Protection of Civilians (POC) sites and their impact on the broader protection environment in South Sudan

In the days following the outbreak of violence in South Sudan on 15 December 2013, tens of thousands of people fled to UN peacekeeping bases across the country seeking protection. While the trend of civilians sheltering in UN bases is by no means unprecedented, what evolved over the next three years was unlike anything humanitarians had experienced previously: the emergence of a new type of protracted displacement site that blends the UN ‘safe areas’ of the Balkans in the 1990s with the traditional camps found elsewhere in Africa.

These so-called Protection of Civilians (POC) sites came to define the response in South Sudan. With the total population of the sites peaking at over 200,000 in September 2015, protection needs were not only tremendous, but the people taking refuge in the sites were also highly visible and accessible. As a result, much of the humanitarian and peacekeeping effort focused on the POC sites, despite them housing less than 10% of the total displaced population. This imbalance sparked debates about the appropriateness of the response and whether the attention to the POC sites came at the expense of meeting needs elsewhere.

In April 2015, the UN Department of Peacekeeping Operations made it a global policy that UN missions with a protection of civilians mandate must be prepared to open their gates to civilians in extreme situations of violence. The inclusion of this language, and the precedent set by the experience in South Sudan, make it likely that the POC site model could emerge in other situations of conflict in the future. With this in mind, it is important to take stock of the lessons from South Sudan, and ensure that they are applied going forward.

The impact of the POC sites on protection elsewhere in South Sudan

Ever since it became clear that the POC sites were not a short-term feature of the South Sudan landscape, the UN Mission in South Sudan (UNMISS)’s approach to protection of civilians has fundamentally changed. Rather than analysing how civilians could be protected both inside and outside the sites, attention appears to have focused disproportionately on the POC sites alone. Some UNMISS personnel have even indicated to visiting researchers See CIVIC’s October 2015 report, Within and Beyond the Gates: The Protection of Civilians by the UN Mission in South Sudan. that they believe protecting the POC sites constitutes the entirety of the Mission’s responsibilities – despite UNMISS having a Chapter VII mandate which covers the whole country.

One of the biggest challenges has come from the fact that UNMISS has insisted that it is unable to deploy additional troops to other hotspot areas due to the number of troops deployed to protect the POCs. While many humanitarians disagree with this assessment (there have been repeated requests for a critical analysis of current troop deployment locations), the existence of the POCs has certainly been used by UNMISS as a pretext not to deploy elsewhere. As a result, civilians in areas such as southern Unity State have been left on their own to try to escape massacres and horrific acts of violence. Between May and October 2015, the Protection Cluster received reports of over 1,200 deaths in three counties alone, during which time there was no static UNMISS presence in the area. As is always the case, the government of South Sudan has the primary responsibility to protect civilians in South Sudan, but where it is unable or unwilling to do so (for example, in the context of southern Unity State), the mission is assigned with an additional protection role.

The protracted nature of the sites has likewise had a stark impact on UNMISS’ willingness to respond to violence or open their gates in areas where they do have a presence. In November 2015, after five months of delays, UNMISS set up a Temporary Operating Base in Leer County. During Protection Cluster assessments in April 2016, conflict-affected people reported feeling that their safety was significantly improved as a result of UNMISS’ presence. See Protection Cluster Situation Update: Leer County, Southern Unity State, May 2016. Just two months later, however, UNMISS troops in Leer refused to open their gates or provide any other form of protection to civilians fleeing nearby violence. One group of civilians who were turned away from the base were attacked shortly afterwards and at least eight women were reportedly raped. It is likely that this number is actually much higher (one humanitarian organisation received secondary reports of over 20 rape cases on the same day), however only eight cases could be confirmed. Had they not lost time approaching UNMISS for protection, they may have run directly to their usual hiding places, and escaped the attack.

This incident is another example of something humanitarians had realised much earlier: despite its mandate, protection of civilians was no longer the priority for the Mission. The first concern was to avoid the creation of new POC sites; everything else (including protecting civilians) came second. In a few instances, this prioritisation was even spelled out by the Mission on paper. In December 2015, UNMISS drafted a concept note entitled ‘Prevention of and Response to Civilians Seeking Protection at UNMISS Operating Base in Leer’. The concept note concluded with the following statement: ‘Until the joint assessment is completed, [the UNMISS Mongolian Battalion] should continue not to protect civilians inside the operating base’, suggesting that, in contravention of the global Protection of Civilians Policy, the default position of UNMISS in this case was to keep the gates closed, even in extreme situations. This has also been observed in other locations in South Sudan, most notably Yambio and Wau, where UNMISS troops actively turned away civilians during outbreaks of violence, See Radio Tamazuj article entitled UN Peacekeepers Turned Away Yambio Civilians Seeking Shelter from Violence, 3 August 2015, and UNMISS Press Release on Wau from 26 June 2016. and in some cases even told them to ask humanitarian organisations for protection (despite the fact that humanitarians have neither the mandate nor the capacity to provide physical security).

The Leer concept note was particularly concerning as it came prior to any UNMISS strategy for the actual protection of civil-ians in the area. Although the note did include some valuable ideas on steps UNMISS could take to prevent and respond to violence, the priority goal (as clearly stated in the title) was to prevent civilians from entering the base – not to protect them from attack. It was only after the Protection Cluster and the Office for the Coordination of Humanitarian Affairs (OCHA) had insisted that this be the focus did UNMISS decide to develop a Protection of Civilians Response Plan for the area.

Protection of IDPs inside the sites

While the impact of the POC sites on protection in other areas has been significant, protection of people inside the sites has also been fraught with challenges. Since the start of the crisis in 2013, four POC sites have been overrun or shelled Akobo (December 2013), Bor (April 2014), Malakal (February 2016), Juba (July 2016). and over 180 IDPs have been killed during attacks on the sites. Although allowing people to enter the sites undoubtedly saved tens of thousands of lives during the first days of the crisis, these repeated, deadly attacks raise the question whether the protective benefits of the sites still outweigh their very tangible risks.

Because the sites are inside UN peacekeeping bases there has been an assumption that the UN will not allow them to be attacked. Unfortunately, much like the ‘safe areas’ of the Balkans, the capacity of peacekeepers to defend the POC sites is limited, and peacekeepers have repeatedly demonstrated that they are unable (or perhaps unwilling) to respond in the event of a serious incident. Despite having a mandate to use all means necessary to protect civilians, up to and including the use of deadly force, time and again peacekeepers have abandoned their posts as soon as fighting nears. In this context, one must question whether characterising these sites as ‘protected’ is giving IDPs a false sense of security, and whether they may have been better off relying on their own coping mechanisms and self-protection strategies.

Some IDPs say that, even though they know UNMISS won’t protect them, their hope is that potential perpetrators will be less likely to attack them in the presence of international personnel. Unfortunately, each time the international com-munity fails to take decisive action in response to these types of attacks, the effectiveness of ‘witnessing’ as a deterrent diminishes.

Ultimately, though, IDPs have voted with their feet, and they have voted to stay. Even after repeated attacks on the sites, tens of thousands of families have indicated that they would rather tolerate the security threats inside the sites rather than risk going outside and being faced with their attackers. Humanitarians and peacekeepers have an obligation to respect this, and should focus on how to strengthen protection both within the sites and outside them, rather than prematurely encouraging people to leave before their safety can be ensured.

Lessons for future operations

Although imperfect, POC sites are an important protection model in extreme situations, and in South Sudan they were undoubtedly responsible for saving tens of thousands of lives. Below are some lessons from South Sudan that could help reduce the risks associated with POC sites in the future. While there are many additional steps that can and should be taken to enhance protection in general, the focus here is specifically on mitigating risks created by the POC sites themselves.

To reduce the negative impact of POC sites on conflict-affected people outside them:

• Maintain an appropriate balance between pro-tecting IDPs in POC sites and protecting civilians elsewhere. The assessment of what constitutes an ‘appropriate balance’ should be based on an analysis of risks across the country, and should be a joint discussion between peacekeepers, humanitarians and affected communities. Steps for achieving this balance should be clearly defined in the Mission’s Protection of Civilians Strategy and resourcing plans, as well as in a Humanitarian Country Team Protection Strategy that articulates how humanitarians will independently work towards improving the pro-tection environment.
• Enhance predictability in opening the gates. To date, there is still no common understanding (either at global or national level) of what constitutes an ‘in extremis’ situation that would trigger the opening of the gates. Not knowing whether they will be offered protection at a base (or what other protection they can expect) significantly impairs civilians’ ability to plan where to flee in times of crisis. DPKO and individual UN missions should outline the process by which these decisions are made and the criteria that determine when the gates are opened. Troops or leaders who fail to open the gates in contravention of agreed policies should be held accountable.
• Determine when not to maintain a peacekeeping presence. If there is a high likelihood that peacekeepers will be unable or unwilling to provide an effective protection response, there is a need to carefully weigh the benefit of maintaining a peacekeeping presence at all. Missions should consult with humanitarians and affected people to discuss whether the benefits of protection by presence are greater than the risks of creating a false sense of security.

To reduce risks to IDPs inside the POC sites:

• Share information about protection capacities so that families can make informed choices. People need to be aware of the limits of protection offered in POC sites so that they can determine whether to remain in the site or move to hiding places elsewhere. Sharing accurate information about protection capacities is critical to avoid creating a false sense of security.
• Strengthen accountability for attacks. Holding attack-ers accountable is directly related to the security of IDPs in POC sites. Many IDPs have taken comfort in the belief that, despite the lack of physical protection from the UN, attackers would be less likely to attack them in the presence of foreigners. Each time we fail to speak, respond or take action, even that minimaldeterrent diminishes. Humanitarians, peacekeepers and the international community need to strengthen their condemnations and responses when POC sites are attacked so as to avoid creating the impression that this can be done with impunity.

To reduce risks to IDPs and conflict-affected people both inside and outside the sites:

• UN missions should develop localised Protection of Civilians Response Plans. Localised response plans are an important tool to complement the national Protection of Civilians Strategy, and should outline specific risks and actions that will be taken at the in-dividual POC site or county level. Peacekeepers should consult humanitarians and affected people when developing these plans.
• Ensure accountability for failures in the protection response. If peacekeepers fail to uphold their protec-tion responsibilities (whether it’s protecting the peri-meter of a POC site or responding to threats against civilians in more remote areas), they must be held accountable. Tokenistic investigations and inquiries are counterproductive. The UN must consider at the highest levels how peacekeepers and troop-contri-buting countries can be more effectively held accountable for failing to uphold their mandates.
• Conduct a rigorous, high-level review of actions and lessons learned from South Sudan. The experience of the POC sites in South Sudan merits much closer examination. The impact of the sites on the protection of people both inside and outside their gates has been significant, and much greater consideration is needed to assess how protective benefits can be supported and negative impacts reduced. A critical examination, similar to the Internal Review Panel Report on UN Actions in Sri Lanka, would be useful.

Caelin Briggs was the Protection Cluster Co-Coordinator from July 2015 to December 2016. She writes here in a personal capacity.