Select Committee on Defence Fourteenth Report



The Negotiations

50 Diplomatic efforts continued in parallel with military preparations. The Contact Group convened the peace talks at Rambouillet on 6 February 1999. The talks were co-chaired by the UK and France who set out a draft agreement for an interim self-administration in Kosovo that was designed to accommodate both the insistence of the Federal Republic of Yugoslavia on its territorial integrity and the expectations of the oppressed majority in the province. The substance of the negotiations at these so-called 'proximity talks' have been dealt with elsewhere.[112] The first round of talks was inconclusive with neither the Belgrade authorities or the Kosovo Albanians prepared to sign up to the proposed settlement. The talks were suspended on 23 February and then reconvened on 15 March at the Kleber Centre in Paris, where on 18 March the Kosovo Albanian delegation accepted the terms and the Serbian Government rejected them.[113] The Rambouillet negotiations failed. Others have concluded, rightly in our opinion, that Milosevic approached the negotiations at Rambouillet entirely in bad faith—he did not believe NATO would carry out its threats and saw no reason to make any attempt to find a diplomatic solution to the crisis.[114]

51 Though the refusal of Belgrade to compromise at the Rambouillet talks may be regarded as the prime cause of their failure, there were other problems which also hampered the negotiations and had the effect of contributing to the closure of other options. The Kosovo Albanian delegation was split into two groups, one led by the veteran LDK leader, Ibrahim Rugova, and the other by leading KLA figures including Hashim Thaci, who had close family and personal links with Albania. Because the Kosovo Albanian delegates still did not provide unchallenged representation for their community, the talks were not unambiguously between two well defined sides. There were also important divergences of view within the international community. There was resentment from countries not represented on the Contact Group. Its membership excluded the majority of members of the Alliance. And though Russia was a member of the Contact Group, it was kept on the periphery of the talks themselves. General Naumann also observed—

    We saw one weakening of NATO's involvement and that was the phase of the Rambouillet and Paris talks where NATO was not admitted to be present. We were not allowed to offer any advice at all. That led to some problems of cohesion within the [North Atlantic] Council since they all knew that five of the Council members had some national information but, as always in international negotiations, the degree of sharing was never 100%.[115]

The Rambouillet talks may have exposed the disorganisation of the Kosovo Albanians and the bad faith of Milosevic. But they also exposed the absence of a single focus for the international peace making efforts. The military and diplomatic tracks diverged at this crucial point and the failure to fully include NATO in discussions at this stage was a mistake.

The Military Annex

52 Another issue that has frequently been raised in discussions of Rambouillet is the so-called 'Military Annex'.[116] Three sections of the Rambouillet draft Accords tabled by the joint chairs of the talks had military implications—Chapter 7 on the military implementation provisions of the Accords, Appendix A which provided for the phased withdrawal of Yugoslav federal forces on Kosovo territory and Annex B on the status of the multinational peace implementation force. Annex B—the Status of Forces Agreement or SOFA—was the key document which has subsequently been the subject of dispute. It was supposedly based on the SOFA accepted by Serbia at the Dayton Talks to allow for the deployment of IFOR and subsequently SFOR in Bosnia. However, there was a crucial difference. At Dayton, Bosnia was treated as an independent state—the SOFA related almost exclusively (with the exception of a few overflight rights and logistical transport provisions) to the territory of Bosnia. In the case of the Rambouillet SOFA, however, the Western nations proposed giving the right to their troops to move throughout the territory of the Federal Republic of Yugoslavia—making no distinction between the province of Kosovo and the rest of the state (although Chapter 7 related more specifically to Kosovo). If read literally, this could have permitted the stationing of Western forces in Serbia or even Belgrade itself. It also required Serbian airspace to be open to NATO forces. In these crucial repects, Rambouillet was indeed different to Dayton. The Serb government, and indeed most Serbian opposition figures, argued that these provisions were unacceptable.

53 There were, probably, two fundamental reasons for these differences. First, the SOFA was undoubtedly drafted in a hurry, and without the assistance of the Russians who strenuously objected even to discussing it. Second, and more importantly, because the Rambouillet Accords tried to preserve the idea of the territorial integrity of the Federal Republic of Yugoslavia and Kosovo's continuing status as part of it, it was technically more difficult to limit the geographical scope of the SOFA. Paradoxically, by virtue of the Western government's efforts to use language which underlined the sovereignty of the Federal Republic of Yugoslavia, an element of imprecision was engendered which could have allowed the SOFA to be read as a charter for NATO's 'occupation' of Sebia.[117] Many critics have claimed that the terms of the SOFA invited Belgrade's rejection. We have examined this issue ourselves.[118] Our colleagues on the Foreign Affairs Committee suggested that 'NATO was guilty of a serious blunder in allowing the Status of Forces Agreement into the package ...'.[119] But this is not true. NATO was not present, formally, at the Rambouillet talks and were not partners to the negotiations. The criticism by the Foreign Affairs Committee should have been directed at individual national delegations of the Contact Group, not NATO.

54 Certainly, the draft of the SOFA was robustly expressed. We would expect that any intervention force should have been given the maximum protection achievable. Nevertheless, the SOFA could, according to Sir John Goulden, have been amended, and NATO would not necessarily have insisted on all its terms in the event of a genuine negotiation.[120] But other participants in Rambouillet have stressed that, during the talks, the SOFA was never raised by the Serbian delegation as an obstacle to agreement to the Accords. Indeed, the Contact Group Ministers specifically agreed on 23 February, when the talks were suspended, that "there should be further negotiations to determine the modalities of the invited [our emphasis] international civil and military presences in Kosovo". This would have been the moment for Milosevic to raise the terms of the SOFA. In fact he did nothing.[121] The evidence points to Milosevic already having decided to put NATO's credibility to the test before the Rambouillet talks began. However, the draft Status of Forces Agreement did give Milosevic a propaganda weapon. Serbian government representatives were citing the military annex as a major reason for the failure of the Rambouillet talks before the resumption of negotiations on 15 March. But Belgrade rejected the political part of the draft Accords before the Status of Forces Agreement ever became the subject of detailed discussion. The subsequent justifications seem largely to have been used as part of the Serbian propaganda campaign during Operation Allied Force.

The Proto-KFOR

55 The Rambouillet Accords provided, in the event of an agreement, for a NATO-led peace implementation force to be stationed in Kosovo. By 10 March some 4,500 UK personnel had deployed, or were preparing to deploy, to Macedonia as the basis of this force, along with French, German and Italian troops. The UK's contribution included the Headquarters of NATO's Allied Command Europe Rapid Reaction Corps (ARRC) for which the UK is the 'framework nation'.[122] In a lecture to the Royal United Services Institute, General Sir Mike Jackson, its commander, described how it was in—

    ... a strange position [in March 1999] as no [NATO] activation order at all was given to the force in FYROM. The force didn't even have a name! We called ourselves KFOR which worried some, but the fact was that we were only there to do one thing which was to go into Kosovo one way or the other. We did not have, therefore, the full authority which comes with an actual mandate. And, perhaps more importantly, we had no operational funding - that also comes with an activation order. So life was interesting."[123]

In evidence to us, General Jackson added that—

    ... the force was pretty small ... in the region I think of 7,000 or 8,000 at this stage. That was by no means, of course, the KFOR that we had designed ... about 25,000. The formal force generation had not taken place. There was no activation order, no formal force generation. In other words, we were a bit of a ... virtual force, not in the sense we did not exist but identity was a bit of a problem.[124]

56 Although, in the event, the peace implementation force was not required to enter Kosovo until June 1999, it is difficult to avoid the conclusion that NATO should have initiated the formal generation of the force earlier. The failure to do so narrowed the range of options available to the Alliance to enforce a settlement. If Milosevic had climbed down because of the threat of air strikes in February, or as a result of agreement at Rambouillet, it is evident from General Jackson's estimate of the forces required that there would have been insufficient soldiers and equipment on the ground to move into Kosovo as soon as an agreement had been signed. The fact that some NATO nations—France, Germany, Italy and the UK[125]—were prepared to act unilaterally before the Alliance collectively had agreed to do so is evidence enough, in our view, that NATO should have initiated the deployment of the peace implementation force earlier. While it would clearly have been a positive result if Milosevic had climbed down in February or March 1999, NATO is perhaps fortunate that the absence of a properly constituted peace implementation force on the ground at that time was not exposed.

After Rambouillet

57 The failure of Rambouillet further narrowed the diplomatic options open to the international community. NATO had threatened coercive action in October 1998 and secured, partly through this means, an agreement from Belgrade which had not been honoured. The KLA had intensified its campaign of insurrection and the Racak massacre had symbolised the unwillingness of Belgrade to moderate its military campaign against the KLA. Now the Rambouillet talks—whatever their limitations as a negotiation—had been treated with contempt by the Belgrade leadership. Intelligence sources confirmed that Milosevic believed that Alliance unity, let alone wider international unity could not be maintained against him.[126] From his perspective, and given the difficulties of the UN in negotiating the wording of UNSCR 1199 and 1203, the EU's difficulties in adopting a united front against Serbian policy, and the undercurrents of Rambouillet, this was not an unreasonable assumption to make. He continued to overplay his hand. After the failure of the Rambouillet talks, compelling Milosevic's compliance with the agreement made in October 1998 became the default policy of the Alliance.

58 In the months leading up to March 1999, military exercises, additional deployments and preparations to engage in phased air operations were all intended to demonstrate NATO's resolve. Similarly, preparations to deploy NATO ground forces into Albania and Macedonia had to an extent (though perhaps insufficiently) indicated a readiness to act to prevent the crisis spilling over from Kosovo into neighbouring countries.[127] However, the MoD's own report on Lessons from the Crisis comments—

    Lessons have been learned at NATO, and planning procedures and the functioning of the headquarters in Brussels are being updated as a result. NATO and the UK have learned practical lessons on the ways in which NATO planning is conducted ... The planning procedures for NATO non-Article 5 (ie non-collective defence) operations are being considerably streamlined, which should result in speedier operational planning and comprehensive and effective contingency planning in peacetime.[128]

We concur with the implicit recognition in these words that NATO's planning procedures in the period leading up to March 1999 proved too reactive and too cumbersome to move at the pace demanded by events in Kosovo. We discuss the wider lessons for NATO's configuration and procedures later.

59 Military preparations were still in progress when attempts to resolve the crisis diplomatically broke down. By 20 March it was estimated that since the start of Milosevic's campaign of ethnic cleansing around 500 ethnic Albanians had been killed in VJ and MUP action and some 400,000 had by then been displaced from their homes.[129] In response to the growing crisis on the ground, the OSCE Chairman in Office decided to withdraw the KVM. On 22/23 March Richard Holbrooke made a final visit to Belgrade, but was still unable to persuade Milosevic to accept the Rambouillet terms, and on 23 March the NATO Secretary General confirmed that he had given authority to SACEUR to begin the air campaign. On 24 March 1999 the then Secretary of State, George Robertson, told us—

    President Milosevic has failed to respond to the most intense diplomatic efforts and the time has now come for NATO to act. Our military objective—our clear, simple, military objective—will be to reduce the Serbs' capability to repress the Albanian population and thus to avert a humanitarian disaster. Military action has the agreement of all 19 NATO nations ... While at this stage I would not wish to speculate on what elements may be involved in any initial military actions, I would like to stress that whatever action is taken, that action is taken on behalf of all NATO Allies with the aim—the clear and, I believe, justified aim—of averting a humanitarian disaster.[130]

NATO commenced air operations against Serbia. As Mr Cohen, US Secretary of Defense, and General Shelton, Chairman of the US Joint Chiefs of Staff, observed in their evidence to the Senate Armed Forces Committee—

    On March 24, 1999, the United States and its NATO allies turned from a path of diplomacy backed by the threat of force to a military campaign supported by diplomacy.[131]

112  Foreign Affairs Committee, op cit, paras 54-70 Back

113  Ev p 200, para 63 Back

114  See eg Foreign Affairs Committee, op cit para 70 Back

115  Q 988 Back

116  See Appendix 2 Back

117  Michael MccGwire,'Why Did We Bomb Belgrade?' International Affairs, 76(1), January 2000, p. 14. Back

118  As did the Foreign Affairs Committee, see eg Foreign Affairs Committee, op cit, paras 62-65 Back

119  ibid, para 65 Back

120  HC (1999-2000) 28-II, Q 72 Back

121  ibid Back

122  Cm 4724, para 8.7 Back

123  "KFOR: The Inside Story", RUSI Journal, February 2000. Back

124  Q 657 Back

125  Q 646 Back

126  Q367 Back

127  Q 1003 Back

128  Cm 4724, para 5.10 Back

129  HC Deb, 25 March 1999, c 536 Back

130  HC (1998-99) 39, Q 356 Back

131   Evidence to Senate Armed Forces Committee, October 14, 1999. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2000
Prepared 24 October 2000