Reconsidering Safe Areas as a Means to Protect Civilians

srebrenica un


Are safe areas an effective way to protect civilians? R2P proponents such as Gareth Evans have suggested “even limited safe havens can help restore peace”[i] while the authors of the Mass Atrocity Response Operations: Military Planning Handbook argue that safe areas could provide both “rapid and direct protection for large numbers of vulnerable civilians” and would “require a relatively small force, concentrated in a few areas.”[ii] Similar arguments have been used with respect to Syria, including by the Turkish government[iii] and former secretary of state Hillary Clinton.[iv] Consequently, it is likely a case of when, rather than if, the international community will authorize another safe area.

Seven safe areas were created in the 1990s. They had little in common, based on differing logics and with significantly different sizes. Open Relief Centres in Sri Lanka, negotiated by UNHCR between the government of Sri Lanka and the Tamil Tigers, may have contained 20,000 civilians. A safe area in Northern Iraq following the Gulf War and supported by American and British troops protected as many as 2.5 million. Yet neither their size nor the money spent determined whether or not they were successful (see Table 1 below).

Table 1: Safe Area Effectiveness

[*Includes costs for country as a whole; **Effectiveness ratio compares direct costs (in millions of US $) to civilians (in thousands) protected]

Years Country Authorization Civilians



Outcome Costs (US$ Million per year) Effecti-veness Ratio**
Military Aid Total
1990- Sri Lanka UNHCR 20 Peaceful N/A 7 7 2.85
2002   MOU (1998)          
1991 Northern Unilateral 2,500 Peaceful 1,700 1,000 2,700 0.93
  Iraq intervention            
1992- Bosnia UN Security 939 Collapse 240 290 530 1.77
1995   Council            
1992- Somalia UNHCR 708 Peaceful Indirect 12 12 59
1992- Liberia ECOWAS 1,200 Collapse 125 65 190 6.315
1996     (1994)          
1994 Rwanda UN Security 1,500 Peaceful 200 50 250 6
1994- Afghanistan Bilateral 300 Collapse N/A 106* 106 2.83
1996 (Jalalabad)   (1996)          


Rather, these safe areas faced three main problems. The first was that were developed to address competing or even incompatible objectives. It was hoped they could contain potentially destabilizing refugee flows even while protecting the civilian population.

The second was that they were built around one of two basic logics: either tactical consent by the belligerents or a credible international military presence established through the deployment of third-party military forces. Consent-based models were established through prior negotiation with the belligerents and reflected a clear doctrine of humanitarian space akin to hospital zones and demilitarized zones.The alternative logic to this was providing protection in the safe area through a credible international military presence, such as in Northern Iraq. Such zones were neither based on consent nor did they have an exclusively civilian character.

By contrast, in the two major safe area failures of the 1990s – Srebrenica and Rwanda- it was presumed that international legitimacy (particularly that of the UN) could secure the safe area in the absence of both consent and a credible international military presence. Such hybrid models were based not on state consent, but instead through an appeal to Security Council resolutions that were then used “as the source of authority.”[v]

These two problems are not disastrous in themselves however. The critical additional third problem was the goals of the belligerents. In most civil wars, tactical consent-based safe areas are effective because they allow the civilian population to remove themselves from the zone of conflict. But such consent cannot work if the belligerents are directly targeting the civilian population. Srebrenica and Rwanda represented much different situations from those in which consent-based safe areas worked: the civilians being protected were still being directly targeted.

It was with the 2000 Brahimi Report that an alternative idea was introduced: that peacekeepers “who witness violence against civilians should be presumed to be authorised to stop it, within their means, in support of basic United Nations principles.”[vi] But peacekeepers with Protection of Civilians (PoC) mandates continued to be limited both by the need for consent and by an informal territorial logic – their presence suggests a higher level of protection to the local population. At the extreme, this may lead to civilians sheltering within peacekeeping bases as is occurring right now in South Sudan, where some 77,000 civilians have sought safety at UN Mission in South Sudan bases.[vii]

The recent creation of the intervention brigade within the UN Organization Stabilization Mission in the Democratic Republic of the Congo (MONUSCO) represents one way to address these issues when civilians are being targeted and the state lacks the capacity to protect them. In this case, a PoC peacekeeping mission could be deployed in a designated area with a direct civilian protection mandate and with formal state consent to operate, but without tactical consent by the belligerents.

In many mass atrocity situations, however, neither tactical nor formal state consent is likely to be forthcoming. An alternative model safe area would see the Security Council invoking the R2P doctrine to either alter the mandate of an existing peacekeeping mission or to trigger a new deployment.[viii] It is here that the safe area concept is supported as a more limited alternative to regime change.

History suggests any new safe area will lack either tactical consent or both that and formal state consent. Therefore, they face three core problems. The first is that any international military force needs to have a credible presence. As the US and UK led intervention in northern Iraq in 1991 demonstrated, in the best case scenario a safe area defended by such a force can lead to negotiations rather than conflict. But this cannot be assumed, which means the international community must be willing to support action, including a military force capable of holding the safe area, and ensure continued delivery of humanitarian assistance.

The second is that a safe area does not provide a solution to conflict or to mass atrocities. As a project independent from either regime change or negotiations, establishing a safe area offers only temporary safety to its inhabitants. Safe area proponents therefore must be prepared to accept significant long-term costs, including both ongoing military support and humanitarian assistance, and to also support alternative dispute resolution mechanisms.

The third is that a safe area will likely trigger significant displacement as civilians from within the country flee into the area for protection. While previous safe areas may not have worked as containment devices, linking future safe areas to a clear means of exit through asylum will help reduce costs and increase sustainability. A successful safe area may lead to reduced refugee flows and, over the longer term, may provide an alternative for returning refugees.


Phil Orchard is a Lecturer in International Relations and Peace and Conflict Studies at the University of Queensland, and a Senior Researcher and Program Leader, Doctrine, Concepts, and Inter-Agency Cooperation Research with the Asia-Pacific Centre for the Responsibility to Protect. He is the author of A Right to Flee: Refugees, States, and the Construction of International Cooperation (Cambridge University Press, 2014). He tweets @p_orchard.


[i]Gareth Evans, The Responsibility to Protect: Ending Mass Atrocity Crimes Once and for All (Washington, DC: Brookings Institution Press, 2008), p. 125.

[ii] Sarah Sewell, Dwight Raymond, and Sally Chin, Mass Atrocity Response Operations: Military Planning Handbook (Cambridge: Carr Centre for Human Rights Policy; Carlisle, PA: US Army Peacekeeping and Stability Operations Institute, 2010), p. 78.

[iii]Michael Weiss, “Snapshot: What It Will Take to Intervene in Syria: First Get the Opposition to Work Together,” Foreign Affairs, 6 January 2012; Ty McCormick, “Turkey Mulls Buffer Zone on Syrian Border,” Foreign Policy, 17 June 2011, 2011/06/17/turkey_mulls_buffer_zone_on_syrian_border, accessed 18 June 2011.

[iv]Rick Gladstone and Neil MacFarquhar, “Rebel-held Land Could Be Safe Haven in Syria, Clinton Says,” New York Times, 24 July 2012.

[v] Yamashita, Humanitarian Space and International Politics, p. 187.

[vi] UN, “Report of the Panel on United Nations Peace Operations, ,” UN Doc. A/55/305, S/2000/809 (21 August 2000), p. x.

[vii]UN News Centre, “UN peacekeeping chief urges South Sudan parties to respect ceasefire amid fresh clashes,” 18 Mar 2014, #.U6oi8LGJofM.

[viii]Hugh Breakey, Angus Francis, Vesselin Popovski, Charles Sampford, Michael G. Smith, and Ramesh Thakur, Enhancing Protection Capacity: A Policy Guide to the Responsibility to Protect and the Protection of Civilians in Armed Conflicts (Brisbane, Queensland: Institute for Ethics, Governance and Law, 2012), p. xxvii.



Author: protectiongateway

Human Protection Hub

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s