Recent history has seen a rapid growth in the involvement of private parties in war conflicts. In 2020, there were almost twice as many private contractors as US soldiers in Afghanistan.1 In the ongoing war in Ukraine, private actors are allegedly deployed by both parties in the conflict.2 Originally hired by states to provide support services from catering to logistics, private military firms (PMFs) have progressively taken on functions, including combat tasks, that were deemed, at least in the last century, inherently governmental.3 The phenomenon amounts to an unprecedented form of corporatized mercenarism.4 The condemnation of mercenarism has an illustrious history. While Machiavelli famously deprecated mercenaries' lack of loyalty and tendency to corrupt the state,5 Rousseau worried that hiring mercenaries, rather than having citizens fight wars, would lead the latter to value comfort more than republican freedom. Recent critics argue, among other things, that fighting for profit is inherently wrong;6 that the privatization of war leads to an unjust distribution of access to security;7 that it allows both states and private parties to escape democratic accountability;8 and that it provides incentives to escalate conflicts and to increase the use of violence in the battlefield.9 Some contemporary philosophers have, on the other side, shown a friendlier face towards mercenarism. Most prominently, Cécile Fabre argues that, at least under ideal circumstances, private parties have a right to sell their soldiering services to states, for the purpose of just defensive killing, and states are at liberty to buy those services from them.10 Departing from Fabre's (qualified) defense of mercenarism, my goal is to provide an account of the wrong of privatized war, which neither rests on the controversial claim that fighting for profit is inherently wrong, nor assumes that privatization leads to unjust distributive outcomes, a lack of accountability, or the disproportionate use of force (although it may). I argue that, even in the absence of such problems, the privatization of (at least some) military tasks would amount to a condition of double domination, whereby both those exposed to the mercenary's use of force and, perhaps more surprisingly, the mercenary themself is dominated: that is to say, subject to the arbitrary will of another. This can occur even within the context of a just war. To make my case, I will first argue that the state's outsourcing of certain military tasks to private parties, including most combat tasks, consists of a system of contracts between states and such parties that contain either invalid (not binding) or seriously problematic promises. Either the mercenary's promissory offer to perform those tasks entails the alienation of certain rights that cannot be so alienated, in which case the promise is invalid, or, if limited to the mere waiving of those rights, then the state is generally under a duty not to accept the mercenary's offer. Therefore, either the mercenary acquires no promissory duty to perform the content of the promise, or, if they do acquire it—because the state wrongly accepts the offer—they should be released immediately. Further, insofar as the mercenary comes to acquire a permission to use force on behalf of the state only as a necessary means to discharge their promissory duty to fight on its behalf, if the promise misfires, the mercenary does not acquire any such permission either. Two implications follow. On the one hand, any attempt, on the part of the state, to compel the mercenary to perform constitutes an instance of domination over them—an arbitrary imposition of duties on them. On the other hand, since the mercenary lacks permission to exercise force on behalf of the state, and since one can arguably legitimately enforce rights on behalf of a state only if one does so in virtue of a grant of permission therefrom,11 those exposed to this exercise are also dominated—subject to an arbitrary will. One important and difficult part of the argument will be to explain why the relationship between states and regular soldiers does not present the same normative problems that afflict the contractual relationship between the state and mercenaries. A concern with double domination provides strong reasons to limit the privatization of war, even if private actors are committed to only fight just wars and to avoid any disproportionate use of force. This does not mean, however, that the privatization of all military tasks, in all circumstances, is equally problematic. My argument provides reasons to condemn the privatization of combat tasks on the ground, and possibly also of non-combat but essential support tasks in dangerous missions for which replacements cannot be easily found, but not the outsourcing of non-essential tasks or of combat tasks with no or very little risk of death to those providing them, such as drone operating. Further, my account condemns the privatization of military tasks by normally functioning states—that is to say, political entities that have both the institutional and economic capacity to maintain a public military—but not necessarily by very weak or failing states. I take these qualifications to be a strength rather than a weakness of the argument, since they reflect the complex nature of the question at stake, and of reality. Although the argument relies on empirical assumptions about contemporary mercenarism, some of which could arguably be changed without compromising the conceptual integrity of the practice itself, nevertheless, the objection is robust insofar as, as we shall see, making the changes needed would undermine the reasons why mercenarism exists in the first place. an individual who offers his military expertise to a belligerent [state] against payment, outside the state's military recruitment and training procedures, either directly to a party in a conflict, or through an employment contract with a private military corporation.13 Also, and importantly, mercenaries do not occupy an official position within a state's armed forces. Although an ethical assessment of the privatization of war is the main focus of this article, it is not its only ambition. The article also provides an opportunity to reflect on what, if anything, makes the occupancy of a public office normatively special, and suggests a plausible, if not conclusive, answer to the broader question of which jobs, if any, should be performed exclusively by public actors. One upshot of my argument is that we have reasons to publicize—to reserve to public agents—the performance of jobs the function of which (1) is both morally permissible and socially desirable, in the sense that a political society has strong reasons to not eliminate the job in question, and (2) can only be fulfilled if employees are bound and can be compelled, by threat of punishment, to stay when an imminent risk of death materializes. Beyond combat tasks, firefighting and, more controversially, also policing and the provision of emergency care during deadly pandemics are plausible candidates. If mercenarism consists in a contractual exchange for services between a government and a private actor, then, mercenarism is a promissory phenomenon. For what is a contract if not an enforceable promise?14 This way of defining contracts, still dominant among philosophers, also reflects the way most legal systems understand contracts. For example, US contract law defines a contract as “a promise … for the breach of which the law gives a remedy, or the performance of which the law in some way recognizes as a duty.”15 Mercenary contracts are legally binding and mutually conditional promises. By outsourcing military tasks (T) to a mercenary (M), government (G) promises to give certain benefits to M if M performs T on G's behalf. Simultaneously, M promises to perform T if G provides, or commits to providing, those benefits to M. A moral assessment of the privatization of war should thus grapple with the question of whether mercenary contracts contain valid promises and, if so, whether such promises are promises that a state can permissibly hold their parties to. For even if the law of contract should not perfectly track the morality of promises—for example, a marriage contract in which Joseph marries Bob mostly for pecuniary reasons may be immoral (unethical), but nevertheless binding—still, the validity, and thus the enforceability, of contracts is conditional on the validity of underlying promises. Further, a state should not use legal norms to force people to comply with promises they made to the state, if the state itself should not have accepted those promises, however valid, in the first place. If it turns out, then, that mercenary contracts contain invalid promises, or promises a state ought not to accept, we would have strong reasons to question the contracts' enforceability. But what does it mean for a promise to be valid? A promise is valid if it successfully imposes on the promisor a new duty, owed to the promisee, such that the promisor would wrong the promisee by failing to perform. I endorse “the authority view”16 in thinking that what generates promissory obligations is not the fact that the promise creates an expectation in the mind of the promisee, for predictions can do this too, but rather the fact that the promisor transfers to the promisee decisional authority over the promisor, including “the right … to be the one to decide whether to act, how to act and on what grounds.”17 Before promising to buy you a book, I had the right to decide whether to buy it or not. After my promise, you have the right to decide that for me. If what generates promissory duties is a transfer of decisional authority, then, to be valid, a promise must meet an alienability condition: the authority that the promisor purports to transfer to the promisee cannot be inalienable. The promise to become someone else's slave is a paradigmatic example of a promise that fails to meet this condition. Why? For one thing, the kind of decisional authority slavery promises attempt to transfer is arguably itself inalienable. If individuals' right to autonomously decide how to live—a right the slave gives up—is grounded on the value of moral agency, where to be a moral agent is to be a person endowed with free purposiveness—that is, the ability to develop long-term projects and to revise purposes over time, as well as to live according to them—it is plausible to see the very grounding of the right to autonomous living as internally constraining its own alienability, to the extent that irrevocably transferring decisional control over one's life to another, while remaining a person, would contradict moral agency.18 someone motivated to insist on the right to decide for themselves what they are going to do will also be motivated to seek the right to require another to behave in a certain way, where their own decisions depend on the actions of that other person.19 Promises that bind someone to irrevocably transfer the right to decide for themselves what they are going to do are invalid because they violate the rationale that justifies having the power to enter into promises in the first place. This invalidates promises to become someone else's slave, and possibly also promises that irrevocably transfer the right to decide on fundamental questions, such as whether one should live or die. With these clarifications in mind, I now turn to ask: can a mercenary validly offer their combat services to the state? Some answer positively to the question above. Fabre, for example, argues that, in the same way in which a weapon manufacturer can permissibly sell guns to those who need them to justly defend themselves, even though the assistance that they provide involves a contribution to an act of killing, private parties also have the liberty and claim, on grounds of freedom of occupational choice, to offer killing services, as long as the cause they serve is just.20 Just defensive killing also provides, for Fabre, a justification for conferring on states the liberty and power to hire mercenaries for killing services. Fabre's argument is intuitively appealing, but a careful analysis of her analogy reveals a problem. Whereas the manufacturer's offer simply transfers to the state a right over the use of an object (a weapon), those who fight for a state generally transfer to it also decisional authority over their life. Take the case of soldiers. By offering their combat services to the state for a just cause, soldiers transfer to it the authority to decide whether and when they should sacrifice their life for that cause. The state acquires the right to demand that, if circumstances necessitate, the soldier stay, even if the risk to their life amounts to almost certainty of death. If the soldier leaves just because things are too risky, they would be seriously wronging the state. The soldier–state contract thus implies the transfer of what Joel Feinberg calls the discretionary right to life—the authority to decide, at any given point in time, and without external interference, whether or not to end or sacrifice one's own life.21 This right transfer is generally considered morally justified, assuming that the state is committed to only fight just wars and to not demand futile sacrifices from soldiers, because without it, the effectiveness of the military would be compromised.22 The transfer is also legally sanctioned. In the US, for instance, the Uniform Code of Military Justice (UCMJ) makes soldiers who disobey an order on the grounds that it demands self-sacrifice liable to punishment, including execution.23 If this is the content of the promise between the state and its soldiers, we have prima facie reasons to think that the content of the promise between the state and mercenaries should be the same. After all, if demanding sacrifice is necessary for military effectiveness in the case of soldiers, why would it not be equally necessary in the case of mercenaries? The mercenary promise would then (attempt to) transfer something that Fabre's manufacturer's promise does not transfer: the promisor's discretionary right to life. But can the promise, so understood, meet the alienability condition? Whereas the manufacturer can alienate through a private contract their property right over the weapon, it is far less clear that the mercenary, as a private individual, can alienate their discretionary right to life through a simple private contract. As we saw, valid contracts must rely on valid promises, but some promises—for example, slavery promises—are invalid if and because they attempt to transfer rights that cannot be so transferred, and this is so even if compliance with such promises could promote a just cause. Many would agree that slavery contracts would be invalid, even if and when a system of voluntary slavery could promote, say, overall welfare.24 Now, the same reasons why slavery promises fail to meet the alienability condition also explain why promises that attempt to alienate one's discretionary right to life fail to meet that same condition—the discretionary authority to decide whether one should live or die being perhaps the most fundamental aspect of moral agency, and definitely something we have an authority interest in controlling, since our ability to embark on meaningful projects and plan long term would be seriously diminished if we transferred to others the right to control that more fundamental decision. Therefore, either the discretionary right to life is inherently inalienable, for its alienability would contradict the grounding value of moral agency, or it cannot, in any case, be alienated through promising, for such alienation would contradict the reason for having the normative power to promise—the authority interest. Note that this does not mean that contracts for, say, voluntary euthanasia are necessarily invalid, for in the case of voluntary euthanasia, I do not transfer to someone else the discretionary authority to decide whether I should live or die, regardless of changes in my future will. To the extent that I can change my mind until the last minute, I am myself exercising that authority.25 By contrast, to (attempt to) transfer to another the right to decide whether I should live or die, regardless of potential changes in my future will, is to (attempt to) abandon that authority. It follows that, if the mercenary's offer is interpreted, on a par with the soldier's promise, as an attempt to alienate the mercenary's discretionary right to life, then, such an offer (unlike Fabre's manufacturer's offer) ought to be regarded as invalid—even if the state accepts it, no promissory duty to stay in the case of an imminent risk would arise. If the mercenary refuses to die, they would not be wronging the state. It could be objected, however, that people sign employment contracts to undertake very risky occupations all the time, and such contracts generally succeed in imposing binding obligations on the employees. Indeed, although mercenaries incur serious risks to their lives,26 professions such as logging or construction working are overall more risky. So why is the mercenary contract, but not, say, the logger contract, invalid? The reason is that the logger's offer does not attempt to alienate their right to life. What the logger agrees to—performing tasks to which some randomly distributed risk of death is attached—is not the same as transferring to another the authority to decide whether they should stay once a risk of death materializes. The logger does not wrong the employer if they leave as soon as they reasonably fear a tree is falling on their head, and this would be so even if they had contractually agreed to do whatever the employer decides. Not only, as Ned Dobos points out, would we “not think it morally justifiable for any civilian employer to demand and enforce obedience unto death”, but also employees, including those in very risky occupations, have a moral and a legal right to disobey an order if they believe that an “imminent danger” exists, where this is legally defined in terms of a reasonable expectation “of death or serious physical harm.”27 In the same way in which a logger's promise to do whatever the employer demands, including staying in case of imminent danger to the logger's life, would be invalid, so too should be the mercenary's promise to do whatever the commander demands in case of imminent risk of death. Later I will explain why soldiers constitute a sui generis exception to this conclusion. It could be objected, however, that in the case of mercenarism, and only in this case, obedience to sacrifice is necessary to fulfill the aim of the profession—an aim that is arguably both morally acceptable and sufficiently valuable.28 Perhaps this fact suffices to justify morally binding the mercenary to do as the commander says.29 The objection, however, encounters a deontological challenge: the expected benefits of alienation would seem insufficient to make alienable what is inalienable. Just as a system of slavery contracts would be invalid, even when necessary to achieve social stability or economic growth, a system of mercenary contracts that entails the alienation of the discretionary right to life would be similarly invalid, even when necessary to achieve better security. It would seem, then, that mercenary promises can be both valid and permissible only to the extent that mercenaries are treated as waiving, rather than alienating, their right to life. Just as, say, sex workers can (arguably) sell sexual services in the market, as long as they retain the right to walk out as soon as they change their mind, mercenaries can sell their combat services, assuming that they retain the right to leave when things get too dangerous. They may have a duty to return the money if they provide no service, but no duty to incur any sanctions for failing to perform.30 [L]eaving a PMF post is not desertion—punishable by prosecution and even death, but merely the breaking of a contract with limited enforceability … As compared to a conscript army, when they return home, contract employees likely face no sanctions for defection as do conscripted soldiers.31 While the fact that walkouts are not punished as desertion indicates that mercenaries cannot be compelled to remain against their will, the fact that they break a contract when they leave indicates that they are understood as having a duty to stay, albeit one with “limited enforceability.” Regardless of how existing law understands the content of the mercenary contract, morally speaking, the mercenary's offer for services can be valid only if it is limited to waiving.32 But this generates a new problem. If mercenary offers are valid only conditionally on private actors retaining the right to walk out in case of imminent danger, then, this fact would seem to compromise the moral permissibility of a state accepting such offers, since acceptance would endanger the success of military operations,33 thereby contravening the state's moral duty to effectively protect its citizens. private security contracts run the risk of walkouts, strikes and dropped contracts, but the consequences of these potential pitfalls are even greater during wartime and contingency operations… [C]ontractual breaches … have occurred regularly enough to warrant investigation.34 It could be argued, however, that even if mercenaries are less reliable than soldiers, they may bring other benefits that could outweigh the heightened risk of walkouts, such that accepting mercenary offers would be, all things considered, the best way for states to fulfil their duty to effectively protect their citizens and their own stability. Of course, when it comes to combat tasks on the ground, reduced reliability would seem to be a particularly weighty consideration, since if combatants walk out in the middle of a military operation, the success of the entire operation would be jeopardized, especially if those who walk out have, say, higher skills or better training. But in order to assess the weight of reduced reliability, relative to other considerations, we should be more specific as to what reasons states may have to prefer mercenaries to soldiers as means to discharge their duty to effectively provide security, compatibly with other duties they may have.35 One often cited reason is that, because of market competition, private companies can provide superior services at lower costs than the state. However, even if we assume, arguendo, that this functional consideration could compensate for reduced reliability, it may simply not apply in the case of most combat tasks, for providing security in conflict zones exhibits, and robustly so, all the features of market failure: scarce competition, a narrow consumer base, and consumers' inability to pass on economic losses.36 A second reason states may have to prefer private companies is “to acquire the services that they want when they want them, rather than having to maintain an ongoing regular military.”37 This is an important consideration, for security is not the only responsibility a state has, and a political society may reasonably opt to redirect military spending to other causes. However, this consideration holds only if we assume that privatizing military tasks will amount to long-term savings that could then be redirected elsewhere, something that is, again, highly questionable, once the costs involved in monitoring and administrating private contracts are factored in.38 Third, hiring private armies allows states to avoid politically unpopular choices such as committing additional soldiers to dangerous missions.39 But either a state should do what the people want and not send the troops, or, if it should pursue the aim of security, despite the people's will, then, other things being equal, it is impermissible to prefer private combatants just to avoid political costs, if there is a higher risk that such combatants will leave when things become too dangerous. Finally, normally functioning states that have the capacity, if not the willingness, to recruit soldiers cannot generally appeal to the scarcity of soldiers as a justification (as different from an excuse) for hiring mercenaries, as a means to discharge their duty to effectively provide security, for in such states the scarcity of soldiers is often a consequence of previous decisions to hire mercenaries instead.40 In sum, insofar as private combatants' reduced reliability cannot be easily outweighed by competing considerations, states with the capacity to maintain or recruit alternatives—which excludes very weak states—will generally have strong reasons, grounded on their moral duty to effectively protect their citizens and defend themselves, not to accept the mercenary's promise.41 But perhaps the assumption that mercenaries are, or can be expected to be, less reliable than soldiers just because they are not subject to threats of punishment is false. After all, economic incentives, such as refusal to renew a firm's contract in case of walkouts, or motives such as comradeship, could and often do suffice to secure the willingness of private combatants to sacrifice their lives.42 In response, we should first notice that incentives can be circumvented. For example, private firms can dissolve and form a different company in order to avoid a ban on bidding for further government contracts.43 As for motives, even if we assume that mercenaries can be motivated, precisely like soldiers, by comradeship and a sense of honor, beyond pecuniary considerations, the threat of punishment may still be necessary to keep cases of desertion under reasonable limits. Otherwise it could not be explained why soldiers, who are similarly motivated, are subject to such threat of punishment. But there is more. A state has a duty to secure not just reliability, as far as the provision of security is concerned, but robust reliability—reliability under a set of possible changes in human motivations. The reason is that a state should ensure that its citizens' security does not significantly depend on the good will of its providers44—in the same way as a state committed to the economic security of its citizens should not leave its provision to philanthropy. If mercenaries, however well intentioned, were allowed to discretionally judge the level of risk, and make decisions about whether to leave or stay on that basis, this would raise important concerns of domination. It is thus impermissible for states to exclusively rely on the comradeship of the mercenary, or on their discretional responsiveness to economic incentives, to ensure the successful performance of justified military operations. To this we can add that, if mercenaries could not be compelled to remain, states would likely have incentives to deceive them about the risks of certain wars in order to prevent desertion.45 This in turn could generate a lack of trust between mercenaries and state officials, as well as other soldiers, thereby further undermining the effectiveness of the military. Finally, the overall likelihood of desertion would impose higher costs on those remaining in the field of war.46 All these reasons, then, suggest that functioning states are under a duty, grounded on their prior moral duty to provide security in an effective, non-dominating and fair way, not to accept mercenary offers. We thus reach the following preliminary conclusion: either the mercenary's promissory offer of military services violates the alienability condition, in which case the promise is invalid, or, if limited to waiving, it is generally impermissible for states to accept the offer. It follows that, if the state does what it ought to do, the mercenary would acquire no promissory duty to perform. If, by contrast, the state (wrongly) accepts the mercenary's promise, it would still be wrong for the state to demand that the mercenary comply with the offer's terms. It should release the mercenary instead. In turn, the state should not consider itself wronged if the mercenary fails to comply, for this failure is only possible as a result of the state's previous wrongful failure to release the mercenary. Note that this argument does not apply to the provision of services, including some combat tasks such as drone operating, that entails no, or a very low risk of death for those providing them, or to services, however essential, for which replacements could be easily and effectively found, because in such cases leaving would not undermine security. It only applies to the provision of services that are both essential and can only be effectively supplied if those who provide them are willing to stay when a risk of death materializes—for example, combat on the ground by front-line infantry or, depending on circumstances, the operation of bombing campaigns. It may also apply to the provision of non-combat support services, as long as such services are essential for the success of a military mission, their provision in conflict zones entails a non-negligible risk of death, and replacements in cases of walkouts cannot be easily secured. But a part of the puzzle is still missing. If mercenaries cannot alienate their discretionary right to life, why can soldiers? Call this the asymmetry question. True, soldiers retain the option of choosing imprisonment as an alternative to making the ultimate sacrifice.47 However, the pair of options between which soldiers can freely choose is imprisonment versus (an imminent and likely risk of) death, as opposed to a free life versus (an imminent and likely risk of) death. Joining the army involves, in this sense, a qualified form of alienation. The literature on war suggests two answers to the asymmetry question, both of which are wanting.48 The first answer points to the distinctively financial cha