1 Introduction
The exact number of deaths in the Darfur
region due to the conflict will probably never be known. But most certainly,
it is far too many. Estimating mortality in conflicts is a notoriously
difficult exercise, even more so in Darfur where the conditions causing
death are extremely variable. Malnutrition, epidemics and violence occur
sporadically, claiming many lives in some areas and none in others. Recognising
the importance of tracking mortality and estimating deaths, humanitarian
aid agencies working in the region have undertaken mortality surveys among
their beneficiaries at different times to assess the condition of their
status and the severity of the crises. These are based on sound statistical
and epidemiological techniques and provide insights into the varying levels
of mortality over the entire region.
Estimating numbers of deaths from surveys depends on representativeness of the sample, double counting of deaths, under or over-reporting by respondents. Another key concern is that intensity of the conflict varies over time and in different areas of Darfur and therefore a blanket application of rates from a few surveys will invariably distort results.
The humanitarian assistance, although slow in early stages, since the first half of 2004 in Darfur has been massive and is widely acknowledged to have saved many lives. As humanitarian needs continue to grow the situation today is deteriorating again and it is clear to the authors that humanitarian aid has to be increased and important international measures to end the aggression must be taken.
From a majority of deaths being caused by military/violence in the wars in the first half of the 20th century, armed conflicts over the last 20 years have taken their toll among the civil populations. Disease and malnutrition have been the main causes of deaths among civilians in most of the major conflicts of the past two decades. These include deaths due to lack of access to health care, to food or harvests leading to starvation, dehydration and disease during displacement. Direct war-related violence on civilians leading to death (massacres, shootings), while heinous, contributes a small part of the total deaths, but remains the only direct evidence of the blunt hostility of armed groups on unarmed inhabitants.
In this paper we present two alternate methods that were used to calculate estimations of mortality in Darfur. The first one was elaborated by epidemiologists at the Brussels-based Centre for Research on the Epidemiology of Disasters (CRED). The second one was performed by the Bureau of Intelligence and Research of the US Department of State.
In summary, the CRED method estimated approximately 134,000 total deaths in Darfur and Eastern Chad over the 17 months from September 2003 to January 2005. Of these deaths, 120,000 were excess deaths directly attributable to the conflict, 35,000 of which were violent deaths. The US State Department method estimated a possible range of 98,000 - 181,000 total deaths over 23 months - from March 2003 to January 2005. Estimates of excess deaths due to the conflict ranged from 63,000 - 146,000 over the same period.
2 Review of conflict related mortality estimates for Darfur and eastern Chad
The conflict in Darfur has given rise to a wide range of estimates, from fairly precise numbers of dead such as 396,563 persons 1 to more general statements such as the UN Office for the Coordination of Humanitarian Affairs' (OCHA) figure of 180,000.2 Regardless of the levels of mortality reported, we assume that none of these sources may be considered as having generated politically motivated or deliberately biased results to make a point. The validity of estimates instead is dependent on methodological rigour and soundness of the assumptions on which they are based. Lower estimations, if indeed valid, do not necessarily diminish the severity of the humanitarian situation or express callousness to the suffering and death in Darfur.
2.1 Building Blocks
There have been numerous estimates put forth on overall death figures in Darfur. Although originating from various sources, all are based in whole or in large part on the same two sources, with extrapolation to the broader Darfur population and the entire length of the conflict. The first source is a collection of 1,132 interviews conducted with Darfur refugees in eastern Chad by a US State Department/USAID/Coalition for
International Justice (CIJ)/American Bar Association (ABA), hereafter referred to as the Atrocities Documentation Team (ADT), in July/August 2004. The second is a UN World Health Organization (WHO) mortality survey of IDP populations of West and North Darfur in August 2004 with a two month recall period and the subsequent WHO mortality projection covering a seven month period (March -September 2003).
2.1.1 State/USAID/CIJ/ABA Interviews
The Atrocities Documentation Team's project was an important effort to reveal the widespread atrocities occurring in Darfur at a time when awareness of the tragedy and international response to the crisis was still very limited. These interviews, however, were not designed in any way to function as a mortality survey nor was there an overall systematic sampling methodology 3 used that could make it representative of the roughly 200,000 refugees that fled to eastern Chad, much less of the entire 2.4 million people affected of Darfur.
The interviews were semi-structured and qualitative in nature and there was neither a sample universe indicated (population of which it is intended to be representative) nor a defined recall period specified in the questionnaire from which a mortality rate could be accurately calculated. These interviews, and the derived mortality rate, undoubtedly represent the most violent period of the conflict, prior to any real international presence or humanitarian assistance, of a cohort tremendously impacted by violence.
The open-ended and qualitative nature of ADT interviews provides not only a sense of the overall scope of atrocities but also details of their brutality. The inappropriate misuse of these interviews, however, as a proxy for the aggregate Darfur population for the entire conflict (despite the availability of other more reliable data, Table 4) has been a major basis of overestimation of deaths (common in most estimates).
2.1.2 WHO Mortality Survey and WHO Mortality Projection
WHO Mortality Survey
The September 2004 WHO mortality survey of displaced populations, found the CMR for North Darfur to be 1.5/10,000/day and 2.9/10,000/day in West Darfur. Non-displaced affected populations were not included in this survey. Deaths attributed to "injury or violence" represented 21% of deaths in North Darfur and 12% in West Darfur. Due to security problems, the survey was completed in only one camp (Kalma) in South Darfur (3.8/10,000/day) representing 73,658 displaced persons.4
Footnotes:
1 Coalition for International Justice "New Analysis Claims Darfur Deaths Near 400,000: Experts estimate 500 people a day are dying." April 21,2005. http://www.cij.org/pdf/Press_Release_CIJ_Mortality_Study_April_21_2005.pdf
2 "180,000 die from Hunger in Darfur." The Guardian, March 16, 2005. http://www.guardian.co.uk/sudan/story/0,14658,1438471,00.html
3 There was no overall sampling methodology that determined selection of settlement/camps and the number of interviews to be conducted in each location. Instead, non-probability sampling, factors such as access, weather, supplies, etc. determined the location and numbers of interviews conducted. In an effort to reduce selection bias at the camp/settlement level, every 10th household was chosen and then one member was randomly selected for interview.
4 World Health Organization. "Retrospective Mortality Survey Among the Internally Displaced Population Greater Darfur, Sudan. 2004" August, 2004 http://www.who.int/disasters/repo/14656.pdf
(pdf* format - 752 KB)