Compensating beneficiaries

Abstract

This paper illuminates a typically obscured ground for rectificatory obligations: harms justified as ‘lesser evils.’ Lesser-evil harms are not the result of overall morally prohibited acts but of acts permissibly carried out to prevent significantly greater harm. The paper argues that harms caused as unintended side effects of acting on lesser-evil justifications, notably in military rescue operations, may give rise to claims to compensation, even if (1) the military acts that caused the harms in question were justified on lesser-evil grounds and (2) the victims in question are no worse off as a result; they may even owe their survival to the act of rescue. The paper defends three claims. First, being better off as a result of a harmful rescue than one would otherwise have been does not preclude claims to be compensated for harms suffered as a side effect. Second, identifying the relevant counterfactual for purposes of compensatory justice is sometimes a prescriptive, rather than a descriptive, matter. Rather than relying on empirical speculations about what would have happened had a harm not occurred, we must, in certain cases, consider what agents ought to have done. Finally, duties of compensation need not fall on those who caused the to-be-compensated harms. That infringing rights is permissible in certain cases does not imply that no compensation is owed, but merely that it is not necessarily rights-infringers on whom duties of compensation fall.

1 Introduction

Even harms justified as ‘lesser evils’ can be significant, most notably in war. Besides physical harm, they may include severe psychological trauma as well as the widespread destruction of homes, businesses, and civilian infrastructure.¹ While much has been written about rectificatory duties resulting from wrongdoing, rectificatory duties that arise from harms permissibly inflicted on morally innocent people who are made overall better off have received only scant attention.² This paper illuminates this typically obscured ground for rectificatory obligations: harms justified as ‘lesser evils.’

Such harms pose at least two challenges to theories of compensation. First, the harms in question are, by stipulation, the result of acts permissibly carried out to prevent significantly greater harm. They are not the result of overall morally prohibited acts but, effectively by definition, all-things-considered justified. Second, not acting on a lesser-evil justification would effectively amount to not preventing a ‘greater evil.’ For example, in the absence of a military intervention, some victims of ‘collateral’ harms might have become victims of atrocity crimes. Being rescued may have overall benefited them.

The purpose of this paper is to explain why compensation may nonetheless be owed under these circumstances. In particular, I argue that harms caused as unintended side effects of acting on lesser-evil justifications, notably in military rescue operations, may give rise to claims to compensation, even if (1) the military acts that caused the harms in question were justified on lesser-evil grounds and (2) the victims in question are no worse off as a result; they may even owe their survival to the intervention.

My argument is threefold. First, the assumption that certain harmful acts were justified on lesser-evil grounds does not imply that no compensation is owed. Rather, in the context of military intervention, it entails that it is not necessarily interveners themselves on whom compensatory burdens fall. Second, the fact that victims may be better off as a result of a harmful rescue than they would have been in its absence does not preclude the possibility of claims to be compensated for harms suffered as a side effect.³ Finally, in cases in which innocent people are harmed on lesser-evil grounds, ordinary counterfactual reasoning is inadequate; it only tells us that a ‘greater evil’ would have occurred. While ordinary counterfactualism relies on empirical speculation about what would have happened had a harmful act not been performed, deontic counterfactualism is concerned with what agents ought to have done. The latter—more effectively than the former—helps illuminate who owes what to whom in cases in which innocent people are permissibly harmed as a ‘lesser evil.’

The issue this paper tackles has a claim on anyone’s attention who believes in lesser-evil justifications. This includes, but is by no means limited to, scholars working on the ethics of war and justifications for defensive harming. Although the main example here is military rescue operations, the claim that even all-things-considered justified and beneficial acts may generate claims to compensation may also extend to other contexts in which lesser-evil justifications for harming may call into question duties of compensation. One reason for focusing on humanitarian intervention is that such wars, in contrast with wars of self-defence, are sometimes thought to permit the shifting of harms away from intervening combatants and towards the intended and expected beneficiaries of such wars.⁴ Proponents of this view may be especially sceptical of the claim that harms justifiably inflicted on lesser-evil grounds in these cases generate compensatory obligations. More broadly, the issues this paper raises are not limited to the particular domain of military rescue operations, but help to advance the broader debate on compensation for permissible rights infringements. Finally, deontic counterfactualism may be of interest to arguments that employ counterfactuals more generally.⁵

A few clarifications before we proceed. My primary concern is with the grounds of rectificatory duties of compensation—more precisely, one moral mechanism by which duties of compensation arise. Part of my claim is that victims of permissible rights infringements are, under certain conditions, owed compensation. My aim is to explain why this can be the case even if the victims in question are incontrovertibly better off than they would have been in the absence of the act that caused the harm to be compensated, and even if those who caused this harm acted all-things-considered permissibly. In cases in which harming morally innocent persons is impermissible and in cases in which innocent victims are made worse off than they otherwise would have been, there is little doubt that compensatory obligations arise. By contrast, my aim is to defend the less intuitive position that victims may acquire claims to compensation even if the harmful acts in question were all-things-considered permissible and made victims better off than they would have been otherwise.

My aim is not to develop a full account of compensation. For example, I will not discuss to what extent the duty to compensate victims is enforceable, or what principles constrain the permissibility of imposing compensatory burdens. Liability to compensation is constrained by various forms of proportionality assessments: for example, compensatory burdens imposed on compensators or third parties must not be excessive in relation to compensatory benefits. It is not my aim here to examine how this works. Nor is my aim to establish what adequate compensation amounts to—how it is to be measured, or what the relevant currency is. I also take no stance on broader questions of how to determine under what conditions compensation is the most appropriate rectificatory measure; what form of compensation may be appropriate; or the degree to which different harms are compensable in the first place.⁶

Finally, I will largely eschew the heavy-handed label of ‘collateral’ harm. For clarity, I will focus on hypothetical cases in which harms to innocent people are, by stipulation, justified on lesser-evil grounds. Lesser-evil justifications essentially permit harming people who are not liable to be harmed—that is, people who possess rights against being harmed—on the grounds that this will prevent some significantly greater harm. Lesser-evil justifications are not wholly consequentialist; for example, they are sensitive to the moral asymmetry between doing and allowing harm, and the difference between merely foreseeable and opportunistic harming.⁷ I will not discuss what conditions must obtain to generate a lesser-evil justification in the first place. For brevity, I will sometimes say ‘lesser-evil harms’ to mean ‘harms justifiably inflicted on non-liable people on lesser-evil grounds.’

2 Permissible rights infringements

One view is that if inflicting a harm is all-things-considered permissible, then no compensation is owed. As John Oberdiek says about Joel Feinberg’s case of the hiker who breaks into a cabin to escape a storm, ‘[i]f the destruction is justified, and therefore no right violated, no compensation is due on moral grounds.’⁸ One response is that compensation may be owed not merely for wrongful rights violations but also for justified rights infringements. The rights infringement/violation distinction has been discussed at length by others, and I shall not defend it here.⁹ Rather, my first aim is to advance the twofold premise that acting all-things-considered permissibly is compatible with wronging people and that these wrongs may generate compensatory obligations.

To say that X is all-things-considered justified in harming Y on lesser-evil grounds is compatible with saying that Y is nonetheless wronged. If Y has neither forfeited nor waived their right not to be harmed, or consented to being harmed, they are wronged by having their right transgressed, notwithstanding the fact that transgressing their right is all-things-considered permissible. In contrast to liability-based justifications for defensive harming, which presuppose that the target of the defensive harm lacks a right not to be harmed and so is not wronged by being thus harmed, lesser-evil justifications presuppose that reasons for harming outweigh the ordinary constraint against harming. Since victims of lesser-evil harms neither necessarily forfeit nor waive their right not to be harmed, there is no inconsistency in saying that, even if agents have a lesser-evil justification for harming, victims of lesser-evil harms may be wronged.¹⁰

Part of the explanation for why victims of lesser-evil harms may acquire claims to compensation is that, unlike in the case of liability, what justifies harming people who possess rights against being harmed—the moral value of averting a substantially greater harm—is wholly external to them. When preventing this greater harm would demand a sacrifice significantly exceeding what can reasonably be expected of non-liable people, forcing them to bear these burdens generates compensatory obligations. Since victims of lesser-evil harms typically have no duty to suffer the harms inflicted on them, they are owed compensation in part for being made to bear a cost they had a right not to bear.¹¹ This goes some way towards explaining why compensation may be owed for permissible rights infringements; not merely for impermissible rights violations.

Oberdiek holds that compensation should be owed only for rights violations because, by demanding compensation for rights infringements, we leave nothing more that could be demanded by rights violations.¹² But insofar as violations of rights are concomitant with all-things-considered wrongdoing, whereas infringements are not, rectificatory obligations generated by violations differ from those generated by infringements. What needs to be addressed in the case of rights violations includes the agent’s overall wrongful act. Rights violations thus require a response that is, to a significant degree, agent-focused; for example in the form of punishing wrongdoers and demanding an apology, which presupposes a certain degree of moral responsibility. Responses to rights infringements, on the other hand, are largely victim-focused and may involve non-punitive remedial obligations to mitigate the harm suffered.

While wrongful rights violations generate rectificatory burdens for the wrongdoer, compensatory obligations for permissible rights infringements in cases in which justified rights transgressors acted on a lesser-evil justification, may, in principle, fall on third parties. The fact that infringing rights is all-things-considered permissible in certain cases, then, does not imply that no compensation is owed, but merely that it is not necessarily rights-infringers themselves on whom duties of compensation fall. Section 4 returns to the claim that duties to compensate victims of lesser-evil harms do not necessarily fall on those acting on lesser-evil justifications.

We can distinguish between two different types of permissible rights transgressions: those justified by agent-neutral, lesser-evil considerations and cases in which the agent’s own interest is what grounds a permission to transgress another person’s right, such as Feinberg’s hiker who breaks into the cabin to shelter from the storm. In the latter type of case, in which permissibly transgressing others’ rights serves one’s own interests, compensatory burdens rest with rights-infringing agents. In cases in which agents act in the interests of others, and it is these others’ interests that ground the lesser-evil justification, agents may be exempt from liability to pay compensation, and compensatory burdens may instead fall on third parties, on which more later. On whom compensatory burdens fall may thus, at least in part, depend on whose interests are served by a permissible rights transgression.

Views about liability to compensate for lesser-evil harms plausibly vary depending on whether we think acting on lesser-evil justifications is required or optional. If acting on lesser-evil justifications is generally required, one view is that this exempts one from liability to compensate for lesser-evil harms.¹³ If the choice is effectively forced by morality, one’s compensatory liability is plausibly weaker than if one had freely chosen to inflict a harm which one could have permissibly refrained from inflicting. At the same time, it seems that, if morality effectively requires infringing people’s rights, morality must also require that they be compensated. On this view, it is coherent for morality to require us to harm people who have rights not to be harmed only if it also requires that they be compensated. One possibility is that those responsible for the ‘greater evil’ whose prevention generates a lesser-evil justification incur the primary liability to compensate. So, even if morality requires agents to act on lesser-evil justifications for harming, who are therefore exempt from compensatory liability, this does not mean that no compensation is owed; it just means that the burden falls on others.

Now consider the possibility that acting on lesser-evil justifications is optional. There are at least two different reasons why acting on a lesser-evil justification may be permissible rather than required: (a) because it is supererogatory, or (b) because, as Kerah Gordon-Solmon and Theron Pummer have recently argued, the two options—acting on the justification and refraining from doing so—roughly balance.¹⁴ Without taking a stance on these views here, consider what they imply for compensatory liability.

On the one hand, if there is no duty to act, the case for liability to compensate seems stronger, given that agents could permissibly have refrained from inflicting an all-things-considered permissible harm. On the other hand, in supererogation cases, if agents already bore greater costs than they were required to bear in acting on a lesser-evil justification, this may weaken the case for compensatory liability. Detailed discussion of whether agents may incur duties to compensate harms caused as a side effect of supererogatory rescues would exceed the scope of this paper. But, in general, there seems to me no inconsistency in saying that, in certain cases, (a) it is permissible not to rescue and (b) one may nonetheless incur duties to compensate victims for harms caused as a side effect of performing the rescue in supererogatory fashion.¹⁵ In parity cases, moreover, the way to decide between options, at least according to Ruth Chang, is to actively commit to one of them.¹⁶ Incurring compensatory burdens might then be one way of taking responsibility for our choices as morally responsible agents.¹⁷

A further complication, in the case of humanitarian intervention, arises from the possibility that it may be permissible for interveners, at least to a certain degree, to reduce risks to themselves at the cost of increasing risks to innocent civilians. If this assumption is correct, much will hang on whether victims would have survived the intervention if it had been carried out in a different way. The case for compensation may thus be weaker if interveners did everything in their power to minimise harms to civilians, and stronger if interveners, albeit permissibly, reduced risks of harm to themselves at the cost of increasing risks of harm to civilians.¹⁸ Besides complexities concerning the just distribution of risks of harm between rescuers and rescuees, a fuller account of compensation for lesser-evil harms will need to be sensitive to backward-looking considerations concerning responsibility for the greater threat of harm generating a lesser-evil justification in the first place, on which more shortly.

3 The benefit objection

That victims may acquire claims to compensation for being harmed, even if agents acted all-things considered permissibly, is only part of the story. Sometimes, victims of permissible rights infringements would have been worse off otherwise: they are better off than they would be if the act of rescue that harmed them as a side effect had not been performed. How, then, can people acquire claims to be compensated for harmful side effects of acts that were not only justified, but, in extreme cases like humanitarian intervention, may have actually saved their lives? If people effectively owe their survival to an act of rescue, how could they be owed compensation for being harmed as a side effect?

One possible view is that people whose lives were saved by being rescued are not owed compensation for harms they suffered as a side effect, because being rescued did not make them overall worse off than they would have been otherwise. This view is an instance of what we might call the

Benefit objection: If victims of lesser-evil harms would have become victims of worse harms if rescuers had not acted on their lesser-evil justification, victims do not acquire claims to be compensated for lesser-evil harms.

According to the Benefit Objection, people are not owed compensation for harms suffered as a side effect of an act of rescue that manifestly benefited them. Correspondingly, in the context of humanitarian intervention, if victims of lesser-evil harms would have become victims of egregious rights violations in the absence of an intervention, they do not acquire claims to be compensated for harms suffered as a side effect of their rescue.

The Benefit Objection rests on the common assumption that the purpose of compensation is to restore victims to the same level of wellbeing, or to a position equivalent to that in which they would have been, had they not suffered the harm in question.¹⁹ Common conceptions of compensation rely on a counterfactual assumption in which the harm was not suffered. As Robert Nozick put it:

Something fully compensates a person for a loss if and only if it makes him no worse off than he otherwise would have been; it compensates person X for person Y’s action A if X is no worse off receiving it, Y having done A, than X would have been without receiving it if Y had not done A.²⁰

The aim of compensatory justice is to make victims as well off as they would have been had a harmful act or event not occurred. The assumption is that, by considering a counterfactual in which the harm did not occur, we can establish the degree to which a victim has been harmed and is consequently owed compensation. We essentially compare a victim’s level of wellbeing resulting from the harmful act or event to be compensated to her level of wellbeing in a counterfactual state of affairs without that act or event. Accordingly, Agent fully compensates Victim for a harm Victim suffered if Agent renders Victim as well off as Victim would have been had the harm not occurred.²¹

Standard accounts of compensation typically assume a counterfactual-comparative conception of harm, according to which one has been harmed by some action if, and only if, one is made worse off than one would have been otherwise.²² The main alternative to the counterfactual account is the causal account of harm, according to which one has been harmed if and only if an action causes a harm to one.²³ A third, less popular, alternative is the temporal account, according to which one has been harmed by some action if, and only if, one is worse off after that action than one was before that action.²⁴ Like most people, I eschew a purely temporal account of harm but will, for the purposes of this discussion, remain neutral between counterfactual and causal accounts.²⁵ So, while I will consider the counterfactual element of standard views of compensation, I will do so without committing to a particular account of harm.

Accounts of harm and accounts of compensation are not the same.²⁶ In particular, counterfactual accounts of compensation may look like they include distinctly temporal elements, without presupposing a temporal account of harm. For example, standard views of compensation are often described as taking the aim of compensatory justice to be that of reversing some deviation from a particular status quo ante; because the purpose of compensation is to bring people back to where they were before they suffered the harm for which they are to be compensated.²⁷ To say that compensation serves to return victims to a position equivalent to that in which they were before suffering the harm to be compensated is different from saying that compensation serves to return victims to a position equivalent to that in which they would be now—or, for what matter, would have been at some relevant point in time—had the to-be-compensated harm not occurred. But neither formulation necessarily presupposes a temporal account of harm.²⁸

Many harms and benefits do not aggregate to yield a totality of wellbeing that could serve as a basis for determining whether a person has been harmed.²⁹ Being overall benefited, for example, is compatible with being wronged. The doctor who administers the life-saving treatment against her patient’s will may wrong her patient by transgressing his right to bodily integrity, even if the patient is benefited by having his life saved. Insofar as being benefited is compatible with having one’s rights transgressed, rights transgressions may generate rectificatory obligations even if victims are not harmed, and potentially even if they are benefited.³⁰ Having been made overall worse off is not a necessary condition for claims to compensation to arise.

Causal overdetermination cases pose a special challenge to standard views of compensation. Harms are overdetermined in cases in which some act or event (AE1) causes a harm, and that harm would also have been caused by some other act or event (AE2), regardless of whether the former act or event (AE1) was performed or occurred. In some cases, victims are overall better off as a result of suffering some harm than they would have been if the harm in question had not been caused. In cases in which victims of lesser-evil harms would have been worse off in the absence of a harmful act of rescue, harms to victims are, in this sense, overdetermined.³¹ For example, while victims might suffer certain harms as a side effect of a military operation—such as severe bodily injuries and psychological trauma—the harms they would have suffered in the absence of an intervention – even worse injuries and worse trauma, perhaps even death – would have exceeded the former.

The standard view of compensation is inadequate in overdetermination cases.³² Consider Jules Coleman’s classic example, in which.

I am scheduled to take a plane from New Haven to Washington. Five blocks from the airport, the taxi hits another car. My leg is broken; I’m taken to the hospital; I miss my flight. The plane I would have taken crashes. There are no survivors. Had I caught the plane, I would have died. Only the taxi driver’s recklessness keeps me alive.³³

In this case, harm is overdetermined, in the sense that I would have been harmed either way. Presumably, I am nonetheless owed compensation for my broken leg, even though I am better off having been injured in the car crash than I would have been otherwise.³⁴ The fact that the driver’s recklessness effectively saved my life does not affect his compensatory duties. Ordinary counterfactualism is ill-equipped to accommodate this notion. Since the aim of compensatory justice is to make victims no worse off than they would have been in the absence of a harmful act, we compare what did happen to what otherwise would have happened. But comparing the harm of a broken leg to the harm of dying in a plane crash will not do. That I am better off as a result of the driver’s recklessness than I would have been had I made it onto the plane is irrelevant for whether the driver owes me compensation for his reckless driving.

Causal overdetermination cases demand a conception of counterfactual compensation that is more complex than the conventional approach.³⁵ Complex counterfactual accounts do not limit themselves to comparing the harmful outcome of an act to the state of affairs that would have obtained if this harmful act had not been performed. They instead seek to accommodate the possibility that compensation may be owed for harming a person, even if one thereby pre-empts a greater harm which the victim would otherwise have suffered. But cases in which victims of lesser-evil harms would, in the absence of being rescued, have become victims of wrongful harm pose a special challenge.

4 The problem of unjust counterfactuals

Mainstream counterfactual approaches seek to determine what would likely have been the case, and how victims would likely have fared, had a particular harmful act or event not occurred. Claims to compensation, recall, are typically determined by reference to victims’ level of wellbeing in a counterfactual world in which they did not suffer the harm for which they are to be compensated. Conventional approaches to compensatory justice essentially adopt a three-step process.³⁶ (1) They presume a state of affairs in which the protagonists are in a state of ‘moral equilibrium’; in which the moral universe is, in relevant respects, in balance, and no rectificatory obligations obtain yet. (2) An agent causes a wrongful harm which upsets the moral equilibrium. (3) As a consequence, rectificatory obligations arise, to set the moral equilibrium right.³⁷

The problem is that, in cases in which conventional counterfactual reasoning yields a counterfactual that is itself characterised by injustice, conventional counterfactual reasoning is inadequate at assessing claims to compensation. How victims of lesser-evil harms would have fared in the absence of those acts that caused those harms is not an appropriate counterfactual. After all, in this counterfactual, the ‘greater evil’ whose prevention generates a lesser-evil justification for harming prevails. For example, in a case in which all requisite jus ad bellum conditions are met and the international community faces a choice between intervening to halt egregious human rights violations or standing by as atrocities are committed, non-intervention describes a state of affairs in which atrocities are not prevented. Insofar as the state of affairs preceding a rescue is itself unjust, lesser-evil harms caused during the rescue do not themselves represent a deviation from a just status quo. Standard counterfactual reasoning, this means, does not tell us what we need to know to assess claims to compensation for lesser-evil harms.

Lesser-evil harms occur within a context that is practically the reverse of that standardly assumed by mainstream accounts of compensatory justice. While mainstream accounts presume a for-all-intents-and-purposes-legitimate state of affairs and assume that the moral equilibrium is offset by an unjust harm which is to be set right through compensation, lesser-evil justifications are sometimes precipitated by an unjust state of affairs in which the moral equilibrium is already out of order. In cases in which the empirical counterfactual situation itself is unjust, it cannot serve as a morally relevant baseline.³⁸ When we assess claims to compensation for harms suffered as a side effect of an intervention that saved victims’ lives, the counterfactual of ‘what if the harmful lesser-evil action had not been performed?’ is out of place.

Certain rescue cases thus differ from typical contexts of compensation not only in that the harms in question are inflicted all-things-considered permissibly—as side effects of acts justified on lesser-evil grounds—but also in that the counterfactual in which rescuers do not act on their lesser-evil justification for harming is itself unjust. In cases in which harms to innocent people are justified on lesser-evil grounds to prevent a ‘greater evil’, we must look beyond ordinary counterfactual reasoning.

5 Deontic counterfactualism

We can distinguish between two different approaches to identifying counterfactuals: empirical and deontic. In seeking to specify a state of affairs that would have obtained had the to-be-compensated harm not been caused, mainstream approaches employ empirical counterfactuals, to bring about a state of affairs as similar as possible to that which would have obtained had a particular harmful act or event not occurred.

Deontic counterfactuals, by contrast, take as their main point of reference agents’ compliance with moral duties. The main question, on this approach, is not what would likely have happened if a harmful act or event had not occurred, but to what degree agents complied with their moral duties; in particular, what duties agents failed to discharge, as a result of which victims are now worse off. So, rather than seeking to determine what would have been the case, deontic counterfactuals focus on what agents ought, or ought not, to have done. The practical aim of compensation remains the mitigation of harm that people had rights not to suffer, but identifying claims to and duties of compensation requires a wider picture of the normative landscape than empirical counterfactuals capture. While empirical counterfactualism is limited to speculation about what would have happened, deontic counterfactualism looks to what agents ought or ought not to have done.

On a deontic conception of counterfactuals, for example, it is clear why Coleman’s taxi driver owes me compensation although his recklessness by chance also saved my life. Insofar as he failed in a duty to drive conscientiously, he owes me compensation for his recklessness, irrespective of the fact that that ultimately happened to make me better off.

Here is one way of thinking about this. Consider a common way of distinguishing between ideal theory and non-ideal theory: ideal theory is concerned with what is required of us when everyone else is doing what is required of them, and non-ideal theory is concerned with what is required of us when others are not doing what is required of them.³⁹ Just like we can distinguish between compliance and non-compliance at the macro level, deontic counterfactualism presupposes that we can identify instances of non-compliance and what is required by way of rectification at the micro level. Deontic counterfactualism seizes upon familiar dimensions of theories of justice, and foregrounds the question of what morality requires, in response to specific instances of non-compliance in the past. Deontic counterfactuals do no more—and no less—than ground the existence of compensatory duties in junctures of injustice in a harm’s causal history.

Deontic counterfactualism, in this rudimentary sketch, is not a distributive principle.⁴⁰ Its purpose is not to tell us how much people should be compensated, or whose claims to compensation to prioritise. Its function is to capture more of the moral landscape than empirical counterfactualism captures. More detailed discussion would exceed the scope of this paper, but a fuller account of deontic counterfactualism might explain how agents’ compensatory liability to mitigate harms is proportionate to the degree of their moral responsibility for and causal contribution to that harm.⁴¹

Against the moral backdrop of the greater evil whose prevention generated a lesser-evil justification for harming in the first place, deontic counterfactualism helps us adequately respond to the lesser-evil harm to be compensated. This also means that Oberdiek’s worry, mentioned earlier, need not arise. Oberdiek, recall, argues that compensation can be owed only for rights violations because, if it were also owed for rights infringements, this would leave nothing more that could be demanded by rights violations. By focusing on compliance, deontic counterfactualism straightforwardly allows us to distinguish between permissible and impermissible rights transgressions, and to respond to them differently.

In addition, deontic counterfactualism has several advantages over empirical counterfactualism. For one, if empirical counterfactualism yields a state of affairs that diverges from what agents ought to have done, it will be normatively uninformative. A distinct advantage of deontic counterfactualism is that it brings into alignment the wrongful harms victims suffered with agents’ failures to discharge their duties.

In addition, deontic counterfactualism accounts for cases in which there exist a range of just counterfactuals. As Daniel Butt puts the point in the context of claims relating to colonialism, sometimes ‘there is more than one just counterfactual to an act of injustice than the simple non-performance of the act in question. Instead, the act might be performed in a different, just way.’⁴² When we consider colonial injustice, for example, the appropriate question is not necessarily whether current generations would be better off had there been no interaction at all between colonial powers and their colonies, but whether current generations would be better off if the relations between the involved parties had been consensual and non-exploitative.⁴³ Designating just interaction rather than non-interaction the relevant counterfactual to unjust interaction shows deontic counterfactualism at work.

Deontic counterfactuals concern what agents ought to have done, rather than what circumstances would have obtained if a particular feature of a situation had been different. If agents violated a negative duty, then the relevant counterfactual is the non-performance of the act in question. In the case of colonial injustice, since the problem was not interaction per se, but unjust interaction, the relevant counterfactual is not a lack of interaction but just interaction; and this might take different forms.

Another advantage of the deontic counterfactual approach is that it can deal with both wrongful instances of doing harm and wrongful instances of allowing harm. Since standard accounts of compensation identify the morally relevant counterfactual as a state of affairs in which a harmful act or event did not occur, they do not have much to tell us if the relevant harm results not from an act done but from a harm one wrongfully failed to prevent. Empirical counterfactualism, recall, identifies the harmful effects of a wrongful act by comparing what actually happened to a possible world in which the harmful act did not occur. As Stephen Kershnar puts it, ‘we determine the conditions in the relevant possible world by assuming that the conditions in it are identical with those in the actual world up until the time of the injury, and then envisioning the most probable outcome if the injuring act had not occurred.’⁴⁴ But, if the ‘injury’ results not from positive agency, but from a wrongful failure to prevent a harm, then ‘envisioning the most probable outcome if the injuring act had not occurred’ requires that we employ deontic counterfactualism, and consider a state of affairs in which agents conformed with their duty to prevent wrongful harm.⁴⁵

One upshot of deontic counterfactualism is that it focuses our attention on the source of lesser-evil justifications that permit harming innocent people. For example, insofar as past failures to prevent threats of wrongful harm may culminate in the threat of harm whose prevention lends normative force to the justification for harming innocent people as a side effect, humanitarian interventions are precipitated by the failure to protect people’s fundamental rights. In these cases, there may therefore exist a direct moral relationship between ex post duties to compensate even for harms caused to innocent people as a side effect of just interventions and states’ ex ante duty to protect people’s fundamental rights, both within and beyond their own borders.

Consider, for example, the core assumption underlying the notion of the ‘responsibility to protect.’⁴⁶ A central idea is that the duty to protect fundamental human rights is universal, such that it is possible to incur a duty to take up the slack when others fail in their obligations.⁴⁷ The thought is essentially that, if country X fails in its responsibility to protect its residents’ human rights, other countries may incur a responsibility to pick up the slack and to protect country X’s residents, possibly from country X’s own government. Arguably, this also means that states may incur rectificatory duties for failing to prevent the humanitarian crisis that justified military intervention to begin with.⁴⁸ Compensatory obligations for lesser-evil harms may thus arise from failures to prevent, mitigate, or address wrongful threats of harm to innocent people at various stages leading up to an intervention. Relevant failures might include states’ failures to take measures to prevent severe human rights violations within and beyond their own borders; the failure to intervene when doing so was morally required; and, where applicable, the failure to distribute risks adequately in bello, where an intervention was carried out but risks of harm were distributed unjustly.⁴⁹

Given the possibility of deontic counterfactualism, matters are not as simple as observing that victims would be worse off if an intervention had not occurred. Whether victims of the intervened party are better off post bellum than they would have been in the absence of the intervention is not the only—in fact, not even the primary—question when it comes to assessing whether compensation is owed.⁵⁰ Other relevant considerations include, first, that victims would have been better off if the international community, potentially including interveners, had prevented the intervened party’s government from perpetrating or condoning atrocity crimes; and insofar as risks of harm could have been distributed more justly in bello, whether victims of the intervened party would be better off post bellum if interveners had distributed risks of harm differently in bello, to further minimise risks of ‘collateral’ harm.

Deontic counterfactualism thus gives us a considerably fuller moral picture than empirical counterfactualism. This helps explain why compensatory obligations for lesser-evil harms may arise even in cases in which victims of those harms are better off than they would otherwise be—including in cases in which victims owe their survival to the acts that caused these harms. The fact that victims of lesser-evil harms would be worse off if others had not acted on their lesser-evil justification does not preclude victims’ claim to be compensated. Indeed, as far as deontic counterfactualism is concerned, the assumption that victims of lesser-evil harms are overall better off as a result of a harmful rescue never enters the picture to begin with.

6 Who should compensate for lesser evils?

That certain military acts were justified on lesser-evil grounds does not imply that no compensation is owed. It rather entails that it is not necessarily those who inflict these harms on whom compensatory burdens fall. Feasibility constraints aside, those responsible for the ‘greater evil’ whose prevention generates a lesser-evil justification bear the primary duty to compensate the victims of lesser-evil harms. According to a familiar principle, those who wrongfully harm others incur more stringent duties than others to compensate those whom they have harmed.⁵¹ This is also reflected in how we think about the just distribution of defensive harm: if harming has become unavoidable, those morally responsible should bear the costs.⁵² Since negative duties not to harm are, generally speaking, more stringent than positive duties to prevent harm, rectificatory duties of those who cause wrongful harms are correspondingly heftier than rectificatory duties of those who fail to prevent wrongful harms. If lesser-evil justifications for harming are grounded in the prevention of a worse harm, those responsible for the ‘greater evil’ bear the primary rectificatory responsibility for the harms caused as a ‘lesser evil’ necessary to prevent the greater one. So, duties to compensate non-liable people for harms justified on lesser-evil grounds fall primarily on those responsible for the threat of that greater wrongful harm the averting of which generated the lesser-evil justification to begin with.

In many cases, however, primary duty-bearers cannot be made to bear the costs of compensation. They may have died or lack the means to compensate victims. In cases in which primary duty-bearers are unavailable, Saba Bazargan-Forward has suggested, compensatory obligations may fall on those who caused and benefited from the harms in question.⁵³ In his words, ‘those who authorize or commit the infringements and who also benefit from those harms will bear that compensatory duty if the unjust aggressor cannot or will not discharge it.’⁵⁴ The assumption is that, in some cases in which primary duty-bearers are not available, ‘the duty would fall on the party who both committed the rights violation and benefited from doing so.’⁵⁵

Bazargan-Forward imagines a trolley scenario, in which a significant number of innocent people are trapped on one set of the tracks and one innocent individual is trapped on another set of tracks. A bystander flips the switch, causing the trolley to injure the one innocent individual to save the greater number. In this case, Bazargan-Forward says, it is not the bystander but the beneficiaries who owe the innocent victim compensation. This, as he puts it, is because ‘[t]heir status as beneficiaries who authorized the rights infringement suffices to impose upon them duties of compensation toward [the innocent victim].’⁵⁶ Thus, beneficiaries may incur compensatory duties for authorising others to commit rights infringements in their (beneficiaries’) defence.⁵⁷

Bazargan-Forward’s point may apply to cases of national self-defence, in which a just defending state both commits rights infringements and benefits from them. But the case for imposing the costs of compensation on rights-infringing beneficiaries is less straightforwardly applicable to rescue cases, such as humanitarian interventions, in which the moral constellations are distinct from those of just wars of national self-defence. In contexts of humanitarian intervention, the intervened party’s own government may bear the primary responsibility for failing to protect its vulnerable populations from threats. And while those who commit the harmful rights infringements in question are just interveners who primarily intend to benefit not themselves but those to be rescued, interveners themselves may be beneficiaries of force protection, which might in turn be a cause for the rights infringements concomitant with ‘collateral’ harms. Besides, some of the rights permissibly infringed by interveners belong to intended beneficiaries of the rescue operation. Unless those committing rights infringements are identical with those who stand to benefit from them, Bazargan-Forward’s argument for imposing compensatory burdens on rights-infringing beneficiaries may not straightforwardly extend to rescue contexts like humanitarian intervention.

Ordinarily, to be sure, it would seem appropriate to focus on those who benefit from rights infringements as potential compensatory duty-bearers.⁵⁸ But some rescue cases raise special challenges. The intended beneficiaries of humanitarian interventions are, effectively by definition, victims of wrongdoing. Since their benefit consists in not suffering harms which they have a right not to suffer, this benefit is not morally salient.⁵⁹ Having ‘benefited’ in the sense of not suffering harms which one has a right not to suffer does not justify the imposition of compensatory obligations. And, insofar as the benefit of not suffering wrongful harms is not morally salient, beneficiaries of the prevention of wrongful harms are not appropriate compensatory duty-bearers.

Claims to compensation, as noted above, are held primarily against those who violate rights—not against those acting on lesser-evil justifications. If those responsible for the unjust harms that necessitated the intervention cannot be made to bear the costs of compensation, the costs of post-intervention reconstruction and compensation for harms and damage justifiably caused by interveners may fall on the international community in general.⁶⁰

As Bazargan-Forward points out, in cases in which victims are especially badly off, there may be a duty of beneficence, which is distinct from the duty to compensate, to pick up the slack when others fail to discharge their compensatory duties. Appealing to such a ‘Samaritan duty to help anyone who is in dire need through no fault of her own’ seems to me perfectly plausible in such cases.⁶¹ Besides, members of the international community might incur remedial obligations as a result of having failed to prevent wrongful harms ex ante. Failures to prevent wrongful harms, which precipitated a humanitarian crisis necessitating intervention, may ground post bellum duties to address lesser-evil harms caused in bello, because these harms would not have occurred if the intervention had not been necessary to stop grave injustice. In this sense, the failure to prevent the ‘greater evil’, which cannot be prevented without causing lesser-evil harms, itself constitutes a reason why the international community incurs compensatory obligations. More generally, even states that lacked an obligation to carry out an intervention may incur duties to assist with the demands of the jus post bellum in the aftermath if this is the fairest way of distributing remedial burdens.⁶² If the protection of human rights is everybody’s business, so is sharing the costs of enforcing this duty and the resulting rectificatory burdens of failures to do so.⁶³

7 Conclusion

This discussion cast doubt on the assumption that compensatory obligations arise only if people are made overall worse off as a result of a harmful act than they would be if the harmful act had not been performed. Even if one owes one’s survival to a harmful act of rescue, this does not automatically preclude one’s claim to compensation. Hence, bringing about a just post-intervention state of affairs and creating the conditions for sustainable peace requires a whole lot more than accounting for rights violations. While enforcing human rights protection may, in certain cases, justify a peremptory and resolute response to atrocities, this comes with weighty obligations to redress the harms and wrongs which even justified humanitarian interventions may entail. Compensatory obligations are not owed merely to victims of wrongful rights violations but, in extreme cases, may also be owed to survivors of ‘collateral’ harms whose lives might have been saved by an intervention.

Whether people acquire claims to compensation depends primarily on whether they were made to suffer harms which they had a right not to suffer. By determining relevant counterfactuals in relation to moral duties rather than empirical speculations, deontic counterfactualism grounds compensatory obligations for lesser-evil harms in previous failures to prevent those wrongful harms whose prevention generates lesser-evil justifications for harming in the first place. The possibility that victims of lesser-evil harms would have been significantly worse off if agents had not acted on the lesser-evil justification poses no obstacle to compensatory obligations.

This paper argued that even acts that are all-things-considered justified and beneficial may generate claims to compensation, because (1) having been benefited is compatible with being owed compensation, (2) identifying the relevant counterfactual is sometimes a prescriptive, rather than a descriptive, matter, and (3) compensatory burdens need not fall on those who caused all-things-considered permissible harms. These duties of compensation fall on those responsible for the unjust harms whose prevention generated a lesser-evil justification and, if wrongdoers cannot be made to pay, on the international community for failing to prevent wrongful harms, which precipitated the crisis necessitating intervention.

What ultimately matters is how prepared we are to defend the significance of individual rights in the face of other, typically consequentialist considerations. In addition to addressing the wrongs of rights violations, theories of rectificatory justice will need to accommodate the moral consequences of permissible rights infringements in the form of duties to mitigate harms justifiably caused to innocent people on lesser-evil grounds.