Writing in this journal in 2013, Daniel Brunstetter and Megan Braun have suggested that some forms of interstate violence—in particular, uses of force that fall short of full-scale war—cannot be judged by the principles that are commonly used to evaluate war.Footnote 1 They argue that such uses of force should instead be evaluated with reference to a distinctive set of principles that, following Michael Walzer, they call jus ad vim.Footnote 2 In this essay I argue that such a proposed set of principles is redundant, and that the jus ad vim project stems largely from an implausible understanding of the principles of jus ad bellum.
I begin by outlining Brunstetter and Braun's arguments in favor of jus ad vim. I then go on to show that each of these arguments fails and that, properly understood, the traditional principles of jus ad bellum do the necessary work in restricting and permitting force. I conclude by considering Brunstetter and Braun's claims about the implications of jus ad vim for the use of drones, arguing that, contra Brunstetter and Braun, jus ad vim does not better capture the harms and goods relevant to uses of limited force, and does not better adhere to the requirement of discrimination.
The Argument for Jus ad Vim
Brunstetter and Braun (hereafter also referred to as “the authors”) distinguish between forceful measures employed as part of war and forceful measures “short of war.” Specifically, they argue that forceful measures short of war, such as the CIA's drone use in Pakistan, cannot be properly evaluated using the jus ad bellum principles familiar from traditional just war theory—that is, just cause, proportionality, last resort, legitimate authority, and a reasonable prospect of success by legitimate means.Footnote 3 Rather, they propose a new, additional set of principles—jus ad vim or “justice before force”—to govern these measures that fall short of war. Jus ad vim allegedly evaluates the gap between domestic law enforcement and war, and is described as a “recalibration” of jus ad bellum principles, plus a novel principle concerning the probability of escalation. By “recalibration,” the authors mean that we might use the same notions as we employ in jus ad bellum—just cause and proportionality, for example—but apply them differently in the context of jus ad vim.
Brunstetter and Braun describe jus ad vim as in some respects “more permissive” than jus ad bellum.Footnote 4 While jus ad vim retains the ad bellum position that self- and other-defense are the only just causes for force, it employs a looser understanding of “defense”:
This means that there are more cases in which injuria justifies some turn to force, but not necessarily war. Within the context of jus ad vim, a state has just cause to use measures short of war when responding to injuria against its interests or citizens. This includes responding to terrorist bombings, attacks on embassies or military installations, and the kidnapping of citizens. These are acts of aggression that justify the right to a forceful response . . . . If a state that has WMD is on the brink of collapsing into a failed state, or if a state is about to use such weapons on its own population, then jus ad vim acts are legitimate.Footnote 5
The authors also suggest that jus ad vim has a “more favorable proportionality calculus” than jus ad bellum. Since it admits of a broader range of causes, there is a broader range of goods that can be allowed to weigh against the harms caused by using force. However, in terms of the measures it permits, jus ad vim is more restrictive than jus ad bellum, since it sanctions only the use of measures short of war.
Brunstetter and Braun stress that jus ad vim is not simply tantamount to the requirement of last resort: the claim is not that one must use measures short of war before one can justify resorting to war. Rather, jus ad vim “should serve as an alternative set of options to the large quantum of force associated with war. This stems from the essence of jus ad vim—its advantage in avoiding the unpredictable and widespread destructive consequences of war.”Footnote 6
With this in mind, the success of ad vim measures is determined by whether they do in fact avoid war. This is why the authors suggest a new principle for jus ad vim that they say is not included in jus ad bellum: the probability of escalation, “defined as the elevation of hostilities to war.”Footnote 7 If it is likely that the use of measures short of war will trigger an actual war, “one could argue that such actions are not justifiable, and must be subject to the stricter jus ad bellum regime.”Footnote 8 They suggest that the category of jus ad vim “provides both analytical clarity and a more nuanced appreciation of the potential ethical pitfalls of drone strikes.”Footnote 9
It might be unclear why Brunstetter and Braun are taking jus ad bellum to be the relevant set of comparison principles, rather than jus in bello. Ad bellum proportionality is about how much harm one is permitted to do to achieve the just cause of the war, but it is not sensitive to the kinds of weapons one will use to achieve that cause. Imagine, for instance, that it would be proportionate to some just cause to kill and seriously injure six hundred people, destroy a thousand buildings, and cause the concomitant disruption of civilian life. Clearly, this limit holds irrespective of how one will cause those harms. However, I think the authors’ reasoning must be that there are cases in which only drones will keep a military action within those limits, and so it is specifically the use of drones that enables the action to satisfy ad bellum proportionality—that is, the resort to force. Note, though, that the use of drones will not affect what it is ad bellum proportionate to do, so much as determine whether one can fight within the limits of what has been independently judged to be proportionate. I do not think this affects the general argument I am making here, which is that the proposed category of jus ad vim fails on two counts. First, it is badly motivated: the apparent shortcomings of jus ad bellum that Brunstetter and Braun identify as creating a need for jus ad vim are merely apparent. Second, the apparent advantages of jus ad vim for evaluating drones strikes are similarly illusory.
The (Alleged) Failings of Jus ad Bellum
Judging Jus ad Bellum
Brunstetter and Braun interpret the jus ad bellum conditions as a one-off judgment, made at the outset of a war. For example, they say that if the drone strikes in Pakistan are part of an ongoing war against the Taliban, “the criteria of jus ad bellum would presumably already have been satisfied, with only the jus in bello rules applying”; and later, “whereas in war, principles such as just cause and last resort need only be satisfied at the outset of a conflict, jus ad vim requires that they be continually reassessed in advance of each use of force.”Footnote 10 They also approvingly cite James Pattison's claim that the just war tradition “lacks the conceptual tools to consider the morality of an intervention that was permissible when it was launched but that later becomes morally problematic.”Footnote 11
But all these claims are mistaken, at least according to any plausible account of jus ad bellum. Meeting the conditions of jus ad bellum is not a single judgment made before the start of a war, but rather an ongoing judgment that must be made throughout the war.Footnote 12 We can see this most clearly with respect to the first condition that the authors mention, namely, just cause. One may not continue to fight a war after its just cause has been secured—not least because in the absence of just cause, no offensive can be militarily necessary for securing the just cause, and thus any fighting is impermissible. But the other ad bellum conditions must also be satisfied throughout the war. If, for example, it becomes apparent during the course of a war that one has no reasonable prospect of success, one may not continue to fight. Nor may one continue to fight if war has become unnecessary—if, say, the aggressor offers terms of peace that will secure the just cause.
The same is true of a war that becomes disproportionate to the just cause. Imagine that an invading force tries to take control of a sparsely occupied piece of our territory. Our evidence suggests that the retention of the land will cost us around three hundred combatant lives and cause serious collateral harms to three hundred innocent people. Say that this is at the limits of what would be proportionate for retaining the land, but it is proportionate, and so we begin a defensive war to repel the invasion. But once the war begins, we realize that the aggressors have been secretly developing drones that enable them to drop bombs on our combatants without endangering their own troops. In light of this new evidence, we realize that the only way to retain our land will be to mount a ground and air offensive of the invading nation's own territory, taking out the military bases from where they operate the drones. We estimate that the cost of achieving this will be around two thousand combatant deaths, plus serious collateral harms to two thousand innocent people on both sides, along with the concomitant disruption of civilian life in both countries. This is disproportionate to the just cause—that is, at the bar of the ad bellum conditions—and thus our war is ad bellum unjust and we may not continue to pursue it.Footnote 13
With this better understanding of jus ad bellum in hand, we can see that Brunstetter and Braun are mistaken to think that jus ad vim differs from jus ad bellum by requiring us to continually reassess our use of force. Properly understood, jus ad bellum also requires this continual reassessment.
Escalation
Brunstetter and Braun are correct that when we are considering using force we ought to take into account the risk that doing so might escalate hostilities. It is implausible, however, to think that this is a new principle absent from jus ad bellum. Considerations of escalation come under the remit of the proportionality requirement. As David Rodin has argued, we are morally responsible not only for harms that we directly cause but also for harms that we foresee we will trigger others to cause.Footnote 14 In other work, I have referred to these as “mediated harms” to capture the fact that they are harms that the agent triggers, but they are mediated through the (wrongful) agency of someone else.Footnote 15 Consider the following examples of Grenade and Provoke:
Grenade: Bully wants to painfully pinch Victim's arm. Victim can dissuade him from doing so by throwing a small object at Bully that will slightly wind him. However, the only small object Victim has is a grenade. When the grenade deflects off Bully, it will detonate next to innocent Friend, killing her.
Provoke: Bully wants to painfully pinch Victim's arm. Victim knows that if he resists, Bully will become so enraged that he will kill innocent Friend.
It is clearly impermissible for Victim to defend himself in Grenade. This is a standard case of causing disproportionate harm, because the good that Victim hopes to achieve is greatly outweighed by the harms he foresees his defense will cause. It is also clearly impermissible for Victim to defend himself in Provoke. Here, too, his defense is disproportionate, because here too the good that Victim hopes to secure is greatly outweighed by the harm he foresees his defense will cause to Friend.
Rodin and I agree that there are differences between Provoke and Grenade. When the foreseen harms of defense are mediated through someone else's agency, as in Provoke, they are heavily discounted in the defender's proportionality calculation. They do not count as much as harms that Victim will directly inflict, as in Grenade. But mediated harms do still matter for proportionality, and when there is great disparity between the good to be secured and the harm one foresees, mediated harms can make defense impermissible.Footnote 16
If these foreseen harms count toward making defense disproportionate (and thus impermissible), then states thinking of using defensive force must, as Brunstetter and Braun argue, consider how likely using force is to escalate a situation. But we can see now that to make this claim is not to add a new governing principle to the use of measures short of war, giving us something distinctive about jus ad vim that is missing in jus ad bellum. Rather, this consideration is part of the proportionality requirement that obtains across the board: it matters for individual self-defense, as in Provoke, and it matters for national defense—regardless of whether the force employed counts as part of war.
I think that jus ad vim, much like traditional collectivist approaches to war, places unwarranted weight on whether something counts as war. The foregoing discussion of escalation is a case in point. Should our concern really be whether measures short of war escalate into war? Or whether they escalate into more serious—and disproportionate—uses of force? It seems to me that what matters is whether we predict that our use of force will cause a situation to escalate until it causes harm disproportionate to the just cause. It does not matter whether this means that we have escalated to a state of war. It would be impermissible to engage in defense if one predicted it would escalate to disproportionate harms even if the types of measures involved remained short of war. Of course, if by war we mean “very widespread and serious harming,” it is obviously worse if things escalate to that point rather than, say, to ten more drone strikes than are proportionate to securing the just cause. Nonetheless, both are impermissible and, pertinent to my argument here, both are prohibited by a single proportionality condition that governs the use of force.
Rejecting the significance of war also circumvents a difficulty that the authors discuss at some length, but do not resolve—namely, how to determine whether something counts as a war.Footnote 17 This difficulty of course applies not only to proponents of jus ad vim but also to those writers, such as Michael Walzer and Henry Shue, who defend the view that war is morally distinctive.Footnote 18 Having moral permissions that obtain only in war provides various perverse incentives for fighters, especially if scale is a factor in determining what counts as war. For example, limiting one's destruction to what is necessary to achieve one's end might keep the conflict below the threshold, thereby excluding one's fighters from combatant privileges. Increasing the scale of the conflict, in contrast, could get it over the threshold to count as war, even if doing so is not necessary or proportionate for achieving one's end. Of course, such a war would be unjust. But for those who tend to view war as morally distinctive, this would not prevent the fighters from enjoying special permissions to do violence. Denying that violence becomes more easily justifiable during war removes these kinds of incentives, and also solves the problem of deciding whether a use of force should be judged by ad vim or ad bellum principles.
Implications for Drone Use
Brunstetter and Braun allege two important implications of jus ad vim proportionality for the CIA's use of drones outside of war. The first is that because the just cause will be less important in cases where war is not justified (and thus any harms caused in pursuing the cause are governed by jus ad vim), less collateral damage will be permitted by the ad vim proportionality conditions than would be permitted under the ad bellum proportionality requirement: “Insofar as the military advantage is smaller, the collateral damage is less justifiable, which means the threshold of acceptable excess destruction—the killing of civilians—is lower than what is allowed in war.”Footnote 19
The second alleged implication is that ad vim's “much more restrictive” interpretation of proportionality means that when using drones, we must “be concerned not only with the loss of civilian life but also the more subtle harms including property destruction, post-traumatic stress disorder, and social disruption.”Footnote 20
Both these alleged advantages of jus ad vim are again undercut by a better understanding of jus ad bellum proportionality. The reason why the “threshold of acceptable excess destruction” is lower “than what is allowed in war” is not that we are using measures short of war and therefore our actions are governed by a different proportionality calculation. Rather, it is that, precisely as the authors say, the just cause being pursued is (by hypothesis) less important and thus warrants less collateral harm. In other words, it is just ordinary proportionality determining what may be done. It does not matter whether it is ad bellum or ad vim; all that matters is the relevant good and the relevant harm.
Noticing this helps explain why the claim about escalation quoted above is confused in two ways. Brunstetter and Braun suggest that when a use of force is likely to cause a situation to escalate to war, “such actions are not justifiable, and must be subject to the stricter jus ad bellum regime.”Footnote 21 But that an action will cause escalation does not show that it is unjustifiable. All it shows is that, according to the authors, it cannot be justified by the principles of jus ad vim. But nothing follows from that for the justifiability of using force. This is like observing that an account of proportionate punishment does not sanction killing a would-be rapist in self-defense. That might be true, because accounts of punishment neither sanction nor prohibit defense. But it tells us nothing about whether killing to avert a rape is justifiable, only that accounts of punishment cannot judge the permissibility of defense—a conclusion that is not philosophically interesting. Furthermore, if what I have argued thus far is correct, moving from jus ad vim to jus ad bellum is not going to involve applying “stricter” proportionality conditions. The increased force will be unjustified if the good sought does not warrant (the risk of) escalating beyond a certain amount of harm. This judgment is unaffected by whether the force to which we might escalate counts as war.
I think this also speaks to the idea that jus ad vim expands the range of just causes compared to jus ad bellum. One can see how this might seem true if, for example, one begins with something like Jeff McMahan's claim that “just causes for war are limited to the prevention or correction of wrongs that are serious enough to make the perpetrators liable to be killed or maimed . . . . The only ends that weigh against the bad effects of war in the proportionality calculation are those specified by the just cause or causes for war.”Footnote 22 Brunstetter and Braun seem to be saying that, unlike in the ad bellum proportionality calculation, jus ad vim proportionality includes other ends—ends that would not justify war. But this is not a different interpretation of proportionality. If what McMahan says is right, it is still true of ad vim that only those goods that warrant the harms inflicted will be able to weigh against those harms. Insofar as jus ad vim is dealing with measures that kill or maim, it too can include only those ends that warrant killing or maiming. McMahan's claim is not that there is something special about waging war per se, which means that only certain ends are relevant to proportionality. His claim is rather that only those ends that warrant force are relevant goods for the purposes of judging the proportionality of force. So jus ad vim is not broadening the range of ends that may be pursued by force. It is simply suggesting that some ends do not warrant, say, vast amounts of killing or destruction, but could warrant some killing or some destruction.
Relevant Harms
We can, contra Brunstetter and Braun, capture more than the mere loss of life in the ad bellum proportionality calculation. Ad bellum proportionality (and in bello proportionality) is sensitive to a range of goods: it covers damage to infrastructure, damage to civilian property, environmental damage, and damage to cultural buildings and objects—and these are just the things listed by the Hague and Geneva Conventions.Footnote 23 It seems very plausible that as a matter of morality (that is, beyond what might be required by legal conventions), proportionality is also sensitive to psychological harms and the general disrupting of civilian life. At any rate, any argument for including them in proportionality calculations for uses of force short of war will speak equally to their relevance for harming during war.
Of course, these harms might not weigh very heavily when one is dealing with just causes for war, because (by assumption) just causes for war are terribly important. Social disruption is not irrelevant to proportionality, on this view, but it is not going to count for much compared to stopping a genocide, violent invasion, or wave of terrorist attacks, and so on. When the just cause for using force is less important, however, such harms will be more able to counteract the good that can be achieved by inflicting them. But this is not to invoke some new, “much more restrictive” interpretation of proportionality. It is just plain old proportionality, according to which securing very important goods warrants foreseeably inflicting more serious harm; and securing less important goods can be prohibited by the foreseeable infliction of relatively minor harms, precisely as the authors envisage ad vim proportionality to work.
We can see this point even more clearly by considering the range of just causes that might justify war. It is not as if, once one has a just cause for war, any harm one causes in pursuit of that cause is justified. Rather, we still have a separate proportionality condition that must also be satisfied if the war is to be just, because even wrongs that meet the level of justifying war can vary in their magnitude. One may do more to prevent genocide than one can to prevent the theft of resources, or the damming of an important river, or the occupation of a small piece of land. (If we did not think that this is the case, there would be no ad bellum proportionality condition at all.) But this does not mean that there is one proportionality condition that applies to genocide, and another that applies to theft, and another that applies to the damming of rivers. There is just a single proportionality condition that requires us (to put it simply but, I believe, correctly) to weigh the good we expect to secure against the harms we expect to cause in a range of situations. The authors’ jus ad vim proportionality adds nothing here: the proportionality calculation for using force short of war is no more permissive, favorable, or restrictive than any other proportionality calculation, because there are no other proportionality calculations.
I suspect that part of this confusion about the harms that are relevant to ad bellum proportionality comes from a conflation of what policymakers and practitioners actually think about when deciding whether to use force, and what the moral principles of jus ad bellum enjoin them to think about. It may be true that, in practice, those making proportionality decisions pay attention only to the killing (and serious harming) of civilians. This could be because they have already decided that if an objective warrants killing anyone it also warrants a fairly generous amount of other kinds of harm—that is, they take the proportionality of those harms to be entailed by the proportionality of incidentally killing even one innocent person. If so, it would usually be the proportionality only of killing that merits discussion. This would be a mistaken view of proportionality, I think, but it is a possible and coherent view.
Alternatively, it could just be that policymakers and practitioners are mistaken about proportionality in a more fundamental way—that they genuinely do not consider these harms as relevant to proportionality at all. But this, too, is just a mistake, and we should not take these mistakes of policymakers and practitioners to reveal some flaw with the ad bellum principles. Philosophically, these failings do not in any way determine the content of the ad bellum principles. And practically, if policymakers and practitioners are not considering these harms when they think about proportionality, simply drawing up another set of principles that (also) tell practitioners to care about nonlethal harms is unlikely to be more successful in getting them to do so. If they ignore these aspects of proportionality in jus ad bellum, we should not expect them to pay attention to these harms when they are packaged as part of jus ad vim.
Discrimination
Brunstetter and Braun also claim that jus ad vim requires stricter adherence to the requirement of discrimination than jus ad bellum:
Scholars such as Walzer have argued that one can imagine a sliding scale that allows for violations of jus in bello principles under certain circumstances—notably supreme emergency—as well as the need, when thinking ethically, to distinguish between intended and unintended effects (in other words, about the doctrine of double effect). Sometimes necessity requires, as the argument goes, breaking the rules of noncombatant immunity, while other, less extreme circumstances occasionally result in unforeseen and nefarious consequences that do not outweigh the overall justness of the use of force . . . . However, in terms of jus ad vim the advancement of technology should make the consequences of the use of force more predictable, which, coupled with the necessarily limited scope of force, means that there should be no recourse to such moral loopholes. In other words, jus ad vim must maintain stricter adherence to the principle of discrimination than jus ad bellum.Footnote 24
It is not clear exactly what the authors have in mind by “unforeseen and nefarious consequences.” The natural reading would be to take this as a reference to collateral damage (if they are unforeseen harms, they are presumably also unintended harms). If they are unintended, however, they do not violate discrimination even if they befall civilians. They might, on an objective account, violate proportionality even if they are unforeseen (or, perhaps, unforeseeable). But to violate discrimination one must be aiming at civilians, not merely harming them. The improved predictability of force could of course limit instances of this kind of disproportionate force, but this would not be a form of greater adherence to discrimination. Moreover, insofar as these more technologically advanced weapons are available in war, they will presumably increase our ability to adhere to proportionality within war as well as outside of war.
Walzer's supreme emergency clause is, of course, a challenge to discrimination. Brunstetter and Braun's thought seems to be that because jus ad vim deals with force short of war, it is not going to encounter situations in which we might contemplate intentionally killing civilians, and thus the temptation to violate discrimination will be avoided. But this is to make a mistake about what jus ad vim is supposed to govern. It is supposed to govern forceful measures short of war—that is, a particular range of means of pursuing goals. It does not restrict the range of goals that may be thus pursued. I suspect this mistake arises because the authors have assumed that if there is a cause so important that it warrants targeting civilians, that cause would warrant war, and so we move beyond jus ad vim to jus ad bellum. But we could easily have a cause that would warrant war, the securing of which requires only measures short of war and that must therefore (on Brunstetter and Braun's account) be governed by jus ad vim (assuming that using force does not seem likely to cause escalation). If we must use a single drone to target a single civilian in order to prevent some kind of humanitarian catastrophe, the mere fact that all-out war (if it were necessary) would be justified to achieve that same end does not mean that measures short of war are suddenly to be governed under jus ad bellum. Given this, jus ad vim is really no better than jus ad bellum when it comes to supreme emergencies and adhering to discrimination.
Conclusion
Prima facie, the proposal to develop specific principles to deal with instances of defensive force that lie between the measures typical of law enforcement and the measures typical of war seems an attractive one. But its appeal is illusory: the putative differences between jus ad vim and jus ad bellum that supposedly warrant a new category of principles rest on an implausible understanding of jus ad bellum. Satisfying the jus ad bellum conditions is not a one-off judgment. Rather, it requires continuous assessment throughout a war. Requiring agents to consider the risks of escalation is not a distinctive, independent element of jus ad vim, but rather part of any proportionality condition, since we are responsible not only for the harms that we directly cause but also for the harms that we foresee our defense will trigger others to cause. Ad bellum proportionality is also sensitive to the range of harms that are to be included in the ad vim proportionality calculation. How heavily these harms will count against causes for war remains open, but that is because causes for war are very important, not because ad bellum proportionality functions differently than ad vim proportionality.
The alleged implications of using jus ad vim to assess the CIA's drone use are not really implications for the permissibility of using drones at all. Ad vim gives us a label for judging measures short of war, but the category lacks any genuine normative role because (a) a use of force that fails ad vim proportionality will also fail ad bellum proportionality, since there is only one proportionality condition, and (b) a use of force that exceeds what can be judged by jus ad vim will not thereby be shown to be impermissible, but will simply be judged by jus ad bellum anyway. Thus, the proposed category of jus ad vim is redundant.
