The British Government Criticises The Lancet
Source: Justice Not Vengeance
By Milan Rai
15 December 2004
CRITICISING THE LANCET
As soon as the Lancet, the world’s leading medical journal, published an estimate that 98,000 Iraqis have died because of the invasion and occupation of Iraq, the British Government attempted to undermine this work.
The Lancet estimate (usually approximated to 100,000 deaths) includes Iraqi civilians and insurgents, and includes all causes of death, whether violent or nonviolent, and whether they were caused by foreigners (such as US pilots) or by Iraqis themselves.
For reasons explored in an earlier briefing (JNV 74), the 100,000 figure is likely to be an under-estimate of the death toll associated with the conflict. The British Government, however, has sought to portray the Lancet study as a massive over-estimate of the likely death toll in Iraq.
THREE DIFFERENT ESTIMATES MEASURING THREE DIFFERENT THINGS
Iraq Body Count (IBC) estimates ‘media-reported civilian deaths in Iraq resulting directly from military action by the USA and its allies’, including ‘civilian deaths resulting from the breakdown in law and order, and deaths due to inadequate health care or sanitation’: between 14,284 and 16,419 (over a period of 18 months). This is based on both direct reports by journalists, and records from morgues and hospitals.
The Iraqi Ministry of Health (MOH) claims to have recorded civilian and insurgent deaths due either to terrorist incidents or to US-led military action, based on bodies delivered to hospitals and deaths in hospital, between 5 Apr. and 5 Oct. 2004: 3,853 (over 6 months).
Finally, theLancet estimates extra deaths of all Iraqis from all causes since the invasion, based on a survey of 988 households containing 7868 residents throughout Iraq: 98,000 (over 17.8 months).
It is hardly surprising that the estimates differ. They each measure something different (and the MOH estimate covers only part of the post-invasion period).
ESTIMATING VIOLENT POST-INVASION DEATHS
The IBC and MOH figures are not necessarily inconsistent: extrapolating the MOH estimate backwards (in a linear, unscientific way) to the entire post-invasion period generates 12,000 violent deaths, near the IBC figure.
On the other hand, the Lancet study does not estimate the number of Iraqis (civilians, soldiers and insurgents) who have died a violent death in the post-invasion period. It must be larger than either the IBC or the MOH estimates.
Comparing IBC with the Lancet, it is not surprising that the IBC record of violent deaths of civilians and insurgents reported in the international press is much smaller than a Lancet-based estimate of the total number of violent deaths (reported and unreported), including those due to crime.
THE MINISTRY OF HEALTH DISCREPANCY
Comparing the MOH record and the Lancet estimate, even after adjusting for the fact that the latter has a time period three times as long, and includes the invasion itself, there would seem to be a sizeable gap, one that the Blair Government finds damning: ‘Hospitals in Iraq have no obvious reason to under-report the number of dead and injured. The Lancet article does not explain this discrepancy.’
In fact, the discrepancy was explained by Iraqi officials: ‘Iraqi health and hospital officials agreed that the statistics captured only part of the death toll… The numbers also exclude those whose bodies were too mutilated to be recovered at car bombing or other attacks, the ministry said. Ministry officials said they didn’t know how big the undercount was. “We have nothing to do with politics,” [Dr Shihab Ahmed] Jassim [member of the Ministry of Health Operations Section] said.’ (Knight Ridder, 25 Sept.)
Knight Ridder noted that, ‘most of the dead are believed to be civilians’. ‘Many Iraqi deaths, especially of insurgents, are never reported, so the actual number of Iraqis killed in fighting could be significantly higher.’ ‘Iraqi officials said the statistics proved that U.S. airstrikes intended for insurgents also were killing large numbers of innocent civilians.’
‘Iraqi officials said about two-thirds of the Iraqi deaths were caused by multinational forces and police; the remaining third died from insurgent attacks.’ ‘The ministry is convinced that nearly all of those reported dead are civilians, not insurgents.’
STATISTICS AND LIES
The Foreign Office attacked the Lancet study in a statement on 17 Nov., saying that the authors themselves ‘noted that the data on which they based their projections was of “limited precision”.’
But the ‘limitations’ identified by the authors included more factors that might have increased the estimated death toll, than might have lowered it. (JNV 74)
The Foreign Office cast doubt on the calculation of the conflict-related death toll: ‘This limited precision is reflected in the very large range which they use for their estimate of excess mortality (8,000 - 194,000).’ This is indeed the range of possible death tolls given by the Lancet study.
But it is highly misleading to say, as the Foreign Office does, that ‘Although the levels of probability vary across its range, any figure within this range is consistent with the data.’
The 98,000 figure is not only ‘consistent’ with the data, it is the most likely figure based on the information gathered in the study.
NORMAL DISTRIBUTION
The authors of the Lancet study wrote later: ‘If we had just visited the 32 neighborhoods without Falluja and did not look at the data or think about them, we would have reported 98,000 deaths, and said the measure was so imprecise that there was a 2.5% chance that there had been less than 8,000 deaths, a 10% chance that there had been less than about 45,000 deaths… all of those assumptions that go with normal distributions.’
A ‘normal distribution’ of (in this case) deaths, is a bell-shaped curve, highest in the middle, marking the most likely number of deaths, and falling away on both sides as the number becomes less and less likely.
If the rules of a ‘normal distribution’ were applying in a normal fashion, there would be a 2.5% chance that the number of conflict-related deaths was over 194,000 - or below 8,000; and a 10% chance that the number of ‘excess’ Iraqi deaths was below 45,000, or above another figure (around 155,000).
So it is highly misleading to say that ‘any figure within this range is consistent with the data’. In the absence of any further information or context, the levels of probability ‘vary’ in a particular way, becoming more likely as you approach 98,000, and less likely as you move away from this figure.
MORE LIKELY TO BE ABOVE THAN BELOW 100,000
But there is more information and context which alters the probabilities.
One of the Lancet authors notes: ‘We had two other pieces of information. First, violence accounted for only 2% of deaths before the war and was the main cause of death after the invasion. That is something new, consistent with the dramatic rise in mortality and reduces the likelihood that the true number was at the lower end of the confidence range. Secondly, there is the Falluja data, which imply that there are pockets of Anbar, or other communities like Falluja, experiencing intense conflict, that have far more deaths than the rest of the country.’
Although excluded from the headline estimate, US epidemiologist Richard Garfield continues, the Falluja data ‘tells us that the true death toll is far more likely to be on the high-side of our point estimate [98,000] than on the low side.’ (See JNV 74 for more details.)
FALLUJA FOOLING
The Foreign Office play fast and loose with the Falluja data, claiming, for example that the data from the town would ‘result in an estimated 200,000 excess deaths within Fallujah alone over the past 18 months… almost two-thirds of the total population of the town - which is just not credible.’
But the Lancet study says the 200,000 excess deaths would have taken place ‘in the 3% of Iraq represented by this cluster’, not just Falluja, but also its outlying areas, with a total population of 739,000.
While the Foreign Office notes that ‘The authors of the study understandably discounted the data’ from Falluja, the FO statement does not actually acknowledge that the 98,000 estimate excludes all Falluja data.
SHIFTING RESPONSIBILITY
Having tried (and failed) to cast doubt on the scale of conflict-related deaths uncovered by the Lancet, the Foreign Office concentrated on trying to shift responsibility for them, pointing out, for example, that the greatest increase in reported deaths was among men between 15-59 years old - possible insurgents.
But the study points out that men were also more likely to be in public and exposed to danger: 58% of vehicle accident-related fatalities also involved men between the age of 15 and 59. The gender-age combination proves nothing.
The FO tried other gambits: ‘Only 2 deaths are attributed in the survey to “anti-coalition forces”… an astonishingly small proportion’ of violent deaths.
Furthermore, ‘Since 58 of the 61 deaths attributed to Coalition forces were said to have been caused by “helicopter gunships, rockets or other forms of aerial weaponry”, it cannot have been possible for the families in every case to have known for certain who was responsible.’ Families may have been afraid to blame ‘the deaths of their relatives on the insurgents.’
All reasonable points, but they do not diminish the scale of extra deaths since the invasion, nor the enormous escalation in violent deaths.
The FO’s doubts about the balance of responsibility for violent deaths in Iraq lead inevitably in the direction of further inquiry.
Following the publication of the Lancet estimate, an eminent group of retired British senior officers and diplomats wrote to Tony Blair requesting a ‘comprehensive, independent inquiry’ into Iraqi deaths and injuries since March 2003 – ‘and the cause of those casualties.’
The signatories pointed out that Mr Blair had rejected the Lancet findings, but offered no comparable assessment of his own (the MOH record is hardly adequate). This Mr Blair refuses to do.
PLEASE 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:
‘JNV’, 29 Gensing Rd, St Leonards-on-Sea, East Sussex TN38 0HE.