Likely Death Toll In Iraq
14 December 2004
PDF of this briefing is available here
Posted: 18 December
OVER 100,000 DEAD?
The Lancet, the world’s leading medical journal, has published
an estimate that 98,000 Iraqis have died because of the invasion
and occupation of Iraq. This estimate (usually approximated
to 100,000 deaths) includes Iraqi civilians and insurgents,
and includes all causes of death, both violent and nonviolent.
The 100,000 figure is likely to be an under-estimate.
The Lancet estimate is based on a survey of 988 households containing
7868 residents in 33 ‘clusters’ throughout Iraq.
One of the 33 ‘clusters’ visited during the study
was in the wartorn city of Falluja.
The number of violent deaths in Falluja was much higher than
in the other areas visited: ‘More than a third of reported
post-attack deaths (n=53), and two-thirds of violent deaths
(n=52) happened in the Falluja cluster.’ Because of this,
the Falluja data was excluded in calculating the 100,000 figure.
HOW THE ESTIMATE WAS COMPILED
The estimate was based on visiting 30 households in 32 randomly
selected ‘clusters’ throughout Iraq. (Note that
Falluja, the 33rd cluster, is not included in the following
calculations.) In each house, questions were asked about births
and deaths during the period from 1 January 2002 to mid-September
2004 (when the interviews were carried out).
The interviewers asked who had lived in each house and how many
of these residents had died during this period - and the cause
of their death.
Crucially, the survey only counted the deaths of people who
had lived in the house for at least two months before their
death. It did not count the deaths of family members who were
living elsewhere, or visitors who stayed for less than two months.
DEATHS PER PERSON-MONTH
The next step was to calculate the number of ‘deaths per
person-month’ for each household, the building blocks
of the national estimate.
In other words, if five people had lived in a house throughout
the whole post-invasion period (17.8 months), and a sixth resident
died after living in that house for eleven months of the post-invasion
period, that would be calculated as one death for 100 person-months,
or 0.01 deaths per person-month. (5 x 17.8 = 89 person-months,
then add 11 person-months to make 100).
The estimate was compiled by adding together the ‘deaths
per person-month’ for all households in each cluster,
and adding together all 32 clusters, for the period before the
war, and for the period after the invasion.
Then the pre-war figure (0.00042 deaths per person-month) is
subtracted from the post-invasion figure (0.00066), and multiplied
by the number of people in Iraq (24.4 million) and the length
of the post-invasion period (17.8 months).
This gives an estimate of the number of extra deaths after the
invasion, or what the Lancet study calls ‘the death toll
associated with the conflict’. The people who would not
have died if the death rate had continued to be the same as
before the war: 98,000 Iraqis.
(To repeat: not all of these are violent deaths; not all of
them were non-combatant civilians).
There was a curious aspect of the Lancet results: ‘It
is surprising that beyond the elevation in infant mortality
and the rate of violent death, mortality in Iraq seems otherwise
to be similar to the period preceding the invasion’, despite
the degeneration in drinking water quality and in health care,
This stability in nonviolent death rates seems implausible,
but does not indicate that the 100,000 figure is an under-estimate,
rather the opposite.
EXCLUDING MILITARY DEATHS
On the other hand, the fact that the study only counted deaths
of people who had been resident in the house for two months
beforehand ‘might have under-represented groups such as
the homeless, transients, and military personnel’, said
the survey team. This rule ‘probably excluded most military
In the absence of data about military deaths, we do not know
the scale of the under-estimate that this represents.
EXCLUDING ALL HEAVILY DAMAGED AREAS
The cluster selection process was a random one. Apart from Falluja,
none of the other heavily damaged Iraqi cities was chosen for
the survey, cities such as Ramadi, Najaf, or Tallafar.
One of the clusters surveyed was in the Shia slum in Baghdad
known as ‘Sadr City’, the scene of intense fighting
since the war. But the Sadr City cluster was ‘by random
chance in an unscathed neighbourhood with no reported deaths
from the months of recent clashes.’
And, of course, the only heavily damaged cluster that was selected
(in Falluja), was excluded from the estimation process.
CHECKING FOR AN EXAGGERATED DEATH RATE
The estimate would be seriously affected if families over-reported
recent deaths. In order to check whether over-reporting was
occuring, investigators asked for death certificates at the
end of interviews twice in every cluster.
‘In 63 of 78 (81%) households where confirmations were
attempted, respondents were able to produce the death certificate
for the decedent. When households could not produce the death
certificate, interviewers felt in all cases that the explanation
was reasonable - eg, the death had been very recent, the certificate
was locked away and only the husband who was not at home had
‘Interviewers also believed that in the Iraqi culture
it was unlikely for respondents to fabricate deaths... We believe
it is unlikely that recall bias existed in the reporting of
non-infant deaths [over-reporting recent deaths and under-reporting
earlier deaths], because of the certainty and precision with
which these deaths were reported, and the importance of burial
ceremonies in the Iraqi culture.’
THE INCENTIVES TO HIDE DEATHS 1
There would be a motivation to hide the deaths of insurgents
from investigators: ‘The under-reporting of adult deaths
recently or since the invasion to hide combatant deaths would
lead us to underestimate the death toll associated with the
invasion and occupation of Iraq.’
THE INCENTIVES TO HIDE DEATHS 2
Another distortion, in the opposite direction, would occur if
there were more residents in some households than were reported
to the survey - if, for example, families were hiding insurgents
This would mean that the postwar ratio of ‘deaths per
person-month’ should be lower, because there were actually
more ‘person-months’ in some households than were
The Lancet study concluded that this possible ‘hidden
insurgent effect’ was outweighed by a more powerful ‘food
ration’ effect: ‘Although certain individuals might
wish to remain hidden, the study team thought that respondents
would claim to have more household members than were actually
present to justify more ration distributions.’
Families in Iraq have food rations that are issued per person
(the ration card is the most important identity card in Iraq,
and is the basis for the electoral registration and voting process).
Every death in the family that is reported reduces the food
that is available to the family.
This would give an incentive for both under-reporting deaths,
and for claiming non-existent living members, both of which
would ‘have the effect of lowering mortality estimates
and thus lowering our estimate of the death toll associated
with this conflict.’
The survey team speculated that its sampling strategy might
have ‘missed small areas where a disproportionate number
of deaths occurred, or conversely, selected a neighbourhood
that was so severely affected by the war that it represents
virtually none of the population and thus has has skewed the
mortality estimate too high.’
In Kosovo, one study apparently showed that ‘there can
be a dramatic clustering of deaths in wars where many die from
Falluja may have been a relatively ‘small area where a
disproportionate number of deaths occurred’. It is not
clear, however, that it was the only such ‘small area’.
Excluding Falluja may have caused a major under-estimate: ‘In
our Falluja sample, we recorded 53 deaths when only 1.4 were
expected under the national pre-war rate. This indicates a point
estimate of 200,000 excess deaths’ (in the Falluja cluster
of 739,000 people).
Including this figure would nearly triple the estimate of deaths
associated with the conflict. The survey team comments: ‘the
uncertainty in this value is substantial and implies additional
deaths above those measured in the rest of the country.’
By effectively excluding military deaths, the survey is very
likely to have produced an under-estimate of the post-invasion
death toll in Iraq.
Given the incentive for households to under-report deaths in
general, in order to maximise food rations, and to under-report
deaths of insurgents in particular, in order to avoid the attention
of the authorities, there is reason to suspect further under-estimation
of the death toll.
While there may have been serious uncertainties associated with
the significance of the Falluja data, the complete exclusion
of Falluja, the inclusion of only a rare, undamaged part of
Sadr City, and the exclusion of all other heavily-damaged cities,
means that the Lancet survey almost certainly under-estimated
the death toll in Iraq.
Thus the conclusion that ‘the death toll associated with
the invasion and occupation of Iraq is probably about 100,000
and may be much higher’.
Counter Terror: Build Justice an international month of peace
action Sign the statement at <www.j-n-v.org> Sponsored
by Noam Chomsky, Kathy Kelly, Bruce Kent, Caroline Lucas, George
Monbiot, John Pilger, Haifa Zangana, Howard Zinn and others.
SUPPORT JNV (Justice Not Vengeance) 0845 458 9571
We are making
as many briefings as we can. Please help with printing/distribution
by sending cheques to:
Gensing Rd, St Leonards-on-Sea, East Sussex TN38 0HE.