How many people have died as a result of Khartoum’s genocidal counter-insurgency campaign in Darfur? What is overall mortality since February 2003? These questions have been much in the news recently, particularly in the wake of a decision by Britain’s Advertising Standards Authority (ASA) that an advertisement by the Save Darfur Coalition and Aegis Trust had inappropriately represented as fact a death toll of 400,000, when this was a matter on which opinions diverged. Notably, the ASA did not find, as erroneously asserted by Sam Dealey in the New York Times (August 12, 2007), that the advertisement “violated codes of objectivity and truthfulness.” Nor is the ASA likely to be the best source for understanding the complexities attending the competing claims of various mortality estimates, ranging from Khartoum’s figure of 9,000 to the figure of well over 450,000 generated by this writer.
Why does any of this matter? Here it’s useful to recall that in February 2004—one year into the most violent and destructive phase of the Darfur genocide—the official UN estimate for total human mortality was 3,000. In retrospect this is of course an absurdly low number, although there certainly was no effort to deceive by the UN. But only activist efforts—not those of professional epidemiologists—succeeded in compelling a closer examination of the data available, which were in fact extremely limited. Activist pressure also helped ensure that subsequently a significantly wider and more authoritative set of data would become available, although not always meeting specific epidemiological standards.
The most controversial data came from a study overseen by the nongovernmental Coalition for International Justice (CIJ) in August 2004—not “summer of 2003,” as erroneously asserted by Conor Foley in this space (August 17, 2007). Nor was it a study of “morbidity” (degrees and extent of illness) as Foley asserts, confusing this basic term with “mortality.” Rather it was a study designed to determine whether genocide had occurred in Darfur. Those conducting the 1,136 carefully randomized interviews among Darfuri refugees at various locations along the Chad/Darfur border were professionals drawn from a wide range of backgrounds, including law enforcement, previous genocide investigations, and human rights work. They had extensive resources, including a full complement of translators.
The overwhelming consensus among the investigators, according to one genocide scholar on the team, was that genocide had been committed and was continuing. This was the basis for the US determination rendered by then-Secretary of State Colin Powell in Congressional testimony of September 2004. But in addition to making clear the genocidal intent of the Khartoum regime, the CIJ study also yielded highly significant data about human mortality in Darfur, if in tantalizingly insufficient form.
The CIJ data is particularly important because of its comprehensiveness as well as its timeliness: humanitarian organizations have reported that through summer of 2004 the overwhelming cause of death in Darfur was violence. At some point, perhaps late summer 2004, the primary causes of death became disease and malnutrition, often directly related to antecedent violence, and thus also genocidal deaths. But violence and its direct effects were the chief causes of death for nearly all the “look-back period” in the CIJ study.
Three studies have attempted to take account of the CIJ data. All suggest that present mortality, from all causes, is 350,000-400,000 or greater. While all have been judged harshly for this use of CIJ data by the US Government Accountability Office (GAO), the basic statistical challenge confronting any mortality assessment remains: without consideration of CIJ data, there is no effective way to calculate violent mortality on a global basis for Darfur. Because the CIJ study was not specifically designed as a mortality study, a range of assumptions must guide use of the data. But unless these assumptions are shown to be unreasonable—something not attempted by the GAO, the ASA, or any other investigator—the choice confronting those who would make a reasonable estimate of total Darfur mortality, including violent mortality, is stark: make use of the comprehensive CIJ data as reasonably as possible—or ignore it. The latter decision may result in greater methodological hygiene; it also ensures that violent mortality will be very dramatically understated.
We should also consider the time-frame for various assessments. The conflict in Darfur has now raged for 54 months; indeed, ethnic violence orchestrated by the Khartoum regime through its Arab militia proxies had claimed thousands of lives before the standard terminus a quo for the conflict, February 2003. No study considered by the ASA or the GAO is temporally inclusive; indeed one study favored by the GAO includes data that reflect only about one third the duration of the conflict.
Here it is important to understand the consequences of the last UN World Health Organization study of global mortality rates (published initially in spring 2005). A senior UN official told this writer at the time, in emphatic terms, that there would be no further global mortality studies done because of severe, sometimes violent harassment by Khartoum. The regime was clearly determined to make global mortality assessments impossible. And even the 2005 UN data and excess morality-rate study excluded most of South Darfur state because of insecurity; yet South Darfur has approximately half the population of Darfur as a whole.
Clearly there can be no certainty about Darfur mortality totals. But, for different reasons, we need both an authoritative lower limit and a credible upper limit. An authoritative “floor figure” for Darfur mortality was provided by a study published in Science (September 2006), one of the most distinguished peer-reviewed journals in the world. Professors John Hagan and Alberto Palloni are authors of the study, which established the currently most commonly cited figure for Darfur mortality, 200,000 dead.
The Hagan/Palloni study excluded CIJ data, but used other, more precise, though less revealing data; this produced a significantly smaller figure for violent mortality than had appeared in Hagan’s earlier co-authored study (spring 2005), which estimated that approximately 400,000 people had died from all causes at that point in the conflict. But the more astringent study in Science—not considered by the GAO—concludes with a significant statement about the range of mortality upwards from the “floor figure” established: “It is likely that the number of deaths for this conflict in Greater Darfur is higher than 200,000 individuals, and it is possible that the death toll is much higher.” Hagan declared to the New York Times (September 15, 2006), “We could easily be talking about 400,000 deaths.”
Using primarily CIJ and UN World Health Organization data, this writer has concluded that, as of April 2006, upwards of 450,000 people had died (http://www.sudanreeves.org/Article102.html). An assessment of this work was offered at the time by the member of the GAO panel most experienced working in Darfur, Francesco Checchi of the London School of Hygiene and Tropical Medicine. Checchi declared of my estimate that it is “‘mathematically correct’ and ‘sufficiently legitimate’ to establish a high-end count.”
Why should we care about credible estimates for either a lower or upper range? Without a solid lower estimate of the sort provided by Hagan and Palloni in their Science article, there was no real corrective to previously common news misreporting of “tens of thousands of deaths in Darfur.” But without a credible upper estimate of human mortality in Darfur we risk seeing a reprise of Rwanda, where mortality was underestimated in ways that worked to sustain international paralysis in the face of a cataclysm of human destruction that claimed some 800,000 lives.
400,000 deaths in Darfur is a fully credible estimate. If not demonstrable fact, it is far more than mere “opinion.”
[Eric Reeves is author of A Long Day’s Dying: Critical Moments in the Darfur Genocide]