A gypsy named Melquiades who died many years ago in Singapore returned to live with the family of Colonel Aureliano Buendia in Macondo, because he could no longer bear the tedium of death. These are the kinds of characters that populate Gabriel Garcia Marquezs magnificent work One Hundred Years of Solitude. Today they also seem to occupy the tribal badlands of Pakistans north-western frontier. On June 3, when Ilyas Kashmiri was killed in a US drone strike, he had already been dead for over a year. In September 2009, the CIA claimed that it killed Kashmiri along with two other senior Taliban leaders in North Waziristan. But the lure of the limelight was seemingly irresistible even in death, because on October 9, Kashmiri returned to give an interview to the late Syed Saleem Shahzad of Asia Times Online. Baitullah Mehsud, the former commander of the Tehreek-e-Taliban Pakistan (TTP), also rose from the dead many times. On at least 16 occasions, Mehsud was in the gun-sights when CIA drones loosed their Hellfire missiles. Yet, until August 2009, he proved unable to settle into the afterlife. Mullah Sangeen also experienced at least two resurrections. Death is clearly not what it used to be Or perhaps the people who were killed in the other attacks were not Kashmiri, Sangeen or Mehsud. Indeed, the attack on a funeral procession on June 23, 2009, which killed Sangeen was supposedly aimed at the TTP chief. It killed 83 people who certainly were not who they were supposed to be. These are not isolated events. At the end of 2009, the Pakistani daily Dawn calculated that, of the 708 people killed in 44 drone attacks that year, only 5 were known militants. Earlier that year, The News, Pakistans other major English-language daily, had calculated that between January 14, 2006, and April 8, 2009, 60 drone attacks killed 701 people - of whom only 14 were known militants. The US has come a long way since July 2001 when it rebuked the Israeli government for its policy of targeted assassination, which it said were really extrajudicial killings. In September of that year, CIA director George Tenet confessed that it would be a terrible mistake for someone in his position to fire a weapon such as the predator drone. By 2009, such qualms were obsolete. Indeed, the new CIA director Leon Panetta declared predator drones the only game in town. The catalyst was 9/11 - and lifting the ban on extrajudicial killings was just one of the many illegal policies it licensed. Many of the post-9/11 criminalities were eventually rolled back, yet the policy of extrajudicial killings not only survived the Bush years, it was intensified. During his eight years in office, Bush ordered a total of 45 drone strikes in Pakistan; in fewer than three years, Obama has ordered more than 200. On his third day in office the president ordered two drone strikes, one of which incinerated a pro-government tribal leader along with his whole family, including three children. Obama has since also expanded the drone war in Afghanistan. The politics of body counts The new tactic has many sceptics, and not all of them are antiwar activists. Criticism has also been voiced from within the CIA and the military. Yet drones have been embraced with remarkable warmth by Obama and the US intelligentsia. This partly has to do with an existing US tendency to see technology as a panacea for all problems, including military ones. But the tactic is also made palatable by a routine exaggeration of its accuracy and a downplaying of its human cost. The statistics have been often quoted in the Western media though all they show is the boundless credulity of LWJ proprietors. Relying solely on media reports - which in turn rely almost exclusively on unnamed Pakistani and US officials - the website claims that a mere seven percent of the 1,954 people killed in Pakistan so far have been civilians. It claims - for example - that, of the 73 people killed in 2007, none were civilians, even though it couldnt name a single individual killed. The more widely cited New America Foundation (NAF) study fares only slightly better. Employing a seemingly rigorous method, the project records every drone attack along with its intended target and presumed outcome. Of the 1,542 to 2,541 people killed in Pakistan by drones since 2004, it claims between 1,249 and 1,960 were militants. Like the LWJ, the NAF also relies on media reports and errs conspicuously on the side of official claims. For example, its data shows that, of the 287 Pakistanis killed so far this year, 251 were militants. This of course cannot be true, since a single incident - the March 17 killing of 38 pro-government tribal elders at a gathering in Datta Khel, North Waziristan - undermines these calculations. The slaughter even managed to provoke a rare outburst from the Pakistani military chief General Ashfaq Kiyani, a tacit supporter of the drone war. These civilian deaths were only acknowledged because the victims were known notables with favourable relations with the Pakistani government - otherwise, as Wazir Malik Gulabat Khan has pointed out, the government never investigates how many of those killed are actual militants. But beyond the reliance on official sources, there is also a fundamental question of honesty. Take two of the most tragic incidents of the drone war. On January 13, 2006, a drone struck the village of Damadola in Bajaur, killing 18 villagers, mainly women and children. US and Pakistani officials initially claimed that four al-Qaeda terrorists were among the dead, a claim which they later retracted. Yet if you visit the NAF database, you will find that it lists all 18 as militants - and none as civilians. On October 30, another drone strike hit Chenagai, also in Bajaur, killing 82 children at a seminary. But if you visit the NAF database, you find that it lists up to 80 militants killed - and again no civilians. The editors, however, note that they have excluded these figures altogether from their list of fatalities. These two incidents alone should void the NAF studys credibility, but there are other reasons why its figures should be taken with a grain of salt. In its annual report on the CIA assassination program, the Islamabad-based Conflict Monitoring Centre highlights several. Besides the tendency to exaggerate success and downplay failure in order to avoid adverse public reaction, neither the US nor the Pakistan government has a mechanism in place to verify the identity of those killed. There is also a concern that the drones are no longer targeting only high value suspects; under expanded authority granted by Bush and continued under Obama the agency can target all suspected militants based on pattern of life analysis collected from surveillance cameras. In the tribal areas, where traditionally most adult males carry guns and ammunition, this makes everybody a potential target. A year before Osama bin Laden was killed, a CIA officer told Jane Mayer of the New Yorker, that because of the drone surveillance, no tall man with a beard is safe anywhere in Southwest Asia. But human intelligence is no less defective, since in Pakistan as in Afghanistan, informants have often settled scores with rival tribes by denouncing them as Taliban' None of this, however, has deterred the NAF projects Peter Bergen from making confident claims about the presumed success of the strategy. He now claims that only six per cent of those killed were civilians, even though he can only name 35 high value targets among the more than two thousand killed. A rare case-by-case analysis of nine attacks by the Campaign for Innocent Victims in Conflict (CIVIC), however, uncovered 30 civilian deaths, including 14 women and children, unreported in the media. Testimonies of survivors collected by Voices for Creative Non-Violence (VCNV) paint an even bleaker picture. My own conversations around Peshawar with FATA residents and Frontier Constabulary (FC) men revealed that the drones are sometimes successful in reaching their targets - but the human cost is invariably steep. There has been no accounting of the psychological costs of the war. Because of the secrecy around the program, there is no way to confirm if there are any safeguards in place to avoid civilian casualties; or, if there are, how well they are being enforced. As a consequence, there is no oversight, accountability or redress. The drone war in Pakistan is, in this respect, very different to the drone war in Afghanistan. The latter is under the command of the military and is therefore subject to the minimal constraints of military rules of engagement. The CIA however has none, so is entirely unaccountable. The possibility of oversight is further diminished by the fact that the CIA employs private contractors (read mercenaries) who operate in an even murkier legal terrain. With no democratic checks or institutional barriers, the drones are, in effect, operating in a heart of darkness. This was brought home last year when the CIA went on a rampage after one of its bases in Khost was blown up by a Jordanian militant. The politics of expertise The pro-war propaganda is not always successful in maintaining its veneer of sophistication. Last May, during an exchange on MSNBC between Colonel Tony Shaffer, a Defence Intelligence Agency veteran advocating boots on the ground, and Christine Fair, an eccentric US academic much in favour in national security circles for her ultra-hawkish views, it dropped altogether. When Shaffer suggested that civilian casualties resulting from the drone attacks were increasing anti-Americanism in Pakistan, Fair took extreme exception and retorted categorically that the drones are not killing innocent civilians. She dismissed Pakistani press reports as deeply unreliable and dubious and claimed that a number of surveys on the ground in FATA had shown that residents generally welcome the drone strikes. As a matter of fact, the only known survey on the ground in FATA at the time was carried out by a letterhead organisation named the Aryana Institute for Regional Research and Advocacy whose conclusions can fairly be described as deeply unreliable and dubious. It claimed that 55 per cent of respondents in a survey it carried out in parts of FATA that are often hit by American drones (among which it curiously included Parachinar, which has never been hit and whose overwhelmingly Shia population is deeply hostile to the virulently anti-Shia Taliban) did not think that the attacks caused fear and terror in the common people; 52 per cent found them accurate in their strikes; and 58 per cent did not think they increased anti-Americanism. The survey got much play in the media, quoted among others in both the New York Times and the Los Angeles Times. Its conclusions were found particularly agreeable by proponents of drone escalation and the label of an institute gave them an ostensibly academic pedigree. Few wondered why the surveys claims were so at odds with known public opinion in the wider region where, according to a Gallup/Al Jazeera poll conducted around the same period, only nine per cent of people showed support for the drone attacks. Those who did wonder, such as the journalists I spoke to in Peshawar, were universally dismissive. But the Institute had served its purpose and, typical of many LHOs, it vanished after a year (Web Archive shows that its website only existed between 2008-2009). Ironically, Aryanas claims were discredited just a year later by a survey in FATA by an institution no less enthusiastic about the drones. Muhammad Idrees Ahmad is a Glasgow-based sociologist, born in Chitral and raised in Abbottabad and Peshawar. He is the co-editor of Pulsemedia.org. Aljazeera