Double Effect And Civilian Immunity

Below are three paragraphs from an edited anthology on civilian immunity, Civilian Immunity in War. These particular comments are from Colm McKeogh in his chapter “Civilian Immunity: Augustine to Vattel” (80-82). Here he offers three problems with the use of double effect to justify attacks which are foreseen to have collateral damage. Now, I think that he goes too far when he implies that there could never be a situation that could be free from the problems he raises, but what his points do establish, I think, is that the bar for permitting actions of the type he is discussing is very high, much higher than modern armies are willing to concede.

First, the PDE requires that the good end sought be proportionate to the harm done. In an attack on a military target in which (it is known) civilians will be killed, what is the good end that is to be proportionate to the harm that is the killing of civilians? Is it the saving of combatant lives that would otherwise be lost if such an attack did not take place, or if alternative means were used to the same tactical end? It is not (and in any case, civilians ought not to be killed as a side-effect of an action to save one’s own combatants, for combatants may be treated as instruments, but civilians remain persons; it is not the case that two combatant lives saved outweigh one civilian life lost). Rather, the attack on the military target is only an intermediary step, a means to the end of victory in the war. Its worth cannot be assessed without reference to the ad bellum end for which the war is waged by that party to the conflict. Yet, if the focus is switched from the tactical to the strategic level, then the end for which civilians are killed collaterally is not so clearly good or right; indeed, one party to the conflict firmly believes that end to be wrong or bad (in other cases to which the PDE is applied, the good end is clear: in the performance of a craniotomy, it is the saving of a woman’s life; so too when a bystander pulls a lever to divert a run-away trolley on to a track where it will kill only one person rather than five: five lives are saved and only one lost).

Secondly, not only is the goodness of the ultimate end in doubt; so too is the probability of attaining it. Again, this distinguishes war from other cases: in the craniotomy and trolley cases, the good or right end that is sought can be attained. The surgical intervention will save the woman’s life at the cost of her pregnancy; pulling the lever will save five lives and lose only one. The probability of attaining the good end is much lower in war. Chance, as Clausewitz asserted, is essential to the nature of war; in war, no good at all may come from the harm that is done. In fact, no good at all will come from the harm that is done by one side— the side that loses; at least one side sacrifices the lives of civilians for no good or right outcome: civilians killed collaterally by the losing side are, ultimately, killed for nothing.

There is a third difficulty in applying the PDE to tactical operations in war that cause civilian deaths. This is that the sum of civilian deaths in such tactical engagements can exceed the maximum believed to be proportionate to the war’s end. This is because war is a serial killer, a chain of violent events. Prior to each tactical operation in war, the question facing the military command concerns only the civilian deaths expected from the proposed future engagement. The PDE asks whether the next loss of civilian life would be outweighed by the expected good end. Past non-combatant deaths are past; civilians’ lives already lost are out of the equation. If the expected good end has taken longer to achieve than initially anticipated, a disproportionate number of civilian lives may already have been lost. Yet the question facing the military command remains whether the loss of civilian life anticipated in the next operation is outweighed by the expected contribution to the war’s end of the tactical operation. The anticipated loss of civilian life in each tactical event will nearly always be outweighed by the strategic good that is the ad bellum end of the war. But because war is a chain of such events, the cumulative loss of civilian life can become disproportionate to the strategic end of the war.