Autonomy is the legally entrenched power of communities to exercise public policy functions of a legislative, executive and/or judicial type independently of other sources of authority in the state, but subject to the overall legal order of the state. Autonomy as a strategy of preventing and settling self-determination conflicts is based on the recognition of group-specific concerns alongside and on par with concerns of individuals (independent of their ethnic identity) and the state. It is equally based on accepting that to endow an ethnic group with legislative, executive, and judicial powers to address these concerns effectively will contribute to individual, group, and state security, and thus to preventing the disruption of the territorial and/or social integrity of a given country.
In practice, autonomy arrangements incorporate executive, legislative, and judicial powers to varying degrees. In cases where it is used as an instrument for self-determination conflict prevention and settlement, autonomy ideally includes such a mix of the three that enables the ethnic group in question to regulate independently the affairs central to the concerns of its members, which are normally easily identifiable as they manifest themselves in concrete claims. However, as autonomy falls short of full sovereignty, this happens within the broader constitutional and legislative framework of the minority’s host country and under the supervision of a central government or similar agencies ensuring the compliance of all actions of the autonomous institutions with the regulations set up for the operation of autonomy.
 This definition draws in part on Wolff and Weller (2005).
Autonomy is neither a new phenomenon, nor has it been understudied. Even though the application of autonomy arrangements in the sense that they are defined here only started in the 20th century, there are some earlier antecedents, including Hungarian self-rule in the Habsburg Empire after the 1868 Ausgleich, the gradually eroded autonomy that Poland had within the Russian empire following the Congress of Vienna (so-called Congress Poland), as well as the Grand Duchy of Finland between 1809 and 1917, also within the Russian Empire.
However, up to the period of time when the post-Cold War transitions in Central, Eastern and South Eastern Europe were beginning, it appeared to be at best a highly unusual tool of state construction, or at worst a highly dangerous one. It was seen to be unusual, inasmuch as autonomy generally seemed to be attached to fairly obscure, historical examples, born out of very distinctive historical settings. Often autonomy regimes operated in remote or otherwise geographically unique locations, such as islands (for example, the Åland autonomy) or enclaves (for example, Klaipeda). These types of cases, it was widely believed, could not offer a great deal by way of guidance in less unique circumstances. Even the few new autonomies that were established after World War II, such as South Tyrol, were until recently taken to be too dependent on specific local conditions as to be of wider interest. Similarly, the Soviet and other socialist autonomies were seen to be rooted too deeply in Leninist/Stalinist ideology, rather than in a genuine practice of accommodating self-determination claims, and thus not considered to be of wider applicability. The various forms of territorial self-governance—from autonomous districts to Union Republics—were notional and devoid of any real, substantive powers of self-governance.
Autonomy was also not given a great deal of consideration because the concept was, rightly or wrongly, associated with self-determination struggles. Outside of the colonial context, any self-determination discourse was viewed with great suspicion by governments, seeing it as a first step onto that slippery slope that inevitably leads towards irredentist or secessionist claims. Thus, autonomy was widely regarded as a somewhat dangerous concept that a state would only employ at its own peril.
Since the end of the Cold War, this climate has changed somewhat. In the transitional states of Central and Eastern Europe, the almost simultaneous breakdown of mechanisms of external (through the Warsaw Pact Organisation) and internal control (through dictatorial forms of government) led to the re-emergence of the so-called national minority question. In Georgia, Moldova, the new Russian Federation, and in relations between Armenia and Azerbaijan, the doctrine of territorial integrity was undermined by intense armed conflict. These conflicts, framed in the rhetoric of self-determination, and the prospect (and subsequently the reality) of the dissolution of Yugoslavia added to the perceived threat to the principle of territorial integrity.
In response, autonomy was re-discovered as a potential remedy to self-determination claims (Kymlicka 2007). While post-communist governments across Central and Eastern Europe remained reluctant at best, and openly hostile to such ideas at worst, western mediators embraced autonomy more enthusiastically. They rejected the claim that autonomy was nothing but the secessionists’ stepping stone towards independence but, instead, considered autonomy as a possible tool in accommodating secessionist movements without endangering the continued territorial integrity of an existing state. In 1990, the member states of the Conference on Security and Cooperation in Europe (CSCE), as it then was, were still cautious when noting ‘the efforts undertaken to protect and create conditions for the promotion of the ethnic, cultural, linguistic and religious identity of certain national minorities by establishing, as one of the possible means to achieve these aims, appropriate local or autonomous administrations corresponding to the specific historical and territorial circumstances of such minorities and in accordance with the policies of the State concerned’ (Article 35 (2), 1990 Copenhagen Document of the Conference on the Human Dimension of the CSCE). But by 1991, the governments of the member states of the European Community (EC) went further in endorsing autonomy as a means of addressing minority issues and self-determination conflicts when establishing conditions for recognition of the new states of Central and Eastern Europe emerging from the dissolution of the Soviet Union, the separation of Czechoslovakia and the disintegration of Yugoslavia. In two declarations on European Political Cooperation, one addressing all of Central and Eastern Europe, the other pertaining to Yugoslavia, minority rights, and to a certain extent, autonomy for national minority groups was prescribed as part of the price that the states of that region might have to pay for diplomatic relations with the member states and institutions of the EC. This demand built upon the work of the EC Peace Conference on Yugoslavia. Through that conference, the EC member states attempted to achieve an agreed dissolution of the Socialist Federal Republic of Yugoslavia. Serbia was the only republic vigorously opposed to this approach. In an effort to address Serb concerns, two successive peace plans provided by Lord Carrington, the Chair of the Conference, offered autonomy arrangements for Serb communities living outside the boundaries of the Serbian Republic within the crumbling Socialist Federal Republic of Yugoslavia.
In the meantime, autonomy as a tool of state construction was of course being applied, discussed and analysed outside the context of the former Yugoslavia as well. Some states in Western Europe have embraced autonomy (or devolution) as a means of maintaining their territorial integrity. In addition to the more established case of Belgium, Spain and the United Kingdom have also made startling progress in this direction. Even centralist France has attempted to move towards autonomy as a means of addressing the Corsica conflict. A number of innovative settlements have been adopted in relation to other areas of conflict or ethnic tension, some of which are principally autonomy-based, such as Gagauzia in the Republic of Moldova or Crimea in Ukraine. Autonomy structures also play a part in several new models of more complex forms of power-sharing that can be found in Northern Ireland and in the framework agreement for Sudan, as well as in Bougainville and Mindanao.
Keating (2001), among others, makes the point that the European Union has created a geo-political and economic environment that is generally conducive to the establishment of territorial self-governing entities at the sub-state level. Yet the EU’s relative reluctance to commit itself and its (existing) member states more explicitly to recognizing and protecting national minorities and stateless nations combined with the resistance of many of its member states to granting specific autonomy rights to their minorities has meant that the framework that EU generally provides has not been used to its fullest possible extent for establishing autonomy regimes. A widely recognizable trend towards multi-level governance across most EU member states either does not result in concrete autonomy gains for minority populations (e.g., France, Greece, Romania, Slovakia, Bulgaria) or where such gains are discernable they cannot be immediately and directly linked to an EU impact (Denmark, Finland, and Spain, for example, have had most of their autonomy arrangements in place before accession to the EU). Nonetheless, the point remains that the EU provides a permissive environment for sub-state territorial self-governance, even if the opportunities arising in this environment do not always translate into concrete results.
More widespread implementation of autonomy regimes as mechanisms to address self-determination conflicts have been complemented by an increased scholarly interest and output in this respect, with several significant scholarly works on autonomy published over the past decade. In terms of standard-setting, the Organisation on Security and Cooperation in Europe (OSCE, the successor organisation of the CSCE) has also maintained an interest in the issue. Its Lund Recommendations provide perhaps the most comprehensive reference to autonomy as a means of good governance and state construction in an authoritative international document thus far. Moreover, the United Nations General Assembly has addressed itself to this—previously altogether too delicate—topic in the shape of the Liechtenstein Initiative on Self-determination through Self-administration, which seeks to offer autonomy as an alternative to secessionist self-determination claims. The United Nations Working Group on Minorities has also been cautiously addressing the issue of autonomy.
Thus, developments over the past decade and a half seem to indicate that there is renewed interest among scholars and practitioners to engage with the thorny issue of autonomy alongside an apparently increasing willingness among major actors in the international community to recommend, and where necessary impose, autonomy regimes on states that might otherwise collapse under the pressure of self-determination conflicts. Increasingly, autonomy is also made available in situations where a self-determination conflict has not yet fully developed. While self-determination conflicts are characterized by a claim to a unilateral change in status, in other instances, ethnic groups may merely be seeking a greater expression of their identity within the state. This can take the form of a claim to enhanced regional or local self-governance.
Taken together, these two trends make it now possible to consider autonomy as a means of state construction that does not always, and of necessity, has to raise the spectre of self-determination struggles and ultimate secession. Instead, autonomy is just seen as one element of state construction addressing the needs of diverse communities.
 This was led by the path-breaking survey by Hannum (1990) and accompanying documents (1993).
Tim Potier, among others, (2001: 54) has noted that ‘…international lawyers have failed to come to any agreement on a “stable” workable definition for autonomy. … it escapes definition because it is impossible to concretise its scope. It is a loose and disparate concept that contains many threads, but no single strand.’ In political science, too, the difficulty to pin down and conceptualise autonomy has been recognised: two of the most eminent scholars in the field, Brendan O’Leary and John McGarry, observed in 1993: ‘Overlapping cantonisation and federalisation there exists a grey area of territorial management of ethnic differences which is often found in conjunction with external arbitration. International agreements between states can entrench the territorial autonomy of certain ethnic communities, even though the “host state” does not generally organise itself along either cantonist or federalist principles’ (McGarry and O’Leary 1993: 32).
Despite this appreciation of the difficulty to define clearly what autonomy is, political scientists and international lawyers have not hesitated to propose a variety of definitions. Michael Hechter (2000: 114) describes ‘political autonomy’ as ‘a state of affairs falling short of sovereignty’. In Ted Robert Gurr’s (1993: 292) understanding ‘autonomy means that a minority has a collective power base, usually a regional one, in a plural society’. Hurst Hannum and Richard Lillich (1980: 859) stated in their influential essay on ‘The Concept of Autonomy in International Law’ that ‘autonomy is understood to refer to independence of action on the internal or domestic level, as foreign affairs and defence normally are in the hands of the central or national government, but occasionally power to conclude international agreements concerning cultural or economic matters also may reside with the autonomous entity.’ In similar terms, Tim Potier (2001: 54) makes the point that autonomy ‘should be understood as the means whereby an authority, subject to another superior authority, has the opportunity to determine, separately from that authority, specific functions entrusted upon it, by that authority, for the general welfare of those to whom it is responsible.’ In her extensive study on autonomy, Ruth Lapidoth draws a clear distinction between ‘territorial political autonomy’ and ‘personal autonomy’. To her, ‘[t]erritorial autonomy is an arrangement aimed at granting a certain degree of self-identification to a group that differs from the majority of the population in the state, and yet constitutes the majority in a specific region. Autonomy involves a division of powers between the central authorities and the autonomous entity.’ (Lapidoth 1997: 174-175) In contrast to this territorial conception, ‘[p]ersonal autonomy applies to all members of a certain group within the state, irrespective of their place of residence. It is the right to preserve and promote the religious, linguistic, and cultural character of the group through institutions established by itself.’ (Lapidoth 1997: 175)
More important than these definitional concerns about the nature of autonomy is that in recent legal and political science discourses a greater appreciation has emerged of the fact that autonomy in and by itself may not be sufficient to provide a framework in which self-determination conflicts can be resolved. Thus, several scholars have begun to argue that complex institutional designs, including in some cases autonomy provisions, are more likely to provide stable, long-term solutions to self-determination conflicts. This has been increasingly understood by practitioners of conflict resolution and has led to an emerging practice of conflict settlement that can be referred to as “complex power sharing”.
Complex power sharing, as it relates to autonomy, refers to a practice of conflict settlement that has a form of self-governance regime at its heart, but whose overall institutional design includes a range of further mechanisms for the accommodation of ethnic diversity in divided societies, including those recommended by advocates of consociationalism, such as McGarry and O’Leary, and power-dividing, such as Roeder and Rothchild, as well as integrationism, such as Horowitz (1985, 1990, 1991, 2002, 2003, 2004, 2006), Reilly (2001), Sisk (1996), and Wimmer (2003). Complex power sharing is thus the result of the implementation of a self-governance regime whose success as a conflict settlement device requires a relatively complex institutional structure that cannot be reduced to autonomy/(ethno-)federalism, (traditional) power sharing or power-dividing.
In a sense, then, many contemporary theoretical interpretations of autonomy lag behind current practical conflict resolution efforts that have moved beyond many of the ‘purist’ approaches to conflict resolution and prevention recommended by different schools of thought. This is not to say that autonomy has lost its relevance as a conflict settlement mechanism. On the contrary, autonomy is more than ever a required component in most such settlements, providing a framework within which other conflict resolution mechanisms acquire meaning and can reach their potential to contribute to achieving stable and durable democratic solutions to self-determination conflicts short of secession.
 This distinction is made by a number of scholars, including Heintze (1997: 37-46), Hechter (2000: 72ff.) and Potier (2001: 55f. and 59f.)
 Cf., for example, Danspeckgruber (2005), McGarry and O’Leary (2004) and contributions therein, O’Leary (2005), Roeder and Rothchild (2005) and contributions therein, Weller and Wolff (2006), and Wolff (2004).
 The term ‘complex power-sharing’ was first used in a research project funded by the Carnegie Corporation of New York (“Resolving Self-determination Disputes Through Complex Power Sharing Arrangements”). In this project, complex power-sharing regimes are distinguished “in that they no longer depend solely on consociational theory, or solely upon integrative theory”, involve international actors that “are often key in designing, or bringing experience to bear upon, the structure of the eventual agreement, or its implementation” and “consider a far broader range of issues … and … address structural issues as diverse as economic management, civil-military relations and human and minority rights, and … do so at many different levels of government”, thus recognising “that at different levels of government, different strategies may be more, or less, applicable, and consequently more, or less, successful, in engendering peace and stability” (Kettley, Sullivan, Fyfe 2001: 4-5). O’Leary (2005: 34-5) uses the term ‘complex consociation’ in a similar manner.
 This is the definition of complex power sharing as used by Wolff (2007a and b).
There is a large number of such settlements that provide evidence for the continuing importance of autonomy as a conflict settlement mechanism. This can be demonstrated with reference to self-determination conflicts in North America (Canada), Central and South America (Panama, Colombia, Mexico, Ecuador and Nicaragua), Africa (Sudan, Zanzibar), Asia (Iraq, Indonesia, Papua New Guinea and Philippines), and Europe (Belgium, Bosnia and Herzegovina, Macedonia, Moldova, Russia, Serbia and Montenegro, Ukraine and United Kingdom). In addition, proposals for autonomy regimes also figure prominently in proposed peace agreements, including in the Annan Plan for Cyprus, the Georgian president’s peace initiative for South Ossetia, and Sri Lanka. Thus in virtually every conflict situation involving self-determination claims by territorially relatively concentrated identity groups at least proposals for autonomy have been made. In many of them, these proposals have been implemented.
A second more specific trend is the combination of autonomy in particular with power sharing at the level of the central government and/or within the autonomous entity. This, however, only tells half the story. It is the specific nature of both of these macro-political conflict resolution techniques—autonomy and power sharing—and the range of ‘supplementary’ mechanisms—be they specific electoral systems, human and minority rights legislation, or coordination and arbitration mechanisms—that need to fit the specificities of the particular case to which they are applied, but also, and importantly, have to fit each other. This means that there are limits to the extent to which the design of complex power sharing settlements can choose at random from the available menu of mechanisms and techniques. This can be illustrated with specific examples and on the basis of broader comparative observations.
Autonomy at Work: The Northern Ireland Settlement
In order to appreciate in full why the Agreement in Northern Ireland reached in 1998 and modified in 2006 provides a reasonable framework for a successful peace and political process, it is necessary to get a more detailed understanding of the conflict as such. The conflict on Northern Ireland is about national belonging. Two communities – Unionist/Loyalist and Nationalist/Republican – have conflicting visions of Northern Ireland’s constitutional status as part of the United Kingdom or the Republic of Ireland, respectively. Overlapping with this fundamental political divide are further divisions that cut through Northern Irish society: religion (Protestant vs. Catholic), language (Ulster Scots vs. Irish), culture (“British” vs. Gaelic).
Any settlement of the conflict, therefore, required addressing the following sets of issues: relations between the communities in Northern Ireland; the status of Northern Ireland within the United Kingdom; relations between Northern Ireland and the Republic of Ireland; and relations between the UK, its constituent entities, and the Republic of Ireland. The Agreements of 1998 and 2006 provide a comprehensive institutional answer to these issues, an answer moreover that was negotiated between the main conflict parties, rather than imposed by well-meaning outsiders.
The provisions of the 1998 Agreement extend to executive and legislative power sharing, and additionally brought to Northern Ireland a wide range of human rights legislation and provisions for victims. The design of the executive power sharing government relies upon the d’Hondt procedure for appointing ministers, thus reflecting party strength in the Assembly as a mechanism to ensure not only proportional representation in the Assembly, but equally in the executive. Under the 1998 Agreement, the First Minister and his (or her) co-equal Deputy were to run on a joint ticket and, to be elected, needed concurrent majorities within both traditions (i.e., 50 percent plus one of members present and voting in both the Nationalist and Unionist communities) and emphasizing individual ministerial and collective Executive accountability. Under the 2006 St. Andrews Agreement, the application of the d’Hondt procedure was extended to cover the offices of First and Deputy First Minister as well. Further changes that affect the Executive include a greater emphasis on the collective nature of government. While the 1998 Agreement had vested strength and independence into individual ministerial portfolios, the St. Andrews Agreement introduces a statutory ministerial Code that requires ministers to ensure that all sections of both traditions in Northern Ireland can participate in the functioning of the devolved power sharing institutions and that there interests are protected. It also opens up the opportunity for Executive decisions to be taken by concurrent majority voting: when no consensus can be achieved in the Executive, a minimum of three ministers can require the employment of this procedure and a decision by the Executive could only be adopted if an overall majority and a majority among ministers from both communities voted in its favor.
The terms of the St Andrews Agreement also establish more clearly than the 1998 Agreement that the Executive is a collective organ of government. In particular, they determine that the Executive be the forum for discussion and decision-making on any issues in the responsibility of more than two ministers, on the prioritization of executive and legislative proposals, and adopting an annual Programme for Government. Crucially, as another illustration of the commitment to power sharing, the ministerial code containing all these provisions can only be adopted and changed on the basis concurrent majorities in the Assembly.
The legislative branch of government in Northern Ireland, the Assembly, is elected by Single Transferable Vote on the basis of six seats being contested in all eighteen general election constituencies in the region. Provisions for legislative power sharing include qualified and concurrent majority voting procedures, committee oversight of ministerial portfolios, and self-designation of elected members of the Assembly.
The most significant amendments to the functioning of the Assembly include a new power of the Assembly to refer individual ministerial decisions back to the Executive for collective consideration. Furthermore, the already existing Committee of the Centre is now on a statutory footing, and thus equal to other departmental scrutiny committees. At the same time, a special standing committee and an efficiency review panel were established to review the working of the institutions and, where necessary, recommend changes to their structures and procedures. Finally, members of the Assembly can no longer change their designation (“Nationalist”/“Unionist”) during an assembly term, except when changing party membership.
Northern Ireland as a whole is a self-governing territory with extensive devolved legislative and executive powers, including in the areas of agriculture and rural development; culture and arts; education, employment and learning; enterprise, trade, and investment; environment; health, social services, and public safety; regional development; and social development. Special arrangements prevail in relation to policing and justice, as well as national security. While the UK government will retain all competences in relation to external defence and significant powers in relation to combating terrorism, there is, under the 2006 St. Andrews Agreement, a provision that makes possible the devolution of powers in the area of policing and justice.
Northern Ireland’s status within the UK is regulated by Act of Parliament (there is no written constitution), and any changes to its status as part of the UK (or the Republic of Ireland, for that matter) is subject to a referendum (“consent principle”).
A North-South Ministerial Council (NSMC) allows for policy coordination between Northern Ireland and the Republic of Ireland. This involves primarily subject ministries, but in Northern Ireland under the terms of the St. Andrews Agreement also gives the Executive collectively a greater role in preparing such meetings, while guaranteeing a right of attendance to the minister into whose competence a particular issue to be discussed at NSMC meetings falls. The two parliaments (and/or their relevant committees) retain rights to scrutinize the work and decisions of the NSMC and its implementation bodies. A Review Group, to be appointed jointly by the Northern Ireland Executive and the Irish Government is to examine the efficiency and value for money of existing NSMC implementation bodies and establish whether there is a case for additional bodies and areas of cooperation within the NSMC. In addition, under the terms of the St. Andrews Agreement, plans exist for the creation of a North-South Parliamentary Forum (bringing together equal numbers from the Irish Parliament and the Assembly, and operating on an inclusive basis) and an Independent Consultative Forum (to represent civil society).
Also established under the 1998 Agreement, is the British-Irish Council (BIC) that brings together the British and Irish governments and relevant executive bodies from all self-governing territories in the UK (Northern Ireland, Scotland, Wales, Isle of Man, Channel Islands) to coordinate policy and cooperate on issues of mutual interest. The St. Andrews Agreement mandates the circulation of all papers in preparation for BIC meetings among all members of the Executive and an entitlement for relevant subject ministers to attend. Plans also exist under the St. Andrews Agreement for the creation of a permanent Secretariat of British-Irish Council and an East-West Inter-parliamentary Framework.
Beyond Northern Ireland: A Comparative Perspective on Applications of Autonomy
Recent conflict settlements, such as in Sudan, Iraq, Papua New Guinea/Bougainville, Philippines/Mindanao, Belgium, Bosnia and Herzegovina, Moldova/Gagauzia, and Ukraine/Crimea establish autonomy regimes that combine forms of horizontal and vertical power sharing and power dividing in an effort to establish stable political and institutional processes conducive to resolving self-determination conflicts. Vertical power sharing and power dividing prove necessary complements of autonomy in two ways: autonomy regimes cannot be established in specific territorial entities without it, and unless such entities become a locus of power, no power can be shared at the sub-state level. Power sharing and power dividing in the Bosnian-Croat Federation, in Bougainville, in the Autonomous Region of Muslim Mindanao, and in South Sudan thus would not be possible if these regions had not been established as legal-political entities and powers had not subsequently been devolved to them.
Mechanisms of power dividing exist in all these cases as well. Apart from the vertical division of power, i.e., the distribution of powers between different vertical layers of authority, one also finds a range of further horizontal mechanisms: most obviously there is an emphasis on independent judicial institutions tasked with the upholding of the constitutional order and the enforcement of human and minority rights legislation. Division of power between executive and legislative branches of government exists as well, but is not as universal. Indeed, parliamentary systems are marginally more common both at central and regional levels of government. Where these systems are integral part of conflict resolution efforts, they are strongly correlated with the establishment of executive power sharing: they are prescribed in Belgium, Brussels, the Federation of Bosnia and Herzegovina, and South Tyrol, but emerge voluntarily in Crimea. By the same token, presidential systems, favored by power dividers, do not preclude executive power sharing. Bosnia and Herzegovina (albeit with a semi-presidential system), Sudan and South Sudan serve as illustrations.
From this degree of variation across the case studies one can draw a number of both analytical and empirical conclusions about the continued importance of autonomy as a conflict resolution mechanism. Empirically, there are four important lessons. First, autonomy will only serve its purpose of contributing to conflict resolution if it goes hand in hand with the division of power along a vertical structure of institutions. It can then serve as a useful substitute for formal horizontal power sharing at either national or regional levels, provided that national or regional ethnic demographies create suitably homogeneous territories and that substantial powers are devolved from the centre. Second, no attempt was made in any of the cases mentioned to create heterogeneous entities as subjects of territorial self-governance. Heterogeneity, where it exists, was addressed by means of consociational powersharing within the self-governing territorial entity. Third, coordination between different vertical layers of authority and the establishment of clear hierarchies are important to ensure that vertical layering of authority remains meaningful and can contribute to the long-term sustainability of a particular conflict settlement. Where there is a danger of eroding the degree of autonomy enjoyed by specific territorial entities and their populations created as a particular layer of authority with the specific purpose of conflict resolution (such as Mindanao, South Sudan, and with some qualifications, Crimea), conflict settlements may not be sustainable in the long term.
This means, fourth and finally, that without safeguards against arbitrary government interference, it is unlikely that the conflict parties will develop a sense of satisfactory permanence and predictability in relation to a particular autonomy regime established as a conflict settlement. Legal and constitutional entrenchment, possibly alongside international guarantees, is thus one important mechanism for the stabilization of institutional structures. These and other strategies that provide checks and balances on the exercise of power serve to ensure that principles of liberal democratic state construction shape conflict settlements and enhance their longer-term legitimacy.
However, from the perspective of the minority community another mechanism can be equally important, namely the option to secede in case of major constitutional, demographic or political changes. Thus, Bougainville has a future option for a referendum on its independence from Papua New Guinea. South Sudan is set to have a referendum on independence in January 2011, after an interim period of six years.
These two observations on entrenchment and popular consultation also underscore again that the preservation of democratic procedures is a key factor for stabilizing autonomous institutional structures created for the purpose of resolving self-determination conflicts, because it is through this longevity that institutions acquire their legitimacy. While democratic institutions in themselves are not necessarily and automatically technically viable, compliance with rules and regulations agreed between all conflict parties and their democratic accountability to voters increases the survival chances of smooth and efficient institutional processes. Any form of autonomy combined with power sharing regime will always present a modification of, and constraints to majoritarian forms of democracy, but this does not mean that its institutions can or should be run without popular support. Any autonomy regime does depend upon the willingness and ability of elites to cooperate and make compromises, but also depends on the willingness of the people to support their respective elites in this process and to uphold a settlement negotiated to bring about a non-violent, stable and predictable political process.
 The 2003 constitution of the Union of Serbia & Montenegro provided for a bi-national federation between the two entities and included an option for Montenegrin independence after three years if at least 55% of people participating in a referendum would opt for it. The referendum was held on 21 May 2006, and Montenegro declared its independence on 3 June after the country’s referendum commission confirmed as official the preliminary result which had already been recognized by all five permanent members of the UN Security Council on 23 May.
 This is not meant to be a comprehensive list of cases. For an analysis of some examples and general trends in the spread of autonomy regimes as part of conflict settlements, see contributions in Weller and Wolff (2005).
 The Agreement Reached in Multi-Party Negotiations; The Agreement at St. Andrews, http://www.nio.gov.uk/Publications/Article/St-Andrews-Agreement
The cases referred to in this article are all similar to the extent that they comprise self-determination claims by territorially concentrated identity groups that lent themselves to the establishment of complex power sharing regimes with autonomy regimes at their heart. Some of them have proven relatively stable over time (i.e., over ten years): Belgium, Brussels, Bosnia and Herzegovina, Crimea, Gagauzia, and South Tyrol. Northern Ireland has achieved not only a very significant reduction of violence but also a relatively stable political process at the beginning of the second decade after the conclusion of the 1998 Agreement. Others, including Bougainville and South Sudan are too short-lived to provide reliable data about their long-term stability and moreover might, in accordance with their respective settlements see a consensual secession of the autonomous entity from the state. Mindanao has only seen partial success in bringing peace to a troubled region of the Philippines. Thus, autonomy as part of more complex power sharing regimes or in rarer cases on its own, remains an essentially democratic, stabilizing, violence-reducing and thus highly desirable approach to the resolution of self-determination conflicts.
Danspeckgruber, W. (2005) ‘Self-governance plus regional integration: a possible solution to self-determination claims’, in Autonomy, Self-governance and Conflict Resolution: Innovative Approaches to Institutional Design in Divided Societies, ed. by Marc Weller and Stefan Wolff. London: Routledge, 26-48.
Diehl, P. F. and Lepgold, J. eds. (2003) Regional Conflict Management. Lanham, MD: Rowman and Littlefield.
Gurr, Ted Robert (1993) Minorities at Risk. Washington, D.C.: United States Institute of Peace Press.
Hannum, H. (1990) Autonomy, Sovereignty and Self-determination: The Accommodation of Conflicting Rights. Philadelphia: University of Pennsylvania Press.
Hannum, H. ed. (1993) Basic Documents on Autonomy and Minority Rights. New York: Springer.
Hannum, Hurst and Lillich, Richard B. (1980) ‘The Concept of Autonomy in International Law’, in American Journal of International Law 74, 858-89.
Hechter, Michael (2000) Containing Nationalism. Oxford: Oxford University Press.
Heintze, Hans-Joachim (1997) ‘Wege zur Verwirklichung des Selbstbestimmungsrechts der Völker innerhalb bestehender Staaten’, in Heintze, Hans-Joachim, ed., Selbstbestimmungsrecht der Völker – Herausforderung der Staatenwelt. Bonn: Dietz, 16-59.
Horowitz, D. L. (1985 ) Ethnic Groups in Conflict. Berkeley, CA: University of California Press.
Horowitz, D. L. (1990) ‘Ethnic Conflict Management for Policymakers’ in Conflict and Peacemaking in Multiethnic Societies, ed. by J. V. Montville. Lexington, MA: Lexington Books, 115-30.
Horowitz, D. L. (1991) A democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley: University of California Press.
Horowitz, D. L. (2002) ‘Constitutional Design: Proposals versus Processes’, in The Architecture of Democracy, ed. by A. Reynolds. Oxford: Oxford University Press.
Horowitz, D. L. (2003) Electoral Systems and Their Goals: A Primer for Decision-Makers, Journal of Democracy, vol. 14, no. 4 (October 2003), 115-27.
Horowitz, D. L. (2004) ‘The Alternative Vote and Interethnic Moderation: A reply to Fraenkel and Grofman’, Public Choice, vol. 121, no.s 3-4 (December 2004), 507-17.
Horowitz, D. L. (2006) ‘Strategy Takes a Holiday: Fraenkel and Grofman on the Alternative Vote’, Comparative Political Studies, vol. 39 (June 2006), 652-62.
Keating, M. Plurinational Democracy: Stateless Nations in a Post-Sovereignty Era. Oxford: Oxford University Press.
Kettley, C., Sullivan, J. and Fyfe, J. (2001) ‘Self-Determination Disputes and Complex Power Sharing Arrangements: A Background Paper for Debate’. Cambridge: Centre of International Studies, available online at http://www.polis.cam.ac.uk/research/cps/download/background1.pdf.
Kymlicka, W. (2007) Multicultural Odysseys: Navigating the New International Politics of Diversity. New York: Oxford University Press.
Lapidoth, Ruth (1997) Autonomy. Flexible Solutions to Ethnic Conflicts. Washington, D.C.: United States Institute of Peace Press.
McGarry, John and O’Leary, Brendan (1993) ‘Introduction. The Macro-Political Regulation of Ethnic Conflict’, in McGarry, John and O’Leary, Brendan, eds., The Politics of Ethnic Conflict Regulation. London: Routledge, 1-40.
McGarry, John and O’Leary, Brendan (2004) The Northern Ireland Conflict: Consociational Engagements. Oxford: Oxford University Press.
O’Leary, B. (2005) ‘Debating Consociational Politics: Normative and Explanatory Arguments’, in From Powersharing to Democracy, ed. by Sid Noel. Montreal and Kingston: McGill/Queen’s University Press, 3-43.
Potier, Tim (2001) Conflict in Nagorno-Karabakh, Abkhazia and South Ossetia. A Legal Appraisal. The Hague: Kluwer International Law.
Reilly, B. (2001) Democracy in Divided Societies: Electoral Engineering for Conflict Management. Cambridge: Cambridge University Press.
Roeder, Philip G. and Rothchild, Donald (2005) Sustainable Peace: Power and Democracy after Civil Wars. Itahca and London: Cornell University Press.
Sisk, T. D. (1996) Power Sharing and International Mediation in Ethnic Conflict. Washington, D.C.: United States Institute of Peace Press.
Wimmer, A. (2003) ‘Democracy and Ethno-religious Conflict in Iraq’, Survival, vol. 45, no. 4 (December 2003), 111-34.
Weller, M. and Wolff, S. eds. (2005) Autonomy, Self-governance and Conflict Resolution: Innovative Approaches to Institutional Design in Divided Societies. London: Routledge.
Wolff, S. (2004) “The Institutional Structure of Regional Consociations in Brussels, Northern Ireland, and South Tyrol”, Nationalism and Ethnic Politics, 10(3), 387-414.
Wolff, S. (2007a) ‘Complex Power Sharing as Conflict Resolution: South Tyrol in Comparative Perspective’, in Tolerance Established by Law Self-Government and Group Rights: The Autonomy of South Tyrol, ed. by Joseph Marko, Francesco Palermo, Jens Woelk. The Hague: Martinus Nijhoff Publishers, 329-70.
Wolff, S. (2007b) ‘Conflict resolution between power sharing and power dividing, or beyond?’, Political Studies, 5(3), 363-79.
Wolff, S. and Weller, M. (2005) ‘Self-determination and Autonomy: A Conceptual Introduction’, in Autonomy, Self-governance and Conflict Resolution: Innovative Approaches to Institutional Design in Divided Societies, ed. by Marc Weller and Stefan Wolff. London: Routledge, 1-25.
Wolff, S. and Weller, M. (2006) ‘Bosnia and Herzegovina ten years after Dayton: Lessons for internationalized state building’, Ethnopolitics, 5(1), 1-13.