A New York Times article this past Sunday noted a serious uptick in CIA drone attacks in Pakistan over the last few weeks. Ninety suspected militants have apparently been killed in eleven strikes since Dec. 30, 2009. Yet as the Times also reported, the ACLU and others have voiced concerns about these attacks. We are not at war in Pakistan, critics argue; the process for choosing targets is secret and suspect; the collateral damage of civilian deaths is too high. Yet the CIA contends its drone attacks are fully legal by the standards of international law. Who is correct?

First, we must analyze what is occurring in Pakistan. Is Pakistan engaged in an internal war, or is it just having a particularly hard time maintaining law and order?  Since 2004, the Pakistani army has been fighting tens of thousands of insurgents with a loose but effective command structure, control of territory, and a plentiful stockpile of weapons. This situation easily meets the legal criteria for an internal war. The targets of the attacks are organized and  their activities are distinct from banditry, short-lived insurrections, and isolated terrorist activities.

With this internal war in mind, we can analyze whether the U.S.’s very presence in Pakistan is legal. Stemming from the United Nation’s Charter, international law contains a general prohibition on the use of force, except in cases of self-defense against an armed attack or by Security Council authorization. One might argue that the U.S. has breached this rule by sending military force abroad and engaging in an international conflict without the permission of the Security Council. But this misuses the term “international.”

The U.N. Charter’s use-of-force rules apply only to international wars—that is, conflicts between states, not just on foreign soil. The Charter does not require the authorization of the Security Council before engaging in an internal war—a conflict within but not between states. Rather, it is within any state’s sovereign rights to oppose revolutionaries in its own borders with military force if necessary. Furthermore, another state is free to assist in such an internal armed conflict so long as it does so with the host state’s consent. The assisting state need not obtain any other international entity’s permission since this conflict is not international—it is not between states.

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

Does the U.S. have Pakistan’s consent to help with this internal war? Most likely yes, although this high-level diplomacy is often hidden by political maneuvering. There is evidence of a secret deal whereby Pakistan publicly complains about U.S. attacks to avoid a backlash from segments of its population while secretly cooperating with U.S. activities in the region. Also, it is widely reported that the drones have operated from bases within Pakistan. The totality of the evidence indicates that the U.S. drone attacks in Pakistan constitute consensual participation between these states against non-state internal warring groups.

Yet even if the U.S. drones are in Pakistan legally, to assist with an internal war, the law of war still binds the host nation and its allies as they engage in the conflict. The law of war traditionally requires military force to be used in accord with the principles of distinction, necessity, and proportionality. In other words, the attacking party must target only combatants, have no reasonable, less lethal alternative, and seek a goal proportionate to the damage it expects to cause. For this analysis, let’s turn to a specific and unusually tragic example.

The date is June 23, 2009. A crowd in South Waziristan, Pakistan has gathered to mourn the deaths of people killed in U.S. drone attacks. Intelligence suggests that Baitullah Mehsud, the leader of the Taliban in Pakistan, is among the mourners, so after the funeral prayers, two circling drones fire at least three missiles into the crowd. Around 45 militants and perhaps as many as 41 civilians are killed, but Mehsud survives.

Did the June 23 attack respect the principle of distinction? It would be illegal to target non-combatants. But who is a combatant in an internal war? Usually, combatants are defined as those who wear a uniform, bear arms openly, and operate under a responsible command. Yet this classic concept breaks down in internal wars. Non-state insurrectionists almost never meet the classic combatant criteria. Thus in an internal war, the operative criterion is combat membership: is there sufficient evidence that a target plays a combat role in a warring group? If so, he or she can be legally targeted.

Available evidence clearly implicated Mehsud—the leader of the Taliban in Pakistan—as a member of a warring party against the de facto government of Pakistan. It seems probable that the drone operators aimed at the crowd with the intention of killing Mehsud but without intending the deaths of civilians (though presumably thinking them highly probable). So long as the civilian deaths were not intended and reasonably could not be avoided, the principle of distinction was respected.

Did the June 23 attack follow the principle of necessity? Even if a state’s participation in an armed conflict is legal, it should not kill combatants if there is a reasonable possibility of incapacitating or arresting them before they resume combat activities. In the case of the June 23 killing, there was at the time no effective control over South Waziristan and thus no reasonable possibility of apprehension. Moreover, Mehsud had been hunted for years and could easily disappear into the populace, only to direct further strikes and militant operations. The June 23 strike would have been the last opportunity to stop him before he directed further strikes. On this evidence, the attack was in accord with the principle of necessity.

Perhaps most controversially, was the June 23 attack proportionate? Although civilians can never be the objects of a military strike, some civilian deaths are inevitable in the successful execution of military operations. Thus, the law of war requires the attacking party to balance its objective with other damage inflicted according to a rule of reason. This decision must be made in good faith on the information available before the attack.

The proportionality analysis is dauntingly complex. It governs reasonably understood goals and not unforeseen effects. It involves many possible factors and unlike, apparently incomparable values such as the number of anticipated casualties, the target’s short- and long-term military importance, his location and surrounding, whether the means to destroy him exceed those necessary, and the feasibility of alternatives.

Since evidence is sparse, it is difficult to convincingly condemn the June 23 attack as disproportionate. Did the drone pilots know that the rest of the crowd was mostly militants? If the pilots thought them civilians, was the value of the target from a long-term perspective worth a high loss of civilian life? A good case can be made that it would have been proportionate to accept many dozens of civilian deaths to incapacitate or kill a leader like Mehsud.

On proportionality, reasonable minds will disagree. Some will recoil at accepting such a high loss of civilian life. But in the final analysis the June 23 attack and those like it should be judged legally proportionate. Not everything permissible will be pleasant, and some unpleasant things will be necessary. For a just warrior to prevail over his enemy, he will have to make difficult decisions that hinge not only on principle but also on prudence and the facts at hand.