John McGinnis, a Northwestern University law professor (and my coauthor on several papers), raises this question in the process of recounting a deliberate violation of the Geneva Convention by the US military during World War II:

I begin with some interesting information that I recently learned on a trip to the Chicago Museum of Science and Industry...: the United States once decided, consciously and at the highest levels of government, to violate the Geneva Convention. The incident occurred when in 1944 our navy captured a German submarine and a version of the enigma machine—the equivalent of the German codebook. While the navy also captured the entire crew, it did not notify the Red Cross of their capture or identities. The exhibit itself states that this failure violated the Geneva Convention. Notification would have tipped the Germans off to the substantial possibility that the allies now had a means for breaking their code.

How, if at all, would this decision then differ from a decision today to violate some aspect of the Geneva Convention in order to achieve an objective as important as avoiding disclosure of capturing a code? Certainly, the violation in 1944 was clear and premeditated. It also cannot be argued that it was not certain to inflict very substantial harm. As a result, the relatives of the German seamen thought they were dead for a substantial period, and, I believe in at least one case, the wife of one of the sailors remarried. Such grief and disorientation of lives might be thought of even greater consequence than the humiliation of a few unapproved interrogation techniques. Would the decision be different because of a difference in the level of threat we now face? Who makes that determination?

This incident was just the tip of the iceberg of a vast number of Allied violations of international law during World War II. In the case of the United States, the most serious WWII-era war crime was probably the deliberate killing of hundreds of thousands of German and Japanese civilians during the strategic bombing campaigns of 1942-45. These efforts very likely violated international law; indeed, President Franklin D. Roosevelt had denounced as illegal the Germans' use of similar tactics on a much smaller scale in their bombing of Guernica during the Spanish Civil War.

Both the strategic bombing campaign and the far less serious incident recounted in John's post can be justified on the grounds that they were necessary to avoid an even greater evil: an Axis victory in the war. Defeat for the Allies would almost certainly have led to far worse atrocities than those we inflicted in order to achieve victory. Assuming for the sake of argument that the illegal strategic bombing campaign really was essential to the defeat of Nazi Germany and Japan, it is possible that it was justified (at least in part).

If so, however, that opens up two other vital issues. First, if even very grave violations of international law (such as the deliberate killing of hundreds of thousands of civilians) were justified in order to defeat the Nazis, it is possible that comparable or smaller violations might be justified if necessary to vanquish other evil regimes and/or terrorist organizations. Whether or not they are depends on how evil the enemy is, how great a threat they pose, and how necessary the violation in question is to the attainment of victory. The threat posed by, say, Al Qaeda, is far smaller than that posed by the Axis. But it nonetheless might be great enough to justify some (smaller) degree of illegality in our response.

Second, this line of argument should lead us to question the traditional view that international law - and the laws of war - should apply equally to all combatants regardless of the cause that they fight for. The difference - if there is one - between the German attack on Guernica in 1937 and the far bloodier Allied attacks on Berlin and Hamburg during WWII lies in the ends they served, not the means. The Allied cause was vastly more just than that of the Germans and therefore arguably justified the infliction of far greater harm (including, in some cases, harm to innocent people) if necessary to achieve its ends.

That said, there are three important caveats to the above reasoning:

First, in this post, I have deliberately abstracted away from institutional, slippery slope, and public choice considerations. Even otherwise justified violations of international law might lead to unjustified ones in the future and should perhaps be curtailed for that reason. The risk of slippery slope problems, public choice problems, and the institutionalization of atrocities, argues for tighter constraints on combatant behavior than might otherwise be justified.

Second, even if we can sometimes justify violations of international law that are needed to achieve victory in a just war, there is no defense for gratuitous atrocities that have no connection to military necessity. Even if the Anglo-American strategic bombing campaign was defensible, there is no justification for such Allied war crimes as the rape of some 2 million German women by Soviet soldiers.

Finally, I emphasize that the above references to Allied WWII-era war crimes in no way justify the considerably greater war crimes of the Axis powers. The latter are all the more reprehensible because they were committed by regimes fighting for an unjust cause and in most cases had no connection to any military necessity.

In response to my earlier post on justifications for violating international law, some commenters point to the danger that "necessity" arguments can be used to justify virtually any illegal conduct. For example, one argues that:

Necessity is a poor defense for war crimes. Barring the truly gratuitous (like the Soviet rapes [of German civilians during WWII]), the violating power can almost always cobble a "necessity" argument.

The whole point of declaring conduct X a "war crime" is presumably to declare it off limits, even when one could get arguably good results thereby.

The argument is not without some force. Indeed, some Soviet officials even tried to justify the mass rape of German and other women by their troops by claiming that it was necessary to keep up the soldiers' morale. However, the fact that many necessity arguments for violations of international law are bogus excuses doesn't prove that all of them are. For example, there is pretty strong evidence that the illegal targeting of German and Japanese civilians by the Allied strategic bombing campaign in World War II played an essential role in the defeat of the Axis (see, e.g., Richard Overy's book, Why the Allies Won). Perhaps the risk of allowing states to get away with bogus justifications is so great that a categorical ban with no exceptions is the only solution. In my original post, I emphasized that:

I have deliberately abstracted away from institutional, slippery slope, and public choice considerations. Even otherwise justified violations of international law might lead to unjustified ones in the future and should perhaps be curtailed for that reason. The risk of slippery slope problems, public choice problems, and the institutionalization of atrocities, argues for tighter constraints on combatant behavior than might otherwise be justified.

Such concerns justify a high burden of proof for necessity arguments, especially in cases such as the strategic bombing campaign where the illegal conduct causes great harm. They may also justify other institutional safeguards to reduce the risk of bogus necessity-mongering. Nonetheless, I am skeptical of claims that these dangers justify a categorical rejection of conduct that violates international law. In some important cases - including World War II - rigorous enforcement of such a rule would lead to far greater atrocities than the it prevents. For example, this rule would mean that outside forces could never intervene to oust a repressive regime engaged in mass murder so long as its crimes were confined to its own territory (because nondefensive war is considered illegal under the UN Charter).

We should certainly guard against bogus justifications for violating international law. But we should also guard against the sort of cult fetishization of international law that unthinkingly elevates adherence to international legal rules above all other considerations.