Albania + 4 more

The new Kosovo protectorate

Executive Summary
The ICG paper "Kosovo - Let's Learn from Bosnia" of 17 May 1999 looked at how experience in Bosnia could be useful in Kosovo, and also at the extent to which the Rambouillet agreement of 23 February 1999 resembled the Dayton agreement of 21 November 1995.

Now, following the campaign of NATO air strikes against Yugoslavia, a new framework of military and civilian involvement in Kosovo has been established by means of UN Security Council Resolution 1244 of 10 June 1999. This mechanism ensures the legality of the operation and contents the Russians and Chinese. It has also left the UN in charge of civilian implementation.

The new model is reminiscent of the old League of Nations mandates in that it charges the foreign administrators with preparing Kosovo for self-government. It is unlike the mandates in that it is vague about the final status of Kosovo, but in the present volatile state of the Balkans that is sensible: the environment in which the future of Kosovo is finally decided will be very different from the present.

The operation in Kosovo offers the international community a new chance to deserve that name. This new opportunity for international organisations (UN, OSCE, EU and so on) to work in partnership rather than rivalry must not fail, if the world is to develop a means of handling similar future catastrophes.

Sarajevo, 20 June 1999


The ICG paper "Kosovo - Let's Learn from Bosnia" of 17 May 1999 looked at how experience in Bosnia could be useful in Kosovo, and also at the extent to which the Rambouillet agreement of 23 February 1999 resembled the Dayton agreement of 21 November 1995.

Now, following the campaign of NATO air strikes against Yugoslavia, a new framework of military and civilian involvement in Kosovo has been established by means of UN Security Council Resolution 1244 of 10 June 1999. It is worth comparing the new structure with that envisaged at Rambouillet.

Why the UN?

During the air campaign the situation in Kosovo changed so much that the Rambouillet text could no longer be enforced. In theory the Rambouillet method could still have been available: once the Belgrade government had capitulated to NATO's five conditions, it could have been required to sign a new agreement with the Albanians, brokered by a benevolent third party (the US). But no-one wanted to hand the great prevaricator, Slobodan Milosevic, new opportunities to seize the initiative and confuse the allies. And after their sufferings the Albanians too would have been in no mood for talking. In the absence of such an agreement, however, a resolution passed in the UN was the only way to guarantee international legal respectability for the foreign civil and military forces to be deployed in Kosovo.
The great advantage of using the UN is that neither the Belgrade government nor anyone claiming to represent the Kosovo Albanians needs to agree a text. Moreover, the Russians and Chinese1 are automatically bound to the common objectives expressed in the UN Resolution, and the legitimacy of overriding Yugoslavia's sovereignty is no longer in question.

But another consequence of using UN mechanisms is that the UN itself then becomes well placed to insist on a lead implementing role. Thus, whereas under Rambouillet the OSCE was to run most of the civilian aspects of the agreement, with the UN confined to UNHCR's task of leading refugee returns, now the UN is to lead the process, with the OSCE confined to a subordinate role.

The New Kosovo Protectorate

The final text of UNSCR 1244 incorporates as Annex 1 the statement by G8 foreign ministers at Petersberg on 6 May 1999, and as Annex 2 the Ahtisaari-Chernomyrdin text approved by the Serbian Parliament on 3 June 1999. In addition to calling for the complete withdrawal of all Serbian security forces, the SCR through its Annexes authorises an "international security presence with essential North Atlantic Treaty Organisation participation" to "secure a safe environment for all people in Kosovo."2 It also calls for "the establishment of an interim administration for Kosovo,"3 which will wield far-reaching powers. Paragraph 10 of the SCR:

"Authorises the Secretary-General, with the assistance of relevant international organisations, to establish an international civil presence in Kosovo in order to provide an interim administration for Kosovo under which the people of Kosovo can enjoy substantial autonomy within the Federal Republic of Yugoslavia, and which will provide transitional administration while establishing and overseeing the development of provisional democratic self-governing institutions"

The Yugoslav and Serbian governments are left almost without any role, except that a token security presence will be allowed for certain specific tasks.4

Although de jure still a part of Yugoslavia, Kosovo will fall under completely separate administrative, political and security arrangements, which will operate as a de facto independent state under international supervision. Meanwhile the role of the international administration will be to govern the country, in the absence of indigenous authorities, while at the same time developing indigenous structures which will in due course be capable of providing self-government. This is a job most reminiscent of the League of Nations mandates after 1919, and entitles the model now being established in Kosovo to be described as an international protectorate.

The UNSCR is agnostic about the final destination of Kosovo, but leaves this to be decided in future work (all responsibilities are allocated to the international presence 'pending a final settlement'). Possible futures are discussed below in section VIII.

Previous Balkan Protectorates: Austria-Hungary and UNTAES

The Treaty of Berlin in 1878 made Bosnia and Herzegovina a protectorate of Austria-Hungary (who formally annexed it in 1908).5 This Austro-Hungarian protectorate, although de jure under Ottoman Turkish sovereignty, became de facto a part of the Austro-Hungarian Empire's polity and economy. Bosnia's experience as an Austro-Hungarian protectorate, like the later experience of the mandated territories, shows that when properly administered a protectorate can provide security, stability, and economic growth in the protected region. But claims that Bosnia and Herzegovina has recently again become a protectorate, this time of the international community, are misplaced.6

A more recent Balkan protectorate was the UN Transitional Administration for Eastern Slavonia (UNTAES), which included complete and vigorous UN control over all civil affairs, as well as a strong, pro-active multi-national military component. UNTAES was a short-lived, two-year project with specific time-lines and aims, with the ultimate goal of a peaceful transition from home-grown (Serb) to Croatian administrative rule. UNTAES succeeded in its aim of a peaceful, just transition, while offering maximum protection to Eastern Slavonia's Serb community. Under UNTAES, a Transitional Administrator (TA) maintained complete control over both civil and military affairs. The TA controlled not only the civilian aspects of the agreement, but also aggressively used the military forces at his disposal to disperse unruly crowds, provide security in areas of potential ethnic unrest, and arrest indicted war criminals.

Civil-Military Separation
The success of UNTAES showed what can be achieved by an able, motivated administrator who commands both civil and military implementation in his patch. The well-documented disasters of UNPROFOR in Bosnia and Herzegovina show how badly things can go wrong without such an administrator and with the wrong mandate. Civilian implementation of the Dayton Peace Accords in Bosnia and Herzegovina shows what can go wrong when the civil and military arms are kept apart.7

A logical conclusion might have been to unite the civil and military arms in Kosovo under the right leader and with a sufficiently vigorous mandate. The mandate is certainly vigorous enough: the responsibilities of KFOR will include:

"(a) deterring renewed hostilities...enforcing a ceasefire... (b) demilitarising the Kosovo Liberation Army... (c) establishing a secure environment in which refugees...can return home (d) ensuring public safety and order until the international civil presence can take responsibility for this task..."8

Early signs were that the implementation of this mandate would need to be vigorous too. Sporadic hostility needed to be firmly countered, and ethnic clashes and vendettas defused. Demilitarising the KLA presented a task requiring a rare combination of firmness and diplomacy. Fortunately the KFOR commander, General Jackson, possesses both in addition to his Balkan field experience in Bosnia: his agreement with KLA leaders announced on 20 June is a very encouraging sign that military implementation (just as in Bosnia) might after all be less painful than many feared.

Nonetheless the UNSCR establishes a civil-military divide between the UN-appointed administrator and the armed force KFOR which 'member states and relevant international organisations' are authorised to establish.9 Why has the UNTAES precedent not been followed? One reason is obvious: the only NATO member state providing troops to UNTAES was Belgium. In Kosovo the leading NATO members will bear most of the weight. Relations between NATO and the UN are sufficiently delicate that neither will accept subordination to the other in so high-profile and political an operation as this. Result: separate command.

Memories of the hapless impotence of UN forces in a 'non-permissive environment' in Bosnia and Herzegovina from 1992 to 1995 may also have contributed to NATO insistence on keeping the two commands separate.

But the lesson from post-Dayton Bosnia is at least acknowledged. Paragraph 6 of UNSCR 1244:

"requests the Secretary-General to instruct his Special Representative to co-ordinate closely with the international security presence to ensure that both presences operate towards the same goals and in a mutually supportive manner."

How this works will depend heavily on the relationship established in the field between the civil and military commanders. But it will also depend on how much interference they suffer from their respective headquarters. UNTAES suffered little, perhaps because it was a quiet success and attracted few headlines. UNPROFOR had been plagued by an excess of managers wielding the 'dual key' decision process which tied the hands of military commanders on the ground. NATO lost a day going into Kosovo, apparently because of orders from Brussels, and as a result was beaten to Pristina airport by a Russian contingent which had raced through Serbia from Bosnia. The lesson seems clear: once you appoint your leaders, let them lead. Any person judged worthy of such heavy responsibilities should be trusted enough to carry them out. If they fail replace them; don't second-guess them.

Concentration of Civilian Powers: Better than Bosnia, at least

The UNSCR itself does not assign roles to other named organisations. In paragraph 10 (quoted in section III above), the 'assistance of relevant international organisations' is mentioned. The intention is clearly that the Secretary-General's Special Representative should be in overall control of all civilian implementation. So he/she will not be bound by the status of a mere 'first among equals' as has been the High Representative in Bosnia, and so should be less vulnerable to rivalries among organisations and conflicts of institutional dignity.

The Secretary-General immediately presented a structure which looked revolutionary. Speaking on 14 June, he said the Special Representative would have four deputies: (i) from the UN, in charge of the civil administration; (ii) from the UNHCR, in charge of refugees and humanitarian affairs; (iii) from the OSCE, in charge of institution building; (iv) from the EU, responsible for reconstruction. If these latter two organisations were really prepared to accept subordination to the UN on the ground, this would be a huge step forward towards the establishment of a genuine international community in favour of the uncoordinated group of well-intentioned actors working in Bosnia.

But this was a unilateral statement by the UN, based on the authority conveyed in the SCR. Very soon Knut Vollebaek, the Norwegian national Chairman-in-Office of the OSCE was holding talks with the UN to clarify the OSCE role, and apparently if possible to expand it.10 On 17 June Vollebaek announced that the OSCE's role would be 'restructuring of a well-functioning civilian community, monitoring human rights and maintaining an ombudsman arrangement for Kosovo, in addition to preparing for and carrying out elections.'11 In short the OSCE will have a role mirroring rather closely its tasks in Bosnia and Herzegovina, plus extra duties relating to preparing a new judicial system and training the police (handled by the UN in Bosnia).

But it still seems that the OSCE in Kosovo will be hierarchically subordinate to the UN Special Representative. It remains to be seen how this structure will function. On the face of it it should run better than in Bosnia, since a single authority will exist to co-ordinate civil administration. But it will be a historic attempt at inter-institutional co-operation, and a lot will depend on the individuals appointed to the key roles and their ability to work together.

All of this should not be taken to guarantee that the UN will always occupy a hierarchically superior position to other organisations. What is important is the principle that international organisations should be prepared to accept subordination to each other in a common cause. The UN will not always be the natural leader, and should be prepared to accept subordination in its turn in suitable circumstances.

ICG, for example, has argued12 that the OSCE itself should lead the process in Kosovo, suggesting that the UN bureaucracy lacks political sophistication and is also too heavy, providing an extra layer of decision-making and paperwork between the field operation and its sponsoring governments. Also, the UN's Balkan track record is too uneven. The OSCE has not yet developed such a large secretariat (and should not be allowed to), so that an OSCE operation could if properly set up have more direct contact with governments, as does the OHR in Bosnia through the structure of the Peace Implementation Council. Indeed the reduction in the OSCE's envisaged role between Rambouillet and the present is drastic. But any arrangement which really does lead to better co-operation among international organisations is welcome.

The Kosovo Protectorate: What Will it Look Like?

The Ahtisaari-Chernomyrdin text and the UN resolution clearly set forth a number of criteria for the international presence in Kosovo. The "international security presence, with substantial NATO participation" will be placed under a "unified command."13 Following the surprise deployment of Russian forces, the question of this unified command is a lively issue:14 a fragmented command structure would naturally be weaker and would give national governments plenty of opportunity to detract from the common purpose - but at least the intention of the SCR is clear. In addition, there will be "an interim administration for Kosovo...under which the people of Kosovo can enjoy substantial autonomy within the FRY, to be decided by the Security Council of the United Nations."15 This interim administration will function while "democratic self-governing institutions" are developed.16 In addition, there must be a "political process towards the establishment of an interim political framework agreement providing for a substantial self-government for Kosovo, taking full account of the Rambouillet accords."17

This reference to Rambouillet, and others scattered about the SCR text, are little more than lip-service to the earlier agreement. Firstly, the special representative enjoys far greater powers than Rambouillet's 'Chief of the Implementation Mission' did. Second, under Rambouillet, it is the OSCE which is " co-operation with the European Union, to constitute an Implementation Mission in Kosovo."18 As already noted, the OSCE will no longer take the lead in this process. Thirdly, the thrust of Rambouillet was to rescue a situation not yet out of control; now the task is to re-establish control in an environment of total chaos.

Final Status: Autonomy, Republic or Independent State?

Two important questions remain regarding the Kosovo Protectorate: what are the long-term goals for Kosovo; what is the exit strategy? The Dayton Peace Accords envisioned a one-year NATO military presence in Bosnia-Herzegovina,19 a period which proved too optimistic. The UNSCR on Kosovo calls for an "initial period of 12 months to continue thereafter unless the Security Council decides otherwise."20 In other words, the Kosovo Protectorate's mandate is indefinite. Given the experience in post-Dayton Bosnia and Herzegovina, the Kosovo Protectorate will last many years. To avoid drift concrete goals and time-lines must be spelled out. Effectiveness of the protectorate must be judged by the ability of the UN Special Representative to meet the goals within the given time framework.

Article 10 of the UNSCR specifically refers to the UN administration in Kosovo as "transitional", which under article 11 will oversee "the transfer of authority from Kosovo's provisional institutions to institutions established under a political settlement."21 If the protectorate is indeed transitional, the question must be asked: towards what?

There are three possible outcomes. The first is to revert to the situation before 1989 and give Kosovo the status of an autonomous province of Serbia. This will not be acceptable now to most Kosovo Albanians. So too, it will complicate the reconstruction of Kosovo, deterring private investment and keeping the territory dependent on aid. It will also raise internal constitutional issues in FRY and Serbia.

A second option, the creation of Kosovo as a third Yugoslav Republic alongside Serbia and Montenegro, is possible under the text of the UN resolution, which reaffirms "the sovereignty and territorial integrity of the Federal Republic of Yugoslavia," but makes no mention of the sovereignty and territorial integrity of the Republic of Serbia.22 This will be difficult to sell in Serbia, which views third republic status as a stepping stone to Kosovar independence. Montenegro -- whose Albanian minority recently called for sweeping autonomy -- would also oppose such a move, not only on the grounds that it will make Montenegro the smallest of the three republics, but also out of fears that it will create an Albanian irredentist movement within Montenegro. Macedonia will also be concerned, as such a move will strengthen irredentist claims of its Albanian population.

The third option, the one most desired by the Albanians and most feared by Serbs, Montenegrins and Macedonians, is complete independence from Yugoslavia. Although the third option appears to be out of step with the guarantees of sovereignty in the preamble of the UNSCR,23 other language in the text holds out the possibility of eventual independence, notably paragraph 11(e), which gives the UN civil administration the duty of "facilitating a political process designed to determine Kosovo a future status."24 Kosovo independence will certainly provoke strong responses from the Serbs, including the possibility of renewed war. Macedonia too will fear that the prospect of an independent Kosovo will lead to calls for a Greater Albania and unrest among Macedonia's ethnic Albanian minority. In general a new ethnic Albanian state (for how many Serbs will remain in an independent Kosovo?) will be a destabilising factor in the region.

None of these outcomes is trouble-free, and none need be chosen at this early stage - but this reinforces the expectation that the international presence will need to stay in Kosovo for some time to come. Most likely Kosovo's future status cannot be decided in isolation, but will need to be determined as part of some general settlement for south-eastern Europe.


The devastation of Kosovo has made necessary a stronger model of international intervention than was envisaged at Rambouillet. The model now adopted is a protectorate in all but name, reminiscent of the League of Nations mandates. The United Nations mechanisms, used to establish the new protectorate, are the only ones in existence which preserve the legality of the operation and also keep the Russians and Chinese on board. This gives the UN a chance to recover the reputation it lost in Bosnia: in this context it would do well to learn from UNTAES and UNPROFOR the lesson that HQ staff should not try to second-guess the commanders in the field: the proper role of supervision is to ensure that resources are adequate (neither too much nor too little) to the task.

The operation in Kosovo offers the international community a new chance to deserve that name. This new opportunity for international organisations (UN, OSCE, EU and others) to work in partnership rather than rivalry must not fail, if the world is to develop a means of handling similar future catastrophes.

With the UN in charge of civilian implementation, it is probably good that NATO, with its stronger stomach for firm action, is leading the military side (leaving aside the question of the status of the Russians). But separation of these two arms is a dangerous model to perpetuate, since it puts extra responsibilities of co-ordination on the field commanders, and the need to report to different headquarters opens a potential for rifts to develop.

Finally, the long-term future of Kosovo is best left alone for now. None of the options on the table looks enticing, and in any case the final decision will surely be taken in local and regional circumstances very different from those prevailing at present.


  1. The Chinese abstained in the vote on 10 June, but did not use their veto and so are bound by the result.
  2. UNSCR 1244, Annex 2.4.
  3. ibid, Annex 2.5.
  4. ibid, Annex 2.6. The tasks include mine clearance and 'maintaining a presence' at Serb patrimonial sites and key border crossings.
  5. Created in 1878 at the Congress of Berlin, this protectorate included not only Bosnia-Herzegovina, but also the Sandzak of Novi Pazar, a predominantly Muslim Slav region straddling the Serbian and Montenegrin border, stretching from Bosnia to Kosovo. The Dual Monarchy governed Bosnia-Herzegovina with its own civil administration, while allowing the Turks to maintain administrative control over the Sandzak. The Austro-Hungarian administration introduced new laws, economic and administrative reforms throughout Bosnia and Herzegovina. Under Austro-Hungarian rule the province prospered, began to modernise its infrastructure, and developed an industrial base.
  6. Kosovo: lets learn from Bosnia: models and methods of international administration, ICG report No. 66, 17 May, 1999.
  7. Ibid.
  8. UNSCR 1244, paragraph 9. (d) in particular is a big step forward from Dayton and Rambouillet, see ICG, op.cit, page 9.
  9. UNSCR 1244, paragraph 7.
  10. Reuters, 14-17 June.
  11. Reuters, 17 June
  12. Op. Cit. in section I above.
  13. UNSCR 1244, Annex 2.4.
  14. Resolved, at least for now, at the G8 meeting in Cologne on 19 June 1999.
  15. Ibid, Annex 2.5. The language is awkward but the sense is clear.
  16. Ibid, paragraph 10.
  17. Ibid, Annex 1.
  18. Rambouillet Interim Agreement for Peace and Self-Government in Kosovo. Chapter 5, Article I.1.
  19. Dayton Peace Accords, Annex 1A, Article I:1.
  20. UNSCR 1244, paragraph 19.
  21. Ibid, paragraph 11.f.
  22. UNSCR 1244, Preamble.
  23. Ibid.
  24. Ibid, paragraph 11.e.