Uncovering the deception
Following the US/UK invasion of Iraq in March 2003, there was concern in the coalition’s military occupation authorities to assemble information on Iraq’s population. In this context, it became known that a census had been undertaken in the centre and south of Iraq in 1997. Moreover, the census had included some very basic questions on child mortality. A copy of the census report was unearthed and forwarded to the US Census Bureau in Washington, DC. And, significantly, an article published in the New York Times in August 2003 quoted a Census Bureau official as stating that ‘it looks like child mortality may not have been quite as high during the mid-to-late 1990s as has been thought.’25
Then, in 2005, the report of a Working Group of an Independent Inquiry Committee established by the Secretary General of the UN to investigate the OFFP was published. This report contained an analysis of the child mortality data from the 1997 census.26 The data suffered from serious errors and limitations. Nevertheless, and to some surprise, the Working Group questioned whether there had been a huge rise in child mortality around 1991 as was indicated by the IST and, more especially, the ICMMS. The Working Group suggested that the Iraqi government might have tampered with the ICMMS data. These ideas were rejected by those who saw no reason to question the ICMMS child mortality estimates on the basis of the 1997 census data.27 28
However, since 2003 three major household surveys covering the whole of Iraq and collecting full birth histories from adult women have been conducted. The first survey was the 2004 Iraq Living Conditions Survey (ILCS). The ILCS was funded by the United Nations Development Programme and organised with Norwegian technical assistance. The main purpose of the ILCS was to provide general socioeconomic data. But it also collected birth histories from women in 21000 Iraqi households. When the birth history data began to arrive from the field, the organisers of the survey decided that child deaths were being greatly under-reported. Accordingly, they resolved ‘to re-interview all households again with (a) small questionnaire’ containing the birth history.29 Nevertheless, the U5MRs for 1999–2003 obtained from the reinterviews were still surprisingly low. Indeed, in its report of 2005, even the Working Group of the Independent Inquiry Committee largely dismissed the ILCS results.
The second major survey was the 2006 Multiple Indicator Cluster Survey (MICS).30 This entailed collaboration between Unicef and Iraq’s Ministry of Health, and it collected birth histories from women in 18 000 households. The third survey was the 2011 MICS. This involved a similar collaboration and interviewed women in about 36 000 households.31 It should be noted that the MICS surveys developed by Unicef have been used in over 60 countries to provide key indicators relating to the state of young children.
With this as background, figure 1 compares the U5MRs calculated from the ICMMS of 1999, the ILCS of 2004, and the MICS surveys of 2006 and 2011. It shows that, in contrast to the ICMMS, the U5MRs from the three recent surveys reveal no sign of a huge rise in child mortality after 1990. This is compelling evidence that it simply did not happen. In short, as suggested with notable intuition by the Working Group of the Independent Inquiry Committee in 2005, the ICMMS data were evidently rigged to show a huge and sustained—and largely non-existent—rise in child mortality. The falsification might have occurred during the data entry stage at the behest of the Iraqi government. However, recall that the government provided the field staff for the ICMMS in the centre and south, so it may be that under the strong-arm influence of government authorities the deception happened during the fieldwork stage (as seems to have occurred during the much smaller 1995 FAO survey in Baghdad). The objective of Saddam Hussein’s government was to heighten international concern and so get the economic sanctions ended.
Figure 1Annual estimates of the child mortality rate in Iraq from four retrospective surveys (under-5 deaths per 1000 live births), 1974–2010. ICMMS, Iraq Child and Maternal Mortality Survey; ILCS, Iraq Living Conditions Survey; MICS, Multiple Indicator Cluster Survey.
That is not to say that there was no rise at all in the U5MR in Iraq between 1990 and 1991. All three major surveys conducted since 2003 suggest that there was a slight increase. Moreover, recall the substantial rise indicated by the ICMMS for the period 1988–1992 in the north—reflecting the oppression of the Kurds. In this context, the rise indicated for January to August 1991 by the modest IST survey was also much greater in the north. That said, for Iraq as a whole the IST survey probably understated the level of child mortality that prevailed during 1985–1990 and overstated the level during 1991.
If we ignore the—false—U5MRs for 1991–1998 from the ICMMS in figure 1, and take account of the fact that older women are more likely to under-report child deaths (indicated most clearly in figure 1 by the U5MRs from the ILCS for the 1980s) then a significant degree of agreement actually emerges. The three surveys conducted since 2003 all put the U5MR in Iraq during 1995–2000 in the vicinity of 40 per 1000. In relation to the difficult conditions of the 1990s, it should be noted that Saddam Hussein’s government became increasingly proficient at evading the sanctions and Iraq’s basic food rationing system probably also helped the situation.
Lastly, for 1970–2015 figure 2 plots the UN’s latest U5MR estimates by 5-year periods for Iraq and neighbouring countries.32 The estimates for Iraq are based mainly on the 2004, 2006 and 2011 surveys we have discussed, with consideration of the ICMMS results for before 1990. In the 1970s, there was considerable variation in child mortality, the trend in Iraq broadly paralleling the trends in Syria and Jordan. From the early 1980s onwards, however, progress in Iraq seems to have been much slower. Indeed, by 2010–2015 the U5MR in Iraq was roughly twice that of the other countries. The period of miserable progress in reducing child mortality broadly corresponds to the period of Saddam Hussein. He brought a host of troubles and disasters to his country of which, however evaluated, the economic sanctions constituted a very small part.
Figure 2United Nations estimates of the child mortality rate in Iraq and neighbouring countries by 5-year periods, 1970–2015.