Baghdad's Murder Rate Irresponsibly Distorted by Lott

Last time I commented on Lott's claims about the Baghdad murder rate, I noted his pathological refusal to admit that he was wrong about the rate. Even though dozens of newspapers have reported that there are hundreds of murders each month in Baghdad (see the table with some of the stories at the end of this post), Lott insisted that the one single report he found that claimed that there were only 24 murders in October must be right and all the others must be wrong. He has now drafted on op-ed repeating his claim and complaining that the New York Times refused to print a "correction".

Lott's arguments about why all the other reports are wrong have to be seen to be believed:

I contacted the authors of both pieces. Albuquerque and O'Hanlon, who wrote the Times piece, provided two sources for their murder rate numbers: An article by Neil MacFarquhar in the New York Times (Sept. 16, 2003) and a piece by Lara Marlowe in the Irish Times (Oct. 11, 2003). Yet, both references clearly stated that much more than murder was included in the reports that they used from the Baghdad morgue. MacFarquhar notes that these deaths also included "automobile accidents" and cases where people "were shot dead by American soldiers," cases that clearly did not involve murders. The Irish Times piece mentions that "up to a quarter of fatal shootings [in the morgue] are caused by U.S. troops."

All right, suppose we want to exclude shootings by US troops from the total. If "up to a quarter" are by US troops, then, at least three quarters are not by US troops. The Irish Times article reports that 518 people were shot dead in August. Three-quarters of 518 is 388. That is a very conservative estimate for the number of murders since it doesn't count those murdered by other means, or those victims whose bodies did not make it to the morgue and that it was "up to a quarter", so the true number could well be higher.

Lott doesn't mention it, but MacFarquhar's piece also reported that 70% of the deaths were from gunfire, so you can do a similar calculation with his data to also see that there were hundreds of murders a month.

Anyway, that's the logical way to correct the figures, but look at the way Lott "corrects" them:

For some perspective, in D.C., murders account for fewer than 5 percent of all deaths. Even counting only the types of deaths explicitly mentioned in the stories citing the Baghdad morgue (accidental deaths, murders, suicides) and assuming that soldiers were engaged in the same type of fighting in D.C. as they are in Iraq, murders in D.C. would account for just a third of deaths. (The respective numbers for the U.S. as a whole are even lower: a half of one percent and 11 percent.) Obviously, counting these other deaths as "murders" in D.C. would imply that murders were three to 20 times more common than they actually were.

The first thing to note is that Lott is trying to pretend that the body count at the morgue is for all deaths, even though the articles say that it is violent deaths. And if there are only 800 deaths a month of all kinds in a city of five million people then that translates to a life expectancy of about 400 years. But never mind that. I want you to admire the circularity of his argument. He assumes that murders in Baghdad would be about 5% of all deaths, just like in DC. But that will only be true if the murder rate in Baghdad is similar to that in DC. In other words, if you assume that the murder rate in Baghdad is similar to that in DC, you can show that the murder rate in Baghdad is similar to that in DC. If, instead of assuming what you want to show, you use the facts given in the articles, it is clear that the murder rate in Baghdad is much greater than than that in DC.

Lott continues:

The Wall Street Journal Europe instead relied on the U.S. Army 1st Division stationed in Baghdad. A public affairs officer with that division, Jason Beck, confirmed for me that a large part of the Iraqi legal system is being overseen by the U.S. JAG officers, and they are using the same standards for murder rates as used in the U.S. and separating out murders from other deaths.

And pretty clearly, they have only been able to record a small fraction of the murders that have occurred.

Lott finishes up by demanding that the New York Times print a "correction":

When a publication of record such as the New York Times gets Baghdad's October murder rates wrong by up to a factor of 28 to 1 and no correction is issued, the consequences are significant.

You have to admire Lott's chutzpah, but he should think bigger than this---he should demand that all the other newspapers listed below also issue "corrections".

Blogger Michael Williams is fooled by Lott, while spc67 is little more skeptical.

Update: More bloggers fooled by Lott: Del Simmons, Little Green Footballs and Perry.

Update 2: Two more suckers: Say Uncle and Ranting profs.

Stories reporting on murders in Baghdad are listed below:

22 May The Wichita Eagle In a month of operation since the war ended, morgue officials recorded 191 deaths from gunfire, compared with 10 to 15 per month before the war. ... In three of the 191 bodies that came to the morgue, medical examiners found smaller-caliber bullets that appeared to have come from weapons used by U.S. soldiers, Obaidi said. But, by far, most of the bullets found in the dead came from AK-47's, which U.S. troops don't normally use.
24 Jul The Times By the time [the morgue] closed at 8pm it had received 23 bodies; 18 were shooting victims.
9 Aug Associated Press

The morgue, which handles all violent or suspicious deaths, recorded 10 gunfire deaths in July 2002. This July it handled 470, said the director, Dr. Fa'aq Amin Bakr.

In normal times, Bakr said, gunshots account for less than 10 percent of Baghdad's unnatural deaths, with the bulk coming from traffic accidents, drownings, burns and asphyxia. In July, more than half of the 702 bodies brought in had died from bullets, he said.

28 Aug Agence France Presse

"We used to deal with about 3,500 suspicious deaths here a year," says Faik Amin Bakr, the head of Baghdad's Medical-Legal Institute, adding that only about 10 percent of those were due to gunshot wounds.

"In July this year alone, we got 780 deaths, with 460 of them due to gunshot wounds. That's the equivalent in one month of deaths due to gunfire that we would normally get in a year and a half," he said.

Bakr adds that August produced similar figures, and notes that it is impossible to estimate the number of people who die violent deaths and whose bodies are not brought to the mortuary.

29 Aug National Post "murder capital of the world: 35 killings a day"
3 Sep The Boston Globe

In 2002 the Baghdad morgue conducted autopsies in 3,500 suspicious deaths; 350 died from gunshot wounds.

But in the three months since the fall of Hussein's government - May, June, and July - the morgue has logged 1,169 shooting deaths out of a total of 1,868 suspicious fatalities, according to the morgue's director, Dr. Faik Amin Baker.

17 Sep The New York Times 462 people dead under suspicious circumstances or in automobile accidents in May, about 70 percent from gunshot wounds; 626 in June; 751 in July; 872 in August. ... By comparison, last year there were 237 deaths in July, one of the highest months, with just 21 from gunfire.
25 Sep The Seattle Times

The number of reported gun-related killings in Baghdad has increased 25-fold since President Bush declared an end to major combat May 1. Before the war began, the morgue investigated an average of 20 deaths a month caused by firearms. In June, that number rose to 389 and in August it reached 518. Moreover, the overall number of suspicious deaths jumped from about 250 a month last year to 872 in August. ...

"I've been working in this morgue for 29 years," said Dr. Abdul Razzaq Ubaidi, a pathologist. "It used to be accidents and natural deaths. Now, there are too many weapons in society. We used to dissect six or seven bodies a day, but now we do 25 to 35 a day and 80 percent of them are bullet injuries. We have more freedom, but with the absence of security there is more freedom for murder."

10 Oct The Independent During September, civilian deaths by gunfire in Baghdad totalled 518. Under Saddam, deaths from gun violence in Baghdad averaged 6 per month. According to the central morgue in Baghdad, violent deaths reached 872 in August. The highest monthly toll in the previous year was 237 deaths, with just 21 from gunfire.
10 Oct Irish Times

Police brought 667 bodies to the morgue in the month of September. Of those, 372 - including 50 women - died of gunshot wounds, says Dr Qais Hassan, also a forensic pathologist and the director of the morgue's statistics department. The worst month this year was August, when 518 Baghdadis were shot dead, compared to 10 fatalities from bullets in August 2002.

The statistics tell the story of Baghad's descent into cold-blooded mayhem. In all of 2002, 174 people died of gunshot wounds in the capital. This year, until the end of September, pathologists recorded 2,173 deaths by firearms in Baghdad alone, almost all of them since May. The institute closed down for ..10 days when the regime fell in April, so dozens if not hundreds of deaths during that period were not counted.

"It's a disaster," Dr Hassan says. "At the end of the war, the Iraqi army left weapons all over the place. US forces could have collected them, but they didn't do it. Security is getting a little better, because there are more Iraqi police now." Dr Hassan estimates that up to a quarter of fatal shootings are caused by US troops.

12 Oct Newsday The morgue counted 667 homicide victims in September, down from 800 in August and 702 in July. But even the September rate is 42 times the rough average recorded last year. And the bloodshed - from crime, revenge killings, shootings of civilians by U.S. troops or guerrilla fighters - is some degree higher than what the morgue measures, because an unknown number of homicide victims are buried without being brought to the morgue.
13 Nov Zaman

According to the Baghdad Institute of Forensic Medicine, between 20 and 25 bodies are brought to the morgue daily. Most of the victims are killed by gunfire.

Chairman of the Institute Faik Emin Bekir said that 3,513 murders by unknown perpetrators have been officially registered since the day the war ended. This number includes bodies found by the police and brought to the forensic medicine institute. The institute, which mainly dealt with deaths resulting from work and traffic accidents during Saddam's rule, sheds some light on murders after the war. Bekir notes that murder statistics have reached the highest level in 40 years. Deaths caused by gunshot wounds were almost none before the war, he said

According to statistics, 872 bodies were brought to the institute in August alone. 850 of the deaths were caused by firearms.

Tags

More like this

Lott responds on his blog to Wyeth's accusation that he had no evidence for his claim about Baghdad murders. (My earlier comments are here.) Notice that Lott responds on a minor point, once again ducking the question of the coding errors. And while he links to Wyeth and responds to…
In July last year, Lott, armed with no evidence at all, claimed that Washington DC had a higher murder rate than Baghdad. Faced with overwhelming evidence to the contrary, Lott stuck to his guns, even demanding that the New York Times "correct" an article and use Lott's…
For those who still want to peddle the "why doesn't the media tell the truth about all the good things going on in Iraq" line, check out this story. Remember last month when the Pentagon claimed that the number of deaths in the Baghdad area had dropped dramatically since they launched a security…
I was curious to see what kind of defense Matt would put on against my suggestion of additional regulations to address the problem of gun violence and homicide in the US, and I was a bit disappointed to see that the response is largely a "no problem" argument. I had actually come into this debate…

Well, the how lott argument seems to be who do you trust.
The Baghdad morgues or the U.S. Army 1st Division stationed in Baghdad and the U.S. JAG officers.

Well...I guess the US army could be making up numbers and falsefing evidence. And i guess that Bush could have know the exact time and place that the Sept. 11 attackers would attack, and he is trying to cover up that fact buy sending Americans to die in another country.

Right and i am Osama Bin ladin.

If you like you can leave a comment on my site for Sandcastle on the number of dead bodies and number of gunshots he hears in Baghdad. I will tell him you stopped buy. I thin

Thanks for the link, I guess, but oh brother. Lott says that the murder rate is distorted by between 300% and 2800%, and you latch right onto the higher number. You also then look at August -- the month with the highest number of recorded gunshot deaths -- and make your argument from there, rather than from the average number.

Plus, the vast majority of dead people don't end up in the city morgue, so your number of 800 a month is totally meaningless.

You're grasping at statistical straws at least as much as Lott was. Frankly, in a city where people routinely fire automatic weapons into the air to celebrate birthdays and soccer games, it's pretty hard to glean a "murder rate" from gunshot death statistics alone.

Michael, His lower "300%" number is rubbish for the same reason -- it assumes that murder rates in DC are similar to those in Baghdad. I worked out the August number as an example of how to correct things. It should be pretty that obvious that the murder rates in the other months are in the hundreds as well. I specifically mentioned that not all deaths are counted, but that just means the murder rate is even higher. It's pretty easy to see by looking at the body if the person was killed by an "air-bullet". The victims are described as being murdered, rather than being killed by idiots firing AKs into the air.

cubicle, I don't think that either the morgue or the 1st ID are making things up. Why do you think that the 1st ID is able to detect every single murder that happens in Baghdad?

Tim, thanks for the link. I do still consider myself "fooled" by Lott's info, however.

And I was surprised for a moment to see my home town of Merced, CA listed in your list of topics.

I notice that you do not produce any evidence of the murder rate in Baghdad. Instead, you have chosen to attempt to manipulate the statistics in a manner even more blatant than the way you accuse John Lott of having done. I don't know Lott from Adam, and his previous conclusions seem a little too convenient to me, but yours are downright preposterous. Do you have some sort of evidence that the murder rate in Baghdad is not similar to the District's? Are you an expert on sociology or homicide? Then present that evidence, do not rely on unproven assertions of fact (as you are quick to accuse Lott of doing).

It appears to me that in your zeal to skewer someone you see as ideologically opposed to you, you have thrown out any pretense of rational academic debate. That's fine, if you want to attract defamation suits for maliciously accusing people of unethical conduct, but not so good if you intend to convince others of the truth of your arguments.

Ryan, it is beyond bizarre to claim that I don't present evidence. I presented twelve independent reports saying that Baghdad had a very high muder rate.

Did you even read my post before commenting?

Yes, I did read the piece. Very carefully. You provided clippings from unreliable news "reports" making uncited, unsubstantiated claims of casualties for which you provided no context to support your arguments. You did not make an attempt to link to the original sources of the facts from these news reports (probably because they are sketchy at best), and the overall impression you give your readers is that the issue isn't worth more to you than a five minute "Google News" search (excluding all news sources that do not agree with "The Guardian"). It is almost insulting to the reader that you couldn't even be bothered to provide links to the "reports" temselves.

Your work is seriously deficient. It amounts to an attack piece that is so raw and unsubstantiated it is almost malicious in nature, and I for one do not take it seriously. If you feel strongly enough about it, do the research...and show your work. In short, ask yourself if your off-the-cuff theorizing would stand up for five minutes to the scrutiny you are urging others to aim at Lott.

Ryan, so you concede that I did provide evidence. Your objection is now that I don't provide links to the original articles. I found the articles via a Factiva search you can see the original articles here
and here.
Instead of asking to see the original articles, or conducting your own search, you falsely accused me of not having any evidence. It appears to me that you are not interested in looking at the evidence, but are instead trying to pick a fight.

Oh, my apologies. You are actually a serious academic. Thank you so very much for your follow up. I can see that it includes both the original source material you relied on to assert that a respected academic is not just mistaken, but lying. In addition, you provided both definitive evidence and the context for that evidence showing that your hypothesis that Baghdad's murder rate is not similar to Washington's.

And you were able to do so with only ten minutes of work using an internet news search engine. Amazing!

Seriously. I am treating your piece exactly as it deserves, like a polemic against someone you seem to have a personal problem with. Nothing wrong with that, but don't think for a moment that posting a blurb on the university's website amounts to a refutation of anyone's serious research. It is what it is, and it certainly isn't a sufficient basis to assert that someone else is guilty of academic dishonesty.

If you really believe your research supports such a charge, do the heavy lifting and show your work. If you want others to believe you, I recommend submitting your work to your peers and publishing it for public review. I highly doubt you would be willing to endure the public opprobrium such an avenue would undoubtedly dump on your head, but until you do so, why should anyone who hasn't already made up their mind believe you when you claim that someone who has done so is not just wrong, but a liar?

Lott misses something in the news article--the number of bodies is the number brought by POLICE to the morgue. If you die of natural causes (and likely if you die from a gun shot wound in the hospital), the police don't bring you to the morgue (at least in the US).
The rest of his statistics are laughable examples of math abuse

By Stop math abuse now (not verified) on 17 Dec 2003 #permalink

...

Ryan-

Lambert has done the footwork needed to show that, at the least, there is a legitimate question over the murder rates cited. You attempt to dismiss his claims by an appeal to the authority of governmental reports.

There is no reason to take US government reports more seriously than those of journalists. It is a fact that on numerous occasions government officials have been contradicted by journalists, and later verified to have given false information, intentionally or otherwise. This has occurred in the Nixon, Reagan, Bush, Clinton, and Bush administrations. Many examples are common knowledge. I invite you to research instances of false reports from government sources at your leisure.

As you point out, facts must be verified. Assuming you can verify the source of the information, there is no quantitative difference between reports from one source or another. Verification is the important step.

Further, it appears that you regard Lott himself as an authority. Lott may be regarded as an authority by you, but Lambert has identified serious flaws in his investigative technique and has found credible evidence that is directly contrary to Lott's assertions. Lott's authority is specifically the subject of this article. This makes it inappropriate to appeal to his authority in his defense; to appeal to his authority in a question regarding his authority. This part of your argument is akin to saying "He is correct because he says that he is correct." That is not valid reasoning.

It seems quite clear that your assumption has been that Lambert is incorrect, that Lott is correct, and that the murder rate of Baghdad is similar to that of Washington DC. Your arguments here rely on disrespectful ad hominem phrasing as well as an implicit assumption that Lott is correct based on your perception of his authority. You give no evidence to support Lott, nor to refute Lambert.

You have attacked Lambert's sourcing in a deprecating and disrespectful manner, only to be confronted with credible sourcing that you might have found for yourself, as you claim, in a few minutes. In essence, you have not been diligent enough to check his argument before attacking, revealing your bias against his proposition. You chose to believe that he was incorrect without doing any research of your own into his argument.

At this point you became more insulting, rather than conceding the point or rejoining the argument with new information. If you are concerned that Lott may be incorrectly represented by Lambert, please, as you say, do the heavy lifting, and work to dismiss Lambert's counterclaim, using evidence that refutes his assertions rather than hyperbole and unsupported assumption.

Happy hunting.