INTRODUCTION

The decade-long crisis in the Balkans has confronted Europe with a seemingly everlasting series of challenges in the field of conflict prevention, peacekeeping, peace enforcement, and post-conflict peace support. The process is still under way, and nobody knows the final outcome of the crisis that has been referred to as the Yugoslav War(s) of Dissolution.1 At the time of writing, it is clear that some of the peace settlements will in effect only last as long as there is a substantial international peacekeeping presence in the region, whereas new developments — particularly in the southern Balkans — may still lead to new confrontations, bloodshed, and the need for new peacemaking efforts and subsequent peacekeeping attempts.

In an African context, there may be a series of lessons to be drawn from Europe’s many attempts to make peace in the Balkans. While every crisis is unique, and every region has its own characteristics, crosscontinent exchanges of experience may be constructive in a global quest for a better understanding of these issues. They should also be of mutual interest to peacekeeping practitioners, as well as scholars specialising on individual regions.

This is in no way intended to suggest that Europe has anything like an established model of peacekeeping and peace support operations that simply can be transferred to other parts of the world so that they can ‘learn’ how to deal with their own conflicts. Nothing could be further from the truth. To be honest, Europe does not really know how to handle contemporary conflicts either. A realistic look at the achievements suggests some occasional, limited success among cases of complete failure, rather than a streamlined regional conflict-handling capacity. It appears that Europe has at least as much to learn from African experiences in conflict prevention as Africa has to learn from Europe.

But Europe’s experiences matter, not least because the reference to the events in the Balkans is such a central part of contemporary doctrine development in the field of peace support operations. Whether this means that any lessons have actually been learned, is another question. A distinction may indeed be made between lessons observed and lessons learned, as the latter implies that these have been taken into the recognised, collective wisdom of the people, states or organisations for whom this insight is relevant, in order to affect future behaviour. The handling of the 1998-1999 crisis in Kosovo repeated several of the mistakes committed during previous attempts to prevent the escalation of conflicts in the Balkans. Other actions were based on false (and overly optimistic) conclusions about what had worked before. As a consequence, a serious question mark should be added when reference is made to lessons learned and not merely observed.

Still, some trends and developments may indeed be identified in the way Europe attempts to deal with its own crises, and these deserve a closer look by anyone dedicated to the enhancement of African capacities in this very crucial area.

One of the rather obvious developments in Europe is the trend towards the regionalisation of peacekeeping efforts. Throughout the 1990s, there was a clear move from the global responsibility of the UN towards more regional efforts and engagements. This process reached at least a preliminary climax when the North Atlantic Treaty Organisation (NATO) decided to attack Yugoslavia in order to influence its policy on Kosovo, without an express mandate from the UN. Strictly speaking, this was a violation of the fundamental principles of the UN Charter — despite the fact that NATO referred to UN Security Council Resolutions 1160 and 1199 (demanding an end to the fighting in Kosovo, but without expressly authorising military intervention).

The UN was clearly marginalised by this act and by the subsequent handling of the war over Kosovo. In the aftermath of the operation, however, the UN re-entered in the form of a UN-led civilian administration in Kosovo, and a UN Security Council mandate for the NATO-led peacekeeping force, KFOR. It may therefore be too early to conclude that the UN has become irrelevant to European crisis management.

Another trend is the recognition that peacekeeping in civil war-like situations is substantially different from overseeing interstate cease-fires or peace agreements. This was bitterly experienced by the UN-led, but predominantly European Protection Force (UNPROFOR) operation between 1992 and 1995. Later engagements by the UN, NATO and the OSCE have led to another important recognition: that peace takes time in coming, and that commitments that are too short may undermine fragile peace settlements. Europe is beginning to learn that intervening in other people’s wars is a very difficult business.

Maybe the key lesson to be learned, is that the level of ambition has been too high. For those unprepared to draw that lesson, there is really only one viable alternative: to engage in serious and structured exchanges of views, experiences and insights in an attempt to make the tools better understood in order to be able to design better responses to future crises in the Balkans, as well as in other regions of the world.

With the end of the Cold War, the superpower ‘overlay’ on European security was largely removed, and the tense but stable environment that had characterised the continent from the late 1940s to the late 1980s went with it.2 Europe went from relative peace in the shadow of the threat of World War III to a post-Cold War situation where new challenges and risks emerged in the wake of the tug-of-war between the superpowers. It quickly became clear that the threats to Europe’s prosperity and security no longer came from the danger of a full-scale war between East and West, but from the potential for secessionist conflict, troubled or disrupted democratisation processes, ethno-nationalist conflicts, and so forth. The forecasts for the near future raged from super-optimism to extreme pessimism, from those heralding ‘the end of history’ to those issuing warnings that the Cold War would soon be missed.3

The concern over European security thus went from being subordinated to the superpower confrontation to the handling of small and medium-sized crises as they occurred (particularly in Central and Eastern Europe) in the short run, and attempting to contribute to crisis prevention in the long run. All the existing security-related organisations of Europe lost their old rationales. It quickly became clear that they would have to adapt to the changing circumstances or become obsolete and perish. While the Warsaw Treaty Organisation was dismantled by its own members, the key Western European organisations embarked on the long road of adapting and restructuring for new realities. Among those that chose to adapt to a new role, NATO is of particular interest in the context of this article. NATO set out in the early 1990s on a path that has largely transformed it from predominantly being a collective defence alliance to a broad-ranging security organisation that can be used for collective security operations out of area. Secondly, it is attempting to ‘project’ security and stability eastwards through the process of expanding its membership to former members of the Warsaw Treaty Organisation.

The development of the Conference on Security and Co-operation in Europe (CSCE) into the Organisation for Security and Co-operation in Europe (OSCE) is also relevant in this context, as it is sometimes seen as the archetype of a ‘regional organisation’, with reference to Chapter VIII of the UN Charter. In 1990, the CSCE solemnly presented its Charter of Paris for a New Europe, and declared that it was now to become a Europe-wide organisation rather than a semi-permanent diplomatic conference between East and West. It was not, however, before 1994 that the name change actually took place.

With the tragic developments that followed the dissolution of Yugoslavia and parts of the Soviet Union, war returned to Europe as a contemporary reality, and with it came the question how to deal with such crises arising in Europe itself. In fact, this meant that Europe had ‘caught up’ with a general global trend towards protracted and savage internal wars with international repercussions. Thus, just as in the case of Africa, the question now was who were to be responsible for dealing with these situations. Should there be a preference for regional solutions, through the use of the OSCE, the European Union (EU), NATO and the Western European Union (WEU), or should Europe, too, rely on the UN and the global level to take a lead role? After all, with the exception of Cyprus, peacekeeping was traditionally something that Europe contributed to in other parts of the world, not something the UN came to Europe to undertake.

Almost a decade later, this question is as relevant as when the Cold War ended. The relationships between regional European organisations, the UN and the leading states in Europe are in the midst of a long period of transition.

Late twentieth century Europe is the world’s most institution-rich region. In principle, this provides an excellent testing-ground for the question how the relationship between the UN and regional organisations could work. The European experience might be a learning process for other parts of the world in this respect. However, in the field of peacekeeping, a regional European role is actually quite new, and still under development. Despite claims to the opposite, with closer scrutiny, the experience may turn out to be less of a ‘success story’ than what the regional organisations themselves may care to admit.

Given its institutional strength, combined with better financial resources than most other parts of the world, Europe might be a logical place to start relieving the UN of the burden of peacekeeping activities, allowing the UN to concentrate on other regions, particularly those where resources are more limited and/or where regional institutional arrangements are more scarce.

The events relating to the collapse of Yugoslavia largely halted the almost euphoric celebration of the era of peace that was to follow the end of superpower conflict. A full-scale war took place before the eyes of a horrified European and world community that proved largely incapable of preventing it from starting or bringing it to an early end. The attempts at political, economic and humanitarian intervention made by all the security-related institutions on the continent ended up proving merely that Europe had not yet achieved a ‘security architecture’ able to prevent and curb major conflicts. Andreatta, for instance, argues that European institutions in 1991 were too weak and, had they been stronger at the time, might have led to a more comprehensive management of the crisis.4 A more pessimistic conclusion could be that, when things get rough, national interests will again dominate, despite previous commitments to international principles and organisations. This may be particularly true when formal institutions have not reached the level of cohesion necessary to remain united even in times of crisis.

The first attempts by the European Community (EC) and CSCE at brokering a peace treaty took place in Slovenia, and were in fact relatively successful. Slovenia and Croatia declared independence on 25 June 1991. On 27 June, Slovenia sent its new armed forces to take over border and customs controls on its international border. Fighting in Slovenia began two days later, as the Yugoslav National Army (JNA) moved to take back control. Both the EC troika and, on Austria’s request, the newly established ‘Conflict Prevention Centre’ of the CSCE were mobilised. Jointly, the EC and the CSCE negotiated a cease-fire, the Brioni agreement, between Yugoslav and Slovene authorities effective from 7 July. The cease-fire, however, did not imply that the EC recognised the formal sovereignty of Slovenia. That decision was to be taken later.

The Brioni agreement was seen as a success for both the EC and the CSCE. For the first time, they could test their ability to mediate a ‘real’ international conflict successfully. Jacques Poos, Foreign minister of Luxembourg and head of the EC troika at the time, triumphantly announced, "This is the hour of Europe."5

The Slovene ‘war of independence’ had little in common with what was to come in the wars in Croatia, and Bosnia and Herzegovina. The war in Slovenia lasted for ten short days and might not even have begun if not for the intention of the Slovene forces to restrain the retreat of the JNA. The fighting killed 39 Yugoslav soldiers, eight members of the Slovene forces, two Austrian journalists, two Slovene civilians and eight foreign truck drivers.6 The JNA was neither enthusiastic about, nor well prepared for fighting its own population, and most of the actual fighting took place because of a few instances of bad communication of orders to hold fire.7 With hindsight, however, the June-July 1991 events in Slovenia opened the door to further military conflict. Woodward argues that, as the internationally brokered settlement defined the JNA as the aggressor and recognised the Republic’s borders as legitimate, the actors in the Yugoslav struggle were set.8 While celebrating their apparent success, the EC and CSCE delegations seemed to pay little attention to the further implications of these decisions for the other republics of Yugoslavia. From then on, the internal constitutional struggle was transformed to one of defending sovereign borders.9

Although the Croat and Slovene issues were ‘settled’ by the 15 January 1992 recognition, EC attempts to negotiate a political solution in Bosnia and Herzegovina continued during February and March 1992. This time, the United States entered the scene, supporting a policy of early recognition of Bosnia and Herzegovina. The US had not yet recognised Croatia and Slovenia. Now, it wanted to restore unity with its European allies on this issue, but at the same time stick to its basic position that if recognition of post-Yugoslav states were to take place, such recognition should be issued to anyone who wanted it.10

Since 1991, France had pushed for the introduction of a European peace force to stop the fighting, but this had not been accepted by its allies. This was partly due to British resistance, as well as US opposition to the development of an independent European military role. The US government was opposed to the introduction of an ‘out of area’ capacity for NATO at the time.

Despite all the efforts to find a ‘European’ solution to the war of Yugoslav dissolution, the initiative reverted to the UN. The cease-fire brokered by Cyrus Vance in Croatia in January 1992 was to be monitored by a UN-led peacekeeping force. The 14 000-strong UNPROFOR was deployed to Croatia in February 1992. According to the plan, the JNA was to withdraw; the Serbs in Krajina, Eastern and Western Slavonia were to be safe in UN protected areas, and a programme of refugee return was launched. Fighting had ended for now, but thirty per cent of Croatia was left outside the control of the Croatian government. Political settlement was deliberately postponed.

The war in Bosnia and Herzegovina flared up in March 1992. On 17 March, an EC plan for a new constitution for Bosnia — based on a cantonment model — was presented, but it was rejected by the parties. On 6 April, the EC recognised the sovereignty of Bosnia and Herzegovina. The day after, the US recognised Croatia, Slovenia, and Bosnia and Herzegovina. The war in Bosnia, soon to be the bloodiest in Europe since World War II, was not to be brought to an end for three and a half long years.

With the advent of UNPROFOR, the EU had abdicated the lead role in the search for a settlement to the Balkan crisis. The time had come to bring in another regional organisation. NATO had been brought into the Balkan theatre as a subcontractor to the UN in 1992. It provided maritime forces to monitor compliance with the weapons embargo on the former Yugoslav region, and air power to enforce the no-fly zone over Bosnian airspace.11 From July 1992, NATO maritime forces were engaged in the supervision of the arms embargo imposed by the UN and economic sanctions in the Adriatic. The operations were initially named Maritime Monitor and Maritime Guard. Maritime Monitor was NATO’s first ‘out of area’ operation. With the introduction of Maritime Guard, the use of force to enforce UN Security Council Resolution 787 was authorised.

At the outset, NATO’s maritime operations were conducted parallel to those of the WEU, but in 1993, they were combined in Operation Sharp Guard after a step-up of the UN boycott. The conducting of these maritime operations, however, did not bring NATO particularly close to the UN at an operative level. Although NATO operated with a UN mandate, beyond the necessary top-level political contacts, the maritime operations were executed autonomously and interfered little with what happened on the ground in Croatia and Bosnia.12

With the introduction of NATO-led air operations into the theatre, the level of co-operation increased. Operation Sky Monitor was initially set up to monitor the no-fly zone established over Bosnia by Resolution 781. In March 1993, the mandate was expanded to authorise active enforcement of the no-fly zone. Later the same year, through Resolution 836, the use of air power was mandated in support of UNPROFOR’s ground forces and in defence of the six safe areas that were designated in April and May 1993. The UN attempted to use NATO air power assets for deterrence, but the strategy eventually failed.13

NATO’s air assets were instrumental in denying the Bosnian Serbs the possibility of using Bosnian airspace for military purposes. The system passed its first real test when four of six Galeb fighter aircraft, violating the no-fly zone, were intercepted and shot down by two US F-16Cs on 28 February 1994.14 This was NATO’s first combat action ever.15 The no-fly zone was, in fact, quite effective in restricting the Bosnian Serbs’ use of fixed-wing aircraft for military purposes, whereas technical and political considerations restricted the use of force against helicopter traffic.16

NATO’s close air support for UN troops was based on requests in situations where UN commanders were to decide on the targeting and execution of such support. This reactive system already existed when the proactive air strikes system was developed in February 1994. The threat of NATO air power was utilised to coerce the belligerents to comply with the twenty kilometre heavy weapon exclusion zone around Sarajevo. However, neither reactive close air support nor proactive air strikes could deter the belligerents in the longer term. Coercion was not possible owing to the incremental use of force demonstrated by the UN and NATO, hardly capable of influencing the cost-benefit calculus of the main aggressors, the Bosnian Serbs. The UN-NATO double key arrangement for authorising air strikes was strongly in focus at the time, and has since repeatedly been referred to as the main obstacle to the use of credible force. If NATO were given authorisation to use air power at will, NATO aircraft would still have to ensure that they did not cause collateral damage to the Bosnian civilian population or UN ground troops. In addition, NATO would have to avoid the various representatives of the numerous international humanitarian organisations and the UN civil police monitors. UN delays in authorising the use of air power were due to checking and double-checking that no collateral damage would be inflicted on any of the representatives of the international community. NATO could hardly have circumvented this issue.

However, the largest obstacle to the use of force with NATO air power was the tight integration of the political, military and humanitarian components. The Bosnian Serbs left the negotiating table when they were attacked, and impeded the distribution of humanitarian relief aid. The vulnerability of the fragile ground operation was exploited through the taking of hostages after an attack. These problems had to be solved before NATO could orchestrate a prolonged air campaign.

NATO was formally less involved in UNPROFOR’s ground operations than in the air and sea operations. However, a clear NATO-UN connection can be found here, too. Many of the main contributors to UNPROFOR were NATO member states (in particular, France and the United Kingdom, which were the main troop-contributing countries). Their military establishments developed substantial Bosnia experience throughout the conflict. NATO’s planning capacity was utilised almost during the entire conflict to prepare for a series of different operational scenarios. From 1993, planners at the Supreme Headquarters Allied Powers in Europe were charged with planning for possible NATO contributions to implement an eventual peace agreement. For instance, the 1993 Vance-Owen plan included a substantial NATO presence on the ground. Massive air strikes were on the planning table from August 1993. Other plans were made for extraction purposes, that is, if the situation in Bosnia were to deteriorate even further and the UN decided to give up its military presence there. Operation Deliberate Effort was the name of NATO’s plan for the active withdrawal of personnel deployed in UNPROFOR and the staff of other international agencies in such a situation. Although it was never put into practice, it provided a good foundation for the planning of Operation Joint Endeavour, which eventually implemented the military aspects of the Dayton Peace Agreement. Throughout the process, the flow of information to NATO was substantial (not least to maintain readiness for a major military involvement). The first IFOR commander, Admiral Leighton Smith, had been kept updated on the developments in the former Yugoslavia on a daily basis in his NATO Southern Command (AFSOUTH) headquarters in Italy for years before he was actually installed in Sarajevo.

The Allied Rapid Reaction Corps (ARRC), which was to become the ground element of IFOR, had also been planning for Bosnia for a long time before they were charged with providing the enabling forces and then the command of the land elements of IFOR in late 1995.17 Throughout 1994 and 1995, UNPROFOR staff officers maintained close links to the ARRC in case of a possible NATO involvement.

Towards the end of the war, NATO air power (which had been available throughout the conflict to support the UN) was used in a coercive diplomacy campaign that led to the cease-fire of 12 October 1995 and the Dayton negotiations in November. The key player in this process was the US special envoy, Richard Holbrooke. In the Holbrookian reading of what happened, air power was the decisive element.18 However, the role of the air strikes in August and September should not be overestimated. They worked because they came in a setting where:

The US accepted, for the first time during the war, the Bosnian Serb claim on a Serb republic (Republika Srpska). The US’ encouragement of the Bosniac (Bosnian Muslims) position between 1993 and 1995, that only a unified Bosnia would be acceptable, is often cited as one of the key obstacles in arriving at an earlier peace settlement. This new US policy was in fact more accommodating towards the Bosnian Serbs than before. Thus, the Bosnian Serbs were bombed into accepting what they had been fighting for all along, albeit within the limits set by the international community.

In other words, the frequent references to the 1995 air strikes as the decisive development in terminating the war may be an overestimation of their relative importance. A false ‘lesson learned’ was drawn, and this proved dangerous years later when NATO was planning for the Kosovo operation, as it created the impression that air power had achieved the status of a magic tool for terminating conflicts.19 Nevertheless, the use of coercive force did indeed contribute to the eventual termination of the Bosnian war, to the cease-fire of 12 October and to get the parties to Dayton for negotiations.

The NATO-led Implementation Force (IFOR) that was sent to Bosnia in December 1995 was the first example of a NATO land military engagement anywhere in the world. This experience has since contributed greatly to the redefinition of NATO’s role and the shaping of the alliance’s identity.

Although IFOR was decided upon at Dayton, without the UN being invited by the US hosts, it was given a Security Council mandate (Resolution 1031 of 15 December 1995). Thus, legally speaking, it was clearly in line with the basic principle of Chapter VIII. Even though it obtained a Chapter VII mandate and the most robust force structure of any peace operation in history, it was still a peacekeeping force in that it was implementing an agreement signed by the parties and was based on the consent of the former warring factions.

Furthermore, IFOR was very much the continuation of UNPROFOR under a new name and with a new and much more operational mandate. For most European troop contributors, 20 December 1995 simply meant switching hats or helmets ‘from blue to green’ and painting the white vehicles camouflage green. The only large country that was coming to the theatre for the first time was the US. However, as it suddenly became the main player in the Balkans operations, the US was eager to underline that IFOR had little to do with the preceding UN operation. For instance, an offer to include people from the experienced and highly skilled UN Civilian Affairs Division, already in place, was turned down, and many people were flown into main headquarters in a country about which they knew little or nothing.20 Other nations were eager to make sure that the transition went as smoothly as possible, while portraying a step change in posture.21 Ultimately, substantial knowledge gained during the UNPROFOR operation was trasnferred to IFOR through this smooth transition.

Other elements also turned out to be beneficial to NATO. Firstly, there were logistical benefits — as the UN had been in place for years, installations and equipment could simply be taken over by NATO rather than brought into a volatile conflict area in the midst of a Balkan winter. Secondly, the operation was to take place in a theatre very close to NATO itself (particularly Italy) and with the friendly support of Hungary, at the time a NATO candidate country that allowed its territory to serve as a staging area and rear base for IFOR. NATO also achieved a very good Status of Forces Agreement with Croatia, through which most transit had to take place, either over land from Hungary or by sea from Western Europe via the Croatian harbour of Split. All this provided an ability to become operational quickly, thus avoiding a period of uncertainty. Thirdly, a very effective ‘psychological operations’ campaign was conducted, sending the message to the parties that they had better not mess with IFOR because it could and would strike back. ‘Robust’ behaviour from the very first days was part of this approach.

Furthermore, NATO had no hesitation in using intelligence resources, which is an integral part of NATO’s force structure. The pre-existing resources in the area (which, of course, officially did not exist when the UN was in charge) could now co-operate better, thus enhancing the possibility of a unified analysis and assessment of the situation. It was not that no intelligence efforts had taken place before, but troop contributors could now be more open with each other on this issue and reap the benefits of joint action, which had been made more difficult by the UN’s formal insistence on openness in all areas and the non-use of intelligence resources.

These were exceptionally good conditions for a NATO force. The former head of the Bosnia Task Force, Greg Schulte, rightly pointed out that, if NATO could not do its job in Bosnia, he would wonder where the organisation would be able to work at all.22

In the design of IFOR, NATO was able to integrate a large number of non-NATO states, including Russia, several former Warsaw Treaty Organisation members and most of Europe’s neutral countries. Thus, NATO had taken a big step forward in becoming a forum for de facto military integration in Europe. IFOR also became the first full-scale test of NATO’s Partnership for Peace agreement.

Beyond strengthening relations with Russia and Central and Eastern European countries, the arrival of IFOR and particularly the US contribution to it led to a noticeable improvement in intra-alliance relations after years of transatlantic quarrelling over what should be done in Bosnia. Now, all the leading countries in the Contact Group had committed forces on the ground, thus paving the way for further integration of policies and action.

One year later, however, it had become clear that the task of securing peace in Bosnia would take more time, and IFOR was replaced by SFOR (Stabilisation Force). Its mandate was given for one and a half years, after which a renewal was given, this time without fixing an end date.

Inter-institutional relations in the Dayton package

SFOR is only one of many agencies involved in the Dayton peace process. The Dayton Agreement stipulates a series of very detailed military tasks for the peacekeeping force. In addition, it is called upon to "provide a secure environment for the civilian implementation process."23

A study of the implementation process suggests that, in the initial stage, it suffered from a noticeable lack of co-ordination. A position of high representation, with a corresponding Office of the High Representative (OHR), was established on Europe’s request, but the US disliked it from the outset. In reality, the high representative received only weak executive powers, and although he was supposed to have primary responsibility for overseeing the implementation process, his only prerogative was that he could call the involved agencies together for a meeting.24 Furthermore, relations between the OHR and IFOR were quite poor during the initial months, and much time was lost.25 In many respects, the civil-military relationship that existed as part of the international community’s endeavours in Bosnia deteriorated initially because of the fragmentation of the various efforts following Dayton. Whereas the UN had provided an integrated framework for the military, the United Nations High Commission for Refugees (UNHCR), the civilian affairs unit and, to a large extent, the main non-governmental organisations (NGOs) involved, all these efforts were now put on separate tracks.26

The situation improved, however, during the second half of 1996, particularly after the introduction of the so-called principals’ meetings, in which the six key organisations began meeting frequently. This was the in-theatre adaptation of the non-existing framework for co-ordination that the international community received at Dayton.

When NATO first started operations, it was frequently pointed out that it had to do its best to avoid so-called mission creep — that is, the tendency for a force to take on tasks other than those for which it was set up. However, this proved very difficult and led to much bad press for NATO. For example, there was the early 1996 burning of Serbian Sarajevo suburbs by radical Serbs who wanted to prevent their fellow nationals from staying on in a Muslim-controlled town, as well as the frequent quarrels with several civilian agencies about access to heavy resources in areas like transportation and engineering. Soon, the NATO-led force was compelled to accept mission creep anyway, as several of the most important challenges were to be found in the interface with the civilian process. In some ways, this ran contrary to what the NATO planners had expected, but was in fact also a consequence of IFOR’s initial military success: The successful implementation of the more precisely defined military aims led to a change of focus now that the war itself had been brought to an end. Relations between the military and civilian dimensions of the Dayton peace process improved further when IFOR became SFOR; SFOR put an increased emphasis on the civilian interface. One example is the augmenting of civil-military co-operation, CIMIC in military slang, in areas such as road construction and gas and water supply.

The initial separation of the civilian and military tasks that came with Dayton was in fact just part of a larger tendency to give every conceivable organisation in Europe a role in the peace implementation process. In some cases, this happened to their outright surprise. The following are a few examples.

The OSCE was given the task of organising elections, something it had never done before. It was not a task the OSCE Secretariat or presidency had asked for, but one allocated to it by Holbrooke and the US delegation at Dayton. The US was eager to promote the role of the main regional civilian organisation, while simultaneously reducing the role of the UN, which in many quarters was seen as partially responsible for the inability to stop the wars in the former Yugoslavia.

Similarly, it came as a surprise to the UN that it was not called upon to organise elections — after all, this was a job the UN had often done before and one for which it is seen as relatively well suited. Instead, the UN was given the task of reconstructing and monitoring the police forces of Bosnia through the International Police Task Force (IPTF). The UNHCR was given the important task of refugee repatriation, but lost much of the role of overall humanitarian co-ordination that it had had during the war. The EU was brought in to help with civilian reconstruction and economic revitalisation, and it was the EU’s Carl Bildt who was appointed as high representative.

Despite its shortcomings, the Dayton process has been a very important learning exercise for Europe, and it remains a model in which co-operation between the UN and a number of regional organisations and agencies remains a key issue. Thus, all in all, this first major experiment in subcontracting does indeed confirm the basic principle of delegating the Security Council’s authority.

If Dayton-Bosnia is an example of a co-operative mode of regionalisation, then the recent developments over the issue of Kosovo may herald a more purely European approach to crisis management in which there is less of a role for the global level and the UN. What was originally presented as an exception — NATO threats, and subsequent use of force without UN Security Council authorisation — may soon turn out to be the beginning of a new pattern in the UN-NATO relationship.

Kosovo is in many ways the place where the current crisis in the Balkans began, and many analysts agree that the crisis will not end before some kind of final settlement has been found in Kosovo.

During the winter of 1997/1998, a guerrilla organisation known as the Kosovo Liberation Army (KLA) took up arms and began fighting the Serb police forces in the region. The war escalated after a Serb attempt to crack down on the military structure of the KLA. Fighting during the summer of 1998 led to a humanitarian crisis that threatened to turn into a catastrophe the following winter, with thousands of internally displaced persons living in the highlands of Kosovo and facing very harsh winter conditions. The violence had a shocking effect on Europe, as many people thought that the crisis in the Balkans was effectively over and that what remained to be done, was long-term peacebuilding assistance, not renewed crisis management.

The UN Security Council was alerted to the Kosovo situation, and sanctions, political pressure and clear demands were issued throughout 1998.27 Several Western powers, Russia and European organisations such as the EU and the OSCE all made a number of diplomatic attempts to achieve the jointly stated aim of the international community: constitutional autonomy for Kosovo within the boundaries of the Republic of Serbia itself, or at least within those of the Federal Republic of Yugoslavia.

However, the agreement in the Security Council stopped short of authorising enforcement action, despite British and US proposals. A dispute developed between the Western members of the Security Council on the one side, and Russia and China on the other. Throughout the year, the US, the UK and NATO had issued a long series of ‘final warnings’ to President Milosevic. NATO was facing a credibility problem over this issue: how far can an alliance like NATO go in threatening to use force when it never materialises?

In the autumn of 1998, preparations were made within NATO to take enforcement action without a UN mandate. Just as they did in the Security Council, the US and the UK took the lead in insisting on the need for a formal NATO decision to authorise air strikes against the Federal Republic of Yugoslavia to put extra weight behind the negotiations in Pristina and Belgrade, which were now led by the chief negotiator from the Dayton talks, Richard Holbrooke. The purpose of NATO’s threat and potential action was to change the calculations of Milosevic towards accepting the demands put forward in UN Security Council Resolution 1199. While most alliance members agreed on the need to put extra pressure on Milosevic, the attitudes towards threatening, and possibly using, NATO military power against Yugoslavia provoked very mixed reactions among the allies. A number of countries, including Germany, Italy, Spain, Denmark, Greece and Luxembourg, expressed grave concerns about authorising NATO action without a UN mandate.28

However, the US used exceptionally strong pressure to get the reluctant NATO members in line. US Secretary of State Madeleine Albright wrote personal letters to the foreign ministers of NATO countries, arguing that "[t]his is not a time to back down’ and that ‘NATO’s credibility is now on the line." According to contemporary press comments, she was trying to invoke the traditional fear in NATO of a withdrawal of US interest in European security.29

During the night between 12 and 13 October, the North Atlantic Council of NATO authorised an activation order, after long deliberations and after hearing fresh reports from the Belgrade negotiations through Holbrooke. In principle, NATO was now ready to attack the Federal Republic of Yugoslavia. However, force planning was primarily conducted on the air power side, as few, if any, NATO countries were prepared to deploy forces on the ground in a war against the Federal Republic of Yugoslavia.30

Obviously, NATO’s aim was to force a pacific settlement of the Kosovo situation through creating a credible threat of force. After the decision on the activation order, and with substantial physical preparations under way for a possible attack, the threat was indeed made more credible. On the other hand, NATO now found itself in a situation where it could no longer step down: it would either have to strike or declare a diplomatic breakthrough, and the alliance did not have much time to lose.

A temporary agreement was reached, however, between Holbrooke and Milosevic on 13 October 1998. It stipulated a partial withdrawal of Serbian special police and military troops from Kosovo, renewed negotiations over partial autonomy for the province, the establishment of a new Kosovar Albanian-dominated police force within the province and the deployment of up to 2 000 OSCE verification officers in the Kosovo Verification Mission. Before this, it had been very difficult to get any OSCE mission established in Kosovo. Yugoslav authorities had blocked the missions of, among others, OSCE/EU negotiator Felipe Gonzales on the grounds that the Federal Republic of Yugoslavia itself had been suspended from the organisation and thus was not prepared to co-operate. The agreement of 13 October was initially hailed as a diplomatic breakthrough — predictably so, as this was NATO’s only alternative to launching the planned military campaign. Right after issuing the Activation Order, the North Atlantic Council decided first to postpone the air strikes for ten days and then, on 27 October, to call them off.31

The OSCE’s Kosovo Verification Mission

After the Holbrooke-Milosevic agreement, the task of verifying compliance with the deal was handed over to the OSCE by the US government, and on 16 October, an agreement was signed between the chairman-in-office, Polish Foreign Minister, Bronislaw Geremek and the Yugoslav Foreign Minister, Zivadin Jovanovic. A 2 000-strong Kosovo Verification Mission was now to be developed, partly on top of the existing Kosovo Diplomatic Observer Mission — which was sent to Kosovo in late spring 1998 — and the European Community Monitoring Mission, the EU analysis and monitoring operation that has been working in several parts of former Yugoslavia since 1991. However, from the outset, it became clear that the observers were facing a difficult and dangerous job. After all, the deal is highly unpopular among the local majority (the Albanian population), as well as the local minority (the remaining Serb population, estimated to number less than ten per cent of the population). While the Kosovar Albanians found that the agreement fell far short with respect to their demand for full autonomy, many Serbs were angry at Milosevic for ‘selling out’ their ‘homeland’.32 According to Fehmi Agani, a key representative of the Kosovar Albanians’ self-styled president, Ibrahim Rugova, the Holbrooke delegation had informed them about the deal with ‘extreme arrogance’, arguing that there was no alternative but to accept it.33 The KLA stated that its aim was independence and that it would remove everything that stood in the way.34 Thus, an unarmed observer force was sent into an ongoing conflict in a setting characterised by extreme polarisation between two groups, the Rugova moderates and some KLA militants, as well as an open internal divide within the majority.

These circumstances have prompted questions about the observers’ security. Holbrooke did not succeed in getting Milosevic’s support for the provision of a lightly armed force in Kosovo to protect the observers. They were to be protected by the very same Serbian police they were to monitor. In practice, protection in such a situation requires a sound working relationship with both the police and the KLA, the two entities among which power is divided in the province. In addition, a NATO extraction force was set up in neighbouring Macedonia, led by French forces. The US refused to send troops to yet another operation in the Balkans at the time.35

NATO was given a role in Kosovo, however, in that it was allowed to conduct air surveillance operations over the area with unarmed aircraft. For this purpose, a separate agreement was signed by the supreme allied commander in Europe, General Wesley Clark, and the Yugoslav army chief of staff, General Momcilo Perisic, on 15 October 1998.36

The Kosovo Verification Mission (KVM) was the first major operation in the Balkans in which the OSCE was involved since the beginning. This is a very different starting point from that of Bosnia and Croatia, and a substantially more demanding one. In Bosnia, the OSCE was brought in after four years of UN involvement. In the Danube region of Croatia (formerly Eastern Slavonia), the OSCE took over the task of monitoring the local police force from the UN, which had been in the region for a number of years. When the transition was made on 15-16 October in the Danube region, it was little more than a change of uniforms, as many OSCE observers were already involved in the UN police monitoring operation there. In addition, most of the practical planning for the new operation had been done by the UN staff rather than by the OSCE itself, as the UN had far more experience in such matters.37 The Kosovo operation was very different: not only was it the largest OSCE operation to date, but there was no predecessor to learn from, and the OSCE was no longer one of several operators, but the main co-ordinating body.

Several critics argued over the winter of 1998-1999 that the apparent reduction in violence in Kosovo was a temporary affair, partly because wars tend to limit themselves in harsh Balkan winter climates, and partly due to temporary self-moderation when checking out the limits of the KVM operation. The developments from January to March 1999 proved the critics right. It became increasingly clear that the OSCE’s civilian monitoring presence would not be a sufficient deterrent to the renewal of fighting. Both sides violated the peace agreement. The Serbian forces continued sporadic attacks on KLA strongholds, whereas the KLA triumphantly marched into positions abandoned by the Yugoslav National Army when they had to remove one-third of their forces and return the rest to barracks. Technically speaking, only the Serbs could violate the agreement, as the Albanians had not signed the Holbrooke-Milosevic agreement, but there had been some kind of implicit understanding that the KLA was also to show moderation in light of the agreement and the presence of the OSCE. However, nothing of the sort happened, and the situation was already deteriorating when 45 men and young boys were found massacred in the little village of Rajac at the end of January 1999. Everything seemed to indicate that the atrocity had been committed by Serbian forces in revenge of the killing of two policemen in the same village. However, the possibly premature statement by the head of the OSCE verification mission, William Walker, who accused the Serbs of the event straightforward, led to a serious deterioration of OSCE/KVM/Serbian relations. With hindsight, this was the beginning of the end for the OSCE-KVM in Kosovo and the beginning of the second countdown to NATO attacks on Yugoslavia (the first ‘countdown’ had been suspended by the Holbrooke-Milosevic agreement).

In order to give diplomacy yet another try, the Contact Group for the Former Yugoslavia, comprising Russia, the US and key Western European states, took the initiative to convene a series of negotiations about an end to the conflict and about the future status of Kosovo at the Rambouillet palace outside Paris in France. Both parties to the conflict were at first extremely reluctant to take part in the negotiations, as they appeared to be more eager to win than to enter into any kind of compromise. Real negotiations hardly took place, as the delegations never physically met, but were given US-initiated proposals which were then carried back and forth by the negotiators. Towards the end of the process, the US negotiators, led by Madeleine Albright, launched a strategy of trying to secure at least one signature — that of the Kosovar Albanians — and then to use this in a direct threat against the Yugoslav authorities: sign here, or you will be bombed. The strategy was even openly confirmed by Madeleine Albright:

"... it is now to a great extent up to the Kosovar Albanians to create this black-or-white situation, the extent to which they now follow up on the political part and continue to make clear that a Nato implementation force is what they want ... the Kosovars have a responsibility in fulfilling their part to make their answer a clear and unequivocal ‘yes’."38

In the end, a half-hearted Kosovar signature was indeed attained. The process developed into a showdown between NATO and Yugoslavia, where none seemed prepared to flicker first. Yet another round of very specific warnings to Yugoslavia’s President Milosevic — several of them issued with a clear deadline — created a situation where both Yugoslavia and NATO painted themselves into a corner and where there was no other way out than either a Yugoslav signature to the Rambouillet proposal that the Kosovars already had accepted, or a NATO attack on the country.

A further deterioration in the security situation on the ground, and the definite collapse of the Rambouillet process, forced the chairperson of the OSCE, Norway’s Foreign Minister, Knut Vollebæk, to order the full evacuation of the KVM from the province. The verification officers had their permits cancelled on leaving Yugoslav territory for Macedonia, and although the KVM was formally upheld in case of a possible re-entry into Kosovo, this was the effective termination of the OSCE’s first experience with a type of peacekeeping operation of its own. The KVM never reached its agreed strength of 2 000 (at maximum strength, they still lacked several hundred verifiers) and important time had been lost due to their belated entry into the theatre, logistic problems, and to internal discussions of leadership structure and the role of the various contributing countries.

NATO’s air strikes began on 24 March 1999. Contrary to what the most optimistic proponents of air strikes had thought, the government of Yugoslavia did not back down at the time, but escalated its operations in Kosovo instead. In many ways, what was experienced in the March-June period of 1999 was not one, but two wars. The first was NATO’s war against Yugoslavia, to which the Yugoslavs offered surprisingly little resistance, even to the point of very sparse use of its air defence systems. The other war was Yugoslavia’s full-scale war against the KLA and against Kosovar Albanian civilians in large parts of Kosovo, to which neither the KLA nor NATO were initially able to put up much effective resistance.

NATO’s air war against Yugoslavia proved to be poorly planned and poorly linked to the political purposes of the alliance. It was criticised for overlooking a number of golden rules of military theory. In the opinion of many analysts, it represented an exaggerated reliance on air power alone, it lacked effective back-up plans in case the original strategy failed, which for instance led to the issue of ground troops being postponed very long into the campaign. NATO also seemed to be caught by surprise by the refugee crisis, despite early intelligence warnings of this as a very real scenario if air strikes were launched.

On the other side, Yugoslavia concentrated on the final onslaught against the armed KLA and civilian Kosovars alike. A massive campaign of systematic and violent expulsion of the civilian population was launched. This war appeared to have been well planned and prepared, and conducted according to a clear, albeit extremely malign, political purpose.

NATO’s plan was to conduct its initial operations during what was described as Phase I (predominantly attacking air defence systems in Yugoslavia) and Phase II (attacks on the Serbian police, military and paramilitary forces active in Kosovo). However, both phases proved rather unsuccessful. The systematic expulsion of Kosovars continued unabated, and when assessing the overall battle damage after the war, it became obvious that the ability to conduct tactical operations only from the air had been substantially over-estimated.39 What was successful in technical terms was NATO’s Phase III: the strategic bombing of infrastructure (like bridges, railways, roads and power plants), communications centres and industry throughout Yugoslavia. Together with Russian diplomatic pressure on Milosevic, and several European efforts to find a compromise solution to the war, a cease-fire was agreed upon at the beginning of June, and the war was terminated on the following conditions:

Post-war Kosovo: A fragile peace

After the withdrawal of the Yugoslav troops, and the successful and undramatic entry of the new peacekeeping force (KFOR) and the civilian UN administration (UNMIK), it quickly became clear that several major issues remained to be solved. These unresolved issues threatened to give European actors headaches over Kosovo for several years to come. Indeed, the future setting in which the UNMIK and KFOR operations are to take place, may end up proving the most difficult and potentially dangerous peace support environment in the Balkans to date.

The fundamental problem in Kosovo after the NATO air war and the withdrawal of the Yugoslav security forces was that nobody had really won the war. Yugoslavia all but lost Kosovo. At the least, it had to hand the province over for an undetermined time to an international administration run by the UN and underpinned by a de facto NATO occupation.40 While Yugoslavia — through the G8 agreement in June 1999 — received formal international recognition of the sovereignty and territorial integrity of the Federal Republic of Yugoslavia, it remains unclear whether the country will ever again exercise any actual control over the disputed province. But, neither did its main opponent, the KLA, win its war. It did not achieve the full independence of Kosovo that it had hoped for, nor did it achieve a situation where it was authorised by the international community to set up a new state administration. Quite the contrary, it now has to endure a long-term international administration of Kosovo based on an agreement where the KLA is supposed to cease to exist in its core function as an army, and to transform itself to a ‘regular’ political movement instead. This runs at odds with what seems to be the bottomline strategy of the KLA: first to liberate the country, and then to install itself as its natural masters in opposition to other (and more liberal) forces. Its final goal is clearly a fully independent state.41 The Democratic League of Kosovo (LDK) used to act as a party-state in the informal home-grown Republic of Kosovo, taking care of many basic state functions from 1989 to 1998, despite Serb insistence that the province was run from Belgrade. Many Kosovars used to see it as the natural dominating force of a future independent Kosovo. The KLA has openly challenged this assumption since the armed struggle began.

With this as a starting point, it was completely predictable that, after having forced the Serbian security forces out, the UN, KFOR and the rest of the international community’s main argument would be with the KLA. Seen from the standpoint of the KLA, the only thing that now stands between them and full independence is the international military and civilian presence.

The relationship between the KLA and UN/KFOR is complex. On the one hand, NATO is seen as a liberator, an ally in the struggle for independence to which gratitude should be expressed and with which, ideally, the best of relations would have to be maintained. Simultaneously, it is the main embodiment of the reality that Kosovo has not been handed over to Kosovar rule after all. Predictably, the tensions between KFOR and KLA will deteriorate, but only slowly, as the second concern eventually takes over from the first. To use a Western European metaphor: the KLA would like to see itself in the position of 1945 France. On behalf of France, General-cum-President Charles de Gaulle could give his allies, the US and Great Britain, their well-deserved warm thanks and then wish them a safe journey home. Instead, the Kosovar leadership is finding itself more in the position of post-1945 Germany or Japan. In Germany and Japan, the new political forces that emerged after the war were subordinated to a long-term benign occupation which was there to ensure the prosperous development of democratic rule and which would only withdraw when this transition was completed to the satisfaction of the occupying forces.42 However, this can hardly be described as the ideal position for the KLA. A gradual normalisation based on UN/KFOR terms may open a broader political spectrum among Kosovars than what the KLA would really like to see before political power is formalised through elections, for example. The KLA’s relative advantage of being the liberating force will then be starkly reduced.

It quickly became clear that the problems in Kosovo were not yet over when a series of what has been described as ‘revenge attacks’ were committed by returning Kosovars against Serbs, gypsies and other non-Kosovar groups. As the Kosovars were moving back into the region from the south, caravans of Serbs were leaving towards the north, and autumn 1999 estimates suggest that as many as 250 000 people were driven away after the war began. As the international peacekeeping presence was unable to prevent this new exodus, parallels were drawn with the burning of the Serb suburbs of Sarajevo in January 1996. This time, however, KFOR at least attempted to curb the flow somewhat. If correct, the reported number of non-Albanians that has left Kosovo since the end of the war are actually the same as the number of Kosovars who were internally displaced by the war and by the actions of Yugoslav forces during the autumn of 1998. It was, to a large extent, this situation that triggered NATO’s decision to prepare for using force against Yugoslavia to prevent ethnic cleansing.

A careful and gradual process may prove to be the only way to ensure that one dictatorial regime is not substituted by another, or that the ethnic cleansing of Albanians is substituted by the ethnic cleansing of Serbs and gypsies instead. But in this very recognition lies the fundamental structural difference between the position of the UN/KFOR and the KLA.

NATO’s understanding of this kind of peacekeeping environment has obviously improved substantially since the first NATO peacekeeping operation was launched with IFOR. In this operation, ‘no mission creep’ was the answer to all the questions raised by the civilian pillar of the implementation process. With KFOR, the understanding that it is in place to help and underpin all the civilian reconstruction efforts, was initially introduced by the first commander, General Michael Jackson, and continued to be in place with the second KFOR commander, General Klaus Reinhardt.43 Indeed, the whole KFOR operation is in several respects more ‘European’ than IFOR, as the American ‘quick fix’ approach is much less visible in Kosovo than it was in Bosnia. Significantly, the first two commanders have been European, the force composition of KFOR is overwhelmingly European, and the operational concepts are much more European than American. This is a better starting point — but it remains to be seen whether this will keep violent conflict out of Kosovo and the countries surrounding it.

At the beginning of the new millennium, the western Balkans remains a highly unstable region. With the exception of Slovenia, few final settlements have been found. The end of the Franjo Tudjman period opens new issues of internal stability in Croatia. A fragile peace is in place in Bosnia, but only due to SFOR’s presence. The final solution to the Kosovo issue remains open — shall it stay an integral part of Serbia, become a republic in the Federal Republic of Yugoslavia, achieve independence, or even join Albania? NATO and the UN are presiding over the formal continuity of Kosovo within Serbia against the will of most Kosovars, while Serbian sovereignty is temporarily suspended. The issue of Kosovo’s future is again in a large part related to the internal developments in the Federal Republic of Yugoslavia, where Montenegro’s move towards de facto independence threatens further disintegration. Inter-ethnic relations in Macedonia are rapidly deteriorating, and are linked to the situation in Kosovo. The EU and OSCE-initiated Stability Pact and other urgently required regional initiatives suffer from the impasse in Belgrade, as a regional approach without the Federal Republic of Yugoslavia, which geographically links the other parts together, seems to prove rather complicated. Thus, the risk of further conflict is definitely open. Yugoslavia’s dissolution has led to four wars so far, and it is far too early to conclude that that the last one has been seen. If further conflict, suffering and disintegration are to be avoided, genuine regionbuilding policies must be introduced, supported and defended.

At this point, it is possible to point to some general lessons that might form part of a potential future set of lessons actually learned in the Balkans, lessons that are also applicable to Africa and other regions of the world.

This article has had a dual purpose. Firstly, it aimed to illustrate the gradual regionalisation of intervention, crisis management and peacekeeping that is under way in Europe. Indeed, such a trend can be found in the present case of UN/NATO/OSCE/EU relations in the Balkans. It is too early to conclude, however, whether this will happen within the co-operative model provided by the UN Charter, or as a further development of Western organisations, like NATO, assuming greater regional or even global responsibilities outside the UN framework.

Secondly, it aimed at presenting some of the specific experiences of the European actors in trying to bring peace to the Balkans, arguing that this might prove relevant to operations in Africa and elsewhere. As stated initially, Europe does not have the answers to these issues any more than other regions. However, the paths taken in the future development of peace support activities in Europe will also be of the utmost importance to the rest of the world. After all, Europe does provide an important testing ground for a cluster of issues of global relevance.

Espen Barth Eide is Director of the ‘UN programme’ at the Norwegian Institute of International Affairs (NUPI) and is responsible for the Institute’s research on collective security and related issues. This article draws partly on E B Eide, Regionalizing intervention: The case of Europe in the Balkans, PRIO-report 4/99: Sovereign Intervention, PRIO, Oslo, 1999.