Dead Afghan Civilians: Disrobing the Non-Counters

by Marc W. Herold1
Departments of Economics and Women's Studies
Whittemore School of Business & Economics
University of New Hampshire

POSTED AUGUST 20, 2002 --

When the U.S. bombing started at 9 p.m. on October 7, 2001, an official 'counting of the dead' in Afghanistan was deemed largely unnecessary. The public was assured that U.S.- U.K. military planners would go to great lengths to avoid civilian casualties and the use of newer, precision-guided munitions would allow only killing the 'bad guys.' The Pentagon has steadfastly refused to deal with 'body counting', fearing an epiphany of the ghost of Vietnam.

And the tone of the U.S. mainstream reporting was revealed at the start of the third week of bombing when William Arkin, a frequent contributor on military issues to major U.S. dailies and author of a study purporting to count civilian deaths in the NATO campaign upon Yugoslavia in 1999, penned a piece on October 21 reassuring the American public that the bombing in Afghanistan will result in low civilian casualties because of precision targeting and "much of the bombing [is] taking place in remote areas."2

But facts in Afghanistan on that day and others in October proved him wrong. On October 21, between 60-80 innocent Afghan civilians were killed by U.S. bombs dropped in five provinces during six bombing incidents.3 The dead civilians included persons fleeing on a tractor in Torai, Uruzgan [21-32 dead], families having breakfast in the Parod Gajaded neighborhood of Kabul [9-18 dead] when two 'smart bombs' dropped by a Navy F/A-18 strayed, a seven-year-old girl who died in the Macroyan housing project in Kabul4 and 25 persons who succumbed in the bombing of a one-time military hospital and mosque in Herat. The Afghan Islamic Press also cited three dead and eight injured in Kandahar, and the Italian NGO Emergency mentioned one dead in Shakar Dara village in Parwan. The U.S. press was silent, but the British press provided details:

"Not long after 7 p.m. on Sunday, Oct. 21, the bombs began to fall over the outskirts of Torai village. Maroof saw a massive fireball rising from the ground." He realized that "bombs had fallen over the little cluster of houses a mile away where his sister and his other relatives were living."

So wrote the Times of London, describing the destruction of an entire family.

"The roll call of the dead read like an invitation list to a family wedding: his mother-in-law, two sisters-in-law, three brothers-in-law, and four of his sister's five young children, two girls and two boys, all under the age of eight."5

Lamentably, most mainstream U.S. reporters have been satisfied to listen to the daily press briefings at the Bagram air base from various U.S colonels - whether 'Buster' Hagenbeck or Robert King - and dutifully transcribe the 'truth' about Afghan civilian casualties.

Recording the 'Truth' at Bagram

Recording the 'Truth' at Bagram [photo by Regan Morris, AP]

To date [August 20, 2002], 10 studies have been undertaken seeking to count the dead civilians in Afghanistan caused by U.S. bombing. Eight have been publicly released and two are in-progress. Five present a picture of the total universe of bombing incidents, whereas five are based upon sample populations.

Table 1. The Map of Studies Counting Afghan Civilian Casualties

Author Affiliation Date released Methodology Published raw data? Deaths
Marc Herold Faculty at University of New Hampshire Dec. 10, 2001 and updated continually Relied upon media and NGO reports Yes 3,100-3,600
Francoise Chipaux Reporter at Le Monde Dec. 13, 2001 ? No Over 1,000 during first 2 mos.
Reuters Reuters news wire service Jan. 3, 2002 Sample of 14 incidents No 982
Carl Conetta Project on Defense Alternatives Jan. 18, 2002 Relied upon only Western news reports No 1,000 - 1,300
Laura King et. el. Associated Press writers Feb. 11, 2002 Press reports, interviews No 500 - 600
John Donnelly and Anthony Shadid Reporters for Boston Globe Feb. 17, 2002 Sample of 14 incidents Yes 830
David Zucchino Reporter for Los Angeles Times June 2, 2002 Universe of 194 incidents No 1,067-1,200
Marla Ruzicka Staff member of Global Exchange July 20, 2002 -- in-progress Sample of 11 bombing 'sites'interviews No 812
Dexter Filkins Reporter for New York Times July 21, 2002 Sample of 11 incidents Yes 396
Marc Herold Faculty at University of New Hampshire in-progress 40 in-depth case studies Yes 870-1,000

Note: sources and further details available upon request from the author.

On December 10, I released my dossier presenting an analysis and disaggregated data stating that over 3,500 Afghan civilians had perished under U.S. missiles and bombs.6 The study relied upon wire service reports, newspaper stories filed worldwide [from Australian to Indian, Pakistani to British], and some reports by NGOs. A weakness of the study - since corrected - was some double-counting due to confused site names. The data base has been continually updated, corrected and now incorporates civilian deaths resulting from British and U.S. special forces attacks.

A couple days later, the French daily Le Monde, published a brief article based upon its own research, which bluntly stated that in two months the air war had cost at least 1,000 civilian lives.7 Some specific incidents were mentioned but little more. In early January, Canada's premier daily referred to a Reuters accounting based upon 14 bombing attacks which had caused 982 civilian deaths [or an average of 70 per attack]8.

In a memo written in January 2002 after a one month stay in Afghanistan, Gene Stolzfus, director of Christian Peacemaker Teams, wrote about having been told "by one organization with monitoring capability that an estimated 5,000 homes valued at $2,000 each were destroyed and an estimated 2,500 people were killed in four of the 30 provinces near Kabul [by U.S. bombing] where careful monitoring has been done."9 The author has confirmed to me personally that this organization felt certain about the estimate for these four provinces [Kabul, Nangarhar, Logar and Paktia].

A second major study was released by the Project on Defense Alternatives [Cambridge, Mass.] in mid-January which stated that U.S. bombing in Afghanistan killed civilians at a rate four times higher than the NATO bombing of Yugoslavia in 1999.10 Between 1,000 - 1,300 civilians were killed through January 1, 2002, according to the report. It pointed out that the bombing campaign "failed to set a new standard for accuracy" because of the mix of weapons used [away from laser-guided to cheaper GPS-guided less precise munitions], the unreliable nature of intelligence, and the decision to bomb al-Qaeda and Taliban leaders in their houses where little margin of bombing error existed.

The study compiled statistics from only Western sources. Whereas this made it more palatable to the mainstream Western media, it also necessarily involved a smaller universe than my own dossier, and it indirectly further strengthened a pernicious view in the United States that truth can only come through a Euro-American lens.11

By early February, stirrings were being reported in the mainstream U.S. establishment. An article appeared in the Wall Street Journal announcing that ,em>Human Rights Watch would soon be sending a team of three researchers to Afghanistan - headed by William M. Arkin - to come up with the correct counting of the Afghan dead.12 Human Rights Watch officials 'said privately' they estimated the civilian death toll between 100 - 350 in early December, figures utterly consistent with their established record of severe under-counting.

On February 11, the Associated Press released its counter-study, boldly proclaiming and reassuring an increasingly alarmed public, "Hundreds Lost, Not Thousands."13 The AP report reached its figure of 500-600 "by examining hospital records, visiting bomb sites and interviewing eyewitnesses and officials." Many methodological problems diminish the report:

For reasons explored elsewhere, I have the same doubts about studies of bombing incidents carried out long after the event, as the Pentagon's Rear Admiral Quigley, who said "If you can't get there real quick, there's not much you are going to learn that you have any confidence in."15

The AP's Laura King then put forth the standard line that the Taliban routinely inflated civilian death figures, and that some news reporters and researchers had uncritically accepted these numbers. As I have argued elsewhere, a more rigorous analysis which explores the incentives to over- or under-report by the authorities of a bombed nation might well reveal a powerful incentive to under-estimate such civilian casualties, as I argue was the case with the Taliban.16

A fortnight after the less-than-satisfactory AP report, John Donnelly and Anthony Shadid of the Boston Globe published their excellent survey of 14 sites bombed by U.S. warplanes, which resulted in 830 civilian deaths [or about 60 per incident]. The authors divulge details on each site based upon accounts of locals, visual evidence and tallies by non-governmental organizations. They note that "Because the 14 sites represent only a small fraction of the total sites targeted...since October, the total is estimated at 1,000 or more."17 The prime culprits for civilian deaths are: faulty intelligence; imprecision of aerial warfare; and "a large number of deaths can be attributed to the selection of targets in civilian areas."

A four month hiatus then ensued in the counting of the dead, broken in early June by David Zucchino of the Los Angeles Times.18 The Times study reviewed newspaper and news wire reports, identifying 194 incidents of civilian casualties between October 7, 2001 and February 28, 2002. The reported death toll was between 1,067 and 1,201, excluding 754 deaths reported by the Taliban as well as 497 deaths not identified as civilian or military. Unfortunately, again, the blame here was put upon faulty intelligence rather than bombing inaccuracy and, upon the Taliban's supposed inflation of casualty figures.

On Saturday, July 20, a staff member of Global Exchange in Kabul released a preliminary report counting civilian casualties caused by U.S. bombing of Afghanistan. The study of "eleven sites" purported to document "At least 812 deaths in the first three months of the U.S. bombing campaign."19 The study did comprehensive surveys in four provinces [Kabul, Kandahar, Mazar and Kunduz - note: Mazar is not a province] and partial surveys in four others [Khost, Ghazni, Paktika and Herat]. The data was compiled by Afghan interviewers though no information is provided on the number of interviewers, their qualifications, or, most importantly, which bombed places were visited. No disaggregated data is produced and not even the number of bombing incidents is divulged.

The 812 civilians killed figure fits poorly with sample studies reported by Reuters, Donnelly, and Filkins. The accompanying commentary attributed to the Global Exchange staffer raises serious doubts: She alleges the deaths were 'errors' and that they were caused by 'faulty intelligence' provided by Afghans. U.S. bombing based upon so-called intelligence provided by Afghans only began around December 20, 2001, and by that date, well over 2,000 Afghan civilians had already perished under U.S. bombs. These people died not because of 'errors' but rather, as I have persistently stated, primarily because U.S military planners dropped big bombs in civilian-rich areas.

Some references to the Global Exchange study have implied that studies carried out on-site through interviews will necessarily and unquestionably be more accurate than other studies based upon other sources of information, for example, that of reporters who were at or spoke with people who had been at the site shortly after the bombing. Research on civilian casualties who died many months ago is flawed for obvious and less obvious reasons. Obvious reasons are that people have moved away from bombed-out neighborhoods or destroyed villages and that remaining ones fear retribution in speaking out against the U.S. and Karzai regime. A less obvious reason has to do with patterns of memory after traumatic events of which the bombing of one's family certainly qualifies. The psychological literature suggests, as I read it, that people's memory of surrounding events becomes cloudy with a focus upon immediate kin - I call this centripetal memory loss. The implication is that persons interviewed later will clearly remember what happened in their own homes, but not in their neighborhoods. This leads to a gross underestimate in the counting of casualties.

A day later the New York Times' Dexter Filkins reported on 11 bombing sites where 396 Afghan civilians had perished [or about 36 per attack].20 Filkins attributed the use of overwhelming force by the U.S. as causing many of the casualties. My own database reveals that in the same 11 incidents, 408-509 civilians had perished. The Filkins study drew an immediate sharp rebuke on the following Monday from U.S. Defense Secretary Rumsfeld.21

During late July - early August, three mentions of Afghan civilian casualties have been made in the mainstream press. These are interesting less for their numerical accuracy and more for their political content. The Bush Administration has persistently proclaimed the accuracy of U.S. bombing, and acknowledged the few though 'regrettable' civilian casualties.

Disregarding widespread and increasing popular revulsion at home, the Karzai regime has also emphasised these low numbers. Exactly three weeks after the deadly U.S. bombing attack of a wedding party in Uruzgan which was roundly criticized for relying upon flawed intelligence and exhibiting a 'shoot first and ask questions later' U.S. military mentality, a Karzai spokesperson in Kabul told the BBC that fewer than 500 civilians were believed to have been killed in U.S. airstrikes during the entire Afghan air campaign.22

A couple of days later, Gary Thomas of the U.S. Government's propaganda outlet, Voice of America News, mentioned that "Published reports suggest that as many as 400 civilians may have been killed by errant U.S. bombs."23 And on the National Public Radio show 'The Connection' on an August 2 segment devoted to the topic of civilian casualties, Sarah Sewall, program director of the Carr Center for Human Rights Policy [whose director is Michael Ignatieff, primus interpares of the interventionist New Humanitarians24 ], quickly agreed with her two fellow guests - retired military officers - that Afghan civilian casualties were in the low hundreds, and then went on to chatting about which weapons systems can best be used to attain U.S. military goals in the area.25

Political counting of the dead is revealed at its very best: Data, facts and research are irrelevant.

Lastly, I have completed an in-depth study of 40 bombing cases occurring between October 7, 2001 and July 1, 2002, carefully chosen to reflect the diversity of geography and daily living prevalent in Afghanistan. I let the voices of those being bombed describe what happened and how it felt. The research, titled "Mosaics of Pain in 40 Pieces," estimates that 870-1,000 civilians died in these 40 bombing attacks [or about 21-25 per attack]. The five studies relying upon sample data show rather high numbers of Afghan civilians killed per bombing attack - the range being a low of 21-25 [Herold] to a high of 70-73 [Ruzicka and Reuters].

Those who generate low overall numbers of civilian casualties stress the faulty intelligence provided by Afghans [they absolve themselves of responsibility], point to an alleged proclivity of the Taliban to inflate such figures, and uncritically accept that the new precision-guided munitions kill mostly the 'bad guys.'

Others like myself point to a very simple, powerful single explanation: The thousands of Afghan civilians who perished under U.S bombs did so because U.S. military and political elites chose to carry out a bombing campaign using extremely powerful weaponry having high margins of error and with huge killing and blast radiuses in largely civilian-rich areas [of course, the isolated training camps being an exception but these were largely destroyed during the first week].

As I have argued elsewhere, the U.S. war upon Afghanistan is the exact opposite of that carried out a decade earlier in Iraq: Afghanistan has been an air war characterized by a low bombing-intensity and high civilian casualty-intensity while Iraq was the reverse.26 The large number of dead civilians I report is the accumulation of many small numbers of people killed regularly in the daily bombing of Afghanistan.

As just one example, a F/A-18 Navy jet was reported to be hitting three to four targets per sortie. To speak about precision-bombing outside of locational context is about as foolish as arguing that the transfer of First World technology to the Third World will necessarily soon result in all the nations there becoming Japans.

I believe that we must care about the level of civilian casualties for many reasons. To diminish or to overlook them gives free play to the enthusiasts of precision-guided weaponry, who may then reassure the public that wars can now be carried out in such ways that only those 'deserving to die' will do so. This is an invitation to proliferation of war. Secondly, the high level of civilian casualties exposes the arrogance of U.S. and U.K. military planners who believed they could drop bombs into urban neighborhoods or mountain villages and actually kill only al-Qaeda or Taliban cadres.

Thirdly, many of the individual instances of bombing attacks - including those upon civilian infrastructure [civilian cars and trucks, clinics, radio stations, hydropower stations, bridges] and during November-December of anything rolling on the roads of southern Afghanistan - can [and should] be seen as violating the humanitarian rules of war and the charge of war crimes against the U.S. seems appropriate.

Lastly, the families of those wrongfully killed are deserving of compensation, just as the Chinese in the Belgrade embassy, the Italians on the ski lift, and most recently, the families of two South Korean teenagers, all killed by U.S. negligence, have been indemnified.27

But everything suggests that should such compensation ever be offered it will be a pittance compared to what the Italians, Chinese and South Koreans received. Just as in 1919, when the British general staff was considering using poison gas upon the unruly Pathans, the marginal cost today of killing an Afghan or Iraqi is zero to the Western military and political elite so long as the human carnage on the Afghan soil can be hidden from the Western public.28 Afghan bodies are largely unworthy of being seen or recorded.

The U.S. air campaign and the subsequent Special Forces attacks must be 'marketed' to the American public. The U.S air war is a military 'Mister Clean.' The carnage must be cleansed and sanitized of detached limbs, trunks of torsos, disfigured and burnt bodies. Every effort has been made to do so - including ignoring civilian casualties - and when Al-Jazeera presents images of gore and blood, the managers of the U.S. propaganda establishment - whether corporate media, human rights groups, or academics - rise in a chorus of righteous condemnation. A simple question might be asked: How many photos of Afghans dismembered by U.S. bombs or missiles have you seen in the corporate media?

The non-counters are disrobed.

-- 30 --


1. The author is a Professor of Economic Development & Women's Studies, University of New Hampshire, U.S.A.. He has written extensively on Third World socio-economic development and has put together large-scale, empirical data bases on the foreign subsidiaries of U.S. multinational corporations [1856-1970] and on the formation of modern business in Brazil [1840-1960].

2. William M. Arkin, "Civilian Casualties and the Air War," Washington Post [October 21, 2001].

3. Data from Marc W. Herold, "Appendix 4: Daily Casualty Count of Afghan Civilians Killed in U.S. Bombing and Special Forces Attacks, October 7 until present day" available at

4. "US Planes Hitting Residential Areas: UN," Dawn [October 24, 2001].

5. Catherine Philip. "War Descends Like a Scythe on Innocent Afghan Village," The Times [October 26, 2001], reprinted at

6. See "A Dossier on Civilian Victims of United States' Aerial Bombing in Afghanistan: A Comprehensive Accounting" at

7. Francoise Chipaux, "Afghanistan: Deux Mois de Frappes Auraient Fait au Moins un Millier de Vies Civiles," Le Monde [13 decembre 2001].

8. In Murray Campbell, "Thousands of Afghans Likely Killed in Bombings," The Globe and Mail [January 3, 2002], p. A1.

9. Gene Stolzfus, "House Rebuilding in Afghanistan," Christian Peacemaker Teams memo [dated January 28, 2002] at

10. Carl Conetta, Operation Enduring Freedom: Why a Higher Rate of Civilian Bombing Casualties [Cambridge, Mass.: Briefing Report #11. January 18., 2002]

11. This topic is explored at length in my forthcoming "Truth About Afghan Civilian Casulaties Comes Only Through American Lenses for the U.S. Corporate Media [our modern-day Didymus]," in Peter Phillips [ed] and Project Censored, Censored 2003: The Year's Top 25 Stories [New York: Seven Seas Publishing, forthcoming 2002].

12. Chip Cummins, "Human-Rights Group to Estimate Civilians Killed in U.S. Campaign," Wall Street Journal [February 7, 2002]: A18.

13. Laura King, "Review: Afghan Civilian Deaths Lower," Associated Press [February 11, 2002 at 2:28 PM ET] and Deb Riechmann, "AP: Hundreds Lost, Not Thousands," Associated Press [February 16, 2002].

14. "Afghans Defend US Air Strikes," BBC News Online [July 21, 2002]

15. Cummins, op. cit.

16. Herold, "Truth About Afghan Civilian," op. cit.

17. John Donnelly and Anthony Shadid, "Civilian Toll in US Raids Put at 1,000. Bombing Flaws, Manhunt Cited," Boston Globe [February 17, 2002]: A1.

18. David Zucchino, "The Untold War. 'The Americans...They Just Drop Their Bombs and Leave'," Los Angeles Times [June 2, 2002]: A1.

19. John Yang, "An Early Estimate. Rights Group Offers Estimate of Afghan Civilian Casualties in U.S. Attacks," [July 22, 2002].

20. Dexter Filkins, "Flaws in U.S. Air War Left Hundreds of Civilians Dead," New York Times [July 21, 2002].

21. Tom Shanker, "Rumsfeld Calls Civilian Deaths Relatively Low," New York Times [July 22, 2002].

22. "Afghans Defend US Air Strikes," BBC News Online [July 21, 2002]

23. Gary Thomas, "Afghan Official Calls for Investigation of Civilian Casualties," VOA News [July 25, 2002].

24. For example see his "The Case for a Committed American Imperialism," New York Times Magazine [July 28, 2002], which is in turn, appropriately criticized in Gilles d'Aymery's column in Swans Commentary [July 29, 2002] at

25. "Civilian Casualties," National Public Radio The Connection [August 2, 2002].

26. See "Afghanistan War Produces High Civilians Killed-Per-Bomb-Dropped Ratio" at

27. See my "The 'Value' of a Dead Afghan: Revealed and Relative" at

28. See Edward M. Spiers, "Gas and the North-West Frontier," Journal of Strategic Studies 6,4 [1983]: 94-112. I am grateful to Edward Herman for the insight regarding marginal cost.


Professor Marc Herold's Afghan Canon