Reparations to Whom?
So who, then, has reparative rights once the war is over? Victorious belligerents do not have those rights qua victors, since their war might have been unjust. Moreover, not all individual members of just belligerent communities have such claims either, since they themselves, in their individual capacity, might have been harmed in the course of committing wrongful war acts. Post-war reparative rights track wrongful harms—whether claimants belong to the victorious side, the vanquished side, the unjust side, or the just side. Consider for example combatants who were fighting for the Rwandan Patriotic Front in 1994 as extremist Hutus were carrying out a genocide against Tutsis and moderate Hutus. Many of those combatants were wounded by the Rwandan Army as they were fighting the latter’s genocidal forces. Others, however, were wounded as they were committing atrocities against Hutu civilians. In so far as they were acting wrongfully, they do not have reparative claims in relation to those wounds—whereas their victims do have such claims in relation to their own predicament. In this respect, restitutive claims interestingly differ from reparative claims: for whereas (as we saw in ch. 5) a party may be under a duty at t1 to restitute a good which they rightfully seized at t2, a legitimate reparative claim necessarily rests on the commission of a wrongful harm.
That being said, the notion of wrongful harm needs disambiguating. As I noted in s.1.3, we must distinguish between violating a right and justifiably infringing it. A violates B’s right not to suffer a loss if B has not done anything to warrant losing that right and if A lacks a moral j ustification for imposing that loss on B. Sometimes, however, A may have such a justification (typically, his life is at stake) even though B has not done anything to lose his right. Most obviously, individuals whose rights are being violated have reparative rights. But so do individuals whose rights have been justifiably infringed. Suppose that some agent A takes from B a good, g, which is in B’s prima facie rightful possession, but which A needs in order to survive. In Joel Feinberg’s well-known example, A is a stranded mountain-hiker who will die unless he finds shelter very soon. He chances upon a cabin, owned by B and currently vacant, breaks in, eats the food left by B and burns one of the wooden chairs to warm himself up. Feinberg holds—as does, e.g., Judith Thomson—that A’s dire predicament furnishes him with a justification for acting as he does, but that B retains ownership rights over both the cabin and its contents, such that A owes her reparation for the food he ate and the wood he burned.
In this particular case, I disagree with Feinberg and Thomson. At the bar of the ethics of assistance which is part and parcel of my account of cosmopolitan justice, B is under a duty to A make her cabin available to him and A acts within his rights by consuming her resources. But consider the following variant of the case, whereby B is on the scene and is not physically able to open the cabin to A. The only way for A to get into the cabin is to trample over B, as a result of which she would become permanently paralyzed from the neck down. In this particular case, in so far as A ‘merely’ tramples over her and thus is not using her body as a means to get in, and as he would die if he does not get in, he has a justification for so acting; at the same time, the no-undue-sacrifice proviso yields the conclusion that B is not under a duty to let A so harm her and that she thus retains her right that he not harm her. In so far as she does retain her right against harm, she has a reparative right against A.
Let us apply this to war cases. Assuming here again that the rough and ready intuition is correct, victims of rights-violations, who thus have rights to be compensated for the resulting harms, include: combatants and non-combatants who are deliberately killed or maimed without justification—for example in prosecution of a war of unjust aggression, or in a war which has a just cause but which nevertheless breaches some other requirement; individuals (usually non-combatants) who are killed unintentionally, though foreseeably, but without justification; individuals, whether combatants or not, whose houses, land, livelihood, and collectively owned property have been deliberately and unjustifiably destroyed. Victims of justified rights infringements, for their part, include: those who suffer grievous harm, to which they are not liable, as a result of the deliberate killing or maiming of liable agents (for example, the dependent children of combatants who fight for an unjust cause); those who suffer grievous harm, to which they are not liable, as a result of the deliberate destruction of privately and collectively owned property and infrastructure; those whose life, limbs, and (collective and private) property are destroyed as justified collateral damage; those whose property, both privately and collectively owned, is justifiably taken by the enemy in prosecution of its just war or a justified occupation.
The case of military occupation is particularly interesting. As we saw in ch. 3, a just belligerent, A, which takes position in B’s territory, may have a powerful justification for requisitioning and/or exploiting resources which are in citizensB’ possession—food, clothing, dwelling, natural resources, agricultural land, and so on. Even if some members within B are significantly causally responsible for providing A with a just cause for war, not all are. Moreover, among those who are not responsible for As predicament (for example, because they could in no way influence their regime’s bellicose policy vis-a-vis A), there are many who are not under a duty and Duties of Compensation’, Philosophy & Public Affairs 13 (1984): 79—88; J. J. Thomson, ‘Rights and Compensation’, Nous 14 (1980): 3—15; M. J. Zimmerman, ‘Rights, Compensation, and Culpability’, Law and Philosophy 13 (1994): 419—50; N. Davis, ‘Rights, Permission, and Compensation’, Philosophy & Public Affairs 14 (1985): 374—84. Thomson defends a Feinbergean line, while Montague and Zimmerman argue, against Feinberg, that the hiker does not owe compensation to the cabin owner.
to assist A’s troops in their just war against their own regime and unjust compatriots—notably because they would jeopardize their own prospects for a flourishing life by providing such assistance. By implication, they retain their rights not to be so harmed, and thus not only have restitutive claims to their land, houses, and so on (under the conditions set out in ch. 5): they also have claims to share in some of the benefits which A’s troops derive as a result of extracting wealth from B’s collectively held resources, as well as reparative claims for the harms they suffered as a result of A’s requisitions.
Some loose ends need tying. First, I stated above that the dependents of those who are justifiably and deliberately killed or maimed can have rights to be compensated for the harms which they themselves suffered as a result. The point applies to descendants of war victims even if they owe their existence to the war itself (as we saw in s.5.2.2), so long as they can trace the harms from which they suffer to continuing rights violations or justified rights infringements committed during the war after they were conceived, and subject to the usual caveat pertaining to the passing of time, the number of claimants, etc.
Sometimes, however, there are no descendants left, or at least no easily identifiable ones—as when whole families are killed in war-related genocides or in systematic and destructive bombing. There are two ways to deal with this particular case, which incidentally was at the heart of reparations negotiations conducted by (West) Germany, Israel, and representatives of the Jewish diaspora in the early 1950s.5 One can either hold that no one has a reparative right to this particular package, which can then be used by the duty-holder for non-reparative purposes, including distributive justice purposes; or one can hold that other victims of this act of war have a claim to it. On a strict construal of the demands of reparative justice, only the direct victim of a given wrongdoing and her heirs are owed reparation for this wrongdoing. Suppose, then, that given available resources, a peace settlement allocates a certain sum of money (say $5000) to the surviving heirs of those who were wrongfully killed by A in bombing campaigns, and that 5000 people have died heir-less. On that strict construal, those $25 million dollars are not to be used for reparative purposes, even to help surviving victims of that bombing campaign (who thus incurred a harm lesser than death and would have been entitled to less than $5000 when all the dead’s heirs are taken into account). However, reparative justice really is a genuine moral imperative, and thus dictates that we confer on those other victims a right to increased payments until the demands of reparative justice as pertains to this particular act of war have been satisfied—subject to meeting the competing demands of distributive justice.
Second, victims of a rights-violating regime who benefit overall from third parties’ decision to wage a war of humanitarian intervention do not have reparative rights against interveners with respect to the justified harms which the war may nevertheless have caused them (though they do have claims against their erstwhile regime both for the pre-war harms it caused them and for the in-war harms that its 
troops also imposed on them by unjustly fighting the interveners.) As Frances Kamm notes, the claim on the one hand draws strength from the view that beneficiaries of an intervention are sometimes under a duty to share some of the costs of the latter, and on the other hand is subject to the proviso that interveners did not violate their rights in the course of fighting that war.
As should be clear, given the multiplicity of victims and of the harms which they incur for which reparations are warranted at the bar of the Reparative Principle, the epistemic hurdles which any institutional scheme for implementing the principle face are so high as to be ‘unclearable’. For not only would one need to know whether the war which claimants’ side fought was just in toto: one would also need to know whether the specific campaign within the war during which they were harmed was just, and whether they themselves committed war wrongdoings which may well cancel out their reparative demands. I mentioned earlier the case of combatants who were wounded by the Rwandan Army as they themselves committed atrocities against Hutus civilians, and averred that they do not have reparative claims—whereas their victims did. The problem is that the victims of those atrocities might themselves have been genocidaires, and might have violated the rights of those combatants or their families. Who, then, if anyone, has reparative claims in such cases?
Those obstacles are not purely practical. To be sure, if we were ever in a position to know relevant facts about claimants which would support or on the contrary undermine their claims for reparations, we would be able to implement the Reparative Principle. But given that we do not know those facts and are unlikely to be able to acquire such knowledge, we have to decide whether to withhold reparations altogether, at the likely cost of not helping individuals who in fact have a claim to reparations, or to replace the Reparative Principle with another principle, at the likely cost of giving reparations to individuals who in fact do not have a claim to them, and in so doing withdrawing from the reparative pot resources which could and should go to genuine claimants. The decision we have to make— of which moral cost we should bear—is of course an eminently moral one. It is also eminently complex. To answer it, however, we need to have a clearer picture of who owes reparations, and why.
-  See J. Feinberg, ‘Voluntary Euthanasia and the Inalienable Right to Life’, Philosophy & PublicAffairs 7 (1978): 93—123. For discussions of this particular example, see, e.g., P. Montague, ‘Rights
-  See, e.g., Barkan, The Guilt of Nations, ch. 1; Colonomos and Armstrong, ‘German Reparationsto the Jews after World War II’.
-  Kamm, TheMoralTarget, 158—63. I defend that view in CW, s.5.5.