How the IBC number is reported

In an earlier post on the IBC I wrote:

Sloboda says:

We've always said our work is an undercount, you can't possibly expect that a media-based analysis will get all the deaths. Our best estimate is that we've got about half the deaths that are out there.

OK, then why does the IBC page say "Iraq Body Count: Max 38661"? That's not really the maximum possible number of deaths, is it? Why not report their estimate that the true number of deaths is 70,000 or so?

IBC's Josh Dougherty responded with:

No. It's the maximum number of reported civilian deaths, as is stated on their homepage, counters, database..etc., and based on the methodology posted openly on their site.

But the number that people are interested in is the actual number of civilian deaths, and that is how the IBC number is described the vast majority of the time. I took a sample of 36 of the press mentions of the IBC in 2005 from the IBC's page on press coverage. In only five of these was the IBC described as an undercount and as just the number of deaths reported in the media. One of these wasn't press coverage at all but Milan Rai's report that the IBC criticised. Then we have Socialist Worker, a student newspaper and the World Socialist Website. The only one in major media was Carl Bialik in the Wall Street Journal. A further seven ten mentioned that it was compiled from media reports so a reader had some sort of chance of figuring out the relation between the IBC number and the actual number of deaths. The other 24 21 just reported it as the number of deaths with the IBC maximum often reported as an upper bound on the number of deaths. The full results are in a table below the fold.

Now I'm sure that no matter how hard they tried, some of the press would misreport what their number means, but they don't even seem to be trying. This has been going on for years and they must be aware of it because they compiled the list of press reports of their work. Why not say on the front page that they think the total number of deaths is twice as high? Why not contact reporters who get it wrong and set them right?

IBC number reported as number of deaths Mention that IBC is compiled from media reports Mention that IBC is an underestimate
"The Iraq Body Count website calculates that the total number of Iraqi civilians killed by military intervention could be as many as 17,721." The Independent - 28 January "The campaigning group Iraq Body Count says it recorded 11,163 civilian deaths using media reports over the same period." Channel 4 News (UK) - 30 October 2005 "Careful and conservative work by IBC principal researchers Hamit Dardagan, John Sloboda, Kay Williams and Peter Bagnall, showed that there had been 24,865 civilian war-related deaths, almost all of them as a direct result of violence, reported between 20 March 2003 and 19 March 2005." Milan Rai (Iraqmortality.org) - 14 October 2005
"Iraq Body Count, an organisation that tracks civilian deaths, indicates that 26,000 to 30,000 civilians have died since the war started." Irish Independent - 31 October 2005 "In one count, compiled by Iraq Body Count, a United States-based nonprofit group that tracks the civilian deaths using news media reports, the total of Iraqi dead since the American-led invasion is 26,690 to 30,051." New York Times - 26 October 2005 "The most quoted is Iraq Body Count (www.iraqbodycount.net). This currently gives a figure of between 26,092 and 29,401 civilians killed and is regularly updated. This figure is based on media reports, and those who produced the figure accept that it underestimates the real number of deaths." Socialist Worker - 01 October 2005

"Visit any anti-war website and you will see an Iraq Body Count counter with a ticking toll of the civilians killed over the previous 24 hours." Spiked-Online - 28 October 2005

"Nearly 25,000 civilians have been reported killed in the first two years, according to the U.K. group run by John Sloboda and Harmit Dardagan." The Village Voice - 21 July "As of the middle of March of this year, 37 percent of all non-combatant deaths were caused by U.S.-led coalition forces -- compared to 9 percent caused by insurgents. I'll repeat that: caused by U.S. forces. According to Iraq Body Count and the Oxford Research Group, the two independent researchers behind the study, the figures in the report should be regarded as the "baseline of the minimum number of deaths."" Purdue Exponent - 24 August 2005

"IraqBodyCount.net estimates Iraqi civilian casualties at between 26,000 and 30,000" Antiwar.com - 27 October 2005

"The report, based on analysis of civilian casualties reported in the news media, states that 24,685 civilians were killed and about 42,500 wounded." New York Times - 20 July "For example, the widely cited number last month of about 25,000 counts only violent deaths that have been reported to the media." Wall Street Journal - 05 August
"The Iraqi Body Count website says that as of today a minimum of 26,690 and a maximum of 30,051 Iraqi civilians have been killed." Media Monitors Network - 26 October 2005 "Nearly 25,000 civilians have been killed since the US-led invasion of Iraq in 2003, a US-British non-governmental organisation said, based on survey of media reports." Financial Times - 19 July "The dossier examines 24,865 civilian fatalities and 42,500 injuries in Iraq between March 19, 2003 and March 19, 2005. The study's methodology was meticulous but conservative, meaning the figures underestimate the extent of the carnage inflicted on the Iraqi people. Only deaths that were reported by at least two out of 152 English-language news sources have been counted." World Socialist Web Site - 26 July
"No one knows an exact number of Iraqi deaths, but there is some consensus -- including from a U.S. military spokesman and outside experts -- that an independent count of roughly 30,000 is a relatively credible tally of Iraqi civilian deaths." Associated Press - 26 October 2005 "As many as 17,721 civilians have been killed as a result of the invasion and subsequent violence, according to Iraq Body Count, a London-based group that opposes the war and compiles its casualty toll from media reports." Bloomberg - 27 January
"The price that the Iraqi people might have to pay for this venture, now more than 25,000 civilians according to Iraq Body Count, a U.S. and British civilian organization, was apparently not of sufficient concern." Cavalier Daily - 12 September 2005 "The number of dead Iraqis is sometimes estimated at over 100000, of which 14000+ are estimated to be civilian deaths; that you can check here- http://www.iraqbodycount.net/" The Washington Dispatch - 03 January
"some 25,000 Iraqi civilians and 2,000 American and British troops have since been killed in chaos that may yet wreak civil war." The Economist (book review) - 01 September 2005 "Iraq Body Count, an organisation that tracks civilian deaths through news reports, indicates that 26,000 to 30,000 Iraqi civilians have died since the war started in March 2003." Daily Telegraph (London) - 31 October 2005
"At press time, the organization reported a minimum of 22,850 and a maximum of 25,881 Iraqi civilian deaths resulting from the invasion and occupation of Iraq." American Journalism Review - August/September "Some news outlets cite estimates compiled by the website Iraq Body Count based on press reports, which are consistently lower (17,000 to 20,000 Iraqi civilian deaths since the invasion) than the Lancet study's findings. " Inter Press Service - 22 April
"A study released last week put a figure to the disturbing images: 25,000 Iraqi civilians have died violently since the U.S.-led invasion in March 2003." North Jersey Media Group - 25 July "Iraqbodycount.net, an anti-war Web site which catalogues the number of Iraqi civilian casualties reported in the media, gives the number as somewhere between 17,061 and 19,432." UCLA Daily Bruin - 21 March
"Almost 25,000 Iraqi civilians have died since US and British troops invaded the country, an average of 34 every single day, a British study said Tuesday." Agence France-Presse - 23 July
"It reveals a total of 24,865 civilians were killed and 42,500 injured between 20 March 2003 and 19 March 2005." Daily Mirror - 21 July
"Lord Ramsbotham makes a cogent case for putting the number of civilian deaths in Iraq since the invasion at 24-25,000." Times (London) - 02 August
"US-led forces, insurgents and criminal gangs have killed nearly 25,000 civilians, police, and army recruits since the war began in March 2003, according to a survey by Iraq Body Count" ABC News (Australia) - 20 July
"Nearly 25,000 Iraqi civilians have died in the two years since US and British troops invaded Iraq - an average of 34 people a day - according to figures published yesterday by British academics." Daily Telegraph - 21 July
"The nonprofit Iraq Body Count Web site on Friday said it was between 21,523 and 24,415 - which reflects uncertainty whether some of the dead were civilians or insurgents." Philadelphia Inquirer - 15 May
"The number of Iraqi civilians killed in the war remains unclear. Data compiled by the Web site iraqbodycount.org suggests that between 21,000 and 25,000 civilians have been confirmed killed." CNN - 09 May
"According to the independent website Iraq Body Count, between 21,000 and 24,000 civilians have been killed as a result of fighting since the start of the war." Agence France-Presse - 30 April
"A different figure comes from an independent team of researchers running the website Iraq Body Count, which counts only civilian deaths, and only those confirmed by press reports or hospital and morgue officials. Yesterday it was between 16,069 and 18,339." Sydney Morning Herald - 24 February
"An independent project Iraq Body Count (http://www.iraqbodycount.net/) estimates that more than 17,000 civilians have been killed by military intervention in Iraq. " Radio Free Europe/Radio Liberty - 13 April
"According to recent figures announced by academics and peace activists broadcast on the Internet site www.iraqbodycount.net based on two media institutions, between 17,053 and 19,422 civilians have lost their lives since the beginning of the US occupation in Iraq." Zaman Online - 18 March

Methodology: I visited every fourth link on the IBC's list of press coverage links starting at the first one. I stopped when I got to the ones from 2004, so that I had one quarter of the links from 2005. I ignored broken links.

Correction and Clarification: Anna Plurabella has identified a few errors in the table. I placed three of the stories in the first column when they should have been in the second one because they mentioned that the IBC number was compiled from media reports. I've fixed this, as well as an incorrect link and giing the total number as 37 instead of 36. Note that if I missed these ones despite actually looking for them, what chance does the average reader have?

There are also four stories in the first column that eventually get around, several paragraphs later, to mentioning that the IBC number is compiled from media reports. I put them in the first column because I didn't read that far. And they belong there -- readers often won't read that far either and even if they do, they are unlikely to go back and reassess the IBC number in the light of the new information.

Tags

More like this

"One of these wasn't press coverage at all but Milan Rai's report that the IBC criticised."

That is very misleading in this context Tim, as our criticism had nothing to do with the issue you are raising in this posting. We did not criticize it on its description of IBC, but rather on its comparison of ILCS/Lancet which contained two calculation errors originating from you, which you have now conceded were indeed errors.

We also praised the piece for being an overall fair-minded analysis.

"but they don't even seem to be trying"

Read:
http://www.iraqbodycount.org/editorial/defended/app.5.a.php

Our main work is compiling our data, which itself takes up most of our members' available time, not educating the media, but we have made many efforts in this area. We will continue these efforts on the side, taking all reasonable suggestions into account, but our main focus must continue to be the main data compiling.

"Why not say on the front page that they think the total number of deaths is twice as high?"

Because we don't know that for certain to be the case, among other reasons I gave to you in a previous thread, and which you seem to have wholly ignored.

We say the appropriate point in our FAQ, and in numerous of our online editorials and press releases. So our view that IBC is bound to be below (and not exactly or above) the full toll is all over our website. It does not specifically say that on the first page itself, but rather indicates clearly that the total is of *reported* civilian deaths, not an estimate from a sample.

Also, on your list. I have found several errors there too.

Cavalier Daily - 12 September 2005
This is a perfectly appropriate reference as it says "more than" our figure.

Radio Free Europe/Radio Liberty - 13 April
This again seems like an essentially accurate remark as it indicates at least that we would "estimate" (as in our FAQ and elsewhere) that the true total is "more than" our figures of reported deaths.

Inter Press Service - 22 April
This is again an accurate remark. IBC is lower than Lancet, the reason being that the two different methodologies are trying to measure different things, one the total of known and documented deaths, the other the best guess at this and the remaining unknown and undocumented deaths.

UCLA Daily Bruin - 21 March
Sydney Morning Herald - 24 February
These two are in your left column, but should be in the middle one as they both describe that IBC is compiled from media reports.

I have long held that the IBC number is widely misused, and thus is doing the world a disservice.

I do appreciate the hard work, though.

However, If the purpose of ongoing research is to constantly publicize results, then the organization doing so has a responsibility to reasonably ensure that the information is being properly interpreted by the media/public. If that is not the case, as we have here, then either the material needs to be presented in a different format, or the organization should be more active in pursuing retractions and corrections.

wrt counting Iraqi dead, the argument that the 'IBC Contradicts the Lancet study' has got to be the most widely peddled piece of mis/disinformation out there.

I used to work in PR. If I was grading the IBC on it's work, I'd give it an A for presentation, C- for results.

JoshD - If IBC were concerned about misreporting of their figures they would rebutt incorrect media reports. Instead you send them traffic, with no warnings about inaccuracy, no criticism, no caveats - just traffic. You are providing publicity to reports which misrepresent your work and, in all the cases I highlighted, downplay Iraqi casualties .

Tim - I also compiled a sample of the highly misleading articles referenced from the Iraq Body Count in the Press page. JoshD's reply was quite revealing -

We don't do a litmus test first to see if we agree with all the claims therein or the wordings.

IBC have now had it pointed out to them on several occasions - here and on the Medialens board - that many of the articles they promote would fail such a "litmus test" because they make claims derived from IBC's work that simply aren't true.

Despite this, niether Josh, John Sloboda, nor anyone else has ever stated that they will do anything about it.

Like you said Tim, they don't even seem to be trying.

I must say I'm mystified by the IBC position on this. Why not just replace the misleading "min" and "max" with a fuller description. For that matter, why not pick a single definition and stick to it, then make it clear that this is a lower bound.

It seems unlikely that IBC wants to encourage understatement of the number of civilian deaths, but clearly this the main effect of its current position.

By John Quiggin (not verified) on 01 May 2006 #permalink

I think the medialens criticism has had the effect of making IBC intransigent on this issue. They could have a prominent link on the front page (like they have for the refutation of their critics) outlining why it is likely that their number is an undercount and by how much it might (in their view) be undercounting. But it's maybe become a matter of principle not to give in to their critics, some of whom have gone too far. (I've been a little short with them myself, but don't think I was personally abusive.) Sloboda is understandably upset. Now having put out that strongly worded refutation and then with Sloboda's BBC interview with its fulsome praise of the media (he seems quite taken with the symbiotic relationship he's developed and very reluctant to risk giving up any of the prestige IBC has acquired), I think it's going to be difficult for them to make any concession to the unwashed hordes on the left.

By the way, the one reply I received from someone at IBC was an attempt at humoring me. IBC forwarded an email by an anonymous supporter of IBC who essentially agreed that the media is under-reporting US-caused deaths and thought that IBC's data was valuable for media criticism purposes. Who, this anonymous antiwar IBC supporter wrote, could possibly doubt that the Western press was biased in just the ways medialens was claiming?

Well, John Sloboda, just for starters.

By Donald Johnson (not verified) on 01 May 2006 #permalink

So, 'Iraq Media Count' it is then? The IMC figure.

I continue to be amazed by joshd's responses.

>We say the appropriate point in our FAQ, and in numerous of our online editorials and press releases. So our view that IBC is bound to be below (and not exactly or above) the full toll is all over our website. It does not specifically say that on the first page itself, but rather indicates clearly that the total is of reported civilian deaths, not an estimate from a sample.

**SO WHY NOT PUT IT ON THE FRONT PAGE INSTEAD OF A MAX NUMBER THAT IS USUALLY MISUNDERSTOOD?**

well, the project is called "Iraq Body Count" isn't it? It isn't called "Estimated Iraq Body Count". The name of the project makes one instantly think of counting, not estimating numbers.

People misread and consciously and unconsciously misrepresent information all the time, in every field. This is not a unique occurance that only happens to IBC. People still get Lancet wrong ('and don't you critizise it now!'), and they still get UNDP wrong ('don't tell me this is more reliable than Lancet, ok!'). People get quotes of writers wrong all day long, and people misrepresent statements of politicians every second. And politicians spin and misrepresent what other politicians say every moment of every day. And every day, people who were corrected still refuse to accept their mistakes. This is human nature, not an exclusive IBC-only reality.

From the horrible "we don't do body counts" soundbite (and before actually) it has been clear that IBC was created to keep track of the actual reported number of actual dead civilians, it is not a survey group that estimates how many have been killed, and it never pretended to be one. If people don't know this, after more than three years, especially journalists, one should perhaps not listen to much what they have to say about anything.

One gets the feeling that people think that IBC somehow wants to keep the number of civilian dead as low as possible. It has been hinted at scores and scores of times throughout this Media Lens 'alert' campaign, by them, and by the people who comment on their 'message board'. I have always looked at the way IBC goes about their invaluable business with great respect. The number is what the number is, they do their work quitely and seem incredibly dedicated, and professional (sic!), but they're counting the number of reported deaths, they're not in the business of estimating what the number of dead could possibly be. They don't operate under some kind of plan where they think 200,000 have died, now let's just try to make that add up somehow. The number of dead is what it is, 50,000, 100,000, 400,000. It is my hope that one day we will know with more certainty how many lives have been lost in this war. And isn't this all we are trying to accomplish? To have more certainty, to know more. Lancet seems to have a number that people 'like' but it is an estimate, and a flawed estimate. We should all be looking at Lancet seriously and use it, but we need bigger and better and more contemporary surveys. And we need to be able to understand that IBC is doing something completely different.

Asking IBC to start posting estimates is like asking Lancet to announce in big bold letters next to their estimate the actual true number of dead people that they can verify. Was that 21 dead or something? What good would that do anyone?

The deeper and more personal this Media Lens vs IBC nonsense goes on it becomes clearer and clearer that most these people are not interested in knowing the actual number of dead at all. They just want to see the highest number they can find (it must be 300,000 - at least!!) in order to score cheap political points on blogs. It's disgraceful to see it play out like this.

(I don't count Tim Lambert as one of those people, just want to make that clear.)

Swift, you are undoubtedly right about some of the disreputable motives behind some of the criticism of medialens, but when you reduce the whole controversy to the lowest possible motives on one side and assume the best possible motives on the other, you aren't rising above the fray, but just flinging some mud of your own.

I had no quarrel with IBC whatsoever for the first two years. I got a little annoyed with them with their two year analysis last summer. I didn't expect them to endorse the Lancet number. I expected them to have a long section talking about the possible inadequacies of their approach if one wants to know the true figure and I wanted them to talk about how it's entirely possible that the sample of deaths they do report may be a biased sample, since it is likely we'll hear more about civilians killed by insurgent terrorists than it is that we'll hear about civilians killed by US forces. Instead they had this section singing the praises of the press--Sloboda strengthening that symbiotic relationship that he tells us he values so much in the BBC interview. Of course there are heroic reporters in Iraq. Reporters only rarely see for themselves who is dying--they have to depend on what various other people tell them. Most of these other people are government spokesmen or people who might have to fear retribution if they say the wrong things. And there have been cases where reporters have managed to investigate a US claim to have killed X insurgents, only to find the locals telling a very different story involving civilian deaths. I tend to bring up the Vietnam War or the French/Algerian war in this context--nobody really knows how many civilians died in either conflict because the US and France had no interest in carefully counting the number of civilians they'd killed and reporting it to the press, and it's sheer idiocy to think that reporters are necessarily going to be able to do this on their own. There are some wars where alternative sources of fairly reliable information are available--in El Salvador I think there were church groups and human rights groups that could track the true death toll, and consistently gave numbers higher than those of the State Department. There've been a couple of cases where Iraqi groups claim to have done this--that one group that said 37,000 deaths by late 2003 and the other last summer that claimed 128,000. But nobody knows how credible they are.

Anyway, I don't expect IBC to put up an estimate next to their counter--I expect them to put a statement right next to the counter which says their "maximum" figure could be a serious undercount. The IBC numbers are very commonly cited as an authoritative estimate of the true civilian death toll and IBC has to know about this misuse better than anyone. It's their responsibility to put a big fat disclaimer right next to their count and if that looks like a capitulation to all the nasty medialens people, well, maybe they ought to think of what they are doing this for.

By Donald Johnson (not verified) on 02 May 2006 #permalink

"SO WHY NOT PUT IT ON THE FRONT PAGE INSTEAD OF A MAX NUMBER THAT IS USUALLY MISUNDERSTOOD?"

And when misreporting of the kinds you speak continue, what then Tim?

I can only guess where the demands on IBC will go next.

How about a terms of use window that will pop up and block access to our homepage until or unless readers sign this online contract agreeing not to misquote our figures?

Or, we'll get that the disclaimer is too small, and it should be in three-foot high letters! ..etc. etc. Maybe our motives will be questioned for not having it in three-foot high letters dominating the homepage ..etc. etc. etc.

We've said the same thing for years, and claims that IBC records all deaths that take place in Iraq can be easily disproven by numerous of our own on-line publications where we state exactly the contrary. If this hasn't been sufficient, it's unlikely that putting one of these many disclaimers directly on our homepage would be. (swift's posting gets its points about right imo)

However, IBC does take all such suggestions, including these, into consideration and will continue to consider such things. Though it will not promise to implement any particular ones.

If people still misreport it, you contact them and ask them to make a correction. If you don't have the resources for this, I imagine the Media Lens folks would help.

JoshD -

However, IBC does take all such suggestions, including these, into consideration and will continue to consider such things.

That's good to hear. Thanks.

Tim -

If you don't have the resources for this, I imagine the Media Lens folks would help.

I can't speak for Medialens, but know that several people who post there, including myself, do contact journos/editors/officials over this issue, and will continue to do so. I believe one Medialens supporter mailed IBC offering to work on their website. The point remains, however, that IBC are making work by promoting misleading articles from their website. Only IBC can resolve that.

Another: "SO WHY NOT PUT IT ON THE FRONT PAGE INSTEAD OF A MAX NUMBER THAT IS USUALLY MISUNDERSTOOD?"

joshd: "And when misreporting of the kinds you speak continue, what then Tim?

I can only guess where the demands on IBC will go next. "

Ah, the old trick - when basic competency is demanded, start crying about the slippery slope to demanding perfection.

Josh, when Sloboda himself admits the true number might reasonably be twice as high as the IBC "maximum", then he himself ought to favor making that very clear with a simple disclaimer right next to the numbers. I know I've seen people in the NYT citing the IBC number as though it were an estimate and not a count. People often cite it as the most authoritative estimate of the death toll. You know they do, so why is it relevant that you'd probably continue to get demands from people to say what you don't believe (that the violent death rate might be ten times higher)? You're not on thin ice here--you're on no ice at all.

By Donald Johnson (not verified) on 03 May 2006 #permalink

I think that you've succeeded in illustrating that you are prepared to go to any length to discredit IBC. Starting with article 1, whose first 12 words would put it into the middle column of your table but ends up in the column which reinforces your prejudices.

The weakness of your "methodology" has been explained here:

http://members.boardhost.com/DT3rd/msg/1163587883.html

Perhaps you need to add a caveat to your blog entries?

Clearly not as desperate as a computer scientist who is unable to come up with a method to handle articles that match multiple criteria.

I'd have expected more of an effort on scienceblogs.com...

Tim Lambert claims that in his sample of 37 press mentions of IBC, 25 misrepresent IBC's figure (by failing to mention that it's based on media reports or represents an undercount). But Lambert's analysis is riddled with errors. Of the "25", I count only 4 which come close to misrepresenting IBC in the manner stated by him.

Of the "25":
9 clearly state that IBC uses media reports.
4 validly give the IBC figure as "confirmed killed" or that "have died" or "have been killed".
2 state that "more than" the IBC figure have been killed.
2 are unavailable for analysis (dead links).
1 presents the IBC figure as the "baseline of the minimum number".
1 doesn't mention death figures at all.
1 doesn't mention IBC at all.
4 misrepresent.

[That gives a total of 24, not 25 - Tim seems to have either miscounted or forgotten one. He tabulates 36 examples total, not 37 as he states].

So, 4 misrepresentations out of 36 press mentions. (Possibly 6, if you give Lambert the benefit of the doubt on those dead links). Now, let's look at these "damaging" misrepresentations...

The first of the 4 misrepresentations is from an antiwar.com article. The second is from an anti-war piece at Media Monitors Network. The third is from an article titled "Iraqi coffin-makers yearn for peace" (from Agence France-Presse). The fourth is from an analysis in the Independent (Patrick Cockburn is a co-contributor).

Some really "damaging" coverage there. No doubt it's being used by war apologists to "damage the anti-war movement" and to "hurt the Iraqi people".

Incidentally, there are more press reports in Tim's sample which omit to mention the Lancet study than which misrepresent IBC. If this were a rational debate, without double standards, MediaLens followers would be demanding that Les Roberts spends more of his time emailing media outlets to complain about these omissions.

By Robert Shone (not verified) on 15 Nov 2006 #permalink

[The above post, and the following, were originally from the POV board, but relate directly to Tim's analysis. http://members.boardhost.com/DT3rd/msg/1163587883.html ]

Here are quotes from the 24 press reports which Tim Lambert erroneously claims misrepresent IBC's figure (by failing to mention that it's based on media reports or represents an undercount).

The 9 that clearly state that IBC uses media reports:-

http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2005/10/31/wirq31…
Iraq Body Count, an organisation that tracks civilian deaths through news reports, indicates that 26,000 to 30,000 Iraqi civilians have died since the war started in March 2003.

http://www.wtnh.com/Global/story.asp?S=4029519&nav=3YeX
Iraq Body Count, a British research group that compiles its figures from reports by the major news agencies and British and U.S. newspapers, has said that as many as 30,051 Iraqis have been killed since the war started. Other estimates range as high as 100,000.

http://www.ajr.org/Article.asp?id=3919
Iraq Body Count, a group that has tracked thousands of civilian deaths based on media reports.[...]The Brookings Institution's Iraq Index provides a monthly estimate, but it's a modified version of Iraq Body Count's numbers. Then there's the estimate of a Johns Hopkins research team whose report was published by the Lancet medical journal in October 2004. Using a random sampling technique, the researchers estimated about 100,000 civilian deaths.

http://www.abc.net.au/news/newsitems/200507/s1417965.htm
US-led forces, insurgents and criminal gangs have killed nearly 25,000 civilians, police, and army recruits since the war began in March 2003, according to a survey by Iraq Body Count.[...] The US-British non-government group has monitored media reports to compile the tally.
[...] Another survey, published in Britain's Lancet medical journal last October, found nearly 100,000 deaths in the 18 months after the invasion, more than half due to violence.

[Note: Lambert makes a further error with the following. The report he quotes is not the report he links to. The latter is the correct one according to his "methodology", but it contains references not to IBC's accumulated figure, but to daily details of atrocities. Lambert instead quotes a different article as a supposed example of a misrepresentation of IBC. But this, also, is incorrect - as you can see from the text emphasised in bold]

http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2005/07/20/wirq20… (quoted by Lambert, and by me, below)
http://tinyurl.com/c8srh (linked to by Lambert)

Nearly 25,000 Iraqi civilians have died in the two years since US and British troops invaded Iraq - an average of 34 people a day - according to figures published yesterday by British academics. [...] Iraq Body Count draws up its figures by compiling deaths reported in the stream of international media reports, and by collating data from other sources such as Iraqi morgues. Given the limited sources of reported deaths, Iraq Body Count says its figures are likely to be an underestimate of the real totals.

http://www.ipsnews.net/interna.asp?idnews=28425
Some news outlets cite estimates compiled by the website Iraq Body Count based on press reports, which are consistently lower (17,000 to 20,000 Iraqi civilian deaths since the invasion) than the Lancet study's findings.

http://www.dailybruin.ucla.edu/news/articles.asp?id=32581
Iraqbodycount.net, an anti-war Web site which catalogues the number of Iraqi civilian casualties reported in the media, gives the number as somewhere between 17,061 and 19,432.

http://www.zaman.com/?bl=international&alt=&hn=17589
According to recent figures announced by academics and peace activists broadcast on the Internet site www.iraqbodycount.net based on two media institutions, between 17,053 and 19,422 civilians have lost their lives since the beginning of the US occupation in Iraq.

http://www.smh.com.au/news/Miranda-Devine/Critics-are-willing-Iraq-to-f…
A different figure comes from an independent team of researchers running the website Iraq Body Count, which counts only civilian deaths, and only those confirmed by press reports or hospital and morgue officials. Yesterday it was between 16,069 and 18,339.

The 4 which validly give the IBC figure as "confirmed killed" or that "have died" or "have been killed":-

http://edition.cnn.com/2005/WORLD/meast/05/08/iraq.main/
The number of Iraqi civilians killed in the war remains unclear. Data compiled by the Web site iraqbodycount.org suggests that between 21,000 and 25,000 civilians have been confirmed killed.

http://www.unison.ie/stories.php3?ca=33&si=1497312
Iraq Body Count, an organisation that tracks civilian deaths, indicates that 26,000 to 30,000 civilians have died since the war started. Of those, around 9,000 are reported to have been killed by the American military itself.

http://www.economist.com/books/displaystory.cfm?story_id=4342329
"AT 5:34am, on Thursday March 20th 2003, the United States began a war of its own choosing, buoyed by grand ambition and perhaps folly." Its declared reason--to wrest terrifying weapons from Saddam Hussein--was bogus or a delusion; its plan for the occupation that followed a scribbled afterthought; some 25,000 Iraqi civilians and 2,000 American and British troops have since been killed in chaos that may yet wreak civil war.

http://tinyurl.com/wzkfj
A study released last week put a figure to the disturbing images: 25,000 Iraqi civilians have died violently since the U.S.-led invasion in March 2003. Most of them, however, were killed not by insurgents but by American troops. In fact, more than four times as many deaths were caused by military operations than by insurgents.

The 2 that state "more than" the IBC figure have been killed:-

http://www.rferl.org/featuresarticle/2005/4/1A7E927A-4306-4650-9E35-27B…
An independent project Iraq Body Count (http://www.iraqbodycount.net/) estimates that more than 17,000 civilians have been killed by military intervention in Iraq. Blomfield estimated that Iraqi Body Count assessments are "probably 60 percent accurate," and the Iraqi Health Ministry's statistics "70 percent accurate."

http://www.cavalierdaily.com/CVArticle.asp?ID=24335&pid=1335
The price that the Iraqi people might have to pay for this venture, now more than 25,000 civilians according to Iraq Body Count, a U.S. and British civilian organization, was apparently not of sufficient concern.

The two that are unavailable for analysis (dead links):-

http://www.philly.com/mld/philly/news/11648292.htm
http://www.manilatimes.net/national/2005/may/02/yehey/opinion/20050502o…

The one that presents the IBC figure as minimum baseline:-

http://www.mirror.co.uk/news/tm_objectid=15761949&method=full&siteid=94…
The detailed report provided an "absolutely firm, unshakeable baseline of the minimum number of violent deaths", he added.

The one that doesn't mention death figures at all:-

http://www.spiked-online.com/Printable/0000000CADFE.htm
Visit any anti-war website and you will see an Iraq Body Count counter with a ticking toll of the civilians killed over the previous 24 hours. Others have US military death-counters which hotly anticipated the two-thousandth death earlier this week.

The one that doesn't mention IBC at all:-

http://www.timesonline.co.uk/article/0,,59-1717031,00.html
Sir, Lord Ramsbotham makes a cogent case for putting the number of civilian deaths in Iraq since the invasion at 24-25,000. Are any figures available on the number of civilian deaths during a comparable period in Iraq prior to the invasion?

The four that misrepresent:-

http://www.antiwar.com/pena/?articleid=7792
According to the Iraq Index, Iraqi civilian casualties have climbed relatively steadily from day one of the war to an estimated total of between 15,000 and 27,000 (IraqBodyCount.net estimates Iraqi civilian casualties at between 26,000 and 30,000, while other organizations put the number at over 30,000).

http://usa.mediamonitors.net/content/view/full/21892
In the meantime, our men are "sacrificing" their lives and we can't even put a definition as to why. For the cause of humanity and war crimes, and to trample down a man who murdered his own people in the name of democracy? Who is committing the war crimes here? The Iraqi Body Count website says that as of today a minimum of 26,690 and a maximum of 30,051 Iraqi civilians have been killed.

http://sify.com/news/fullstory.php?id=13902177
Iraqi coffin-makers yearn for peace
With a daily fare of bombs and shootings up and down the country, barely a day goes by with out fresh victims of the violence. Almost 25,000 Iraqi civilians have died since US and British troops invaded the country, an average of 34 every single day, a British study said Tuesday.

http://news.independent.co.uk/world/middle_east/story.jsp?story=605327
The Iraq Body Count website calculates that the total number of Iraqi civilians killed by military intervention could be as many as 17,721. There are no reliable figures for Iraqi military casualties but we do know that at least 6,370 Iraqi soldiers died during the war itself. The fact that the allies have never bothered to count the dead is seen as an insult in Iraq. Analysis by Rupert Cornwell, Andrew Grice, Patrick Cockburn, Anne Penketh, Andrew Buncombe, Ben Russell, Stephen Castle and Elizabeth Davies.

By Robert Shone (not verified) on 15 Nov 2006 #permalink

Some of those posts clearly indicate that the IBC number is the bare minimum. Others do not. I don't really care about the issue of how IBC is cited, so long as the Lancet numbers are cited as well, but I'd say that more than 4 of those stories would leave a false impression in the minds of the average newspaper reader--that would include most of the people not obsessed with the subject as people here tend to be.

By Donald Johnson (not verified) on 16 Nov 2006 #permalink

Thanks but I for one read them the exact same way Tim Lambert does Robert, I've no idea how or why you wouldn't. In short the IBC's behavior is encouraging the misrepresentation of its own data in the media and that misrepresentation is just what I see in the extracts you've kindly provided.

Donald wrote:
>> more than 4 of those stories would leave a false
>> impression in the minds of the average newspaper reader

Possibly, but the question is whether emails from IBC to the media outlets would change this (in cases where there's no unequivocal misrepresentation to start with - ie the vast majority of cases). Basically you wouldn't be asking IBC to complain about media mistakes (since they're not mistakes), but to attempt to get media to alter wording (to give a different "impression"). Based on my own experience trying to get media to present a different "impression" (as opposed to correcting mistakes), I think this would be wasted time.

I see few unequivocal mistakes in the coverage of IBC (I see more of those on the ML board, for example, than in media coverage). Even in a few of the 4 cases mentioned above, I think the reporters (actually anti-war campaigners in two of the cases) could make a "semantic" argument that there's no misrepresentation involved.

Anyone who doubts this should compare the (mis)representation of IBC by the MediaLens editors with a "misrepresentation" in the mainstream media (as claimed by critics of IBC):

"[IBC] puts the number of civilian deaths, based on reports by the media, at 43,000-48,000" (Edwards/Cromwell)

"[IBC] put the number at about 50,000 civilian deaths, although its estimates rely on previously published sources." (Times, 12/10/06)

If the Times is misrepresenting IBC, then so are the ML editors. There's a lot of irony here - Edwards/Cromwell regularly castigate IBC for not writing more often to complain to media outlets about these "misrepresentations". And meanwhile, they are authoring these very "misrepresentations" themselves (the above was from their New Statesman article). Actually, it's not just irony - it's gross hypocrisy.

Donald wrote:
>>I don't really care about the issue of how IBC is
>> cited, so long as the Lancet numbers are cited as well

Yes, I think the problem for many people is omission of any mention of the Lancet study. In Tim Lambert's sample, there are more omissions of the Lancet study than misrepresentations of IBC. To me, the implication is that all those people demanding that IBC write to media outlets should in fact be demanding the same of Les Roberts et al.

By Robert Shone (not verified) on 16 Nov 2006 #permalink

I put three articles in the first column which should have been in the second column. I've fixed this. The rest of Plurabella's criticism is misdirected and/or wrong. Stories in the second column mislead their readers about what the IBC number means, they are just not as badly misleading as those in the first column.

Plurabella also claims that four stories that present the IBC number as the number that "have died violently" (or similar language) are valid. They are not. The IBC number is the just the number of violent deaths that have been reported in the media, not the total number of violent deaths.

She also points to one story that she says doesn't mention the IBC at all. But the IBC included this story in their list of press coverage links, so they felt that the story was using the IBC number.

It's good to see Tim Lambert correcting some of his errors, but it's clear he hasn't gone far enough. If he's moved 3 articles to the middle column, that still leaves 6 which need moving, since they clearly "mention that IBC is compiled from media reports".

Perhaps a better solution would be for him to define exactly what his categories are. Because at the moment they seem to consist of whatever subjective opinion he has for deciding whether they are "valid".

The last point he makes (about the article which doesn't mention IBC) is irrelevant to his analysis. The fact stands that it's not a misrepresentation of IBC (perhaps he shouldn't have included it in his analysis).

By Robert Shone (not verified) on 16 Nov 2006 #permalink

should in fact be demanding the same of Les Roberts et al

Yes, the marketing of scientific research is the thing isn't it? There's a damning equivalence between IBC's enthusiasm for promoting itself, its encouragement of the misrepresentation of its wares by the media, and your belief that the scientists ought to be trying harder to competitively market themselves to the same inept or complicit media. Bewdiful Robert.

There are four articles that eventually, several paragraphs later, mention that the IBC number is compiled from media reports. I put them in the first column because I didn't read that far, and neither will many readers. I hope that clarifies the classification scheme for you.

The other article doesn't mention the IBC by name, but they are obviously referring to it. Certainly the IBC thinks so.

Tim Lambert wrote:
>> I put them in the first column because I didn't
>> read that far, and neither will many readers.

Thanks for your honesty, Tim. Perhaps you should state in your "methodology" that you didn't even bother to read, in full, the articles that you claim to analyse.

Even allowing for this, there remain many errors in your analysis:- the remaining articles that mention IBC's figure is based on media reports (and not "several paragraphs later"), the articles which state "more than" IBC's figure, the article which states that IBC's figure is for "confirmed killed".

And then there's the article which says: "Iraq Body Count says its figures are likely to be an underestimate of the real totals" (which is still in the first column of your table). And the article which states that IBC's count represents a "baseline of the minimum number" (also still in your first column).

I guess your excuse for classifying the latter two as misrepresentations is (again) that you never read to the end of the article. If this is how you conduct your "analysis", you should say so clearly and prominently (rather than leave it until way down in the comments section before making this admission).

By Robert Shone (not verified) on 17 Nov 2006 #permalink