Friday, May 1, 2009

How Many Have Died In Iraq And By What Means?

In April 2009, there were two reports on Iraqi casualties. One was a study by the Associated Press estimating how many Iraqis have died in the war, and the other was by the New England Journal of Medicine on what types of attacks were the deadliest. Both show how devastating the fighting in Iraq has been for its citizens.

On April 23, 2009 the Associated Press wrote about its tally for how many Iraqi citizens have been killed since the U.S. invasion. Many of the news wire services have regular reports on Iraqi deaths each month. Most use numbers released by the Iraqi government. Agence France Presse for example prints statistics from the Iraqi Defense, Interior, and Health ministries. The Associated Press is one of the few that keeps its own independent count. They have traditionally depended upon those same ministries as sources, but also look at press reports, and hospital records. Recently the Associated Press was given access to new data collected by the Iraqi Health Ministry, and searched through morgue records along with burials at the Shiite cemetery in the holy Shiite city of Najaf. All together the Associated Press estimated that about 110,600 Iraqis have died since the U.S. invasion. It's speculated that the actual number might be 10-20% higher since many deaths were not recorded. The majority of these, 87,215, were killed since 2005. Out of those 59,957 died during the height of the sectarian war from 2006-2007.

The Associated Press' count is close to other studies on the subject. In 2007 the United Nations' World Health Organization conducted a cluster study estimating that 151,000 Iraqis, including insurgents, died from 2003-2005. The Brookings Institution's Iraq Index, which relies upon press reports and the Pentagon counted 110,387 civilians killed from May 2003 when the invasion ended to March 2009. Iraq Body Count, one of the major sources for casualties in Iraq, found 100,346 dead from March 2003 to April 2009. Icasualties.org is another major organization that keeps track of deaths in Iraq, but they did not start counting Iraqis until 2006.

The most well known study of Iraqi deaths is called the Lancet report by Johns Hopkins University and al-Mustansiriya University in Baghdad. That July 2006 study led by Gilbert Burnham published in the Lancet medical journal claimed that 601,027 Iraqis had died since the U.S. invasion. The report has faced major criticisms since then that puts into question its authenticity. For one its margins of error are so large that its numbers could be off by tens of thousands. The study only polled 1,800 Iraqi families, an estimated 12,800 people out of 29 million total. Some have said that is too small a sample. Others claim the numbers were inflated for political reasons. Most importantly however the American Association for Public Opinion Research asked Burnham in early 2009 to answer some basic inquiries about what questions he asked, the instructions given to Iraqis, etc., and he only gave partial responses. He ended up being censured by the group for not cooperating, and violating basic research standards. The Lancet numbers always seemed incredibly high as it would mean only 1/6 of the deaths were ever reported or found. Where would all the other bodies be? Not to mention that since at the height of the fighting the insurgents and militias wanted people to see the death and destruction to intimidate them and force them out of their neighborhoods it wouldn't make sense to have so many unrecorded casualties.

Grand Totals of Iraqi Deaths

Associated Press (March 30, 2003 to April 2009): 110,600
Brookings Institution (May 2003 to March 2009): 110,387
Iraq Body Count (March 2003 to April 2, 2009): 100,436

Iraqi Civilian Deaths By Month

Brookings Institution Iraq Index


2003

2004

2005

2006

2007

2008

2009

Jan.


831

1,448

1,778

3,500

600

270

Feb.


938

1,599

2,165

2,700

700

230

March


1,190

1,333

2,378

2,400

750

260

April


2,104

1,200

2,284

2,500

950


May

866

1,627

1,777

2,669

2,600

550


June

1,026

1,021

1,517

3,149

1,950

490


July

935

932

1,658

3,590

2,350

500


Aug.

1,292

1,517

3,303

3,009

2,000

450


Sep.

860

1,434

1,964

3,345

1,100

400


Oct.

825

1,329

1,376

3,709

950

350


Nov.

677

2,638

1,640

3,462

750

270


Dec.

817

1,333

1,348

2,914

750

350


Totals:

7,298

16,804

20,163

34,452

24,550

6,360

760

Iraq Body Count


2003

2004

2005

2006

2007

2008

2009

Jan.


568

1,035

1,430

2,806

742

275

Feb.


604

1,201

1,449

2,536

1,007

343

March

3,976

957

786

1,789

2,614

1,538

408

April

3,437

1,256

1,025

1,590

2,436

1,260

20 to 4/2

May

545

619

1,226

2,103

2,766

759


June

593

833

1,215

2,426

2,086

669


July

650

762

1,444

3,159

2,536

583


Aug.

790

823

2,165

2,743

2,325

591


Sep.

553

943

1,330

2,408

1,221

535


Oct.

493

947

1,201

2,924

1,185

527


Nov.

478

1,533

1,208

2,969

1,043

472


Dec.

529

906

996

2,662

903

521


Totals:

12,044

10,751

14,832

27,652

24,457

9,204

1,406

The other major report that came out this month was in the New England Journal of Medicine on the causes of death in Iraq. It looked at casualties from the beginning of the U.S. invasion on March 20, 2003 to March 19, 2008 using data from Iraq Body Count. During that time the group recorded 91,358 Iraqi civilians and police that were killed. The study included police because they are considered normal parts of society, but did not include soldiers, militias, or insurgents. The study discounted prolonged periods of fighting such as the two battles for Fallujah, along with body counts from Iraqi morgues because there was no accounting for how exactly they died. That dropped 30,877 from the original total.

The remaining 60,481 deaths were then broken down by how they were killed, where the event happened, and how many casualties occurred in each incident. That analysis found that executions after being kidnapped or captured were the most common form of death in Iraq, accounting for 19,706 fatalities or 33% of the total. 29% of those showed signs of torture. Small arms fire, 11,877 or 20%, followed by suicide bombers, 8,708 or 14% of the total, were next. The majority of the latter, 5,401, were car bombers. Air attacks by Coalition forces, 3,050 or 5%, roadside bombs, 2,854 or 5%, and mortar fire, 2,079 or 3%, were at the bottom of the list. The most common incidents were gunfire, executions, and roadside bombs, while air attacks with ground fire, 17 civilians per event, bombing runs by planes, 17 civilians per, and suicide bombers on foot, 16 civilians per, resulted in the most casualties. This breakdown shows that while bombings usually make the headlines, more than 50% of the deaths were either up close and personal with executions or when civilians were caught in the cross fire of armed groups.

The New England Journal of Medicine study also broke down forms of death by age and gender. For men the deadliest forms of attack were executions with torture, 97%, executions, 95%, and gunfire, 91%. For women it was small-arms fire, 28%, air attacks by the Coalition, 14%, and executions, 13%. Gunfire, 19%, suicide bombers, 16%, and air attacks, 16%, were the costliest for children. Being captured or kidnapped was basically a death sentence for any man, while women and children appeared to be caught in gunfire, air raids, or bombings most of the time.

New England Journal Of Medicine Report On Deaths In Iraq 3/20/03-3/19/08

Method

Total No. of Civilians Killed (% of Total)

Execution – Any

19,706 (33%)

- With torture

5,760 (10%)

Gunfire

11,877 (20%)

Suicide bomb – Any

8,708 (14%)

- Bomber in vehicle

5,401 (9%)

- Bomber on foot

3,293 (5%)

Vehicle bomb

5,360 (9%)

Roadside bomb

2,854 (5%)

Mortar fire

2,079 (3%)

Air attack without ground fire –Any

2,363 (4%)

- Bomb only

479 (1%)

- Missile only

357 (1%)

Air attack with ground fire

687 (1%)

Total for all methods

60,481 (100%)



Both of these reports highlight the costs of war in Iraq. While violence has taken a dramatic drop from its height during the sectarian war of 2006-2007, there are still several hundred casualties per month. Millions of Iraqi families have also been affected, either through having lost a loved one or friend, or being forced from their homes because of the fighting. The conflict is likely to last several more years as well, only adding to these counts.

SOURCES

Associated Press, “AP count: Iraqi civilian, security details drop to near lowest level,” 3/3/09
- “Iraq study: Executions are leading cause of death,” 4/15/09

BBC, “Iraqi death researcher censured,” 2/4/09

Gamel, Kim, “Secret Tally Shows 87,000 Iraqis Have Been Killed Since 2005: AP,” 4/23/09

Hicks, Madelyn Hsiao-Rei, Dardagan, Hamit, Serdan, Gabriela Guerrero, Bagnall, Peter, Sloboda, John and Spagat, Michael, “The Weapons That Kill Civilians – Deaths of Children and Noncombatants in Iraq, 2003-2008, Abstract,” New England Journal of Medicine, 4/16/09

Karim, Ammar, “Iraq death toll for February ‘rises to 258,’” Agence France Presse, 3/1/09

O’Hanlon, Michael Campbell, Jason, “Iraq Index,” Brookings Institution, 4/30/09

6 comments:

Jeffrey said...

motown67,

This is good stuff. I'm going to add it to my "Look Back: 2008-9" under the "Articles and Documents" section, which I will slowly add to when I have time.

*

Buckles said...

I've seen a couple articles on this subject, but none compiled with the same depth.

thanks

nadia said...

Huh, I thought one of the huge factors for the big numbers and margin of error of the Lancet study was that it was estimating the number of excess deaths since '03, not just violent deaths which is what all the other studies AFAIK were trying to measure, which are quite different things.

motown67 said...

my understanding of the Lancet report is that it said 600,000 died from violence and an additional 50,000 died from the bad conditions in the country resulting from the war.

motown67 said...

Wikipedia has a good rundown of all the different criticisms made of the Lancet study.

http://en.wikipedia.org/wiki/Lancet_surveys_of_Iraq_War_casualties

Joel Wing said...

This is from a poster SLOVOSOD over at Tom Ricks' Best Defense Blog at Foreign Policy from July09. He listed some criticisms of the Lancet study as well:



epidemiologist who says the Lancet is wrong:

• Paul Spiegel, an epidemiologist at the UN, commented on IFHS (which estimated 151,000 violent deaths over the same period as Lancet 2006): “Overall, this [IFHS] is a very good study [...] What they have done that other studies have not is try to compensate for the inaccuracies and difficulties of these surveys.” He adds that “this does seem more believable to me [than Lancet 2006]“. http://tinyurl.com/53s82b

And more...

• Mark van der Laan, an authority in the field of biostatistics (and recipient of the Presidential Award of the Committee of Presidents of Statistical Societies) has written, with Leon de Winter, on the Lancet 2006 study:

“We conclude that it is virtually impossible to judge the value of the original data collected in the 47 clusters [of the Lancet study]. We also conclude that the estimates based upon these data are extremely unreliable and cannot stand a decent scientific evaluation.” http://tinyurl.com/4txbpw

And more...

The Journal of Peace Research Article of the Year Award has gone to Neil F. Johnson, Michael Spagat, Sean Gourley, Jukka-Pekka Onnela & Gesine Reinert for ‘Bias in Epidemiological Studies of Conflict Mortality’ (Journal of Peace Research 45(5): 653–663). http://jpr.sagepub.com/cgi/content/abstract/45/5/653

According to the jury who awarded the prize, the peer-reviewed study on msb:

…provides an important advance in the methodology for estimating the number of casualties in civil wars. The authors show convincingly that previous studies (The Lancet) which are based on a cross-street cluster-sampling algorithm (CSSA) have significantly overestimated the number of casualties in Iraq.



"Motown, there are no epidemiologists saying the Lancet study was wrong."

This is completely wrong. First of all, you can count the number of epidemiologists who work in the area of conflict mortality maybe on one hand. So it is not like some wide field, nor one that has some kind of monopoly on research in such fields. But here is a page collecting much criticism of the Lancet junk-science:

http://dissident93.wordpress.com/2008/10/20/project-censored-as-censors/

One of the links is from a report by the Centre for Research on the Epidemiology of Disasters, run by ....epidemioligists, who say the Lancet study was wrong:

"Research by Debarati Guha-Sapir and Olivier Degomme, from the Centre for Research on the Epidemiology of Disasters (CRED) estimates the total war-related death toll (for the period covered by Lancet 2006) at around 125,000. They reach this figure by correcting errors in the Lancet 2006 survey, and triangulating with IBC and ILCS data. http://tinyurl.com/3mlz5w"

Among many other such citations, the page also cites this paper which lays out a mountain of evidence that the Lancet study was, simply, a fraud. No wonder the authors of the Lancet study were hiding basic information about their methods from AAPOR:

http://personal.rhul.ac.uk/uhte/014/Ethics%20and%20Data%20Integrity_8_09_08.pdf