UNHCR and Kosovo: a personal view from within UNHCR

In this article, the author discusses how the mass arrival of refugees in Albania and Macedonia challenged every aspect of the international community’s ability to respond.

Since early 1998 international attention to Kosovo has had a very strong humanitarian focus. The problems faced by the majority Kosovan Albanian inhabitants and the international response have often been presented primarily in terms of the need for humanitarian assistance and reconstruction. Yet, as the UN High Commissioner for Refugees noted when reporting on her September 1998 mission to the province and region, “Kosovo is a political problem, with devastating humanitarian consequences, for which there is only a political solution”. At the heart of this problem have been long-standing abuses of human rights. The first phase of this tragedy, the crisis in Kosovo itself, was above all a challenge of protection: the over-riding concern of the victims of the conflict was security, not material assistance. The second phase, the mass arrival of refugees in Albania and Macedonia during the NATO air campaign, challenged every aspect of the international community’s ability to respond.

The early stages

The current crisis began in late February 1998 when the first serious clashes between the Yugoslav and Serbian security forces [henceforth ‘security forces’] and the Kosovo Liberation Army (KLA) and its suspected sympathisers led to dozens of civilian deaths at the hands of the security forces. Localized severe human rights abuses by the security forces, the emergence of the KLA and its increasing control of territory and roads characterized this stage. There was only limited need for relief assistance. A new phase began in the second half of July 1998 with the expected major counter-offensive by the security forces against the KLA. The security forces re-established control over a number of key areas, committed wide-spread human rights abuses and intensified a campaign of terror, intimidation and forced displacement to subjugate the civilian population. This massive collective punishment for perceived support of the KLA assumed that every able-bodied male Kosovan Albanian was a suspect. For its part, the KLA was also responsible for human rights abuses. While there was an increasing need for assistance, the greatest need remained for physical security. In addition to mobilizing assistance, the key elements in UNHCR’s strategy during this stage were information, advocacy and a strong emphasis on protection.

By the time of the High Commissioner’s mission to Kosovo in the last week of September 1998 (the second of three such missions in the year), the conflict had already displaced over 300,000 persons inside and outside the province. The central conclusion of her mission was that no just and lasting solution would be possible without a fundamental change in Belgrade’s attitude towards the Kosovan Albanians. Without this change, of which there was little prospect, the ability of the international community to help ensure protection was clearly limited. Observers noted that the muted international reaction prior to September reflected concern at the emergence of the KLA and the resulting ‘independence’ of some areas of Kosovo.

With the adoption of Security Council Resolution 1199 on 23 September 1998, the international community significantly increased its engagement. This resolution, required, among other demands, that the security forces oppressing the civilian population be withdrawn from Kosovo. In his joint declaration with President Yeltsin in Moscow on 16 June, President Milosevic had already made a commitment that “there will be no repressive actions against [the] peaceful population”. In late October 1998, as the security forces began a partial withdrawal, the KLA reasserted its presence. With the deployment of the unarmed OSCE Kosovo Verification Mission (KVM) and the onset of winter, a significant number of those displaced within Kosovo returned home (another demand of S/RES/1199), or at least to the vicinity of damaged homes, and in the short term security improved. All parties understood that this was not itself a political solution - a little time had just been bought in which to find one. By late December 1998 the ceasefire was breaking down. With apparent impunity, the security forces embarked on a series of ‘winter exercises’, using live ammunition, which were clearly aimed at KLA strongholds. This caused new displacement of civilians, including from areas hitherto not badly affected.

Engagement of the international community

A combination of the breakdown of the ceasefire, security threats to the KVM and highly publicised events such as the Raçak massacre on 15 January 1999 led the international community to engage more seriously. The resultant Rambouillet negotiating process began in February and effectively lasted until 19 March 1999. Despite the talks, violence and displacement continued and accelerated markedly after the Rambouillet negotiations ended without agreement on 23 February 1999. By the time the UN humanitarian organizations and our NGO partners had to suspend operations in Kosovo on 23 March 1999, there were thought to be over 260,000 persons displaced within Kosovo, over 100,000 elsewhere in the region and over 100,000 others who had sought asylum outside the region since early 1998.

By early 1999 the humanitarian operation inside Kosovo had assumed a number of the characteristics of UNHCR’s operation during the war in Bosnia, including international convoy teams.(1) Unlike in the earlier operation, access for the convoys and our staff was rarely denied and the delivery of assistance was much easier. By this time some Serb civilians in Kosovo were in need of protection from the KLA. As during the earlier conflicts in Bosnia and Croatia, the humanitarian operation made no distinctions except on the basis of need among the displaced, returnees and others directly affected by the conflict but not displaced. Some 400,000 people in Kosovo were receiving assistance from an operation widely seen as effective. But the limitations of humanitarian action in the absence of successful political action had again been made starkly clear.

During the eleven weeks of NATO air action, from late March to early June 1999, some 900,000 new refugees fled or were expelled from Kosovo. UNHCR found itself engaged in an operation as difficult and complex as any we have faced. Great practical problems of assistance were compounded by major protection problems. Macedonia was reluctant to give asylum. In Montenegro the Kosovan Albanians found a government ready to protect them if it could but were threatened by the presence of Federal security forces. KLA recruitment gathered pace. Many families were separated and the number of missing family members grew. A multitude of practical, protection and political problems had to be addressed in a highly charged political environment in which the stakes for the governments concerned were very high indeed.

Lack of preparedness

The most immediate problem was that we were not prepared for what happened on the scale it happened. Like almost every Western decision maker and commentator, and indeed like most Kosovan Albanians, UNHCR did not predict the mass expulsion of the majority of the ethnic Albanian population of Kosovo. That we were in such company is no excuse. However, until days before the exodus began, the international community, particularly the key Western governments, were banking on peace, and urging UNHCR to get prepared for the early implementation of the Rambouillet accords. It seems unlikely that these same governments - some of which have been sharply critical of our lack of preparedness on a sufficient scale - would have responded to a request from UNHCR for massive contingency preparations predicated on the failure of their own peace efforts. To have been prepared for what actually happened, such a request would have had to have been already met, not just made, at a time when the success of these peace efforts looked possible and what actually happened was still all but unthinkable.

Our key concern was less the failure of contingency planning and lack of stockpiles (these were in the wrong place - inside the Federal Republic of Yugoslavia), than lack of staff. This phase was very staff-intensive from the outset. Getting small groups safely across the border, starting relief and setting up the humanitarian evacuation programme all demanded a high level of experience, keen political judgement and media skills of almost everyone in the field. In the course of a single day any UNHCR field officer could face a host of acute and often highly political problems (and a stream of high-level visitors). We redeployed the team withdrawn from Kosovo, among them veterans of our operation during the war in Bosnia, and mobilized colleagues from across the world. UNHCR’s emergency response team system worked well and made a marked difference. Nevertheless we had insufficient numbers of our most experienced field staff.

Despite the constraints, the rapidly mobilized response to the influxes did meet immediate material needs with minimal avoidable deaths. Much of the credit for this goes to the resilience of the refugees themselves and the hospitality of their immediate hosts. That alone would not have been enough. The international response was generous. Material resources were quickly mobilized, even if some bilateral initiatives were initially uncoordinated. Working very long hours, seven days a week from the start of the exodus, the colleagues that UNHCR and our UN and NGO partners had on the spot in time somehow managed to cope. The contribution of our national colleagues, including those who had themselves just become refugees, was critical.

Request for NATO assistance

A week after the influxes began, and with 300,000 new refugees already in Albania and Macedonia, UNHCR turned to the military for help. We recognized that we were becoming overwhelmed by the political problem of asylum in Macedonia and the practical logistical problems in Albania. The problems of humanitarian organizations working with a military that is or becomes a party to a conflict are well documented.(2) UNHCR requested NATO's assistance only after careful thought. The agreement with NATO explicitly recognized the primacy of the humanitarian organizations, as did NATO’s actual support on the ground. The immediate reason for UNHCRs request to NATO on April 3rd was that there was no other way to unblock the political impasse that had left 65,000 persons perilously stranded in sight of safety on the Kosovo-Macedonia border. NATO's readiness to build camps for them and the start of the humanitarian evacuation programme was the ‘package’ that Macedonia required to allow asylum to these refugees. We should much have preferred such a package not to have been necessary. It was, in the circumstances, literally vital.

As our request to NATO has been criticized in some quarters it may be relevant to note that both the Albanian and Macedonian governments had themselves earlier asked for NATO's ‘humanitarian’ assistance. The belief that an overwhelmed humanitarian response required NATO assistance was shared by Alliance member governments with troops originally deployed to implement the Rambouillet accords, public opinion in most member states and UNHCR. We considered coordination of this assistance preferable to the bilateral arrangements that would otherwise have been inevitable. We had no illusions as to our ability to influence governments on matters where their national interests were so powerfully engaged.

If the need for NATO’s involvement was a combination of the imperatives of the political situation in Macedonia and the scale of the influxes, the humanitarian evacuation programme was a response to the former. For years, the case for ‘burden sharing’ has been argued in meetings such as those of UNHCR’s Executive Committee. Those countries of asylum who wish the donors to share the asylum and not just the financial burden, have had little if any leverage. With the NATO presence on its territory, Macedonia not only had a good case but also it had leverage. The humanitarian evacuation programme was the condition for continued asylum. Faced with this political reality, UNHCR argued for the programme, though - like NATO support to the humanitarian operation - it would have happened anyway.

Selecting refugees for departure on humanitarian evacuation was fraught with difficulties and open to abuse. The concept itself was new. Potential protection problems were many. What was on offer was not resettlement and not even temporary protection. Some governments offering places sought to limit their responsibilities by refusing to allow the programme to be used for immediate family reunion, even of spouses, because this could have given those evacuated the full rights of refugees. Indeed, one of the ironies of this phase was that governments that had respected UNHCR’s requests with regard to the protection and non-return of Kosovo asylum seekers in the phase before NATO action became more restrictive after it had begun. UNHCR has even had to refute the argument that as the deportees had not themselves fled persecution they were therefore not entitled to the full range of protection as refugees.

Because for some commentators (as within UNHCR) the earlier conflict in Bosnia has been a point of reference, the fundamental differences in the contexts of these two operations have tended to be overlooked. In Bosnia the UNHCR operation was, in a sense, a substitute for political action. Ensuring its success, and its perception as a success, was important to key governments. By contrast, the Kosovo exodus was in a sense the consequence of - but most certainly not caused by - political actions of key governments. There was suddenly a massive new humanitarian crisis which governments (and NATO) urgently needed to be seen to be containing. Major and sometimes competing political considerations were at stake for these governments. The humanitarian operation was at times simultaneously a vehicle for and subordinated to these concerns. There was no shortage of claims of achievement and an obvious candidate on hand to blame for whatever perceived shortcomings did become apparent.

UNHCR coordination

Far from pushing UNHCR up front, as in Bosnia, some donor governments themselves took humanitarian initiatives and became directly involved operationally. Needing their own visibility they sought quick solutions that were simply not available. In Bosnia UNHCR had effectively controlled access to, and participation in, the humanitarian operation. As a result coordination was relatively simple. Although the local military forces could deny or restrict access they had limited impact on other aspects of the operation. In the case of the refugee influxes from Kosovo, however, UNHCR coordination of the humanitarian response, though a mandated responsibility that was unchallenged in theory, proved very difficult in practice.

Initially it was well nigh impossible, a massive supply-driven ‘free-for-all’ as donor governments keenly competed with each other. UNHCR was urged to coordinate more effectively while host and donor governments made bilateral arrangements for assistance, camp construction and the running of camps. UNHCR learnt of these post facto, at times only when those NGOs with whom we had agreed provision services came to tell us that they had been replaced. Standards of bilateral assistance varied widely. NGOs who had not previously been active in Kosovo or the region arrived anxious to help. Some insisted on ‘visibility’. Some lacked a clear understanding of the context or needs and not all had the necessary experience.

In Albania, the situation was further complicated by the government's initial wish to work bilaterally where possible and the fact that the immediate response (unlike that in Macedonia) did not have the benefit of experienced ex-Kosovo NGOs. UNHCR helped the large numbers of NGOs that arrived in Albania to set up their own NGO coordinating cell. The sectoral coordinating mechanisms with UNHCR’s direct partners that were already in place for the refugees who had arrived in 1998 were adapted to the new situation. UNOCHA staff seconded to UNHCR played a key role in the Albanian Government's Emergency Management Group, within which responsibility for coordination was effectively shared with UNHCR. In Macedonia UNHCR's NGO ex-Kosovo partners looked to UNHCR from the outset for coordination at the sectoral and overall operations level. New arrivals slotted into already established arrangements, some more quickly and readily than others. UNOCHA staff seconded to UNHCR in Macedonia contributed to the coordination of the key logistics sector and to daily situation reporting (as they had from the UNHCR Office in Belgrade during the previous phases of the crisis).

Such problems are characteristic of the early response to most high-profile humanitarian emergencies. Within a few weeks, as donor liaison mechanisms also gained in effectiveness, a more coherent and consistent needs-driven approach was evident. Co-ordination, nevertheless, proved particularly difficult to handle in this emergency. In the critical early days a host of pressing tasks, and the insufficient numbers of experienced staff available, set back the development of coordination mechanisms. We should have articulated the many problems and constraints more clearly from the start. Another important lesson which has been reconfirmed is that refugee emergencies of such a scale and complexity require a team/consortium approach with our UN and other partners of the kind we have had during this crisis.

There were very significant achievements, often against considerable odds, in the response of the humanitarian organizations, first within and then outside Kosovo. There were failures too and UNHCR is commissioning an independent evaluation. This and other evaluations should help ensure that the lessons are taken on board. It is to be hoped that lessons will also be learnt from the cumulative failures to take resolute political action that contributed to making such a difficult operation necessary in the first place.


Nicholas Morris was UNHCR Special Envoy in the Balkans in 1993 and 1994 and again from 1 April 1998 until 30 April 1999. The article was written in early June 1999. The views expressed in this article are those of the author and not necessarily shared by the United Nations or UNHCR.


  1. For an eye-witness account, see Fernado del Mundo’s “Kosovo Diary” in the UNHCR publication Refugees, vol 1, no 114, 1999, pages 24-25.
  2. For a perceptive analysis in this context, see Elizabeth Becker, “Aid Group’s [UNHCR’s] Partnership with NATO Raises Concern Over Tradition of Neutrality”, New York Times News Service, 9 April 1999.


Opinions in FMR do not necessarily reflect the views of the Editors, the Refugee Studies Centre or the University of Oxford.
FMR is an Open Access publication. Users are free to read, download, copy, distribute, print or link to the full texts of articles published in FMR and on the FMR website, as long as the use is for non-commercial purposes and the author and FMR are attributed. Unless otherwise indicated, all articles published in FMR in print and online, and FMR itself, are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) licence. Details at www.fmreview.org/copyright.