Drone Strikes Linked to More Civilian Deaths

Despite claims that drone strikes in Pakistan have been effective and efficient, new reports are set to come out later this week that link the drone campaign with high civilian casualty rates, raising questions regarding the United State’s transparency in the ongoing drone war.   Continue reading

Signature Strikes are as Old as War

Arianna Huffington recently wrote on the Huffington Post about the Obama Administration’s use of so-called “signature strikes” by drones in Pakistan and elsewhere. Underlying much of her criticism is a basic assumption that signature strikes are a new form of targeting that is more pernicious than the more targeted strikes that we are used to. This assumption is perfectly true, if your sense of history goes back no more than about decade.

Signature Strikes are the norm, not the exception

The truth is that what we now call “signature strikes” used to just be called “targeting the enemy.” The practice is in fact so old that it is impossible to trace its origins. In fact, some of the earliest accounts of armed conflict speak about the use of “signatures” or qualities that allow a soldier or government agent to identify an unknown person as a probable member of an enemy group. The Old Testament is replete with examples, including David’s use of foreskins to identify Philistines, and the Gileadites’ use of the word “shibboleth” to identify and kill the Ephraimites.

In fact, knowing the identity of your enemy with any precision was extraordinarily rare until the last ten years. Granted, targeted killing of one sort or another (even leaving aside clear cases of political assassination) has existed for time immemorial as well, but was definitely not the norm. It was instead an unusual undertaking, and attempts rarely led to success. For eons, the standard approach to warfare has been to assemble together a large group of men, few of whom are expected to know any of the enemy by name (with the exception of their king or president), and effectively set them loose on a collection of the enemy population. In later more “civilized” times, these armed groups would mostly only attack one another, although this was never a perfectly uniform practice. These soldiers, whether professionals or conscripts, would not identify an enemy by name, position, or other individualized characteristic. It was always enough that they “look like the enemy.”

This is what has always been expected of soldiers, and to a great extent, still is. We now have more rigorous standards of conduct, laws to mitigate civilian damage done during an attack on enemy forces, and clearly spelled out rules of engagement. Even still, in my four years in US Army’s 82nd Airborne Division, including deployments to Iraq and Afghanistan, I was shown a picture of an individual enemy that we planned to kill or capture on a handful of occasions. The rest of the time we were just expected to use the same standard as is used for pornography: you know it when you see it. In other words, we were expected to use the ground combat equivalent of signature strikes. Warfare is just not that individualistic, even today.

Signature strikes may not be new, but our new techniques are helping reduce civilian harm, despite the “evidence”

Still, the US military and intelligence community is becoming impressively proficient at identifying and locating individual members of the enemy. Despite the much publicized (but mostly classified) report recently published by the Center for Naval Analysis’ Larry Lewis, most reliable evidence says that drones, even when used for signature strikes, are much more discriminant and precise, killing or injuring fewer civilians per strike, and far fewer per militant killed. Much of the evidence to the contrary is in fact gained from anonymous sources linked to the Pakistani military or to the militants themselves. Given the sourcing–the enemy and their patrons–it is a little surprising that the claimed number of civilian deaths is in fact so low, even including signature strikes. Maybe these sources are willing to double the number of civilian casualties, but tripling it seems a bridge too far (except for a stalwart few who transparently either fudge their numbers or need to retake middle school pre-algebra).

As a case in point, even the Bureau of Investigative Journalism, who appears at first to be an honest broker of information on drone strikes, upon deeper investigation, are highly biased. They reported yesterday that the number of civilian deaths due to drone strikes in Pakistan has been underreported by 81. They describe this as a “high civilian death toll,” but for a campaign waged over the course of nine years, this is actually an extremely low death toll. I challenge anyone to find another military campaign that lasted nine years yet killed fewer than 45 civilians per year (according to their numbers). However, even these numbers are suspect, as can be seen from their source document, a leaked Pakistani summary of drone strike casualties. Setting aside the arithmetic problems (4+5=8? Apparently, in Pakistan. See line 78), several problems indicate unreliability.

First, the number of civilians killed is not always as clear as the Bureau would have it. In the vast majority of “civilian” casualty cases, the number of civilians are not specified, and is indictated only by inclusion of the word “civilian” in the remarks column. It is not clear whether this means all dead and wounded were civilians or whether one or some of them were. However, in other cases, the number is given, or the remarks indicate that they were “all civilians.”

Second, and more damning, only two references were made to militants out of the nearly 750 dead in 84 attacks. However, one entry refers to “miscreants,” and several others list foreigners as among the dead. This raises the suspicion that the number of militants killed is actually under-reported, and thus the possibility that some listed as civilians are in fact militants.

Third, and worst of all for the reliability of this information, it is apparently obtained secondhand and from afar. Several of the entries indicate that the information is “reported” or “yet to be received.” Local elders and even local political administrations are often pro-Taliban or otherwise compromised (e.g. by bribery which constitutes the majority of their income).

The hard truth is that coming to even a good, ball-parked number of civilian dead is extremely difficult. But what is clear is that the US drone campaign, even when using signature strikes, has a remarkably low rate of civilian casualties when compared to other available options, given their advanced optics, long loiter times, and precision munitions. Those who advocate against drones tend to focus on the technology, when what they are actually opposed to is the use of force in general. This is a laudable sentiment, but these same folks never seem to be willing to offer workable alternatives.

The truth is that drones are here to stay, because they are an extremely useful, discriminating weapon system. Given that we are involved in a protracted conflict with an enemy that hides among the civilian population, our drones and drone pilots are doing a remarkable job. Of course, as I will describe in a later post, relying solely on drones to win this conflict for us is in the end counterproductive. More on that soon.

 Paul W. Taylor, Senior Fellow
Center for Policy and Research