vendredi 5 octobre 2007

Iraq Body Count: "A Very Misleading Exercise"

By Media Lens
05 October,2007

Media Lens


The mainstream media are continuing to use figures provided by the website Iraq Body Count (IBC) to sell the public a number for total post-invasion deaths of Iraqis that is perhaps 5-10% of the true death toll.


As we recently reported, only a handful of media outlets covered a new ORB poll revealing that 1.2 million Iraqis had been murdered since the 2003 invasion. BBC Online provided a rare example:
“A UK-based polling agency, Opinion Research Business (ORB), said it had extrapolated the figure by asking a random sample of 1,461 Iraqi adults how many people living in their household had died as a result of the violence rather than from natural causes.

“The results lend weight to a 2006 survey of Iraqi households published by the Lancet, which suggested that about 655,000 Iraqi deaths were ‘a consequence of the war’.

“However, these estimates are both far higher than the running total of reported civilian deaths maintained by the campaign group Iraq Body Count which puts the figure at between 71,000 and 78,000.” (BBC Online, ‘US contractors in Iraq shootout,’ September 17, 2007)
BBC’s Newsnight programme used IBC’s figures in the same way:
“More than a million Iraqis have been killed since the invasion in 2003, according to the British polling company ORB. The study’s likely to fuel controversy over the true, human cost of the war. It’s significantly up on the previous highest estimate of 650,000 deaths published by the Lancet last October… The independent Iraqi [sic] Body Count group puts the current total at closer to 75,000.” (Newsnight, BBC2, September 14, 2007)

These reports again raise serious issues about what IBC’s figures actually mean, how they are being used and misused to cast doubt on higher numbers, and about what IBC is doing to promote or reduce the confusion. (See our 2006 Media Alerts archive for previous analysis, beginning here.)
Just “Care And Literacy” — No Extrapolations Required

In its latest press release, ‘The State of Knowledge on Civilian Casualties in Iraq,’ IBC explains ‘What IBC Does’:
“Provides an irrefutable baseline figure”
Similarly in 2006, IBC wrote: “We are providing a conservative cautious minimum.”

These both describe laudable objectives involving little more than accurate data collection. IBC co-founder John Sloboda made the point in a BBC interview in response to criticism that he and his colleagues were “amateurs” in the field of mortality studies:
“Our position is, and always has been, that reading press reports, which is what this job is, requires nothing other than care and literacy. The whole point about it is that it doesn’t require statistical analysis or extrapolations.”
And yet in their latest press release (September 3, 2007), under the title, ’How plausible is 600,000 violent Iraqi deaths?’, IBC devote five pages to wide-ranging criticism of the 2006 Lancet study which estimated 655,000 excess deaths in Iraq.
IBC’s conclusion:
“Our own view is that the current death toll +could+ be around twice the numbers recorded by IBC and the various official sources in Iraq. We do not think it could possibly be 10 times higher.”
In similar vein, the Toronto Star quoted IBC co-founder John Sloboda as saying:
“The death toll could be twice our number, but it could not possibly be 10 times higher.” (Haroon Siddiqui, ‘How many civilians have died?’ Toronto Star, September 20, 2007)

This last comment was reported less than a week after the publication of ORB’s poll revealing 1.2 million Iraqi deaths.

Two questions arise: Why is it important for IBC — providing an “irrefutable baseline” based on data collection — to challenge the methodology and conclusions of epidemiological studies published in the Lancet which go far beyond data collection and which do not in any way challenge their baseline as a “cautious minimum”?

Secondly, while IBC’s self-described task does indeed require only “care and literacy”, does not the task of challenging peer-reviewed science published by some of the world’s leading epidemiologists require very much more? Does it not, in fact “require statistical analysis or extrapolations,” and much else besides?

In a 2006 addition to their website, IBC wowed visitors with scientific jargon:
“Our data is very rich, because it provides a large subset of what is happening.
“It has high spatiotemporal specificity. Post-event interviews are always hampered by the fact that people tend to move on, and may not remain in the area or even in the country. Our data is recorded as close to the time and place of death as possible, and so has ‘forensic’ elements.”

It seems that IBC have used their credibility as data collectors to ‘cross sell’ their credibility as commentators on peer-reviewed epidemiology to the media community. But this second task is unrelated to their task as data collectors, and is an area in which, to our knowledge, none of the co-authors of their press releases have any research record or publication history in any relevant scientific discipline.

In a 2006 BBC interview, John Sloboda said of the 2004 Lancet study:
“Some critics of the Lancet study have said it’s like a drunk throwing a dart at a dartboard. It’s going to go somewhere, but who knows if that number is the bulls eye.

“Unfortunately many many people have decided to accept that that 98,000 figure is the truth - or the best approximation to the truth that we have.”
Sloboda was here endorsing a claim based on a failure to comprehend even the basic meaning of the Lancet study’s range of figures — the “drunk throwing a dart at a dartboard” analogy was and is absurd. No qualified epidemiologist would countenance making such a comment.

Unsurprisingly, most journalists reporting on international affairs appear unable to distinguish between the task of “reading press reports” on the one hand, and engaging in “statistical analysis or extrapolations”, on the other.

Reporters naturally assume that, given its “high spatiotemporal specificity”, IBC’s credibility is on a par with the world’s leading experts in the field published in the world’s leading scientific journals and subject to an exacting system of peer review.

Certainly IBC do nothing to discourage, and everything to encourage, such a view. Wouldn’t it be more reasonable for IBC to point out in commenting on the Lancet studies to highly influential media that they are in fact +not+ especially qualified to comment on the science of epidemiology?

The Problem Of Relying On The Journalistic Record
IBC also move far beyond data collection in this latest addition to the site:
“Those who suggest that the IBC data-base is likely to contain only a tiny minority of actual deaths generally argue three things. First, they say that IBC only records deaths in areas where Western journalists are present; second they propose that there have been at least seven credible studies which suggest up to ten times as many deaths as we have recorded; and third they assert that an alternate media world exists containing a professional Arab-language press which continually reports far more deaths than the sources we monitor in English.

“We have dealt with the first two claims in detail on the public record and will be happy to answer questions about them in the discussion. IBC in Context (Feb 2006)”
IBC omit to mention the most obvious and telling criticism: that the credibility of their database as an approximate guide to levels of violence in Iraq — i.e., “The death toll could be twice our number, but it could not possibly be 10 times higher” — is undermined by the fact that conditions in Iraq are so lethal that journalists are unable to discover many violent deaths of civilians.

Consider that a study of deaths in Guatemala from 1960 to 1996 by Patrick Ball et al at the University of California, Berkeley (1999) found that numbers of murders reported by the media in fact decreased as violence increased. Ball described the “problem of relying on the journalistic record” in evaluating numbers killed:

“When the level of violence increased dramatically in the late 1970s and early 1980s, numbers of reported violations in the press stayed very low. In 1981, one of the worst years of state violence, the numbers fall towards zero. The press reported almost none of the rural violence.” (Patrick Ball, Paul Kobrak, and Herbert F. Spirer, ‘State Violence in Guatemala, 1960-1996: A Quantitative Reflection’, 1999)

Ball added:
“Throughout the 1980 to 1983 period newspapers documented only a fraction of the killings and disappearances committed by the State. The maximum monthly value on the graph [see link above] is only 60 for a period when monthly extra-judicial murders regularly totaled in the thousands.”
Ball explained that “the press stopped reporting the violence beginning in September 1980. Perhaps not coincidentally, the database lists seven murders of journalists in July and August of that year”.

The significance is indicated in a Reporters Sans Frontieres (RSF) report (September 7, 2007), which described how the number of journalists and media workers killed in Iraq since the start of the 2003 invasion had reached 200. According to RSF, 73 per cent of journalists killed had been directly targeted, a figure which was “much higher than in previous wars”. RSF also reported that more journalists had been taken hostage in Iraq than anywhere else in the world. A total of 84 journalists and media workers had been kidnapped in the previous four years.

Lancet study co-authors Les Roberts and Gilbert Burnham wrote recently:
“A study of 13 war affected countries presented at a recent Harvard conference found over 80% of violent deaths in conflicts go unreported by the press and governments.” (Roberts and Burnham, ‘Ignorance of Iraqi death toll no longer an option‘)

We contacted the author of the study, Ziad Obermeyer, for details. Demonstrating a level of scientific caution that is absent from some of IBC’s bold pronouncements, Obermeyer responded that because his manuscript was progressing through the peer review process he could not provide anything for “formal citation”. He added:
“It is safe to say, however, that our estimates of violent war deaths, based on nationally representative surveys, are substantially higher than those commonly cited for most of the 13 countries we study.” (Email to Media Lens, September 24, 2007)

Roberts and Burnham continued:
“City officials in the Iraqi city of Najaf were recently quoted on Middle East Online stating that 40,000 unidentified bodies have been buried in that city since the start of the conflict. When speaking to the Rotarians in a speech covered on C-SPAN on September 5th, H.E. Samir Sumaida’ie, the Iraqi Ambassador to the US, stated that there were 500,000 new widows in Iraq. The Baker-Hamilton Commission similarly found that the Pentagon under-counted violent incidents by a factor of 10.” (Roberts and Burnham op. cit)

IBC’s methodology was devised by Marc Herold, a professor of economics at the University of New Hampshire. Herold has tracked deaths in Afghanistan since the US-led invasion of 2001. It was Herold’s Afghan Victim Memorial Project that inspired John Sloboda to set up IBC. Herold’s “most conservative estimate” of Afghan civilian deaths resulting from American/NATO operations is between 5,700 and 6,500. But, he cautions, this is “probably a vast underestimate”. (Haroon Siddiqui, ‘Counting the casualties in Iraq and Afghanistan,’ Toronto Star, September 23, 2007)

There is no reason to believe that the application of the same methodology in Iraq is generating very different results. But IBC has never, to our knowledge, accepted that their own count is “probably a vast underestimate” of the total death toll.

In the past, IBC’s response to the suggestion that violence prevents journalists from capturing many deaths has been, in effect, ‘Prove it!’ Well, the bureau chief of one of three Western media agencies providing a third of IBC’s data from Iraq sent this email to a colleague last year (the latter asked us to preserve the sender’s anonymity):
“iraq body count is i think a very misleading exercise. We know they must have been undercounting for at least the first two years because we know that we did not report anything like all the deaths we were aware of… we are also well aware that we are not aware of many deaths on any given day.” (Email sent October 25, 2006)

Despite IBC’s claims, nowhere in their discussion do they deal with the problem that journalistic reporting of violent deaths can decrease as violence increases, particularly when that includes violence against journalists, as is very much the case in Iraq.

More to the point, as data collectors, IBC are not in a position to comment authoritatively on the impact of violence on the capacity of journalists to report accurately from Iraq. As data collectors, they have no more insight, no deeper understanding, than anyone else.

The reasonable response to the question of political impacts on their database is not for IBC to authoritatively suggest that they “have dealt with” the problem of lack of journalistic coverage — to conclusively declare: “The death toll could be twice our number, but it could not possibly be 10 times higher” — but to openly acknowledge that their task is limited to the monitoring of media reports.

For leading mainstream journalists, and for IBC themselves, to present IBC as an informed and credible source on political realities on the ground in Iraq is highly inappropriate.

A good example of this distortion was provided on September 7 by Michael Gordon of the New York Times. Gordon offered positive spin on the ‘progress’ of the ’surge’:
“The most comprehensive and up-to-date military statistics show that American forces have made some headway toward a crucial goal of protecting the Iraqi population.” (Gordon, ‘Assessing the “Surge” - Hints of Progress, and Questions, in Iraq Data,’ New York Times, September 8, 2007)

In assessing evidence for this humanitarian “headway”, Gordon turned to IBC:
“Iraq Body Count, a British-based nongovernmental group that monitors civilian deaths, notes that the number of civilians who were killed by shootings, executions and bombs has declined from January through July.”

He quoted IBC:
“’Levels of violence reached an all-time high in the last six months of 2006… Only in comparison to that could the first half of 2007 be regarded as an improvement.’”
The last caveat was unimportant, the word supporters of the occupation were looking for was “improvement”.

But there is a problem with IBC’s evidence and with Gordon’s analysis of its significance. In fact, IBC have not at all found that “the number of civilians who were killed by shootings, executions and bombs has declined”. The website has found fewer +reports+ of deaths of civilians killed by shootings, executions and bombs in “information gathering and publishing agencies, principally the commercial news media who provide web access to their reports”.

While a significant proportion of the deaths recorded or corroborated by IBC come from “cumulative totals reported by official Iraqi sources, in particular the Medico-Legal Institutes (morgues) and, for corroboration purposes, the Ministry of Health”, IBC describes the commercial news media as their “main sources”. (Ibid)

And Les Roberts has commented:
“Media and government reports catch only the tip of the iceberg.” (Siddiqui, op. cit.)

For IBC to emphasise that “the first half of 2007 [could] be regarded as an improvement” on the basis of their data collection is therefore misleading. Indeed the whole basis of IBC’s comment was misleading:
“Levels of violence reached an all-time high in the last six months of 2006.”
In fact, levels of media +reporting+ of civilian deaths was at an all-time high in the last six months of 2006 — that is not the same thing. As a consequence, and as the material cited above from Patrick Ball and RSF makes clear, IBC are in a position to comment +only+ on numbers of media reports of deaths, not on the inferred significance of those numbers for political realities on the ground.

The Failure To Challenge Media Distortions
What is so disappointing is that while IBC are willing to stray radically beyond merely “reading press reports” with “care and literacy” to challenge scientific studies that do not in any way challenge their “irrefutable baseline figure”, they are apparently not willing to challenge media reports that in effect do challenge that figure. The New York Times report above was a good example. Another appeared in the Financial Times on September 10:

“The war has already cost the lives of 3,760 US troops, and wounded 28,000 more. Iraq Body Count, a group that monitors Iraqi deaths, estimates that 70,000 Iraqis have been killed. It says there has been a ‘modest improvement’ in security compared with the bloody second half of 2006….” (Demetri Sevastopulo, ‘Echoes of Westmoreland and Vietnam,’ Financial Times, September 10, 2007)

But IBC is +not+ “a group that monitors Iraqi deaths”; it is a group that monitors media reports of Iraqi deaths. And IBC does not monitor “Iraqi deaths”; it monitors media reports of Iraqi +civilian+ deaths as a result of violence. IBC does not monitor reports of war-related deaths due to disease, lack of food, water and medicine, and so on. IBC also does not collect reports of Iraqi military deaths.

Because IBC’s “irrefutable baseline” figure refers only to violent deaths of civilians reported by the media, the Financial Times in effect challenged that baseline by asserting that 70,000 Iraqis — i.e., civilians and military — had died. Readers might well have construed that some of these “Iraqi deaths” must have been military deaths, for example, and therefore will have come away from the article believing that many less than 70,000 civilians had died from violence.

The Financial Times could hardly be a more prestigious, influential and high-profile media outlet. And this kind of distortion has been repeated innumerable times, globally, since 2003. Notice, again, the complete inappropriateness of quoting IBC as an authoritative source reporting “a ‘modest improvement’ in security” on the basis of its data collection. As the Guatemala study above indicates, the drop in media reporting could be interpreted as indicating a +worsening+ of security, not least for journalists, leading to a drop in reporting of violent deaths.

Whereas IBC have responded vigorously, indeed tirelessly, in responding to the 2004 and 2006 Lancet studies (and to our criticism), to our knowledge they have all but ignored these actual challenges to their baseline figure — a figure which seeks to establish a “cautious minimum” for violent deaths of Iraqi civilians +alone+, not for “Iraqi casualties” in toto, as the Financial Times report suggests.

Indeed, far from exposing these abuses of their work, under ‘Press and media uses of IBC,’ IBC provide not a single word of criticism of media use of their work. Instead, one of the examples they choose to highlight is an Independent article from July 2005. The first sentence of the article reads:
“Almost 25,000 Iraqi civilians have been killed during the two years of war and insurgency that began with the US-led invasion in March 2003. More than a third have died as a result of action by allied forces.” (Terry Kirby and Elizabeth Davies, ‘Iraq conflict claims 34 civilian lives each day as “anarchy” beckons,’ The Independent, July 20, 2005)

It is striking that IBC link to a high-profile media report that so badly misrepresents its figures. As so often, this opening sentence gave the impression that IBC are recording the total number of civilian deaths, rather than merely recording deaths from violence as reported by the media. The extreme gravity of this distortion in downplaying the true extent of Iraqi casualties to the British public is clear enough, given, for example, Patrick Ball’s work.

Elsewhere, IBC write:
“A large number of press and media reports have cited our figures, discussed and assessed our work. Nearly all mentions have been in the context of drawing attention to the human cost of the war.” (John Sloboda, February 17, 2006)
Again, this is not mere data collection; it is political analysis of media performance. Having ourselves studied media reporting on Iraq closely over the last four years, we arrive at a very different conclusion: media reports have often cited IBC’s figures in the context of +burying+ the human cost of war.
As numerous studies over many decades have shown, it is quite simply the structural role of the corporate media to defend established power by minimising, as far as possible, public perception of the costs to civilians of US-UK state violence. This role has not suddenly changed in regard to Iraq. On the contrary, media performance on Iraq has been a text book example of a corporate propaganda system acting to protect allied elite interests.
Finally, the danger of moving beyond data collection is emphasised in this comment on IBC’s website in response to media reports of the “surge”:
“Despite any efforts put into the surge, the first six months of 2007 was still the most deadly first six months for civilians of any year since the invasion.” (’The Baghdad “surge” and civilian casualties‘)

This was also highly politicised analysis. IBC’s framing of the issue exactly matches that found in the pro-war Observer:
“Despite the surge, violence remains roughly at the same levels.” (‘Iraq benchmarks,’ The Observer, September 2, 2007)
Imagine what Western journalists would have made of a Soviet organisation observing that a particular period of time had been “the most deadly” for civilians in Afghanistan in the 1980s “despite” a massive surge in Soviet military activity.

And yet this is currently the standard line in mainstream reporting, part of a wider attempt to present the occupation as a well-intentioned effort to achieve peace and democracy, rather than conquest and control.

To their credit, IBC have made an improvement to their website. Their “counter”, which previously recorded “Minimum” and “Maximum” deaths in Iraq, has been changed. Viewed alongside the name Iraq Body Count, visitors were likely to assume that the “Maximum” category referred to the maximum possible number of civilian deaths in Iraq — the full body count — rather than the maximum number of deaths recorded in media reports. The counter now reads:
“Documented civilian deaths from violence 74,432 – 81,120”
IBC comment:
“The change to a simple unlabeled range is intended to help avoid misinterpretation or misrepresentation of these numbers as (for example) the ‘maximum possible’ death toll, or IBC’s ‘estimate’ of it.”

This is a welcome change. However, the very name of the website remains misleading. IBC is, in truth, an Iraq Reported Body Count — nothing more.


Media Lens is a UK-based media watchdog group headed by David Edwards and David Cromwell. The first Media Lens book, Guardians of Power: The Myth Of The Liberal Media, is now available (Pluto Books, London, 2006).

Aucun commentaire: