Working Paper # 4

A "Safety-First" Approach to
Physical Protection in Refugee Camps

Karen Jacobsen(1)

May 1999


Many refugee camps today are places of insecurity and outright danger, both for refugees and relief workers, and, by virtue of their destabilizing effect, for those living around the camps. Camps are an essential element of the humanitarian response to refugees, and cannot and should not be eliminated (as has been suggested elsewhere). However, they need to be rendered secure in order to ensure the safety of displaced people and others living and working in and around the camps. During the past few decades, the main focus of the international humanitarian response in asylum countries has been to emphasize assistance at the expense of protection. Particularly in the initial emergency phase, physical assistance ("biological needs") is given priority over protection and security concerns. It is proposed in this paper that this focus be re-directed, particularly in the contingency planning and emergency phases, so as to stress security and physical protection needs before assistance: that is, adopt a ‘security-first’ approach. Such an approach would be the basis of an overall strategy to ensure security and protection in refugee hosting areas. This strategy should be ‘regionally appropriate,’ that is, designed in accordance with the needs and capacities of the various state, international and nonstate actors that participate in the specific refugee situation. An important part of the strategy, it is argued, is the presence of a security force in the camps. Camps cannot be made secure without armed backup, but it is crucial that such a force be appropriately trained and prepared for refugee situations, and carefully controlled and monitored. The specific composition and mandate of such a force would vary from one host country to another, depending on needs and capacities.

This paper begins with an analysis of security and safety problems in camps and refugee hosting areas, and then outlines some ideas about a security-first strategy, including the political feasibility of a camp security force in the current international context, and how such a force would be composed, monitored and controlled. All of these, but especially the latter, are issues that warrant further discussion, and it is hoped that this paper will motivate such discussion. The following qualifications should be noted at the outset:


Physical Safety Problems in Refugee Camps and Hosting Areas (RHAs)

A survey of RHA and camp conditions in 1997, set out in Table 1, revealed three main sources of physical safety problems: a) external military attacks or raids on camps and surroundings; b) violence and intimidation occurring from sources inside or outside the camps; and c) a breakdown of law and order in the camps that gives rise to crime and associated problems.

A. Military Attacks or Raids by Armies, Bandits, Militias, Rebel Groups, etc.

In almost every refugee situation in the past five years (see Table 1), camps have been subject to some form of military engagement, ranging from artillery bombardment of camps in eastern Zaire by Rwandan government forces, bombing runs by the Turkish air force of Kurdish camps in northern Iraq, raids by rebel forces of Sudanese camps in northern Uganda, and ‘hot pursuit’ raids by Myanmar government forces across the border into Thailand. Around the Somali refugee camps in eastern Kenya, bandits have operated with impunity, raping women, hijacking relief vehicles, and kidnapping relief workers. When camps are attacked or preyed on by armed gangs, it is not only camp populations and relief workers who are at risk, but also surrounding communities of local people and self-settled refugees.

  Camps are targets for military attacks for two main reasons:

  1. By containing combatants in their midst, camps are perceived by antagonistic forces, either in the country of origin or in the host country, as giving assistance and protection to their enemies, and are therefore targeted. Large numbers of combatants amongst the refugees can lead to camps becoming militarized, with accordingly increased likelihood of attack. In addition, the presence of combatants in camps undermines civilian authority and sources of law and order, and can lead to camps falling under the control of political or military elements. Refugees are then more likely to be deprived of their rights and otherwise subject to violence and intimidation.
  2. Camps are largely undefended repositories of resources, including food, vehicles and relief supplies, as well as people, who can be forcibly recruited for military or sex or labor purposes, or taken hostage.

 In recent years, a third reason for military attacks on camps has emerged. Governments involved in internal or regional conflicts have deliberately targeted refugees and camps as part of a military strategy to weaken and demoralize opponents, and to promote ethnic cleansing. This has occurred most recently in Kosovo, where Serbian president Milosevic has pursued such tactics, but also occurs in Sudan and elsewhere, as part of a larger pattern of conflict in which civilians are targeted for military purposes. Camps are particularly likely to be involved in armed engagements when they are located close to the border or in conflict-prone or sensitive areas. Then, even if camps are not directly targeted, their location is more likely to expose refugees to crossfire, or to landmine fields. 

B. Violence and Intimidation

 Inside many camps, refugees are subject to intimidation, violence, and harassment from a variety of groups and individuals. These include other refugees, who use violence for reasons of ethnic conflict, or political pressure; and camp guards or other host government authorities, who use physical intimidation to extort resources or sex from refugees, or to pressure refugees to leave the camps, or enter them, or to repatriate. Clashes can also occur between refugees and local people, usually outside the camp, and most often when there is resentment by locals towards refugees for perceived wrongdoings, such as theft or immoral acts, or for inequities resulting from refugees’ access to relief resources.

C. Breakdown of Law and Order

A third reason for camp insecurity is the absence of law and order. There are two main reasons for this occurrence: 

  1. Especially in the emergency phase, soon after an influx, camp populations consist of uprooted, often traumatized or destabilized people. Many refugees are rural people with little education, who have lost their ties to families and villages, and who find themselves cast adrift in an alien, unstructured shantytown-like culture. The result is often increased crime and violence, or increased likelihood of recruitment into militias or organized crime.
  2. Transgressions go unpunished because there is no adequate force to back up what rule of law does exist in camps. In the absence of effective rule of law, petty and violent crime flourishes and can lead to camps becoming zones of drug smuggling, human trafficking, organized crime, illegal logging, and gun running, with the attendant problems of violence. Relief supplies are diverted to enrich those in control or support the war effort, and the perpetrators are able to elude justice by hiding amongst the refugee population. As in any high crime area, the non-criminal population is subject to more generalized violence, and the climate of violence leaches out into the surrounding community. The presence of weapons (even when hidden) increases the combustibility of the situation in and around the camps, as does the problem of bored and frustrated young men in camps, who are candidates for involvement in crime or recruitment to militias.


Making Camps Safe and Secure: Measures and Actors

Securing camps and implementing physical protection requires that the following measures be undertaken in camps: 

If these measures were appropriately implemented, they would address many of the safety problems associated with refugee camps and hosting areas. However, adherence to these principles and the implementation of these measures has been patchy or absent in a number of situations, notably the Great Lakes camps but also elsewhere, largely because host governments have been unwilling or unable to take the necessary steps to secure the camps. Before considering strategies for implementing such measures, it is worth examining the perspectives of the main actors concerned with security in refugee camps, because doing so will give insight into the important political context in which solutions to safety problems are developed.

Two sets of actors have responsibility for undertaking these tasks: the host government and UNHCR. According to international agreements, including the OAU Refugee Convention, the Law of Armed Conflicts, the UN Charter, human rights law, and other instruments, the physical protection and security of refugee camps is in the hands of host governments, who are supposed to work with UNHCR to ensure the safety of refugees.(3) These principles require the host government to ensure that refugee camps remain civilian and humanitarian in nature, that military elements have been disarmed and removed, and that camps are situated away from border areas. In some cases, these principles are put into practice by host states: police or army forces are positioned in and around camps, with the duty to ensure the security of the camps and the safety of refugees. However, in many other cases, this system of protection does not work, or does so only poorly. To understand why this is so, it is necessary to look at the political context of camps, and the position of the main actors in this context.

A. The Host Government

The physical safety problems identified above will not always be of concern to the host government and local authorities, indeed some of the second and third type may even be perpetrated by them. Safety problems are most likely to be addressed when they pose direct security threats or economic problems for the central government, or sometimes for the local authorities. In many host countries, developed and developing alike, the political stance of the central government vis a vis refugees and locals is quite different from that of the local government or authorities, including the military, which control the refugee hosting area.(4)

Refugee movements occur in periods of regional instability and intensified political maneuvering, where the governments of both host and sending countries have a variety of political, security and economic interests at stake, as do local authorities. As zones of concentrated resources and people, camps become integrated into the political and security context of the region, making it unlikely that they could be neutral, nonpoliticized places. Instead, camps are often viewed as components of the conflict, and as resources to further political ends.

Most of the worst cases of camp insecurity occur in regions where the refugee flows and refugee camps have a long history of militarization. During the Cold War, the militarization of refugee groups and of camps, such as occurred with Afghan groups in Pakistan or the Khmer Rouge on the Thai-Cambodian border, was ignored or condoned because of the role they played in the superpower struggle,(5) or because host state supporters of liberation struggles, particularly in Africa, condoned and enabled the use of refugee camps for use by liberation armies, as occurred by the Mozambican FRELIMO in Tanzania, by the South African ANC and Zimbabwean liberation forces (ZAPU) in Zambia.

This situation has changed only in a few ways since the end of the Cold War. Camps and refugees continue to be used and targeted as part of the conflict,(6) except that in most cases, superpower involvement has been reduced, and camp militarization has become less acceptable to the West. In the post-Cold War context, host governments’ view of refugee camps and their willingness to prevent militarization of camps, is motivated by regional and border politics. This was graphically illustrated during the Great Lakes refugee crisis from 1994-1997,(7) but the situation also occurs elsewhere, for example in southern Sudan, along the Thai-Burmese border, and along the Tanzanian-Burundian border, to name only a few. Now, as ever, refugees and camps are part of the political strategies of host and sending governments, as well as whoever may be supporting these governments behind the scenes.

That governments can determine the fate of camps, including whether or not they become militarized, is illustrated by the fact that a host government can respond to different refugee groups and their camp situations in different ways. A few examples illustrate:

As a consequence of their history of militarization, most refugee camps are plagued by the problem of "mixed populations," that is, the presence of combatants and criminals among the refugees. In most camp situations, the refugee population includes both non-combatants (usually women, children, elderly) and combatants, either actual or potential. Even if efforts are made to exclude actual combatants from camps (as is required by UNHCR mandate), it is difficult to enforce such exclusion for any length of time. Combatants are easily able to shed their weapons and uniforms in order to enter the camps. If camps are close to the border this enables refugee combatants more easily to mobilize and conduct guerilla forays across the border. The militarization of camps leads to all manner of security problems, as discussed above.

Even if host governments do not actively encourage the militarization of camps, there may be a general reluctance to host refugees, stemming from concerns about the economic or environmental burden posed by refugees or from potential security concerns. These concerns are often accompanied by the desire to see rapid repatriation, which can in turn translate into deliberate non-protection ("humane deterrence") as a way to encourage repatriation. This absence of will to protect refugees on the part of governments can also lead to camps and refugees being located too close to borders, or in zones of conflict, and thus vulnerable to ‘hot pursuit’ raids or artillery fire. Badly located camps are targeted by rebel groups or bandits for their resources, especially vehicles, but also food and medical supplies, or for the purposes of forced conscription. Even nonmilitarized camps or refugee settlements can thus become part of the war economy. This is a common occurrence, for example, in northwest Uganda, where rebel groups have attacked refugee settlements, sometimes shooting relief personnel or taking hostages, stealing relief supplies and hijacking vehicles. More examples are found in Table 1.

Host government failure to provide physical protection is not always only an issue of absent political will, but sometimes occurs because of lack of capacity. It is not always possible to locate camps in a deliberate way; refugees often spontaneously set up their own camps close to the border in order to facilitate return or monitor the situation in their home region, and then it is difficult to relocate them. In situations where armed combatants are mixed up with refugees, the only effective way to screen them out and keep them out is by means of an armed force. Although initial screening and disarming of incoming refugees sometimes is carried out at border crossing points by host authorities, it is not always effective—and of course, not all refugees come through border crossings. The authorities can be overwhelmed by numbers, inadequately trained, or otherwise unwilling or able to accomplish this task, and many arms or combatants "get through" and mix with bona fide refugees. Unless entering combatants are willing to yield their arms, it is often difficult for unarmed border officials or UNHCR Protection Officers to disarm combatants. Similarly, without adequate military backup, it is difficult to ensure that the civilian nature of a refugee camp is sustained, and that weapons are eliminated from and kept out of camps.

Host governments often do not have sufficient or adequately trained and equipped forces, either police or military, to provide adequate and appropriate physical protection in camps. Lack of capacity is especially problematic in less developed host countries, particularly in border regions, which are less accessible and often beyond the reach of central governments.

B. UNHCR and other relief agencies

When host governments will not or cannot implement protection measures, or in cases where there is no effective or functioning host government, who has responsibility for the physical protection of refugees? The responsibility for protection is often relegated to UNHCR, which is also blamed for physical protection problems,(8) despite their not having either the mandate or capacity for physical protection, to say nothing of other security concerns. In some cases, UNHCR has tried to take on responsibility for camp security, as in the Goma camps(9) in 1994-96, but these have been ad hoc responses and although partially successful in the Goma case, on the whole they have been less than satisfactory responses.

In most cases, UNHCR, NGOs and donors can only pressure host governments to provide the needed protection. When diplomatic pressure fails to produce results, relief agencies are faced with the problem of what to do. When camps become militarized, or when attacks on personnel occur, relief agencies must decide whether to remain in the camps and continue to provide assistance, but turn a blind eye to the violations taking place, or to withdraw from the camps. It is difficult to discern a pattern of withdrawal: much depends on the politics of the situation as well as the likelihood of things improving. In northern Iraq, UNHCR withdrew from Atroush camp when it became too militarized; in eastern Zaire, they did not withdraw. In May 1993, Medecins sans Frontières (Belgium) withdrew most of its women workers from camps in northeastern Kenya after reports of a high incidence of assaults on refugees and aid workers. In northern Uganda in 1996, Oxfam, the largest NGO operating there, withdrew its staff after a series of assaults including abductions and robbery. In eastern Zaire, however, MSF withdrew, as did other NGOs, but Oxfam remained.

Apart from withdrawal, other options for UNHCR are rather meager. The use of the Exclusion Clause as a way to keep camps free of combatants has not been widely used, and also requires the presence of a military or police force to carry out the ejection of combatants from the camps.(10)


Ensuring Physical Protection in Camps: Security and Protection First

Until recently, the main focus of the international humanitarian response has been to emphasize physical (biological) assistance at the expense of physical protection and human rights.(11) In the past year or so, largely as an outcome of the 1994-97 crisis in the Great Lakes, there has been widespread condemnation of the presence of combatants in camps, and calls for more stringent efforts to separate combatants from refugees. UNHCR has sought to address physical protection problems more directly. The High Commissioner(12) and the Director of the Division of Protection(13) have made emphatic statements to various UN bodies, including the Security Council, and to host states, about the need for mechanisms to improve the physical protection and security of refugees and camp populations. The High Commissioner has stressed the need for practical ways to address security problems, and has proffered such concepts as a "ladder of options," which sets out a variety of responses depending on the security problems and needs of the refugee hosting area.

The leadership of UNHCR and other agencies has thus demonstrated a heightened concern about security problems and refugee protection. However, as yet these concerns are not reflected in UNHCR’s recent contingency plans. A draft of the contingency plan for the Great Lakes Region in the context of the current insurgency in the DRC (October 1998), prepared by the Office for the Coordination of Humanitarian Affairs (OCHA), contained barely a mention of security and protection needs and was entirely focused on plans for providing cross-border assistance and humanitarian needs.(14)

Ways must be found to support and strengthen the recent calls for renewed emphasis on physical protection and safety in RHAs. The balance between protection and assistance must be re-set. The remainder of this paper suggests a strategy for pursuing this goal.

A. A Rights-based, Safety-First Strategy: The Need for Force

It is argued here that the primary focus of UNHCR’s response, both in the contingency planning and in the emergency phases, should be changed so as to stress safety and protection before assistance. This may be called a ‘safety and protection-first’ approach.(15) From their inception, camps must be located, set up, and secured, to the extent possible, so as to ensure the physical safety of refugees and all camp denizens, prevent future problems for the surrounding area, and to make the delivery of assistance safe (and perhaps more efficient). At the same time, safety measures must be tempered by legal protection considerations; that is, the rights of refugees must undergird the safety-first approach.

The only truly effective way to secure camps is by means of a force that is trained in issues of refugee protection and has the mandate and capability to use armed force in a refugee context. Many within the relief community have recognized the need for a security force in camps, or a more generalized ‘humanitarian protection force.’(16) Various international and state actors have also recognized the need for such a force. In his April 1998 Report on the Situation in Africa, the UN Secretary General urged "the establishment of an international mechanism to assist host governments in maintaining the security and neutrality of refugee camps and settlements." "Such a mechanism," he continues, "might encompass training, logistics, financial support, the provision of security personnel and the monitoring of national security arrangements."

The UN High Commissioner for Refugees has sought to refine such a mechanism and has advocated a ladder of options. In calling on donor states for "concrete action on behalf of victims of violence," she stated that such action will have to be in the form of security support:

Every situation...requires a specific response. It should not be a matter of all or nothing—or, in other words, a massive military operation with overwhelming force or no action at all. There should be a ladder of options—from military operations to peacekeeping operations, to civilian police, armed or unarmed, UN Guards, or the kind of civilian arrangements we have been involved in ourselves. What we need is a kind of Rapidly Deployable Arrangement that can intervene in order to create a safe environment for humanitarian action.(17)

However, even if such a force were to be available, it is not enough simply to plug it into any situation and hope for the best. Camp safety and protection are best approached from a sub-regional, strategic perspective, rather than as an international mechanism designed ‘from above’ that can be plugged in anywhere. A sub-regional perspective means assessing the area-specific physical protection needs and capacities of the host and sending countries, and devising goals and plans accordingly.

 In some cases, a safety-first plan could be piggybacked on state or regional security initiatives. Recently, there have been several regional intergovernmental initiatives aimed at pre-empting or addressing security and other regional concerns. For example, in 1998, in response to the Secretary-General’s Africa Report, ECOWAS drafted a proposal for the establishment of a "Mechanism for the Prevention, Management Resolution of Conflicts, Peace-keeping and Security". This mechanism envisages the establishment of an ECOWAS Mediation and Security Council, which will be empowered to authorize intervention to restore or maintain peace.(18) This kind of political initiative and will must be taken advantage of in planing for refugees’ safety.

A safety-first strategy will include different elements depending on the requirements of the specific receiving area, but one essential element is a unit trained to use armed force when necessary and to back up security measures. What is envisioned is a country-based Camp Security Force, which would capable of carrying out the protection tasks identified earlier, namely:

B. Planning a Strategy for Camp Safety

Two phases to a safety- and protection-first strategy are envisioned: a pre-influx, contingency planning phase, and a post-influx phase including both the emergency and post-emergency, care-and-maintenance periods. 

I. Contingency Planning Phase

At present, in order to prepare for refugee emergencies, UN Country Teams in host countries have contingency plans (CPs), which are regularly updated. Many of these CPs do not presently contain physical protection components, but it is entirely possible to adapt existing CPs to a safety-first orientation. Many of the elements of such a strategy are already in place, or are in the process of being set up, as a result of the new approach to coordination of humanitarian assistance embodied in the UN’s Office for the Coordination of Humanitarian Affairs (OCHA). In the Great Lakes Region, for example, there are planned UN Joint Logistics Centers, as well as existing UN Country Teams working within an inter-agency framework that includes major NGO partners and host government authorities.

In pursuing a safety- and protection-oriented approach, UN Country Teams could add or reinforce the following components: 

  Contingency plan goals include:

a) Identification, training and supervision of the Camp Security Force. Host government authorities, and particularly local authorities, will play a major role in deciding where the force would come from, that is, whether and which police or military forces could be used. Training, capacity-building and supervision of the force would occur during this period. The roles of UN agencies and NGOs in this aspect of the plan would be advisory or monitoring only, since they have no comparative advantage when it comes to police work. However, DPKO officials can potentially play an active role in training and supervision. In situations where the host government lacks the capacity to create such a force, decisions must be taken as to which international organization or donor state could take on this task.

b) Logistical review of options for screening posts, transit centers and camp location, utilizing past experience where possible, and incorporating input of local population. Camp location and layout would be planned with security as a priority.

c) Decisions about what will be done with screened out nonrefugees and combatants.

II. Post-influx Phase

During and after a refugee influx, not all refugees enter camps, but in sustained complex emergencies a significant number spend some time in camps. There are normally two operational phases to camp implementation: the initial emergency stage, when a refugee flow first occurs, followed by the ‘care and maintenance’ phase. The lengths of these phases vary, depending on circumstances, but the emergency phase is generally only a few weeks long, while the care and maintenance phase can stretch into decades. Each phase requires different security interventions, but if adequate security and protection are accomplished early in the emergency phase, the ensuing security goals will be more easily achieved. This occurs because, if camps can quickly be demilitarized and rendered as civilian and relatively neutral zones, two of the main causes of insecurity will be minimized: first, camps will not be seen as places harboring combatants and therefore will be less likely to be attacked for military reasons, and second, in the absence of combatants, camps are more likely to become places of law and order where crime does not flourish. Demilitarization of camps will not necessarily address the third cause of insecurity—their targeting by rebels and bandits for their resources—but reducing two out the three causes of insecurity will allow a focus upon the third cause.



In recent years, a number of researchers have suggested that camps themselves are at the root of protection problems, and that the eradication of camps will go a long way to resolving these problems.(19) However, camps neither can nor should be eliminated; they are an essential element of the humanitarian response to refugees, both for efficacy of assistance programs and because they are potentially a major source of protection and safety for refugees. What is required is a re-orientation of humanitarian assistance away from a focus on meeting the physical needs of refugees and towards ensuring their security and protection.

This paper has argued that the only practical way to ensure the safety of camps and protection of refugees, what might be called security, is to pursue a security strategy underpinned simultaneously by a trained and armed Camp Security Force and by a rights-based orientation. There are a variety of precedents for a Camp Security Force, ranging from the US Military Police used in the Cuban and Haitian camps at Guantanemo Bay, to the Zairean Presidential Guard used in the Goma camps. These and other cases must be further examined to determine which elements can feasibly be extracted and used or adapted for forthcoming refugee camp scenarios. In turn, human rights monitoring has become much more effective and widespread in recent years, and human rights organizations are increasingly astute at pursuing their goals. It is timely and feasible now to combine force and human rights to protect refugees and those caught up with them. The biggest constraints, as always, are the political ones. It will require the sustained will of donors and host states, working with UNHCR and NGOs, to implement such a safety-first, rights-based strategy. Given the apparently serious and widespread concerns about the security problems associated with refugees, there might presently be a window of opportunity for such an endeavor.


Security Threats to Refugees in Major Host Countries in Africa and Asia,1997


This table summarizes the actual and potential physical threats to the main refugee groups in host countries in Africa and Asia with refugee populations of over 20,000 in 1997. The table is not exhaustive, but is intended to give some idea of the types of threats confronting refugees. The table excludes refugee populations in Europe and Palestinians refugees in the Middle East.

[Note: Since this is a ‘snapshot’ of refugees in host countries in 1997, many refugee situations have changed, and in some cases, full repatriation has occurred (eg. Chakma returned from India to Bangladesh). If refugees entered a host country and then repatriated during 1997, they are not included in the table.]


Types of Threats and Likely Causes


Nature of Threat to Refugee Settlement Site/Camp

Likely Causes

Direct military attack or bombardment

presence of combatants among refugees;

to force repatriation or break up (militarized) camps

Caught in cross-fire or armed conflict

sites too close to border or located in zone of conflict or civil war in host country

Armed raids by rebel groups or enemy forces

for resources or hostages;

to force or prevent repatriation;

forced conscription;


Ethnic or political (factional) conflict between refugees or between refugees and locals

poor organization or management of camps;

ineffective policing by host authorities;

refugees and/or locals are dissatisfied or resentful about camp conditions;

violent crime inside or outside settlement

absence of law and order or ineffective policing of settlement area

abuse or intimidation by camp authorities or refugee leaders

to prevent or encourage repatriation;

absence of law and order or ineffective policing of settlement area


Table 1

Main Refugee Populations and Security Threats, 1997

I. Africa

Of the 51 states in Africa, 20 hosted refugee populations of over 20,000 in 1997, and of these host countries, refugee experienced serious security problems in 10 of them, as summarized in this table.

Note: many of these host countries hosted multiple ‘case loads’ i.e. refugee populations from different sending countries, but if a particlar ‘case load’ did not experience significant security problems it is not listed in the table.




(UNHCR Total Refs)a

Country of

Origin Total/UNHCR-assistedb

Refugee Distribution in Host Country

Threats to refugees

(actual and potential)




Eritrea 315,000/119,800

One third dispersed among 25 settlements in eastern region; remainder urban

civil war in Sudan might affect settlements


Ethiopia 44,300/14,800

15,000 in settlement sites and camps; remainder urban

civil war in Sudan might affect settlements





4 camps in w

Sherkole (in Assosa)

none reported/unknown, but effects of Sudan civil war might affect camps;

Camps divided along ethnic lines to avoid hostilities among refugees



Somalia 174,100/134,100

120,000 in 3 camps in Dadaab (ne region)

rapes, robbery, carjackings by bandits outside camps


Sudan 37,400/all

Kakuma camp (nw) region

Sudanese rebels in camps exert pressure on refugees



Sudan 160,400/all

Arua/nw region

Ugandan rebels attacks on settlements;

Sudanese rebels in settlements exert pressure to repatriate



Burundi 6,900/2,300

3 sites in south

harassment and threats of expulsion from Rwandan officials


DRC 27,100/26,600

2 camps on western border

Camps attacked by Hutu insurgents, general insecurity/civil war of region



Burundi 459,400/259,400

8 main camps along border

spilling over of conflict in Burundi into camps;

political factions among refugees led to tension and violence, in turn this led to refoulement by Tanzanian authorities


DRC 74,300/all

2 large camps on border

hardline elements sought to inhibit repatriation; Tanzanian authorities crackdown led to aggressive refoulement

Dem. Rep. Congo (DRC)


Rwanda 37,000/2,900

"the lost refugees"

3 makeshift camps subject to "insurmountable logistical difficulties" (UNHCR) then overtaken by war

thousands fled camps after outbreak of civil war and trekked into jungle, massive loss of life


Burundi 47,000/5,900

sites and self-settled along border

civil war in DRC led to widespread insecurity of refugees


Sudan 61,200/all

farming sites in ne

DRC civil war led to repatriation but effects unknown


Uganda 44,300/all


DRC civil war led to repatriation but effects unknown





6,000 in camps near city of Point-Noire

not directly affected by Congo civil war, but concern about presence of Angolan forces hostile to refugees



Sierra Leone 126,800/25,000

sites along border

sites shelled by forces in Sierra Leone



Liberia 243,000/all

sites in Forest region along border

Guinean police and military have subjected refugees to harassment and extortion


Sierra Leone 192,200/180,200

sites in Forest region along border

Guinean police and military have subjected refugees to harassment and extortion



II. Asia

37 states in Asia hosted refugees in 1997, and of these, 13 hosted populations of over 20,000. Of these, refugees experienced security problems in 6 of them.




(UNHCR Total Refs)a

Country of

Origin Total/UNHCR-assistedb

Refugee Distribution in Host Country

Threats to refugees

(actual and potential)






many camps along Burmese border

attacks from Burmese forces;

refoulement by Thai forces



Afghanistan 1,411,800/all

self-settled and 32 government-run camps

none reported/unknown, but possible forced repatriation




mixed with Afghans

none reported/unknown, but possible forced repatriation





Kurds in Atrush camp near border

cross-border attacks by Turkish forces




Kurds in local settlements and a camp in northern Iraq

Factional fighting amongst Kurds and attacks by government forces





9,000 in AL-Gahin camp, rest urban

none reported/unknown, but possible forced repatriation





2 camps

clashes between refugees and authorities sparked by forced repatriation by authorities





camps in Tripura

pressure from Indian authorities to repatriate


a Number refers to total number of refugees in host country, according to UNHCR 1997 Statistical Overview, Table 1.

b Considers only refugee populations over 20,000. UNHCR 1997 Statistical Overview, Table 3. The usual caveats about estimating refugee numbers apply.



World Refugee Survey, U.S. Committee for Refugees, Washington D.C. 1998.
UNHCR 1997 Statistical Overview , UNHCR, Geneva 1998.



* The Mellon Reports series, and the studies upon which they are based, are supported by a generous grant from the Andrew W. Mellon Foundation.

1 Assistant Professor, Political Science Department, Regis College and Adjunct Professor, Fletcher School of Law & Diplomacy. My thanks to the many UNHCR officials whom I interviewed, and especially for the ideas and input of Tahir Ali, Kofi Asomani, Joel Boutroue, Jeff Crisp, Bernie Doyle, Anki Erikkson, Fillipo Grandi, Pablo Mateu, Ekber Menemgioclu, Craig Sanders, and the late Pierce Gerety. Thanks also for the input of the Mellon-MIT Inter-University Program on NGOs and Forced Migration seminar, especially John Hammock, Jennifer Leaning, Sharon Russell and Myron Weiner.

2 "Nonrefugees" are exiles who should be excluded from refugee status by virtue of their involvement in crimes against humanity.

3 For a discussion of the international legal roots of refugee protection, see Pirkko Kourula, Broadening the Edges: Refugee Definition and International Protection Revisited. (Boston: Martinus Nijhoff Publishers 1997) especially Chap. V.

4 This difference has manifested itself in the United States, when the government of Florida took a stand against the federal government’s Cuban immigration policy in 1994, and forced the federal government to reconsider its position. Presentation by Rick Nuccio at The Fletcher School of Law and Diplomacy, October 9, 1998.

5 Much has been written on the political impact of refugees. See Zolberg, A. R., Suhrke, A. and Aguayo, S., Escape From Violence: Conflict and the Refugee Crisis in the Developing World. New York: Oxford University Press, 1989; Weiner, M., The Global Migration Crisis: Challenges to States and Human Rights. HarperCollins, 1995. Loescher, Gil. 1993. Beyond Charity: International Cooperation and the Global Refugee Crisis. New York: Oxford University Press.

6 " many instances, large-scale movements of people - whether from, to or within their country of origin - have been deliberately provoked or engineered by the parties to armed conflicts, with the specific intention of furthering their political and military interests." UNHCR/OAU 1998. Regional Meeting on Refugee Issues in the Great Lakes. Kampala, 8-9 May 1998. Paper 1, Regional Protection and Security in the Great Lakes Region, p.2

7 Boutroue, Joel. 1998. Missed Opportunities: The Role Of The International Community In The Return Of The Rwandan Refugees From Eastern Zaire. Center for International Studies, M.I.T. Cambridge, Massachusetts.

8 See for example, Bonaventure Rutinwa, "Refugee protection and security in East Africa." Refugee Participation Network, No. 22, October 1996.

9 See Boutroue 1998.

10 For a discussion of the use of the Exclusion Clause, see Lawyers Committee for Human Rights, "Safeguarding the Rights of Refugees Under the 1951 Exclusion Clause," August 1997.

11 For a discussion of the historical roots of the assistance-first orientation of UNHCR, see E. Mtango, "Military and armed attacks on refugee camps." In G. Loescher and L. Monahan (eds.). Refugees and International Relations (New York: Oxford University Press. 1989) pp.87-121; and P. Kourula, Broadening the Edges: Refugee Definition and International Protection Revisited. (Boston: Martinus Nijhoff Publishers, 1997).

12 See for example the following statements by the High Commissioner, Mrs. Sadako Ogata: Statement at the Open Debate on the Secretary-General’s Report on the Situation in Africa (New York, 24 April 1998); "Humanitarian Action in Conflict Situations" (Panel discussion organized by Institute of Policy Studies, Singapore Red Cross Society and Society of International Law, Raffles Hotel, Singapore, 9 January 1998); Statement to the Foreign Policy Society, Copenhagen, 20 October 1997; Opening Statement to the 48th Session of the Executive Committee of the High Commissioner’s Programme, 13 October 1997.

13 McNamara, Dennis. 1998. "The Future of Protection and the Responsibility of the State. Statement to the 48th Session of the UNHCR Executive Committee." International Journal of Refugee Law, Vol. 10 No.1/2 pp. 230-235.

14 Some country-based contingency plans do contain references to the need for screening or other security-based measures. Much depends on whether or not the host government is perceived to be able to implement protection measures. When they are, there is less need to insert security concerns into the contingency plan. OCHA’s is the first regional contingency plan, and is still in the drafting stage.

15 For a brief discussion of the security-first concept, see Jeremy Ginifer, "Protecting Displaced Persons Through Disarmament," Survival, 40(2) Summer 1998, pp. 161-76.

16 See Thomas G. Weiss and C. Collins. Humanitarian Challenges and Intervention: World Politics and the Dilemmas of Help. Boulder, CO: Westview, 1996.

17 Statement by Mrs. Sadako Ogata, United Nations High Commissioner for Refugees to the Foreign Policy Society (Copenhagen 20 October 1997).

18 ECPS & ECHA Joint Working Group, "Proposed Substantive Follow Up Actions to Section III of the Secretary-General’s Africa Report", draft 19/8/98.

19 For a review of this approach, see Forced Migration Review, No.2, August 1998.