Compensation for Historic Injustice: Does it Matter how the Victims Respond?

Abstract

When states are required to compensate victim groups for the historic wrongs they have committed, how should the compensation due be calculated? It seems that alongside the counterfactual world in which the wrongdoing never occurred, we should also consider the counterfactual world in which the wrongdoing has occurred, but the victims have responded to it in a prudent way. Under tort law, the damages a victim can claim are reduced if they are judged to have been contributorily negligent, thereby exacerbating the harm they have suffered. The paper considers four reasons why victim groups could not be expected to respond collectively to injustice in a way that outsiders might judge to be prudent. 1. The wrongdoing has destroyed their capacity to act as a group agent. 2. The wrongdoing has inflicted psychological traumas that lead them to make poor decisions. 3. They would need to violate a moral norm that is deeply embedded in their culture. 4. They would need to adopt policies that risk inflicting severe injustice on an internal minority. In many instances of historic injustice, one or more of these reasons apply. So although in principle victims’ behaviour is relevant when compensation is being calculated, we must be very cautious in using that doctrine to assess real cases.

I

It is widely accepted that political communities owe duties of compensation to groups whom they have wronged in the past, and whose members continue to suffer today from the effects of that historic wrongdoing. Examples of this phenomenon are many and varied, ranging from the waging of unjust wars, the imposition of colonial or imperial rule on indigenous people, the economic exploitation of weaker nations by stronger ones, and the emission of greenhouse gases, causing global warming with seriously damaging consequences for vulnerable people. These cases differ in many respects, but the common factor is that there is an identifiable agent, or set of agents, whose wrongful actions in the past have caused an identifiable group in the present to have fewer resources than they ought to have—fewer resources than they would have enjoyed today had the wrongdoing not occurred. So they are owed compensation by the wrongdoers or their descendants as a matter of corrective justice. The compensation, to be clear, is for the resource deficiency they are currently experiencing. They may also, independently and depending on the case, be owed reparation of a different type for the original wrongdoing—for example an apology, or some other kind of symbolic acknowledgement of what has happened in the past. I do not address other forms of reparation here, and I do not assume that material compensation is always going to be the most important demand put forward by victims of wrongdoing or their descendants. Nevertheless, material compensation is often what such groups are claiming, and (much less often) it is also what the perpetrators of wrongdoing have offered by way of reparation—the best-known example probably being the payments made from 1953 onwards by the Federal Republic of Germany to the victims of Nazi persecution, which by the end of the century totalled as much as 80 billion Deutschmarks.¹

If we are to take paying compensation seriously as a matter of corrective justice—rather than as merely a symbolic expression of remorse for what has occurred in the past—then we face the difficulty of determining how much is owed. Compensation is meant to undo the material effects of the wrongdoing and restore the victims as nearly as possible to the position which they would have been in had the wrongdoing not occurred, but to calculate what needs to be paid, we need to identify that position. How is that to be done?

The answer that most commentators give is that we have to engage in counterfactual reasoning, using the best available evidence to determine what historical trajectory the group would have followed in the absence of the wrongful intervention. There are well-known problems involved in choosing the relevant counterfactual (see Waldron 1992; Simmons 1995; Butt 2009, ch. 4; Butt 2013). We have to imagine a world in which the wrongdoing we are targeting did not occur, but in other respects remains as close as possible to the actual world. But how should we characterise the ‘normatively correct’ relationship that should have obtained between the group that perpetrated the wrongdoing (henceforth the Ws) and their victims (henceforth the Vs)? Should the relevant counterfactual be a world in which the Ws never interacted with the Vs at all, or should it be a world in which W interacted with V, but on morally permissible terms—for example, engaged in trade relations with V, but now on terms that were fair rather than exploitative? Should it be a world in which Europeans never colonised America, or a world in which some colonisation occurred but with the full and informed consent of the indigenous peoples?²

These difficult questions must be tackled at some point if we are to establish what victim groups and their descendants are owed by way of compensation. For present purposes, I am going to assume that the relevant counterfactuals can be established in order to ask a different question, one that has not to my knowledge been discussed in any detail in the literature on historic injustice.³ This is how, if at all, the response of the Vs of the wrongdoing, both during the time at which it occurred and subsequently, should factor into our calculation of the compensation that is owed to them. For the extent of the overall loss imposed by the Ws on the Vs will very often depend on what the Vs decide to do as a result of the Ws’ actions. For example, when the arrival of settlers compels an indigenous group to abandon its traditional lands and move elsewhere, its members have to decide how best to adjust to their new circumstances. Should they, for instance, continue to try to raise the crops that they have traditionally raised, or should they aim to get their subsistence in some other way? Of course, this is not a choice which they should have to make: it is forced upon them by the wrongful actions of the settlers. But given that they have to make it, we may think that their response should play a part in determining the compensation that they are owed. If they have made bad decisions, should the effect of these be discounted when calculating the loss that the Ws have inflicted, and therefore the financial reparations that the Ws should pay?

To perform that calculation, we have to introduce a second counterfactual, namely how the world would be now if the injustice had occurred as it did, but its Vs had responded to the wrong inflicted on them in a reasonably prudent way. To see what difference that makes, let us denote the Vs’ actual level of resources in the present day as R_a and the level of resources that the Vs would have in the relevant counterfactual world in which the historic injustice has not occurred as R_c. Then on the conventional approach the compensation owed by W to V is R_c–R_a. Now introduce the possibility that V has failed to respond to the injustice in a reasonably prudent way, and we have a second counterfactual to consider: the world in which the injustice has occurred, but the members of V have adapted their behaviour in a way that is judged to be reasonable. In that world the resources now available to them would be R_p. Then we can assess the compensation now owed to V not as R_c– Ra but as R_c– R_p.⁴ The proposal, therefore, is that we should calculate the compensation due by comparing the world in which no historic injustice occurred with the world in which the injustice happened as it did in the actual world, but the members of V responded to it in a reasonably prudent way.

Many readers will instinctively recoil from this suggestion, since it may sound like a case of ‘blaming the victim’. The issue, however, is not one of moral blame. Although there is certainly a normative element in our assessment of the Vs’ response, the norm we are applying is a norm of prudence rather than morality: we are asking ‘what could the Vs reasonably be expected to do in the circumstances in which they found themselves?’ Yet we still need to know whether this is a question we should be asking, and it can be broken down into two sub-questions: why should the Vs’ response matter when compensation is being calculated and when should the Vs’ response matter?

To help answer the first question, I will turn in section II to Anglo-American tort law to throw light on the principles that should guide us when assessing compensation claims. Tort law is illuminating because it has to deal with real cases in which the amount of compensation payable by the wrongdoer is in dispute. According to the associated doctrines of ‘contributory negligence’ and ‘mitigation of damages’, the damages that the tortfeasor is liable to pay by way of compensation for his wrongdoing can be reduced, or even in the extreme case entirely eliminated, if it can be shown that the victim’s negligence has contributed to the losses he or she has suffered. I do not propose applying tort law directly to cases of historic injustice, although this has occasionally been attempted.⁵ Instead I treat it as a source of principles of corrective justice that can potentially be brought to bear on larger-scale forms of wrongdoing such as those listed in the opening paragraph of the article.

A limitation of tort law, however, is that in order to decide the amount of compensation owed to the V, it has to assume that the latter is a competent agent with the capacity to make prudent decisions. It asks ‘what would a reasonable person do in this situation?’ and treats any shortcomings as potentially a case of contributory negligence (or failure to mitigate). But in the case of those who suffer historic injustice, we are dealing with groups whose collective agency may be impaired by the injustice itself, and for whom the question of what should count as reasonable prudence may be contested. In section III I explore these complications in some detail. It may turn out that in many actual cases of historic injustice, we have no reason to hold victims to an externally imposed standard of ‘reasonable prudence’ and therefore no grounds for adjusting the compensation they are owed in light of the way they have actually responded to the injustice. By undertaking this analysis, we will understand better when it is justified to hold Vs responsible, and when it is not.

Section IV concludes the analysis by returning to the reasons that lie behind the tort law principles of V responsibility, and explaining why these reasons will usually carry much less weight in cases of historic injustice.

The aim of the article, in other words, is not to reach some blanket conclusion on the question of whether Vs’ responses should affect the compensation that they are owed. It is rather to contribute to debates on historic injustice by separating, in a principled way, cases in which the way that Vs have responded should matter from cases in which it should not. But before embarking on the analysis of tort law, there are two further clarificatory points about the scope of the article that need to be made.⁶

First, as hinted at earlier, I am not attempting to give a comprehensive answer to the question of what wrongdoers and their descendants are required to do in order to repair historic injustice. I am interested only in one part of the reparative process, the part that involves paying material compensation to the Vs and their descendants for the loss they have suffered, which I treat as a matter of corrective justice.⁷ There is a wide range of views about what role, if any, compensation in the sense I am using it—non-specific resource transfers to the groups who are currently suffering the effects of historic injustice—should play in the overall process of coming to terms with the past (see O’Neill 1987; Minow 1998, ch. 5; Vernon 2003; Satz 2007; Kumar 2014). I assume only that paying compensation will sometimes be (part of) the process of making redress, so that investigating how compensation claims should be calculated is a worthwhile endeavour. In particular, I make no attempt to address the issue of supersession—how the passage of time may alter circumstances in such a way that what originally counted as an unjust seizure of property or territory ceases to be so. I am not here concerned about restitution—the handing back of things unjustly taken—which is the context in which supersession questions arise, even though that is also often part of the reparative process in the real world.⁸ Nor am I concerned with cases in which resource transfers are made as a way of expressing apology for some historic act; here what is at stake is not the material loss suffered by the present generation, but recognition of past wrongdoing as a prelude to reconciliation, and so the amount that is transferred matters only insofar as it is significant enough to be taken seriously as an apology.

I am also going to sidestep the difficult question of identifying the present-day victims of historic injustice. I approach the issue by assuming that Ws and Vs form groups whose identity persists over time, so that an individual person can make a claim today by virtue on belonging to such a group, but I do not try to specify the nature of that identity, which anyway is likely to differ from case to case. On the side of the Ws we will usually be dealing with groups that are politically organised as nation-states, so in practice it is not difficult to work out who should be held liable to pay the compensation that is due. There remains, of course, the much-debated issue of how individual members of W today can justifiably be held remedially responsible for paying compensation for injustices perpetrated by their ancestors. I assume that this question can be answered positively and will not discuss it further here.⁹

On the side of the Vs, by contrast, we cannot assume that they will have remained an organised group in the time that has elapsed since the original wrongdoing took place. So their membership has to be established through descent: the present-day Vs are those who are linked to the original Vs by familial or communal ties, such that they are the expected inheritors of the property and other resources held individually or collectively by the earlier generations of Vs. In the real world, establishing their identity will be complicated by the fact of mixed descent and/or the fact that some descendants of the original Vs will have chosen to cut their ties with the group, thereby stepping outside the line of inheritance. These complications, it should be said, threaten the whole project of corrective justice, since we cannot begin to calculate compensation until we know to whom it is owed. Since my concern is not so much to defend that project in its own right as to ask how it should be carried out, I set the complicating factors aside and assume that the identity of the Vs across time can be determined in the way suggested.

II

In this section I explore how the issue of Vs’ responsibility has been handled in Anglo-American tort law. My assumption is that the evolving body of tort law captures shared intuitions about what is fair when someone has sustained a loss by another’s hand and the cost of repairing the loss has to be divided between them. Looking at the problem in this more restricted setting can provide answers to several questions, such as when the tortfeasor should be obliged to repair all of the loss irrespective of the victim’s contribution to it, and what standard should be applied when judging whether the victim’s actions count as negligent. There is of course debate about whether tort law in the form it now takes is the best way to deal with cases in which W causes V to suffer a loss: a scheme of social insurance that compensates V regardless of what W has done might be proposed as an alternative.¹⁰ Nevertheless for the problem addressed in this article, the tort law analogy seems to fit best, since there is no global institution currently in a position to compensate all of those who have suffered historic injustice, so no equivalent to social insurance as practised at national level. If any compensation is going to be paid, it will have to be by the perpetrators themselves—typically individual nation-states, or combination of such states.

For readers not well versed in tort law, a brief sketch of the relevant aspects follows. Given that W has negligently injured V’s person or property, under what circumstances should V be held responsible for part of the injury such that the damages he can claim from W are reduced? Two doctrines are relevant here, and there is some disagreement about whether they should be regarded as wholly separate, or as companion doctrines that both give expression to the same underlying principle.¹¹ The first of these is ‘contributory negligence’ and here the main focus is on the role of the V in the event that caused the damage—though the doctrine has also been applied in some cases to her subsequent behaviour. Thus if a speeding driver hits a pedestrian who has wandered into the road without paying attention, the pedestrian may be judged to have negligently contributed to her injuries; moreover (and regardless of whether she was negligent in causing the accident), if she fails to have the injuries treated in good time, as a result of which they are exacerbated, she will again be seen as negligent. Contributory negligence of any kind was once held to block payment of damages outright, but current tort law uses a principle of comparative negligence, whereby the costs of the wrongful event are apportioned between the two parties according to the degree of fault that each has displayed.¹² The formula used to make the apportionment varies between jurisdictions, but the general principle is that the V’s overall claim for compensation is reduced in proportion to the extent of her negligence compared with that of the W.¹³

The second relevant doctrine is ‘mitigation of damages’. Here the focus is exclusively on the V’s behaviour after the tort has occurred. He is said to have a ‘duty to mitigate’, i.e. to take steps to ensure that the resulting damage is minimised. If he does so, he can claim the costs of mitigation against the tortfeasor, and if the attempt to mitigate is unsuccessful, can then claim compensation for the whole of the damage. But if he makes no such attempt, then the law will try to separate that part of the damage that is directly due to the W’s behaviour from that part that is due to the V’s neglect of his duty to mitigate. So whereas contributory negligence distributes costs according to relative degrees of fault, mitigation of loss does so on the basis of causal contribution: which parts of the damage can be attributed causally to W’s behaviour, and which parts to V’s negligent failure to mitigate?¹⁴

The effect of these doctrines is to reduce the compensation that W is required to pay to V in cases where W is ordinarily negligent: he has acted in a way that carelessly and wrongfully risks imposing costs on V or others like her, as for example does the speeding driver or the householder who leaves a trip hazard outside his property. Neither applies, however, in cases in which W either deliberately or culpably recklessly imposes costs on V. Here V can claim the full costs of repair regardless. The rationale for drawing such a line may seem obvious. As one tort law theorist puts it, ‘if a person intends to harm another, it does not seem fair to allow that person to argue that the victim did not take sufficient care to protect himself or herself from being harmed’ (Cane 1997, p. 59).¹⁵ As I suggest below, preserving fairness between the parties is one of the two main rationales that can be used to support the negligence doctrines, and a W who deliberately or recklessly wrongs his V forfeits his claim that the V should co-operate to reduce the damages he must pay.¹⁶

This restriction has clear implications for the use of tort law to throw light on Vs' responses to historic injustice. If deliberate or culpably reckless wrongdoing exempts the Vs from any duty to mitigate, then this will apply to many of the entries that appear in the conventional record of historic injustice: genocide, ethnic cleansing, and slavery, for instance. So we have already reached our first conclusion: Vs’ claims to compensation should not be affected by how they respond when the injustice they or their predecessors have suffered is of this kind. I return to this issue at the beginning of section III.

But this still leaves, regrettably, many instances of historic injustice to consider: cases in which victims have been badly treated by being deprived of resources, exploited, displaced, and so forth by Ws who pursue their own interests without sufficient regard to the impact they will make on the Vs, but without deliberately or recklessly causing them harm. We need to look more closely at the normative reasons for applying contributory negligence/mitigation of damages to compensation claims. Is this simply motivated by efficiency concerns—encouraging victims to take steps to avoid being harmed, thereby reducing the number and the costs of accidents—or are there deeper reasons behind the practice? Why is it fair to ask the V to bear part of the cost when it is the W, and not the V, who is in breach of a duty? If the W had not acted as he did, there would be no costs to apportion.

This is the challenge posed by Klar, for example, who puts the case against applying contributory negligence by highlighting morally relevant asymmetries between W (here ‘the defendant’) and V (here ‘the plaintiff’):

It is also clear that the positions of the two parties are quite different. The defendant’s negligent conduct endangers others, which is clearly more egregious and worthy of sanction than conduct which only fails to protect oneself, but endangers no one else. In addition, plaintiffs whose negligent conduct has failed to prevent or minimise their injuries already bear a heavy price—they must suffer these injuries. The consequences to a defendant of causing injuries to others are solely financial, if that (Klar 2019, p. 144).

The key point here is that while both W and V are negligent, W’s negligence consists in a failure to take sufficient care to safeguard the interests of others, whereas V’s negligence consists in a failure to safeguard his own.¹⁷ Although in both cases we can talk about ‘faulty behaviour’, the two types of fault are really incommensurable, critics suggest, and therefore, the idea of distributing costs according to the degree of fault that W and V respectively display, as the doctrine of comparative negligence requires, makes no sense on closer inspection.¹⁸

According to current legal doctrine, the criteria for judging conduct negligent should be the same for both W and V.¹⁹ The standard in each case is what a ‘reasonable person’ can be expected to do, whether to safeguard others’ interests or her own. Yet in practice, it seems, courts afford greater latitude to the V, taking account of subjective factors when deciding whether she has behaved reasonably in the circumstances, such as personal phobias or lapses of attention that would not excuse the W (Simons 2015, pp. 36–38). So applied tort law acknowledges some degree of asymmetry between W’s negligence and V’s, but still allows V’s negligence to reduce the compensation she which can claim from W. Moving from practice to principle, how can this be justified? On two grounds, I suggest.

The first is simply fairness. A loss is going to occur, but the extent of that loss—the final injury suffered by V—depends on the behaviour of both W and V. Although W is morally at fault in a way that V is not (assuming here that V has only a prudential and not a moral duty to protect himself from injury), both have behaved in a way that fails the ‘reasonable person’ test, and so it is fair to allocate the loss between them in a way that reflects the extent of their fault.²⁰ So even if W is held to a higher standard of behaviour than V, where V can be shown to have been negligent he should be expected to carry some part of the overall loss. This aligns tort law with the general principle that where a person suffers a loss that she could easily have avoided by taking certain steps, she has no grounds for claiming compensation from others.

A second ground has a more communitarian character. Each of us has a social responsibility to avoid creating costs for others if we can do so at little cost to ourselves. So, for example, in a society that operates a public health service, each person has a responsibility to take reasonable steps to stay healthy so as to avoid overburdening the system—such as agreeing to be vaccinated against serious diseases. We can extend this principle to the behaviour of tort Vs. The extension is easiest to justify in the cases that fall under ‘mitigation of damages’ where the V has to decide how to act after the original tort has occurred. If he fails to mitigate, he increases the costs that the W is liable to pay. The proposal is that he owes it to the W to keep the costs down, provided the necessary repairs are easy to organise and their (lesser) cost can be recouped from W as part of the settlement.

Whether we find this reasoning compelling may depend on how we view the relationship between W and V. Here it is important that the relationship should not have been severely impaired by the W having acted in a way that shows a callous disregard for the welfare of the V. We are considering cases where W’s behaviour is negligent, but not culpably so, and negligence of this kind does not cancel duties that are otherwise owed to W—for example it does not entail that he has forfeited his right to be rescued at modest cost.²¹ The suggestion, then, is that although in tort cases the W must of course pay for the damage he has caused, the V owes it to him to keep the cost down when she can do so by taking reasonable steps to mitigate. Failure to mitigate is not treated by the law as itself a form of wrongdoing (in that respect talk of a ‘duty to mitigate’ is misleading), but it is liable nonetheless to attract a penalty in the form of reduced damages.

That concludes my reconstruction of the principled case for reducing the compensation payable to Vs who are contributorily negligent or fail to mitigate damage. One further aspect of the V’s responsibility is worth exploring briefly before returning to cases of collective historic injustice. This is the question of whether he can escape being judged contributorily negligent if he refuses to take cost-reducing steps because of some moral or religious objection to the actions he would need to perform. The general limits to the V’s responsibility that tort law recognises have been summed up as follows:

If the effort, risk, sacrifice, or expense which the person wronged must incur in order to avoid or minimise a loss of injury is such that under all the circumstances a reasonable man might well decline to incur it, a failure to do so imposes no disability against recovering full damages (McCormick 1935, p. 133).

How, then, have courts decided cases in which on religious grounds a V refuses to take steps to limit his injuries, such as refusing to accept a blood transfusion? It appears that they have generally not been willing to allow this as grounds for increasing the V’s damages: he will only be compensated for the injuries he would have sustained if he had accepted the transfusion or other such medical aid (see Parobek 2006; Ramsay 2007). In other words, the ‘reasonable man’ test that is applied to decide whether sufficient mitigation has been attempted brackets off the religious belief; it is a generalised reasonable person, not, say, a reasonable Jehovah’s Witness, who is appealed to in order to decide whether it is an excessive sacrifice for an injured person to receive a blood transfusion. Critics have argued that tort law should be reformed so as to take greater account of the V’s existing moral and religious commitments when assessing mitigation of damages, but so far, it appears, it has been disinclined to do so.²² What role such commitments should play when claims for compensation for historic injustice are being assessed is an issue to be addressed in the following section.

III

There are several points of disanalogy between the claims that are typically made in tort law, and the compensation claims that can be made by the victims of historic injustice. One already alluded to is that in the latter case those affected make their claims not on an individual basis but by virtue of their membership in a large group that collectively has borne the impact of the injustice over time. A second difference is no less important: whereas typical tort cases arise from a specific event, such as a car accident, the episodes of historic injustice that mainly concern us occur over lengthy periods of time. They usually involve policies systematically pursued by the Ws, such as expansionist territorial policies or aggressive trading policies. This makes it harder to apply the distinction that tort law makes between conduct that is ordinarily negligent and conduct that involves deliberate or culpably reckless harm to its V. As noted already, following tort law reasoning entails that Vs’ responses should make no difference when they are subject to gross historic injustices such as invasive war or genocide whose perpetrators act with the deliberate intention of inflicting damage on their Vs.²³ So if the analogue to contributory negligence applies at all, it will be to cases in which Ws act in their own interests and without sufficient regard to the interests of the people who will bear the impact of their behaviour, but without the intention to cause damage.

But are there any such cases? Looking at the historical record, what we often find are policies pursued by the Ws that qualify as negligent overall, though punctuated by specific events that amount to deliberately harmful or culpably reckless behaviour. Consider, for example, the case of colonisation. The immediate aim of the settlers may simply be to acquire land to support themselves in a place that they regard either as unoccupied, or as so sparsely occupied that the colonists can live alongside the indigenous people without unduly disrupting their lives; their negligence lies in failing to understand the impact that their arrival will have on human communities whose territorial needs are very different from their own. To look at colonisation in this way is not, however, to deny that its actual historical record in North America and elsewhere is also blighted by massacres, forced removals, and other events that straightforwardly count as deliberate wrongdoing. Or take the case of imperial rule that emerges out of a desire to stabilise what originally were trading relationships, which those who undertook them may have understood to be advantageous to both parties, even if in fact they were not. Here again, the injustice (overall, and leaving aside specific episodes) involves a kind of negligence—a failure to consider, on the one hand, the material impact of establishing trade relations between a developed industrial economy and a subsistence economy, and, on the other hand, the corrosive cultural impact of the traders’ presence in the society in question.

These cases are admittedly controversial, since the intentions of the perpetrators may be contested. I will continue to use examples drawn from the history of colonialism, in particular, for the light that they throw on the question of Vs’ agency, but others might wish to restrict the analysis that follows to behaviour that unequivocally counts as negligent. The emission of greenhouse gases by industrial nations, leading to global warming that seriously harms people living in the global South, certainly meets this description, at least after the point at which the causes and effects of climate change were scientifically established. So do cases of trade relations between independent nations that create significant levels of hardship on one side, even though they were undertaken in the expectation that they would be mutually advantageous.

The main question to be addressed in this section is whether there are likely to be impediments to the collective agency of the Vs such that they cannot be held responsible for failing to mitigate the effects of the Ws’ wrongdoing.²⁴ Tort law treats victims as responsible agents with the capacity, if not always the will, to take steps to avoid or lessen the impact of the damage inflicted on them. Can we assume the same of those who have suffered historic injustice? To see why group agency matters, consider cases in which a collective loss occurs as a result of a series of isolated individual acts. For example, the land that an indigenous group occupies and on which it depends for its existence as a community might be sold off parcel by parcel to colonisers, and each sale might be a rational choice on the part of the individual group member who makes it. To prevent its effective dissolution, the group needs to decide collectively which, if any, parts of its land should be treated as transferable to outsiders. But in order for that to happen, the group must be able first to reach a decision, and then exert sufficient authority over its individual members to ensure compliance. That is one necessary condition for its being a group agent. There is no need to go further here and stipulate features of the group that might be desirable but are not essential to its agency, such as making its decisions through democratic deliberation. A group that acts on the commands of a headman can count as a group agent so long as the rest of the group recognise the headman’s authority and for the most part comply willingly.

Why might the group agency condition fail in the case of groups that are suffering, or have suffered, from historic injustice? In what follows I consider four reasons why V groups may be prevented from acting in ways that might appear from an external perspective to be ‘rational’ as a response to the loss that they have incurred (more than one reason may apply in a particular case). This first is that, among the other harms that they have inflicted, the Ws have undermined the previously prevailing system of authority that would have enabled the Vs to act collectively. For example they may have co-opted a traditional leader to serve as their henchman—a common feature of imperial rule—or they may have forced the Vs to relocate to different territory, in the process breaking down the social relationships that had enabled the group to co-ordinate.

As an example of the latter, consider the forced removal of the Cherokees from Georgia to what is now Oklahoma in the 1830s, often referred to as ‘The Trail of Tears’.²⁵ The pressure applied by the state of Georgia divided the Cherokees between hardliners who believed that on no account must they abandon their ancestral lands, and pragmatists who came to think that removal was the only way in which their nation could be preserved. When the leaders of the second faction, known as the Treaty Party and headed by John Ridge, arrived in the west, they accepted the system of government established there by previous Cherokee settlers. But they were soon challenged by the larger hard-line group led by John Ross, and condemned for having sold land in Georgia, with three prominent leaders including Ridge himself being executed according to a tribal law that forbade the unauthorised sale of Cherokee land. The bitter divisions between the two factions that resulted from these events lasted for decades. As one historian puts it:

Unification of the Cherokee Nation proved illusory. Disorder continued for years, something like a Corsican vendetta raging between the Ross and Ridge adherents, with a mournful succession of murders. The Ridge faction killed their oppressors without warrant of law, the Ross supporters committing ‘similar excesses, sometimes in the name of the law. Life and property were not safe and a state almost of civil war existed’ (Wilkins 1986, pp. 342–343).

A further consequence of their internal division was that the Cherokees were unable to present a united front in negotiations with the federal government—essential if they were to receive the financial compensation they were owed for the costs associated with their removal—since each faction sent its own delegates to Washington, allowing successive administrations to play them off against each other, meanwhile withholding payment (McLoughlin 1993, ch. 1). Although a treaty signed in 1846 brought a period of relative stability, the outbreak of the Civil War reignited the sectarian divide, with partisans of the Ridge and Ross factions enlisting on opposing sides. Indeed ‘a legacy of hatred smoldered on for generations, coloring Cherokee history even after tribal government was superseded in 1907 by formation of the State of Oklahoma’ (Wilkins 1986, p. 344).

I have cited the case of the Cherokees to illustrate how historic injustice, here in the form of the forcible appropriation of a people’s inherited territory and their physical displacement, can fracture its Vs’ institutions of self-government in such a way as to render them largely incapable of collective action. Under these circumstances, it is question-begging to ask whether the Vs ought to have responded differently, since their capacity to respond, collectively, has largely been destroyed by the Ws themselves. So this gives us a first reason to believe that the hypothetical responses of Vs—what they might have done had they been able to act as a group agent—is irrelevant to any judgement about what they are owed by way of reparation.²⁶

Consider now a second way in which the impact of the Ws might cause the V group to act in ways that are detrimental to its own interests, even after the actual Ws have left the scene. Assume that Franz Fanon’s diagnosis of the psychologically debilitating effects of colonial rule on its subjects is correct (Fanon 1968, 2001). Particularly where the division between settlers and natives is also a racial one, as in the African case, the colonised, Fanon claims, suffer from a ‘dependency complex’. As black people, they idealise and seek to emulate the whites, while also realising that they can never in fact become white themselves. As he puts it:

When the native confronts the colonial order of things, he finds he is in a state of permanent tension. The settler’s world is a hostile world, which spurns the native, but at the same time it is a world of which he is envious. We have seen that the native never ceases to dream of putting himself in the place of the settler—not of becoming the settler but of substituting himself for the settler (Fanon 2001, p. 41).

Fanon argues that the only way to escape from this mentality is through violent struggle in the course of which the colonised will come to reject the beliefs and values that make up the colonial world-view in its entirety. (The alternative is that the colony becomes formally independent, but the ‘national bourgeoisie’ simply take over the positions of power abandoned by the colonialists, so the structure of colonial rule remains intact.) However there is no reason to think that beliefs formed in the course of violent struggle will be a reliable guide when post-colonial institutions and policies have to be adopted. For example, suppose it is indeed the case that private property, the rule of law, and an open economy are key determinants of a society’s wealth, as economic historians such as Acemoglu and Robinson argue (Acemoglu and Robinson 2012), but these are rejected by the recently decolonised as institutions that only serve the interests of the settlers; then they may choose to adopt alternative institutions and practices that perform badly, and make the economic legacy of colonialism even worse than it needs to be.²⁷

If what Fanon says is true, therefore, the experience of colonial rule may lead to two forms of distorted consciousness, each with own damaging effects: first, while the regime is still in place, failed attempts by its victims to emulate the behaviour of the colonisers; then second, when it is removed, a wholesale rejection of the institutions and practices that the colonisers had introduced, even when these might (perhaps with some modification) have been beneficial to them. So here the prior behaviour of the Ws is at least in part responsible for the way in which the Vs respond to the wrongdoing. This contrasts sharply with tort law cases, where the V is regarded as being wholly independent of the tortfeasor. His choice of how to respond to the wrong that has been inflicted on him—whether prudently or imprudently—is his alone. W is not held responsible for V’s poor decisions if that is what ensues.

We should not assume that Fanon’s diagnosis, even if true of the particular form of colonial rule that he experienced in Martinique and Algeria, generalises to all cases of historic injustice: it is unlikely to.²⁸ Nevertheless we have uncovered a second reason why the Vs’ response to injustice should be discounted when the compensation they are owed is being calculated. Even if their collective agency has not been completely destroyed, their ability to respond effectively in the aftermath of injustice may be undermined by distorted beliefs formed during the time that they suffered it, responsibility for which lies at least partly with the Ws.²⁹

A third reason for discounting the Vs’ response to injustice concerns the standards which it is appropriate to use when judging whether victim groups have behaved in a reasonable way given the circumstances they face. As we saw, tort law asks what a ‘reasonable person’ would have done in order to judge whether a victim has behaved negligently. We also saw that there has been some debate over whether this reasonableness test should take into account commitments such as religious beliefs that would rule out taking actions that other people might be expected to take, such as receiving certain forms of medical treatment for injury. However, for understandable practical reasons, the law has been wary of making such adjustments. Vs are judged in the light of prevailing social norms that set standards for what a person can reasonably be expected to do by way of avoidance or mitigation. When we turn to assess the collective response of a V group, however, we will often be making judgements about people whose shared culture and its accompanying social norms are markedly different from those of the Ws (and from our own). This will have implications for what counts as reasonable behaviour in the aftermath of the wrongdoing. As an example, consider a Muslim society that has freed itself from colonial exploitation. Suppose its economic recovery is slowed by the fact that it adheres to the Islamic prohibition on lending money at interest. Since the norm prohibiting usury is one of the central elements in Islamic doctrine, it would be wrong to judge this behaviour as unreasonable and apply a counterfactual in which the society develops using banking and other money-lending practices modelled on those developed in the West. A fortiori, it would also be wrong to apply a counterfactual in which the society abandons Islam altogether. Instead we should assume that adherence to Islam is one of the ties that binds it together and allows it to function as a collective agent.

The fourth and final reason why we should be hesitant in applying counterfactuals when assessing the behaviour of V groups concerns the distribution of gains and losses inside of the group: this again draws our attention to a significant difference between an individual tort V and the group having to respond to historic injustice. Consider the following example: a society whose agriculture has traditionally been organised on the basis of collective property-holding continues with this practice following decolonisation. Western advisors point out that economic growth would be faster if farms were privatised, making it possible for the economic damage inflicted by the colonial regime to be repaired more quickly—in other words R_p > R_a. However their advice is rejected on social grounds: the concern is that privatisation, even if economically efficient overall, would remove the element of mutual insurance contained in the collective arrangement, and leave some villagers vulnerable to becoming destitute. There is no parallel to this in tort cases in which the V is a single person and the only question is whether she has acted in a reasonable way to avoid or lessen the damage she has suffered. Groups cannot be held negligent for seeking to protect the interests of their individual members. A response on the part of the group that imposed severe risks on some of its members can be ruled out as unreasonable even if the response would have been maximally beneficial to the group as a whole.

It is important to underline here that we are investigating responses to injustice. The Ws have unfairly damaged the earlier Vs in such a way that the later Vs are now deprived of resources that they should have had: R_a < R_c. We are asking whether they should have behaved differently, such that their reparative claim would be smaller. But clearly, if that different behaviour would have involved the unjust treatment of a sub-group of V—say the villagers who become destitute—injustice has not been repaired but merely shifted between groups: the Vs collectively would have behaved unjustly in their attempt to repair the damage that they have suffered at the hands of the Ws. This is not how the Ws can reasonably expect the Vs to have behaved. Even though they are entitled to ask whether the Vs have acted prudently when calculating the compensation they owe, the standard of prudence must be set so as to exclude imposing unfair costs or undue risks on individual members of the group.

IV

Let us take stock of the discussion so far. We have been considering cases in which V groups appear to have responded sub-optimally to the injustices that they have suffered when measured by some standard of reasonable prudence. They find themselves at R_a when they might have been at R_p had they behaved differently. The Ws, let us suppose, are claiming that the compensation they are required to pay is R_c– R_p and not the larger amount R_c - R_a. I have looked at four challenges to that claim—four reasons why the Vs could not be expected to reach R_p.

• 1. The treatment experienced by the Vs at the hands of the Ws has destroyed their capacity to act as a group agent, which they would need in order to reach R_p.

• 2. The treatment experienced by the Vs at the hands of the Ws has inflicted psychological traumas that prevent them from adopting the institutions and policies that would lead to R_p.

• 3. In order to reach R_p, the Vs would need to contravene a moral norm or principle that is deeply embedded in their culture, and not open to revision.

• 4. In order to reach R_p, the Vs would need to adopt policies that risk exposing some of their number to severe injustice.

It is plausible that in many actual cases of historical injustice, one or more of these reasons will apply, so claims that the Vs’ compensation should be reduced to R_c– R_p should be dismissed. This is particularly the case when the injustice is ongoing, or has only recently ceased. When it is further in the past, reasons 1 and 2 in particular are less likely to apply to the more recent behaviour of the Vs. Accepting this narrowing of scope, it is still worth examining cases in which none of the four extenuating factors obtains: the Vs have collective agency, they are not traumatised, but they have a shared culture that prevents them from reaching R_p, a culture that they could choose to adapt without violating a fundamental belief (so 3 does not apply) and without the risk of internal injustice (so neither does 4).

To address such cases, I start from a view that I share with, among others, John Rawls, namely that inequalities between groups and between societies can be unobjectionable when they arise from cultural differences; in other words, a collective cannot claim compensation if they are materially worse off than a second collective just because they have a different culture that leads, for example, to their being less economically productive over time.³⁰ Although this view faces certain objections (in particular it entails that later generations in one society can justifiably be worse off than their contemporaries elsewhere simply because of decisions that their predecessors made and for which, therefore, they themselves were not responsible), I will take it for granted here and not attempt to reply to such objections. My interest is in whether the principle of collective responsibility that can justify inter-group inequality in general also applies when what is at stake are claims to compensation. To see whether it does, we need to return to the reasons that I proposed in section II to explain why it was reasonable for tort law to hold Vs responsible for their failures to mitigate.

The first of these involved an appeal to fairness. When V fails to take reasonable steps to prevent or mitigate his loss, both he and Ware to some extent negligent. So although W as the wrongdoer bears primary responsibility for the overall loss, it is fair for V to assume some part of it in the form of reduced damages. However, this fairness claim looks most plausible when the two parties are equally able to bear a share of the loss—for example, they are two individual car drivers, and there is no general reason to think that the careless driver who causes the collision will be in a better position than the owner of the damaged car to pay the repair bill. Tort law does of course also apply where we should expect the parties to be unequally placed in this respect—such as employers and workers in industrial accident cases—but in such cases it may recognise a duty of care on the part of the stronger party that precludes finding the victim contributorily negligent.³¹ In the cases of historical injustice that concern us, it is very unlikely that there will be rough equality between the Ws and the Vs—indeed it will be the Ws’ stronger position that has enabled them to pursue policies that impose significant material losses on the Vs (such as exploitative trade policies). So the fairness rationale for holding negligent Vs responsible for part of the loss is going to be very much weaker, since in these cases it conflicts with a rival principle of fairness: the principle that where costs are now unavoidable, they should be allocated in such a way that they fall mainly or entirely on the shoulders of those who are best able to bear them.

The second rationale I described as communitarian in character, appealing as it does to the responsibility which we have not to increase the burden that falls on others’ shoulders when it is easy for us to do so. Although you have carelessly damaged my car, if I can reduce the hefty bill you would have to pay by taking timely reparative action now, I should do so. That makes sense if we are neighbours or fellow-citizens with general social responsibilities to one another. But generally speaking the relationship between the Ws and the Vs in cases of historic injustice is not of that kind. Their interests are likely to be conflictual, for example if the Vs have suffered the impact of colonisation, or have been the weaker parties to a trade negotiation. Under these circumstances, although the Vs will still have prudential reasons to mitigate the effects of the injustice, they will not have moral reasons arising out of their relationship with the Ws. Moreover the idea that V owes it to W to reduce the overall amount of compensation that needs to be paid makes sense on the assumption that W will actually pay up. In tort cases, the fact that payment of damages can be legally enforced should provide a sufficient level of assurance that this will happen. But when we move to cases of historic injustice, the default position is that the Ws will deny their liability (the payments made by West Germany to Holocaust victims stand out as an exception) and there is of course no mechanism—no international reparations court with enforcement powers—for making them pay compensation.

Does it then follow that the answer to my title question is an unequivocal ‘No’: we should never consider how the victims of historic injustice have subsequently behaved when calculating how much compensation they are owed? That conclusion is too strong. We have found that there are several reasons why the Vs in responding to injustice could not be expected to behave in ways that might at first sight seem to be prudent. We have also seen that the justifying reasons that can be appealed to in support of the doctrines of contributory negligence as used in tort law will not apply in many instances of historic injustice. But there may still be cases in which the fairness reason in particular does apply, and in which the factors that would disqualify the Vs as group agents are absent. Typically they will be cases in which the Vs form a politically organised group that is not dominated by the Ws and has the capacity to make real choices. Consider two examples from the contemporary era. If an oil tanker registered in state A founders and discharges its oil on the shores of state B, the compensation owed to B (including for clean-up costs) may depend on whether B has taken reasonable steps to contain the disaster by deploying containment booms or skimmers. Or consider a small island state threatened by sea level rise as a result of climate change. The rich states chiefly responsible for global warming may propose subsidising adaptation measures that require some changes in the islanders’ lifestyle (though without radically disrupting their culture). If the islanders decide collectively to reject these proposals, and as a result their island later becomes uninhabitable, they cannot, at the bar of justice, claim to be fully compensated for the costs of their relocation (they might still have a needs-based claim, but recall that throughout this article I have been treating claims for compensation as claims of corrective, not distributive, justice).

These examples show that it is a mistake to dismiss the issue of Vs' responsibility just because the Vs form a collective that has been exploited or damaged by another collective. Critics, however, will say that they are quite different in character from the episodes of historic injustice that feature most prominently in public debate (and the academic literature that it generates). This is broadly true. As I have indicated at various points in the article, these episodes often involve the intentional harming of Vs, or forms of domination that undermine the latter’s collective agency. So the conclusion we should draw is not that, on principle, the response of victims should never matter when the compensation they are due is being calculated; rather, that it is unlikely to matter in the most flagrant cases of historic injustice, which are also the cases that most often give rise to reparative demands in the world we inhabit.