Literature review on Iraq mortality

Two main categories of relevant research emerged with regards to the Iraq context: statistical data and analysis, including censuses, surveys, and other reports; and other public health research, especially deep analysis into specific topics such as sanctions, malnutrition, genocide, or political dysfunction.

Statistical reports were the hardest to come by, although they were the most relevant to this report. The Annual Abstract of Statistics for 2012 published by the Central Statistical Organization (2012) was the primary resource for this report, and even that required complex processes to access the data through the broken links on the Central Statistical Organization (CSO) website.

Other public health research was informative in identifying what issues were given the most coverage (and which were overlooked). Also, it provided a framework to build upon regarding the many complex issues that make demographic analysis of Iraq particularly challenging.

Statistical Reports

Beyond the Annual Abstract of Statistics, the resources available were sometimes sparse or incomplete. “Iraq Mortality Study Authors Release Data, but Only to Some,” read a slightly spiteful headline (Kaiser, 2007). Thankfully, some researchers have done impressive work. Fayyad (2012) focused on fertility (as opposed to mortality) but provided a compelling example of work using similar data sets, demographic approaches, and decompositions according to geographic divisions and social groupings. Also, Fayyad created a thorough overview of existing literature with few or no parallels in existing literature.

Fayyad conducted a fascinating study of the Demographic Transition, beginning with broad and widely-accepted definitions of the main phases, followed by an analysis of how this phenomenon has taken place in Arab societies. “When discussing the applicability of the theory of demographic transition to developing countries … where the theory is applied to Arab states, three main issues emerge,” explained Fayyad (p. 7). The first issue, described below, is the most compelling,
The first is that it is very clear that the patterns of socio-economic and political transformations in developing countries during the second half of the 20th century, including Arab states, were often different from those taking place in advanced Western countries. For instance, great differences existed in terms of the role of women in economic and political life, the level of industrialization, the sources of government income, the average income, and the political structure. (Fayyad, 2012, p. 7)

The second and third issues mentioned by Fayyad refer to how Arab societies are unique in contrast to the eurocentric baseline in the framework of the Demographic Transition. For example, Fayyad interrogated whether the conditions necessary for the Demographic Transition have precipitated due to internal development or external influence, and whether these conditions have taken root selectively or have been transformational at all levels of the society. This sort of dissection of what is happening in Iraq was reflected in the regression modeling conducted for this report, where categorical variables along sectarian lines were examined for their statistical significance and led to quite powerful results.

Notably, Fayyad focused on national reporting, especially from the Ministry of Planning for the central Iraq government, and omitted the Kurdistan Region (p. 2). This seemed to reflect a pattern where researchers focused on the fifteen federal governorates, or the four semi-autonomous governorates, but seldom both. This is not necessarily prejudice: the two administrations have separate statistical offices with separate approaches. Fayyad’s firm decisions and, subsequently, his development of frameworks for interpretation which are localized to Arab, Muslim, and specifically Iraq contexts, was a powerful achievement upon which to build.

Another remarkable report was by Cetorelli (2014), who also used, if not architected, compelling arguments around fertility in Iraq. “The number of studies assessing fertility changes during war in low- and middle-income countries is scant,” described Cetorelli, alluding to the struggle for data from a country in nonstop wars for decades (p. 582). Nonetheless, Cetorelli continued with a powerful analysis of impacts on fertility from education and age grouping, using age-specific fertility rates but alluding to the type of indicator and response relationship used in a regression model. A major takeaway from Cetorelli’s work was that meaningful statistical investigations using standard techniques may be possible in conflict situations, albeit with limited explanatory power.

Ali (2003) also conducted a thorough investigation on data (and on mortality, no less) but the time period of 1991 to 1998 was from several decades earlier; anyhow, this was a useful example to study when conducting research in the 21st century.

Other Public Health Research

There are major issues in public health in Iraq which may not be as obvious to demographers elsewhere around the world. As such, these issues should be reviewed as a baseline of knowledge, even if sophisticated statistical or even just theoretical links are beyond the scope of this report. For example, a significant number of doctors have fled Iraq for economic reasons in the 1990s and security reasons after 2003; perhaps, emigration may be unevenly distributed across the governorates of Iraq, and lead to problems with healthcare in some governorates (Al-Khalisi, 2013, p. 363). This is important to note, although further investigation would be an entire report in itself. According to Lal (2006), “there were 34,000 physicians in Iraq before the invasion of 2003, and already by mid-2004 some 2,000 had been murdered while another 12,000 had emigrated” (p. 5028).

It is possible that the impact of conflict, including conflict-related deaths and displacements, may rest more heavily on certain provinces and demographic groups — but many researchers prefer a country-level analysis. According to Abu-Sittah (2019), “more than a decade after the American invasion of Iraq, health and medical services in the country remain in tatters” (p. 137), although this focus on the public healthcare system overlooks that the private healthcare system and medical tourism have both skyrocketed by 2019 (and have become more accessible) according to my own anecdotal but lived experience in the country.

Not only is granular information sometimes hard to come by, but even understanding what is happening is hindered because the corresponding theoretical principles and frameworks are weakly academized. According to Harding and Libal (2019), “public health is directly shaped by war, conflict, and capitalism, yet exploring the connections between these processes remains neglected in scholarship and policymaking arenas,” (p. 111). Such a catastrophic situation may make it hard to extract a regression model that is both statistically significant and has enough of a foundation to be meaningfully interpreted. Furthermore, the conflicts added variables that are unique to Iraq, such as proximity to chemical weapons sites (Sim, 2009, pp. 725-726). On the other hand, perhaps the role of the conflict may be trackable using a key indicator such as extrajudicial killings. Taback (2007) explored killings of journalists. Reading this, it appears possible that an area’s exposure to conflict and dysfunction may be measurable through something like incidence rates of extrajudicial killings, and a continuous variable such as that may have a statistically significant relationship to overall mortality.

At the same time, the situation in Iraq is changing so frequently that models and analysis have limited predictive power. For example, Cordesman (2014) analyzed mounting risks and civil unrest, just months before Iraq largely fell to the Islamic State. Many other changes have also happened in stunningly quick periods. Referring to food insecurity, Field (1993) wrote, “victimization occurred in a time frame so short as to be riveting” (p. 185), but this phrase could fit into reports on many other catastrophes as well in Iraq.

Unreliable Data

Some scholars have flagged that Iraqi data may be unreliable, especially due to political goals. Dennis Halliday, the UN Humanitarian Coordinator in Iraq from 1997 to 1998, remarked, "I had been instructed to implement a policy that satisfies the definition of genocide, a deliberate policy that has effectively killed well over a million individuals, children and adults" (Fabian, 2002, p. 1). But his claim may have been based on unreliable information. According to Garfield (2002), "claims of a million-plus deaths of Iraqi due to economic sanctions are based on Iraqi ministry of Health information provided since 1993 on the number of deaths recorded in hospitals" (p. 16). "Among children under five years of age, all deaths due to respiratory infections, diarrhea, gastroenteritis and malnutrition are counted as deaths caused by sanctions. Among those other five years of age, all deaths due to cardiac diseases, hypertension, diabetes mellitus, renal diseases, liver diseases and cancers are counted as caused by sanctions" (p. 16). Put simply: "Not all the deaths from these causes are related to sanctions" (p. 16).

Another major statistical issue was conflict-related deaths after 2003, where the United States possibly wanted to have the most conservative and limited counts possible. Lal (2006) noted that there were two credible reports that provided an estimated total number of Iraqi fatalities due to conflict in the post-invasion period: Baghdad’s Mustansiriyah University estimated 654,965, while Iraq Body Count estimated 49,697 (p. 5028). “The human thirst for truth cannot be satisfied by both numbers,” wrote Lal about these two figures (p. 5028).

Sometimes, it can be hard to tell whether significant changes are due to killings, emigration, or chicanery. According to Barringer (2003), “In 1987 census takers recorded about 106 men for every 100 women; a decade later they found just under 99 men for every 100 women, a very big change in demographic terms. … Demographers at the United Nations and the United States Census Bureau say that perhaps the men were casualties of war or Mr. Hussein's repression, or became exiles. Or perhaps the men simply made themselves scarce at census time” (paragraphs 3-5).

Clearly, data that is used for political messages should not be presumed to be entirely accurate. There should be an interrogation of its validity. This is not to say that questionable data is totally without value. After all, deaths with misattributed causes may still be useful for some purposes. The uncertainty surrounds the cause of death, not whether there was any death whatsoever. (So, in fact, more thorough reporting of deaths occurring may even have some advantages for certain types of analysis.) Overall, it is clear that data should be questioned with an intense lens — do any major stakeholders benefit from a certain outcome? are there blind spots in the context of a failed state? — as part of diligently pursuing truth.