As a result of the unprecedented 41 drone strikes into Pakistan authorized by the Obama administration, aimed at Taliban and al Qaeda networks based there, about a half-dozen leaders of militant organizations have been killed--including two heads of Uzbek terrorist groups allied with al Qaeda, and Baitullah Mehsud, the leader of the Pakistani Taliban--in addition to hundreds of lower-level militants and civilians, according to our analysis.[1]

The number of civilian deaths caused by the drones is an important issue because in the charged political atmosphere of today's Pakistan, where anti-Americanism is rampant, the drone program is a particular cause of anger among those who see it as an infringement on Pakistan's sovereignty. A Gallup poll in August found that only 9 percent of Pakistanis favored the strikes, while two-thirds opposed them.

An important factor in the controversy over the drones is the widespread perception that they kill large numbers of Pakistani civilians. Some commentators have asserted that the overwhelming majority of casualties are civilians. Amir Mir, a leading Pakistani journalist, wrote in The News in April that since January 2006, American drone attacks had killed "687 innocent Pakistani civilians." A month later, a similar claim was made in the New York Times by counterinsurgency experts David Kilcullen and Andrew Exum, who wrote that drone strikes had "killed some 700 civilians. This is 50 civilians for every militant killed, a hit rate of 2 percent." In other words, in their analysis, 98 percent of those killed in drone attacks were civilians. Kilcullen and Exum advocated a moratorium on the strikes because of the "public outrage" they arouse.

A very different picture was presented earlier this month by the Long War Journal, an American blog that closely tracks terrorist groups, in particular al Qaeda and the Taliban. Bill Roggio, the editor of Long War Journal, concluded that according to his close analysis of the drone strikes, only 10 percent of those killed were civilians.

Our analysis suggests quite different conclusions than those of either Kilcullen and Exum or the Long War Journal.

But first, a word on our methodology. Our analysis of the drone campaign is based only on accounts from reliable media organizations with substantial reporting capabilities in Pakistan. We restricted our analysis to reports in the New York Times, Washington Post, and Wall Street Journal, accounts by major news services and networks--the Associated Press, Reuters, Agence France-Presse, CNN, and the BBC--and reports in the leading English-language newspapers in Pakistan--The Daily Times, Dawn, and The News--as well as those from Geo TV, the largest independent Pakistani television network. (Links to all those individual reports can be found in Appendix 1 of this paper.)

The news organizations we relied upon collectively for our data cover the drone strikes as accurately and aggressively as possible. And though we don't pretend that our study is accurate down to the last civilian death in every drone strike, we posit that our research has generated some quite reliable data on the number of militant leaders killed, a fairly good estimate of the number of lower-level militants killed, and a reliable sense of the real civilian death rate.

Since 2006, our analysis indicates, 82 U.S. drone attacks in Pakistan have killed between 750 and 1,000 people. Among them were about 20 leaders of al Qaeda, the Taliban, and allied groups, all of whom have been killed since January 2008. (A list of their names, as well as links to stories about the drone strikes that targeted them, can be found in Appendix 1.)

It is not possible to differentiate precisely between militant and civilian casualties because the militants live among the population and don't wear uniforms, and because the militants have the incentive to claim that all the casualties were civilians, while government sources tend to claim the opposite. However, of those killed in drone attacks from 2006 through mid-October 2009, between 500 and 700 were described in reliable press reports as militants, or some 66 to 68 percent.

Based on our count of the estimated number of militants killed, the real total of civilian deaths since 2006 appears to be in the range of 250 to 320, or between 31 and 33 percent.

That finding tracks with polling by the Aryana Institute for Regional Research and Advocacy, a think tank that works in the Pakistani tribal region along the Afghan border where the drone attacks have consistently taken place. It found that more than half the people surveyed in the winter of 2008 in this region, which is known as the Federally Administered Tribal Areas, said the drone strikes were accurate and were damaging the militant organizations based there.

Under President Obama, the strikes have taken out at most a half-dozen militant leaders while also killing as many as 530 others. Of those, around 250 to 400 are reported to have been lower-level militants, about three quarters, and about a quarter appear to have been civilians. The strikes appear to have killed a slightly lower percentage of civilians in the past nine months than during the earlier years of the American drone campaign in Pakistan.

Obama, far from curtailing the drone program he inherited from President George W. Bush, has instead dramatically increased the number of U.S. Predator and Reaper drone strikes. There have been 43 strikes in Pakistan this year (two while Bush was still in office), compared to 34 in all of 2008. None of the strikes under either Bush or Obama has targeted Osama bin Laden, who seems to have vanished like a wraith. U.S. intelligence officials say they have not had a solid lead on the whereabouts of the al Qaeda leader since the battle of Tora Bora in eastern Afghanistan in December 2001.

The leader of the Pakistani Taliban, Baitullah Mehsud, the mastermind of Benazir Bhutto's December 2007 assassination and many of the suicide bombings in Afghanistan, was a frequent target of the drone attacks. Under Obama, according to our analysis, 15 drone strikes specifically targeted Mehsud.

But he still didn't see it coming. On August 5, Mehsud, a diabetic former gym instructor, was receiving a leg massage on the roof of a house in South Waziristan when a drone slammed into his hideout, killing one of his wives, her father, and the terrorist chief himself.

The Pakistani press was jubilant. "Good Riddance, Killer Baitullah" was the lead headline in the quality Dawn newspaper. Much of the previous coverage in Pakistan of U.S. drone strikes in the tribal region had ranged from critical to downright hostile. But in the case of Mehsud, U.S. strategic interests and Pakistani interests were closely aligned because the Pakistani Taliban's victims included not only Bhutto, the country's most popular politician, but also hundreds of Pakistani policemen, soldiers, and civilians.

*

By July 2008, Bush administration officials had tired of Pakistan's unwillingness or inability to capture or kill the ever-expanding number of militants in the Federally Administered Tribal Areas (FATA). So they decided to ramp up the CIA's drone program targeting al Qaeda and Taliban leaders in the tribal regions.

What had particularly alarmed Bush administration officials over the previous three years was the mounting evidence that al Qaeda and affiliated groups were using the FATA to train Westerners for attacks on American and European targets. For instance, the masterminds of the July 7, 2005, transit system attacks in London, which killed 52 people, had trained in the tribal regions. So too had two Germans and a Turk who were planning to bomb the U.S. Air Force base in Ramstein, Germany, in 2007. And, during this period, both Osama bin Laden and his top deputy, Ayman al-Zawahiri, who are generally presumed to be living in or around the FATA, continued to release a stream of audio- and videotapes demonstrating that al Qaeda's leadership was very much intact.

Over the summer of 2008, the Bush administration authorized a sharp increase in the number of drone attacks. In the first six months of the year there were six such strikes, while in the second half there were 28.

According to Pakistani and U.S. officials and media accounts, drone strikes in 2008 killed about a dozen senior or mid-level leaders of al Qaeda or the Taliban. They included Abu Laith al-Libi, who orchestrated a 2007 suicide attack at Bagram Air Base in Afghanistan during a visit by Vice President Dick Cheney. Al Qaeda's resident authority on weapons of mass destruction, Abu Khabab al-Masri, was killed by a strike in South Waziristan in late July, and Abu Haris, al Qaeda's chief in Pakistan, was felled in September in North Waziristan.

The drone campaign certainly has hurt al Qaeda's leadership, which increasingly has had to worry about self-preservation rather than planning attacks or training recruits. One measure of the pain is the number of audio- and videotapes that the terrorist group has released through its propaganda arm, As Sahab ("the clouds" in Arabic). Al Qaeda takes its propaganda operations seriously and in 2007 As Sahab had a banner year, releasing almost 100 tapes. But the number dropped by half in 2008, indicating that the group's leaders were more concerned with survival than public relations. This year, however, As Sahab has already released 60 communiqués, according to IntelCenter, a government contractor that tracks jihadist propaganda, indicating that al Qaeda's propaganda machine has recovered.

As the drone attacks have put al Qaeda and allied groups under increased pressure, law enforcement authorities have uncovered few serious plots against U.S. or European targets that are traceable back to militants who had trained in Pakistan's tribal regions since the summer of 2008, when the drone program was first expanded.

A major exception is Najibullah Zazi, a onetime coffee cart operator on Wall Street who later worked as a shuttle van driver at the Denver airport. Zazi, an Afghan immigrant, was allegedly planning to launch this fall what could have been the deadliest terrorist attack in the United States since 9/11. He had traveled in late August 2008 to Pakistan, where by his own admission he was trained in explosives by al Qaeda members in a tribal area along the Afghan border. The FBI found pages of handwritten notes on Zazi's laptop computer about the manufacture and initiation of explosives and the components of various detonators and fusing systems--technical know-how he had picked up at one of al Qaeda's training facilities between late summer 2008 and January 2009, a period of intensified drone attacks.

*

Although the drone campaign has certainly killed terrorist leaders, it also presents a number of tactical and strategic problems that must be part of the debate about its efficacy.

The first is that the strikes, which almost inevitably kill civilians, may be on shaky legal ground. Columbia Law School professor Matthew Waxman points out that this is a tricky judgment call: "The principle of proportionality says that a military target may not be attacked if doing so is likely to cause incidental civilian casualties or damage that would be excessive in relation to the expected military advantage of the attack.... But there is no consensus on how to calculate these values (how do you compare the value of civilian lives versus the value of disrupting high-level terrorist operational planning?) Nor is there consensus on what imbalance is ‘excessive.' It's very hard to draw definitive conclusions because it requires assessments about such things as the expected military gain from neutralizing the target, the likely civilian harm, and the availability of alternative means of attacking that could save innocent lives."

Second, conscious that the drone attacks are generally unpopular with the Pakistani public, militants have used them as an excuse to strike government targets in the country's Punjabi heartland. The Taliban's March attack on a Lahore police academy, in which 18 people were killed, was "in retaliation for the continued drone strikes by the U.S. in collaboration with Pakistan on our people," Baitullah Mehsud said at the time. Several weeks later, Hakimullah Mehsud, Baitullah Mehsud's deputy and now reportedly his successor, told reporters, "We will continue to launch suicide attacks until U.S. drone attacks are stopped."

Third, the strikes no longer have the element of surprise. It is highly unlikely that the drone program will be expanded from FATA into other, non-tribal regions of Pakistan because of intense Pakistani opposition to such a move. Understanding that fact, some militants have undoubtedly moved out of FATA and into safer parts of Pakistan, potentially further destabilizing the fragile Pakistani state.

Fourth, although drone attacks often kill low-level militants, they also destroy the computers, cell phones, documents, and "pocket litter" that can provide myriad other leads to investigators. Dan Byman of Georgetown University has correctly observed that drone strikes are a "poor second to arrests [because] dead men tell no tales."

Fifth, the drone program is a tactic, not a strategy. Bruce Hoffman, a Georgetown University professor widely regarded as the dean of terrorism studies, says, "We are deluding ourselves if we think in and of itself the drone program is going to be the answer." He points out that the 2006 U.S. airstrike that killed the leader of al Qaeda in Iraq, Abu Musab al-Zarqawi, did not exactly shut down the organization. Following Zarqawi's death, violence in Iraq accelerated.

Sixth, while there islittle doubt that the strikes have disrupted al Qaeda's operations, the larger question is to what extent they may have increased the appeal of militant groups and undermined the Pakistani state. This is ultimately a lot more worrisome than anything that could happen in Afghanistan, given that Pakistan has dozens of nuclear weapons and is one of the world's most populous countries.

Drone strikes will remain an important tool to disrupt al Qaeda and Taliban operations and to kill the leaders of these organizations, but they also consistently kill Pakistani civilians, angering the population and prompting violent acts of revenge from the Pakistani Taliban.

For the time being, however, they appear to be the least bad option the United States has for reducing the threat from Pakistan's militants, given that an American ground assault into Pakistan's tribal regions is out of the question, and that U.S. and Pakistani strategic interests are more closely aligned today than they have been in years because of the two countries' shared interest in attacking the Pakistani Taliban and their al Qaeda allies.

[1] This information is accurate as of October 19, 2009.