By James Woodcock
Downloading PDF. Please wait... Issue 1970

How Iraq’s dead were counted

This article is over 16 years, 3 months old
A letter in last week’s Socialist Worker asked what evidence there was for the figure of over 100,000 dead in Iraq, which has frequently been quoted in this paper.
Issue 1970

A letter in last week’s Socialist Worker asked what evidence there was for the figure of over 100,000 dead in Iraq, which has frequently been quoted in this paper.

Two main sets of independent figures exist. Both show the murderous nature of the war and occupation. However, the estimates differ substantially. This is because they use different methods and measure different things.

The most quoted is Iraq Body Count ( 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.

It only counts civilian deaths inflicted by coalition forces, so excluding soldiers and members of the resistance killed, and civilians killed by suicide bombers, insurgent attacks, or violent crime. Yet all these deaths are direct consequences of the invasion.

Their figure also excludes non-violent deaths, for example those caused by damage to sanitation, power supplies and health services during the invasion and occupation. To be included in Iraq Body Count, each death must be reported by two separate media sources from an approved list.

The figure of 100,000 deaths comes from a survey conducted by Les Roberts and a team of Iraqis. It was published in the Lancet, a top medical journal (Mortality before and after the 2003 invasion of Iraq: cluster sample survey. Roberts L, Lafta R, Garfield R, Khudhairi J, Burnham G. The Lancet, Volume 364,  9448,  1857-1864 They estimate how many more deaths there were in Iraq after the invasion than before — counting deaths from all causes. First they went to the Iraqi ministry of health and asked them how many people were in each city and each village on 1 January 2003.

Then they randomly picked 33 neighbourhoods to visit. In each of these neighbourhoods they randomly selected a cluster of 30 houses. These households were interviewed to see how many people died in the 15 months after the invasion compared with the 18 months beforehand. They then extrapolated from this sample to the whole Iraqi population.

Avoiding distortion

Roberts’ team excluded the city of Fallujah from the final result because the number of deaths was so high that they feared it would distort their whole analysis.

The overall estimate, excluding Fallujah, of 98,000 dead is given with a range of statistical uncertainty from 8,000-194,000. This roughly means the authors are 95 percent confident the true figure is in this range. This size of the range has led to ill informed criticisms. But it is not unusual to have imprecise results when conducting studies, particularly in a country as dangerous as Iraq.

We also know that the figure is more likely to be nearer 100,000 than at either extreme. And, given our knowledge of Fallujah, where thousands are thought to have died, the real figure is likely to be far higher.

Roberts’ team found that the major causes of deaths changed from being heart attacks and other chronic diseases to violent deaths, the vast majority caused by the coalition forces and from gunships, rockets or other forms of aerial weaponry.

Supporting evidence for a high figure comes from an article published on 1 July 2004 in the New England Journal of Medicine. Based on interviews with troops returning from Iraq, doctors found that 14 percent of the ground forces and 28 percent of US Marines had killed a non-combatant. Given the numbers of troops in Iraq, and considering that this figure would exclude aerial bombardment, it suggests a higher figure than the Roberts study.

The Roberts study only estimates deaths up to September 2004, so the figure now is likely to be much higher. Part of the horror of war is how many victims become at most statistics rather than names and faces. Statistics come with uncertainty, but we must not let that uncertainty cloud the knowledge that tens and probably hundreds of thousands have been killed.

James Woodcock is an editor of Clinical Evidence


Sign up for our daily email update ‘Breakfast in Red’

Latest News

Make a donation to Socialist Worker

Help fund the resistance