When Saviors Become Victims: Trends in Attacks Against Aid Workers and Peacekeepers
Here’s a paradox: declining levels of worldwide conflict and increasing violence against aid workers. The Human Security Report has documented a worldwide decline in the number and intensity of armed conflicts since 1992. This is a dramatic and consistent, though the decline has slowed recently.
Meanwhile, analysis of data by the Overseas Development Institute indicates that attacks on aid workers have risen even sharply, from about 4 per 10,000 aid workers in the field in 1997 to 9 per 10,000 in 2008 (268 incidents in which humanitarian workers were killed, seriously injured or kidnapped). 
(For ongoing discussion of violence against aid workers, also see Patronus Analytical.)
Currently, the worst places for violence against aid workers are Somalia and Afghanistan, which account for a disproportionate number of attacks, and where belligerents have ceased to respect the neutrality of humanitarians. Whether these two extreme cases account for the whole phenomenon, is not clear.
Sudan ranks in third place. It’s now clear that the level of lethal violence against civilians in Darfur is down. But that is not the only index of insecurity: there is also forced displacement, sexual violence and robbery as well””and attacks against aid workers and peacekeepers. Anecdote suggests that international workers in Darfur have never had it so bad, with a new spate of hostage-taking alongside ongoing banditry and government expulsion of NGOs.
In this posting I examine the trends in violence against aid workers and peacekeepers in Darfur, using the UNAMID dataset. Unlike reports of violence against Darfurians, the dataset for attacks on UNAMID, UN agencies and international NGOs is comprehensive and reliable, as incidents are reported through a centralized system. Attacks on Sudanese NGOs and their staff are not included in these data, and some agencies such as the ICRC fall outside this reporting system. Notwithstanding these shortcomings, the trends are clear.
The summary data for incidents in 2008 are as follows:
Carjacking: 195 incidents (50 against UNAMID with 57 vehicles; 52 against UN agencies with 61 vehicles; 93 against international NGOs with 98 vehicles).
Hostage taking : zero.
Armed attacks: 108 (20 against UNAMID, with 4 deaths and 3 injuries; 10 against UN agencies with no casualties, 78 against INGOs with no casualties).
The summary data for 2009 to date are as follows:
Carjacking: 83 incidents (55 against UNAMID, 59 vehicles; 5 against the UN with 5 vehicles; 23 against INGOs with 28 vehicles).
Hostage taking: 6 (2 UNAMID international civilian staff, 4 INGO).
Armed attacks: 26: (22 against UNAMID, with 3 deaths and 15 injuries; 2 against UN agencies and 2 against INGOs, no casualties).
These data suggest that security is indeed improving for peacekeepers and international aid workers. However, all is not so simple.
Violence against UNAMID is at about the same level in 2008 and 2009. In the meantime, the number of UNAMID peacekeepers and staff has more than doubled, and they are much more active in mounting patrols. This could be interpreted to mean that UNAMID members are only half as much at risk now, representing a success. Or it could be interpreted to mean that despite the doubling of UNAMID strength, attacks have not diminished””a failure.
While the trend of carjackings and armed attacks against INGOs is sharply down, this might be partly attributable to the fact that the number of international aid workers dropped dramatically in March this year when the Sudan Government expelled many of them. However, the same trend is also apparent with respect to UN agencies, which were not expelled. What accounts for this decline? Is it better security procedures including increased police efficiency? Is it a bigger UNAMID presence? Is it a different operational pattern””perhaps selectively avoiding the more dangerous areas? Or is it that the general improvement in security is also reducing crime against international agencies?
Hostage taking is a new phenomenon. There have only been a few cases so far, so it is hard to say whether it is a political or a criminal activity, or a bit of both.
Also, restrictions on UNAMID movement and operations have markedly increased between 2008 and 2009, mostly obstruction by the Sudan Government.
Across the border in Chad, 2009 has been a more dangerous year than 2008. A recent report from Abeche suggests that carjacking has become endemic there, with UNHCR losing as many as two vehicles per day. Local conditions, it seems, are everything.
Alex,
Excellent post! Considering this data, is it possible that the rising need for more troops in places like Darfur might be fueled by the rising threats to the troops themselves? In other words, considering Darfur’s current “low-level” of violence (excluding violence towards troops and aid workers), could the UN be sending troops to protect its troops?
Or, is the UN sending more troops in order to protect aid workers, because Darfurians are not getting the humanitarian aid they need?
This year’s UN work plan shows that the mortality rate in Darfur (excluding violence) has now returned to prewar levels (0.3), meaning that the need for more humanitarian aid is not as pressing as it was before the war. (See page 140 of the 2009 UN Work Plan for Sudan)
Therefore, the rationale that the UN should send more troops to protect aid-workers doesn’t add up either. That is, if the overall goal is to decrease the total number of casualties.
Is there a point at which the whole international operation in Darfur becomes more focused on self-maintenance than it does on the problems it originally sought to address? Or is the POTENTIAL for disaster so great that further deployment of troops is necessary?
Alex,
If I may add a few details to your post here. One of the keys to the lower level of attacks against INGOs in Darfur is that they almost all now use old beat up rented vehicles that no one would want. This and the expulsions are the main reasons for the drop in car jacking incidents.
As for the UN they are hardly going on field missions by vehicle at all. Most of the UN Agencies travel by helicopter outside of the main towns. This is the main reason for the drop in UN agency incidents.
I am very leery of the mortality statistics because I don’t think they include the “non-government” controlled areas at all. These statistics as best as I can tell only reflect the situation where humanitarian and UN agencies are operating and thus reporting. The large bulk of areas go unreported and thus the statistics are highly questionable.
The aid community finds it increasingly difficult to “safely” reach the “deep field” locations due to the increasing severity of security incidents aimed at them. This is also a factor in the decreasing number of incidents aimed at humanitarians.
A more accurate statistic (if it were available) would be what percentage of field trips have a security incident. This number may show a small or large increase even though the overall numbers are down.
Hope this helps the discussion
RR
Dear Ricardo
Many thanks for your comment. The idea of using the latest most sophisticated off-road vehicles as a ‘security measure’ has long been absurd and I am sure you are right that the strategy of using old vehicles that won’t last long if they are stolen is a good one. Perhaps the UN will learn that trick in due course.
It would be useful to try to correlate the level of incidents against the INGOs, UN and UNAMID against the types of vehicle they use and their patterns of travelling in rural areas. Then we would be able to get a more reliable picture of whether danger levels are changing, and if so, how.
The UNAMID statistics for violence against the Darfurian population are clearly incomplete, and I suspect have become more so in recent months. However, it is worth pointing out that many of the initial reports of incidents that come in tend to have exaggerated fatality figures, which get cut down when there is a proper investigation. I am less than confident about making claims about recent (i.e. 2009) reductions in fatality levels. But I am confident that (a) the level is far lower than in 2003/04, (b) there have been no sharp rises in the last couple of years, and (c) the pattern of violence that the UNAMID statistics show is approximately correct.