CIAO DATE: 02/02

Demilitarisation and (Lack of?) Transformation in Kosovo

Robert Neil Cooper
Senior Lecturer in Politics and International Relations, University of Plymouth

1999 - 2000

The Geneva Centre for Security Policy (GCSP)

Introduction

Security sector reform in post-conflict societies faces a number of challenges. Of particular significance however are the challenges involved in demilitarisation. The conventional approaches to this issue have combined disarmament and demobilisation agreements between warring parties on the one hand, with supply side initiatives to prevent the excessive or inappropriate acquisition of arms by actors in post-conflict states on the other. Both of these approaches, however, are problematic. Disarmament and demobilisation initiatives are difficult to embed in post-conflict societies given that the proximity of forces and the easy availability of small arms create particularly acute security dilemmas for parties to a post-conflict peace. In addition, the economic destruction wrought by the legacy of war does not provide an auspicious environment in which to demobilise ex-combatants. At their worst, peace agreements can thus become temporary pauses in a cycle of conflict which allow actors to re-group and replenish arms supplies.

Equally, supply-side initiatives to control the supply of arms to post-conflict societies are being rendered increasingly ineffective by the contemporary conditions of the global system. In particular, the globalising dynamic inherent in the contemporary international defence industry, the increasingly porous nature of national borders (particularly but not exclusively in weak states), the commercial imperatives created in an industry suffering from reduced demand and over-capacity and the growing salience of civil technology in even low level contemporary conflicts all place question marks over the effectiveness of strategies of technology control.

Thus, traditional approaches to post-conflict demilitarisation are problematic. Moreover, as the 'security first' approach recognises, demilitarisation also has to proceed alongside the development of effective law and order and defence capabilities for the political entities that emerge out of conflict. At best this presents complex challenges for the management of peacebuilding processes, at worst it can become a contradiction which undermines peacebuilding.

Towards A Structural Arms Control Agenda

This therefore implies the adoption of an innovative approach to the dual but related problems of weapons proliferation and post-conflict demilitarisation. In particular, there is an important role for 'structural arms control' strategies. Traditionally, commentators have used the concept of structural arms control and also that of 'structural disarmament' to describe the way in which economic or political factors have acted to inhibit the arms dynamic. The concept of structural disarmament has, for instance, been used to describe the way in which the ever rising costs of defence equipment has forced reductions in the absolute numbers of weapons systems held by armed forces. However, this quantitative reduction in weapons systems has also been accompanied by a qualitative increase in lethality. It would thus seem more appropriate to describe the process as one of structural arms control. This latter term has also been used by Stephanie Neuman to describe the process by which changes in economic structures (e.g. global recession) or political structures (the end of the cold war) have limited the ability and willingness of states to raise money to be spent on defence. Structural arms control strategies, therefore, are intended to re-balance economic and socio-political structures in a manner that reduces the incentives for excessive or inappropriate arms expenditure, and/or increases the incentives for and the capacity to promote, a more balanced approach to the different security needs of a society (military, environmental, economic and human). I have discussed elsewhere how structural arms control strategies might be relevant to the general control of the global arms trade. For instance, through emphasising restrictions or bans on the use offsets, export credits and arms brokers.

With respect to post-conflict societies, structural arms control initiatives are likely to encompass two broad categories of initiatives.

1. A range of initiatives designed to manipulate the balance of economic incentives and disincentives for actors to engage in arms purchases and/or military commitments that are excessive and/or inappropriate to the real security needs of a country. Many of these initiatives may be undertaken by external actors with a view to influencing the general conditions of the global arms market or to manipulate the structure of incentives within post-conflict societies in a way that encourages demilitarisation rather than remilitarisation.

2. A range of initiatives aimed at fostering the development socio-political structures within post-conflict states in a way that both promotes meaningful demilitarisation throughout society, whilst also allowing for the development of appropriate police and defence functions. One element of this relates to the role and capacity of parliaments and civilian governments/ administrations to oversee and control the security sector (military, police and judiciary, customs). A second element relates to the role of civil society in both monitoring the operation of the security sector and influencing policy decisions. The civil society groups might comprise veterans associations, Helsinki Citizens' Committees, women's groups, community associations and academic security study groups.

The aim in this paper is to consider how the process of disarmament and demilitarisation in Kosovo can be understood through the lens of structural arms control. This will be undertaken through, first, an analysis of the political-economy of the KLA's ascent. Second, examination of the political support provided to militarised and non-militarised actors in Kosovar society prior to NATO intervention. Third, an analysis of the approach to disarmament and demilitarisation of the KLA adopted by the international community, focusing in particular on the creation of the KPC and the policing of Kosovo. The final section of the paper will then attempt to consider how we can understand the approach to demilitarisation and peacebuilding in Kosovo through the lens of structural arms control.

The Militarisation of the Albanian Struggle

The speed with which the KLA grew and achieved its political goals was quite remarkable. That it was able to do so was a function of both Western policy and the nature of the contemporary arms market, and any attempt to understand the challenges of demilitarisation in 'post-conflict' Kosovo needs to begin by taking into account the context in which the KLA flourished.

The Political Economy of the KLA's Ascent

As has been well documented, by the early '90s Serb rule in Kosovo had created a highly repressive system. From 1988 to '91 Serbian president Slobodan Milosevic had withdrawn the autonomy allowed to Kosovo under the 1974 Yugoslav constitution. This was followed by demands for loyalty oaths that led to most Kosovo Albanians being sacked from the jobs they had in public administration or publicly owned enterprises. Over the decade, approximately 140,000 Albanians lost their jobs, and by 1995 unemployment was reportedly as high as 70%. This pool of unemployed Albanians formed a natural recruiting ground for the KLA. Indeed, in a survey of demobilised KLA conducted after the end of the conflict, over a quarter (26%) indicated they had been unemployed prior to joining the organisation.

The decade of enforced measures was also characterised by systematic corporate looting of companies by Belgrade. In consequence, many public enterprises had all their monetary assets and even equipment removed. Coupled with the impact of conflict and of international economic sanctions, this led to Kosovo's GDP contracting by 50% from 1990 to 1995, falling to less than $400 per capita, lower than Albania, Europe's poorest country and lower than other regions of Yugoslavia.

At the same time, the Serbs in Kosovo were also suffering economically. By 1980, the Serbs share of unemployment relative to their population was actually higher than that of the Kosovo Albanians. Indeed, the effect of war, sanctions and hyperinflation (which by January 1994 reached a monthly rate of 313,563,558%) meant that Serb salaries had dropped from roughly DM 1000 per month to just DM100. The Albanians may have burned with resentment at their exclusion from employment in the public sector, but the Serbs equally resented the Albanians relative success in the private sector and the funds they received from the Albanian diaspora abroad. As Mertus concludes, 'in sum, the economic situation was bad for everyone'.

It is certainly the case that Serb repression in Kosovo provided a stimulus to Albanian nationalism in the '90s. Similarly, the discourse of 'ancient ethnic hatreds' mobilised by both sides (and reaffirmed by the international community) arguably took on such force as to become an autonomous factor driving polarising nationalisms. At the same time, however, the role played by the economic factors noted above should not be underestimated, both as an independent factor stimulating the rise of nationalism on both sides, and as an explanation for repression in the first place. Indeed, as Alcock has noted, 'There is an instructive correlation between the major periods of civic unrest in Kosovo and the periods of greatest economic dislocation in the country as a whole'. The disturbances of 1968 followed from the attempted economic reforms of 1965 and the riots of 1981 followed the second oil shock of 1979 to which Yugoslavia as an energy importer (and Kosovo as the poorest region of Yugoslavia) was particularly vulnerable. Equally, Strazzari has argued that the relatively peaceful situation that existed in Kosovo until 1997 can be explained by reference to the fact that the worst effects of economic discrimination were mitigated for the Albanians by funds from the diaspora and by money earned from illicit trading activities. In addition, there was also a degree of co-operation between the regime in Belgrade and Albanian smugglers in both Kosovo and Albania proper, co-operation that extended to both sanctions-busting activity such as the smuggling of oil and the trade in illicit goods such as arms and heroin. Indeed, there were times when fuel in Albania was scarce because so much had been exported to Serbia. Ironically, the lifting of sanctions in 1995 may thus have had a negative impact on the economic well-being of at least certain sections of Kosovar society, as did the fact that from 1997 other drug routes opened up that by-passed Kosovo.

The militarisation of the Albanian struggle was also, in large part, a function of the political economy of the KLA's arms network. Notably, the group was able to draw funds for arms from levies on the earnings of the Albanian diaspora in Europe and the US, as well as profits acquired from various black market activities. However, whilst the KLA's funding network was certainly extensive, the key event that enabled it to blossom as a military force came in 1997 when widespread rioting in Albania led to a breakdown of government control. In the process, looters seized several hundred thousand weapons and 20,000 tons of ammunition. Up until this point attempts to train and prepare for conflict in the 1980s and 1990s had not only been foiled by Serb police but were equally constrained by Kosovo's position as a landlocked region and Albania's reluctance to jeopardise aid from the West or invite reprisals from Serbia by providing the large quantities of weapons required for an insurgency. In 1997, however, the absence of government authority in Albania, coupled with the flood of weapons onto the market, meant that it was now relatively easy for the KLA to use the funds collected from the Albanian diaspora to obtain arms at knock-down prices (in particular large numbers of Kalashnikovs were available for as little as $10 each).

The role of the international community

The Albanians, who comprised 82% of the population, responded to Serb oppression by setting up their own 'parallel state' with its own educational, fiscal, health, media and other structures. In 1991 they held a referendum which resulted in an overwhelming vote in favour of an independent state and a government was formed. They also declared independence and in 1992 held a clandestine referendum which elected Ibrahim Rugova (the leader of the LDK - Democratic League of Kosovo) as president. Rugova was then re-elected in March 1998 with 99% of votes whilst the LDK won 84 of the 89 seats available.

Throughout most of the nineties the Albanians, under Rugova's leadership, pursued a strategy of passive resistance. Whilst this partly reflected a pragmatic assessment of the likely consequences of military action, as much as it did any principled attachment to non-violence, it was nevertheless a remarkable approach to adopt in the context of the Balkans in the early '90s. For Rugova and his supporters then, it was hoped that this example of civility and tolerance would earn the backing of Western democracies. However, whilst the West was prepared to offer rhetorical support, the fact was that the strategy of passive resistance produced little in the way of concrete initiatives from the international community. In particular, the 1995 Dayton agreement on Bosnia required the co-operation of Slobodan Milosevic and it was not therefore deemed politic to raise the question of Kosovo. Indeed, Bujar Bukoshi, the Prime Minister of the government-in-exile, arrived at Dayton to raise the issue but was turned away at the gates. In contrast, the Serbs, who had committed various atrocities in Bosnia, were rewarded with their own political entity within Bosnia. Even after Dayton, when the High Representative in Bosnia, Carl Bildt, tried to interest the West in negotiations with Belgrade about Kosovo, he met with little interest. The Albanians were deeply disappointed, labelling themselves as 'the forgotten of Dayton'. Thus, from early 1996, influential intellectuals began to challenge Rugova's strategy of passive resistance arguing, with some justification, that it had failed. It was only in 1998, though, that popular support began to shift towards the KLA. Indeed, even at the start of 1998 their numbers were only estimated by outside observers at between 100-1,500 fighters whilst just 279 (or 2%) of demobilised KLA have stated they joined prior to 1998. However, in March, the Serbs responded to KLA actions with a vicious counter-insurgency campaign. The Kosovar villagers, particularly in the western areas, now began to link their self-defence units up with the command structure of the KLA. From this point the outside world began to discuss the KLA as a popularly based organisation. Even at this stage though, as Susan Woodward has noted, the basis of the KLA's support was not political but military - it was aimed at physical security and self-defence. Many of the self-defence units were in fact supporters of Rugova's LDK. Indeed, particularly in the aftermath of the Serb offensive against the KLA launched in July 1988, these same concerns about physical security led some villages to ask the KLA to leave, for fear of being targeted with them, and to hand in to the police weapons distributed by the KLA.

Moreover, whilst there was some movement of LDK officials to the KLA, it was also the case that there were deep divisions between many of the KLA and the LDK. Indeed, throughout 1998 and 1999, many KLA actions were actually undertaken against Albanians labelled as Serb conspirators and as Clark has noted, 'the fact that so many of the UCK's initial attacks were against Albanians ...suggested that perhaps personal scores were being settled'. The LDK actually set up an alternative armed faction, the FARK, and for a brief period the KLA and the FARK fought each other in the mountains of Albania, whilst the commander of the FARK, Ahmet Krasniqi was actually assassinated in Tirana by two masked gunmen reputed to belong to the KLA.

Despite these divisions, the general Western policy towards the KLA was notable for the way it moved from initially labelling it as a terrorist group to the point at which, during the discussions on post-war demilitarisation, the group was essentially treated as the sole representative of Kosovo. At the start of 1998 the KLA was still classified by the US State Department as a terrorist organisation. Indeed, the US Special Envoy to the Balkans, Robert Gelbard, gave what many viewed as a green light to a Serb counter-insurgency campaign by stating in Pristina in February 1998 that the KLA "is without any question a terrorist group". However, in mid-June, responding possibly to a request by Moscow, Serb forces moderated their actions and the KLA took advantage of Serb passivity, seizing control of large swaths of territory. Even at this stage though, there is a sense in which both the military and political significance of the KLA was externally defined by both western politicians and the media. As one British official told the House of Commons Defence Committee,

What really happened was that the Serbs were not really contesting certain areas. What they had not realised was that what in effect had been a vacuum had been filled by these people who were getting outside commentators to look, so they could walk around say "We are an army".

Nevertheless, responding to what appeared to be the new reality on the ground, the US dropped its resistance to including the group in an Albanian negotiating team, recognising that as a party to the conflict it needed to be included. The first formal US contact with the KLA took place at the end of July when the US diplomat Christopher Hill met with KLA representatives - a day later a British official also met with them. Emblematic of this new shift in Western attitudes was the position of Bob Dole, previously a presidential candidate and an advocate of the Albanian cause. In May of 1998 Dole could describe the KLA as a terrorist group to a US Congressional Committee, yet by June, in evidence to the same committee, he considered them to be beleaguered 'rebels', who

have no choice but to continue to defend their people. And I would assume there are some KLA membe[r]s who are, while I would not say terrorists, are capable of terrorist-like acts.

Indeed, once NATO went to war with Serbia the KLA was totally reimagined, with Madeline Albright insisting to the same committee that 'we do not consider them a terrorist group...We consider them the military arm of the Albanians'. As will be noted below, it is this latter view of the KLA that persisted through to the negotiations on demilitarisation.

However, whilst there was a new recognition in the summer of 1998 that negotiations with the Albanians had to include the KLA, this new attitude did not extend to including either them or any other Albanian group in the ceasefire agreement reached with Milosevic in October of that year. Indeed, they were not even informed of the content of the military undertakings or the precise nature of the overall package. Consequently, the cease-fire was a one-sided affair, it did not formally require anything from the KLA and more to the point the KLA did not feel particularly bound by the agreement. Thus, whilst the Serbs, initially at least, broadly abided by the agreement, the KLA took the opportunity to fill the void left by withdrawing Serb forces.As one KLA representative has noted "the cease-fire was very useful. It allowed us to consolidate and grow".

Moreover, the possibility that KLA 'terrorists' would move into territory vacated by Serb forces had been Milosevic's main concern during the negotiations, and he had not only been assured that compliance on Serbia's part would place the onus for any violence on the KLA, but that NATO would try to control the organisation. It was even suggested that the OSCE mission sent in to verify commitment to the agreement (The Kosovo Verification Mission (KVM)) could do this. Albania did come under pressure to curb arms smuggling to Kosovo, but it was hostile to the idea of introducing an international force which would monitor its border region with Kosovo. The failure to take action to limit the KLA's operations left Milosevic with a sense of betrayal that would cloud future policy. As one Belgrade commentator noted,

He [Milosevic] thought the US would close the border with Albania to prevent arms smuggling, that the US would freeze the KLA's assets and make arrangements to terminate the KLA's influence...When Milosevic understood that Holbrooke would not fulfil such a 'promise' he went for war'

Moreover, the 200 strong OSCE Kosovo Verification Mission (KVM) created to observe the ceasefire was slow to form. Although it formally began on 25 October 1988, by December the KVM had only 50 staff, and even by the time of its withdrawal prior to NATO military action, its strength had still only reached two thirds of the number envisaged. As the House of Commons Foreign Affairs Committee have noted, 'deployment of the OSCE verifiers was an important part of the Holbrooke agreement, and...the international community did not display sufficient seriousness or urgency in fulfilling this side of the argument.' In addition, the mission consisted of unarmed monitors who could do little more than report violations of the ceasefire. In a context where one of the main combatants was not even a party to the agreement and where there were question marks over the good faith of the other, the wisdom of this approach is debatable. As Daalder and O'Hanlon have noted, however, US reluctance to risk troops on the ground precluded the deployment of a more muscular force able to provide security to all threatened parties and thus mute the security dilemma confronting them. The British, for instance, had already concluded that ground troops would be needed and there was an expectation, at least in some quarters of NATO, that the Holbrooke agreement would result in the deployment of a force similar to SFOR in Bosnia. Western intelligence had even picked up indications that Milosevic would not reject such a proposal out of hand.

Although the cease-fire more or less held for two months, increased KLA activity (including continued kidnappings and violence against local Albanians) prompted a predictable Serb response, and as the situation in Kosovo deteriorated the Contact Group now summoned the parties to the Kosovo conflict to Rambouillet to work out a peace deal.

Prior to Rambouillet, the Kosovo delegation that had commented upon the draft that would form the basis for discussion had consisted of individuals primarily nominated by the elected government of president Rugova. As Robert Gelbard noted in May 1998, the US view was that 'Dr Rugova, as the person who has been elected by about 85% of the Kosovar Albanian population, is the legitimate representative of the Kosovar Albanian people'. In the run-up to Rambouillet, however, the US now began to press for a delegation that would include the KLA. This was certainly necessary if a meaningful agreement was to be reached. However, it was also reported that at a meeting in November 1998 KLA leaders had traded a commitment not to widen the conflict to Macedonia and to avoid contact with Islamic extremist groups for involvement in the diplomatic process. Whatever the rationale, the Albanian delegation at Rambouillet was broadened so that it consisted of roughly one third LDK representatives, one third LBD (the more radical United Democratic Movement) and one third KLA, plus two independents. The effect was the creation of a delegation in which the KLA and its supporters now constituted a majority leaving the government, which had been elected by an overwhelming majority of the Kosovo population, in a minority.

Moreover, whilst the composition of the negotiating parties was being extended to bring in the military actors on the Albanian side, it was, in other respects, being de-limited. For instance, the Serbs arrived at Rambouillet with representatives from the various minority communities in Albania (Roma, Turks etc) who were widely perceived to be stooges and who were disowned by the ethnic groups they were supposed to represent back in Kosovo. Yet no comparable attempt appears to have been made to draw in legitimate representatives from these groups, who surely had as much a stake in the ultimate outcome of Rambouillet as any of the parties. Similarly, on the Serb side, Bishop Artemije (a member of the relatively moderate Serbian Resistance Movement) turned up at Rambouillet but was refused entry to the conference. In contrast, other peace negotiations have been notable for the way in which they have included representatives from civil society. For instance, negotiations over the Good Friday Agreement in Northern Ireland included the Northern Ireland Women's Coalition, and womens groups also participated in the Arusha Peace talks on Burundi.

Of course, within the logic of the conference this is not surprising. At best it was conceived as an archetypal great power exercise aimed at restoring order rather than establishing a process that could build towards conflict resolution. For a conference conducted within the paradigm of high politics diplomacy, any actors other than a narrow group of political elites from the conflicting parties were essentially irrelevant. At worst, critics have argued that Rambouillet was designed to fail in order to legitimise the subsequent bombing of Serbia. However, succeed or fail, Rambouillet was always going to be part of an ongoing process which, at some point, had to promote peacebuilding in Kosovo. In contrast, it had the effect of legitimising the political power and the voice of the KLA whilst marginalising the politicians and wider civil society. Whatever kind of Kosovo was to be built after Rambouillet, this offered a problematic precedent which would be further entrenched in the demilitarisation negotiations after the war.

Arms Control and Demilitarisation in Kosovo

There had already been some arms control initiatives undertaken or attempted prior to the NATO action against Serbia. An arms embargo had been imposed on Yugoslavia in 1992 as a consequence of the conflict in Bosnia. However, this was lifted in 1995 with the signing of the Dayton agreement. In contrast, the US blocked Yugoslavia's return to the World Bank and the IMF, citing the issue of Kosovo as one reason for maintaining an 'outer wall' of sanctions. In March 1998, with conflict escalating in Kosovo, the UN Security Council imposed a comprehensive arms embargo against the FRY including Kosovo. This included restrictions on the sale of arms and related material of all types and restriction on the provision of arms and training for terrorist activities. This was followed in July by a statement from the Contact Group urging all those supplying financial support, arms or training for what were now termed 'armed Kosovo Albanian groups' to desist and further requested all states to prevent funds collected on their territory from being used to contravene the UN arms embargo. In addition, as part of the diplomacy leading up to Rambouillet, NATO issued a statement which noted that it was 'studying how to support measures to curb arms smuggling into Kosovo'. However, this does not seem to have been followed by any concrete action and, as one diplomat has noted, "whilst Kosovo was under the cosh there was no mileage in getting at the arms flow problem". Indeed, as US policy shifted towards support for the KLA in 1998 and '99, CIA operatives working in the OSCE verification mission reportedly provided US military training manuals and field advice on fighting the Yugoslav army. As the monitors left in the period immediately preceding the start of the NATO bombing campaign, they also handed over satellite telephones and global positioning systems so that the guerrillas could stay in contact with NATO and Washington. Indeed, several KLA commanders reportedly had the mobile phone number of the NATO commander General Wesley Clarke.

After Rambouillet, of course, there was much debate over whether the military annexe of the agreement was structured in such a way as to engineer a rejection by Milosevic that would then provide a pretext for NATO bombing. Less attention, however, has been given to the demilitarisation requirements imposed on the KLA by Rambouillet, and it is these that are more pertinent to the discussion here.

The military annex was not actually presented to the negotiating parties until the day before the initial deadline before agreement, although in an act that would raise suspicions amongst the Albanian delegation, the whole peace plan, including the security annex, was actually presented in separate negotiations with Milosevic in Belgrade by the US diplomat Christopher Hill. The delay in presenting the security annex was partly due to the fact that the Contact Group itself was divided on the question of military implementation, and partly because it had been previously announced that no changes to the proposals for implementation would be countenanced anyway. As a number of commentators have suggested, this delay was, in retrospect, a mistake.

The parties had received one briefing on military implementation before the security annex was presented to them but, according to Weller, this had left the KLA and its supporters in a state of shock and did not appear designed to reassure the military elements of the Kosovo delegation. However, an effort was now made to address the concerns of the KLA over the details of the agreement. This included briefings by military experts and by General Wesley Clarke, NATO Supreme Allied Commander Europe, through which 'the KLA was to be assured as to the actual meaning of demilitarisation in practice'. Madeline Albright also offered the Kosovars a form of security partnership by promising to assist with the transformation of the KLA from a military to a civil force. What this meant in practice was left, in public at least, vague, although Albright did note that the US was prepared to help retrain qualified members of the KLA who wished to join the new multi-ethnic Kosovo police service proposed under Rambouillet.

At the same time as the KLA was being offered a variety of incentives to persuade it to accept the demilitarisation requirements of Rambouillet, it was also being threatened with reprisals if it did not. These included steps by NATO states to cut off funding from the Albanian diaspora, the re-labelling of the KLA as a terrorist organisation and pressure on the Albanian government to close logistical and training facilities. It was even threatened with the possibility that NATO might patrol the Albania-Kosovo and Macedonia-Kosovo borders in an attempt to inhibit the supply of weapons and troops. In other words, the KLA were now being threatened with precisely the kind of initiatives that Milosevic had expected when he signed up to the Holbrooke Agreement. As Weller notes,

...if implemented, this would have significantly reduced the effectiveness of the KLA as a fighting force. In fact, the organization would probably have had to abandon its ambition to function as an organized underground army capable of engaging Yugoslav military and paramilitary forces.

Whilst there had been some discussion about the possibility of requiring the KLA to disarm, ultimately the final agreement was one under which it was required to 'publicly commit to demilitarise on terms to be determined by COMKFOR [and] renounce violence'. It was also required to refrain from committing any hostile or provocative acts or smuggling across international boundaries. By K + 1 day, what was termed in the agreement as 'other forces' (in other words forces other than KFOR, VJ, MUP or local police forces) were required not to carry weapons within 1 km of VJ and MUP cantonments and also designated main roads and border areas. After 30 days they were obliged to have placed all prohibited weapons and automatic weapons in storage sites, to cease wearing military insignia and to cease carrying prohibited weapons. After 90 days, authority for the storage sites was to pass to KFOR and it would be illegal to posses prohibited weapons and automatic weapons. After 120 days demilitarisation was to have been completed, although the agreement did not stipulate any additional measures that had to be taken after the 90 day deadline. There was also a provision in the document for the creation of communal police units. These would total some 3,000 personnel and there was a requirement that the national communities in each commune would be fairly represented. Communal police officers were allowed a sidearm and baton and one long-barrelled weapon for every fifteen police officers, the latter to be securely stored at communal headquarters or at municipal stations. An earlier draft of the agreement had also included a provision for the establishment of a Kosovo Security Commission and Communal Security Commission which were to be forums for co-operation, co-ordination and the resolution of disputes concerning law enforcement and security in Kosovo, and whose functions included the ability to consider complaints regarding police practices filed by individuals or national communities. The membership of these commissions was to be broadly based. For instance, the Kosovo Security Commission would consist of a representative of each commune and a representative of each national community, as well as representatives from the FRY and the international community. The final text, however, made no reference to such commissions. It did propose to establish a Joint Military Commission to address any military complaints, questions or problems that required resolution, such as allegations of cease-fire violations. This would be chaired by the Commander of KFOR and would include representatives of the Kosovo civilian leadership as well as representatives of the FRY and KLA forces.

The refusal of the Serbs to sign the agreement, however, meant that the arms control provisions of Rambouillet were never implemented, and instead NATO began the bombing of Serbia on March 24th. The NATO bombing was finally halted by the Ahtisaari-Chernomyrdin-Milosevic agreement of 2 June which became known as the Kosovo Accords of 4 June. This agreement was supplemented by the Kosovo Military-Technical Agreement of 9 June between Yugoslavia and NATO and UN Security Council Resolution (SCR) 1244 of 10 June. In contrast to Rambouillet, which gave the Serb forces 180 days to withdraw from Kosovo, the Military-Technical Agreement only allowed 11 days for them to complete the withdrawal of some 20,000 soldiers, 400 tanks and over 200 artillery pieces, a deadline they achieved with eight hours to spare.

Negotiations with the KLA over their post-war demilitarisation had begun in early June and were led by the commander of AFOR (Albania Force). Madeline Albright had already secured a commitment on 8 June from three of the main Kosovar leaders (Hashim Thaci of the KLA, Ibrahim Rugova of the LDK and Rexep Qosja of the United Democratic League of Kosovo) that the KLA would demilitarise and that KLA forces would not fire on withdrawing Serbs. Two days later UNSC 1244 also demanded that the KLA and other armed Albanian groups 'comply with the requirements for demilitarisation as laid down by the head of the international security presence in consultation with the Special Representative of the Secretary General' (a day after this Sergio Vieira de Mello of Brazil was named as Acting Special Representative for Kosovo). From this point on, however, the negotiations leading to the KLA's Undertaking on Demilitarisation were conducted exclusively between the KLA and NATO representatives.

In contrast with (and possibly as a consequence of) the intense diplomatic effort put into the drafting of UNSCR 1244, there seems to have been far less thought given to the modalities of the KLA's post-war demilitarisation either by major Western governments or by NATO itself. UNSCR 1244 required the demilitarisation of the KLA, but beyond that, as the Commander of KFOR has noted, until an agreement was reached with the KLA 'I had no policy, I had no directive as to how to treat the UCK'. There was then a sense in which Western policy, particularly with respect to the KLA's transformation, was being made-up on the spot by those on the ground.

Although some exploratory contacts had been made in early June, the first formal meeting on the question of demilitarisation took place on June 14th (5 days after the signing of the Military Technical Agreement by the Serbs) with Xhavit Haliti, one of the original founders of the KLA who was flown from Slovenia. This meeting focused principally on the mechanics of arranging discussions with Agim Ceku, the leader of the KLA's military wing. These took place in Albania on June 15th after Ceku had been collected from the KLA's General Command HQ in Kosovo by an SAS helicopter. A further meeting was held in Kosovo on 17 and 18th June with Ceku, Hasim Thaci and all the KLA's zone commanders. James Rubin from the US State Department was also present. These negotiations were apparently difficult with a concession on one aspect of the agreement encouraging the KLA to try and gain further concessions on aspects already settled. Finally, after much haggling, the negotiators took Ceku and Thaci outside and impressed upon them that the bottom line had been reached in terms of concessions to the KLA. A text would be drafted and a deadline of 1600 on June 20 was set for the KLA to agree to the draft text that had been produced. Eventually, an Undertaking of Demilitarisation and Transformation was formally signed by the KLA at ten minutes past midnight on 21 June, one week after NATO troops arrived in Kosovo and after the completion of the Yugoslav withdrawal.

The semantics of international diplomacy meant that even the title of this document had significance. It was not, formally at least, an agreement between NATO and the KLA. It was, rather, a unilateral undertaking on the part of the KLA which was received by NATO. This was of importance to both parties. For the KLA it was important to be able to demonstrate that, as victors, they were disarming voluntarily and without being pressured by NATO. For NATO it meant that the organisation was not entering into a formal agreement with either an unrecognised political entity or a group with a rather shady reputation. In practice, if not in law, the Undertaking was an agreement. Hashim Thachi, the leader of the KLA who signed the agreement, was immediately rewarded with phone calls from Bill Clinton and Madeline Albright hailing his political courage.

There were a number of issues. First, as under the Rambouillet text, the process of demilitarisation was to be undertaken in phases. Most notably, the KLA, with immediate effect undertook not to carry any weapons of any type within 2km of specified main roads and towns and the external borders of Kosovo. After 7 days automatic small arms weapons not placed in secure storage sites were only to be held inside authorised KLA assembly areas. Within 30 days the KLA was required to have handed in all prohibited weapons, with the exception of automatic small arms. Of the latter, 30% were to be handed in after 30 days, 60% after 60 days and the rest within 90 days (apparently the KLA had declared a figure at the outset of negotiations). All KLA were also required to cease wearing military uniforms or the insignia of the KLA after 90 days. Notably, then, whilst under Rambouillet KLA demilitarisation was to be completed after 180 days, they were essentially provided with a much more leisurely timetable to complete the key elements of demilitarisation. Indeed, they were allowed an additional two months before all their weapons had to be in secure sites and before they were required to cease wearing their uniforms. Press reports have suggested that General Jackson, the KFOR commander actually wanted a faster timetable in order to limit the exodus of Serbs, but James Rubin who, as already noted, attended the final stages of the talks with the KLA, had urged that concessions be made to the KLA. Others, however, have suggested that Rubin merely attended as an observer and contributed little to the negotiations themselves. Whatever the case, it is clear that during the course of negotiations with the KLA the original Rambouillet model of 30 days was extended.

It has also been argued subsequently that given the multiple demands placed on KFOR during its entry into Kosovo, the 90 day deadline provided NATO forces with a more realistic timescale for overseeing the demilitarisation of the KLA. Also, in contrast to Rambouillet, the KLA were being asked to hand in their weapons against the backdrop of what had been a vicious intensification of Serb abuses in Kosovo, and in a context where their numbers had grown quite substantially. Moreover, whereas Rambouillet would have been a negotiated agreement (albeit one with the threat of military action hanging over the Serbs) the Serbs had now been forcibly evicted. Under these circumstances there were concerns at the time that some Serb forces might have stayed behind with the aim of attacking either KFOR, the KLA or civilians. Nevertheless, given that the agreement was only signed once Serb forces had departed, given the concern about KLA reprisals against Serb civilians, and given the political symbol that a tight deadline would have sent to the KLA, the wisdom of affording the KLA two months more than under the Rambouillet agreement is debatable.

Demilitarisation vs Disarmament

Much was made of this issue in the press, although some officials have suggested the distinction between the two terms was not much of an issue. According to press reports, whilst normally demilitarisation might be considered to be a more far-reaching requirement than disarmament, in the context of the negotiations, this was not the case. Demilitarisation was a vague and unspecific term whereas disarmament implied the complete surrender of KLA arms. The KLA's defence minister neatly summed up their view on this issue: "to demilitarise an army means just to reduce the size of forces that exist". Ultimately, the KLA were merely required to demilitarise. Of course, the provisions for weapons hand-in outlined above did require specific disarmament initiatives from the KLA, and by the time of the 90 day deadline the KLA had handed in: 9000 small arms, over 800 machine guns, 300 anti-tank weapons, 178 mortars over 27,000 hand grenades, 1,200 mines, 1000 kg of explosives and nearly 5.5 million rounds of ammunition. However, following the experience of other post-conflict disarmament agreements, the equipment handed over was often either broken or of poor quality. There is also widespread recognition amongst all parties in Kosovo that the KLA still retains a large number of arms. Indeed, the number of small arms handed in contrasts sharply with the number of former KLA registered for demobilisation which now stands at over 25,000. It also contrasts with Hedges estimate of 30,000 KLA arms and the view of one UN official who noted on 20 September, when the KLA's demilitarisation was officially deemed to be complete, that "we are satisfied ...the UCK has handed in 10,000 arms even if we know they still have 100,000". Perhaps the most accurate description of the situation with regards to the number of KLA weapons still in circulation was provided by Mike Jackson, the now-ex-commander of KFOR who has noted :

'The inevitable question is how many weapons are out there which were not declared. Frankly, I have no idea and the UCK probably doesn't either [although it probably has a better idea than Jackson]. There are lots'.

It is certainly the case that KFOR is still finding weapons caches in Kosovo. Indeed, the largest to date was uncovered on 21 June 2000 - exactly a year after the Undertaking of Demilitarisation was agreed - when four bunkers of weapons were discovered. The first two bunkers alone were estimated by KFOR to contain enough weaponry to fully outfit two heavy-infantry companies, eliminate the entire population of Pristina and destroy 900-1000 tanks.

For KFOR, however, complete disarmament of the KLA was of less concern than establishing a process of demilitarisation. This was for a number of reasons.

First, the importance of the demilitarisation agreement was not in getting the KLA to abandon all their weapons, an unrealistic goal, but the act of political symbolism implied by the KLA not only signing up to, but visibly adhering to the disarmament provisions contained within the demilitarisation agreement. In essence, the fact that the KLA made a commitment, and delivered on it, was taken to represent a sign of its good faith vis à vis KFOR and its willingness to begin the process of transition to normal politics.

Second, it was also the case that obfuscation on the issue reflected disagreement amongst the main Western powers over the long-term future of Kosovo. As Woodward has noted, for at least some Western officials 'the Yugoslav forces remain the primary threat. The possibility that they could re-enter the province and take control after KFOR departs should be central to the current programme. For this group, security requires deterrence, which a KLA force exposed to Western values and trained to NATO standards is best positioned to provide'. Thus, retaining the option of a local army represented a potential exit strategy for KFOR.

Third, the compromise on demilitarisation (as indeed other compromises, for instance on the KPC) was partly a function of the fact that the Undertaking was on the one hand a document that, ideally, was required to take account of a number of different security relationships at the same time: one between the KLA and the rest of Kosovar society (including local Serbs, other minority communities in Kosovo and non-KLA Albanians), one between the Kosovo Albanians and Serbia and one between the KLA (along with its supporters) and KFOR. On the other hand, for KFOR, the main concern was with these latter two relationships, particularly with the way in which they impinged on the security of KFOR troops.

KFOR were of course liberators who had defeated the Serbs and, in this sense, the allies of the KLA. However, the KLA, and most Albanians in Kosovo, not only wanted independence from Serbia (an option officially precluded by the West) but they also faced the possibility that KFOR might ultimately depart leaving them to the mercy of the Serbs once again. An overly intransigent approach on the part of KFOR thus had the potential to create a very real security dilemma for the KLA, one which might only be resolved by the latter taking up arms against those trying to disarm them. Consequently, for KFOR, once it crossed into Kosovo the KLA turned from an informal military ally to a potential security threat and one over which its leverage had arguably decreased. There was thus a delicate balance to be struck between developing an approach which successfully controlled the KLA's latent military threat, and one which would so alienate the KLA as to risk them turning their guns on KFOR soldiers. For an Alliance that had just fought a post-heroic war governed by the need to avoid military casualties, the prospect of asking KFOR troops to become neo-heroic peacekeepers was not enticing.

Fourth, for KFOR, the concern was not with the number of weapons still circulating but with developing a domestic arms control regime that would control their use. This involved the confiscation of openly carried weapons, demobilisation of the KLA, strict control over the arms held by the Kosovo Protection Force and the police (see below) and an attempt to establish a secure environment for the citizens of Kosovo.

Finally, the compromise on demilitarisation essentially represented a recognition of the realities of the contemporary market, particularly in the Balkans. Whilst the hurdle set for the KLA to demonstrate their compliance with demilitarisation (the hand-over of some 10,000 arms) was arguably less rigorous than it might have been, the essential principle adopted here was certainly correct. As Susan Woodward has noted, whilst

the international decision to allow obfuscation with the term demilitarisation ...will create serious political trouble down the road ...the decision not to disarm the KLA was a pragmatic, even prudent choice. Given the circumstances, such as KLA networks abroad, the well-established and well organised smuggling routes, the contest within Albania over support for the KLA, and the failure of previous regimes in the area to disarm the Albanian population of Kosovo disarmament would never actually succeed.

Or, as one Western official put it more succinctly, "you'll never police the border". The difficulties involved are perhaps well illustrated by the case of a 200 strong FARK unit which was sent into Kosovo after the signing of the Undertaking of Demilitarisation. These were disarmed by the Albanian army before entering Kosovo, only to be discovered with more weapons (presumably obtained from an arms cache within Kosovo) by KFOR troops who once again disarmed them. Despite this, the unit had to be disarmed a third time by the KLA.

Thus, whilst the hurdle set for the KLA to demonstrate their compliance with demilitarisation (the hand-over of some 10,000 arms) was arguably less rigorous than it might have been, the essential approach here was certainly correct. The problem was that for this approach to work it needed to be supplemented by strict application of the domestic arms control regime outlined above and also a regional approach designed to prevent the flow of weapons into Kosovo. In addition, those elements of civil society more inclined to peace than violence had to be empowered. As will be noted below, the fact that this approach has been implemented in a half-hearted manner has contributed to the continuing violence and lawlessness in Kosovo today.

The Kosovo Protection Corps

The debate about demilitarisation vs disarmament was intrinsically related to the concern of the KLA to retain an army. This was justified on the grounds that the Albanians needed a deterrent force against the Serbs as an insurance against a possible KFOR departure at some point in the future. As Hashim Thaci told the Pristina daily Koha Ditore in mid-June (prior to the signing of the Undertaking), 'The KLA will be transformed into an army ...that will defend the territory of Kosova'.

It was also related (as is everything in Kosovo) to the Albanian desire for an independent state. Armies, like flags, are important symbols of state-hood and their possession, irrespective of their effectiveness, thus represents an important affirmation of sovereignty. Thus, whilst formal Western policy precluded recognition of an independent Kosovo, the retention of anything that could be remotely labelled as the nucleus of an army both resolved a security concern for the KLA and represented a further step along the path towards its ultimate goal of independence.

For KFOR, as already noted above, there were a number of factors to take into account. On the one hand, allowing the KLA to retain an army at the very least muddied any future negotiations over Kosovo's ultimate status, and at the very worst could be taken to imply a form of de facto independence for Kosovo. On the other hand, if Kosovo was ultimately to be rewarded with independence, then retaining the nucleus of an army represented one component of a future exit strategy for Western forces. Similarly, allowing the KLA to retain even a proto-army posed a potential long-term threat to KFOR, yet insisting on complete disarmament risked provoking a confrontation with the KLA, a problem neatly illustrated by one KLA commander, who noted in August (a month after the Undertaking had been signed) "If two months from now. KFOR tells us, this is it, from now on you have no army, no national guard, nothing, then they are making a mistake. We will build up an army and we will fight against KFOR. Our people did not die to let the Serbs come back. If that happens we will attack everywhere, starting with the Russians".

Negotiations on this issue centred around the KLA's aspiration for a civil emergency corps, on the lines of the US National Guard, which could then form the basis of an army of national defence. The National Guard issue was a subject of contention and the para in the agreement dealing with it was reportedly taken out at one stage, only to be put back in because of its importance to the KLA. In the end the Undertaking noted

...that the international community should take due and full account of the contribution of the UCK during the Kosovo crisis and accordingly give due consideration to ....the formation of an Army in Kosovo on the lines of the US National Guard.

At a press conference after the signing of the Undertaking, James Rubin suggested that the para merely represented the aspirations of the KLA (which, given the wording of the document and its status as a unilateral undertaking, was technically correct), but the KLA were more forthright on the issue. Hashim Thaci declared the KLA would be transformed into a Kosovo National Guard, and a KLA spokesman noted 'in peacetime we need a regular army in the barracks and this National Guard will be defensive in nature'.

Further discussions on this issue occurred up until September when the ratification of KLA demilitarisation and the agreement on the establishment of what was labelled as the Kosovo Protection Corps (KPC) was delayed for 48 hours by disputes over the weapons regime for the KPC, over whether the KLA could retain their uniform and red and black emblem and over the name itself. On the latter issue, for instance, the KLA argued for the more military sounding label of 'Kosovo Army' or 'Kosovo National Guard'.

For the various international actors in Kosovo the creation of the KPC had two main advantages. First, it was the principal mechanism by which the KLA would be drawn into a domestic arms control regime. As one Western official in Kosovo noted, "the Jackson argument is that the National Guard is good because you can keep an eye on them rather than having them walk away in the hills". Indeed, at one point during the negotiations that took place on the creation of the KPC, the KLA did walk out and announce they would return to the hills. The Corps was to be allowed 2,000 weapons of which 1,800 would be 'held in trust' in KFOR secure weapons facilities. The reminder would be available for the guarding of installations and security when units were deployed. Additionally, sidearms could be carried by authorised officers and close protection units for key members of the Corps. The KPC was strictly defined as a civilian emergency force under the authority of UNMIK, with no role in defence, law enforcement or internal security. Its role was to provide disaster response services, humanitarian assistance, perform search and rescue and contribute to rebuilding infrastructure and communities. Under the agreement members of the Kosovo Protection Corps were precluded from holding public office or from actively engaging in political affairs. It was to be limited to 5,000 members of which 3,000 would be permanent and the remainder reserves (although this was still a larger force than strictly necessary for the purposes envisaged). All members of Kosovar society were eligible to join but 'a significant portion [would] initially come from the leadership and ranks of the demilitarised UCK'. Thus, not only was any future force subject to verifiable limits, but the creation of the KPC could be understood as an integral part of a demobilisation process designed to create employment for former KLA who might otherwise return to violence or crime, as had been the experience in other post-conflict demobilisation processes.

Second, it can also be argued that the creation of the KPC, along with other initiatives in Kosovo, has the benefit of giving the KLA a stake in the political system created in Kosovo and thus provides the West with a source of leverage over the KLA which it did not originally have.

Equally, however, despite the formal dissolution of the KLA, there still remain a number of problems related to the creation of the KPC. First, the KPC, formally at least, is supposed to be an entirely new creation open to all residents of Kosovo. In the words of the UN, the official constitution of the KPC on 21 January 2000 meant that 'the Kosovo Liberation Army (KLA) was transformed into a civilian agency designed to provide emergency response and reconstruction services to Kosovo'. However , as the ICG have noted,

...no one seriously believes that the KPC is anything but a new manifestation of the KLA, inheriting its leaders and loyalties...Even UNMIK's own officials and some KFOR officers admit (though never in public) that the KPC is, and will probably remain a military style organisation.

Indeed, even the Albanian title of the KPC (Trupat Mbrojtëse të Kosovës (TMK)) reflects this, as the phrase Mbrojtje' can mean protection but it can also mean defence, an ambiguity that General Jackson was well aware of when negotiating the agreement to set up the KPC. In addition, whilst the old KLA uniform has been outlawed, the new KPC one looks very similar. Even the organisational structure of the KPC draws directly from the KLA, with former zone commanders servings as Regional Task Group (RTG) commanders, and their staff being drawn from their original units as well.

The creation of the KPC also undermined the faith of the Serb community in the impartiality of KFOR and UNMIK. Indeed, the two Serbs serving on the UN Transitional Council withdrew from the body in protest at the establishment of the KPC, arguing that the transformation of the KLA into the KPC had violated the declared multi-ethnic nature of Kosovo. Some Serbs were prompted to put forward the idea of a rival Serb Protection Force, an idea rejected by the UN who could point to the fact that 10% of the positions are to be filled by civilians and minorities. However, by October 2000 only about 100 posts had been filled. Partly this is a consequence of the scepticism of those communities about the KPC, with just 96 out of the 20,000 KPC applications coming from minorities. However, there is also evidence of intimidation. For example, in October a local UN official, whose jobs was to help recruit minorities into the KPC, was attacked and beaten.

The KPC have also been reluctant to accept the de-induction of the 2,000 members initially scheduled to become reservists in April and which still had not occurred by the time of writing in October. Whilst this probably will happen after the up-coming local elections, it is indicative of the fact that many in the KPC (and in Kosovo generally), have still not resigned themselves to a vision of the KPC as a civil emergency force.

It is also the case that, ironically, the funding for the KPC is both too much and too little. The operating, training and equipment costs of the KPC are funded by voluntary contributions from UN Member states. The total budget for the financial year 99/2000 is 20.7 million DM. In contrast, the total budget for UNMIK is 562 million DM (about $299 million) of which some 221 million DM will come from donors (the rest being made up through the collection of local taxes etc). Thus, 9.3% of donor funding for the UN administration in Kosovo has been earmarked for the KPC. If the UN administration in Kosovo was well funded this would not be a problem, but it is not. Instead UNMIK has developed a reputation as a near insolvent operation barely able to pay its bills or to provide adequate policing, basic services, and reasonable levels of pay to public employees. Moreover, the lack of resources available for other areas of UNMIK's operations is an issue that also has implications for the success of arms control. UNMIK's inability to provide for the civilian population has an impact on its perceived legitimacy and authority which, in turn, as Eide and Nustad have noted, has a crucial bearing on the prospects for successful domestic arms control. Just to give one example, the pensions of Serbs in Mitrovica are roughly double those paid by the UN, as the Serbs receive extra money from Belgrade. If security is to be established in Mitrovica, then clearly a pre-requisite is for local Serbs to develop a sense that they have a vested interest in the promotion of good relations in Kosovo and in working with the UN. Indeed, if the argument that the KLA needs to be drawn into domestic institutions in order to provide leverage over them has any validity, then so does the argument that the civilian population as a whole also needs to develop a stake in these same institutions. At present, the limits on UNMIK's funding has meant that one arms control initiative is actually undermining UNMIK's ability to devote funds to long-term structural arms control. It should also be noted that where UNMIK has been able to improve services, it has sometimes had favourable effects on the level of violence in Kosovo. For instance, when the electricity in Pristina was re-established, the murder rate dropped from roughly 30-40 per week to the high twenties.

However, having created the KPC as a vehicle for demobilisation, it is equally the case that funding from the international community has been intermittent and salaries are low. For instance, the average salary paid to members is just 292 DM per month. Whilst this is slightly higher than the average for other public employees, it is hardly a level of payment designed to discourage a resort to corruption. Indeed, many members of the KPC already engage in other economic activities both licit and illicit to supplement their income. In addition, KPC funding actually threatened to dry up completely in the summer. Whilst external donors eventually came up with more money, the situation has created disillusion amongst some members of the KPC. Indeed, it is notable in this context that the KPC has suffered from a significant drop-out rate. Whilst there are probably a variety of factors which explain this phenomenon, the funding problems and poor salaries certainly do little to allay this trend. Moreover, the Commander of the KPC, Agim Ceku, has in the past expressed his preference for all the 20,000 applicants to the KPC to be given training on a rotating basis. Whilst this is not going to happen, it is equally the case that if the high drop-out rate continues, it does risk creating, albeit by default, a third group of what might be termed unofficial KPC reservists.

A further issue relates to the role of the KPC in post-conflict Kosovo. It is becoming increasingly clear that former KLA, including members of the KPC, have been involved in both ethnically motivated violence and more conventional criminal activity. For instance, the introduction to the OSCE's report on human rights abuses in Kosovo in the period June to October 1999 notes

A consistent reporting feature has been assumed UCK presence and control. The report is littered with witness statements testifying to UCK involvement, both before and after the demilitarisation deadline of 19th September ranging from reports of UCK 'police' to accusations of intimidation by self-proclaimed members of the provisional Kosovo Protection Corps (TMK).

More recently the ICG has commented on the proclivity of KPC regional commanders 'to carve out local fiefdoms for themselves' and rumours of covert military training for KPC members in the town of Skenderaj. In August 2000 the UN Secretary General reported 164 cases of KPC non-compliance with UNMIK standards since February and noted that 'the frequency of weapons violations and incidents involving threatening behaviour remain at a significant level'. In one incident, for instance, a member of the KPC shot at police officers after he had asked them to let him off a traffic violation because he was part of the KPC and they had refused. Indeed, The Observer newspaper in the UK has claimed that a confidential report written for the UN Secretary General identified the KPC with criminal activities, killings, ill-treatment, torture, illegal policing, abuse of authority, intimidation, breaches of political neutrality and the running of protection rackets across Kosovo. Similarly, when reports appeared suggesting that Ceku, the leader of the KPC, was under investigation for war crimes committed in Croatia, one UN diplomat could note; "if we lose him it will be a disaster. When you get to the second level of the TMK, you're down to a bunch of thugs".

Moreover, it may be that some military units of the old KLA still exist outside the KPC entirely. In particular, towards the end of 1999, a KLA 'splinter group', the Liberation Army of Presevo, Medvedja and Bujanovac (UCPMB) began operating in a region of Serbia known to Albanians as 'Eastern Kosovo'. These guerrillas have engaged in paramilitary activities including mortar attacks against Serb police and attacks against those labelled as 'collaborators' with the Serb regime. This has resulted in an exodus of Albanians into Kosovo, fleeing both the activities of the UCPMB and the resulting Serb response. KFOR troops in the area have now started to conduct raids on the hideouts of this new group. In one raid, for example, they seized 200 uniforms, 22 crates of rifles and machine gun ammunition, two mortars, 28 hand grenades, seven rifles, six landmines and other military supplies. Reports from local villagers suggest that this new group is a local branch of the KLA receiving backing from inside Kosovo.

Law and Order

A further crucial element in any process of demilitarisation and peacebuilding is the provision of law and order. In particular, it is important to establish a secure environment so that a cycle of revenge and hatred is halted and so that those elements of society more willing to contemplate reconciliation can begin the process without fear of intimidation. However, Kosovo has experienced serious flaws in the provision of security and enforcement of the law.

The international community's approach to the provision of policing in Kosovo has been usefully characterised by the ICG as a three stage strategy: 1. soldiers acting as police, gradually handing over to 2. UN police drawn from overseas gradually handing over to 3. an indigenous police force recruited by the UN but trained by the OSCE. Unfortunately, there have been flaws in the provision of policing at every one of these stages.

First, a number of local and outside observers have criticised many (though not all) KFOR contingents for a lack of robustness in their approach to crime and ethnic violence. For instance, Human Rights Watch have complained that concerns about the safety of its own troops, coupled with a shortage of available personnel, have rendered KFOR both unable and unwilling to take the initiatives necessary to build confidence, particularly amongst the minority communities now facing persecution for their perceived complicity in the victimisation of Albanians. For example, French troops in Mitrovica have been criticised for relying mainly on guards at a few key points to provide security, rather than engaging in more regular foot patrols, whilst the US has refused to allow its troops to be used in Mitrovica for the fear of incurring casualties.

The situation with respect to the UN police is little better. The deployment of UN police was painfully slow, with only 210 pledged by various governments by mid-July 1999. As one French Colonel noted in June '99, the absence of both UN police and a functioning judiciary meant "we cannot detain anyone for more than twenty four hours, so we just took the looters names and let them go the next day". Even when they have arrived the numbers supplied have remained pitifully low. An initial assessment made in June had estimated a requirement for 3,100 police but it was soon realised this figure was unrealistic. Subsequently, Bernard Kouchner requested some 6,000 police, but against the backdrop of competing demands for the operation in East Timor, a force of only 4,700 was eventually authorised by the UN. Even this figure has still not been reached, with numbers only at 4,155 by the beginning of October 2000 and certain areas remaining severely under-policed. For instance, in Stimlje a population of 18,000 is served by a sub-station consisting of only 3 UNMIK police officers and 25 cadets from the locally recruited Kosovo Police Service (KPS).

Some international observers have also complained that many of the police sent to Kosovo don't always represent the cream of their national forces - in one case a police dentist was sent as a patrol officer. Some police recruits were actually ordered home but their countries, more interested in preserving the income flow accruing from their deployment, refused their return. Moreover, whilst the largest contributions have come from the US and Germany, it is also the case that a number have been provided by states whose police forces have a less than savoury human rights record themselves, e.g. Kenya, Nigeria, Pakistan, Turkey, Zambia and Zimbabwe. Inevitably of course, this international force faces the challenge of policing in an alien society with a different language and in a cultural environment traditionally closed to outsiders, a problem exacerbated by the fact that secondments are likely to be for a limited time. Thus, as soon as officers are beginning to understand the environment they have been working in, they are likely to be on their way home.

The third element in the approach to policing in Kosovo has been the development of a local Kosovo police force. Again, however, the creation of this police force has been marked by a number of problems. From the start the process was dogged by in-fighting between the OSCE and the UN over who was to control it. Initially, responsibility for local police training had been vested in the OSCE which drew up a planning guide for the introduction of local police forces that was developed as early as April 1999. Under this initial plan, the OSCE would conduct the recruitment, selection and training of the Kosovo Police Service. However, a new division of tasks was agreed between the UN Secretary General and the OSCE chairman-in-office in June. As a result, the UN is now responsible for recruitment and selection whilst training is undertaken under the auspices of the OSCE at the Kosovo Police Service School in Vucitrn, which was opened in September 1999. Moreover, under the original OSCE training plan, local recruits would not receive their official appointments until they had completed three years of training and probationary work, a period which has now been reduced.

A further issue is the role of the KLA in the new Kosovo Police Service. The policy of the KLA at the end of the conflict with Serbia was to transform itself into three new component parts: one would be the new Kosovo Protection Corps, the second would be a new political party the PPDK (Party of Democratic Progress of Kosovo). The third would join the new Kosovo Police Service. This latter aspiration, like that regarding the KPC was reflected in the demilitarisation agreement negotiated by the KLA, under which it was noted that

the international community should take due and full account of the contribution of the UCK during the Kosovo crisis and accordingly give due consideration to [recognition that the UCK] is committed to propose individual current members to participate in the administration and police forces of Kosovo, enjoying special consideration in view of the expertise they have developed.

In practice this has meant that, on the one hand, the KLA have not been allowed an automatic block entry into the KPS and former members of the KLA have thus had to apply as individuals. On the other hand, however, a matrix was created for the selection process that gave a tremendous weight to prior military experience. Indeed, at least one senior member of the UNMIK police complained in September 1999 that he was being pressured by US diplomats to ensure that roughly 50% of the KPS consisted of former KLA. As of September 2000 former KLA accounted for roughly 43% of KPS recruits. The UN has also faced problems in screening recruits. All prison and educational records are gone, making it difficult to verify the information given by applicants and, initially at least, the psychological screening applied was only constructed to weed out the pathological. Moreover, as with the KPC, the problem of low pay is an issue. Indeed, some members have even made formal applications (all turned down) to be allowed to do additional work (as security guards for instance), in order to supplement their income. As one international official noted with respect to low pay and the KPS "it does lead to corruption and it will lead to corruption".

Thus, a military force that comprised less than 1% of the population of Kosovo currently comprises over 40% of its future police force. On the one hand, of course, it can be argued that, together with the creation of the KPC (and also with the efforts of the IOM to find civilian employment for former KLA), it represents an important strand in the UN's efforts to re-employ former KLA. This is an argument that certainly has some validity. Other post-conflict demobilisation processes have demonstrated the importance of ensuring that former combatants have secure jobs to go to and are thus not tempted to either return to violence or to take to criminal activities. Moreover, the UN's target of 15% minority recruitment is currently being met (with 7% of KPS graduates consisting of Serbs and 6% consisting of other minorities). It is also the case that the KPS has not acquired the same dubious reputation as the KPC.

However, the role of former KLA in both the Kosovo Protection Corps and the Kosovo Police Service means that the KLA not only has a political voice through the PPDK but it has a disproportionate influence in both official local security structures and unofficial ones, such as the UCPMB. Of course, it could be argued that given the presence of the UNMIK police and KFOR, in the short-term at least, this will not matter. However, many contingents in the latter seem more concerned with protecting themselves and the former is chronically under-equipped and still under-staffed. The implications this has for both security in Kosovo and for the region's long-term democratic development is not good.

A Structural Arms Control Perspective on Kosovo

The aim in this section is to ask how we can understand demilitarisation in Kosovo from a structural arms control perspective. It should be emphasised, however, that the current paper is not intended as an exhaustive answer to this question. Instead, what follows should be read as indicative of the issues that might be raised rather than a definitive list. The best way of undertaking this analysis is to consider the question in terms of the two broad categories of structural arms control initiatives outlined at the outset of this paper.

A. The first aim of structural arms control is to manipulate the balance of economic incentives and disincentives for actors to engage in arms purchases and/or military commitments that are excessive and/or inappropriate to the real security needs of a country. How can policy on Kosovo be understood in this light? There are a number of points to note here.

First, the development of conflict in Kosovo can be understood as a failure to implement structural arms control on the part of both the Serbs and the international community. A key factor driving the Albanians to conflict was the combined impact of the enforced measures and external sanctions on their economic well-being. In effect the dramatic contraction of Kosovo's economy in the early '90s hurt the living standards of both Serbs and Albanians and in particular created a large pool of unemployed and disillusioned Albanians who formed a natural recruiting base for the KLA. To the extent that this situation was exacerbated by fitful Western sanctions, it required a more carefully targeted approach to the question of both sanctions and economic help for civil society in Kosovo. Indeed, it can be argued that Milosevic's eventual overthrow was in part a function of the exemplary effect of more proactive attempts subsequently made to help ameliorate the effect of sanctions on those regions of the FRY opposed to Milosevic's rule, as well as efforts to help build capacity amongst Serb civil society. In contrast, the reward of the Kosovo Albanians for adopting a strategy of passive resistance was not only political repression, but a dramatic decline in living standards.

The effect of sanctions, however, was certainly not one way. They also contributed to the development of a criminal network which, in part at least, was directed at sanctions busting. In the short-term this may even have had a positive effect in that, along with remittances from the Albanian diaspora and the profits from involvement in other illicit trade (drugs for instance), it helped to ameliorate the worst effects of both the decade of enforced measures and sanctions. In some respects this trade may also have helped mute animosities between Albanians and Serbs by creating a form of economic interdependence between Kosovo and Belgrade, albeit one that existed in the informal sector and one which ran parallel to the formal structures of repression and economic competition. In the longer term however, sanctions busting contributed to the development of a criminal network which the Albanians would draw on to help fund arms acquisitions. Moreover, the lifting of sanctions, coupled with the re-routing of the drugs trade, may have hurt economic activity in the informal sector and also the illicit networks of co-operation established between Belgrade and Kosovo. At the same time, the short-term effect of the end of sanctions was to re-open Serb businesses to competition from more efficient foreign producers, and thus exacerbating the decline in living standards in Serbia generally.

In economic terms, the key to the growth of the KLA as a military force was the looting of Albanian military arsenals (an event itself precipitated by economic collapse in Albania), which released a mass of weaponry available at bargain basement prices for transport across a porous Albanian border. From a structural arms control perspective, there was a need then not only to recognise the long-term security implications of Albania's potential and actual economic collapse through greater support for the economy but, in particular, to address the problem of arms supplies from Albania. Specifically, if this issue had been tackled as vigorously as Milosevic had been led to believe in the negotiations over the Holbrooke agreement, and in the way that the KLA were threatened with at Rambouillet, then the balance of incentives for war and peace would have swung more substantially towards the latter option. Similarly, greater effort could have been made to address the way in which funds from criminal activity or from the diaspora were used to finance arms acquisitions.

In contrast, whilst the call to Rambouillet was underpinned by a NATO commitment to address the problem of arms smuggling, and whilst the UN imposed sanctions on arms supplies to the whole of the FRY, there was little interest in Western capitals in attempting to rigorously limit the KLA's supply of either arms or funding (and to the extent that there was, it tended to be a function of Western concerns about the role of Albanians in the criminalisation of their own economies). Thus, whilst Germany's main credit watchdog, the Federal Credit Supervisory Agency, instructed banks to block accounts used to finance KLA activities, reporters from The Times subsequently found that donations to the Homeland is Calling Fund were still being accepted by the bank holding the account. This failure to clamp down on arms trafficking to the KLA prior to, and during the conflict, can partly be explained by the ambivalence of Western politicians to the KLA. On the one hand there was concern about the illicit activities of Albanian mafia groups throughout Europe, on the other, there was also sympathy for the KLA's cause and hostility to Serbia - an ambivalence neatly illustrated by the Interior Minister of Bavaria in Germany, who banned KLA fund-raising in October of 1998, but who also urged colleagues to "allow Kosovo Albanians to return to fight for their homeland". Of course, it can be argued that however rigorous the initiatives undertaken, they would have never prevented arms supplies to the KLA. However, this is precisely the starting point of structural arms control, which accepts that this is unlikely. What is important though is the balance of economic incentives for peace, and those for war. Coupled with a failure to provide more funding to control Albanian arms stocks, this meant that the economic costs of a strategy of conflict were relatively low, whilst the economic costs of passive resistance were becoming intolerable.

What of the post-conflict position? First, whilst UNMIK only represents one source of aid from the international community, it is notable that a significant proportion of external aid to what is currently the de facto government of Kosovo is being spent on the creation of the Kosovo Protection Corps. Whilst this can be justified as an important element in an employment strategy for demobilised KLA, it is also the case that, in the context of a constrained budget, it has a negative impact on UNMIK's ability to provide material benefits to civil society, which in turn impacts on UNMIK's perceived legitimacy amongst the local population. Moreover, the KPC will ultimately only employ 3,000 out of the 25,723 demobilised KLA. Even when one takes into account those former KLA allowed preferential entry into the police service, the majority of ex-KLA will in fact be found employment in the civil sector through the programme established by the IOM on behalf of UNMIK. At the same time, the intermittent funding of the KPC is creating resentment amongst its members who are also resorting to work in the informal economy in order to supplement the rather low salaries they are paid. Thus, its usefulness as an employment mechanism is limited, the conditions under which it is funded may be creating a force which is resentful of the international community's high-handedness towards it, and its members currently have every incentive to maintain links with the informal economy. Of course, the creation of the KPC can be justified as expedient necessity given the reluctance of many KLA commanders to disarm at the end of the war, the support for the idea of an army amongst Albanians generally and the potential threat the KLA posed to KFOR. However, the KLA could only have threatened KFOR through sustained guerrilla actions and thus only really posed a threat to KFOR if the logic of a post-heroic war, in which it was unacceptable to incur even minimum casualties, was extended to post-conflict peacekeeping - which of course it has. Moreover, given that the Albanians were now dependent upon KFOR to defend them from Serbia, the KLA's real scope for biting the hand of their liberators is debatable.

Second, whilst it is certainly the case that former combatants need to be re-employed, it is also the case that the rest of civil society needs to be provided with an economic stake in both peace and the politics of pluralism. Yet, the international community seems intent on making the same mistakes in Kosovo as in other regions of the world when it comes to economic reconstruction. The emphasis, as ever, is on the structural adjustment mantra of a privatisation cum international management, elimination of subsidies and job cuts in inefficient industries. Similarly, donor governments have already made it clear they will not pay any more current expenditure after this year and will only pay for large investment projects. At the same time, UNMIK's entire budget for the year 2000 equates to the cost of half a days bombing during the war.

This is short-sighted. Whilst the Kosovo Albanians have benefited from remittances from the diaspora, and whilst numerous restaurants and other shops have sprung up to service the needs of international personnel, unemployment is still a problem. Indeed at 50%, Kosovo's official unemployment rate is the highest in Europe. Even if one accepts that a move to a market economy represents the best strategy for advancing the material benefits from peace, there is still a balance to be struck between the long-term goal of economic reconstruction and the short-term requirement to provide employment, and thus a stake in peace and reconciliation, for Kosovo's civilian population. Moreover, coupled with the poor salaries paid by UNMIK to local public employees, this approach creates incentives for continued participation in the informal economy. Of course, to the extent that the continued strength of the informal economy in Kosovo helps raise living standards, it may, in the short-term at least, play a positive role in alleviating discontent in Kosovo generally, and amongst the political elite in particular. In the long-term, however, illicit economic activity is likely to fuel the growth of criminal violence and be counter-productive to the development of an efficient economy.

At the regional level, there is also an urgent need for the international community to deliver on the pledges made under the Balkans Stability Pact. To date, there has been little evidence of achievement on the part of the Pact which has moved at a sclerotic pace. Moreover, whilst the Pact has been advertised as a Marshall Plan for the Balkans, the funding committed does not stand comparison with the latter. Indeed, the funding pledged under the Pact amounts to just 0.2% of EU GDP, compared with the 1.3% of US GDP devoted to the Marshall Plan. Moreover, much of the financial assistance to transition economies since 1989 has consisted of loans - many at market rates of interest. It is debatable whether increasing the foreign debt of the region is the best way to promote contended societies more likely to live in peace with each other.

Third, the international community's approach to the KLA has partly been a function of its policy over the question of Kosovo's independence, partly a function of the perceived public support for the KLA and its contemporary manifestations, but also a function of the latent threat it can pose to KFOR and an assessment of the realities of the contemporary arms market. This latter meant, quite rightly, that the disarmament provisions in the Undertaking of Demilitarisation were viewed as purely symbolic and not designed to mop up the whole of the KLA's arms. The thrust of the demilitarisation agreement was aimed at subjecting the KLA to a domestic arms control regime through the dissolution of the KLA, the creation of the KPC, entry into force of a police service subject to control by UNMIK and the promotion of security for the general population. The creation of the KPC aside, the main problem here was not in the conception but in the implementation.

First, within Kosovo itself this domestic arms control has not been applied rigorously. Whilst weapons searches have been undertaken in Kosovo, KFOR has been criticised by the ICG for not adopting a more assertive approach. For instance, whilst the existence of arms dumps in the town of Skenderaj was an open secret in Kosovo, it was only in June 2000 that KFOR launched a major search for weapons in the area. Although denied by UNMIK and KFOR, the timing seemed to be driven by political considerations - in particular as a response to Serb anger at a rise in violence against them which led to the Serb National Council walking out of the Interim Administrative Council (IAC). Of course, using arms searches as a tool of political pressure which can be turned on or off as necessary has a certain logic. The problem is that, in this context, political pressure is not just one way. Thus, partly in response to the weapons discovery, KFOR access to KPC camps was delayed on a number of occasions. Notably, although KFOR detained the owner of the land and although it 'strained credulity' that the Commander of the KPC was unaware of an arms cache in an area where he had his wartime headquarters, by the end of August no action had been taken against anybody for what represented a violation of the Undertaking. In contrast, if structural arms control is to work then all aspects of the domestic arms control regime need to be implemented on a consistent basis and the costs of violating the Undertaking need to be significant. By turning the issue of arms caches into a political tool that can be turned on and off, the issue becomes one that can be subject to pressure by domestic parties with an interest in ensuring the free availability of arms inside Kosovo.

Moreover, although the failure to rigorously restrict arms supplies to the KLA from outside Kosovo during the conflict might be justified on the grounds that it was fighting a war of liberation against an oppressive regime, the post-war situation is arguably little better. This is not to deny that a number of organisations have taken aim at the issue of potential arms supplies into Kosovo either directly or indirectly. For instance, in Albania the UN, EU, OSCE, and NATO all now provide frameworks for programmes designed to address the proliferation of small arms in the country. Consequently, however, there is a risk of duplication and lack of co-ordination. Moreover, many of these initiatives are either under-funded or currently appear largely symbolic. There is, instead, a need for a regional approach to the problem of small arms in the Balkans (and indeed in specific centres for black market weapons outside the Balkans), co-ordinated by a single organisation which aims to raise the costs of illicit arms supplies by generally increasing the security and accounting procedures at national weapons depots and improving the capacity of national law enforcement and customs officials. The challenge of this task should not be underestimated, particularly given the lawlessness that still prevails in much of Albania. For instance, a visit by one international official to customs officers in Albania was cut short by the fact that they were being held hostage by the local mafia. The model here should be the US Co-operative Threat Reduction Programme, which provides aid to the nuclear successor states of the former Soviet Union to comply with their arms control requirements and to employ former nuclear scientists. By 1996 the programme had provided $1.5 billion. A similar kind of programme needs to be operated throughout Europe. At the very least there needs to be one overarching programme for the Balkans. One forum for such a programme might be the Working Table on Security which meets under the umbrella of the Balkans Stability Pact and already has a mandate to address the control of arms sales.

B. The second range of initiatives that come under the rubric of structural arms control are those aimed at fostering the development of socio-political structures within post-conflict states in a way that both promotes meaningful demilitarisation throughout society, whilst also allowing for the development of appropriate police and defence functions. How can we understand events in Kosovo from this perspective?

Prior to and during the conflict with Serbia those elements of society committed to peaceful resistance and change went signally unrewarded by the West. In contrast, the most militaristic section of Kosovar society was rewarded from 1998/99 when it was rapidly transformed in Western policy from a terrorist organisation to the externally defined political voice of the Kosovars, despite the fact that the numbers in the KLA were still actually quite small. More support could have been given to Rugova's LDK prior to, and during Rambouillet. Instead, it was relegated to the role of subsidiary partner at Rambouillet, despite being the elected government of the Kosovo Albanians. This process has continued in the negotiations on the post-war Undertaking of Demilitarisation, which was conducted exclusively between NATO forces and the KLA, with James Rubin flying in to seal a deal. Even in the subsequent negotiations over the creation of the KPC, finally concluded in September, only KFOR, UNMIK and the KLA were directly involved. Indeed, more effort appears to have been made to keep the Serb government on-board with the agreement than other political groups in Kosovo or wider civil society. At one point, for instance, Belgrade had been persuaded to accept the more civilian sounding Kosovo Corps as the name for the KPC. Similarly, the Undertaking on Demilitarisation also set up a Joint Implementation Commission (JIC) solely comprised of representatives from KFOR, KLA and now UNMIK. In contrast, of course, the Joint Military Commission proposed under Rambouillet would have included representatives of the Kosovo civilian leadership.

In some respects this is understandable. During the conflict many elements of the FARK had actually fought under the banner of the KLA and at some point both the KLA and the FARK agreed that the former would speak on their behalf. Indeed, the FARK subsequently supplied their own written undertaking in which they confirmed that all their members had demobilised in KLA assembly areas and that they were also prepared to abide by the Undertaking of Demilitarisation. Similarly, Rugova, who spent the latter days of the NATO campaign in Italy, was less than proactive during this period, only returning to Kosovo permanently on 3 August. It was also the case that as the NATO bombing campaign had progressed, the divisions both within the KLA itself and between it and the LDK had become less obvious. At the same time however, these same divisions very quickly, and predictably, re-surfaced with the end of the war. Indeed, intimidation and even murder of LDK members, allegedly by members of the KLA or its successor organisations, is again a feature of Kosovo's political landscape. Similarly, Albanian attacks on minority groups very quickly became a feature of post-war Kosovo, and not just against Serbs. For instance, on 18 June armed clashes took place between KLA fighters and members of the Gorani community and the next day a previously unheard of organisation, the Gorani Defence Front, called for KFOR to halt the infiltration of KLA 'terrorists'.

Moreover, the Undertaking of Demilitarisation was not just a technical agreement outlining the modalities of disarmament and demobilisation, but also a political agreement for post-war Kosovo. Indeed, as a result of the Undertaking, former KLA now dominate local Albanian security organisations - the KPC and the Kosovo Police Service - in a context where concerns about corruption and political intimidation particularly on the part of the former are high. Given the continued political divisions in Kosovo between Albanians and Serbs and also within the Albanian community itself, this does not bode well for the future of either an autonomous or independent Kosovo.

The process of negotiation on demilitarisation and the composition of the JIC also contrasts markedly with the philosophy of security sector reform for post-conflict societies advocated by the UN and key NATO nations such as the UK. This places a strong emphasis on civil society oversight of the security sector, which in Kosovo seems to be interpreted as meaning oversight by the international community with little role for the Kosovars themselves. This approach also contrasts with the recommendations on the disarmament and demobilisation of ex-combatants produced by the UN's Department for Peacekeeping Operations which emphasises both that there be 'as early involvement as possible in the negotiating process by UN negotiators supported by technical experts' and that 'every effort should be made to involve civil society'. Indeed, it notes that

'local inputs and capacity building are not only important to the reconciliation process but to the entire post-conflict reconstruction effort of which the reintegration of ex-combatants is but one part. The enhancement of local, and national capacities through active consultation engagement and participation should be considered in the planning and implementation of DD&R [Disarmament, Demobilisation and Reintegration]'

This has not really happened in Kosovo. For instance, the negotiations on demilitarisation might have provided an opportunity to develop a version of the Kosovo Security Commission and Communal Security Commissions proposed in one of the early drafts of the Rambouillet text, and whose membership was defined to include representatives of each national community. In contrast, not only was the membership of the JIC narrowly drawn but other efforts to address security issues have been notable for their lack of inclusivity. Most notably, an MOU signed with the leader of the relatively moderate Gracanica Serbs, Bishop Artemije, and which the UN described as 'a partnership for security', elicited protests from Albanian politicians, partly at least, on the grounds that they had not been consulted over the document.

That this has occurred is not perhaps surprising, as the demilitarisation of the KLA was ultimately structured more by the international community's concern with preventing the emergence of a KLA threat to KFOR, than as part of a broader process of conflict resolution and capacity building amongst civil society. The lack of inclusivity also reflects a broader imperial approach to peacebuilding which views the Kosovars on both sides of the ethnic divide as the outsiders and as part of the problem rather than part of the solution.

In addition, the failure of at least some KFOR contingencies to be more assertive in policing the region, the chronic under-resourcing of the UNMIK police, as well as the delay in creating a functioning judiciary have contributed to a climate in which security for both Serbs and Albanians is at a premium. As the OSCE have noted

'it is clear that the deficiency in the law enforcement capability provided by the international community and the lack of sufficient assistance in the administration of justice has fostered the climate within which...human rights violations...have taken place. Impunity for the acts committed has resulted from failures to conduct serious investigations and this impunity, in turn has perpetuated the violence'.

Far from encouraging those elements of civil society committed to pluralism and peace, then, the process of demilitarisation in post-war Kosovo actually served to reinforce the positions of those elements in Kosovo society more closely associated with militarism and violence, both towards other ethnic groups and even towards more moderate Albanians.