Self-Determination and the Right to Do Wrong

I. THE BOUNDARIES OF SELF-DETERMINATION

It is relatively uncontroversial to suppose that autonomous individuals occupy privileged positions of moral dominion over their self-regarding affairs. It is far from clear, however, when a matter qualifies as self-regarding. The principle of affected interests suggests that an action is self-regarding just in case it does not impact the interests of others. If your X-ing would harm me, for instance, then your X-ing does not appear to be self-regarding. As Robert Nozick famously illustrated, though, this account generates awkward conclusions.¹ Imagine, for instance, that I am such a huge fan of Travis Kelce and Taylor Swift that my happiness depends upon their marrying me. I would be elated if they included me in a throuple and inconsolable if their marriage excluded me. Despite the extent to which my welfare hangs in the balance, Kelce and Swift seem to be solidly within their rights to exclude me. Even if they would ultimately be happier in a throuple with me than they would as an exclusive couple, Kelce’s and Swift’s rights to self-determination entitle them to decline my marriage proposal.

It thus appears that we should follow Nozick’s lead and focus on rights rather than mere interests. On this construal, an action is self-regarding just in case it does not trespass on another party’s moral rights. The reason that Kelce’s and Swift’s rights to self-determination entitle them to exclude me from their marriage despite the extent to which my interests are affected is because I have no right to associate with them in this manner. If I had not merely an interest but a moral right to join their marital union, Kelce and Swift would not be entitled to exclude me. In short, we must reject the principle of affected interests in order to avoid an implausibly deflated notion of self-determination.

While the principle of affected interests leaves insufficient room for self-determination, Nozick’s alternative threatens to saddle us with an implausibly expansive account of our self-regarding affairs. Imagine, for instance, that Kelce and Swift use material they acquire without violating the Lockean proviso to invent a cure for a disease that will otherwise painfully kill me. On Nozick’s view, Kelce and Swift have no moral duty to share this medicine with me. I obviously have an urgent interest in obtaining the medicine, but as the marriage example reveals, a pressing interest does not necessarily amount to a moral right. Just as focusing on rights allows us to see that Kelce and Swift may exclude others from their marriage, this approach implies that Kelce and Swift are at liberty to deny others access to the medicine. While unflinching libertarians may be comfortable with this solution, most regard it as unpalatable. Even if Kelce and Swift need not open their marriage, many believe they are required to share their medicine.

It thus appears that defining our self-regarding affairs in terms of other’s interests generates an implausibly deflated account of self-determination, while explicating our self-regarding affairs in terms of other’s rights leads to an unacceptably inflated construal of self-determination. One way to negotiate this impasse is to refine the principle of affected interests.² Perhaps not all interests will suffice; only the vital interests of others limit our right to self-determination. If my interest in their medicine is vital whereas my interest in joining Kelce and Swift’s marital union is not, for instance, a distinction like this might allow us to thread the needle between under- and over-inclusive accounts of self-determination.

Although I understand the temptation to pursue this strategy, I would recommend an alternative approach. I think we should combine Nozick’s understanding of our claim to self-determination as bounded by the rights of others with a different account of rights. This is not the place to develop and defend a comprehensive account of our moral rights, but I do want to register my conviction that Nozick wrongly denies the existence of positive rights. If I am sufficiently imperiled and Travis Kelce can easily save me, then I have a right to his easy rescue. If Kelce could save my life merely by calling 911, for instance, then I believe that he is obligated to do so. By affirming samaritan rights, we can construe self-determination as restricted by the rights of others without implying that an action qualifies as self-regarding just in case it does not harm anyone else. Even if Kelce and Swift do not harm me when they withhold their medicine from me, they may violate my rights by wrongly failing to benefit me. And if their refusal to share medicine violates my samaritan right, Kelce and Swift cannot justify their conduct as merely exercising their rights to self-determination.³ In sum, the crucial question is not whether others would be affected or harmed, but whether they would be wronged by one’s behavior.

II. THE RIGHT TO DO WRONG

One objection to conceiving of self-determination as limited by the rights of others is that doing so appears to leave no room for the right to do wrong. If the boundaries of our self-regarding behavior are determined by the rights of others, then we can never justify violating the rights of others merely by appealing to our right to self-determination. We might permissibly infringe rights in extraordinary circumstances when these rights are overridden, but the permissibility of our doing so would depend upon the existence of a lesser-evil justification rather than an appeal to our self-determination.⁴ Some will regard this implication as problematic, though, because many follow Waldron in supposing that we sometimes enjoy a “right to do wrong.”⁵

There is a lot to like about Waldron’s landmark introduction to this idea, but it is important to appreciate two things about his thesis. First, (as is widely recognized) Waldron never meant to defend the permissibility of disrespecting a duty; that claim would involve an elementary Hohfeldian error. Instead, Waldron suggested that wrongdoers might have a right against external interference. As Waldron puts it, “(t)he cutting edge of the claim that P has a right to do A is the correlative claim that other people are morally required to refrain from interfering with P’s performance of A. If P has a right to do A, then it follows that it is wrong for anyone to try to stop P from doing A.”⁶ In short, Waldron’s thesis is about enforceability rather than permissibility.

Second (and less often appreciated), Waldron’s thesis plausibly applies to suberogatory but not to impermissible acts.⁷ In an effort to motivate the idea that others may not permissibly interfere even when another person acts wrongly, Waldron begins his article with a number of cases, including the following:

Someone uses all the money that he won fairly in a lottery to buy racehorses and champagne and refuses to donate any of it to a desperately deserving charity.

A man refuses to give a stranger in the street the time of day when he asks for it or coldly rebuffs attempts at conversation in a railway compartment.⁸

I agree with Waldron: even though the lottery winner ought to be more charitable and the man in the street or in the railway compartment ought to be less rude, third parties may neither force the lottery winner to donate to the charity nor force the man on the street to tell a stranger the time. Imagine, however, if the man on the street wanted to punch a stranger in the face. Would it still be impermissible for a third party to forcibly interfere? Presumably, not. But why not? If there is a right to do wrong, why does a person not enjoy a right against external interference when he seeks to assault an innocent victim?

The crucial factor, I believe, is whether one poses a threat to the rights of others. We may interfere with someone’s attempt to punch an innocent victim in the face because the prospective victim has a right against being punched, but we may neither force the lottery winner to donate to a worthy charity nor force someone to tell a stranger on the street the time because the charitable institution has no right to this donation and the person on the street lacks a right to be told what time it is. To confirm this, consider two instances when I have moral reasons to allow someone to go ahead of me in the line at the grocery store. In the first scenario, imagine that it is late at night, I have cart full of groceries, and I notice that the weary-looking woman directly behind me is buying only one item. In the second case, I quickly step in front of a man in line when he is not looking. Intuitively, it seems as though the man in the second case might permissibly use proportionate force to get in front of me in line but the woman in the first case may not. What explains this difference?

In both scenarios I would act wrongly if I refused to let the person behind me go before me, but the types of moral reasons at play are importantly distinct. When I have cut in front of someone already waiting in line, I act contrary to duty-imposing reasons that correspond to this person’s right that I respect his place in the queue. I also have moral reasons to let the woman buying only one item go ahead of me, but these are virtue-making rather than duty-imposing reasons.⁹ I act impermissibly when I cut in front of someone already waiting in line, whereas I act permissibly but in a suberogatory manner when I do not offer to allow the weary person to check out before me. Put another way, I can act in a manner that reflects poorly on my character even when I do not act unjustly. Most importantly for our purposes here, I warrant moral criticism in either case, but others may forcibly interfere only when I disrespect duty-imposing reasons that correspond to another person’s moral right.

Not everyone will share the same views regarding where to draw the line between the deontic versus aretaic domains of morality. Despite this room for debate, I do not take this general picture to be controversial. As Jeff McMahan says, “(i)n common sense morality, what it is permissible to do to a person is not determined solely by what the consequences of one’s action will be for all those affected. It depends also on what rights the person has and whether he or she has done anything to waive or forfeit them.”¹⁰ And following Judith Jarvis Thomson, I believe that one becomes liable to defensive force just in case one threatens to violate the rights of others.¹¹

This analysis illustrates how nicely Waldron’s and Thomson’s iconic articles complement one another. Whereas Thomson contends that third parties may resist when someone poses a threat of violating a right, Waldron emphasizes that we may not forcibly interfere whenever someone is not violating the rights of others. Even if all rights are enforceable, there is plenty of room for a right to do wrong because no one has the moral standing to forcibly interfere when one acts in a suberogatory manner. Put in terms of Waldron’s examples, the lottery winner and the man on the street or in the railway compartment remain morally invulnerable to coercion despite exhibiting vices, because their vicious behavior is within their rights.

Given that many have failed to recognize that Waldron’s view must be restricted in this way, I want to emphasize that my construal is by no means novel. In an early response to Waldron, for example, William Galston observes that “only moral wrongs that do not involve the violation of others’ rights can possibly fall within the purview of Waldron’s thesis.”¹² And in his own analysis of the right to do wrong, Gerhard Overland insists that “(t)his kind of right would not amount to a right to do wrong, only a right not to do what is morally best.”¹³ While I understand why Overland might reserve the term “wrong” for only impermissible conduct, I am happy to extend it to behavior that is either impermissible or suberogatory. Like Galston and Overland, though, I simply want to emphasize that Waldron’s important thesis is most plausibly understood to apply only to conduct that does not violate the rights of others. Finally, while Waldron did not explicitly endorse this construal in “The Right to Do Wrong,” he did so in his later essay, “Kant’s Legal Positivism,” where he stressed that “(o)ne of the ways in which we delineate the domain of rights and justice from the rest of morality is that rights and justice may properly be enforced.”¹⁴

Not everyone will agree with how I have put things to this point, but I hope that most will accept the general picture, which amounts to little more than observing that Waldron never meant to contest the old adage that “the right to swing my fist ends where the other man’s nose begins.”¹⁵ In any event, I hope that the preceding is not terribly controversial, because I offer it merely as prelude to the main focus of this essay: whether groups can acquire a more robust right to do wrong. Pretheoretically, I think we should be skeptical. After all, if A, B, and C each individually lack the right to punch me in the face, it is hard to see how they could gain this right as a group. If, after sequestering in a room with B and C, A emerged and announced that she now has a right to punch me in the face on behalf of the group of A, B, and C, one would presumably be mighty curious to hear what transpired in that room to generate this newfound right. We should keep an open mind, though, because our moral relations may change once political states are introduced, and leading political theorists such as David Estlund, Anna Stilz and Jonathan Quong have all defended an official’s right to do wrong on behalf of her political state. Let us examine their arguments in turn.

III. ESTLUND ON FOLLOWING UNJUST ORDERS

Estlund does not invoke Waldron’s phrase, “the right to do wrong,” but in “On Following Orders in an Unjust War” he is explicit that “even acts that would wrong their victims can sometimes be rendered permissible by suitable authoritative sources.”¹⁶ More specifically, he argues that unjust combatants may permissibly kill just combatants—even though doing so wrongs the just combatants—as long as the unjust combatants are ordered to fight by a suitably democratic state. Estlund’s analysis is subtle and nuanced, but the crux of his argument is that, just as we believe that a jailer may permissibly enforce an unjust prison sentence against someone who was either wrongly convicted or received an unjustly severe sentence, we should conclude that soldiers may permissibly follow orders to target enemy combatants who have done nothing to make themselves liable to attack.

Estlund does not believe that we should always follow orders. More minimally (and more plausibly), he suggests that we are obligated to follow orders issued from sufficiently epistemically authoritative sources. If a jailer works for a corrupt criminal legal system that routinely convicts defendants merely because they are political dissidents, for instance, the fact that someone was convicted would not sanitize the jailer’s actions. If the legal system is properly designed to convict all and only those who are guilty of wrongdoing and the officials have conscientiously sought to follow the rules, however, then presumably the jailer would be wrong to replace the system’s verdict with his own judgment.¹⁷ No court system is infallible, so there will inevitably be cases in which an innocent person is wrongly convicted or a guilty person is sentenced too harshly, but as long as the verdict is not patently unreasonable and all of the officials have honestly done their best to follow the rules, the jailer is morally obligated (and thus permitted) to impose the unjust sentence, even though this punishment wrongs the victim.

Estlund presumes that few readers will contest this assessment of the jailer in a just criminal legal system. His analysis becomes more original and interesting when he explains why the jailer is obligated to defer to the system’s judgments, and the implications this has for the authority of democratic decision-making. According to Estlund, the jailer must defer to the system’s verdicts because of this system’s epistemic authority. If the rules of the system have been properly designed and are conscientiously followed, we should expect it to reliably convict culpable wrongdoers and exonerate those who are innocent. No system can rule out all mistakes, but if criminal legal procedures are properly designed and conscientiously followed, they enjoy an epistemic authority that warrants respect. The jailer is thus not at liberty to pick and choose when to act in accordance with the system’s verdicts. Provided that the verdicts are not unreasonable, the jailer is obligated to comply with the system’s judgments, even when they are incorrect.

Estlund then argues that this analysis has important implications for the authority of democracy. Just as properly designed criminal legal systems do an exemplary job of sorting the innocent from the guilty, properly designed democratic processes reliably distinguish between just and unjust wars. Of course, no political system is error-proof, but just as jailers should defer to the legal system’s verdicts, soldiers are obligated to fight even when the democracy has mistaken an unjust war to be just. As Estlund puts it, “this is the reason I would offer for thinking of certain democratic political arrangements as the epistemic authoritative procedure that is the counterpart to the jury trial: it is the epistemic device that allows the authoritative order to go to war to sanitize the soldier’s (otherwise impermissibly) knowingly killing innocent people.”¹⁸

One might question whether existing democracies have in fact been designed to reliably refrain from engaging in unjust wars.¹⁹ I am sympathetic to this critique, but I am more interested in focusing on the underlying moral questions than interrogating the extent to which democracies are epistemic authorities. With that in mind, imagine that Ursula is of average intelligence with an average education and access to the information required to assess the morality of any given war. Let us stipulate that, if left to her own devices, Ursula would correctly assess whether any given war is just eighty percent of the time. To be maximally concessive to Estlund, let us suppose that Ursula lives in a model democracy that correctly distinguishes between just and unjust wars ninety-nine percent of the time. Finally, let us suppose that, in compliance with an order to fight in an unjust war, Ursula kills Jane, a just combatant who has done nothing to make herself liable to attack. Estlund believes that Ursula was obligated (and thus permitted) to kill Jane even though Ursula’s aggression wronged Jane. I disagree. In my view, even if Ursula should not be blamed for deferring to her government’s orders, she acted impermissibly. To appreciate why Ursula acted impermissibly despite the fact that she ought to have deferred to her orders, it is important to distinguish among three questions: (1) Who is responsible for Ursula’s action? (2) Is Ursula blameworthy? (3) Is Ursula’s killing permissible?²⁰

Given that Ursula was following orders, why not conclude that Ursula’s government rather than Ursula is responsible for killing Jane? I am open to assigning responsibility to Ursula’s government, but it is important to note two things. First, that other parties are responsible for Ursula’s killing Jane would not entail that Ursula is not also responsible. Indeed, it does not even entail that Ursula is less responsible for the killing than if she was solely responsible. Moral responsibility is not a zero-sum matter; if ten accomplices are responsible for a murder, it does not follow that the murderer is not also fully responsible for her wrongdoing. And if Ursula freely chose to follow orders, she may be no less responsible for killing Jane than if she did so on her own initiative. Second, even if Ursula were somehow less responsible for killing Jane, this would have no bearing upon the permissibility of her doing so.²¹ If the fact that a democratic government ordered Ursula to fight an unjust war entailed that the government was entirely responsible for Ursula’s killing Jane, for instance, it would still not follow that Ursula’s act was permissible.

Here one might protest that, given the government’s expertise, Ursula should follow her government’s lead on this matter. Indeed, we might criticize Ursula if she refused to defer to an established moral expert on such an important question as whether to fight in a war. Whether or not we should blame Ursula if she did not defer to her government, I grant that we should not blame her for following her state’s directive. It is important to distinguish, however, between whether Ursula’s action was excused and whether she acted permissibly.

To see that Ursula’s action might have been fully excused but still impermissible, consider Jeff McMahan’s case of the Resident:

The identical twin of a notorious mass murderer is driving in the middle of a stormy night in a remote area when his car breaks down. He is nonculpably unaware that his twin brother, the murderer, has within the past few hours escaped from prison in just this area, and that residents have been warned of the escape. The murderer’s notoriety derives from his invariable modus operandi: he violently breaks into people’s homes and kills them instantly. As the twin whose car has broken down approaches a house to request to use the telephone, the resident of the house takes aim to shoot him preemptively, believing him to be the murderer.²²

Even if the Resident’s actions were fully excused, few deny that he is liable to defensive force.²³ Thus, just as we conclude that the Resident acts impermissibly despite being fully excused, there is ample room to acknowledge that Ursula may have been fully excused when she impermissibly killed Jane. Neither Ursula nor the Resident is blameworthy because each acted in a manner that would have been permissible if the evidence available to them had been accurate, but both acted impermissibly because (through no fault of their own) their evidence was misleading. To confirm this construal, notice that, although neither Ursula nor the Resident is morally liable to punishment, both are liable to defensive force because both are in fact posing a threat of killing an innocent victim.

Here Estlund might object that Ursula’s case is importantly distinct from Resident’s because Ursula morally ought to defer to her government, and how can someone act impermissibly when they act as they ought? Responding to this concern requires appreciating that moral principles can function either as standards or as guides.²⁴ Notice, for instance, that some sophisticated utilitarians suggest that the best way to maximize happiness may not come from aiming directly at maximizing happiness.²⁵ Even though such a utilitarian would deny that we have a moral right to self-determination, for instance, she might recommend that we should act as if each of us enjoys a privileged position of moral dominion over our self-regarding affairs, because doing so maximally promotes happiness over the long haul. For such a consequentialist, rights theory functions as the moral guide even while utilitarianism remains the moral standard. In other words, while we should be guided by the idea that people have rights to self-determination, the correct moral action remains whatever maximizes happiness. Utilitarians of this stripe recognize that there will occasionally be cases when the guide misleads agents to act in a manner that does not in fact maximize happiness, but they nonetheless recommend allegiance to this guide because of its promise to maximize happiness over the long haul. In other words, these utilitarians continue to endorse the guide despite the fact that it will occasionally lead agents to act wrongly.

In my view, following Estlund in supposing that (suitably organized) democratic states are epistemic authorities on the permissibility of waging war shows at most only that we should be guided by their orders to fight in a war, not that the moral standard thereby changes from “do not kill innocent victims” to “follow democratically-derived orders.”²⁶ The moral standard remains to refrain from killing those who are not liable; it is just that average combatants such as Ursula should be guided by their democratic state’s orders rather than their own judgment on this matter. Most importantly, on the rare occasion that Ursula’s state mistakenly orders her to fight in an unjust war, she still acts impermissibly when she kills a just combatant. We do not blame Ursula for trusting that her state is not ordering her to act impermissibly, but this is a matter of agent-appraisal rather than act-evaluation. The judgment that Ursula is excused for impermissibly posing a threat to Jane’s life is confirmed by our twin convictions that Ursula has not forfeited her right against punishment but has become liable to defensive force.

As noted above, the essence of Estlund’s argument is that, just as we believe that the jailer permissibly imprisons an innocent person who was duly convicted, we should acknowledge that an unjust combatant may permissibly kill a just combatant when ordered to do so. Because I have not contrasted the unjust combatant’s moral position with that of the jailer, I should acknowledge that, unlike Estlund, I do not regard it as absurd to allege that the jailer acts impermissibly in this case. The jailor should not be blamed for being guided by the verdict of the exemplary legal system, but an excuse does not amount to a justification and thus does not give the jailor the right to imprison an innocent person. Once again, this conclusion is confirmed by our convictions that the jailer is not morally vulnerable to punishment but is liable to proportionate defensive force needed to liberate the wrongly incarcerated victim. As in Ursula’s case, if we take care to distinguish between standards and guides as well as between agent-appraisal and act-evaluation, we can see that the jailer’s wrongdoing may be excused, but an excuse does not amount to a right to do wrong.

IV. STILZ ON LEGITIMACY

In Territorial Sovereignty, Anna Stilz explains a legitimate state’s right to do wrong as follows:

Because of the importance of their members’ interests in political autonomy, I believe that self-determining peoples, like self-determining individuals, sometimes have a “right to do wrong.” An individual is permitted to make a wide range of decisions regarding the shape of her life, but she has a moral duty not to lie, not to break promises, and not to be unkind to others. When considering her domain of choice, she ought to view options that involve deception, promise breaking, and spitefulness as morally “off the table.” Yet, if an individual does lie, act unkindly, or break a promise, her liberty may not be interfered with. We protect her liberty in these wrongful choices because of the importance of allowing her an independent domain in which to shape her life in the light of her values. Similarly, legitimate peoples should be allowed to govern themselves independently—even when the laws they enact depart to some degree from ideal justice—because their members have morally significant political autonomy interests in shaping their collective life on the basis of their own sense of justice.²⁷

Like Stilz, I believe that autonomous individuals and legitimate political states are morally entitled to a sphere of self-determination. As outlined above, though, I am skeptical that one can appeal to self-determination to justify violating the rights of others. In response to this worry, Stilz has since clarified and further defended her stance. In this later work, Stilz distinguishes between basic versus non-basic justice and between “one-off” versus ongoing enforcement.²⁸

Stilz agrees that outsiders may coercively interfere if a state is perpetrating a basic injustice (as it would if it instituted slavery, for instance), but she contends that foreigners may only try to persuade a legitimate state to refrain from committing non-basic injustices (such as privileging the majority cultural group, excluding harmless immigrants, or refusing to fairly share the profits from natural resources). I appreciate that force which would be proportionate when employed to prevent basic injustice might be disproportionate if imposed to prevent non-basic injustice, but I doubt that we may not in any way forcibly resist rights violations that do not qualify as basic injustices. If August is attempting to steal April’s lunch money, for instance, I believe that April as well as third parties may permissibly use proportionate force to prevent August from doing so. It strikes me as counter-intuitive to suppose that April would wrong August if she did anything more than try to persuade him not to violate her rights. Similarly, if Stilz is correct that foreign countries like Chad are morally entitled to some of the profits that Norway is currently hoarding from the sale of the North Sea oil, I fail to see why no one may employ any coercive force whatsoever to prevent this theft. To reiterate the core theme of this essay: why think that Norway’s right to self-determination provides moral immunity in its attempted theft?

In fairness to Stilz, she buttresses her contrast between basic and non-basic justice with an appeal to the distinction between “one-off” interventions as opposed to inescapable ongoing coercion. Whatever qualms she may have with the former, Stilz emphasizes that she is particularly concerned with how one’s independence can be threatened by the latter, which she characterizes as “alien rule.” Returning to the example of August’s stealing April’s lunch money, imagine that August is forcibly prevented from doing so by Albert, who is considerably bigger and stronger than everyone else. However problematic Albert’s coercion may be in this instance, Stilz would be much more worried about an ongoing situation in which Albert forcibly prevents August not only from stealing anyone’s lunch money, but from violating any rights. According to Stilz, this unilateral subjection of ongoing and global coercion is doubly problematic: it compromises August’s “independence” and it “sets up a hierarchical relationship” between Albert and August.²⁹

I do not share Stilz’s concerns. To begin, I am curious whether appealing to August’s “independence” is anything other than invoking his right to self-determination. If independence is just another word for self-determination, then one can no more appeal to independence than to self-determination to justify violating another person’s rights. Of course, Stilz could mean for independence to be importantly distinct from self-determination, but, if so, she does not explain this. In addition, if there is nothing wrong with Albert’s preventing August from stealing April’s lunch money, why is there anything wrong with Albert’s also preventing August from stealing Amy’s lunch money or from vandalizing Adam’s property? If each of Albert’s interventions is permissible, what is morally problematic about their conjunction? Stilz alleges that their conjunction creates a hierarchical relationship between Albert and August, but this claim invites new questions. If it is not objectionably hierarchical for Albert to prevent August from stealing April’s lunch money, why is it objectionably hierarchical for Albert to subsequently prevent August from stealing Amy’s lunch money? And given that Albert forcibly employs his superior size and strength against August only when the latter attempts to violate someone’s rights, August could avoid all such interactions by simply respecting everyone else’s rights. It is hard to see how August can rightfully object to repeated unpleasant interactions with Albert any more than a recidivist in a legitimate state can complain about how often her fellow citizens call the police on her.

Having registered my disagreement with Stilz, let me acknowledge that I am more sympathetic to other claims she advances regarding punishment and the indeterminacy of rights:

Suppose that an enforcer does not simply engage in a one-off defense: rather, she imposes her conception of justice on others through relations of rule that are difficult for them to escape. She may threaten her neighbors with serious penalties if they do not comply with her directives, where she has the known power to follow through and where her neighbors cannot easily move away from her vicinity. (Note the difference between imposing rules through threats of punitive force vs. engaging in limited acts of self-defense or resistance.) Further suppose she is enforcing not just basic bodily rights but a wide range of claims, including some that are at least partly convention dependent (that is, rights that could be specified in a variety of morally acceptable ways). Finally, suppose the interpretation of justice is substantively correct (she enforces everyone’s objectively justified moral rights).³⁰

To appreciate the force of these concerns, consider them in terms of August, April and Albert negotiating a potential conflict that Stilz offers in her first book, Liberal Loyalty.³¹ Stilz stresses that property rights may not be fully specified in the state of nature. If August and April own property adjacent to one another, for instance, in the absence of institutional specification, there may not be a determinate answer to the question of whether August is entitled to build a huge building on the very edge of April’s property. Without institutional specification, August might reasonably believe that he is entitled to do so, and April might reasonably believe that August’s doing so would impermissibly encroach upon her property right. Suppose that after August builds this building, Albert sides with April and punishes Albert for violating April’s rights.

Following Stilz, this scenario may appear doubly problematic. First, might does not make right, so the mere fact that Albert is stronger than everyone else does not entitle him to unilaterally specify the precise contours of this indeterminate property right. It would be morally preferable if this indeterminacy were resolved omnilaterally through something like a democratic process. Second, given that August reasonably believed that his conduct was within his property rights, even if we assume for the sake of argument that he violated April’s rights, August did not do so culpably. Put in terms of actus reus and mens rea, even if we assume that August did a bad act (actus reus), he lacked a guilty mind (mens rea). And because excused wrongdoers do not forfeit their rights against punishment, Albert’s punishment violated August’s rights.

Given my inclination to grant all of this to Stilz, why am I so skeptical that Albert and August might have a problematically hierarchical relationship? The key is in the parenthetical sentence from Stilz’s quote: “Note the difference between imposing rules through threats of punitive force vs. engaging in limited acts of self-defense or resistance.” Put simply, I can share Stilz’s worry about Albert’s punishing August because the imposition of punitive force involves reactive attitudes and thus raises different issues than the imposition of preventive force, which is the focus of this paper. This difference is confirmed by our analysis of McMahan’s example of the Resident invoked above. Because the Resident was excused in mistaking the murderer’s twin as the murderer, he would not be liable to punishment for trying to kill the twin. Excused or not, though, the Resident posed a threat of violating the twin’s right to life and therefore was liable to defensive force. Thus, even if Stilz’s doubts about the permissibility of Albert’s punishing August are well founded, they are irrelevant to this discussion, which concerns preventive rather than punitive force.

Here one might wonder why our worries about Albert’s punishing August for violating an indeterminate right do not also apply to Albert’s preventing August from violating an indeterminate right. The idea is that, just as Albert objectionably acts unilaterally if he punishes August for violating an indeterminate right, he objectionably acts unilaterally if he prevents August from violating an indeterminate right. The preventive force may be less bad (because it does not express moral condemnation for August when he is not culpable), but it remains problematic insofar as Albert unilaterally imposes his contestable moral interpretation upon August.

I have some sympathy for this objection, but it will not vindicate Stilz, because her thesis is that legitimate states have the right to do wrong as long as they do not perpetrate basic injustices, not that legitimate states are morally immune from interference when it is indeterminate whether they are committing an injustice. Recall her claims about a state’s privileging the majority culture, excluding harmless immigrants, or refusing to fairly share the profits from selling natural resources. Stilz does not argue that it is indeterminate whether these practices are unjust. On the contrary, she concludes that, even though each of these practices violates determinate moral rights, legitimate states enjoy the right to do wrong in these cases because none of these wrongs amounts to a basic injustice. To emphasize: even if Norway was violating Chad’s right to its share of the profits from the North Sea oil, Stilz insists that it would be impermissible for Chad to forcibly resist this injustice. Given that Chad’s resistance would not constitute punishment, it would be a “one-off” interaction, and it would be in defense of a determinate right, Stilz’s concerns about an enforcer’s on-going punishment of agents who violate indeterminate rights seem beside the point.

V. QUONG ON THE PUZZLE OF LEGITIMATE INJUSTICE

In Political Philosophy: The Puzzle of Legitimate Injustice, Jonathan Quong reviews and rejects all extant efforts to explain why some unjust laws can remain legitimate before offering his own account, which begins with the observation that states will inevitably fall short of perfect justice:

Democratic institutions, even those operating entirely in good faith and doing the best that we can reasonably expect, are not infallible in discerning the substantive content of justice, and thus some amount of unjust legislation is a permanent burden of our cooperative life. Like the other burdens and benefits of that cooperative life, principles of distributive justice must assign rights and duties concerning this burden. I argue that since some unjust laws are an unavoidable burden arising from our shared attempt to live together on fair terms, justice requires that we share in the responsibility for this unavoidable burden.”³²

I have no quarrel with this important point. This alone will not suffice, however, because a legitimate state’s errors will require some victims to endure more than their fair share of injustice. For the remainder of this discussion, then, I will explore whether Quong can satisfactorily explain why we may not resist rights violations that would require prospective victims to absorb more than their fair share of injustice.

Quong also emphasizes that officials act on behalf of the entire political community: “State officials who enforce such (unjust) laws do not commit injustice while pursuing their own private goals. They are instead authorized agents acting on behalf of the political community. Thus, the idea of holding people substantively responsible for the costs of their chosen plans and projects does not support allocating the costs of injustices to state officials.”³³ This too is an important point. Quong and I derive very different conclusions from this observation, though, so it is worth reviewing his reasoning closely. According to Quong,

state officials who implement democratically authorized directives when those directives fall within the scope of reasonable disagreement are not specially liable to harm. When the laws or directives these officials enforce are unjust, the error is not theirs, but rather one that we collectively make. To hold state officials uniquely or specially liable would be deeply unfair—it would be to hold state officials substantively responsible for an unavoidable shared or collective burden of our cooperative political life.

In sum, on my view, a state official is nonliable to defensive harm when they enforce an unjust law and

• A) the official is non-negligently acting within the scope of her authority in the service of

• B) enforcing or otherwise carrying out democratically authorized law or policy that

• C) constitutes the resolution of a reasonable disagreement.³⁴

It is striking that Quong shifts in the first paragraph from his claims that state officials are neither uniquely nor even specially liable to his conclusion in the second paragraph that they are nonliable. I agree that state officials are often not uniquely liable, but presumably one can remain liable even when one is neither uniquely nor even specially liable. With this in mind, let me explain how I regard the liability of officials who enforce their political community’s unjust directives.

When discussing Estlund’s views, I noted that the presence of others who share moral responsibility can neither change the act’s deontic status nor preclude the possibility that an agent remains morally responsible. It does not follow that shared moral responsibility is of no moral consequence, however, as it may impact who, if anyone, is liable to defensive force. It does not seem implausible to suppose that the jailer has a heightened responsibility for imprisoning a wrongly convicted defendant, but let us be maximally concessive to Quong and assume either that (despite his agency and special causal responsibility) the jailer is no more morally responsible or that, even if he is more morally responsible, he is no more liable than all of his compatriots. Even so, I believe the jailer remains liable to defensive force. Let me now explain what I believe may and may not follow from his equal liability.

First, it seems to me that any compensation owed to the wrongly imprisoned person should come from the political community as a whole, rather than solely (or even disproportionately) from the jailer. Second, if the jailer is harmed by proportionate defensive force employed by the defendant, I believe the entire political community should make the jailer whole; it would be unfair to require the jailer to absorb all of these costs on his own.³⁵ And finally, if we are supposing that the jailer is no more morally responsible for this rights violation than any of his fellow citizens, I am open to the idea that all of these citizens are equally liable to defensive force and thus that the defendant could permissibly employ defensive force against any of the citizens. What ultimately distinguishes the jailer, then, is not a principled fact regarding his liability, but the contingent (but nonetheless morally relevant) fact that resistance employed against him—unlike resistance against some random citizen who has nothing to do with the prison—is likely to be especially, if not uniquely, effective in escaping the prison.³⁶

I suspect that Quong would accept my first two points; he would likely contest only my third claim. Because Quong denies that the jailer is specially liable and infers nonliability from equal liability, he would insist that neither the jailer nor any other citizen is morally liable to defensive force. It strikes me as highly counterintuitive, however, to insist that the wrongly convicted defendant must refrain from resisting this severe rights violation merely because the jailer is not specially liable.

Here Quong might respond that his stance on liability is not so counterintuitive once one recognizes that the jailer acts permissibly. As he puts it, “I have argued that state officials are not liable to harmful interference when they enforce reasonable but unjust laws and policies that arise from the democratic process. But legitimacy, as we have been using the term, implies not merely that state officials are immune from liability, but also that those officials act permissibly when they enforce laws or policies.”³⁷ Quong contends that the permissibility of the official’s enforcement of unjust laws follows from the democratic imperative to respect each citizen’s equal right to exercise political power. He formulates his argument as follows:

P1. Reasonable disagreement about justice is possible, even when there are determinate answers about what justice requires.

P2. Justice requires giving people equal opportunity to exercise political power in cases of reasonable disagreement.

P3. If the state fails to enact and pursue a reasonable position that emerges from a fair democratic process, it violates the requirement in P2, and so acts unjustly.

P4. At least one option from an agent’s feasible set must be morally permissible.

P5. A fair democratic process sometimes results in the selection of a policy that is reasonable but substantively unjust.

Therefore,

C1. it must be permissible for the state to act unjustly, either by violating the principle requiring equal opportunity to exercise political power, or else by imposing a substantively unjust policy.³⁸

Put in terms of Estlund’s case, Quong contends that if the jailer enforces a sentence handed down in response to a violation of an unjust law, he would thereby violate the rights of the convict, but if he allowed the duly convicted person to escape, the jailer would violate his compatriots’ democratic rights to equal opportunity to exercise political power.

Following Zofia Stemplowska and Adam Swift in “Dethroning Democratic Legitimacy,” we might counter that this argument shows only that officials should compare the weight of these two injustices on a case by case basis, not (as Quong alleges) that officials should uniformly obey all reasonable directives with the appropriate democratic pedigree.³⁹ This is a plausible position, but I contend that officials who disobey directives to violate rights do not act unjustly because their fellow citizens lack even a pro tanto claim to deference when they direct state officials to violate rights.

I appreciate that my critique is more ambitious than Stemplowska and Swift’s, but Quong cannot summarily dismiss my stance as outlandish because it is actually a close cousin of his. Quong and I both deny that democratic law is globally authoritative; the only difference is the content of our respective defeaters. Whereas Quong concedes that unreasonable unjust laws lack the moral authority they would otherwise enjoy, I contend that laws requiring the commission of injustice lack moral authority. While Quong allows that we may fight back against the imposition of an unreasonable injustice, I submit that we may proportionately resist any injustice.

At this stage of the essay, it should be clear that I favor injustice over unreasonable injustice as a defeater because it squares with Thomson’s compelling theory of the permissibility of defensive force, the best understanding of Waldron’s defense of the right to do wrong, and a proper appreciation of the right to self-determination. It also aligns with the widespread conviction (which Quong shares) that McMahan’s Resident is liable to defensive force even though his mistaking the murderer’s twin as the murderer is in no way unreasonable.⁴⁰ In addition, it has other salient advantages. Following Joel Feinberg, moral rights seem precisely the type of entities on which prospective victims can stand.⁴¹ Perhaps Quong is correct that we may require that others be reasonable, but I am more confident that each of us is entitled to insist that others respect our rights. Most importantly, though, my position parallels how we analyze other content-independent sources of obligations. To see this, imagine that Jane promises Jennifer that she will violate one of Joan’s moral rights. One might suppose that the stringency of Jane’s duty of fidelity to Jennifer must be weighed against the stringency of Jane’s duty not to violate Joan’s right, but the standard approach to promising is to insist that whatever moral weight Jane’s promise to Joan would otherwise have is undermined in this case by the content of the promise. In other words, promises to commit injustice are typically thought to have no binding force.

Along these same lines, consider the content-independent duty that children may have to obey their parents. I am not sure that I must follow my mother’s directives, but it seems reasonable to think that I should do things like call my brother if she asks me to. I doubt that I have any moral reason to violate someone’s rights, however, even if my mother directs me to do so. My mother might feel disrespected if I refused to commit an injustice as she instructs (especially if she does not regard it as an injustice), but I am not sure that my disobedience actually disrespects her. Most importantly, though, even if my refusal to follow my mother’s unjust orders somehow disrespects her, it does not wrongly disrespect her. Whatever psychological reaction we might expect my mother to have, she has not been wronged by me. I would analyze a state official’s moral relationship to her compatriots analogously. If an official is democratically directed to commit an injustice, her refusal to comply may disrespect her fellow citizens, but it does not wrongly disrespect them. To emphasize: I do not deny that officials have a general, content-independent obligation to comply with democratically-derived laws; I merely contend that the normal moral force that laws of this pedigree enjoy is undermined when obedience would require the official to violate someone’s right. Quong’s caveat that officials need not obey unreasonable unjust laws mitigates the extent to which his argument generates unpalatable implications, but because reasonable injustices can still be severe, Quong’s account does not ensure that all such injustices will be modest enough for us to require that citizens endure them as their fair share of unavoidable injustice absorption.

Quong anticipates this type of objection and responds by invoking the right to do wrong: “Even if state officials act impermissibly in enforcing unjust laws, they can still have protected claim rights in doing so. There’s nothing incoherent about having a claim right that protects one from harm while being engaged in wrongful conduct, and indeed a right to do wrong in this sense is familiar from ordinary interpersonal morality.”⁴² As I have sought to illustrate above, however, the right to do wrong cannot vindicate Quong, because agents are not morally immune from external interference when their wrongful conduct violates the moral rights of others.

Before closing, it is worth considering one last potential response. To be clear: Quong offers the following in reply to someone who believes that victims of reasonable injustices are owed compensation, but someone sympathetic to Quong’s approach might repurpose it as a potential counterargument against theorists who deny P2. The crucial point is that, even if we are in theory tempted by the idea that all injustice is impermissible, there will in practice remain intractable disagreement about precisely which laws require officials to violate rights. As Quong explains, “The fact of reasonable disagreement is assumed to be permanent. We will thus never agree on which (reasonable) laws are unjust…The proposal to simply compensate those who have been the victims of unjust laws is thus practically hopeless—we might as well just assume away the problem of unjust law altogether.”⁴³

I acknowledge that it is unrealistic to suppose that state officials will invariably be able to discriminate between just and unjust directives. Given this, perhaps officials should simply defer to the (reasonable) results of the democratic process. If this sounds familiar, it is because we have come full circle to a reiteration of the Estlundian point that state officials will typically do better over the long haul if they defer to their democratic authority. Even if we concede arguendo that democracies are epistemically superior to the average state official, though, this implies at most that officials in democracies should be guided by their political directives, not that the moral standard has somehow changed from “violate no rights” to “always obey (reasonable) democratically-derived commands.” Quong was thus wise not to employ this tactic against critics of P2, as doing so mistakes a moral guide as the moral standard.

VI. POTENTIAL REJOINDERS

Before concluding, let us consider three ways in which someone sympathetic to Estlund, Stilz, or Quong might counter my critique: (1) one might deny that all rights are enforceable; (2) one might argue that moral rights function as guides rather than standards; or (3) one might object to my position’s revisionist implications regarding democracy.⁴⁴

While few deny the general connection between rights and enforceability, some allege that there are exceptions, such as the minor wrong committed by the grocery shopper who cuts in the line at the checkout. Trivial rights violations might seem too inconsequential to justify inflicting harm.⁴⁵

I defend the enforceability of rights elsewhere; here I make only two points about minor rights violations.⁴⁶ First, I follow authors such as Douglas Husak and Jessica Flanigan in believing that de minimis wrongs generate practical but no principled barriers to enforcement.⁴⁷ When discussing the failure to fulfill a relatively inconsequential promise, for instance, Flanigan suggests that: “(i)t is only because there are so few means of interference that would be necessary, effective, and proportionate to minor acts of wrongdoing…that we usually think that minor acts of deception are unenforceable rights violations. But these rights violations are not in principle unenforceable, only in practice. In many cases, interfering to prevent minor violations would be disproportionate to the nature of the violation.”⁴⁸

Even if minor rights violations were unenforceable, though, this would be irrelevant to our debate here. Given that I agree with Quong that citizens must accept their fair share of the unavoidable injustice of cooperative life, my view is that we may forcibly resist only when refraining from doing so would require one to endure more than one’s fair share of injustice, which would be far more than a de minimis rights violation. To emphasize: even if I am wrong about the permissibility of enforcing minor rights, this would not show that we may not resist jailors who seek to wrongly imprison us, combatants who seek to wrongly kill us, or foreigners who seek to wrongly deprive us of our share of the profits garnered from natural resources.

A related objection questions my presumption that rightholders may always enforce their own rights. After all, if we may unilaterally defend our moral rights, there would be nothing wrong with vigilante resistance, which seems counterintuitive. Are we not morally required to enlist police officers rather than unilaterally resort to violence, at least when the authorities could reliably protect our rights?

A philosophical anarchist such as John Simmons would not be troubled by this implication, but I share this critic’s aversion to vigilantism. In my view, we are obligated to obey the just laws of legitimate states as our fair share of the collective samaritan chore of rescuing others from the perils of the state of nature, and, as I have argued elsewhere, our political obligations include a duty to defer to the legal authorities.⁴⁹ The position I favor thus sits comfortably between those of Simmons and Estlund. Unlike Simmons, I believe that citizens may not personally punish those who have wronged them, even if state officials incorrectly exonerate these wrongdoers. Unlike Estlund, though, I contend that we need not acquiesce if we are wrongly convicted and can evade the unjust punishment without employing disproportionate force. Because I believe that citizens in legitimate states lack some of the rights to enforce morality that they would have in the state of nature, my position does not open the door to vigilantism in the manner this objection suggests.

A second line of critique protests that moral rights must function as guides if they are to give agents a satisfactory degree of control over their lives. If rights functioned as moral standards (as I suggest), then agents could be liable to defensive harm despite having followed the appropriate moral guides. If I concede that the jailor in Estlund’s case ought to have been guided by the court’s verdict, for instance, my insistence that he nonetheless remains liable to defensive force objectionably provides the jailor with insufficient control over his life.

Before exploring ways to ensure that luck plays no role in one’s liability, it is worth noting that Estlund’s, Stilz’s, and Quong’s proposals can also leave agents liable, even when they reason as they ought to. This is because these authors concede that we may defend ourselves against unreasonable injustices, and a state agent’s reasoning as she ought to provides no guarantee that she will not perpetrate an unreasonable injustice. Just as one might rationally believe that the earth is flat, one might rationally mistake an unreasonable injustice to be reasonable or even entirely just. For a concrete example, recall that Estlund specifies that a jailor should defer to the court’s verdict only if the criminal legal system’s rules are properly designed and conscientiously followed. Imagine that a jailor works for a just legal system whose rules are almost always conscientiously followed, but in this instance a rogue prosecutor has framed an innocent defendant. Because the jailor has no reason to suspect that this defendant has not been conscientiously convicted, if he reasons as he ought to, the jailor will imprison this wrongly convicted person. In cases such as this, mere luck results in the jailor’s acting wrongly and thus being liable to defensive force.

Those troubled by these types of cases have two ways to ensure that agents cannot be rendered liable merely because of bad luck: (1) endorse subjectivism regarding rights and duties, or, more narrowly, (2) insist that only culpable agents are liable to defensive force. Both options have proponents, but each comes with costs.

If our duties and rights are evidence-relative rather than fact-relative, the jailor in the preceding case would not act wrongly. Rather than try to adjudicate the debate between subjectivism versus objectivism here, I merely note that the excellent analyses of Peter A. Graham and Julia Driver make plain why most contemporary moral theorists are objectivists regarding act evaluation and subjectivists regarding agent appraisal.⁵⁰ Among other things, objectivism explains why we sometimes revise our judgments about an action’s deontic status in light of new evidence.

Given the problems associated with wholesale subjectivism, one might more minimally follow Kimberly Ferzan in opting for the culpability account of liability to defensive force precisely because it insulates agents against bad moral luck.⁵¹ On this view, because Estlund’s jailor is not culpable, he would not be liable to defensive force. As noted above, however, virtually everyone who writes on the morality of defensive force (including Quong) rejects the culpability account because of its implication that fully excused agents (such as the Resident in McMahan’s case) are not liable to defensive force. In sum, while the temptation to insulate agents from moral luck understandably attracts some to subjectivism and/or the culpability theory of liability, there is reason to be wary of both approaches. To be clear: I am not accusing Estlund, Stilz, or Quong of having endorsed either of these theories; I am merely highlighting the costs for anyone who might be tempted to do so in response to my critiques.

The third objection is that my view is insufficiently deferential to democratic participation. This is an important and thorny issue, but space constraints permit me to cover only the basics here. To begin, notice that the battle lines on this issue have already been drawn between Michael Walzer’s view that democratic governments “have a right to act wrongly—in much the same way that they have the right to act stupidly” versus Gerhard Overland and Christian Barry’s stance that “while there are many reasons not to interfere actively with the decisions made within democratic societies, none of these is based on the rights of its members to do wrong.”⁵² It should come as no surprise that I favor the latter; one can appreciate the value of democracy without insisting that democratic states may violate rights with impunity. Imagine, for instance, that the only way to prevent an unjust war would be to interfere with a democratic election that would otherwise result in the installation of a leader who would wage this destructive military campaign. If this electoral interference could be achieved without disproportionate damage, I believe it would be justified.

Do I really want to insist that we may permissibly interfere with democratic procedures whenever these processes would otherwise result in a rights violation? Estlund, Stilz, and Quong avoid this extreme position by insisting that we must respect unjust democratic results so long as the injustice is not unreasonable. For two reasons, though, I doubt that reasonableness is the correct place to draw a line in the sand. First, virtually everyone who works in the morality of defensive force denies that anything like reasonableness renders agents immune to liability in other contexts, so I believe we should be similarly skeptical that it has this sanitizing power in democratic deliberation. Second, a reasonable injustice can be quite grave, so insisting that individuals must passively endure all reasonable injustices demands too much of innocent victims. To emphasize: one of the recurring themes of this discussion has been that there is not a perfect correlation between an injustice being unreasonable and its being grave, and where these two characteristics come apart, we should focus on the gravity, rather than the putative unreasonableness, of the injustice.

Given my disinclination to tolerate all “reasonable” injustice, how can I avoid the unpalatable implication that we may resist all rights violations? Put another way, why do I agree with Quong that citizens must accept their fair share of the unavoidable injustice of cooperative life? As I have argued elsewhere, the best reason to affirm that states can remain fully legitimate despite falling short of perfect justice is because the functions states perform are valuable enough to justify at least some rights violations.⁵³ Just as a military campaign might secure sufficiently important results to justify some collateral damage, a political state does not necessarily become illegitimate as soon as it violates a moral right. I thus submit that the extraordinary benefits that legitimate states secure explain why it is not too much to require that citizens endure their fair share of the inevitable injustice. Importantly, though, this concession is compatible with insisting that no one is obligated to endure rights violations that exceed this threshold.

VII. CONCLUSION

This essay is littered with specific accusations. I have accused David Estlund of conflating guides and standards; I have accused Anna Stilz of shifting the debate from resisting nonbasic injustices to punishing violations of indeterminate rights; and I have accused Jonathan Quong of mistaking equal liability for no liability, incorrectly presuming that democratic directives to violate rights retain the authority they would otherwise have, and relying upon an implausible understanding of the right to do wrong. In closing, let me now conclude in much more general terms. It is tempting to suppose that one somehow disrespects an agent whenever one interferes with her actions and that it is therefore wrong to prevent an agent from violating another person’s rights. I urge us to resist this temptation. Consider, for instance, whether Albert should prevent August from stealing April’s lunch money. Even if Albert’s intervention would somehow disrespect August, it would not wrongly disrespect him. If anything, we should worry that, if Albert could easily defend April, his refusal to intervene would wrongly disrespect her. And as for April herself, rather than object that April’s resisting August would be arrogant, we should worry that April’s acquiescence would be objectionably servile.⁵⁴

ACKNOWLEDGMENTS

I am grateful to David J. Clark and Anna Stilz for helpful feedback on an earlier draft of this article.

COMPETING INTERESTS

The author declares that he has no competing interests.