US officials rarely mention civilian casualties by US drone strikes. When they do, they generally offer extremely low estimates in the “single digits.” It is very difficult–given the opaqueness of the US government about its targeted killing program, and the obstacles currently faced by independent observers investigating on the ground–to determine precisely the total number of individuals killed, let alone the number of civilians who have been killed or injured in drone strikes in Pakistan. Yet the numbers of civilians killed are undoubtedly far higher than the few claimed by US officials.
At the same time, however, given the military effect of drone strikes themselves, as well as the political impact caused by reports of civilian deaths from drone strikes in Pakistan, the Taliban and other armed groups have an interest in exaggerating civilian casualty figures. Caution, therefore, must be exercised around all claims, and underlying sources must be scrutinized. It should also be noted that such concerns about both exaggeration and under-counting are not unique to the drone strike context, and are present in many conflict and government use of force contexts around the world.
This section aims to account for the contradictory claims made about drone casualties, and to explain the obstacles to certainty about who has been or is being killed by the US. First, we consider the concerning implications of reducing all casualties to an oversimplified civilian/“militant” binary, as most government and media sources do. We then examine the biases and demonstrated unreliability of government accounts of drone strikes, and explain the various factors that produce conflicting and often unreliable reporting by major media outlets. Lastly, we detail the methods and content of the three most well-known and widely cited strike data aggregators—The Long War Journal, New American Foundation, and The Bureau of Investigative Journalism (TBIJ)–and outline why TBIJ’s data currently constitute the most reliable available source.
Major media outlets in the US, Europe, and Pakistan that report on drone strikes tend to divide all those killed by drone strikes into just two categories: civilians or “militants.” This reflects and reinforces a widespread assumption and misunderstanding that all “militants” are legitimate targets for the use of lethal force, and that any strike against a “militant” is lawful. This binary distinction, in turn, feeds the political discourse around drone warfare, enabling commentators and analysts to make sweeping claims about the program’s efficacy and accuracy. The civilian/“militant” distinction is extremely problematic, however, from a legal perspective, and also because of the questionable reliability of the information on which “militant” determinations are based.
First, in most coverage of drone strike casualties, “militant” is never defined. The term’s use often implies to the reader that the killing of that person was lawful. The frequent use of the word “militant” to describe individuals killed by drones often obscures whether those killed are in fact lawful targets under the international legal regime governing the US operations in Pakistan. It is not necessarily the case that any person who might be described as a “militant” can be lawfully intentionally killed. As discussed in the Legal Analysis section, Chapter 4, in order for an intentional lethal targeting to be lawful, a fundamental set of legal tests must be satisfied. For example, depending on the applicable legal framework (but at the very minimum): the targeted individual must either be directly participating in hostilities with the US (international humanitarian law) or posing an imminent threat that only lethal force can prevent (international human rights law). Thus, for instance, members of militant groups with which the US is not in an armed conflict are not lawful targets, absent additional circumstances (such as evidence that lethal force against that person is proportionate and necessary). Further, simply being suspected of some connection to a “militant” organization—or, under the current administration’s apparent definition, simply being a male of military age in an area where “militant” organizations are believed to operate–is not alone sufficient to make someone a permissible target for killing. Failure by government and media sources to provide any additional details about most of those killed often makes it difficult to assess the legality of any particular attack.
Second, the label “militant” also fails to distinguish between so-called “high-value” targets with alleged leadership roles in Al Qaeda or anti-US Taliban factions, and low-level alleged insurgents with no apparent access or means of posing a serious or imminent threat to the US. National security analysts—and the White House itself—have found that the vast majority of those killed in drone strikes in Pakistan have been low-level alleged militants. Based on conversations with unnamed US officials, a Reuters journalist reported in 2010 that of the 500 “militants” the CIA believed it had killed since 2008, only 14 were “top-tier militant targets,” and 25 were “mid-to-high-level organizers” of Al Qaeda, the Taliban, or other hostile groups. His analysis found that “the C.I.A. [had] killed around 12 times more low-level fighters than mid-to-high-level” during that same period. More recently, Peter Bergen and Megan Braun of the New America Foundation reported that fewer than 13% of drone strikes carried out under Obama have killed a “militant leader.” Bergen and Braun also reported that since 2004, some 49 “militant leaders” have been killed in drone strikes, constituting “2% of all drone-related fatalities.”
Third, major media outlets, the main source for public information on drone strikes, typically cite to “anonymous officials” (generally from Pakistan) for the claim that a certain number of those killed were “militants.” Often, little to no information is presented to support the claim. And, it is entirely unclear what, if any, investigations are carried out by the Pakistani or US governments to determine who and how many people were killed. It is these media reports that are typically compiled by drone strike data aggregators and become the basis for statistical claims about the US drone program.
Underreporting of Civilian Casualties by US Government Sources
While western media outlets are generally quick to report official US accounts of drone strikes and their attendant casualties, those government sources have proved to be unreliable. Civilian death toll figures cited by the Obama administration during the last few years have been so low that even the most conservative nongovernmental civilian casualty estimates—including those released by think tanks such as the Foundation for Defense of Democracies and the Jamestown Foundation—contradict the administration’s claims. Most recently, officials in the Obama administration asserted that civilian casualties in Pakistan have been “exceedingly rare,” perhaps even in the “single digits” since Obama took office. These estimates are far lower than media reports, eyewitness accounts, and the US government’s own anonymous leaks suggest.
A recent exposé in the New York Times partially helped to explain the White House’s astonishingly low estimates by revealing that the Obama administration considers “all military-age males [killed] in a strike zone” to be “combatants . . . unless there is explicit intelligence posthumously proving them innocent.” How the US would go about gathering such posthumous evidence is unclear, in part because drone victims’ bodies are frequently dismembered, mutilated, and burned beyond recognition. And importantly, there is little evidence that US authorities have engaged in any effort to visit drone strike sites or to investigate the backgrounds of those killed. Indeed, there is little to suggest that the US regularly takes steps even to identify all of those killed or wounded.
Consistent with an apparent lack of diligence in discovering the identities of those killed, there is also evidence that the US has tried to undermine individuals and groups that are working to discover more about those killed. In August 2011, the New York Times first reported on efforts by Pakistani human rights lawyer Shahzad Akbar and by TBIJ, an independent non-profit news reporting agency based at City University in London, to document civilian drone casualties. The Times reported then that “anonymous US officials” accused Akbar of “working to discredit the drone program at the behest of . . . ISI, the Pakistani spy service.” The Times further reported that these officials argued that the Bureau’s data were “suspect” because of links to Akbar. TBIJ released a report a few months later on the US practice of targeting rescuers and funeral-goers. Another anonymous official dismissed the report’s findings with the statement, “[l]et’s be under no illusions—there are a number of elements who would like nothing more than to malign these efforts and help Al Qaeda succeed.” The US has never provided any evidence that might link Akbar to the ISI, or that might justify its allegation against TBIJ, relying instead on mainstream media sources to re-publish serious but anonymous accusations made by its own officials.
Even before the Obama administration’s novel definition of a “combatant” was revealed, a number of journalists who regularly cover drone strikes already recognized that the sweeping official claims of all-militant casualties were likely untrue. Nonetheless, most major Western and Pakistani news agencies still tend to rely on anonymous government sources and to report that strikes have killed “militants” or “suspected militants.” Some of the media agencies update their reports later to reflect contrary information if and when it emerges, but others, including major wire services, have at times let their initial reports stand even after credible accounts of civilian casualties have subsequently come to light.
Conflicting Media Reports
Media reports on drone strikes also often contradict one another on a range of strike details, including the nationalities of victims, the number of persons killed, and the types of structures targeted. For example, a May 24, 2012 strike in Khassokhel, Mir Ali was reported by the Associated Press as a strike on a “militant hideout” that killed “10 alleged militants,” most of whom were “Uzbek insurgents.” A Reuters wire released at around the same time reported that the strike was on “suspected Islamist militants” and killed ten people, while the Agence France-Presse reported that there were five “insurgents.” Neither Reuters nor AFP made any mention of the victims’ nationality. The BBC, for its part, reported that the strike was on a “house,” and that it had killed “at least eight people” of “Turkmen origin.” Within twenty-four hours, a number of other reputable sources, both western and Pakistani, reported that the strike had actually hit a mosque during morning prayers, and that some sources, at least, contended that the dead included local Waziri villagers. Some western media outlets updated their reports to reflect these new allegations, while others ignored the new information. The Associated Press referenced the May 24 strike in a separate article four days later, but failed to mention the possibility that a mosque had been struck. Instead, AP wrote that “[t]he attack took place in a militant hideout” and that “[m]ost of those killed were Uzbek insurgents,” citing a Pakistani intelligence source.
The discrepancies in these reports are the result of numerous factors–primarily the US government’s opaqueness, compounded by the investigation obstacles faced by independent actors. As described in Chapter 1 (Background and Context), Federally Administered Tribal Areas (FATA) is closed to all outsiders, including Pakistani citizens from outside the agencies. This means that few researchers or non-local journalists can actually visit North Waziristan to investigate drone strike casualties independently. When they do, they are often accompanied by Pakistani military forces who have an interest in controlling their access to information and influencing their reporting.
Most journalists writing on drone strikes thus rely instead on a combination of intelligence and military leaks, government sources who refuse to go on the record by name, and, sometimes, local Waziri correspondents, or “stringers.” All of these sources have the potential to be unreliable. First, the reliability of intelligence and security reports, especially anonymous ones, should be questioned in light of their political interests and the documented history of such officials incorrectly reporting basic facts. For instance, Pakistani security officials initially reported that the well-known March 17, 2011 drone strike in Datta Khel destroyed a militant “house” where “a group of some three dozen alleged Taliban fighters were meeting.” Convincing evidence indicates that the strike was actually on an open-air bus depot, where prominent civilian tribal leaders were holding a jirga. “Official” reports from the local government are also problematic because they come through the local political agent, an office notoriously insulated from the community in which it sits and which many suspect will report whatever seems politically expedient at the time.
Local stringers are in many ways a significant improvement over government sources because they have access to people and places unavailable to those outside of FATA. Yet they also face a range of unique pressures and challenges that can limit their usefulness to journalists on the outside. First, some locals are reluctant to speak to stringers about strikes at all, because years of living with ISI, Taliban, and US intelligence operatives in their midst have left them justifiably fearful of retaliation from all sides of the conflict. The ISI, for instance, is widely believed responsible for forcibly disappearing and illegally detaining FATA citizens suspected of militant ties. Paid CIA informants are also rumored to have planted drone-targeting chips on neighbors. Lastly, the Taliban is believed to have avenged drone strikes by killing those it believes to be US spies. Like local contacts, stringers themselves are also under strong pressure from competing local interests, living under constant threat of violence from both armed non-state actors and the Pakistani military if they fail to report information favorable to one side or the other. Indeed, the Tribal Union of Journalists FATA reports that at least ten journalists or stringers have been killed since 2005, and that those still working in the area are subject to intimidation and coercion.
While many outside journalists are conscious of these pressures on their local sources and of the hidden agenda behind government reports, they have very limited options for getting information out of FATA. Corroborating or challenging the divergent reports they receive from officials, stringers, and locals is difficult. As a result, journalists often find themselves in the position of having to choose between reporting “official” casualty figures that they consider untrustworthy, or higher numbers from civilian sources that they may be unable to corroborate. Those who work for major news outlets and wire services tend to spend more time embedded with military and intelligence officials and are thus more likely to report “official” accounts. Those who are not escorted into FATA by the military rely more on locals and stringers. The result is that different journalists with different contacts get different stories, make different decisions about who to trust, and frequently end up publishing conflicting accounts of each strike.
Other Considerations that May Lead to Conflicting Reports
Limited First-Hand Knowledge
Even when journalists are able to get information directly from local residents or stringers, there is no guarantee that those locals actually know the full extent of the casualties around them, even among their own neighbors. Many traditional Waziri families live in large, high-walled, multi-family compounds in which women and young children work, eat, and sleep separately from men. It is generally unacceptable to ask direct questions to a male family member about female relatives, or to photograph women. As a result, male community members may not know details about one another’s families or households, including the exact number of people who live there, and so may not be able to say how many people were inside a home before it was hit by a drone strike. The result is that neighbors and second-hand witnesses may, in some cases, underreport drone strike casualties simply because they do not know the full extent of a given strike’s toll.
At the time of this writing, the US is believed to have conducted 344 total strikes in Pakistan, 52 between June 17, 2004 and January 2, 2009 (under President Bush), and 292 strikes between January 23, 2009 and September 2, 2012 (under President Obama). Those numbers, which TBIJ has pieced together from available media reports, may underestimate the total number of strikes, especially during the early years of the drone program.
Between 2004 and 2007, the Pakistani government under President Musharraf attempted to hide the fact of US strikes (and Pakistan’s role in them) by contending that the strikes were either Pakistani military operations, car bombs, or accidental explosions. Many of those claims were contradicted within days or weeks by anonymous leaks and eyewitness accounts, and by local journalists gathering evidence at the scenes of the attacks. In one unusually well-publicized incident, an official in the Musharraf regime reportedly asserted that the Pakistani military had conducted a strike on a religious school in Bajaur that killed over 80 people, including 69 children. One of Musharraf’s aides reportedly told a Pakistani media source that the government believed “it would be less damaging” to claim it had killed 82 people than it would be to reveal that it had agreed to let the US carry out strikes on Pakistani soil. Musharraf’s administration was reported to admit that the strike had been a US operation only after political backlash from the strike turned out to be much greater than the government had anticipated. Considering the Musharraf government’s apparent efforts to cover up the US’s role in drone strikes, and the fact that drones often target remote or isolated areas, it is possible that other strikes from the 2004-2007 period have yet to be identified.
Our team’s fieldwork in Pakistan documented at least one incident that might fit this pattern. We interviewed 15 Waziris, including four survivors and four more who visited the strike site within hours or days of the attack, who described to us what they believed to have been a drone strike that took place on June 10, 2006. The attack took place in the early morning of June 10 on a workers’ bunkhouse in a chromite mining camp in the mountains near Datta Khel. In the bunkhouse, a large group of young miners and woodcutters were asleep. Missiles killed 22 and badly injured four. The press described the incident as a helicopter gunship attack carried out by the Pakistani military, based on statements by Pakistani officials claiming responsibility. The survivors and those killed were asleep before the first explosion and knocked unconscious shortly thereafter. In light of the classification by media sources (helicopter strike), the lack of available physical evidence given the remoteness of the location, the lack of eyewitness testimony to the source of the strike, and the significant passage of time since the attack, our research team could not determine whether this incident was a US drone strike or Pakistani helicopter strike, and so chose not to include this event as a drone strike. Nonetheless, given the extensive loss of life, this incident should investigated thoroughly by competent authorities.
Strike Data Aggregators
The three most well-known and widely quoted sources of aggregated strike data are the Year of the Drone project by the New America Foundation think tank; The Long War Journal, a blog and project of the Foundation for Defense of Democracies; and TBIJ, a London-based journalism non-profit. Each of these organizations, in seeking to track and aggregate strikes and their impacts, fulfills an important public transparency role. Their data have been invaluable in public debates about drone and targeted killing policies. Given the US government’s failure to provide even basic facts about the strikes, these non-governmental sources are essential.
Nevertheless, the data sets of aggregator organizations have limits. Because consistently reliable information on drone strikes is impossible to come by, none of the online databases that track drone strike reports can provide wholly accurate data either. All three aggregators state that their data is sourced from largely the same universe of publicly available press reports in major western and Pakistani media outlets.
Nonetheless, to determine how many people died in a particular strike and determine whether they were civilians or “militants,” each organization must navigate a morass of contradictory press accounts and opaque intelligence reports, and make several subjective decisions about which sources are more reliable than others. Each uses a different set of categories and labels to classify the victims. Long War Journal uses “civilians” or “Taliban/Al Qaeda,” or “leaders and operatives from Taliban, Al Qaeda, and allied extremist groups.” New America Foundation uses “militant,” “unknown” or “civilians.” TBIJ uses total killed or injured and “civilians,” with no express category for non-civilians. Each aggregator places different weight on different types of primary sources. As a result, the three data aggregators each come to different conclusions about who has been and is being killed by US drone strikes in Pakistan.
For instance, New America Foundation’s Year of the Drone project reports that somewhere between 1,584 and 2,716 “militants” have been killed in Pakistan since 2004, and between 152 and 191 civilians (and 130-268 “unknowns”). The Long War Journal (which does not keep data for 2004 and 2005) reports that drones have killed 2,396 “leaders and operatives from Taliban, Al Qaeda, and allied extremist groups” (which we will refer to as “Taliban/Al Qaeda”) in Pakistan since 2006, and 138 civilians. With the exception of high-value named targets (which are few), neither provides information about the “militant” victims that would indicate whether they were actually lawful targets under international law. TBIJ, which does not use the “militant” label in its data sets, reports that drones have killed between 474 and 881 Pakistani civilians since 2004, out of 2,562 to 3,325 total deaths.
To explain the discrepancies in these figures, we briefly analyze in the section below the methodologies used by each of the three strike-tracking sources to cull and categorize strike reports.
The Long War Journal
The Long War Journal, a project run by the Foundation for the Defense of Democracies, claims that 138 civilians have been killed between 2006 and the present. Unlike the New America Foundation and TBIJ, discussed below, The Long War Journal does not make its data available in a strike-by-strike format. Instead, it publishes blog posts about new strikes soon after they are initially reported, and maintains a series of regularly updated statistical graphs. The strike information in its blog posts is based on reports by major media outlets and on the Journal’s own investigations, which appear to consist primarily of conversations with unnamed “US intelligence officials.” One analysis of drone tallies asserts that The Long War Journal’s methodology places great weight on US intelligence sources, especially when distinguishing between Taliban/Al Qaeda and civilian casualties. According to The Long War Journal’s managing editor, Bill Roggio, for the purposes of categorizing strike deaths, all those killed are counted as “Taliban/Al Qaeda” unless “they are identified as civilians.”
This raises two major concerns about the accuracy of The Long War Journal’s statistical claims. First, because The Long War Journal does not make its data visible in a strike-by-strike format, it is impossible to tell whether and where its editors have logged credibly reported civilian casualties, or to tell whether they update older strike data regularly to reflect new information as it comes to light. The only strike-specific information available on its website comes in the form of blog posts written by managing editor Bill Roggio. Those posts usually appear within twenty-four hours of each new strike, citing initial reports from major media outlets that almost invariably assert that only “Taliban/Al Qaeda” were killed. Second, The Long War Journal’s practice of labeling all drone victims as “Taliban/Al Qaeda” unless they are specifically identified as civilians, combined with its reliance on demonstrably untrustworthy government reports corroborated by comments from anonymous US intelligence sources, raises questions about whether its drone strike statistics underestimate civilian deaths.
New America Foundation
New America Foundation’s Year of the Drone project—the most widely cited in the US of the three strike-tracking sources—currently estimates that 152 to 191 civilians have been killed by drones since 2004, only slightly higher than The Long War Journal’s estimate. One of the New America Foundation’s directors, Peter Bergen, has made headlines recently as a national security analyst for CNN, using New America Foundation’s data to argue that civilian death rates due to drone strikes have dropped to single-digit percentages, and that drones have caused no civilian deaths in Pakistan in 2012. Scrutiny of both assertions has since revealed omissions and inconsistencies in New America Foundation’s dataset, calling its widely publicized conclusions into question.
First, contrary to claims made on its website and in its publications, New America Foundation’s strike data do not appear to be “updated regularly” to include the most up-to-date information about the number and identities of victims killed in drone strikes. Several of New America’s strike descriptions going back to 2006 fail to incorporate a number of credible (and in some cases, high-profile) reports of civilian casualties. For example, New America Foundation reports that a strike on October 31, 2011 killed three to four militants, and makes no mention of “civilian” or “unknown” casualties. That strike, however, was widely reported to have killed two civilian teenagers, 16-year old Tariq Aziz and his cousin Waheed Khan—a fact that has been reported in a variety of western and Pakistani media outlets including BBC, ABC, The Guardian, and Dawn. Similarly, the New America Foundation website reports that a June 15, 2011 strike on a vehicle outside Tapi village killed three to eight militants, and makes no mention of “civilian” or “other” casualties.
However, within days of the attack, at least some credible Pakistani media outlets reported that the strike killed civilians, later identified as Akram Shah, Sherzada, Umar (or Amar) Khan, Irshad Khan, and Atiq-ur-Rehman (Tariq). We detail the circumstances of that strike in the Narrative Section of the Living Under Drones Chapter of this report.
In July 2012, an article by TBIJ also pointed out several other glaring omissions from New America Foundation’s data. These included the confirmed deaths of dozens of children in 2006, and seven civilian deaths confirmed by an AP news investigation to which Bergen himself, along with co-author Jennifer Rowland, had cited in their CNN piece. TBIJ had brought several of these errors to New America’s attention over the previous two years, but New America Foundation had not made any changes or updates in response until very recently. In August 2012, possibly in response to TBIJ’s criticisms, New America Foundation updated its website and incorporated some reports of civilian deaths that it had previously omitted, including the 69 children killed in a single strike in 2006. Others, such as the seven civilian casualties on August 14, 2010 that have been confirmed by an independent AP investigation, were still absent at this writing. “The cumulative effect of all these omissions and errors,” observed TBIJ’s Chris Woods, “is that [New America Foundation’s] data substantially under-estimates both the overall numbers of those killed, and the reports of civilians who have died in Pakistan strikes.”
In addition to its failure to update its database regularly, the underlying data relied upon by New America Foundation must be scrutinized. New America Foundation’s Year of the Drone project is a valuable resource. However, because its data consist of a collection of news reports, the conclusions that can definitively be drawn from analyzing that dataset are limited and must be attenuated in important ways. For example, when Bergen and Rowland asserted in their July 14, 2012 CNN column that New America’s data showed no civilian deaths in 2012, our team reviewed every news article New America linked to on its website in support of its 2012 drone strike statistics. The inadequacies in this underlying data (detailed below) mean that it should not be used to support the conclusions drawn by Bergen and Rowland (and New America Foundation) that there have been no civilian deaths in US drone strikes in Pakistan in 2012:
- First, the articles cited by New America Foundation rely to an overwhelming extent on information provided by anonymous officials. Our team’s review of the dataset for 2012 (the most recent strike considered being July 6, 2012) found that anonymous officials are cited as a source for the allegation of the number of “militants” killed in 88% of articles referenced by New America Foundation, and are the only source of this information in 74% of the articles. When framed as a breakdown of sources per strike, anonymous officials are the only source of the number of “militants” killed in 16 of the 27 drone strikes. This heavy reliance on anonymous officials is troubling given the demonstrated unreliability of official reporting;
- Second, the conclusion that no civilians have been killed in 2012 overlooks the problem of identification referenced in a number of the articles in the dataset. In 15 articles, it was noted that those killed could not be identified or that the identities of victims were not known. For example, in one such instance, an anonymous official stated that: “Fifteen militants were killed in a dawn strike on a compound. The bodies of those killed were unable to be identified.” Furthermore, 18 articles in the dataset refer to the object of attack as being “destroyed”, reinforcing concerns about how the number of persons killed and their identities could be known.
Thus, what can fairly be concluded from analyzing New America Foundation’s dataset is that, according to anonymous officials quoted in a set of collected news reports, there have been no civilian deaths reported in 2012.
New America Foundation’s finding of no civilians killed in 2012 is also troubling given that “reputable news sources” have suggested the possibility of civilian casualties in six of the 27 strikes that inform New America Foundation’s 2012 statistics. Those sources include Reuters, Agence France-Presse, The News, and Dawn, all of which New America Foundation has found reliable on other occasions when they reported only “militant” casualties. Bergen and Rowland’s July 14 CNN piece does not explain why they chose to disregard those news sources when they report civilian casualties. Instead, Bergen and Rowland attempt to head off criticism by singling out TBIJ and dismissing their contradictory estimate of three to 24 civilian casualties as coming “in part from reports provided by an unreliable Pakistani news outlet as well as the claims of a local Taliban commander.” TBIJ explained in response that the “unreliable Pakistani news outlet” must refer to either Dawn, The Nation, or The News, all of which New America Foundation draws from on a regular basis, and that the Taliban commander’s claim (which appeared in only one of the six strikes in which civilian casualties were reported, and which referred to only two civilians) appeared in an article from Reuters. Bergen and Rowland did not say where they believe the other part of TBIJ’s estimate came from.
Conor Friedersdorf of the Atlantic Monthly has questioned the reliance of Bergen and Rowland and the New America Foundation on “getting an unnamed official to state the number of deaths” as “deep reporting” worthy of inclusion in their database. In particular, Friedersdorf juxtaposes that reliance with the journalists’ apparent exclusion of further reporting above and beyond anonymous official quotes as unreliable. For example, neither the Year of the Drone website nor any of Bergen and Rowland’s articles mentions the reported deaths of between three and eight civilian worshippers at a mosque on May 24, 2012. The deaths were reported by both The News, a prominent Pakistani newspaper, and the UK’s Channel 4. Both quoted detailed descriptions of the strike and of the civilian casualties directly from a local eyewitness that The News identifies by name. That level of detail and local investigation constitutes a far “deeper” report than the terse descriptions from anonymous officials, with one exception, that appear in the articles relied upon by New America Foundation, which in turn simply state the number of “militants” or “suspected militants” killed and their nationalities.
The Bureau of Investigative Journalism
The Bureau of Investigative Journalism maintains a much more dynamic database than either New America Foundation or The Long War Journal, updating its strike information frequently to reflect new information as it comes to light. This frequent updating, together with TBIJ’s own investigations, makes its data far more reliable than other aggregating sources. While TBIJ’s data are also highly transparent and its investigations more thorough than others, its aggregation of information from news articles faces the same problems as described above, and its full body of strike data is not, and indeed cannot be, wholly accurate (nor does TBIJ purport that it is).
As of August 1, 2012, TBIJ estimated that between 482 and 849 civilians have been killed by drones in Pakistan since 2004. That estimate represents the full range of civilian casualties credibly reported in reliable sources, some of which TBIJ has corroborated with its own field investigations in Pakistan and with information gathered by “credible researchers and lawyers.” The use of these corroborating sources to supplement data drawn from press accounts sets TBIJ apart from both The Long War Journal and New America Foundation.
TBIJ’s media datasets are also more thorough and comprehensive than both New America Foundation and The Long War Journal. As discussed above, New America Foundation linked to only 107 news articles in support of its data on the first 27 strikes of 2012, of which eleven were duplicates. TBIJ, by contrast, links to 344 sources cited in support of those same 27 strikes, and provides information on a handful of additional possible strikes that have not yet been verified. The Long War Journal does not reveal all of the sources used to compile its database, and rarely cites to more than two or three external sources in any given report. TBIJ is also more transparent than either New America Foundation or The Long War Journal in its reporting, providing both high and low estimates of civilian and unspecified deaths for each strike. It also quotes heavily from reports that contradict one another, thus giving a full picture of the range of conflicting stories about each strike.
 Jo Becker & Scott Shane, Secret ‘Kill List’ Proves a Test of Obama’s Principles and Will, N.Y. Times (May 29, 2012), http://www.nytimes.com/2012/05/29/world/obamas-leadership-in-war-on-al-qaeda.html?pagewanted=all; see also infra note 156.
 See David Rohde, The Drone War, Reuters (Jan. 26, 2012), http://www.reuters.com/article/2012/01/26/us-david-rohde-drone-wars-idUSTRE80P11I20120126 (observing, in the context of Afghanistan and Pakistan, that “militants use exaggerated reports of civilian deaths to recruit volunteers and stoke anti-Americanism”).
 See infra Chapter 4: Legal Analysis.
 Becker & Shane, supra note 144.
 Philip Alston, the former United Nations Special Rapporteur on extrajudicial, summary, or arbitrary executions, has explained that a person who merely engages in “political advocacy, supplying food or shelter, or economic support or propaganda” for Al Qaeda or its affiliates is not a legitimate target under international humanitarian or human rights law, because such conduct does not rise to the level of direct participation in hostilities. Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions, Study on Targeted Killing, ¶ 57-69, Human Rights Council, UN Doc A/HRC/14/24/Add.6 (May 28, 2010) (by Philip Alston); see also infra Chapter 4: Legal Analysis.
 Adam Entous, Special Report: How the White House Learned to Love the Drone, Reuters (May 18, 2010), http://www.reuters.com/article/2010/05/18/us-pakistan-drones-idUSTRE64H5SL20100518; see Peter Bergen & Jennifer Rowland, CIA Drone War in Pakistan in Sharp Decline, CNN (Mar. 28, 2012), http://www.cnn.com/2012/03/27/opinion/bergen-drone-decline/index.html.
 Entous, supra note 149.
 Peter Bergen & Megan Braun, Drone is Obama’s Weapon of Choice, CNN (Sept. 6, 2012), http://www.cnn.com/2012/09/05/opinion/bergen-obama-drone/index.html.
 See infra notes 241-269 and accompanying text.
 See infra note 187 and accompanying text.
 Most notably, the President’s top counterterrorism advisor, John O. Brennan, claimed in June 2011 that the US had not killed a single civilian since August 23, 2010. See Obama Administration Counterterrorism Strategy (C-Span television broadcast June 29, 2011), http://www.c-spanvideo.org/program/AdministrationCo; see also Chris Woods, US Claims of ‘No Civilian Deaths’ are Untrue, The Bureau of Investigative Journalism (July 18, 2011), http://www.thebureauinvestigates.com/2011/07/18/washingtons-untrue-claims-no-civilian-deaths-in-pakistan-drone-strikes/.
 The Long War Journal, a project of the Foundation for Defense of Democracies, claims that drones have caused 138 civilian deaths since 2006. Bill Roggio & Alexander Mayer, Charting the Data for US Airstrikes in Pakistan, 2004—2012, Long War Journal, http://www.longwarjournal.org/pakistan-strikes.php (last updated Sept. 16, 2012). Bill Roggio, the Long War Journal’s managing editor, was quoted in 2011 as saying “the C.I.A.’s claim of zero civilian casualties in a year is absurd.” Scott Shane, C.I.A. is Disputed on Civilian Toll in Drone Strikes, N.Y. Times (Aug. 11, 2011), http://www.nytimes.com/2011/08/12/world/asia/12drones.html?pagewanted=all.
 A study released by the Jamestown Foundation in late 2010 found that 68 people killed by drones in Pakistan since 2004 “could be clearly identified as civilians.” Bryan Glyn Williams, Matthew Fricker, & Avery Plaw, New Light on the Accuracy of the CIA’s Predator Drone Campaign in Pakistan, 41 Terrorism Monitor 8 (2010).
 Colonel David M. Sullivan, an Air Force pilot with “extensive experience with both traditional and drone airstrikes” told the New York Times that the US figures “do not sound . . . like reality.” Shane, C.I.A. is Disputed on Civilian Death Toll in Drone Strikes, supra note 157.
 John O. Brennan, Assistant to the President for Homeland Security and Counterterrorism, Remarks at the Woodrow Wilson International Center for Scholars (Apr. 30, 2012), available at http://www.wilsoncenter.org/event/the-efficacy-and-ethics-us-counterterrorism-strategy.
 Becker & Shane, supra note 144.
 In 2009, an unnamed US official told the New York Times that the US had killed “just over 20” civilians in the two preceding years. Scott Shane, C.I.A. to Expand Use of Drones in Pakistan, N.Y. Times (Dec. 3, 2009), http://www.nytimes.com/2009/12/04/world/asia/04drones.html?pagewanted=all. Five months later, officials claimed the number since 2008 remained under 30. David S. Cloud, CIA Drones Have Broader List of Targets, L.A. Times (May 5, 2010), http://articles.latimes.com/2010/may/05/world/la-fg-drone-targets-20100506. A recent article comparing statements given to the press by US officials found that the Obama administration’s civilian death estimates over the last two years have vacillated between 0 and 50. See Justin Elliott, Obama Administration’s Drone Death Figures Don’t Add Up, ProPublica (June 18, 2012), http://www.propublica.org/article/obama-drone-death-figures-dont-add-up.
 Becker & Shane, supra note 144.
 Newspaper accounts of drone strikes sometimes note that the bodies of strike victims are too damaged to be identified. See, e.g., Drone Strike Kills 14 in NWA, News (July, 24, 2012), http://www.thenews.com.pk/Todays-News-13-16297-Drone-strikes-kill-14-in-NWA (“[B]odies were damaged beyond recognition.”); Haji Mujtaba, US Drone Attack Kills 10 in Pakistan: Officials, Reuters (Feb. 8, 2012), http://in.reuters.com/article/2012/02/08/pakistan-drone-idINDEE81701N20120208 (“Almost all the men were burnt beyond recognition.”); US Drone Attack Kills 10 in North Waziristan, Daily Times (Feb. 9, 2012), http://www.dailytimes.com.pk/default.asp?page=2012%5C02%5C09%5Cstory_9-2-2012_pg7_4 (“‘Almost all the men were burnt beyond recognition,’ a villager said.”). Several interviewees also told us that the bodies recovered from strike sites are mutilated and burned beyond recognition. See, e.g., Interview with Ismail Hussain (anonymized name), in Islamabad, Pakistan (Feb. 26, 2012) (“[T]heir bodies were totally destroyed. . . . We can’t say that it is exactly four persons [that were killed]. It could be five or six as well because they were cut into pieces. We couldn’t identify them.”); supra Chapter 3: Living Under Drones.
 US officials told the New York Times that the CIA and NSA investigate drone casualties by watching the aftermath of strikes by video, and “track[ing] the funerals that follow.” Shane, C.I.A. is Disputed on Civilian Death Toll in Drone Strikes, supra note 157. They further “intercept cell phone calls and emails discussing who was killed.” Id. The sufficiency of this method of post-strike investigation is questionable, given frequently poor cell signals in the area, and given that most households do not have the electricity or infrastructure to support an internet connection. See Tayyeb Afridi, Would Social Media Bring Change to Pakistan’s Tribal Area?, KUTNews Austin (May 25, 2011), http://kutnews.org/post/would-social-media-bring-change-pakistan%E2%80%99s-tribal-area (noting that social media and internet service are generally unavailable in FATA due to lack of electricity, and high cost); Interview with Noor Behram, in Islamabad, Pakistan (Mar. 9, 2012).
 TBIJ was founded to produce “high quality investigations for press and broadcast media with the aim of educating the public and the media on both the realities of today’s world and the value of honest reporting.” About the Bureau, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/who/ (last visited Sept. 8, 2012). It was founded in 2010 with a grant from the David and Elaine Potter Foundation, a British charity dedicated to promoting “reason, education, and human rights” around the world. David & Elaine Potter Foundation, http://www.potterfoundation.com/ (last visited Sept. 8, 2012).
 Shane, C.I.A. is Disputed on Civilian Death Toll in Drone Strikes, supra note 157.
 Chris Woods & Christina Lamb, Obama Terror Drones: CIA Tactics in Pakistan Include Targeting Rescuers and Funerals, The Bureau of Investigative Journalism (Feb. 4, 2012), http://www.thebureauinvestigates.com/2012/02/04/obama-terror-drones-cia-tactics-in-pakistan-include-targeting-rescuers-and-funerals/.
 Scott Shane, US Said to Target Rescuers at Drone Strike Sites, N.Y. Times (Feb. 5, 2012), http://www.nytimes.com/2012/02/06/world/asia/us-drone-strikes-are-said-to-target-rescuers.html.
 Scott Shane, the author of both articles, was criticized by Harvard’s Nieman Foundation for Journalism for attributing personal attacks to anonymous sources, which they said violates the New York Times’ ethical policies governing the use of confidential sources. John Hanrahan, Why is the New York Times Enabling a US Government Smear Campaign Against Reporters Exposing the Drone Wars?, Nieman Watchdog (May 11, 2012), http://www.niemanwatchdog.org/index.cfm?fuseaction=ask_this.view&askthisid=00562&forumaction=post. In written correspondence with Nieman Watchdog, Shane defended his use of the anonymous quotes by explaining that these anonymous comments were all he was able to get from the US, and that he has to use them in order to “include some voice from the other side.” Id.
 Becker, supra note 144.
 Interview with W.K., journalist with major western news source (anonymized initials), and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012); Interview with G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 After the New York Times piece, media sources have continued to rely on anonymous government sources and tend to report that strikes have killed “militants” or “suspected militants.” See, e.g., Nasir Habib, Suspected Drone Attack Kills 12 in Pakistan, CNN (July 23, 2012), http://edition.cnn.com/2012/07/23/world/asia/pakistan-drone-attack/index.html; Salman Masood & Ihsanullah Tipu Mehsud, 15 Killed in US Drone Strike in Pakistan, N.Y. Times (July 6, 2012), http://www.nytimes.com/2012/07/07/world/asia/15-killed-in-us-drone-strike-in-pakistan-aimed-at-taliban.html?_r=1; US Drone Strike ‘Kills At Least Five’ in North Waziristan, Express Tribune (June 26, 2012), http://dawn.com/2012/06/26/us-drone-strike-kills-at-least-five-in-north-waziristan/ (reprinting Agence France-Presse story); Drone Strike Kills 4 in Pakistan Ahead of Allen Talks, CNN (June 26, 2012), http://edition.cnn.com/2012/06/26/world/asia/pakistan-drone-strike/index.html; US Drone Kills Nine in North Waziristan, Nation (July 6, 2012), http://www.nation.com.pk/pakistan-news-newspaper-daily-english-online/islamabad/06-Jul-2012/us-drone-strike-kills-4-in-nwaziristan (reprinting Agence France-Presse story); US Drone Kills Seven Militants in North Waziristan: Officials, Dawn (July 29, 2012), http://dawn.com/2012/07/29/us-drone-strike-kills-four-in-north-waziristan-2/ (reprinting Agence France-Presse story); US Drone Strike Kills Six Militants in Pakistan: Officials, Express Tribune (July 1, 2012), http://tribune.com.pk/story/401902/us-drone-strike-kills-six-militants-in-pakistan-officials/ (reprinting Agence France-Presse story); US Drone Kills 12 Suspected Militants in Pakistan, Reuters (July 23, 2012), http://uk.reuters.com/article/2012/07/23/uk-pakistan-drone-idUKBRE86M13G20120723; Mushtuq Yusufzai, US Drone Kills 8 Suspected Militants in Pakistan Hideout, MSNBC.com (July 1, 2012), http://worldnews.nbcnews.com/_news/2012/07/01/12504073-us-drone-kills-8-suspected-militants-in-pakistan-hideout?lite.
 See Conflicting Media Reports, infra Chapter 2: Numbers.
 Rasool Dawar, Pakistan Officials Say US Drone Kills 10 Militants, AP Worldstream (May 24, 2012).
 See Hasbanullah Khan, ‘Five Militants Killed’ by US Drone in Pakistan, Agence France-Presse (May 23, 2012), http://www.google.com/hostednews/afp/article/ALeqM5h_NYzU3o4KUWtTBouMGdzKeOhqjw?docId=CNG.12419227fd3472cf2255b588417525f8.341; US Drone Kills 10 in Pakistan, Irish Times (May 24, 2012), http://www.irishtimes.com/newspaper/breaking/2012/0524/breaking8.html.
 See supra note 177 and accompanying text. The AFP article did mention, however, that in addition to “insurgents” being killed, there were reports that “a nearby mosque where three worshippers believed to be Central Asian nationals were wounded.” See Khan, supra note 177.
 See, e.g., Drone Strike Hits Pakistan Mosque, Say Locals, Channel 4 (May 24, 2012), http://www.channel4.com/news/us-drone-attack-hits-pakistan-mosque; Malik Mumtaz Khan & Mushtaq Yusufzai, 10 Killed in Drone Attack on NWA Mosque, News (May 25, 2012), http://www.thenews.com.pk/Todays-News-13-14861-10-killed-in-drone-attack-on-NWA-mosque; US Drone Strike Hits Mosque: 10 Killed, Nation (May 25, 2012), http://www.nation.com.pk/pakistan-news-newspaper-daily-english-online/national/25-May-2012/us-drone-strike-hits-mosque-10-killed.
 See, e.g., Drone Strike Hits Pakistan Mosque, Say Locals, supra note 180; Khan & Yusufzai, supra note 180.
 See, e.g., Khan, supra note 177; Mushtaq Yusufzai, Pakistan Official: US Drone Strike Hits Mosque; 10 Killed, MSNBC.com (May 23, 2012), http://worldnews.msnbc.msn.com/_news/2012/05/23/11839215-pakistan-official-us-drone-strike-hits-mosque-10-killed?lite.
 Chris Brummitt & Riaz Khan, Pakistan Convicts Doctor Who Helped Find bin Laden, AP Worldstream (May 24, 2012), http://bigstory.ap.org/content/pakistan-convicts-doctor-who-helped-find-bin-laden-0; Haji Mujtaba, UPDATE 2-US Drone Strike Kills 10 in Northwest Pakistan—Officials, Reuters (May 24, 2012), http://www.reuters.com/article/2012/05/24/us-pakistan-drone-idUSBRE84N03I20120524.
 Rasool Dawar, Associated Press, Pakistan: US Missiles Kill 5 Militants in NW, Guardian (May 28, 2012), http://www.guardian.co.uk/world/feedarticle/10261577.
 The Pakistani military occasionally helicopters embedded journalists from American media outlets into FATA for just a few hours at a time. Interview with G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with K.N., journalist with major western news source (anonymized initials) in Islamabad, Pakistan (Mar. 5, 2012).
 Interview with G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 Dozens Die as US Drone Hits Pakistan Home, Al Jazeera (Mar. 17, 2011), http://www.aljazeera.com/news/asia/2011/03/20113178411386630.html. AFP also reported a security official’s claim about missiles striking a “militant training centre.” Militants Killed in Pakistan Drone Strikes, ABC News (Mar. 17, 2011), http://www.abc.net.au/news/2011-03-17/militants-killed-in-pakistan-drone-strikes/2654524.
 See March 17, 2011 Strike Narrative, infra Chapter 3: Living Under Drones.
 Interview with Samina Ahmad, International Crisis Group, in Islamabad, Pakistan (Feb. 28, 2012); Interview with W.K., journalist with major western news source (anonymized initials) and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012); Interview with Noor Behram, in Islamabad, Pakistan (Mar. 9, 2012).
 Interview with journalists W.K., journalist with major western news source (anonymized initials) and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012).
 Id.; Interview with journalist G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with journalist K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 See, e.g., Waseem Ahmad Shah, Illegal Detentions: Court Tells Army to Rein In Errant Agencies, Dawn (Apr. 13, 2011), http://dawn.com/2012/04/13/illegal-detentions-court-tells-army-to-rein-in-errant-agencies/; Declan Walsh, Court Challenges Put Unusual Spotlight on Pakistani Spy Agency, N.Y. Times (Feb. 6, 2012), http://www.nytimes.com/2012/02/07/world/asia/isi-in-pakistan-faces-court-cases.html?pagewanted=all.
 See Beyond Killing: Civilian Impacts of US Drone Strike Practices, infra Chapter 3: Living Under Drones.
 See, e.g., Taliban Shoot Dead Four ‘US Spies’ in North Waziristan, Dawn (Mar. 21, 2011), http://dawn.com/2011/03/21/taliban-shoot-dead-four-us-spies-in-north-waziristan/.
 See, e.g., Amirzada Afridi, FATA Journalists: The Forgotten Scribes of a Secret War, Express Tribune (Sept. 10, 2011), http://tribune.com.pk/story/249142/fata-journalists-the-forgotten-scribes-of-a-secret-war/; Ikram Junaidi, FATA Journalists on Razor’s Edge, Dawn (Mar. 1, 2012), http://dawn.com/2012/03/01/fata-journalists-on-razors-edge/ (“President [of the] Tribal Union of Journalists Safdar Hayat Dawar . . . alleged that both the military and Taliban forced media persons to file stories of their choice, adding [that] both didn’t care about human rights.”); Rahimullah Yusufzai, Pakistani Journalists Under Siege, Newsline (Feb. 29, 2012), http://www.newslinemagazine.com/2012/02/pakistani-journalists-under-siege/; Micah Zenko, The Courage of Pakistani Journalists, Atlantic (Sept. 20, 2011), http://www.theatlantic.com/international/archive/2011/09/the-courage-of-pakistani-journalists/245358/.
 See, e.g., Government Urged to Ensure Security of Journalists in FATA, Daily Times (Mar. 2, 2012), http://www.dailytimes.com.pk/default.asp?page=2012%5C03%5C02%5Cstory_2-3-2012_pg7_19.
 Interview with journalists W.K., journalist with major western news source (anonymized initials) and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012); Interview with journalist G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012).
 Interview with journalists W.K., journalist with major western news source (anonymized initials) and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012); Interview with journalist G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with journalist K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 Interview with journalist G.Z., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 7, 2012); Interview with journalist K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 Interview with journalists W.K., journalist with major western news source (anonymized initials) and journalists with Pakistani news outlets, in Islamabad, Pakistan (Mar. 3, 2012); Interview with journalist K.N., journalist with major western news source (anonymized initials), in Islamabad, Pakistan (Mar. 5, 2012).
 See, e.g., Interview with Ibrahim Shah, in Islamabad, Pakistan (May 9, 2012) (telling us he lives in a large extended family compound of 50-60 relatives).
 Population Demography, Government of Pakistan Federally Administered Tribal Area Secretariat, http://fata.gov.pk/index.php?option=com_content&view=article&id=56&Itemid=92 (last visited Sept. 1, 2012) (noting that in FATA “tribal custom forbids the disclosure of information about women to outsiders”).
 The Bush Years: Pakistan Strikes 2004-2009, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/2011/08/10/the-bush-years-2004-2009/ (last visited Sept. 1, 2012).
 Covert War On Terror—The Data, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/category/projects/drone-data/ (last visited Sept. 16, 2012).
 Covert US Strikes in Pakistan, Yemen, and Somalia—Our Methodology, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/2011/08/10/pakistan-drone-strikes-the-methodology2/ (last visited Sept. 1, 2012).
 Gareth Porter, Why Pakistani Military Demands a Veto on Drone Strikes, InterPress Service (Aug. 16, 2011), http://ipsnews.net/news.asp?idnews=56873; see also Accidental Blast While Assembling Bomb Kills Eight, Gulf News (Nov. 6, 2005) (reporting Pakistani army officials’ claim that the November 5, 2005 strike was caused by militants who “set off a blast while making bombs at their compound”), http://gulfnews.com/news/world/pakistan/accidental-blast-while-assembling-bombs-kills-eight-1.443552; CIA Drone Kills al-Qaeda Operative, MSNBC.com (May 14, 2005) (reporting Pakistani officials claim May 8, 2005 strike was car bomb explosion), http://www.msnbc.msn.com/id/7847008/ns/us_news-security/t/cia-drone-kills-al-qaida-operative/#.T9VqdeJYvAs; Anwarullah Khan, 82 Die as Missiles Rain on Bajaur: Pakistan Owns Up to Strike; Locals Blame US Drones, Dawn (Oct. 31, 2006) (reporting Pakistani officials insist October 30, 2006 strike on a madrassa that killed 69 children was a Pakistan Army operation), http://archives.dawn.com/2006/10/31/top1.htm; Ismail Khan, Senior Al Qaeda Commander Killed, Dawn (Dec. 3, 2005) (reporting Pakistani authorities claim December 1, 2005 strike was “the result of an explosion inside the house”), http://archives.dawn.com/2005/12/03/top4.htm; Iqbal Khattak, Nek Killed in Missile Strike, Daily Times (June 19, 2004) (reporting the Pakistani military and intelligence sources claim to have carried out June 14, 2004 strike using “US-provided night-capable helicopter”), http://www.dailytimes.com.pk/default.asp?page=story_19-6-2004_pg1_1; US Drone Attack? It Was Us, Says Pakistan Army, Reuters (Jan. 19, 2007) (reporting Pakistani military insists January 19, 2007 strike was conducted by “helicopter gunships”), http://www.reuters.com/article/2007/01/19/us-pakistan-usa-idUSSP30752020070119.
 See, e.g., Ismail Khan, supra note 208 (contradicting official reports to quote witnesses saying that both Nov. 5, 2005 and Dec. 1, 2005 strikes were drone operations, and that the first had killed a woman and children); Ismail Khan & Dilawar Khan Wazir, Night Raid Kills Nek, Four Other Militants: Wana Operation, Dawn (Jun. 19, 2004) (speculating that June 18, 2004 strike may have been a targeted missile from a “spy drone”), http://archives.dawn.com/2004/06/19/top1.htm; Dana Priest, Surveillance Operation in Pakistan Located and Killed al Qaeda Official, Wash. Post (May 15, 2005) (revealing that May 8, 2005 strike was conducted by a CIA Predator drone), http://www.washingtonpost.com/wp-dyn/articles/A60743-2005May15.html.
 Local FATA journalist Hayatullah Khan was the first to gather conclusive evidence of US involvement in a drone strike when he photographed Hellfire missile shrapnel in the rubble of a December 2005 strike that killed two children. See A Journalist in the Tribal Areas, Frontline (Oct. 3, 2006), http://www.pbs.org/wgbh/pages/frontline/taliban/tribal/hayatullah.html; see also House-Owner Called After Missile Attack, Dawn (Dec. 5, 2005), http://archives.dawn.com/2005/12/05/top3.htm. Khan was abducted four days afterward on December 5, 2005, and his body dumped in a ditch six months later with gunshot wounds to the back of the head and government-issued handcuffs on his wrists. See A Journalist in the Tribal Areas, supra. Many major Pakistani news outlets speculated that the Musharraf regime abducted and killed Khan in retaliation for exposing their fabrications and complicity with US strikes. See Cable from Ambassador Ryan C. Crocker, US Embassy Islamabad, Subject: Fata: Missing Pakistani Journalist Found Dead in Waziristan (Jun. 20, 2006), available at http://www.cablegatesearch.net/cable.php?id=06ISLAMABAD11675&q=hayatullah%20khan.
 Anwarullah Khan, supra note 208; see Yousuf Ali, Most Bajaur Victims Were Under 20, News (Nov. 5, 2006), http://www.thenews.com.pk/TodaysPrintDetail.aspx?ID=4043&Cat=13&dt=11/5/2006; Porter, supra note 208.
 Americans Bombed the Bajaur Madrassa, Daily Times (Nov. 27, 2006), http://www.dailytimes.com.pk/default.asp?page=2006%5C11%5C27%5Cstory_27-11-2006_pg1_3 (quoting Musharraf aide).
 Id.; Porter, supra note 208.
 See Interview with Yaser Abdullah (anonymized name) in Islamabad, Pakistan (Mar. 9, 2012); Interview with Masood Afwan (anonymized name), in Islamabad, Pakistan (Feb. 26, 2012); Interview with Marwan Aleem (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Aftab Gul Ali (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Khalil Arshad (anonymized name) in Islamabad, Pakistan (Mar. 9, 2012); Interview with Umar Ashraf (anonymized name) in Islamabad, Pakistan (Mar. 9, 2012); Interview with Ajmal Bashir (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Mohsin Haq (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Dawood Ishaq (anonymized name) in Islamabad, Pakistan (Mar. 8, 2012); Interview with Maher Jabbar (anonymized name) in Islamabad, Pakistan (Mar. 9, 2012); Interview with Dannesh Jameel (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Shahbaz Kabir (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Haidar Nauman (anonymized name) in Islamabad, Pakistan (Mar. 9, 2012); Interview with Noor Shafeeq (anonymized name) in Islamabad, Pakistan (Feb. 26, 2012); Interview with Arman Yousef (anonymized name), in Islamabad, Pakistan (Feb. 26, 2012).
 The Pakistani military asserted that the June 10, 2006 attack was carried out by “4 gunship helicopters and artillery,” and that “explosive material in [the building] started to explode,” killing the “militants” inside. Security Forces Kill 20 Militants Near Pak-Afghan Border, Pak Tribune (June 11, 2006), http://paktribune.com/news/Security-forces-kill-20-militants-near-Pak-Afghan-border-146479.html.
 That Pakistani authorities accepted responsibility for the attack should not be viewed as dispositive. In several instances between 2005 and 2007, missile strikes initially claimed by authorities to have been executed by the Pakistani military were later shown to have been drone strikes. See, e.g., David Rohde & Mohammed Khan, Ex-Fighter for Taliban Dies in Strike in Pakistan, N.Y. Times (June 19, 2004), http://www.nytimes.com/2004/06/19/international/asia/19STAN.html; see also Ismail Khan, Senior Al Qaeda Commander Killed, Dawn (Dec. 2, 2005), http://archives.dawn.com/2005/12/03/top4.htm; Ishtiaq Mahsud, Tribe: US, Not Pakistan, Hit Village, Wash. Post (Jan. 19, 2007), http://www.washingtonpost.com/wp-dyn/content/article/2007/01/19/AR2007011900472.html; 3 Killed in Mysterious Explosion in North Waziristan: Tribesmen Warn of Ending Peace Deal, Daily Times (Apr. 28, 2007),http://www.dailytimes.com.pk/default.asp?page=2007%5C04%5C28%5Cstory_28-4-2007_pg7_1.
 One piece of evidence requiring further research is the observation, by one interviewee, that a piece of shrapnel bore British identification. Arman Yousef (anonymized name), who lost his son in the incident, told our researchers, “[w]e collect parts of the missiles. When my son was killed, I saw a part of the missile—it said ‘Made in Britain.’” Interview with Arman Yousef (anonymized name), in Islamabad, Pakistan (Feb. 26, 2012).
 The Year of the Drone, New America Foundation, http://counterterrorism.newamerica.net/drones (last visited July 31, 2012); Pakistan Body Count also tracks suicide bombings and drone attacks. See Pakistan Body Count, http://www.pakistanbodycount.org/.
 See, e.g., Covert Strikes in Pakistan, Yemen, and Somalia—Our Methodology, THE BUREAU OF INVESTIGATIVE JOURNALISM, http://www.thebureauinvestigates.com/2011/08/10/pakistan-drone-strikes-the-methodology2/ (sources include, inter alia, research publications, governmental documents, and media sources that include “CNN, MSNBC, ABC News, Fox News, Reuters, the BBC, Associated Press, the Guardian, the Telegraph, the Independent, TIME, the Wall Street Journal, the Washington Post, the New York Times, the Los Angeles Times, the Nation, the Atlantic, Salon, Xinhua, Army Times, Navy Times, Bloomberg, AFP, NPR, Al Jazeera, and Al Arabiya” ); Roggio & Mayer, supra note 157 (stating that Long War Journal data is obtained from “press reports from the Pakistani press (Daily Times, Dawn, Geo News, The News, and other outlets), as well as wire reports (AFP, Reuters, etc.), as well as reporting from the Long War Journal”); The Year of the Drone: An Analysis of US Drone Strikes in Pakistan, 2004-2012, New America Foundation, http://counterterrorism.newamerica.net/drones (last visited Sept. 16, 2012) (stating that its database “draws only on accounts from reliable media organizations with deep reporting capabilities in Pakistan, including 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 English-language newspapers in Pakistan—the Daily Times, Dawn, the Express Tribune, and the News—as well as those from Geo TV, the largest independent Pakistani news network.”).
 Roggio & Mayer, supra note 157.
 The Year of the Drone, supra note 221.
 Covert War on Terror—The Data, supra note 206.
 The Year of the Drone, supra note 221.
 Roggio & Mayer, supra note 157. Long War Journal does not keep drone strike data for the years 2004 and 2005. Id.
 See Bergen & Rowland, supra note 152; Entous, supra note 149.
 Covert War on Terror—The Data, supra note 206.
 Roggio & Mayer, supra note 157.
 See, e.g., Bill Roggio, Latest US Drone Strike Kills 10 ‘Militants’ in South Waziristan, Long War Journal (June 3, 2012), http://www.longwarjournal.org/archives/2012/06/latest_us_drone_stri.php; Bill Roggio, North Waziristan Drone Strike Kills 4 ‘Militants’, Long War Journal (June 13, 2012), http://www.longwarjournal.org/archives/2012/06/us_drone_strike_kill_7.php; Bill Roggio, US Drones Kill 15 in North Waziristan, Long War Journal (June 4, 2012), http://www.longwarjournal.org/archives/2012/06/us_drone_kill_15_in.php.
 Avery Plaw, Matthew S. Fricker, & Brian Glyn Williams, Practice Makes Perfect? The Changing Civilian Toll of CIA Drone Strikes in Pakistan, 5 Perspectives on Terrorism 51, 58 (Dec. 2011)(observing that “the Long War Journal relies heavily on U.S. intelligence sources.”). Plaw, Fricker, and Williams have generated numerous reports using their own strike database, currently known as the UMassDRONE project, but have not made it available to the public. See, e.g., id.; Williams, Fricker, & Plaw, supra note 158, at 8.
 See Sharon Weinberger, Pakistani Scholar Disputes US Drone Death Tallies, AOL News (May 19, 2010) (quoting Bill Roggio as saying that “I’m using the opposite approach . . . I only count when they are identified as civilians.”), http://www.aolnews.com/2010/05/19/pakistani-scholar-disputes-low-drone-death-tallies/.
 See, e.g., Roggio, Latest US Drone Strike Kills 10 ‘Militants’ in South Waziristan, supra note 231; Roggio, North Waziristan Drone Strike Kills 4 ‘Militants’, supra note 231; Roggio, US Drones Kill 15 in North Waziristan, supra note 231; Bill Roggio, US Drones Strike in Miramshah’s Bazaar, Kill 3 Militants, Long War Journal (June 14, 2012), http://www.longwarjournal.org/archives/2012/06/us_drones_strike_in_1.php.
 Weinberger, supra note 233 (quoting Long War Journal analyst Bill Roggio).
 Civilian death toll estimates are a recent addition to the Year of the Drone website, which, until August 2012, tallied all drone-related deaths as “militant” and “others.” See Year of the Drone, supra note 221 (as it appeared through August 12, 2012) (copy on file with authors).
 Bergen, along with fellow New America Foundation analyst Jennifer Rowland, stated in March 2012 that the 2011 civilian drone strike casualty rate in Pakistan was 7%. Peter Bergen & Jennifer Rowland, CIA Drone War in Pakistan in Sharp Decline, CNN (Mar. 28, 2012), http://us.cnn.com/2012/03/27/opinion/bergen-drone-decline/index.html?hpt=op_t1. In June 2012, Bergen and Rowland said the rate was actually 5.5%, but did not point out the adjustment or explain how they arrived at the lower figure. Peter Bergen & Jennifer Rowland, Obama Ramps Up Covert War in Yemen, CNN (June 12, 2012), http://edition.cnn.com/2012/06/11/opinion/bergen-yemen-drone-war/index.html?iref=allsearch. In July 2012, they raised the 2011 casualty rate figure to 6%, but again did not explain the adjustment. Peter Bergen & Jennifer Rowland, Drones Decimating Taliban in Pakistan, CNN (July 4, 2012), http://edition.cnn.com/2012/07/03/opinion/bergen-drones-taliban-pakistan/index.html?iref=allsearch.
 See, e.g., Peter Bergen & Jennifer Rowland, Civilian Casualties Plummet in Drone Strikes, CNN (July 14, 2012), http://edition.cnn.com/2012/07/13/opinion/bergen-civilian-casualties/index.html.
 See Conor Friedersdorff, CNN’s Bogus Drone-Deaths Graphic, Atlantic Monthly (July 6, 2012), http://www.theatlantic.com/politics/archive/2012/07/cnns-bogus-drone-deaths-graphic/259493/; see also Conor Friedersdorff, Flawed Analysis of Drone Strike Data is Misleading Americans, Atlantic Monthly (July 18, 2012), http://www.theatlantic.com/politics/archive/2012/07/flawed-analysis-of-drone-strike-data-is-misleading-americans/259836/ (citing to primary research carried out by Sarah Knuckey and Christopher Holland, contributors to this report); Chris Woods, Analysis: CNN Expert’s Civilian Drone Death Numbers Don’t Add Up, The Bureau of Investigative Journalism (July 17, 2012), http://www.thebureauinvestigates.com/2012/07/17/analysis-cnn-experts-civilian-drone-death-numbers-dont-add-up/.
 Peter Bergen & Jennifer Rowland, CIA Drone War in Pakistan in Sharp Decline, CNN (Mar. 28, 2012), http://us.cnn.com/2012/03/27/opinion/bergen-drone-decline/index.html?hpt=op_t1 (claiming website is “up-to-date”); Peter Bergen & Katherine Tiedemann, The Year of the Drone: An Analysis of US Drone Strikes in Pakistan, 2004-2010 (2010), available at http://counterterrorism.newamerica.net/sites/newamerica.net/files/policydocs/bergentiedemann2.pdf; The Year of the Drone, supra note 221.
 2011: The Year of the Drone, New America Foundation, http://counterterrorism.newamerica.net/drones/2011.
 See, e.g., Pratap Chatterjee, The CIA’s Unaccountable Drone War Claims Another Casualty, Guardian (Nov. 11, 2011), http://www.guardian.co.uk/commentisfree/cifamerica/2011/nov/07/cia-unaccountable-drone-war; Orla Guerin, Pakistani Civilian Victims Vent Anger Over US Drones, BBC (Nov. 3, 2011), http://www.bbc.co.uk/news/15553761; Nick Schifrin, Was Teen Killed by CIA Drone a Militant—or Innocent Victim?, ABC News (Dec. 31, 2011), http://abcnews.go.com/Blotter/tariq-khan-killed-cia-drone/story?id=15258659#.T8LWW5lYvuV; UK Drone Strikes Must Stop: UK Lawyer, Dawn (Nov. 8, 2011), http://dawn.com/2011/11/08/us-drone-strikes-must-stop-american-lawyer/. New America Foundation claims that its reporting is based “on accounts from reliable media organizations with deep reporting capabilities in Pakistan, including 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, the Express Tribune, and the News.” The Year of the Drone, supra note 221.
 2011: The Year of the Drone, supra note 242.
 See, e.g., NWA Tribesmen Protest Drone Attack Casualties, News (June 17, 2011), http://www.thenews.com.pk/TodaysPrintDetail.aspx?ID=52979&Cat=7&dt=6/17/2011 (describing the victims, which it identified as four, as a “driver,” an “owner of an auto spare-parts shop,” a “student,” and a man “who was running a medical store”); Tribesmen Protest Drone Attacks, Dawn (June 17, 2011), http://dawn.com/2011/06/17/tribesmen-protest-drone-attacks/ (noting, two days after the strike, that “enraged tribesmen blocked Bannu-Miramshah Road on Thursday [June 16] to protest killing of innocent people”). Note, though, that initial accounts in Western media depicted those killed as militants. See, e.g., Drones Said to Kill 15 Militants in Pakistan, Boston.Com (June 16, 2011); 15 Killed in Two Suspected Drone Strikes, CNN (June 15, 2011), http://edition.cnn.com/2011/WORLD/asiapcf/06/15/pakistan.drone.strike/index.html; Hasbanullah Khan, US Drone Kill Eight Militants in Pakistan, AFP (June 15, 2011), http://www.google.com/hostednews/afp/article/ALeqM5jJGoBfRYzeaAzuAc88gIioBa3Ysg?docId=CNG.921d971040a618e5fd16673c1ea984a7.501&hl=en&lr=all; see also infra Chapter 3: Living Under Drones.
 Woods, Analysis: CNN Expert’s Civilian Drone Death Numbers Don’t Add Up, supra note 240.
 At the time TBIJ published its article, New America Foundation’s total overall civilian casualty figures failed to include the deaths of 69 children killed in a single US drone strike in October 2006, whose names and ages had been published by Pakistani newspapers in the weeks after the attack.
 See Sebastian Abbot, New Light on Drone War’s Death Toll, Associated Press (Feb. 25, 2012).
 Bergen & Rowland, Civilian Casualties Plummet in Drone Strikes, supra note 239.
 The Year of the Drone, supra note 221.
 Abbot, supra note 249.
 2010: The Year of the Drone, New America Foundation, http://counterterrorism.newamerica.net/drones/2010 (last visited Sept. 9, 2012).
 Woods, Analysis: CNN Expert’s Civilian Drone Death Numbers Don’t Add Up, supra note 240.
 Bergen & Rowland, Civilian Casualties Plummet in Drone Strikes, supra note 239.
 See The Year of the Drone, supra note 221; At the time our review was conducted, New America Foundation had reported 27 strikes in 2012, the most recent on July 6, 2012. Of the 107 links cited in support of New America’s data, ten were broken, and 11 corresponded to more than one strike. This left 86 articles from 13 western and Pakistani news agencies to support Bergen’s July 14 statement. It bears noting that TBIJ cites 344 sources for its data on the same 27 strikes. See Obama 2012 Pakistan Strikes, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/2012/01/11/obama-2012-strikes/.
 See also supra notes 156-175 and accompanying text (discussing the demonstrated unreliability of US official reports of all “militant” death tolls). Pakistani intelligence officials, who are often cited as sources for strike information, may be similarly unreliable and prone to overstate “militant” casualties and understate civilian casualties because of the negative public perception in Pakistan that they are complicit in US killings of civilians.
 Bergen & Rowland, Civilian Casualties Plummet in Drone Strikes, supra note 239 (explaining that New America Foundation’s data is drawn from “reputable news sources”).
 According to TBIJ, there were indications of civilian casualties in strikes on February 9, 2012; May 5, 2012; May 24, 2012; June 2, 2012; June 3, 2012; and July 6, 2012. Obama 2012 Pakistan Strikes, supra note 256. TBIJ also reports possible civilian casualties in strikes on July 23, 2012 and July 29, 2012, which took place after Bergen’s article was published. Id.
 See, e.g., Hasbanullah Khan, Five Militants Killed by US Drone in Pakistan, Agence France-Presse (May 24, 2012); Khan & Yusufzai, supra note 180; Twenty Die in Double Drone Attack, Dawn (July 7, 2012), http://dawn.com/2012/07/07/twenty-die-in-double-drone-attack/; US Drone Strike Kills Militant in Pakistan, Officials Say, Jerusalem Post (Feb. 2, 2012), http://www.jpost.com/LandedPages/PrintArticle.aspx?id=257117.
 See The Year of the Drone, supra note 221.
 See Bergen & Rowland, Civilian Casualties Plummet in Drone Strikes, supra note 239.
 Woods, Analysis: CNN Expert’s Civilian Drone Death Numbers Don’t Add Up, supra note 240.
 Bergen & Rowland, Civilian Casualties Plummet in Drone Strikes, supra note 239.
 Friedersdorf, Flawed Analysis of Drone Strikes is Misleading Americans, supra note 240.
 Drone Strike Hits Pakistan Mosque, Say Locals, supra note 180; Khan & Yusufzai, supra note 180; Woods, Analysis: CNN Expert’s Civilian Drone Death Numbers Don’t Add Up, supra note 240. French wire service Agence France-Presse reported the damage to the mosque and said that worshippers there may have been injured. Hasbanullah Khan, US Drone Strike Kills 8 in Pakistan, Agence France-Presse (May 24, 2012).
 Haq Nawaz Khan & Richard Leiby, US Drone Strike in Pakistan Kills 10 Suspected Militants, Wash. Post (May 24, 2012), http://www.washingtonpost.com/world/asia_pacific/us-drone-strike-kills-10-suspected-militants-in-pakistan/2012/05/24/gJQAQbpRmU_story.html; Salman Masood, Drone Strikes Continue in Pakistan as Tension Increases and Senate Panel Cuts Aid, N.Y. Times (May 24, 2012), http://www.nytimes.com/2012/05/25/world/asia/pakistan-says-us-drone-strike-kills-suspected-militants.html?_r=1&ref=world; Haji Mujtaba, US Drone Strike Kills 10 in Northwest Pakistan: Officials, Reuters (May 24, 2012), http://www.reuters.com/article/2012/05/24/us-pakistan-drone-idUSBRE84N03I20120524; Pakistan Says US Drone Kills 10 Militants, USA Today (May 24, 2012), http://www.usatoday.com/news/world/story/2012-05-24/Pakistan-drone/55179756/1?csp=34news. The New York Times went deeper than the other reports, and provides information about the strike from local residents reached by telephone, who stated that some of the strike victims were “Uzbek fighters who belonged to the Islamic Movement of Uzbekistan.” Masood, supra.
 For example, TBIJ’s entry for a recent cluster of strikes that took place on July 29, 2012 was updated two days later to include the names of three local villagers killed in the attack, once those names were reported by The News, a major Pakistani daily newspaper. See Obama 2012 Pakistan Strikes, supra note 256; Three Drone Victims Laid to Rest in FR Bannu, News (July 31, 2012), http://www.thenews.com.pk/Todays-News-2-123753-Three-drone-victims-laid-to-rest-in-FR-Bannu. Over two weeks after the attack took place, New America Foundation still had not reported it, and The Long War Journal had limited its report to include only the subset of missile strikes that hit an alleged Uzbek compound. See Bill Roggio, 6 Uzbeks Killed in North Waziristan Drone Strike, Long War Journal (July 29, 2012), http://www.longwarjournal.org/archives/2012/07/six_uzbeks_killed_in.php; The Year of the Drone, supra note 221.
 Covert Strikes in Pakistan, Yemen, and Somalia—Our Methodology, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/2011/08/10/pakistan-drone-strikes-the-methodology2/ (last updated March 27, 2012).
 See The Year of the Drone, supra note 221.
 See Obama 2012 Pakistan Strikes, supra note 256.
 Bill Roggio, US Drones Kill 10 in Mir Ali Strike, Long War Journal (May 24, 2012), http://www.longwarjournal.org/archives/2012/05/us_drones_kill_10_in_1.php.
 Id.; See, e.g., Obama 2011 Pakistan Strikes, The Bureau of Investigative Journalism, http://www.thebureauinvestigates.com/2011/08/10/obama-2011-strikes/ (last visited Sept. 14, 2012).