On the PBS NewsHour evening news broadcast of July 16, reporter Margaret Warner interviewed the Israeli Ambassador to the United States, Ron Dermer. She asked whether the then approximately 200 deaths of Gaza civilians (now estimated over 1000) as a result of Israeli attacks did not show that Israeli use of force is disproportionate. This is an argument that comes up frequently around the conflict in Gaza, and a major criticism of Israel.

Ambassador Dermer first claimed that some of those who spoke of disproportionate force were merely comparing the number of deaths on each side—for example, 200 civilians dead in Gaza with 1 Israeli civilian—to decide if the military action is proportionate and doing so, he said, involves “not understanding the rules of war.” He noted that many more Germans were killed in World War II than Americans, so the imbalance alone was no reason to say the war violated proportionality. “Proportionality,” he said, “has to do with something else entirely. . . . Always make a distinction between combatants and noncombatants. . . . We [Israel] do not deliberately target Palestinian civilians.” He contrasted deliberately targeting the civilians, as he said Hamas does, with causing deaths of civilians only as a side effect, as the Israelis sometimes do.

I found parts of the Israeli Ambassador’s response puzzling and worrisome.

What is Proportionality?

My first concern about his response is that proportionality in war is not about whether one deliberately targets or merely foresees and causes side effect deaths of civilians. Not deliberately targeting civilians either as an end or as a means to an end is a separate condition on the morality of war that Hamas may indeed be violating. Deliberate targeting of combatants or opponent’s military facilities is, of course, permissible.

Proportionality in the standard rules of war is about whether the harm that will be caused by military action is proportionate to the goal of the war or an individual military action. Most importantly, unintended side effect deaths of civilians are included in the harms that must be weighed against achieving the goal of the war or a military action to decide whether the use of force is proportionate. However, harm to enemy combatants, whether deliberate or produced as a side effect, is not standardly thought to count against achieving the goals of a military mission. This is one reason why military actions are not considered disproportionate if many more enemy combatants are killed than one’s own combatants. This is also why it is wrong to compare deaths of civilians on one side with deaths of combatants on the other on a one-to-one basis.

The use of force is not disproportionate merely because more civilians are killed on one side than the other.

In addition, as the Ambassador noted, the use of force is not disproportionate merely because more civilians are killed as side effects on one side than on the other side. But this is not because the deaths are not deliberate. Rather it is because the greater number of deaths on one side could still be proportionate to achieving the goal of military action and that is what the proportionality condition is about.

So while the Ambassador is correct in saying proportionality has something to do with the distinction between combatants and civilians, he is not correct if he thinks that it excludes counting deaths of civilians caused as a side effect and not from deliberate targeting. It is surprising that a high government official would not be aware of what the proportionality condition involves. It would be very serious if other government officials also were not aware of this and it is important that the public monitor this aspect of government awareness. It may be that 200 or more side-effect deaths of civilians (Palestinian or Israeli) are not disproportionate to the Israeli goals, or they may be disproportionate. The proportionality condition does not itself tell us what is proportionate or disproportionate to the goal a country is trying to achieve. However, it does tell us something about what ought to be considered in making that decision, and that can include civilian deaths that are not the result of deliberate targeting. So some of those who say Israeli force is disproportionate may not be comparing civilian losses on both sides but rather claiming that side effect civilian deaths are out of proportion to the goals being sought by the military action.

Permissibility and the Goals of War

However, that a war and actions in it meet the proportionality test in the rules of war is not sufficient for permissibility. This is because the proportionality test is applied independent of other tests that must also be considered. For example, suppose there is some other not too costly and much less harmful way to achieve the goals of war or actions within it. Then, given that proportionality only speaks to the relations between goals and harms, the fact that the more harmful way to achieve the goals is still proportionate does not mean that the more harmful way to achieve the goals would be permissible. This is because they are not necessary and so fail the necessity test in the rules of war.[i] (Furthermore, if proportional and necessary means that impose collateral harm on civilians are unlikely to be successful in achieving goals, this can also make them impermissible.)

One way of thinking of the highly successful Iron Dome antimissile defense system that Israel uses is that it is a less harmful way of achieving many of its goals by comparison to attacks on Gaza. However, it is not perfect and sometimes a missile breaches Israeli defenses. (Even if it were perfect in stopping missiles, it does not stop other infiltration mechanisms such as tunnels and it does not achieve the goal of getting others to recognize Israeli sovereignty.) Nevertheless, suppose, for argument’s sake, that breaches occur rarely and these breaches were the primary source of harm to Israelis. Then it seems that we would have to reconceive Israeli’s goal in attacking Gaza: the goal of the attack would not be to defend many Israelis from attack but only to defend a few from attack. Then it is relative to this “reduced” goal that we would have to decide if civilian casualties as a result of attacks on Gaza were proportionate. 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. Notice that comparing how many civilians on each side would be killed depending on whether the attacks on Gaza take place is not the same as comparing how many civilians have been killed. Hence, though it involves a comparison of expected deaths, it does not involve the sort of comparison of which the Ambassador originally complained.

Deliberate Targeting of Civilians

A final concern raised by the Ambassador’s remarks that I shall discuss here is what is meant by “deliberate targeting.” One of the claims repeatedly made by Israeli leaders is that Hamas uses civilians as “human shields” because it deliberately places rocket launchers in the midst of civilian areas, presumably when it could do otherwise. (Also, presumably, Israeli leaders are implying that Israel does not place its rocket launchers similarly.) Suppose the Gazan civilians really are used as such deliberate shields (or at least are unavoidably put in the position that a human shield would be put in). This suggests that if Israel were to attack the rocket launchers, it would kill civilians in a way comparable to shooting through a hostage attached to the front of an attacking tank; it would need its shots to occupy the same physical space as the human shield occupies, on the shot’s way to reaching the military target. This seems to be a form of “deliberate targeting” of the civilian shield rather than a side effect death, even though it is not as a civilian but merely as an occupant of a certain physical space that the civilian must be aimed at. (In another scenario, civilians might be placed inside the military equipment so that the shots need not pass through where they are in order that the equipment be reached, yet there is no way to precisely target the military equipment that does not include targeting the civilians. Elsewhere I have called this “intratarget killings”.[ii] By comparison, an ordinary side effect death would occur if the civilian was, for example, beside the launcher and not in it or blocking it.)

Hence, it seems that the more accurate the Israeli claim is that Palestinian civilians are used as actual human shields, the less accurate it is to say that those civilians are not being “deliberately targeted,” and the less the permissibility of attacks can depend on the claim that there is no deliberate targeting of civilians, contrary to what the Ambassador emphasized. Nevertheless, this should not, I think, necessarily imply that the attacks are morally impermissible. Suppose a tank with a human shield attached is headed toward killing people and the only way to stop the tank’s attack is to shoot through the hostage placed there by the person who runs the tank. It may well be permissible to shoot if this will stop the attack on many people. (The rules of war should take account of such cases in refinement of its condition on not targeting civilians.) On the other hand, if many civilians are attached as shields to the tank, and the tank would kill far fewer civilians were it to continue, it may violate proportionality to shoot to achieve the goal of stopping the attack. So again the calculation of expected civilian deaths that would occur on each side, depending on whether a military device is attacked or not, is necessary.

I have described some aspects of the proportionality condition as standardly understood. There are those who think it should be revised, for example so that deaths of combatants could also count in deciding whether use of force is disproportionate to the military goal. Further, in connection with the duty not to target civilians deliberately, some think there should be an additional requirement: to aim at not harming civilians rather than merely not aiming to harm them. So when Israel sends advance warning of its attacks on combatants or military facilities that may take civilian lives as a side effect, it shows that it is aiming at not harming civilians, at least if it gives them time to escape.[iii] If Israel did not do this, but merely did not deliberately target civilians, it would still meet the standard requirement not to aim at harming them. But, I have argued, its aiming not to harm or not aiming to harm would not yet tell us whether it satisfied the proportionality requirement, contrary to what Ambassador Dermer suggests, or several other requirements of the rules of war.


[i] Some of these points about what proportionality is and its being distinct from necessity were independently made by Michael Walzer during the 2009 Gaza war in his January 8, 2009 New Republic commentary. (I thank Thomas Scanlon for drawing my attention to that article.) It does not seem that the points have yet been absorbed by all government officials and hence bear repetition as well as further elaboration.

[ii] See my Ethics for Enemies: Terror, Torture, and War (OUP, 2011), Chapter 2.

[iii] Whether Hamas uses civilians as shields and whether Israel gives adequate advance warning was debated on the PBS Newshour on July 24.