How Should States as “Trustees of Humanity” Respond to an Armed Attack? Reflections in Light of the Gaza War 2014

The recent war in Gaza poses an important challenge to the claim that countries must act as “trustees of humanity” rather than as trustees of their own citizens only, and must therefore take the interests and concerns of affected foreigners seriously into account. This claim is reflected in several aspects of international law but not, at least not sufficiently, in the laws that regulate warfare. The traditional laws on the use of force (jus ad bellum) and on the conduct of hostilities (jus in bello) impose several restrictions on states engaging with their enemies, but what informs those traditional requirements and prohibitions is the premise that states are entitled to ensure maximum safety for their citizens regardless of the costs to all the rest.

This traditional premise is reflected in the scope of the right to self-defense. Once a state is faced with an “armed attack,” it is entitled to resort to whatever force is necessary and proportionate to repel the attack (and, some would say, also to deter recurrent aggression). In that quest, whatever costs are imposed on the other side (including on its civilians) do not enter the proportionality calculus because the right of the defending state to ensure its own security is axiomatic. The same applies to the conduct of war: no limit is set on the killing of enemy soldiers and foreign civilians who take direct part in hostilities, unless, of course, they surrender or are wounded. There is a prohibition on directing an attack against civilians, and an attack on military targets must be aborted if it “may be expected to cause incidental loss of civilian life … which would be excessive in relation to the concrete and direct military advantage anticipated.” This proportionality calculus calls for a risk-benefit analysis that takes the “military advantage” as given. It does not require an assessment whether there were alternatives to secure the same advantage, perhaps even by nonmilitary means.

This strict risk-benefit analysis is exemplified by the U.S. Army’s 2012 No-Strike and the Collateral Damage Estimation Methodology, which explicitly refers to the commanders’ duty to “balance the dynamic relationship between three principal categories of risk: risk to mission, risk to forces, and risk of collateral damage.” The traditional approach posits that the legitimacy of the mission and the need to ensure its success are not affected by this proportionality calculus: “The intersection of the three [categories of risk] is determined by the weaponeering restrictions required to reduce collateral damage to an acceptable level while achieving mission success and minimizing operational risk to forces” (my emphasis).

These time-honored tests reflect a state-centric vision of traditional international law, a vision according to which states are the ultimate sources of authority and accountability, and each is responsible to its citizens and only to its citizens. But in our contemporary interdependent world that vision is no longer tenable, neither morally or legally. If sovereignty is not only a right, but instead entails, in Max Huber’s words “as corollary a duty: the obligation to protect within the territory the rights of other States”; and if, as Michael Reisman suggests, the sovereign “has important obligations to other states which are the very basis of its claim to territorial sovereignty”; and if these obligations encompass not only foreign governments but also, and indeed primarily, foreign individuals, than a state should take into account the costs it imposes on foreign civilians as it seeks to safeguard its own citizens.

From this perspective, it is rather startling to find that the traditional tests of the laws on war, particularly the disjunction between the ad bellum and the in bello aspects of military conflict, sanctify state security regardless of the human cost. The separation between the ad bellum proportionality test (“whatever is needed to repel the attack”) and the in bello proportionality (which assumes that the military advantage is legitimate) facilitates the disregard of the distributive question: whose civilians bear the burden of the defender’s attaining the military goal? The ad bellum assessment doesn’t ask how many civilians will be killed by the defending forces; the in bello assessment doesn’t question what the latter are fighting for and why. Neither of the tests takes account of the civilian costs. In contemporary asymmetric wars where military goals are elusive and “victory” never promises absolute peace, raising security on a pedestal is a recipe for ignoring civilian losses.

A trusteeship vision of sovereignty introduces a different type of proportionality assessment, one that characterizes the ubiquitous risk-benefit assessments that governments must undertake when they limit the rights of individuals subject to their authority. Such means-ends proportionality assessments include an inquiry whether the goal pursued reflects a compelling public interest. In the context of warfare such an assessment must mean that the ends of the fighting must be also justified. In David Luban’s words, “military necessity itself requires taking civilian interests into account.” A trivial goal will not justify the infliction of harm on others. Moreover, a trusteeship vision of sovereignty requires commitment to a rigorous decision-making process for identifying targets and for exploring options to reduce collateral harm, even by nonmilitary means.

The adoption of such a public law vision and the recourse to a means-ends proportionality assessment is pertinent especially in the case of asymmetric warfare, as regular armies fight non-state actors who have no commitment to the civilians in their midst and who are at best indifferent to their suffering. The military asymmetry calls for extending the regular army’s responsibility to take into account also the foreign civilians who have nobody else to count on for protection.

Adopting the means-ends proportionality approach calls for merging the ad-bellum and in-bello proportionality assessments. The trustee sovereign would have to explain whether the reason to continue fighting justifies “excessive” civilian losses. In fact, such an approach may be already reflected in certain formal reactions to the military confrontations in Gaza. In a previous round of violence, when asked whether the United States viewed Israel’s attacks as disproportionate, the U.S. Ambassador to the United Nations conflated the two types of assessments:

“[W]e believe that Israel has the right to defend itself against [the Hamas] rocket attacks and we understand also that Israel needs to do all that it can to make sure that the impact of its exercise of right of self defense against rockets is as minimal and no affect [sic] on the civilian population.”

Admittedly, in 2014 key speakers again assumed that the right to self-defense is axiomatic (e.g., the spokesperson for the United States Department of State: “we certainly support Israel’s right to defend itself against these attacks. We appreciate – we’re concerned, of course, about the safety and security of civilians,” and UN Secretary General Ban Ki-Moon stating that “no country would accept the threat of rockets from above and tunnels from below,” but adding that Israel was obliged to protect civilians). But the very mixing of the two questions keeps suggesting that self-defense considerations are relevant also to the in bello proportionality calculus.


How to value foreign lives versus domestic lives during warfare?

The comprehensive means-ends proportionality assessment requires weighing the comparable value of foreign lives against those of one’s own citizens as well as other pertinent concerns. How could this be done?

During the recent war in Gaza the moral philosopher Frances Kamm stated:

“To do this proportionality calculation, we would have to estimate how many Israeli civilians would be killed if the attacks on Gaza did not occur relative to the number of Gazan civilians that would be killed if the attacks occurred. The fewer Israelis that would be saved and the more Gazan civilians that would be killed, the less likely the attack is to be proportionate, at least to the goal of preventing harm to Israelis.”

Note that Kamm does not necessarily imply that the proportionality ratio should be 1:1 or any other fixed number. Does fixing a ratio make sense?

Any ratio-based proportionality analysis would seem to be too crude a tool to assess proportionality. Even if we had a ratio to work with, implementing it would be quite complicated. Any assessment of the relative risks must entail various unknowns and probabilities (how many lives will be saved if we target the head of a terrorist organization that is hiding among civilians? What if the terrorists aim at strategic assets? Are we defending our survival as individuals, perhaps even the survival of our nation, or a much more modest threat?). Moreover, it is necessary also to determine the proper weights to assign to the evidence (quantity, reliability) to apportion the risk of error in conditions of uncertainty that could result in harming or saving civilians, both immediately and later on.

The complexity of the task is exemplified by the detailed procedure that was adopted by the U.S. Army mentioned above. The instructions on collateral damage estimation methodology (CDM) state that “the CDM is not an exact science,” and although it “follows a rigid process and generates estimated values, neither analysts nor commanders should be under the impression that these values in any way constitute ground truth, an exact science, or flawless data.”

Moreover, beyond saving lives and other personal harms, there are other legitimate governmental interests that states seek to ensure and therefore need to be taken into account: bringing criminals to justice; deterring others; restoring authority. This is the reason why even if the two groups of threatened civilians belong to the same state (envision, for example, a scenario where terrorists are holding captives in one part of a town while shelling those in another part) one cannot expect the local police to sit on the sidelines as an impartial third party and refrain from action due to moral argument that “the duty not to kill is stronger than the duty to save.”

In the inter-state context there are also collective rights at stake, primarily the right to national self-determination, that the aggressor may be challenging and that the defender is entitled to protect. As the trustee primarily of its own people, the defending state in no way resembles a third party that must impartially allocate the burden of the risk caused by the aggressor. In a world of states which recognizes states’ positive duties primarily to their own citizens, the state’s obligation to “ensure” the lives of its citizens may weigh heavier than its duty “to respect” the lives of foreigners.

Due to the complexity of defining the proper ratio and implementing it in actual combat, It would seem more effective, even if not more accurate, to give the proportionality test a procedural framing, following David Luban who envisions an “audience sceptical enough of the military … [which] gives equal weight to the lives and other interests of all sides.” Any ex-ante decision of a military commander will internalize its obligation to “take all feasible precautions … with a view to avoiding, and in any event to minimizing, incidental loss of civilian life” if she can expect a “sceptical audience” to second-guess its performance ex-post. Such scrutiny will be able to identify whether there was a failure to take civilian interests seriously into account. Such a scrutiny is necessary as the war in Gaza ends.


Tolerating aggression? or why moral philosophers must take game theory into account

Jeff McMahan not only suggests an answer to the ratio question, but also argues that there must be a threshold for responding to an armed attack, even one which kills many civilians. McMahan believes that a ratio of 1:2 should be set, namely that risking foreign civilians is legitimate only when the attack is likely to save at least double the number of the defending force’s civilians. Until that threshold is reached, no military response risking foreign civilians is justified. McMahan’s argument extrapolates from the situation of individual self-defence:

“[because] individuals have no ‘self-preference’ permission to kill an innocent bystander as a side effect of self-defence, and [because] third parties may cause no more harm to innocent bystanders in defending others than those others would be permitted to cause in their own defence, … it would not be permissible for Israeli soldiers to kill two Palestinian civilians, or even a single Palestinian civilian, as a side effect of preventing the killing of one Israeli civilian. … assuming that the duty not to kill is stronger than the duty to save, the number of Israeli civilians that would be saved must presumably be greater than two.”

It is doubtful whether it is possible to extrapolate from the exceptional one-off situation of individual self-defense to the recurring aggression to which public authorities respond. Asymmetric warfare is an indefinitely recurring bargaining act, and no actor can hazard signaling weakness or lack of resolve. So any practice that involves comparing lives must take a long-term view and factor in the probability that a timid response might generate more aggression and more losses later on. The weakness of McMahan’s argument becomes clearest in the context of responding to recurring acts of terrorism. Tolerating acts of terror encourages them and even invites harsher attacks if, as is often the case, the purpose of the aggressor is to provoke its target into action. The lack of response will only encourage the aggressor to initiate, continue, and even intensify its attacks, until such time as the defender’s losses reach McMahan’s ratio of justified response. By then too many people will have lost their lives in vain.

As von Clausewitz famously noted, “war is simply a continuation of political intercourse, with the addition of other means.” If we want to suppress recourse to such “other means” the law must strive to ensure that they are costly to every actor that contemplates their use.



More on the Gaza War:

Eyal Benvenisti and Georg Nolte, Der ungleiche Krieg, Frankfurter Algemeine Zeitung, 7 August, 2014 (in German)


More on trusteeship for humanity and international humanitarian law:

Eliav Lieblich and Eyal Benvenisti, WPS 2014-10: The Obligation to Exercise Discretion in Warfare: Why Autonomous Weapon Systems are Unlawful


More on valuing foreign lives (also in regulatory context):

Arden Rowell and Lesley Wexler, Valuing Foreign Lives (forthcoming 2014)

Gregory S. McNeal, Targeted Killing and Accountability 102 Georgetown L.J. 681 (2014)

David A. Dana, “Valuing Foreign Lives and Civilizations in Cost-Benefit Analysis: The Case of the United States and Climate Change Policy” (2009).

Eyal Benvenisti Quixotic Altruism or an Effective Investment? On Unilateral Measures to Mitigate Climate Change