The Drone Papers: Intercepting the Nonsense
Editor’s Note: The drone program remains controversial, and critics regularly blast it for creating more terrorists than it kills. Prominent among these critics is Glenn Greenwald, a founding editor of The Intercept, which published an anti-drone report based on leaked documents. C. Christine Fair, a regular Foreign Policy Essay contributor, took on Greenwald on Al Jazeera. In this essay, she goes into detail about some of the myths and misunderstandings about the drone program that came up on the program and argues that many of the criticisms are flawed or biased.
Last week, I had the ignominious distinction of engaging in a discussion with Glenn Greenwald on Al Jazeera’s weekly program, “The Arena.” This particular episode was motivated by a recent report on drones by Greenwald's magazine, The Intercept. When Al Jazeera asked me to participate in a discussion of the drone program in Pakistan, I was suspicious for several reasons. My cumulative experiences with Al Jazeera with respect to its coverage of the various U.S. drone programs have been disappointing, because the network takes an editorial position against “the drone program,” while purporting to report on them fairly.
The host of the show, Mehdi Hasan, also did not inspire confidence in the prospects for a legitimate debate. Mr. Hasan has publicly reviled the drone program and, more worrisome yet, he has charmingly cited the Quran to compare kuffar (infidels like me) to "cattle" who are "of no intelligence." In explaining the innate moral superiority of Muslims to these ruminants, he elaborated that Muslims like himself “know that keeping the moral high-ground is key. Once we lose the moral high-ground we are no different from the rest, of the non-Muslims; from the rest of those human beings who live their lives as animals, bending any rule to fulfil any desire.”
Leaving aside the dubious journalism ethics of the network, whether in either English or Arabic, and the sketchy bona fides of the host, there was my fellow guest: Glenn Greenwald has tirelessly flogged the use of armed drones with Crusader-like conviction. I was equally confused as to why the show was going to focus upon Pakistan, when the latest tranche of pilfered documents released by The Intercept promises to detail “the inner workings of the U.S. military’s assassination program in Afghanistan, Yemen, and Somalia.” However, the producer, Ryan Kohls, assured me that they were, in fact, interested in pushing back on Greenwald’s “black and white” account of the program and bringing nuance to the debate. Also counter to my instincts, the show was taped, rather than live. This allowed the network to selectively edit the conversation, as well as visuals.
In this post, I address many–but by no means all–of the outright falsehoods and empirical obfuscation that Mr. Greenwald brought to the show, including some that were deleted from the show as aired.
Selection Bias Happens
First, a brief word about the nature of Mr. Greenwald’s enterprise, The Intercept. The Intercept has set up a secure drop box to facilitate government employees’ illegally providing classified information to the organization. Mr. Greenwald and his associates refer to these persons as “whistleblowers.” And let's leave aside for a moment whether that's a fair description. Empirically, the documents that have been leaked are riven with selection bias; leakers, driven by whatever personal motives, often selectively leak specific documents. Consumers of these documents have no idea of how representative this sample is of all information that exists about the drone program. There are most certainly classified reports that positively describe the program’s utilization in specific theatres, but these documents have not apparently been leaked. In fact, given Mr. Greenwald’s evangelical zeal against the various uses of drones, I doubt that if the organization received such reports, it would publish them. And if it did publish such exculpatory documents, would The Intercept’s writers simply dismiss them as “rank propaganda”?
Second, these leakers are always anonymous, for obvious reasons: they do not want to be prosecuted for breaking the law. However, anonymous sources cannot be vetted for the sagacity of their interpretation or for their motives. In its most recent product, "The Drone Papers," The Intercept tells readers that the leaker was moved by his moral outrage. But is it not possible that less noble motivations compelled the leaker to provide these particular documents? If the motive was espionage or work-place dissatisfaction, does that change our perception of the documents and the leaker’s explanation of them?
A third problem is the presumption that classification confers some standard of quality. As anyone who has used classified documents can attest, this is not always, or even often, the case. After all, a classified analysis is only as good as the data that undergird it and the skills of the analyst. In my experience, classified products are rarely worth the effort to obtain and read.
Making a Bad Sausage out of Bad Data
In addition to the classified materials The Intercept has garnered, let’s turn to the ways in which Mr. Greenwald utilized these documents in the Al Jazeera debate. The discussion began with Mr. Greenwald making the sweeping assertion that 90 percent of the drone victims are “innocent.” Mr. Greenwald buttressed this claim by citing documents provided by the “whistleblower,” whose leaked materials formed the basis of "The Drone Papers." Obviously, someone who is breaking the law to provide these documents must be presumed to be completely honest and factually correct in his assessments.
There are several problems with this assertion. First, The Intercept claims that the documents it received “show that during a five-month stretch of the campaign [Operation Haymaker], nearly nine out of 10 people who died in airstrikes were not the Americans’ direct targets.” This is not the same as being “innocent.” Moreover, the documents shown in the report do not demonstrate this claim. In fact, one slide detailing the mission statistics for Task Force 3-10 from September 2011 through September 2012–a longer period of time than the five-month period cited above–indicates that only 14 civilians were killed. However, the leaker asserts that “the 14 civilian casualties is highly suspect … I know the actual number is much higher.” Why should anyone believe this claim?
More to the point, how can Mr. Greenwald employ this suspicious assessment by a leaker about a specific operation to make a generalized claim about all drone operations everywhere, whether operated by the CIA or the Department of Defense? Equally problematic, neither the leaked documents nor The Intercept’s analysis of them makes any distinction between the various kinds of aircraft conducting the strikes in Operation Haymaker. The Intercept’s apparent, but unwarranted, presumption is that they are all or mostly drones.
Apart from the problems with the document, the leaker and The Intercept’s interpretation of both, there is another reason to doubt Mr. Greenwald’s claim: the data collected by the New America Foundation and the Bureau of Investigative Journalism–hardly mouthpieces of the Obama administration–on drone strikes in general do not support it by any measure. As reported by CNN:
Over the life of the drone program in Pakistan, which began with a relatively small number of strikes between 2004 and 2007, the estimated civilian death rate is 16%. And in the Obama administration, between 1,507 and 2,438 people have been killed in drone strikes. Of those, 148 to 309, or between 10% and 12%, were civilians, according to the New America Foundation data.
With respect to Pakistan, there is one study that actually comes to the exact opposite conclusion as the one put forward by Mr. Greenwald. A few years ago, a Pakistan-based journalist sent a Waziri stringer into Waziristan to interview locals about who died in the drone strikes in their villages. According to the report from that six-month-long study, villagers claimed that “at least 194 people killed in the attacks, about 70 percent–at least 138–were militants. The remaining 56 were either civilians or tribal police, and 38 of them were killed in a single attack on March 17, 2011. Excluding one catastrophically disastrous strike which inflicted one of the worst civilian death tolls since the drone program started in Pakistan, nearly 90 percent of the people killed were militants.
Mr. Greenwald and Mr. Hasan also asserted vigorously that drones make more terrorists than they kill. To support this claim, Mr. Greenwald cited the opinions of several well-known persons. First, he cited Malala Yousafzai, the teenage girl whom the Pakistani Taliban shot for her promotion of female education. He noted emotively that Ms. Yousafzai explained to President Obama that drones make more terrorists and asked him to stop using drones. Mr. Greenwald defended his reference to her, arguing that she was the spokesperson for Pakistan because she had won the Nobel Peace Prize.
Ms. Yousafzai is a courageous young woman. She is not, however, representative of Pakistani thought. Indeed, Mr. Greenwald was apparently not aware that Ms. Yousafzai has been widely reviled in Pakistan, where detractors accuse her of being a tool in a foreign plot to defame Pakistan. She's also no expert on the matter at hand, her good intentions notwithstanding. Leaving aside her personal tragedy and perseverance, Ms. Yousafzai lived in Malakand, not in the tribal areas where drones exclusively operate. And she did not become an international celebrity for her knowledge of drones or militant recruitment.
Her Nobel Prize may be well deserved, but it is beside the point. Indeed, if views of Nobel Prize winners are a metric of credibility, President Obama also received that distinction. And though he, like Ms. Yousafzai, is widely reviled in Pakistan, his views of drones differ from that of Ms. Yousafzai or Mr. Greenwald.
Mr. Greenwald also cited the opinions of well-regarded generals such as General Stanley McChrystal, the former commander of NATO troops in Afghanistan. Mr. Hasan also cited the views of retired U.S. Lt. General Michael Flynn, former head of the Defense Intelligence Agency (DIA), who said in an interview with The Intercept that launching a drone strike “makes us all feel good for 24 hours” but ultimately “doesn’t matter.” Clearly, there are others who disagree with both of these generals; otherwise the program would be dead, rather than thriving. However, it should be emphasized that the opinions offered by any generals, unless supported by data and rigorous analysis, are simply opinions.
In contrast to the curated quotes of prominent personalities which Mr. Greenwald gathers and describes as a “mountain of evidence” about the dangers of drones, there is actually a robust body of scholarly work that addresses the effects of leadership decapitation on a wide array of militant groups operating in diverse countries and their ability to produce violence. In general, the scholarship produces mixed results, with some work showing the efficacy of leadership decapitation (e.g. Johnston 2012; Price 2012) while other studies find that it is sometimes effective (e.g., Jordan 2009) or even counterproductive (Hafez and Hatfield 2005).
With respect to Pakistan, in particular, Johnston and Sarbhai (2015) found that “drone strikes are associated with decreases in the incidence and lethality of terrorist attacks.” The quantitative work of James Igoe Walsh (a professor of political science at the University of North Carolina whose quantitative work is funded by the U.S. government and was published by the U.S. Army War College in 2013) found that that drone strikes that do result in civilian deaths have little relationship to subsequent militant violence. In contrast, Jaeger and Siddique (2011) find varying effects of drone strikes upon insurgent activity depending upon the time frame examined. (Such nuances will no doubt vex Mr. Greenwald.)
While the scholarly literature arrives at differing opinions about the efficacy of leadership decapitation generally and drone strikes in particular, the militant leaders in Pakistan, including Osama bin Laden and Pakistani Taliban commanders, themselves have discussed drones mostly in terms of fear, rather than as recruitment opportunities.
Using and Abusing Data
During the discussion, Mr. Greenwald and Mr. Hasan made ample reference to a particular advocacy-driven report on drones in Pakistan authored by persons associated with the law school clinics of NYU and Stanford University Law Schools. That report is deeply flawed in every manner of execution. Let's start with the fact, acknowledged by the authors, that the report was commissioned and facilitated by Reprieve. The authors concede that in “December 2011, Reprieve, a charity based in Britain, contacted the Stanford Clinic to ask whether it would be interested in conducting independent investigations into whether, and to what extent drone strikes in Pakistan confirmed to international law and caused harm and/or injury to civilians” (p. i.).
It is important to note that Reprieve, and its Pakistani partner organization The Foundation for Fundamental Rights (FFR), are not merely “charities;” rather, both have vigorously argued for the termination of the drone program. Reprieve’s website explains it position on drones:
The CIA killer drones programme is the death penalty without trial, and the new face of US counter-terrorism policy. Reprieve is assisting survivors and victims’ families in their fight for legal accountability, transparency and justice.
This is a bit like the National Rifle Association (NRA) sponsoring a study on firearms and crime.
It is possible that the report could have been executed rigorously nonetheless, but as I have detailed elsewhere, this was not the case. That report, like virtually every other anti-drone report, assiduously silences the voices of those who are pro-drone because they distort the narrative. In fact, one should be wary of any report that discusses drones without any reference to their domestic proponents.
Speaking of the Pakistan case, which I know well, those in the Federally Administered Tribal Areas (FATA) in closest proximity to the various terrorists view the drones either as the least of all evils or even as their best source of protection. One Pakistani editorial from 2012 summed up Pakistan’s dilemma:
[T]he real threat to our nation comes from the heavily armed outfits marching across our northern areas, rather than the strikes made by unmanned planes. It is true that such strikes ignite a degree of anger and thus spur on militancy but this is a relatively minor matter, blown out of proportion compared to the threat of militancy from within.
Moreover, many residents in FATA vigorously support the U.S. armed drone program and even compare them to ababil, the holy swallows mentioned in the Quran (Surat al-Fil, “The Elephant”) who repelled an army of elephants that invaded Mecca by dropping black stones upon the invaders. When I raised this point on Al Jazeera, Mr. Greenwald dismissed it as “rank propaganda,” even though this view originated from residents of FATA in Pakistan itself. Whether Mr. Greenwald wants to admit it or not, Pakistanis in the tribal areas–who are most affected by the terrorists and drones–are not as opposed to them as he would like his flock to believe.
Another set of studies that Mr. Greenwald and Mr. Hasan distorted were the various surveys by Pew. Mr. Greenwald asserted that “the vast majority of Pakistanis oppose the drones,” according to Pew’s surveys of Pakistan. I attempted to explain the problems with this assertion, beginning with the simple fact that majorities or large minorities of Pew’s overwhelmingly urban sample had never even heard of the drone program. This question about familiarity with the program was then used as a gateway question to solicit respondents’ views of the program. (For some inexplicable reason, Pew stopped asking respondents if they had heard of the program after 2012 even though large minorities were still unaware of the program.)
It is true that majorities of those (large minorities or small majorities of respondents) who had heard of the program opposed it. In a recent peer-reviewed scholarly publication, my colleagues and I found that those who had heard of the drone program in Pakistan were overwhelmingly male, educated, and accessed their news from the Internet. Given the various selection effects inhering in who knew about the program from an overwhelmingly urban sample, one cannot infer from these results that the “majority of Pakistanis” oppose the program. Moreover, large minorities also believed that drone strikes are “necessary to defend Pakistan from extremist groups.” Thus Mr. Greenwald’s characterization of Pew data is inaccurate along virtually every dimension. It should be noted that as I was trying to explain these fraught issues of sample and social desirability bias, Mr. Greenwald complained that I had been talking too much and the host agreed.
Perhaps Mr. Greenwald’s most offensive deception was his claim that drones do not have pilots and that the operators of the drones only see “codes” on their small screens. Curiously, Al Jazeera deleted this instance of Mr. Greenwald’s misinformation. This may have been due to the fact that I tweeted about this upon leaving the studio, before being advised not to do so by the producer. Mr. Greenwald sought to imply the usual canard that drone warfare is remote-control killing where the “operator” is shielded from the consequences of her actions.
As Mr. Greenwald should know, not only do the drones have pilots, those suffer the same stresses that conventional pilots do, and possibly with greater intensity and higher frequency. The reason for this is that drone pilots frequently observe the targets for days to establish a pattern of life. Unlike a conventional pilot who does not see the outcomes of her action, drone pilots must watch it all. On occasion, after the munition has been released, the drone pilot sees a child enter the site and the pilot cannot recall the munition. Drone pilots see the death they cause up close and personal. This issue has been covered extensively by media outlets for which Mr. Greenwald worked (e.g., Salon and the Guardian).
In conclusion, I should have known better than to expect a fair airing of empirical issues inhering in the U.S. government’s use of armed drones in Pakistan or elsewhere given the host, the network, and my fellow guest. In fact, I did not bother watching the show until asked to write this post. I knew it had aired only when Mr. Greenwald’s legion of acolytes began trolling me on Twitter with the familiar litany of often-misogynist rants.