Darfur’s Trends of Violence: UNAMID’s June 2009 Reporting
The incident statistics collected and verified by UNAMID for June show that harassment of UNAMID and international actors continues unabated. The general pattern of violence continues to manifest the characteristics of a low-intensity conflict, with a low level of confirmed fatalities amidst a generalized situation of insecurity.
Sixteen fatalities have been recorded in June. Twelve of these incidents are probably criminal in nature, the only association with the overall Darfur crisis being that the crimes sometimes occur in or around IDP camps. Three killings were reported by the residents of Kalma and subsequently confirmed, including one sub-sheikh in the camp. One further incident was the death of a boy who picked up an unexploded grenade which detonated, killing him. The majority of the violent deaths in Darfur during the month of June 2009 could hardly be linked directly to the ongoing conflict and are more characteristic of common crimes. The killings of one GoS National Intelligence officer, one SAF and one policeman are however more open to speculation, as they could have some connection with the conflict.
There is also one reported incident of homicide in the non-government held areas. As investigations are completed the data will be updated.
This is the lowest monthly tally of violent fatalities on record. Reasons for the continuing decline in lethal violence include the current lull in offensive military activities by the belligerent parties, the reduction in inter-tribal conflict consequent on stepped up local peacemaking activities, and the increasing effectiveness of UNAMID deployments throughout the three Darfur states. It is no coincidence that IDP populations are expressing better confidence toward UNAMID, thus making the interaction with them more productive.
Six cases of rape were reported, confirming that female IDPs remain vulnerable to harassment in and around the camps.
Eight vehicles were stolen at gunpoint, of which three belong to UNAMID, one to WHO, and four to NGOs.
Al Salaam remains a contentious IDP camp, with the conflicting presence of PDF and SLM-Minawi members likely to continue to fuel tension there.
Restrictions continue to be imposed on UNAMID movements and actions, by different armed elements. On numerous occasions these obstructions are carried out in intimidating manner, and might be incited by what is perceived as a lack of consistent reaction from UNAMID side. Those restrictions range from preventing UNAMID from holding meetings/workshops and confiscating UNAMID funds, to delaying/denying access at checkpoints. Examples include: Wada’a, North Darfur, 4 June; Babun-Shagil, North Darfur, 21 June; Kutum, North Darfur, 24 June; Kuma, North Darfur, 28 June; Abou Dam, South Darfur, 29 June.
In the same vein, banditry against UNAMID seems to be a low-risk operation for the perpetrators. The attack on a UNAMID Formed Police Unit convoy in the vicinity of El Geneina camp is the latest example. The current deployment phase of a Chapter VII operation has never been meant to be a context whereby UN assets and personnel would be consistently targeted while wrongdoers are able to roam unhindered. UNAMID is working on improving its response to criminal attacks of this nature. The image and credibility of the Mission will be improved by a prompt and vigorous reaction to such situations.
Moudjib Djinadou is head, UNAMID Joint Mission Analysis Centre, al Fashir.
As reported, these figures fall squarely in the middle range of the homicide rate to be expected in a society not experiencing armed conflict. Taking the total Darfur population as 7.2 million this is equivalent to an annualized homicide rate of 2.83 per 100,000 per year. By way of comparison, the homicide rate in the U.S. in 2006 was 5.95, i.e. approximately twice that level. Homicide rates in the Caribbean, Andean Republics, and South Africa are markedly higher while those in western and northern Europe are lower. In order to make a full assessment of these figures, we need additional information about (i) the reporting and investigation methods (and in particular what criteria are including a reported incident in the dataset) and (ii) (noting that reporting is always likely to be incomplete in sub-Saharan Africa) the medium-term trends and profiles of the incidents.
It is well known that the incidence of homicide during armed conflicts
such as Darfur oscillates considerably owing to seasonal and political
factors. It is therefore risky to develop a per annum fixed homicide rate on
the basis of a data generated for a single month. This is because a
point estimate can be unrepresentative and give a misleading
impression of wider patterns and trends. Nevertheless, six monthly
estimates of violent deaths reported by UNAMID between January and
June 2009 report some 341 “civilian” fatalities and 217 “combatant”
fatalities (acknowledging a blurred distinction between the two).
Assuming that these figures capture the majority of violent deaths,
and discounting the high probability of under-reporting of incidents,
in a population of approx. 7.4 million, the civilian homicide rate for
the year to date amounts to 9.22 per 100,000.
A review of violence-related data featured in the 2008 Global Burden
of Armed Violence (http://www.genevadeclaration.org) can put Darfur’s “homicide rate” in an international perspective. The
report observes that each year at least five times more people are
violently killed in non-war contexts due to homicide (+/- 490,000)
than war-affected societies (+/- 52,000). Drawing on a comprehensive
national data set, the global average homicide rate is estimated to be
7.6 per 100,000. Comparatively low rates (below this baseline) were
reported in North America, Western Europe and East Asia and a
dramatically higher rates were found in countries throughout Latin
America, the Caribbean and Southern Africa.
It is important to stress that the data on which these global and
regional estimates are based are preliminary, rendered on annual
estimates, and nationally (rather than sub-regionally) based. The
Global Burden of Armed Violence will be re-examining national homicide
rates in future editions. Nevertheless, taken together, they suggest
that while Darfur is currently experiencing a homicide rate moderately
higher than the global average, it is considerably lower than the
sub-Saharan average and also lower than the Caribbean, Central and
Latin America. While making direct comparisons is of course tricky, it
is interesting to note that Darfur’s homicide rate is only slightly
more than most recent estimates of the US annual rate of 5 and 6 per
100,000. (See this link for comparative data).
Robert Muggah is research director of the Small Arms Survey (Switzerland), an associate of the Centre for Conflict, Development and Peacebuilding (Switzerland) and a Principle of the SecDev Group
(Canada).
Dear Robert,
Your point is taken. A better comparison than the US would be east Africa which has a rate of 20.8 per 100,000 per year. Even if the UNAMID-reported numbers represent just 50% of the true toll this only brings Darfur up to the east African average. The southern African homicide rate at 30 per 100,000 is higher still.
Searching for other descriptors of the Darfur conflict I stumbled across the Uppsala Conflict Database, which defines the conflict intensity in 2007 and 2008 as ‘minor’. See: http://www.pcr.uu.se/database/conflictSummary.php?bcID=233.
Dear all,
Yes, the violence in Darfur is less than in 2003-4, but to compare it to civilian rates elsewhere seems to miss the point of the UNAMID data. For one, this is only one month.
Which civilian society would have the Chadian army bombing in Sudan (no fatalities according to the Sudan air force in July 2009) or in early June according to JEM 2 people were killed by the Sudanese air force in Furawiya. We need full access to the UNAMID data to find out what they need to confirm a death (two independent sources; own verification?). The UNAMID dat give reliable (confirmed) amount of deaths but not necessarily the most accurate (also including some probable, but not fully confirmed deaths).
Alex de Waal recently wrote that the deaths people mentioned he spoke to in Darfur were often included in the UNAMID database.
However, are deaths in remote placec like Furawiya and rebel held territory like the Jebel Marra included in UNAMID’s data? We don’t know.
In June 2009 12 JEM rebels were sentenced to death in Khartoum (if executed, also related but not counted).
Violent deaths in Eastern Chad will not be included (not UNAMID’s mandate) but still are related to Darfur’s conflict.
So, let’s not establish tenuous violent death rates now for Darfur without the full data and its caveats. Lack of knowledge, reporting and interpretation of the caveats led to the previous disagreements in counting Darfur´s deaths. Let´s not repeat that mistake. Just report the UNAMID´s deaths, but include the caveats.
The bottom line is clear: the conflict in Darfur is low-level now. But it still is a (sometimes violent) conflict looking for a solution.
In total 108 alleged JEM prisoners are imprisoned in Koober, and all of them have been sentenced to death. They have appealed and the process of appeal and no death sentences can be carried out until that process is complete. The prisoners are also the number one priority agenda for the Doha negotiations (from what we hear almost the only agenda item) and it would be as foolish the the Sudan govt to execute the prisoners as it would be to release them.
This is an extremely useful debate. I am hoping to have some cross-checked statistics and trend indicators to post in the next month or so.
In the meantime, several points.
First: as Moudjib notes, these data are for confirmed fatalities. In the last three months, UNAMID JMAC has increased the rigour of its reporting requirements which has had the effect of cutting out some of the reported but unverified fatalities and so lowering the figures. I have no doubt that some of the unverified incidents are real and the figures will increase. On the other hand, the unverified reports are of small numbers, usually single individuals, indicating that the possibility of invisible massacres is small.
Second: as Rob Muggah notes, the civilian-combatant distinction is an imprecise one. Some individuals are clearly one or the other, but those killed in inter-tribal fighting (classified as civilians) or soldiers killed while off-duty in the market (classified as combatants) are in a grey zone.
Also, it is extremely hard to decide whether an act of violence should be classified as political or criminal. Part of the reason for this is lack of accurate or complete information. But the distinction is also arbitrary, as a great deal of the violence is both criminal and political at the same time, or exists in a kind of hazy zone between the two. The comparisons with places such as Colombia and South Africa are interesting in part because they suffer high homicide rates on the fringes of organized war or after an insurgency has formally been ended, illustrating that war cannot be defined simply in terms of numbers of people killed, while neither is there a simple threshold of fatalities which can define a war–‘minor’ or otherwise.
One of the purposes of the statistical collation and analysis is to try to identify and model what is driving the whole spectrum of violent incidents, which would in turn allow for a better understanding of when, where and why violence flares up, and when, where and why it subsides. I am encouraged by the growing ability of UNAMID to undertake this analysis, which may in turn be one reason why, as Moudjib says, there is growing confidence between UNAMID and the Darfurian civilian population, and which could also be a contributory reason for the declining levels of lethal violence.
One broader question at issue here is whether UN peacekeeping missions will be able to base their operational strategies on this kind of systematic evidence gathered and analyzed in real time. For a police service in a developed country it would be entirely normal and standard to collect and process these kind of data on a routine basis and organize force deployment accordingly. For a military force engaged in any kind of large-scale operation such as a counterinsurgency this would also be standard operating procedure. If UNAMID JMAC is initiating this process in Darfur it is an encouraging precedent. Could someone enlighten me whether such analysis and reporting is routinely undertaken by the Force Commander and SRSG and provided to the UN Security Council?
Dear Gregory,
In April, the head of UNAMID, Rodolphe Adada presented data on violent incidents to the UN Security Council and created controversy by describing Darfur as a ‘low intensity conflict’ with declining levels of violence. He was roundly criticized by some activists and some members of the Council, but stood his ground, based on the statistics.
Last week the UNSC renewed UNAMID’s mandate and requested UNAMID to submit a strategic plan with targets and benchmarks. It didn’t make any specific reference to these kind of data on lethal violence, but it would be a logical inference for UNAMID to take this as one of its key benchmarks.
Besides the military troops, UNAMID also has a police component, which, by the way, is the principal apparatus that maintains permanent connection with the local populations; patrolling EVERYDAY the IDP camps and communities, running Community police centres, and gradually establishing a 24/7 presence in IDP camps.
With a presence covering the territories of all of the three states, 52 team sites -Military and Police-, daily confidence building and escort patrols, night patrols, frequent long range patrols, UNAMID is uniquely placed, to witness and verify, on behalf of the international community, the day to day incidents in Darfur. Certainly, a few incidents would always occur without UNAMID’s knowledge or confirmation; and the Jebel Marra comment might be true, although vigorous efforts have been made recently by UNAMID leadership, allowing improved interaction with those present in the area.
CRED, the organisation that has done the most analysis of all available mortality surveys in Darfur has just released more data, extending the period it covers from september 2003 (missing a fair bit of violence in early and summer 2003) to december 2008 in its newsletter available at http://www.cedat.org/sites/default/files/newsletter-july2009.pdf
It concludes:
Overall, we estimate the excess mortality in Darfur to be
approximately 300,000 (180,000-460,000) deaths from September
2003 to December 2008. The main cause of death
is not violence but rather disease such as diarrhea, which
accounted for 80% of the excess deaths. We have further
identified significant differences between IDPs and residents.
Displacement was correlated with higher diarrhea
related mortality but with lower violence-related mortality.
This finding suggests that IDP settings or camps are
protected from attacks while the displaced are living in
very vulnerable conditions and overcrowded settings where
the risk of disease transmission is higher.
If UNAMID (still predominantly (and most effective) in and around IDP camps) starts to use violent mortality as a benchmark we should be aware in advance that “success” defined as low violent mortality now still entrenches IDP camps, an expensive peacekeeping force and humanitarian assistance, all unlikely sustainable beyond more than ? 3-5 years. Helen Young’s work on the IDP camps, their livelihood strategies and the urbanisation of Darfur do give us an idea what will happen to the camps in the future.
So, maybe the 300,000 figure is not too bad to use in general media reporting (with updated range of dates and maybe also the range (180,000-460,000) and the UNAMID data will be good to keep an eye on violent mortality (and as Alex points out try to understand which is political; one murder of an IDP camp leader in Abu Shouk this week probably is)
So, the debate and understanding of mortality in Darfur is moving on, rather than continuing in acrominy. The bottom line for Darfur remains: a political solution is what is needed.
To the non-demographer, the range of estimates for excess mortality in Darfur may be difficult to understand and prone to misinterpretation. The range of figures, 180,000-460,000 cited by Dr Coebergh does not indicate that there has been a headcount and possible undercount but instead reflect the estimation methods employed. Two factors are especially important to determine the scope of the credible estimate: (i) the method used to derive the current death rate and (ii) the estimate for the baseline. According to the data obtainable for Darfur, both of these have big question marks hanging over them.
The method for estimating current mortality is in most cases an emergency-type survey by a relief organization. There are many ways of conducting such surveys but the bottom line is that these are rule-of-thumb estimates for whether a population is experiencing an emergency, or not. Typically the figure is expressed as a rate per 100,000 per day. The figure is not adjusted for demographic structure and is measured against a threshold for an “emergencyâ€. The emergency threshold is usually put high enough to screen out any fluctuations in Crude Death Rate that could be accounted for by, for instance, a change in population structure which increases the number of under-fives (who have higher age-specific death rates) relative to adults (who have lower). Among other things this could explain why camp populations have higher disease-related mortality than non-camp populations. In order to derive a reliable estimate for excess deaths, the CDR needs to be normalized against a standardized population structure (population pyramid). This does not appear to have been tried.
A baseline normalized CDR estimate is essential as without it there can be no estimate of excess deaths. For a population without vital registration and subject to demographic instability it is extraordinarily difficult to derive a reliable baseline and becomes more so as the time since the initial demographic insult extends. It is interesting that Polloni and Hagan, in their article in Nature do not even try to do so: that is the simplest approach, but the least satisfactory. The difficulty that arises is that a small adjustment of the estimated baseline can lead to huge variations in the estimated excess deaths, when summed for a large population over several years. For example, an adjustment in the baseline CDR from 12/1000/year to 11/1000/year for a population of 7 million over 5 years amounts to a difference of 35,000 deaths.
As time passes from the initial demographic insult the question of attributing causality becomes ever-more problematic. The second Congo war broke out in 1998 and 11 years on the CMR is still elevated. A CDR elevated by just 1/1000/yr for 70 million population for 11 years people equals 770,000 “excess†deaths. This is well within the range of variation that could be attributed to factors such as socio-economic change, increasing birth rate, etc., and it becomes difficult to make a causal attribution to “warâ€. In the case of Darfur the changes in settlement patterns will, in and of themselves, have impacts on vital rates. Nomadic populations have lower birth rates than settled populations and so tend to have lower CDRs though also lower life expectancy for adults.
On the other hand an argument can also be constructed that in the absence of war, the demographic transition and other factors (improved health care) were reducing general population mortality and so even a stabilization of CMR represents excess mortality over what would otherwise have been expected. To which must be added the phenomenon, common to famines, of delayed or cancelled births, a number which typically rivals the number of excess deaths. By the time we are attributing responsibility for unborn children (in a still-growing population) the exercise becomes somewhat hypothetical.
When a clearly-identifiable short-term demographic insult is over, it is analytically more meaningful to speak of demographic changes including alterations in vital rates, rather than continuing to try to calculate an increasingly elusive figure for excess deaths or mortality attributable to the crisis.
I agree with Gregory McDonald that analytically it is more meaningful to speak of demographic changes including alterations in vital rates, rather than continuing to try to calculate an increasingly elusive figure for excess deaths or mortality attributable to the crisis. However no average (short)media report can do this, nor would its readers/listeners/viewers easily appreciate it, so the elusive figure is still desired by those big and important groups.
Gregory McDonald is entirely right that baseline mortality measurements in Darfur are difficult to obtain and initial calculations used different ones, from 0.3/10.000/day from a UNICEF/SCF survey before the war or the standard sub-saharan UNICEF 0.5/10.000 day. How applicable this was to Darfur is unclear.
Hagan and Polloni published in science not nature. The rate is normally expressed per 10.000/day not 100.000/day.
The analysis works both ways. Gregory McDonald argues that in a non-war situation improved healthcare would help reduce mortality. One can also argue (without evidence) that more Darfuri’s have access to healthcare now than before the war. There are certainly many many more doctors (and nurses) in Darfur now than before.
The fact that many (including Mamdani) still write that it is the numbers that make it a genocide or not shows they are important. It is of course not, looking at the genocide convictions for crimes in Srebrenica, when the legal jurisprudence concluded that a substantial part of the population needs to have been killed, but not all (of Srebrenica, not the whole Bosnian Muslim population as the defence tried to argue). What a substantial part is remains unclear (in that case 7.000 killed of a pre-existing 20-40.000 people in Srebrenica).
For demographic analysis of the war in Darfur we need to move beyond the absolute numbers, but numbers are still wanted by many and without engagement and caveats the numbers risk running away.
Thank you, Dr Coebergh for the corrections (the perils of writing away from my files). I also consider myself a novice in disaster epidemiology and would hesitate to write on the topic without having actually been in the field and conducted such surveys in person. This has been an unexpectedly fruitful exchange for me.