argument. For him, the natural duty of justice grounds a moral obligation to obey the law. Since his argument states that nothing more than justice is needed, among arguments that include the claim that justice is necessary for a moral obligation to obey the law, it most obviously includes the claim. In this section, I set out Rawls's argument and dismiss an alleged objection to it, before identifying and then offering a solution to an issue that is genuinely problematic but barely discussed. Rawls argues for two principles of justice that apply to the institutions that comprise the 'basic structure' of society. 40 The first principle states that "[e]ach person has the same indefeasible claim to a fully adequate scheme of equal basic liberties, which scheme is compatible with the same scheme of liberties for all." 41 The second principle states that [S]ocial and economic inequalities are to satisfy two conditions: first, they are to be attached to offices and positions open to all under conditions of fair equality of opportunity; and second, they are to be to the greatest benefit of the least-advantaged members of society (the difference principle). 42 Rawls maintains that these two principles, of which the first is prior to the second and, in the second, fair equality of opportunity is prior to the difference principle, are in 'reflective equilibrium' 43 with considered judgments about justice and would be agreed to by parties in a hypothetical situation-the 'original position'-that models these judgments by putting the parties behind a 'veil of ignorance' that ensures the fairness of their agreement. 44 In addition, Rawls contends that parties in the original position would choose principles that apply to individuals in a just society. They would, he says, choose a principle of fairness from which all obligations, including the obligation to keep promises, arise. 45 He thinks that they would choose several principles of natural duty, too. 46 He thus separates obligations and duties, between which a distinction is not usually made (and is not relied on by me in this paper). Whereas, for Rawls, an obligation depends on both a voluntary act, such as a promise, and a social practice, such as promising, a natural duty depends on neither. Hence, natural duties, on his understanding, are owed by all individuals, whatever their voluntary acts, to all others, whoever they are. These duties include "the duty of helping another when he is in need or jeopardy, provided that one can do so without excessive risk or loss to oneself; the duty not to harm or injure another; and the duty not to cause unnecessary suffering." 47 They also include the duty of justice: From the standpoint of justice as fairness, a fundamental natural duty is the duty of justice. This duty requires us to support and to comply with just institutions that exist and apply to us. It also constrains us to further just arrangements not yet established, at least when this can be done without too much cost to ourselves. Thus if the basic structure of society is just, or as just as it is reasonable to expect in the circumstances, everyone has a natural duty to do his part in the existing scheme. Each is bound to these institutions independent of his voluntary acts, performative or otherwise. 48 Rawls denies that parties in the original position would choose another principle instead of this one. He claims that they would prefer it to the principle of utility because the principle of utility would fit badly with the two principles for institutions and so "lead to an incoherent conception of right." 49 Moreover, he says, they would base the requirement to comply with just institutions on it rather than the principle of fairness due to the difficulty of defining the voluntary acts that bind them to the political society into which they were born and, even if this difficulty were overcome, the difficulty of identifying those who have, in fact, bound themselves. 50 Rawls is fairly certain that the natural duties, including the natural duty of justice, would be chosen. 51 "The real difficulty," he says, "lies in their more detailed specification and with questions of priority: how are these duties to be balanced when they come into conflict, either with each other or with obligations, and with the good that can be achieved by supererogatory actions?" 52 He concedes immediately that "[t]here are no obvious rules for settling these questions." 53 In general, when a 'duty other things equal' is a 'duty all things considered' remains to be determined. 54 Rawls does, however, suggest that 'negative duties,' such as the duty not to harm others, "have more weight" than 'positive duties,' such as the duty to help those in need. 55 He also discusses "a few special cases in connection with civil disobedience and conscientious refusal under circumstances of : : : a nearly just regime." 56 These cases involve "a conflict of principles," some of which "counsel compliance while others direct us the other way." 57 In such cases, one must decide whether the law is so unjust that it ought  Walton to be disobeyed, notwithstanding the natural duty to support and comply with institutions that are "reasonably just, as estimated by what the current state of things allows." 58 When deciding whether, in a particular case, one ought to act on "the duty to oppose injustice," 59 one must, says Rawls, take into account whether the injustice was caused by the non-application of publicly accepted standards that are "more or less just" or by the application of publicly accepted standards that are "unreasonable, and in many cases clearly unjust." 60 If the former, "an appeal to the society's sense of justice is presumably possible to some extent." 61 Rawls thinks that acts of civil disobedience involve an appeal of this sort. 62 If the latter, "[t]he course of action to be followed depends largely on how reasonable the accepted doctrine is and what means are available to change it." 63 Rawls acknowledges that his conception of justice, though the most reasonable, is not the only reasonable conception. 64 Yet he stresses that, sometimes, "as when a society is regulated by principles favoring narrow class interests, one may have no recourse but to oppose the prevailing conception and the institutions it justifies in such ways as promise some success." 65 In such instances, the duty to support and comply with institutions that are reasonably just is presumably absent altogether, rather than defeated by a conflicting duty.

The Issue of Particularity 66
Since natural duties are owed by all individuals to all others, natural-duty arguments for a moral obligation to obey the law such as Rawls's are generally thought to be incapable of satisfying a condition of which Simmons is the principal advocate. In the following passage, Simmons notes his influence and provides a statement of the condition: [O]thers writing about the duty to obey have been less persuaded by my negative arguments about a moral duty to obey the law than they have been by the various framing assumptions that I have employed in discussing the problem. Foremost among these assumptions (and probably the most original of them) has been my demand that an account of the duty to obey meet the "particularity requirement." This requires, roughly, that such an account be able to explain why the moral duty (or obligation) to obey is owed specially to one particular society (or to its subjects or governors) above all others (namely, to "our own" societies), rather than offering only some moral reason for obedience that would bind one equally or more imperatively to obey or support the laws or political institutions of other societies. The moral duty to obey the law should be understood to be a duty to specially obey our own laws in our societies, thus tying this duty to the idea of allegiance and to the exclusive relationship of citizenship. 67 For Simmons, a moral obligation to obey the law must be a moral obligation to obey the law of one's own state first and foremost. Some theorists go so far as to say that a political obligation that requires such conduct must be owed to one's own state and none other than one's own state. 68 Although, in this passage, Simmons refers to 'societies' and says elsewhere that they need not be states, neither he nor any other theorist who endorses the particularity requirement ever seems to have in mind anything but a moral obligation to obey state-law. 69 Particularity is necessary, says Simmons, due to the widespread belief that citizens have a special bond to their own state. 70 But, even if this belief is generally held, Simmons's reliance on it is puzzling, given the suspicion with which he regards another supposedly common belief about the state, namely, the belief in a moral obligation to obey the law. 71 His suspicion about that belief is due to "[t]he clear instrumental value to political and legal superiors (and, more generally, to those enjoying positions of privilege) of an inculcated popular sense of duty to obey, along with the wide variety of means of inculcation available to leaders and to the privileged in modern states." 72 If these facts lead Simmons to suspect the belief in a moral obligation to obey the law, they should lead him to suspect other beliefs about the state. Yet he does not query the belief about the state on which he relies in asserting the need for particularity. As William Edmundson comments, "it is not obvious why the particularity intuition should be any more sacrosanct than the prereflective intuition that political obligations exist-an intuition that Simmons himself repeatedly warns against taking at face value." 73 Rather than justifying the particularity requirement, therefore, Simmons actually suggests a reason for rejecting it.
I suspect that the condition can be defended only by supposing that a moral obligation to obey the law is political in a certain way. Its justification, I think, depends on a conception of politics as concerned with a special bond that citizens have to their own state. But must the political character of the obligation be so conceived?
What makes a moral obligation to obey the law political is, rather surprisingly, not discussed by any theorist except Dorota Mokrosińska, who rightly claims to "enter unexplored territory" in considering it. 74 She is clear that the need for particularity follows from her view of politics. 75 Yet her view of politics is unconvincing. In setting it out, she shifts too quickly from what she calls a "descriptively correct" understanding of politics as "a domain of asymmetrical power relations" to a normative conception of politics as just co-operation. 76 She explains this shift by saying that moral questions are raised by the power relations. That they are raised does not mean, however, that her account of politics must itself answer them. Incorporation of a vision of justice into the very meaning of politics is not an inevitable consequence of the need for a "normative counterpart" to a descriptive view. 77 Her view of politics is narrower than it needs to be.
Indeed, any conception of politics from which Simmons's condition follows is bound to be overly narrow. I see no reason not to understand the political character of the obligation as broadly as possible and thus to leave open for discussion as many substantive issues as possible. Although the obligation is moral in addition to being political, what makes it political need not depend on what makes it moral. It might be political, therefore, in the sense that it relates to "the process of influencing people and getting things done (in one way or another) in the realm of collective human action," as Raymond Geuss says. 78 On this view, politics is, in Geuss's words, "about power, its acquisition, distribution, and use." 79 Guess adds: "Rather than following Rawls's injunction (If you want to think about politics, think about our intuitions about justice), I am suggesting a different injunction: If you want to think about politics, think first about power." 80 If Rawls thinks about politics in the way that Geuss alleges, then perhaps Rawls can defend his natural-duty argument for a moral obligation to obey the law only by relinquishing his view of politics.
Of course, a moral obligation to obey the law is concerned with a particular response-obedience-to a particular way of 'getting things done'-legal norms that state requirements. It is not concerned (but is nevertheless consistent) with the less obvious forms of power that Michel Foucault describes. 81 It is not concerned with all politics, but only with politics in the wider of Max Weber's two senses, 82 which, in contrast to the narrower sense, is not limited to the state and, says Geuss, "sees politics as having to do with any set of relations of subordination, that is of command on the one side and obedience on the other." 83 Understood as political in this sense, a moral obligation to obey the law need not require citizens to obey primarily or even exclusively the law of their own state. On this view of politics, particularity is no longer a condition that a moral obligation to obey the law must satisfy but is a potential outcome of an argument for such an obligation. Moreover, on this view, the obligation might require obedience to norms that are not laws of institutions that are not states. On the statist conception of politics that generates the condition of particularity, no one can have a moral obligation to obey, for instance, international law or, at least, one can have such an obligation only secondarily. 84 Yet, if politics is understood as Geuss proposes, and regardless of whether international law is really law, 85 such an obligation is possible and not necessarily secondary. If particularity were no longer a condition that an argument for a moral obligation to obey the law must satisfy, natural-duty arguments would no longer be vulnerable in the way that Simmons and others maintain. If particularity were no longer an issue for any argument, it would no longer be an issue for them, including Rawls's argument. One might worry about the consequences of rejecting the particularity requirement, however. In its absence, one need not treat as secondary or even illusory a moral obligation to obey the norms of an institution other than one's own state in the event of a conflict with a moral obligation to obey the law of one's own state. Simmons concedes that such an event is likely to be rare but is nevertheless troubled by the potential for it. 86 Indeed, he fears that "the position of dual (or multiple) citizenship is simply morally untenable." 87 But his concern is perplexing-Edmundson favours the epithet "hyperbolic" 88given his recognition that a moral obligation to obey the law is not necessarily conclusive. 89 The conflict between a moral obligation to obey the law of one's own state and a moral obligation to obey the norms of another institution is no different from other conflicts between moral principles. In each case, one must decide what one ought to do, all things considered. Simmons's disquiet is puzzling, therefore, and should not prompt scepticism about demoting particularity from an essential to a potential feature of a moral obligation to obey the law. 208 Walton

The Issue of Degree
Although proponents of natural-duty arguments need not worry as much as Simmons and his followers think that they must about the issue of particularity-in fact, they need not worry about it at all-those who are sympathetic to Rawls's natural-duty argument must say more about an issue that is seldom discussed, namely, the degree of justice that is needed for the natural duty of justice to ground a moral obligation to obey the law. This issue is a general one: every argument for a moral obligation to obey the law that includes the claim that justice is necessary must say how much justice is necessary and why that amount of justice is necessary. For instance, a theorist for whom both consent and justice are needed must specify the degree of justice that is needed and explain that specification. Is perfect justice needed? If so, why? Is the standard lower? If it is, why is it not higher or even lower? Such questions require answers, yet theorists provide few. Without answers, one cannot determine the success of an argument that includes the claim that justice is necessary for a moral obligation to obey the law. The issue of degree is, therefore, critical. I offer a response to it here, but only in relation to Rawls's argument. Although the issue is a general one, I see no reason to suppose that a solution for one argument must be a solution for all. The degree of justice that is needed for the natural duty of justice to ground a moral obligation to obey the law might be different from-it might be more, or it might be less than-the degree of justice that is needed for something else, such as consent, to do so. For Rawls, the natural duty of justice cannot ground a moral obligation to obey the law unless the basic structure of society is "nearly just, making due allowance for what it is reasonable to expect in the circumstances." 90 Perfect justice is not necessary, given the impossibility of a "political process which guarantees that the laws enacted in accordance with it will be just." 91 But how much justice is required for institutions to be reasonably just? Although a precise answer cannot be given, not least because it depends on consideration of the circumstances, elaboration on Rawls's standard is needed. Rawls provides some, albeit only a little. Recall that he excludes from the category of reasonably just institutions those that are "regulated by principles favoring narrow class interests." 92 He also remarks that the natural duty of justice cannot ground a moral obligation to obey the law if a minority is permanently deprived of the basic liberties to which all are entitled. 93 David Lyons thinks that, in making this remark, Rawls alludes to "the system of racial subordination known as Jim Crow, which 90. Rawls was enforced by terror, coercion, harassment, and brutality, and which had been tolerated at all levels of government for generations." 94 Perhaps further elaboration on Rawls's standard can be found in his later work on 'political liberalism,' notwithstanding the lack of explicit consideration therein of the problem of political obligation. 95 Perhaps institutions are reasonably just if they are 'legitimate' in the sense that they exercise power "in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason." 96 This possibility is endorsed by Paul Weithman, who claims that Rawls does not explicitly consider the problem of political obligation in his work on political liberalism because he intends merely to rephrase his natural-duty argument in terms of legitimacy. 97 Yet one cannot be sure that Rawls intends to connect the natural duty of justice to the principle of legitimacy in this way. Perhaps, for him, the principle of legitimacy has no bearing at all on the problem of political obligation, though that possibility seems remote in view of the moral force that Rawls ascribes to legitimate law, or, if it does bear on the problem, it does so independently of the natural duty of justice. 98 One might also look to Rawls's work on international relations for help in understanding the degree of justice that, in his opinion, is needed for the natural duty of justice to ground a moral obligation to obey the law. 99 One might think that institutions are reasonably just if the society to which they belong, though not liberal, is 'decent,' which is to say that the institutions, though not legitimate, "meet certain specified conditions of political right and justice (including the right of citizens to play a substantial role, say through associations and groups, in making political decisions)." 100 Decent societies, like liberal societies, are 'well-ordered' 101 and, again like liberal societies, do not violate human rights, including rights to life, liberty, property, and "formal equality as expressed by the rules of natural justice." 102 The legal system of a society of this kind imposes "bona fide moral duties and obligations" on all members. 103 One might think that these moral requirements arise from the natural duty of justice. But Rawls does not offer such an explanation for them. Instead, he points to work by Philip Soper in which the natural duty of justice does not feature. 104 Why, though, should the institutions of decent societies be excluded from the category of those that are reasonably just? If one were aware of Rawls's motivation for adopting his standard rather than another, one would be able to interpret his standard accordingly and make sense of his distinction between it and others, such as the standard of decency. One would also be able to ask whether another standard actually serves his justification better than his standard. But he does not say why institutions must be reasonably just. He does not say why the degree of justice that is needed cannot be lower. He does not say why it need not be higher.
How, then, should one figure out the degree of justice that is needed? Here is a suggestion: the degree of justice that is needed is however much in the circumstances is required to indicate that justice is more likely to be promoted by maintaining current institutions than replacing them with others or none. 105 I regard this suggestion as a compelling response to the problem of degree that threatens to scupper Rawls's natural-duty argument. It might be described as a Humean response to the extent that it follows David Hume in requiring assessment of the probable consequences of maintaining current institutions or replacing them. 106 If maintaining them would be more likely than replacing them to promote justice, the natural duty of justice grounds a moral obligation to obey the law. Conversely, if replacing them would be more likely than maintaining them to promote justice, the natural duty of justice not only fails to ground a moral obligation to obey the law but also, as Rawls says, "constrains us to further just arrangements not yet established, at least when this can be done without too much cost to ourselves." 107 The cost that Rawls mentions here is only one of the circumstances that matters when considering whether justice is more likely to be promoted by maintaining current institutions or replacing them. What others think-not least about justice, about which they are certain to disagree-and how they act also matter. Indeed, any fact at all that relates to the probable consequences for justice of maintaining current institutions or replacing them should be taken into account.
Although I describe this response as Humean, one can recommend it without following Hume in regarding every act of disobedience as rebellious. For him, current institutions cannot be maintained "without an exact obedience." 108 Yet the consequences of disobedience for current institutions need not be as damaging as he assumes. Their continued existence might be compatible with civil disobedience, which seeks to make current institutions more just, and conscientious refusal, which does not. Moreover, a supporter of the response need not, as Hume does, believe that maintaining current institutions is almost certain to be a better option than replacing them. While Hume admits the possibility of justified resistance, he supposes that revolutions are so likely to be detrimental that they should nearly always be avoided. 109 Yet one can both embrace the response and think that maintaining current institutions is less likely to be preferable than Hume imagines.
The response can be distinguished not only from Hume's conservatism but also from the utilitarianism that some ascribe to him. 110 Although consequentialist, the response need not be utilitarian. It is concerned with the consequences for justice. Hence, it might be endorsed by Rawls, who rejects utilitarianism. Indeed, one might think that Rawls would readily endorse the response, given the apparent consequentialism of his natural-duty argument for a moral obligation to obey the law. 111 One might even think that the response should be branded Rawlsian. Of course, if Rawls were to endorse it, he could not be sure that his standard, even if interpreted sympathetically, would be the answer in all circumstances to the question that it asks. Perhaps the circumstances are such that current institutions need not be reasonably just to indicate that justice is more likely to be promoted by maintaining them than replacing them. Then again, perhaps they cannot so indicate unless they are more than reasonably just.
Since the response requires one to take into account all of the circumstances that are relevant to whether justice is more likely to be promoted by maintaining current institutions than replacing them, one might worry that the response demands too much. Must one consider so much to decide whether the natural duty of justice grounds a moral obligation to obey the law? One must, but political decisions, whether made by 'professional' or 'occasional' politicians, are always demanding in this way. 112 Some remarks that Rawls makes about Locke on the right of resistance are instructive here (and also strengthen the case for branding the response Rawlsian): [A]ll things considered, there may be a right of resistance to an illegitimate and sufficiently unjust regime when the likelihood is great enough that resistance will be effective and that a legitimate regime will be established in its stead without great loss of innocent life. : : : Here, of course, we have to balance imponderables: How great must the likelihood be? How unjust the regime?-and much else. These questions have no precise answers and depend, as one says, on judgment. Political philosophy cannot formulate a precise procedure of judgment; and this should be expressly and repeatedly stated. What it may provide is a guiding framework for deliberation to be tested by reflection. Such a framework may include some indication of their relative weight when they conflict, as they are bound to do. There is no avoiding, then, having to reach a complex judgment weighing many imponderables, about which reasonable persons are bound to differ. 113 Even if one accepts as inevitable the complexity of the enquiry and the difficulty of the judgment for which the response calls, one might still be reluctant to undertake it due to concerns about Rawls's conception of justice. Yet one need not agree with it (or, for that matter, Hume's conception of justice). 114 One need not agree with Rawls's exclusive focus on the institutions that comprise the basic structure, his interest in only some humans, or his insistence on the specification of an ideal. One might favour an approach to justice that applies to the actions of individuals as well as institutions, 115 is concerned with all humans and even all animals, 116 or compares alternatives without reference to an ideal. 117 Further disagreements with Rawls's view of justice are also possible. As Kent Greenawalt says, "Rawls's natural duty can be detached from most of what he says about substantive principles of justice." 118 Of course, how one conceives of justice is bound to affect one's assessment of the likelihood of it being promoted by maintaining current institutions rather than replacing them. Yet one cannot conclude that the natural duty of justice grounds a moral obligation to obey the law unless one agrees with Rawls that institutions are crucial for justice, even if one thinks, in addition, that things other than institutions are crucial for justice and that things other than justice are crucial for institutions. The response to the problem of degree can be non-Rawlsian in some ways but not in all.
I bring this section to a close with two final points. First, one need not think, despite my assumption thus far to the contrary, that one must decide between maintaining or replacing all current institutions. Rather, one might judge that justice is more likely to be promoted by maintaining some of them and replacing others. Whether such disaggregation is possible depends on the circumstances. 119 Second, one might object that a moral obligation to obey the laws of a wicked regime is a potential outcome of dealing with the problem of degree in the way that the response suggests. If a wicked regime is capable of suppressing opposition to it so effectively that replacing it is not a realistic prospect, one might conclude that maintaining it is more likely than replacing it to promote justice because maintaining it is the only option. Yet this conclusion does not follow. Maintaining current institutions cannot be more likely than replacing them to promote justice if maintaining them is not likely to promote justice at all.

Conclusion
In this paper, I have examined the frequently made but rarely discussed claim that justice is necessary for a moral obligation to obey the law. I have not determined whether, given the existence of injustice, endorsement of the claim leads, as one might fear, to philosophical anarchism. I have, though, given the claim some of the attention on which this determination is contingent. By reflecting on the concept of obedience, I have identified one version of the claim that must be right and another that must be wrong. I have also offered a response to the most pressing issue-not particularity, as some believe, but degree-for Rawls's natural-duty argument for a moral obligation to obey the law. This is the argument that most obviously includes the claim. Others that include it await similar scrutiny.