International Spectator

The International Spectator

Volume XXXIII No. 4 (October-December 1998)


Kosovo after the Holbrooke–Milosevic Agreement. What Now?
By James Gow


Several weeks after NATO went to the brink of using air power over Kosovo, the security situation in the province located in southern Serbia and within the Federal Republic of Yugoslavia remains fragile. Both the military and the political dimensions of the arrangement agreed by the authorities in Belgrade, the Serbian and Yugoslav capital, to avert NATO air strikes in October have been less expeditiously implemented than was foreseen.

The relatively slow pace of deploying unarmed international verifiers by the OSCE and an armed reaction force to protect them by NATO means that the sense of inevitability which cemented the Dayton peace accords for Bosnia and Herzegovina in 1995 are not clearly present. Without this sense of inevitability, the situation in Kosovo continues to be strategically dynamic, with Serbs and, especially, Albanian extremists believing that the use of violence can still produce benefits for them. This means that if the pace of international verification of the ceasefire and force agreement cannot be increased, the political progress needed in negotiations between the Serbs and Albanians, conducted indirectly through American offices, will not be achievable. Yet, political progress is essential to avert breakdown of the delicate ceasefire which has largely stayed in place, but is insecure and challenged.

Under the terms of the agreement signed between Yugoslav President Slobodan Milosevic and US Envoy Richard Holbrooke in Belgrade on 14 October and the eleven points statement then made by the Serbian authorities, elections should be held in Kosovo within nine months. Agreement on procedures for the elections was due to be announced on 9 November, but was not, already putting implementation behind schedule. The announcement on elections was part of a tightly worked out timetable imposed by the US envoy, with elements entered into in an ostensibly voluntary way by Belgrade. This US approach reflected the substantial preparatory work that had been carried out over a number of months with the Kosovo Albanian leadership by the US Ambassador to Skopje, Christopher Hill.

However, other dates in that timetable had already been allowed to slide backwards, notably the date for agreement on procedures to reach a political solution—2 November. There were two principal reasons for this slippage in the pre-planned timetable. The first was difficulties presented by both the Belgrade authorities and the Kosovo Albanian leaders. The second was the longer time than planned required to put international verification and protection forces in place.

On the first of these problems, while the discussions between Hill and the Kosovo Albanian leadership prior to the agreement had settled on the notion of an interim agreement for a period of three years, political divisions translated into reluctance to accept this interim period without commitment to steps towards independence to be taken later. The three-year interim arrangement is intended to persuade the Kosovo Albanian leadership that any agreement need not necessarily be forever, while ensuring that it would recognise the perspective shared by Belgrade and the international community that Kosovo could not become independent without Belgrade’s consent. Under pressure from figures such as Adem Demaci, political spokesman for the Kosova Liberation Army (UCK), either not to agree to anything at all, or to agree only to an arrangement which carries the promise of a referendum on the status of the province in southern Serbia, the negotiating team found it difficult to follow the line expected of it by US diplomats.

For its part, the Belgrade leadership has been extremely reluctant to endorse a framework document which ensures the territorial integrity of the FRY, but barely mentions either it or Serbia by name and which, among other things, includes provision for inclusion of Kosovo Albanian representation on the Supreme Defence Council of the FRY, the highest decision-making body on military affairs. To mark Serbian government and Yugoslav military reluctance to accept the document, on 20 November, Yugoslav President Milosevic and Serbian President Milan Milutinovic arranged for their ruling Serbian Socialist Party to issue a joint declaration with some other parties in Serbia of principles for agreement which emphasised Serbian terminology and challenged the US proposals. While this presents a challenge to the legitimacy of any agreement based on the Hill proposals, it seems highly unlikely to do more than delay an agreement based on them. This statement does, however, show the way in which delay creates the opportunity for distraction.

In an effort to narrow the positions between the Belgrade authorities and the Kosovo Albanian leadership, the US State Department supplied the Kosovar Albanian side with a lawyer to assist them in negotiating an agreement of which they could be confident and a US commitment to ensuring that it would be upheld. In this context, US officials let it be known that they were not unduly worried by the delay, as Hill’s role in mediation and negotiation between the parties was continuing in a positive way, even if it was behind schedule. This position might have appeared weak, given that only weeks earlier the US and others in the international community had been reinforcing their position that the threat of NATO air strikes would remain in place if the terms of the agreement, as well as the eleven points to be incorporated in any political arrangements for Kosovo, were not maintained. However, given the political difficulties on the Kosovo Albanian side, there has been no reason to rush to use NATO air power against the Serbian and Yugoslav side; nevertheless, the Activation Order for use of air strikes regarding Kosovo was indefinitely maintained by the North Atlantic Council (NAC) on 27 October.

Although all aspects of the agreement, most notably the military ones, ultimately require the political structures agreed upon to be in place, the international community has been relaxed about the pace of implementation as it has not been able to deliver its part of the bargain—a verification force to oversee implementation. While the existing Kosovo Diplomatic Observer Mission (KDOM) continues to operate and experienced a “surge” in numbers, including the recruitment of “local staff” (including some regarded by observers as “private soldiers”) to take on the roles assigned to the OSCE Kosovo Verification Mission (KVM), the KVM itself is not likely to begin operating until late November or early December. Advance parties had been in Kosovo preparing a 300 strong Headquarters for the operation which should total around 2,000 personnel—primarily unarmed soldiers in civilian dress. But progress on the ground, as well as in composing the full force has taken longer than initially hoped.

In addition to the delay in gathering and deploying the KVM, there is also a further dimension to the delay. This concerns the armed protection and extraction force, to be supplied under NATO auspices, for deployment to the former Yugoslav Republic of Macedonia. This force would be a significant safeguard for the operation of the KVM, once both are in place. This can be expected to stabilise the situation sufficiently to facilitate implementation of any political agreement and elections following it.

Almost as significantly, this resolves the critical problem faced by NATO countries over a period of several months regarding appropriate legal authority for the possible use of armed force. When the NAC finally decided to issue its Activation Order to the military commanders in October, the issue of legal authority for use of force had been fudged. While the US and the UK argued that a compound of legal bases, including UN Security Council Resolution 1199, provided appropriate authority, a number of their allies, including The Netherlands, Italy and Germany, strongly believed that an explicit UN Security Council Resolution was required. With the deployment of the KVM, complemented by a new UN Security Council Resolution (1203), using full authority of enforcement powers under Chapter VII of the UN Charter, a previously vague situation has been transformed and two clear legal bases for the possible use of force established.

The first of these is the right to self-defence, meaning the right to protect nationals involved in the KVM. The other is the recognition in Resolution 1203 that the OSCE mission would need to “consider arrangements to be implemented in co-operation with other organisations” to provide for action to ensure their “safety and freedom of movement”. The problem has ceased to be the legal authority to use force. But this is not the end of all problems, as the composition of the force itself has to be decided. This immediately presented NATO with an old transatlantic dilemma: the US contribution to such a force would not be ground troops, despite the presence of US ground personnel in Macedonia already with the UN preventive deployment (UNPREDEP), leaving Europeans to carry what they might interpret as the burden of a US-inspired policy. While planning for various options has been accelerated on the assumption that US ground personnel would not be directly involved, the exact nature of the likely force in the absence of Americans remains open.

The difficulties in establishing the international mechanisms to support implementation of the various agreements made in mid-October, including that between the Supreme Allied Commander Europe, General Wesley Clarke,and the Chief of Staff of the Yugoslav Army (VJ), General Momcilo Perisic, has meant that international actors have not forced the pace to achieve agreements that they are not yet ready fully to oversee. This is despite the deadlines established earlier. The absence of these mechanisms has permitted the pace of political negotiations to falter. In turn, the pressures on the fragile security situation have grown. The October ceasefire and withdrawal of Serbian and Yugoslav forces from Kosovo and from the field has come under pressure at times.


Challenges to the ceasefire

There have been two main challenges to the ceasefire. The first of these is that, in the absence of political agreements according to schedule and effective mechanisms offered by the KVM and its back-up, there is no belief in the likely success of the political negotiations. (By contrast, in Bosnia and Herzegovina at the end of 1995, the existence of a political agreement and an international commitment, including a role for US troops to see it implemented, created a mood of inevitability which greatly facilitated the embedding of the ceasefire.) Unless the political timetable can be recovered soon after implementation provisions are finally in place, it is likely that any belief in the agreements succeeding will be eroded. This will foster a mood among some Kosovo Albanians and Serbs that armed hostilities will return early in 1999. Only international commitment and momentum can prevent this.

The other big challenge, again encouraged by the fact that the prospect of a political agreement has fallen behind timetable and that the verification presence has not arrived, is that backsliding by the Serbs on the military provisions of the October agreements is being implicitly condoned by the international community, while, in the absence of a stronger international presence, elements of the heavily crushed and divided UCK have begun to exert an influence on the ground, both against Serbs and against ethnic Albanians. The UCK, which has been reduced from around 7,000 (some reports suggested 10,000) strong to no more than 2,000 is now deeply divided over immediate political perspectives. However, its various elements could relatively easily take over where Serbian forces have withdrawn.

The October agreements insisted that Belgrade should remove the additional forces it had deployed to Kosovo for its crackdown begun at the end of February. Belgrade had two types of force deployed. The larger and better armed, was the VJ. Over 20,000 additional troops were deployed to the province in the year preceding the crackdown, meaning that over a third of the VJ’s total strength was deployed there. The other element comprised interior ministry (MUP) units. These were of three kinds—regular police (albeit relatively heavily armed), special anti-terrorist units and special combat forces. The first of these constituted the 6,500 Ministry of the Interior personnel considered normally to be based in Kosovo, while the last two, especially the third, constituted the additional units deployed to Kosovo in numbers estimated to have peaked at around 13,000. It was the MUP Special Forces which carried the main thrust of the attacks against the UCK and, above all, against villages and towns, in the course of the seven-month campaign.

By the 27 October deadline set by the NATO air strike ultimatum for withdrawal of Belgrade forces from Kosovo and from the field, the VJ had been reduced to the level that General Clark’s team judged to be its standing force in Kosovo—1,000 personnel on the borders with Albania and Macedonia, as well as 10,600 personnel in barracks at Kosovska Mitrovica, Pec, Djakovica, Prizen, Urosevac, Gnjilane and, the provincial capital, Pristina. The last of these was also headquarters to the 52nd Corps and the joint command for all VJ and MUP operations in Kosovo. The same October withdrawal agreement allowed for 6,500 MUP personnel to remain in Kosovo in 28 locations, meaning that all others had to be withdrawn by the NATO deadline if air strikes were to be averted. In the end, around three companies—360 personnel—failed to meet the deadline. However, there was no question of using air strikes because of this, given that in the last 12 hours before the midnight deadline, Belgrade withdrew around 4,126 MUP personnel from Kosovo, thereby signalling effective compliance.

The withdrawal of forces did not constitute a complete return to the pre-crackdown position, as some key units, such as the VJ 215th Armoured Brigade, did not return to barracks (in this case, at Nis), suggesting that it continues to be deployed in southern Serbia, ready to return to Kosovo should the order be given. In addition, other units within Kosovo continue to be deployed in the field, However, the Belgrade authorities complied with a 29 October deadline to supply to KDOM full information on all MUP and VJ forces deployed in Kosovo and their locations.


Challenges from the UCK

In Kosovo itself, the ceasefire agreement has met challenges, including a number of killings. Nonetheless, overall, for over a month now, it can be judged generally to have held and to have been successful. However, an increasing number of incidents indicate that the security situation is not completely stable. The main element in this appears to be the activity of units of the UCK, although the UCK has cited Serbian provocation for its actions. While the UCK was badly beaten by quantitatively and qualitatively superior Serbian and Yugoslav forces, the ceasefire has enabled parts of it to re-emerge and take up positions where Belgrade forces have been withdrawn. This has resulted in international indulgence for the continuing presence in the field of limited VJ and MUP units from within the standing force for Kosovo. VJ and MUP presence in the field has never ceased on the east side of the road from Djakovica to Klina and along the road from Suva Reka to Stimlje, while MUP units have established new positions in the Malisevo area. In all cases these deployments are judged to be defensive positions.

On 9 November, UCK units attacked the MUP headquarters at Malisevo, while MUP special forces attacked Kosovar Albanian targets at Stimlje. In each of these areas there was tension and incidents occurred. Where MUP units were withdrawn, the UCK moved in to operate, for example in Drenica, Podujevo and the parts of Malisevo not defensively help by the MUP. In some places, there was UCK action against Serb positions. However, there was also growing evidence of UCK action against ethnic Albanians, for example, at Bukos, where a 25 year old was shot in the back of the head. This was an apparent act of intimidation, or disciplining, by the UCK and was not isolated, as UCK elements also fired on and interfered with the freedom of movement of KDOM and KVM preparatory personnel in their clearly marked orange vehicles; VJ forces also later opened fire over the roofs of OSCE vehicles, giving rise to official protests and to a warning by NATO Secretary General Javier Solana in Rome on 17 November that “serious consequences” would follow continued attempts to intimidate international observers, or to impede their freedom of movement. UCK activity, as well as killing ethnic Albanians, also includes renewed attempts to recruit.

In these circumstances, the international community has not rushed to condemn the Serbian and Yugoslav presence, although protests have been lodged. Rather it has temporarily (at least) quietly acquiesced in what has probably involved a small increase in the MUP and VJ presence in the field, in order to provide greater security than would otherwise be the case—NATO air strikes, for example, could not seriously have been employed against UCK forces, which are too small and too disparate to be susceptible of this kind of treatment. The Serbian presence has only served, however, to increase tension with the UCK.



Faster progress on deployment of the KVM and its back-up NATO reaction force in Kosovo is needed to ensure that the delicate security situation in Kosovo will not break down completely. Crucial to ensuring this, is the need for political agreement on the interim arrangements for governing Kosovo which could pave the way for elections.

Political agreement requires the underpinning to be provided by the KVM, inter alia. Both are essential if the momentum for peace is to be maintained in what is not yet a strategically static situation, as both sides, but especially, the UCK, believe that more can be gained through further use of violence. Momentum has partly been lost with slippage in the original timetable regarding agreement on interim political arrangements by 2 November and for agreement on holding elections under those arrangements by 9 November. This slippage is not yet critical—challenges to the ceasefire can be maintained for the time being at current levels. However, gradual erosion will mean eventual collapse. To avoid this, significant political progress from the talks under the aegis of Ambassador Hill is required—perhaps even before the international community’s KVM is ready to go into full operation.

James Gow is Reader in War Studies, King’s College, London.