It is scarcely a secret that the United States is fighting an air war in Pakistan on at least four different levels. It is using unmanned combat aerial vehicles (UCAVs) to support U.S. forces in “hot pursuit” in the border area. It is using them to attack Taliban and other insurgent forces near the border to limit their capability to operate in Afghanistan. It is striking at insurgent and terrorist leaders and training camps inside the tribal areas in Waziristan, and it sometimes supports Pakistani forces in strikes against the Pakistani Taliban. The New York Times , Washington Post , Wired and the Long War Journal have all published articles on the details of these supposedly secret operations.
What has been far less clear, however, is the context. Some reporting makes this look like a massive bombing campaign, and one producing large numbers of unnecessary civilian casualties. Other reporting somehow makes it seem illegitimate or talks about a Pashtun honor code as if U.S. forces can only fight insurgents face-to-face with their weapons on their terms.
One has to be very careful about unclassified statistics, but the Long War Journal reports that the number of strikes against cadres in Pakistan is very limited. It reports only one strike a year in 2004 and 2005, three in 2006, five in 2007, thirty-five in 2008, fifty-five in 2009 and seventy-seven in the first nine months of 2010. This rise in strike numbers is a kind “surge,” but it adds up to all of 175 strikes over the entire war, and these strikes (65 percent) have been concentrated in North Waziristan where the Pakistani army has been unwilling or unable to act, and almost all of the other 35 percent have been in areas in South Waziristan where the Pakistani Army and Air Force cannot bring anything like the same intelligence, targeting and precision-strike assets to bear.
There is certainly a steady rise in strikes, but talking about it as “intense combat” is absurd. Wired says the U.S. Air Force (USAF) reported it flew a peak of 19,500 close-air-support sorties in the Iraq War in 2007, and has flown 4,620 so far in 2010. Wired reports that the USAF has said it flew an average of over two thousand a month in Afghanistan in 2009, and over two thousand five hundred a month in 2010. The total number of UCAV strikes in Pakistan over the entire war is a fraction of the air strikes per month in Afghanistan and Iraq, and a tiny number by the standards of any previous air war.
It is also important to stress that UCAVs are simply the tip of the spear. The UCAV strikes against the insurgent and terrorist networks are the result of one of the most massive and sophisticated targeting efforts in history. They are targeted as a result of the use of virtually every intelligence asset America has from satellites to manned aircraft to human intelligence, and the use of unarmed unmanned combat vehicles. They are subject to careful review to minimize civilian losses, and they still manage to be extremely effective. If one looks at the estimates in the Long War Journal, the seventy-seven UCAV sorties flown through September 2010 killed eighteen senior insurgent leaders, including nine with at least some links to Al-Qaeda.
As for casualties that are inflicted, one only has media reports to draw upon, but there are several things to consider. We have no alternative way to fight and all of the other options would be far worse even if they were available. The United States can sometimes send in small Special Forces elements and specially trained local fighters, but only in very small operations near the border. Moreover, Special Forces are far safer—and inflict far less civilian casualties—when they can use UCAV sorties than when they are in direct combat.
Moreover, any land operation that crosses the border and becomes public, and even the most limited helicopter attacks, become a political crisis. Flying manned U.S. fighter aircraft into Pakistan could push Pakistan into shutting down all of its cooperation, and would inevitably inflict much higher casualties. High-speed jet fighters can’t linger over a target for hours to verify a target in order to do as much as possible to strike at a time that ensures civilian casualties are kept to a minimum. Even if Pakistani land forces did take over the job, we have already seen in Swat and South Waziristan that they would have to fight their way in and the end result would be far more Pakistani casualties—and at least ten times more civilians killed and thousands or ten of thousands displaced.
In contrast, reporting in Wired indicates that all of the UCAV strikes made between 2006 and the present have killed a total of 1,490 insurgents and 104 civilians. Improvements in the rules of engagement have actually cut civilian casualties: The fifty-three strikes in 2009 killed 463 insurgents and forty-three civilians. The seventy-seven strikes in 2010 killed an estimated 546 insurgents and ten civilians, which is 0.12 civilians per sortie versus 0.8 civilians per sortie in 2009. If these numbers are even roughly accurate, no other form of modern war has come close to being this lethal against the enemy and this humane in terms of civilian casualties.
War remains horrible and still kills the innocent as well as the enemy. But, we need to be realistic. Pakistan is at best a tenuous and divided ally. Islamabad was unwilling to attack Afghan Taliban targets and conduct a major campaign against al-Qaeda before the flood; elements of the ISI remain tied to the Taliban and al-Qaeda; and its civilian government has far too many elements that are corrupt, incompetent and unwilling to act. Fighting a war in Afghanistan that has given the enemy a sanctuary in Pakistan, and al-Qaeda immunity in Pakistan, has little point. More bluntly, if Pakistan cannot provide at least enough cooperation to passively allow such strikes, it is not an ally, it is a major strategic liability.