Issue 40 - Article 13

Security for humanitarian organisations in the Kenya crisis

October 9, 2008
Nik Bredholt and Steve Penny, IAWG

This article reflects on experiences in Kenya following the eruption of violence there in the wake of presidential elections on 27 December 2007. The crisis continued for two months until the announcement of an agreement between the two main political parties. However, insecurity persists, arising either directly or indirectly from the initial political crisis. Generally, the post-election crisis was fuelled by ‘old’ grievances that long predate the election, and will not be solved quickly.

The learning points presented here are based on the experiences of the Interagency Working Group (IAWG) Security Subgroup. The group, which was active throughout the crisis, comprises dedicated security officers and more general project staff from a number of international and local NGOs, the Red Cross Movement and representatives of the UN Department of Safety and Security (UNDSS). The IAWG is an innovative structure in the NGO world, bringing together NGOs and UN agencies to work collaboratively to strengthen disaster preparedness in East and Central Africa. Although the IAWG has a regional remit, the Security Subgroup has limited its work to Kenya.

Background

The Security Subgroup started working on a template for a security contingency plan related to possible election violence in November 2007. This work was completed by the beginning of December, when the situation was still peaceful. The plan focused on expatriate staff with regard to relocation, hibernation and evacuation, but considered all staff (national and international) in terms of overall planning. The group started meeting in early January 2008, immediately after the crisis began. In the very tense days that followed meetings were held at least once a week and sometimes more. Some meetings had to take place in cafes or private houses as offices were closed and travel through certain parts of Nairobi was unwise. The following learning points and challenges emerged from this experience.

National staff are more vulnerable than expatriates

Most security plans are written with expatriate staff in mind. The security of national staff is very much a secondary consideration, and yet national staff were the very people most in danger in this crisis. Agencies in the Security Subgroup presented their approaches to issues such as temporary shelter arrangements and extra allowances for national staff, relocation plans and extra time off to accommodate family needs arising from insecurity. This served to establish a common view so that the different agencies took the same line on these delicate matters.

Many agencies preferred to conduct field operations using expatriate rather than national staff, as musungos(Kiswahili for ‘whites’) would not be perceived as implicated in the ethic conflict that was such an important element of the violence, and would generally be less vulnerable to harassment. Depending on their community of origin, national staff already in the field when the violence broke out either remained in place or were relocated.

Hope for the best – prepare for the worst

By setting up a template for an evacuation plan from Kenya to another country, for instance Tanzania, and a policy document for relocation, the group was able to provide all interested agencies with at least a minimum level of preparation. This was particularly important for smaller agencies that did not have dedicated security staff, as they could use the templates to make basic security preparations.

Whilst evacuation seemed an unlikely prospect in November, as the crisis escalated it soon became a very real possibility: some agencies evacuated non-essential staff and dependants, and the preparation of evacuation plans was a sensible and necessary provision. Although the security template prepared before the crisis identified what subsequently unfolded as the worst-case scenario, the impact of the crisis was more serious than expected. Agencies when writing contingency plans rarely act on the worst-worst case scenario.

Establish indicators and keep monitoring them

While drawing up a template for evacuation, it became clear that the group would need to establish and monitor indicators to create a basis on which to take informed, objective decisions. Colour-coded security levels of green, yellow and red were adopted, with the group agreeing on specific indicators that would suggest that the situation was changing from one category to another. Indicators included riots, political and social breakdown, regular and widespread armed conflict, increased checkpoints and security force operations, heightened tension throughout the country and in towns, civilian transport considerably reduced due to security concerns and international staff members restricted to major towns without clearance to move to the field. This was useful for NGOs based in Kenya, as the UNDSS security levels were not uniformly followed by NGOs before the crisis. Obviously, many agencies have their own sets of indicators, but in situations such as this it would make sense if a standardised set of security levels and indicators were used, to avoid confusion.

Involve UNDSS and local Kenyans

Soon after the crisis hit, the UN established cluster coordination for all major sectors. However, there is no cluster for security and it is unlikely that one will be established. Nonetheless, it is advisable to establish a strong relationship between non-UN agencies and UNDSS, ideally prior to any crisis arising. A senior UNDSS official made themselves available for many of the subgroup meetings, providing updates, analysis and advice. This input was extremely helpful. A small but not insignificant success was that, by the end of February, UNDSS had decided to share its sitreps with NGOs, via a list of those agencies that were well-established in Kenya and had been present at the subgroup’s meetings.

The Security Subgroup contained both Kenyan nationals and expatriates. The Kenyans were able to provide much deeper contextual analysis and historical background explaining the crisis in terms of politics, economics and ethnic rivalry, subtleties which expatriate members would not otherwise have fully comprehended.

Setting up an SMS network takes time

The subgroup worked on an SMS system, whereby breaking news could be distributed quickly and efficiently through short messages via the mobile telephone network. While there was agreement about the usefulness of such a system very early on, it took a long time to set it up and get the hardware to work. The system had to be hosted (the International Organisation for Migration (IOM) eventually volunteered), and adequate hardware and software bought and installed. The system was only tested at the end of February, by which time the situation had started to improve. Any such system in the future should have clear standard operating procedures (SOPs) and, information often being sensitive, it should be made available only to a known, defined group of members.

Information-sharing is difficult in a volatile environment

Sharing information about incidents such as attacks on staff and damage to agency offices and vehicles requires a certain level of trust, and establishing this can be a huge challenge when participants keep changing. No two meetings had the same composition. In the end, the group decided that information could be shared between known agencies and in particular a pre-established list of names which would not change from meeting to meeting.

Security in the field varies significantly

The security situation in field locations varied greatly: some provinces were severely affected by violence, while others were not. In some locations, good coordination mechanisms were established, where security issues were discussed and common approaches agreed. One example of this was in western Kenya, where agencies engaged local youth groups to promote peace-building. Elsewhere, however, there was little inter-agency collaboration, and repeated calls to establish the UN cluster mechanism at provincial level were heeded too late or not at all. Agency personnel were often exposed to extreme threats when travelling to distribution sites, especially when aid organisations were seen to be assisting internally displaced people of different origin than the host community. Agencies had to resort to air evacuation of their staff on a number of occasions.

Nairobi: a huge expatriate community

Security contingency plans are often designed for small, contained field locations, for instance where a small team lives together in a compound. Nairobi, however, presents very different challenges. Many agencies have a huge presence, with regional offices and support offices for Somalia and Sudan, and many staff dispersed across the city. The large slums, estates and residential areas spread all over the city mean that Nairobi does not lend itself to easy evacuation route planning. This presents unique challenges in terms of hibernation and evacuation planning, reinforcing the need for adequate plans and preparations prior to any crisis. These plans should not only cover Nairobi and the suburbs, but also the operational environment in Kenya as a whole.

Conclusions

Each crisis is unique, and there will always be shortcomings as we cannot plan for all eventualities. However, it is important that at the least minimal plans and guidelines are prepared in advance, including security indicators, evacuation and hibernation plans and clearly established methods of communication. Such plans should be communicated to and understood by all relevant staff. In fact, while many agencies in ‘Nairobbery’ had plans in place to deal with criminality and terrorist attacks, few had thought about election violence as Kenya, unlike other countries in the region, was generally regarded as politically stable.

Careful attention should be paid to the different risks to national as opposed to expatriate staff. The risks to nationals should not be underestimated and agencies should provide a forum for national staff to express their concerns. Individual agencies should also clearly understand their contractual obligations to staff on different (national, international, consulting) contracts.

Nik Bredholt (nbredholt@cafod.or.ke) is Chair of the IAWG Security Working Group, and Regional Humanitarian Coordinator for CAFOD. Steve Penny (Steve_Penny@wvi.org) is the Vice-Chair of the IAWG and Africa Humanitarian and Emergency Affairs Director for World Vision.

Comments

Comments are available for logged in members only.