Paying the ultimate price: an analysis of aid-worker fatalities
by Dennis King October 2002

Increasing numbers of aid workers are losing their lives in the line of duty. Yet while this trend is of increasing concern to the aid community, much of the evidence is anecdotal, and there is little hard data on which to base policy.

Statistics and documentation concerning fatalities among humanitarian workers are woefully incomplete. In addition to the efforts of individual researchers, the Office of the UN Security Coordinator keeps records and compiles statistics on casualties among UN civilian staff, and the UN Department of Peacekeeping Operations tallies fatalities among UN peacekeepers, military observers and other personnel on UN peacekeeping missions. Likewise, the International Federation of Red Cross and Red Crescent Societies and International Committee of the Red Cross maintain data on security incidents involving Red Cross delegates, volunteers and local staff. However, to analyse the trends and assess the security threats faced by aid workers, we need to track incidents involving all organisations providing some type of humanitarian assistance, not just UN agencies.

Chronology of incidents

The analysis in this article is derived primarily from reports in the ReliefWeb document database for the years 1997–2001. ReliefWeb collects and publishes documents from over 600 sources. Data follows the standard promoted by the Structured Humanitarian Assistance Reporting (SHARE) approach, under which data is presented with its source, the date of the information and the geographic area of the incident reported. The research resulted in the compilation of a detailed chronology of fatalities, which is on the ReliefWeb website.

To be included in the chronology, incidents had to adhere to a number of criteria:

  • The incident took place in a country during an emergency or post-emergency transition period.
  • Death was the result of intentional violence or a work-related transport accident in an aircraft or road vehicle.
  • Recorded deaths were limited to individuals working for civilian humanitarian-related organisations, both local and expatriate, and all passengers/crew on a humanitarian assistance mission, for instance drivers, security personnel and pilots. Incidents involving peacekeeping personnel were included only if such personnel were killed while providing security assistance for a humanitarian mission or convoy.
  • Deaths resulting from natural causes or suicide were not included.

Screen Shot 2012-11-05 at 12.02.34 PM


An analysis of the information collected in the chronology indicates that more civilian humanitarian aid workers were killed by acts of violence than died in vehicle and aircraft accidents. Almost half (47%) of the non-accidental deaths of aid workers were the result of ambushes on vehicles or convoys, carried out by bandits or rebel groups.

The research also produced a list of the ten countries where the greatest number of non-accidental deaths occurred. Unsurprisingly, these countries are in conflict, or have recently been so.

Among the incidents of intentional violence, 74% of fatalities were local staff, and 26% expatriate. More than half (59%) of these victims worked for or on behalf of non-governmental organisations, while 41% were employed or under contract to UN agencies. The number of local and/or NGO fatalities is probably higher, since these incidents are less likely to be reported in public sources than the deaths of UN and/or expatriate personnel.

Screen Shot 2012-11-05 at 12.02.46 PM

Enhancing the security of humanitarian aid workers

The data demonstrates that the dangers faced by civilian humanitarian aid workers are serious. The provision of better protection and security training is being addressed by the UN, as well as by many NGOs and NGO consortia. UN agencies and international organisations have improved staff security monitoring and response mechanisms, and have incorporated security awareness into their training programmes. Agencies, both UN and other, are employing more trained ‘security officers’. Organisations such as InterAction, VOICE, RedR and the International Rescue Committee have developed training programmes and security coordination systems.

Reports in 1999 and 2001 from the UN Secretary-General to the General Assembly on the safety and security of UN personnel have recommended new procedures and coordination structures. The Secretary-General has proposed new measures to improve security, including appointing a Secretary Coordinator at the Assistant-Secretary-General level, allocating new resources for hiring field and headquarters security officers and more training and counselling for general staff. New ways have also been proposed to ensure consistent funding for the UN Security Coordinator. However, these mechanisms, such as the Trust Fund for the Security of UN Personnel and special staff security programmes included in the annual UN Consolidated Emergency Appeals, are seriously under-funded.

International humanitarian law has also sought to address the security of humanitarian personnel. The Rome Statute of the International Criminal Court makes the murder of humanitarian personnel a war crime, but it is not yet in force and very few cases have been prosecuted at national level. In 1999, the UN enacted the 1994 Convention on the Safety of United Nations and Associated Personnel. However, the convention is not applicable to humanitarian NGOs that do not have implementing/partnership agreements with the UN and its specialised agencies; nor does it apply to locally-recruited personnel. In the March 2001 issue of Humanitarian Exchange, Randolph Martin, Senior Operations Director at the International Rescue Committee, looked at ways of strengthening the 1996 Memorandum of Understanding between the UN Security Coordinator and NGOs to provide a framework for a security relationship between UN organisations and their NGO implementing partners. One way of improving this relationship is to develop a system for tracking all security incidents and sharing this information among all the actors concerned.

Data and policy-making

The research on which this article is based was conducted to demonstrate how systematic, consolidated data collection and reporting can be used for analysis. The analysis provides representative statistics to determine how many humanitarian aid workers have been killed, where these incidents have taken place, and the causes of these fatalities. This information can then form the basis for policy-making and improved security awareness and procedures.

Dennis King is Deputy Director of the US government’s newly-established Humanitarian Information Unit. He has worked on several humanitarian information initiatives, including the UN’s ReliefWeb project and the UNICEF Emergency Operations Centre, and was coordinator for the Symposium on Best Practices in Humanitarian Information, which took place in Geneva in February 2002. His email address is:

References and further reading

Dennis King and Maxx Dilley, ‘How To Share Information in a Complex Emergency’, Humanitarian Affairs Review, Summer 2001.

Randolph Martin, ‘A More Proactive UN Role in the Security of NGO Staff’, Humanitarian Exchange 18, March 2001.