Governing in Bosnia, Macedonia, and Northern Ireland

Power-Sharing Executives NATIONAL AND ETHNIC CONFLICT IN THE 21ST CENTURY

Brendan O’Leary, Series Editor Power-Sharing Executives

GOVERNING IN BOSNIA, MACEDONIA, AND NORTHERN IRELAND

Joanne McEvoy

UNIVERSITY OF PENNSYLVANIA PRESS

All rights reserved. Except for brief quotations used for purposes of review or scholarly citation, none of this book may be reproduced in any form by any means without written permission from the publisher.

Published by University of Pennsylvania Press Philadelphia, Pennsylvania 19104- 4112 www .upenn .edu /pennpress

Printed in the United States of America on acid- free paper

10 9 8 7 6 5 4 3 2 1

Library of Congress Cataloging- in- Publication Data McEvoy, Joanne. Power-sharing executives : governing in Bosnia, Macedonia, and Northern Ireland / Joanne McEvoy. — 1st ed. p. cm. — (National and ethnic confl ict in the 21st century) ISBN 978-0-8122-4651-3 (hardcover : alk. paper) 1. Representative government and representation—Case studies. 2. Minorities—Political activity—Case studies. 3. Ethnic groups—Political activity—Case studies. 4. Ethnic confl ict—Political aspects—Case studies. 5. Cultural pluralism—Political aspects—Case studies. 6. Bosnia and Hercegovina—Politics and government. 7. Bosnia and Hercegovina—Ethnic relations. 8. Macedonia—Politics and government. 9. Macedonia—Ethnic relations. 10. Northern Ireland—Politics and government. 11. Northern Ireland—Ethnic relations. I. Title. II. Series: National and ethnic confl ict in the 21st century. JF1061.M44 2015 352.23—dc23 2014012660 For Giancarlo and Luca This page intentionally left blank CONTENTS

1. Power Sharing, Institutional Design, and External Actors 1

PART I. NORTHERN IRELAND 2. Th e Sunningdale Executive: Lessons from Failed Power Sharing 39 3. Th e Good Friday Agreement 1998: An Inclusive Co ali tion 61 4. Th e 2007– 11 Executive: A New Era in Northern Ireland Politics? 87

PART II. BOSNIA AND HERZEGOVINA 5. Power- Sharing Stalemate in Post- Dayton Bosnia 107 6. From Dayton to Brussels? 132

PART III. MACEDONIA 7. Macedonia: From In de pen dence to the Ohrid Framework Agreement 159 8. Toward a Binational Macedonia? 182

Conclusion 204

Notes 227 Index 273 A c k n o w l e d g m e n t s 2 7 9 This page intentionally left blank CHAPTER 1

Power Sharing, Institutional Design, and External Actors

In December 2012, Belfast hit the world headlines again, the story far from positive. More than fourteen years aft er the Good Friday Agreement of April 1998, sectarian politics came to the fore amid loyalist riots lasting several months. Around fi fty police off ers were injured, several politicians received death threats, and property was damaged across the city. Clashes between loyalists and nationalists took place at sensitive interface areas, and a prison offi cer was murdered by dissident republicans. Th e violence was sparked by a vote by Belfast City Council to limit the fl ying of the Union fl ag at City Hall to designated days rather than every day of the year. In response to the ongoing protests, po liti cal leaders called for calm and sought to bolster the power-sharing executive. First Minister Peter Robinson (Democratic Unionist Party) referred to the “historic decisions” his party had made “to build a shared society in Northern Ireland.”1 Deputy First Minister Martin McGuinness (Sinn Féin) urged the parties to work together against “anti- peace pro cess” elements: “We are not going to kowtow or bow the knee to their activities.”2 Th e issue of equal treatment of groups’ symbols became po liti cally contentious in Bosnia and Herzegovina in 2004 following a request from the Chair of the Presidency Sulejman Tihić for the Constitutional Court to review the constitutionality of symbols on the fl ags of the two entities (the Federation of Bosnia and Herzegovina and Republika Srpska).3 Tihić claimed that the gold lilies on the Federation’s coat of arms and fl ag represented Bosniaks while the red and white squares represented Croats, thereby discriminating against Serbs.4 He also claimed that Republika Srpska’s fl ag was based on symbols solely from Serb history and that the entity’s anthem, 2 Chapter 1

“Bože Pravde,” discriminated against Bosniaks and Croats as it asked the Lord “to unite the Serb brothers, save the Serb king and the Serb lineage.” In March 2006, the Constitutional Court held that Articles 1 and 2 of the Law on the Coat of Arms and Flag of the Federation of Bosnia and Herzegovina and Articles 2 and 3 of the Constitutional Law on the Flag, Coat of Arms, and Anthem of the Republika Srpska contravened Article II.4 of the Constitu- tion of Bosnia and Herzegovina in conjunction with the International Con- vention for Elimination of All Forms of Racial Discrimination. Th e Court’s ruling upheld the notion that equal rights must be given to the three constituent peoples (Bosniaks, Serbs, and Croats) and all other citizens in both entities. Flags and symbols have also been the source of politi cal tension in Mace- donia. Th e Ohrid Framework Agreement of 2001, which ended the six-month confl ict between ethnic Macedonians and ethnic Albanians, stipulated that public authorities should have the right to display emblems representing the majority community in their municipality. Many ethnic Albanians prefer to identify with the Albanian fl ag—a double-headed black eagle on a red background—while accepting the Macedonian fl ag as the state fl ag. In June 2005, a government bill proposed that a community constituting more than half of a municipality’s population would have the right to raise its preferred fl ag alongside the Macedonian state fl ag. Although some ethnic Albanians protested that they were prevented from fl ying the Albanian fl ag in municipalities where they made up less than 50 percent, the government’s parliamentary majority ensured the bill was passed, and the opposition boycotted the vote in protest. Th ese three episodes demonstrate how symbols can become highly controversial when groups have opposing identities and nationalist aspirations. In cases of ethno-national confl ict, the contending groups’ identities are oft en bound up in symbols representing their respective cultures and traditions. Marc Howard Ross explores “cultural contestation” in deeply divided places where there are “issues of identity, recognition, and inclusion and exclusion that quickly come into play when leaders and groups evoke cultural images that stir up deeply held and clashing feelings.” 5 As politics in such contexts is oft en a zero-sum game, each group feels that the public expression of the other side’s symbols is a threat to their own identity. Symbols representing historical events and other controversial issues, such as fl ags and language rights, are among the challenges faced by power- sharing executives whereby representatives of the contending groups form a cross- community co ali tion. In this book, I investigate the successes and Institutional Design and External Actors 3 challenges of power- sharing executives in three contemporary cases: North- ern Ireland, Bosnia and Herzegovina, and Macedonia. Ultimately, I seek to add to the debates on what helps power sharing “work.” I proceed from the idea that the institutional rules that govern new politi cal arrangements are central to the realization of demo cratic stability in post- confl ict states.6 But deciding the most appropriate institutions is no easy task. As trust will almost certainly be lacking among the communities previously in confl ict, institutional designers face a considerable challenge in creating institutions that will encourage stability and help prevent the new system from collaps- ing into further violence. Power sharing is a form of government recommended for deeply divided territories where majoritarianism is not a fair, realistic option to promote peace. It brings together representatives of groups previously in confl ict to govern the country together. Th e country may have been devastated by war or wracked by long-running confl ict. Many people may have lost their lives; many more have lost their loved ones, their home, or their job. Th ey may have been intimidated, bereaved, or discriminated against for their identity and emotionally scarred by the damaging eff ects of confl ict. It is hardly surprising, then, that power sharing among communities is not easy. Even in a new po liti cal dispensation, with an end to violence and the setting up of new po liti cal structures, it would be naïve to expect that government by former foes will be plain sailing. By focusing on three power-sharing cases, I want to show what power sharing is really like and to understand what this practice means for power- sharing theory. I demonstrate that power sharing is a diffi cult, challenging form of government. But I also have a normative commitment to power sharing. I believe it is a feasible way to govern a post-confl ict territory with two or more ethno-national groups where majoritarianism is inappropriate. I would happily sit within the camp of “cautious optimism” about the “principal viability of power-sharing institutions.”7 And I view power sharing to be potentially “democracy-enabling” and “conducive to peacebuilding.”8 Th ere are, of course, a number of ways power sharing can be designed. It is clearly not a monolith. Neither is it a panacea for resolving confl ict.

Aims, Methods, and Contribution

My overall aim is to understand power- sharing government in deeply divided places. Understanding po liti cal institutions is, of course, “a serious 4 Chapter 1 endeavor” that continues to preoccupy scholarship and has important real- world implications.9 As Elinor Ostrom writes, “Th e opportunities and constraints individuals face in any par tic u lar situation, the information they obtain, the benefi ts they obtain or are excluded from, and how they reason about the situation are all aff ected by the rules or absence of rules that structure the situation.”10 Having such an impact on the opportunities and challenges faced by politi cal actors in territories wracked by violent confl ict means that institutional rules are crucial for peace and stability. First, I want to explain the institutional design process of cross-community power- sharing executives. As yet, the power- sharing literature does not explore how such systems are formed. Here I respond to the still- relevant call in the academic literature for a greater focus on the “formation and transformation of structures.”11 Questions of institutional formation and change are crucial in post-confl ict situations because the design pro cess arguably has a bearing on the stability of the new system. Is there a process of institutional design in post- confl ict societies? Who are the designers and why do they opt for partic u lar arrangements? What is the role of external actors? And why are power- sharing executives revised once created? Second, I seek to uncover which institutional factors are likely to foster elite cooperation within a power- sharing executive. I focus on two potential explanatory factors for cooperation in power- sharing executives: institutional rules and the role of external actors. Regarding institutional rules, I explore institutional choices over executive formation (the use of a sequential portfolio allocation method versus inter-party bargaining) and the impact of group vetoes. As the three cases use diff erent institutional rules, can we say some are more likely to promote inter-ethnic cooperation? And what role do external actors play? In their peace implementation role, do external actors incentivize cooperation among the local elites? Under what conditions are these incentives eff ective? Th e project is a small- N qualitative study comparing the practice of power-sharing executives in Bosnia, Macedonia, and Northern Ireland. Th e three cases are examples of the general phenomena under investigation— power-sharing executives established aft er an ethno-national confl ict. Th ey are suffi ciently similar, with intriguing variation to be analyzed in a comparative framework. Brendan O’Leary has described these three cases as examples of “complex consociation” due to the creation of institutional arrangements to settle a self-determination dispute, the development of a peace process, an additional strategy such as territorial autonomy, and the involvement of international actors in the implementation of the agreement.12 Th e Institutional Design and External Actors 5 self- determination disputes have been between nationalists and unionists in Northern Ireland; between Bosnian Serbs, Bosnian Croats, and Bosnian Muslims (Bosniaks); and between ethnic Slav Macedonians and ethnic Al- banians. Th e respective peace processes resulted in the Good Friday/Belfast Agreement of 1998, the Dayton Peace Agreement of 1995 (DPA), and the Ohrid Framework Agreement of 2001. Northern Ireland also had an earlier peace agreement, the Sunningdale communiqué of 1973, which is discussed in Chapter 2. An important focus of the study is the role of external actors in the formation and operation of power sharing. External actors have been involved in all three cases, seeking to uphold the respective agreements and promote inter-ethnic cooperation. In Northern Ireland, the British and Irish governments contributed an enormous amount of time and energy to secure the agreement in 1998 and to restore power sharing in 2007.13 A plethora of international organizations has been involved in Bosnia since the outbreak of war in 1992. Th e institution in charge of facilitating the implementation of the DPA, the Offi ce of the High Representative (OHR), has been heavily involved in the governance of the country. In recent years, the Euro pe an Union has become the principal actor guiding the country’s progress toward EU integration. In Macedonia, the EU and the U.S. were instrumental in bringing about the Ohrid Framework Agreement, and the two actors have since played a signifi cant role in encouraging the implementation of the agreement. In partic u lar, the EU delegation has been a key actor in persuading the parties to agree on policy reform. Although the three cases are similar for these reasons, we need to try and understand what accounts for their relative stability. Power sharing in Bosnia and Herzegovina has certainly been diffi cult since the signing of the DPA. Trust among the parties representing the contending groups remains elusive. Th e state-level Council of Ministers remains weak in contrast to the strength of the two entities. Although the OHR has been working toward closure since 2006, it has been cautious of pulling out too soon and has been frustrated at the slow pace of reform. Yet the reforms required by the EU and the lure of membership are not suffi ciently compelling to overcome divisions and po liti cal wrangling among the parties. Some politicians prefer to block rather than work the system in a positive manner. Th e principal problem relates to the parties’ ongoing dispute over the future structure of the state: Although Bosnian Serbs want their entity, Republika Srpska (RS), to be the locus of power and control, Bosniaks want to see a centralized unitary 6 Chapter 1 state, and Bosnian Croats want to ensure equality for their group while also, at times, calling for their own entity. Unsurprisingly, the lack of consensus on state structures, including power-sharing institutions, has led to a pains- takingly slow pro cess of constitutional reform. In Northern Ireland, the record of cross-community power sharing is more positive. Th is was not always the case. Th e fi rst power- sharing executive (1973–74) was hampered by inter- and intra-executive competition and fi nally collapsed under the weight of opposition from more extreme forces. Twenty-fi ve years later, the second (more inclusive) power- sharing executive experienced diffi culties, largely over paramilitary decommissioning, and was suspended on several occasions by the British government between 1999 and 2002. In 2007 and 2011, the third and fourth power-sharing executives came into being. At the time of writing, the two largest parties—Sinn Féin and the DUP— are willing to work the system. Th ere have, of course, been considerable challenges relating to the devolution of policing and justice and intra- executive divergence over controversial policy issues. Never- theless, the assessment of power sharing in Northern Ireland since May 2007 has been a positive one. Th e governing parties have even sought to ex- port the lessons from the Northern Ireland peace pro cess to groups involved in ethnic confl ict elsewhere.14 In Macedonia, power sharing has also been relatively successful. Cross- community executive power sharing between the Macedonian and Alba- nian communities has operated since the country became in de pen dent in 1991.15 In the years since the 2001 confl ict, the parties made progress on the implementation of the Ohrid Framework Agreement (OFA). Yet power sharing was challenged over executive formation procedures in 2006, a series of parliamentary boycotts and stalemate over the name dispute with Greece leading to brakes being set on the country’s Euro- Atlantic integration. Although power sharing appears fairly stable in Macedonia as both communities support the politi cal structures in an increasingly bi-national state, contentious identity politics have served to challenge stability. To sum up, power sharing in Bosnia has been and remains considerably diffi cult while Northern Ireland and Macedonia deserve a more positive assessment. So why is this so? Can this variation tell us anything more general about cross- community executive power sharing? Potential methodological criticisms might suggest that a small- N study based on three cases will not be able to tell us much about what causes Institutional Design and External Actors 7 power-sharing executives to be stable. Some researchers might argue that causal explanations for democratic stability are only possible with a large-N time-series cross-national analysis. Pippa Norris’s work provides such an analysis as well as paired qualitative case studies. She concludes that power- sharing institutions improve the chances of demo cratic governance.16 Al- though she argues that small-N case studies are “invaluable” and “allow researchers to develop theories, to derive hypotheses, and to explore causal mechanisms,” she suggests that the “method fails to resolve the debate between proponents and critics of power sharing, however, since the danger of potential selection bias means that diff erent cases can be cited on both sides.”17 In another large- N study, Michaela Mattes and Burcu Savun explore civil war recurrence with regard to forty- six civil wars and conclude that po liti cal power sharing and third-party guarantees help reduce the likelihood of further confl ict.18 In comparison with these studies, does my research provide enough cases to explain variation in the operation of power- sharing executives? Ar- guably, a time- series cross- national analysis would be inappropriate for the objectives of the study. Concepts such as cooperation, accommodation, or jointness are hard to quantify. Moreover, there is a limited number of relevant cases for any investigation of executive power sharing. Although the book focuses on three cases, each includes diff erent experiences in power sharing (Northern Ireland 1974; Northern Ireland 1998; post-DPA Bosnia; pre-Ohrid Macedonia; post-Ohrid Macedonia). Th ese experiments have also experienced institutional change and a variety of institutional rules. And although the three cases are predominantly consociational, there have been some integrative elements (more “moderate,” exclusive coali tions and partic u lar institutional rules, e.g., the election of the fi rst minister and deputy fi rst minister in Northern Ireland). Overall, I provide a comprehensive, in- depth, and rich qualitative account of the practice of power- sharing executives. I treat institutional rules and the role of external actors as the principal determining factors for cross- community cooperation in power- sharing government. And the study is not about taking “sides” between ad- vocates and critics of power sharing. As already noted, I proceed from the notion that power sharing is an appropriate form of government for post- confl ict societies. Moreover, I develop comparative insights for cases where power sharing has witnessed considerable diffi culties, namely Lebanon and Cyprus. 8 Chapter 1

Th e book’s fi ndings have implications for power-sharing practice and power- sharing theory. In relation to executive design, I demonstrate that power sharing comes about as the outcome of interaction among a multiplicity of actors. Th ese actors may have similar preferences, or their preferences may be extremely divergent. Yet the accommodation of these preferences (through concessions, side payments, or compromise) can lead to executive power sharing. Th e pro cess of executive design will likely be infl uenced by external actors (who may even impose a settlement), the state’s historical experience, or the evolution of power- sharing proposals over time. I also show that executive power sharing as stipulated in a peace agreement is not fi xed forevermore but has the capacity to change. Changing po liti cal circumstances may bring about institutional revision, sometimes in a pragmatic fashion, to allow the continuation of power sharing among the contending groups. Revision of power-sharing rules can also come about through efforts to address shortcomings or dysfunctional rules. Notably, in the context of a fragile post- confl ict environment, parties opposed to institutional revision may resist such eff orts as a threat to group interests. Beyond questions of institutional formation and transformation, the book has implications for ongoing debates on what works in power sharing. Based on rich analysis of the functioning of variable institutional rules, I argue that although inter- ethnic cooperation is far from guaranteed, power sharing can provide po liti cal space for moderation via joint decision making. I argue in support of power sharing, with qualifi cations: Parties in government must have renounced violent means to pursue their politi cal goals, and they must be willing to operate within the politi cal system. I advocate inclusion as a means to provide the opportunity for cooperation and moderation. It can be a mistake to assume that a limiting exclusionary focus on the “moderate middle” will generate cooperation. In relation to executive formation rules, I concur with existing literature that liberal power sharing is preferable to corporate rules that specify positions for certain groups.19 Th ough negotiations are appropriate if parties are willing to compromise over who gets what, a sequential portfolio allocation method may be benefi cial in relieving parties of diffi cult negotiations over the distribution of ministerial seats. Regarding veto rules, the fi ndings suggest that less is more: Veto issues should be clearly defi ned in legislation (limited to identity, education, security, and the budget) and constrained in terms of few veto mechanisms. Finally, I contribute to the power-sharing literature by clarifying the mechanisms by which Institutional Design and External Actors 9 external actors incentivize domestic elites to agree on and maintain power sharing. Th e research relies on and adds to work on rationalist mechanisms of socialization whereby the promotion of power sharing by external actors and the adoption of power sharing by the internal contending groups are explained by instrumental/strategic behavior. External actors’ strategic behavior is based on positive and negative (even coercive) incentives. Th ese incentives will likely be more eff ective when they uphold a peace agreement that satisfi es groups’ preferences on constitutional issues. External incentives can, under certain conditions, lead to internalization and the potential “habitualization” of power sharing as norm-conforming behavior.20 Yet the strategy of external actors will be less eff ective when their socialization eff orts are inconsistent and coercive or are viewed as threatening to one or more of the contending groups.

Experience of Conﬂ ict

All three cases have experienced the destructive consequences of violent confl ict, albeit to a diff erent degree in the number of lives lost. Th e Northern Ireland “Troubles” endured for more than thirty years and claimed more than 3,600 lives in addition to thousands of people injured, bereaved, and intimidated. As a result of the 1992–95 Bosnian war, more than 100,000 people lost their lives and 2.3 million became refugees— more than half of the pre-war population of 4.4 million. Macedonia escaped the devastating violence following the collapse of the former Yugoslavia in the early 1990s until the outbreak of a six- month period of armed confl ict in 2001 that resulted in approximately 130 killed (though each group cites fi gures higher than the offi cial numbers). In the aft ermath of the 2001 confl ict, however, tensions continued to run high and the confl ict served to aggravate relations between Macedonians and Albanians, which had been historically diffi cult. A popu lar explanation for Bosnia’s diffi culties, relative to success in Northern Ireland and Macedonia, is that it is hardly surprising that the system is oft en blocked and that parties do not cooperate, given the impact of the devastating wars from 1992 to 1995. Certainly, the Bosnian wars have had a tremendous impact on the local people as they come to terms with what happened and attempt to move on from this dark period. Positing variation in confl ict as the causal explanation for power-sharing stability is, 10 Chapter 1 however, too simple. To illustrate how the experience of confl ict cannot be seen as a persuasive inde pen dent variable, we need to consider the historical experience of confl ict in each case. Th is demonstrates that all three have witnessed considerable confl ict during several periods over time.

Northern Ireland

Th e roots of the Northern Ireland confl ict trace back to the Anglo- Norman invasion of 1169. Th e conquest of Ireland by En glish monarchs in the sixteenth century led to the Plantation of Ulster in the seventeenth century, which involved the transfer of almost all land from the native Gaelic Irish to Scots and English settlers. In the eighteenth century, a series of rebellions sought Ireland’s inde pen dence from England. Th e United Irishmen staged an insurrection across Ireland in 1798, including the Ulster counties of Antrim and Down. Across the country more than 20,000 people were killed, and the rebels were ultimately defeated by the British forces.21 Th e 1798 rebellion preceded smaller rebellions in 1803, 1848 and 1867 designed to challenge the union of Ireland and Great Britain. In the second half of the nineteenth century, sporadic instances of intercommunal violence took place, including the 1864 riots in Belfast.22 A few decades later, Irish nationalists orga nized a campaign for self-government. Th e Home Rule campaign led to (defeated) bills in the UK parliament in 1886 and 1893. Unionists in the north opposed Home Rule and wanted Ireland to remain part of the United Kingdom under the Act of Union 1801. Th e postponement of the Home Rule issue due to the outbreak of World War I and the re sis tance of Ulster Unionists prompted the Irish Republican Brotherhood (IRB) to or ga nize an insurrection, the 1916 Easter Rising. In response, British troops arrived in Dublin and the rebels surrendered aft er fi ve days. Although the rising failed militarily, the execution of sixteen of its leaders led to growing public sympathy for the goal of Irish inde pen dence. When the Home Rule issue once more dominated the political agenda, the British government sought a compromise in the form of the Government of Ireland Act 1920, creating two new jurisdictions with the respective Dublin and Belfast parliaments. Northern Ireland comprised the six counties of Ulster, excluding Cavan, Donegal, and Monaghan. Th e birth of the new “statelet” met considerable intercommunal violence. In the Institutional Design and External Actors 11 fi rst few years, hundreds of people were killed and as many as 23,000 were driven from their homes.23

Bosnia

Apart from a brief period of inde pen dence in the medieval era, Bosnia has been under the control of several external powers: the Ottomans from 1463, the Hapsburgs from 1878, the Serbs under the fi rst Yugo slavia, the Croatian fascist regime during World War II, and the communists in the second Yugo- slavia. Even since inde pen dence in 1992, and particularly since the DPA, Bosnia can be regarded as a semi- protectorate of the international community. Although the ethnic groups in Bosnia coexisted peacefully for periods throughout its history, the peace has, at times, been punctuated by violence among the diff erent groups and against the ruling powers. For instance, numerous anti- Ottoman rebellions took place, particularly aft er the 1860s. A major uprising was orga nized in 1875 when peasants revolted against taxes imposed on their harvest. Th e resis tance transformed into a movement for liberation and spread across Bosnia, leading to severe retribution by the Ottoman forces with at least 5,000 people killed and possibly as many as 250,000 refugees.24 Austria- Hungary annexed Bosnia following the 1878 Congress of Berlin. Th e Hapsburgs’ occupation of the country met a determined Muslim resis- tance, which was eventually crushed aft er a number of months.25 Th e assas- sination of Austrian Archduke Franz Ferdinand in Sarajevo by a Bosnian Serb, Gavrilo Princip, was the infamous event that triggered World War I. During the war, approximately 10,000 Bosnians were killed or wounded and the country reportedly suff ered a 2 percent population defi cit of 40,000 people between the 1910 and 1921 censuses.26 During World War II, Bosnia experienced a series of inter- ethnic atrocities. Yugo slavia surrendered to Hitler in April 1941, and Bosnia was absorbed by the fascist government controlled by the Ustaše in Croatia. Two main re sis tance groups were set up against the German and Italian occupiers and the Ustaše: the Četniks (Serbian guerrilla units) and Tito’s multi-ethnic Partisans. In addition to the anti- fascist struggle, Bosnia experienced an inter- ethnic war, mainly Serbs fi ghting against Muslims and Croats. More than a million Yugo slavs died in the war and millions were wounded. Almost half of Yugoslavia’s 12 Chapter 1 war deaths occurred in Bosnia.27 In the early period of the postwar regime, as many as a quarter of a million people were killed as Tito sought to remove anti- Partisan forces.28

Macedonia

Macedonia has been a site of confl ict for centuries. Ruled by the Ottomans for more than 500 years from 1371, it has also been the target of territorial and cultural claims by its larger and stronger neighbors: Albania, Bulgaria, Greece, and Serbia. By the late nineteenth century, Macedonian Christians aspired to in de pen dence from Ottoman rule. Th e Internal Macedonian Rev- olutionary Orga ni za tion (IMRO) was set up as a resis tance movement and launched a series of rebellions. According to Robert D. Kaplan, Macedonia “was to become the original seedground not only of modern warfare and politi cal confl ict, but of modern terrorism and clerical fanati cism as well.”29 Th e Ilinden (St Elijah’s Day) Uprising of 1903, which established a short- lived Macedonian republic at Krusevo, holds a special place in Macedonian collective memory. Th e Turks’ retribution for the uprising led to thousands of murdered civilians and rape victims, destroyed villages, and 50,000 refugees. Th e Balkan Wars (1912– 13) were fought largely for the control of Mace- donia when Bulgaria, Serbia, and Greece declared war on the Ottoman Em- pire.30 Th e First Balkan War ended in December 1912; Macedonia had been freed from Ottoman rule but was divided between the victors. As the Greeks and Serbs tried to limit Bulgarian infl uence, Bulgaria retaliated with an attack on Serbian forces in June 1913, triggering the Second Balkan War. At the end of the war just a few weeks later, Bulgaria was forced to surrender almost all land gained. Serbia received part of Macedonia, which became part of the fi rst Yugoslavia. Two hundred thousand combatants had been killed throughout the region, and Macedonia had witnessed the greatest violence and devastation. As World War I broke out, Bulgaria joined the side of the Central Powers (Germany and Austria-Hungary) in an eff ort to regain Macedonia. From 1916 to 1918, the French, Greeks, Serbs, and British Com- monwealth armies fought against the Bulgarians and Hapsburgs in Mace- donia. Bulgaria tried again to take the country by siding with the Germans in World War II. When the fi rst Yugo slavia surrendered in 1941, Macedonia was carved up between Bulgaria, Germany, and Italy. Th e Bulgarian army Institutional Design and External Actors 13 was then driven out by Serb, Greek, and British forces, and Macedonia became one of six republics in Tito’s Federal Yugo slavia.31

Designing and Reforming Power- Sharing Executives

One aspect of this study is to analyze the pro cess of institutional design: how power-sharing executives come into being, why they take the shape they do, and how they change over time. Mentioned above, this objective responds to the institutionalist concern to focus on institutional formation and transformation. Pippa Norris suggests we “need to understand how power-sharing constitutions arise and what process of bargaining and pact- making leads to their accep tance and implementation.”32 In keeping with the “new institutionalism” that entered the lexicon of politi cal science in the 1980s, this book’s approach argues for a “more general view of the place of institutions in politics and the possibilities for a po liti cal theory which is more attentive to them.”33 So an important focus for this book is the formation and change of power- sharing executives. As Paul Pierson notes more generally, “Th e origins of institutions, as well as the sources of institutional change, remain opaque.”34 To clarify the origins of power- sharing executives in the three cases, to what extent can we say a pro cess of institutional design existed? Can we say that the power- sharing executives in Bosnia, Macedonia, and Northern Ire- land were actually designed? We might argue that executive design in these three cases was simply part of a broader peace settlement to resolve ethnic confl ict. Th is description, however, needs to be qualifi ed. In Macedonia, for example, a tradition of power sharing developed in the initial post- independence period. Th e OFA did not address executive design because power sharing had become an accepted and expected convention following the end of communist rule; the agreement cemented power sharing by adding features such as the concurrent majority voting procedure in parliament. In Northern Ireland, executive design using the d’Hondt procedure to allocate ministerial seats in a power-sharing government resulted from policy “learning” by parties and the evolution of po liti cal initiatives from the 1970s.35 Th ere was, however, an element of design in relation to the creation of the ( joint) posts of fi rst minister and deputy fi rst minister agreed by the two main parties in the 1998 negotiations, the Ulster Unionist Party (UUP) and the Social Democratic and Labour Party (SDLP). Th e system of power 14 Chapter 1 sharing in Bosnia was more certainly “designed” at the Dayton negotiations and imposed by the international community, particularly the U.S. Th us, this book considers the various ways in which power- sharing institutions come to be designed and the interaction among a multiplicity of internal and external actors. We might say that the Bosnian arrangements were designed at Dayton, that there was an informal process of executive design in Macedonia, and a largely evolutionary pro cess of institutional learning took place in Northern Ireland. Institutional design has been described as both “a process aimed at producing prescriptions” and “a pattern or plan which can be detected, or imagined, in existing institutional structures without any reference to the processes that produced the pattern.”36 Importantly, design requires political actors to determine “how institutions might be, and ought to be, constructed.”37 Yet the wider literature has since questioned the reality of institutional design in post-confl ict states. Donald Horowitz considers institutional design in deeply divided societies to be an oxymoron. He argues that “constitutions that have been designed, as opposed to merely constructed, are diffi cult to fi nd.”38 For Horowitz, a multiplicity of actors makes design unlikely: “Th e sheer proliferation of participants makes it less, rather than more, likely that a design, with its consistent and interlocking parts, will be produced at the outset and adopted at the conclusion.”39 Robert Goodin notes that the “Myth of the Institutional Designer (still less the Myth of the Institutional Design) is greatly to be avoided in theories of institutional design. Th ere are just lots of localized attempts at partial design cutting across one another, and any sensible scheme for institutional design needs to take account of that fact.” 40 But it is important to be clear on what we mean by institutional design and to make sense of what a design process might involve in an ethnic confl ict. Th e creation of institutions in a post-confl ict situation is usually part of a peace pro cess in which negotiating parties, oft en under the aegis of a third party, seek to arrive at an agreement. Horowitz is right that the project of institutional design oft en includes the proliferation of actors. Yet this proliferation does not necessarily mean that institutional design is unlikely. Much depends on the preferences of the internal and external actors involved in the negotiations and the interaction among those preferences. Actors from diff erent parties may hold common positions on the appropriate institutional framework or they may have divergent positions and expectations on the bargaining pro cess and potential outcomes. Both scenarios involve in- Institutional Design and External Actors 15 stitutional design. Institutional design in post-confl ict situations is unlikely to be a matter of actors agreeing to a blueprint or a single designer working to a preconceived plan. Th e outcome is the result of the interaction and ultimately compromise among actors’ preferences. It is also important that institutional design can be of an evolutionary nature whereby institutional rules carry over from previous negotiations to form part of a new peace agreement. Horowitz suggests that institutional designers need to watch out for a number of common diffi culties in deeply divided societies. For instance, designers may have imperfect knowledge of the range of potential institutions and a “failure of expertise” may occur as “provision merchants” seek to in- stall their preferred institutions that might not fi t the needs of the country. He argues there can be “disjunctions between what severely divided societies require and the methods that are used to decide on the institutions that will govern these societies.” 41 It may certainly be the case that actors will not have a thorough grasp of all the potential institutional solutions. Given this lack of knowledge of all possible options, actors may plump for what they are familiar with from the country’s history or their neighbors’ experience. To Horowitz’s argument for consistency and coherence in institutional design (preferably in terms of an incentives-based ideal type, discussed below), Philippe Van Parijs replies that a mélange or hybrid approach may still produce a centripetal outcome. He suggests that what matters is fi nding a set of institutional rules that together “systematically defuse potential confl ict.” 42 All of these potential diffi culties and challenges lead us to ask, Who are the institutional designers in the study? Do they have choices? Why do they choose a par tic u lar executive design? Th e institutional designers in the three cases reviewed here are the internal and external po liti cal actors who took part in the respective peace processes to reach agreement on po liti cal institutions. Yet these actors may have had diff erent purposes and oft en divergent objectives for the new politi cal system. Th e internal po liti cal parties who represent opposing groups will likely have diff erent positions from one another. If external actors are involved, they may have diff erent preferences from the internal politi cal parties. As noted above, external actors have played an important role in the negotiation and implementation of politi cal institutions in the three cases reviewed here. Th e U.S. (though joined by representatives of the other Con- tact Group countries: the UK, France, Germany, and Russia) was the lead author of the DPA. Th e U.S. and the EU took part in the negotiations at 16 Chapter 1

Ohrid, Macedonia in 2001. In Northern Ireland, the British and Irish governments devoted considerable time and eff ort to the negotiations leading to the 1998 Agreement; the Clinton administration was also infl uential, not least due to the input of former U.S. Senator George Mitchell. To explore the pro cess of why and how actors agree to a par tic u lar institutional framework, it is helpful to consider the theme of intentionality. Goodin suggests that social change may occur (1) by accident, (2) as the outcome of evolution, or (3) as the product of intentional intervention.43 For Goodin, intentionality is a key factor in the emergence and change of institutions. He notes that agents, “individually or more oft en collectively, sometimes fi nd themselves literally asked to decide which sort of social arrangements they would prefer to retain and reproduce.” 44 Importantly, this intentionality does not necessarily include the input of designers because institutions may emerge or evolve at their own pace. Goodin contends that “Institutions are oft en the product of intentional activities gone wrong— unintended by- products, the products of various intentional actions cutting across one another, misdirected intentions, or just plain mistakes.” 45 Th us, institutional design may not necessarily be a clear- cut outcome of actors’ intentional interventions. If we say that institutional design may result from a number of intentional actions or even misdirected intentions, we need to acknowledge the limitations of functional explanations. Jon Elster suggests there are conditions that need to be met in order to say that an institution can be explained by its function for a group. He says a functional explanation “can succeed only if there are reasons for a feedback loop from the consequence to the phenomenon to be explained.” 46 Elster brings us back to consider intentionality as a more appropriate mode of explanation in the social sciences. Intentional explanation “involves showing that the actor did what he did for a reason.” 47 Pierson also critiques the functional explanation that institution X exists because it serves the function Y. He claims that the “most straightforward version of rational institutional design [that] focuses on the intentional and farsighted choices of purposive, instrumental actors” is simplistic.48 If it were simply a matter of rational design, “institutional ef- fects should be seen as the intended consequence of their creators’ actions— and in that sense as supplying the explanation for why the institution takes the form it does.” Pierson advances a number of limitations to the explanation to what he terms “actor- centered functionalism.” 49 He suggests that actors may have multiple goals and that institutional functioning cannot be Institutional Design and External Actors 17 derived easily from the preferences of designers; actors may be instrumental but not farsighted; major institutional eff ects may be unintended; and broader social change may impact adversely on the fi t between actors and institutions. Th is theoretical discussion on functional explanations and actors’ intentionality is important for understanding why executive design took a par tic- ular form in these three cases. We need to remember that actors may fi nd themselves facing their po liti cal enemies across a negotiating table, tasked with identifying and agreeing upon an appropriate institutional framework for a post-confl ict environment. But even though we might say there is a moment of institutional “choice” in the context of peace negotiations, it is important to recall that institutions have the capacity to evolve in their own way over time. We also need to consider the overlapping or diverging preferences of the various actors, as demonstrated by the diffi cult peace talks in Northern Ireland and the “hothouse” atmosphere at Dayton. Moreover, actors’ intentions may lead to unintended consequences that give rise to a need for institutional reform to correct “mistakes” or respond to changes in the wider politi cal landscape. As Pierson writes, “Anyone engaged in empirical research in the social sciences knows that the most instrumental and canny of actors still cannot hope to adequately anticipate all the consequences of their actions. Institutions may not be functional because designers make mistakes.” 50 We need to consider whether institutional change has taken place as the outcome of an evolutionary process, the fusion of actor preferences, or the result of broader institutional dynamics.51 Institutional change has been a feature of power sharing in all three places. Important changes were made to executive design in Northern Ire- land by the St. Andrews Agreement, legislated in the Northern Ireland (St. Andrews Agreement) Act 2006. As the DUP and Sinn Féin were the largest parties of the respective blocs since 2003, inter-party talks focused on how to persuade these two parties to agree to form an executive (including the UUP and the SDLP).52 Under the executive formation rules in the 1998 Agreement, the DUP was not prepared to vote for a Sinn Féin deputy fi rst minister. Re- vised rules, therefore, allowed for straightforward nomination rather than election of the top two posts. Second, amendments to the 1998 Act centered on the accountability of individual ministers to the assembly. For the DUP, these changes were an important precondition for reaching agreement. A number of amendments placed provisions in the 1998 Act on a statutory footing and introduced additional mea sures, including the rule for three 18 Chapter 1 ministers to trigger a cross-community vote within the executive on another minister’s decision when a vote is required. In Bosnia, the nature and scope of executive power sharing has evolved since the DPA. First, the state powers have been strengthened following legislation imposed by the OHR that increased the Council of Ministers portfolios from four to eight.53 Second, a number of revisions were made to the institutional rules for power sharing following the Constitutional Court’s decision in 2000 that declared the ethnic predominance of one or two groups in both entities to be unconstitutional.54 Th e court ruled that the three “constituent peoples” should have equal rights throughout Bosnia, thus requiring reform of the entity constitutions because Bosnian Serbs were not recognized as such in the Federation, and the RS operated without recognition of Bosniaks and Bosnian Croats. Following the court’s decision, the OHR established commissions in the two entities to plan the pro cess of constitutional reform. Th ough inter-party talks on the issue came close to agreement, a decision was eventually imposed by the OHR in 2002 in the form of amendments to the entity constitutions. In the RS, the president and vice-presidents may not be from the same constituent people, and the executive must include eight Serb, fi ve Bosniak, and three Croat ministers. In the Federation, the constitutional amendment stated that the executive shall be composed of eight Bosniak, fi ve Croat, and three Serb ministers. Since the tenth anniversary of the DPA, the international community has pushed for constitutional reform in Bosnia. Florian Bieber refers to “two parallel debates on reform, within the international community and policy circles and among politicians and intellectuals within Bosnia and Herze- govina.” 55 Th e international community promotes centralization and the transfer of powers from the entities to the state level. Following much inter- party wrangling, the parties negotiated a package of constitutional reform in 2006. Known as the “April package,” it failed to pass the state assembly by just two votes. Constitutional reform has remained politi cally contentious as the parties continue to disagree over what kind of state Bosnia should be. Th is question is likely to become increasingly important as the country draws closer to fulfi lling EU requirements and particularly in relation to the implementation of EU law in the acquis communautaire. Although power sharing has been an informal tradition in Macedonia since in de pen dence in 1991, the Ohrid Agreement cemented the power- sharing system. Th e agreement did not introduce legislation requiring the two groups to share power, but it provided for minority protection under Institutional Design and External Actors 19 the locally termed “Badinter principle.” 56 Legislation in certain areas requires an overall majority in the assembly and a majority of deputies not in the majority community. Th e device acts as a mechanism for minority veto. Since the crisis of executive formation in 2006, there has been some discussion on whether the constitution should be reformed to extend the rule to executive formation and approval of the bud get.

Power- Sharing Theory

Th e overarching concern of this book relates to the success of power sharing, a topic that has provoked considerable and, at times, intense debate within the academic literature. For several decades, debate has raged over which politi cal structures provide the most appropriate framework for deeply divided places. Some literature refutes the potential of power sharing as a confl ict resolution measure. Donald Horowitz suggests that consociationalism leads to the entrenchment of ethnic divisions and is “inapt to mitigate confl ict in severely divided societies.” 57 Philip Roeder and Donald Rothchild maintain that power-dividing institutions are more stable than power- sharing institutions in ethnically divided territories.58 Among exponents of power sharing, an important debate exists on what kind of power sharing should be promoted in such cases. Discussion on the most appropriate institutional design focuses on the accommodation-versus- integration debate. Th e background and focus of these debates are dealt with comprehensively elsewhere.59 Nevertheless, we need to consider the normative and prescriptive positions of both approaches relating to executive design. Th ere are two principal frameworks for designing a new po liti cal framework in deeply divided territories aft er the termination of violent confl ict: accommodation and integration. As Sujit Choudhry notes, these two approaches are based on “fundamentally diff erent assumptions over the dura- bility of ethnocultural identities [that] reframe the debate over constitutional design for divided societies.” 60 Accommodationists “insist that in certain contexts, national, ethnic, religious and linguistic divisions and identities are resilient, durable and hard, rather than malleable, fl uid, soft , or trans- formable.” 61 Integrationists, however, “reject the idea that ethnic diff erence should necessarily translate into politi cal diff erences. Th ey argue for the possibility of a common public identity, even in the midst of considerable ethnocultural diversity.” 62 In terms of ethnic diversity, integrationists oppose the 20 Chapter 1 public institutional recognition of group identities while accommodationists promote and respect multiple public identities. Consociation (along with centripetalism, multiculturalism, and territorial pluralism) is an accommodationist option for managing divisions in a multi-ethnic state.63 Consociation is considered a feasible option based on four elements fi rst proposed by Arend Lijphart: cross-community executive power sharing, minority vetoes, segmental autonomy, and proportionality. Representatives of the main groups cooperate in government while each group has control over its vital interests. Lijphart’s classic consociational statement recommends the formation of a “grand coali tion” whereby the representatives of the main segments in society should cooperate in government.64 Th is maximal degree of inclusivity in consociational executives has been revised by Brendan O’Leary and John McGarry, who argue that consociation need not be all-encompassing. Consociations may be “complete,” “concurrent,” or “weak” (later revised as “plurality executives”), and what matters is “meaningful cross-community executive power sharing in which each signifi cant segment is represented in government with at least plurality levels of support within its segment.” 65 McGarry and O’Leary write that a consociation requires “some cross- community jointness and proportionality.” 66 Further distinction can be made between corporate and liberal consociation and between formal and informal power- sharing rules. Executive formation rules may be corporate in that they secure executive positions for the contending groups via predetermined positions according to ascriptive identities such as religion or ethnicity. 67 Alternatively, executive formation rules may be liberal in that positions are accorded to whichever groups are victorious in democratic elections. Bosnia is clearly a case of corporate consociation as positions are assigned to representatives of the three “constituent peoples” and on an entity basis. Northern Ireland has both corporate and liberal power- sharing rules. As set out in the Good Friday Agreement, the positions of fi rst and deputy fi rst minister are assigned to one unionist and one nationalist and are subject to a cross-community vote in the assembly. Discussed in Chapter 4, this rule was amended following the 2006 St. Andrews Agreement. Th e d’Hondt method is used to assign ten ministerial seats to parties on a proportional basis according to their strength in the assembly. Macedonia is almost wholly liberal, save the Committee on Inter- Community Relations, which stipulates positions among Macedonians, Albanians, and smaller minorities. Northern Ireland and Bosnia are cases of formal power sharing (as set out in the 1998 Northern Ireland Act and the Institutional Design and External Actors 21

Constitution of Bosnia and Herzegovina, respectively), while the Macedo- nian power-sharing system is more informal, given the tradition of binational co ali tions since in de pen dence. A central point of divergence within power- sharing debates relates to who should share power and how inclusive the co ali tion should be. Th e centripetalist approach, advocated by Horowitz, recommends an incentives- based electoral system that will promote vote pooling by parties across the ethnic divide.68 When parties are reliant on votes from other groups, power sharing will be based on multi-ethnic coali tions of moderate parties. He recommends the Alternative Vote as the most appropriate electoral system for vote pooling and to make moderation rewarding for politicians who are dependent on votes from other groups. Horowitz argues that the electoral system can incentivize politicians to cooperate: “If the goal is to produce a moderate, interethnic center, it is necessary to provide ethnically based parties with electoral incentives to take moderate positions on issues of inter- ethnic relations and to form electoral alliances and governing co ali tions with moderate parties of other groups.” Yet others argue that centripetalism, in partic u lar the use of the Alternative Vote, is inappropriate in deeply divided places.69 Th e approach advocated by Horowitz proposes a co ali tion of the “moderate middle,” which will likely exclude hardliners or those parties perceived to be “extreme.” In contrast, consociationalists believe that it may be benefi cial to include hardliners, especially when seeking to convince other sections of their group to move away from violence and because participation in government can make radicals less extreme.70 Th e institutional choice between an inclusive executive and a “moderate middle” coali tion is critiqued by Ian O’Flynn, who argues that inclusion and moderation should be viewed as “mutually reinforcing.”71 Adopting a deliberative democracy approach, O’Flynn calls for a concern for democratic equality rather than stability, the latter being the focus of debates between consociational and incentives-based approaches to power sharing. He counters that there is “no argument to be had” between inclusion and moderation and that the process of executive formation should be constrained to include those elites who indicate their willingness to moderate.72 I bear these arguments in mind in the case studies to come. Th e three cases under investigation here are consociational. Yet an exclusionary, “moderate middle” coali tion was formed in Northern Ireland following the Sunningdale communiqué of 1973 that included unionists, nationalists, and the cross- community Alliance Party. Th us, the Sunningdale 22 Chapter 1 executive was not as strongly consociational as the coali tions formed in 1999, 2007, and 2011: It was limited to the “centre” parties because anti- agreement unionists and republicans were opposed to power sharing. Follow- ing the collapse of the Sunningdale executive, a series of po liti cal initiatives throughout the remainder of the 1970s until the late 1990s failed to secure agreement between the two communities. Th e peace process of the 1990s was fully inclusive and led to the creation of a consociational executive, including Sinn Féin and the anti-Agreement DUP, in 1999. Exclusion of republicans from government would arguably have made it impossible for the Sinn Féin leadership to “sell” the Agreement to its members. Th e executive was subject to a number of stops and starts and was eventually suspended for the fourth and fi nal time in October 2002, precipitating a period of po liti cal stalemate until the restoration of an inclusive co ali tion in 2007. In Bosnia, the DPA set up a complex system of power sharing at both the state and entity levels on the basis of the three “constituent peoples.” Follow- ing the 2000 general elections, the international community helped forge a “moderate” executive called the “Alliance for Change” co ali tion, which governed the Federation and formed the state Council of Ministers.73 Despite some success in pursuing a reform agenda, the co ali tion lacked cohesion and ultimately fell apart due to inter-party rivalry ahead of fresh elections. Th e main three nationalist parties—the (Bosniak) Party of Democratic Action (SDA), the Croatian Democratic Union (HDZ), and the Serb Demo- cratic Party (SDS)—were the electoral winners in the 2002 general election. Nationalist parties were again successful at the 2006 general election. Fol- lowing four months of painstaking inter-party negotiations, a seven-party co ali tion formed with Nikola Špirić (SNSD, Bosnian Serb) as Chairman of the Council of Ministers.74 In Macedonia, the Albanian community has participated in government since 1991. What makes the Macedonian model particularly interesting is that there is no legal requirement for power sharing. Th e ethnic Macedonian and ethnic Albanian parties have, however, formed pre- and post- electoral alliances, and parties and citizens have come to expect the formation of a coali tion including representatives of the two main groups. Th e Albanian party set up by former rebel leader Ali Ahmeti, the Democratic Union for Integration (DUI), was a coali tion partner of the Social Democratic Alli- ance of Macedonia (SDSM) from 2002 to 2006. Th ough the co ali tion had some success on sensitive issues such as language rights, problems persisted in relation to corruption, slow progress on economic and security sector re- Institutional Design and External Actors 23 forms, and plans to increase the inclusion of Albanians in the public sector. Th e inclusion of the DUI was arguably signifi cant for promoting stability following the 2001 confl ict. Following the 2006 election, the DUI was excluded from government and became an opposition party; the party reacted bitterly at not being in the executive and later boycotted parliament for four months. Th e three cases demonstrate the inclusion of “hardliners” in consociation. So long as they abide by the institutional rules, the inclusion of the “extremes”— who may have previously been involved in violence— may be central to the pursuit of stability and prevention of further confl ict. Con- sider the inclusion in government of Sinn Féin in Northern Ireland and the DUI in Macedonia, the parties led by fi gures who participated in their respective armed struggles. An interesting question is whether the inclusion of such parties is conducive to peace or whether it makes compromise diffi - cult, which might otherwise be forthcoming from more moderate parties. An argument could be made for inclusion on the basis of democratic legitimacy. To promote widespread support for the system, post- confl ict institutions ought not to exclude parties with suffi cient electoral support, even if previously linked to armed groups. In addition to holding opposing visions of how inclusive a power- sharing executive should be, the literature includes a debate on the nature of ethnicity and whether ethnic groups should be the “building blocks” of power sharing. Accommodationists see ethnicity as durable and must be recognized for refl ecting the reality of the situation on the ground as people in deeply divided places continue to identify in ethnic terms. Consociationalists, therefore, argue that it is necessary to work within this reality and recognize groups as the basis of power sharing. Th ey claim that po liti cal elites will compromise with representatives of other groups and ethnic divisions may become less salient over time. O’Leary maintains it is perverse to “deny the existence and salience of ethnic identities” and that “the dissolution of (undesirable) collective identities and antagonisms may be likely to occur aft er a period of consociational governance.”75 Although Horowitz recognizes power sharing among groups, he calls for “majoritarian decisions made by a moderate, interethnic center.”76 Integrationists argue that by treating ethnic groups as the “building blocks” of po liti cal arrangements, consociation is, in fact, entrenching ethnic divisions and antagonism. Rupert Taylor’s social transformation approach maintains that democracy in post- confl ict societies should break down ethnic divisions and greater attention should be 24 Chapter 1 paid to cross- cutting cleavages such as class and the creation of a common identity.77 But is it really possible, in the short- to medium- term at least, to move toward a system based on cleavages other than communal divisions? Th e evidence presented in this book suggests that such a transformation is unlikely to take place any time soon. Th ough around a third of people surveyed in Northern Ireland describe themselves as neither unionist nor nationalist, it is clear that ethnicity remains politi cally salient. Indeed, election results demonstrate that the overwhelming majority of the electorate vote for ethnically based parties. In the 2011 Assembly Election, for instance, the four main ethno- national parties won 84.3 percent of the vote. Th e bi- communal Alli- ance Party of Northern Ireland received 7.7 percent, albeit an improvement on 5.7 percent in the 2007 elections and 3.7 percent in 2003. Th e salience of ethnic groups is all too evident in Bosnia. Since the DPA, voters have predominantly voted for the more nationalist parties. Although some “moderate” parties were successful in the October 2010 elections, the pro cess of executive formation at the state and Federation levels took several months, due to elite intransigence.78 Th ere is little potential for a move away from a state based on three constituent peoples toward a common identity. Although Bosniak presidency member Haris Silajdžić called for a new constitution based on majority rule with civic rather than ethnic group rights, his community (the largest group in the Federation and the state as a whole) would benefi t over the other two groups.79 Of additional note is that the Council of Europe’s Venice Commission pointed to discriminatory provisions in the election of the three-member presidency (one Bosniak and one Bosnian Croat from the Federation and one member from the RS). As the rule excludes “Others” as well as Serbs from the Federation and Bosniaks and Croats from RS from being elected to the presidency, the Commission proposed an indirect election of a single president by the Parliamentary As- sembly as preferable to direct elections.80 Th e Venice Commission later recognized, however, that a single instead of collective presidency in Bosnia was not yet po liti cally possible.81 In Macedonia, voters continue to vote for parties they believe will best represent their group interests. For most of the 1990s, the more moderate ethnic Macedonian and ethnic Albanian parties shared power. Aft er the 1998 elections, however, a governing co ali tion was formed between the VMRO- DPMNE and the Demo cratic Party of Albanians, the more nationalist parties. Th e September 2002 election saw victory for the Social Demo- Institutional Design and External Actors 25 cratic Union of Macedonia- Liberal Demo cratic Party co ali tion (SDSM- LDP) and the ethnic Albanian DUI. Th e vote for the Social Demo crats was reportedly a protest vote by ethnic Macedonians against the VMRO for its handling of ethnic tensions and corruption, and DUI leader Ali Ahmeti has been a key leader for the Albanian community since the 2001 confl ict. Th e 2006 election saw the return to government of the VMRO, who chose the Democratic Party of Albanians as its partner, relegating the DUI (the largest Albanian party) to the opposition. Th e party reacted by boycotting parliament, leading to negotiations between party leaders Nikola Gruevski and Ali Ahmeti and agreement on further implementation of the OFA. Al- though the agreement did not institutionalize power sharing to the same extent as the DPA in Bosnia and sought to “enhance the civic nature of the state,” it nevertheless “elevates the status of Albanians as a community by aff ording them rights comparable to those of the Macedonian majority.”82 In an increasingly binational state, therefore, a common Macedonian identity remains remote, and executive design needs to focus on how to incentivize inter- ethnic cooperation among elites who represent the contending groups.

Demonstrating Cooperation

A central question is whether and how power- sharing executives promote inter- ethnic cooperation among ethnic elites. But how do we decide that some arrangements promote cooperation while others do not? Lijphart acknowledges that consociation’s “key concepts have been very hard to defi ne and mea sure precisely.”83 Indeed, scholars have critiqued the “thorny issue of how to mea sure the degree of elite cooperation,” understood as the essence of consociationalism.84 McGarry and O’Leary write that “what makes consociations feasible and work is joint consent across the signifi cant communities—with the emphasis on ‘jointness.’ ”85 O’Leary suggests that power sharing means “coordinated jointness in shared decision making.”86 My primary objective is to uncover what factors are likely to produce cooperation, indicated by joint decision making. Inter-ethnic cooperation within a power-sharing executive can be demonstrated with evidence of ministers working together. In the following case study chapters, I track the extent of cooperation evidenced by joint decision making on policy issues, including the implementation of the respective peace agreements. As the start of this 26 Chapter 1 chapter shows, the new executive will likely be faced with the challenge of how to manage sensitive issues relating to culture, symbols, and identity. Th ese oft en ad hoc issues are considered problems of “cultural contestation.”87 And given that the agreements sought to end violence and promised a future peaceful environment, the management of security issues (particularly police reform) is considered a useful indicator of cooperation. Overall, I seek to investigate whether cooperation has been attainable or in short supply.

Executive Formation: Sequential Portfolio Allocation or Negotiation?

Executive formation rules can arguably deliver a process that will help foster improved relationships and cooperation among ethnic parties. Th ese rules vary in terms of whether groups are guaranteed positions in government. Lijphart makes a distinction between predetermination and self- determination. Self-determination “gives various rights to groups within the existing state . . . and it allows these groups to manifest themselves instead of deciding in advance on the identity of the groups.” Pre- determination “means that the groups that are to share power are identifi ed in advance.”88 Lijphart has a clear preference for self-determination in power sharing for several reasons. Among them, he notes the fl exibility of self-determination based on numbers of people identifying with specifi c groups: “It is naturally and continually self-adjusting.” 89 Th e distinction between predetermination and self-determination corresponds to that made by McGarry and O’Leary between corporate and liberal consociation. Th ey write that a “corporate or predetermined consociation accommodates groups according to ascriptive criteria, such as ethnicity or religion, on the assumption that group identities are fi xed and that groups are both internally homogeneous and externally bounded.” By contrast, a liberal or self- determined consociation “rewards whatever salient politi cal identities emerge in democratic elections, whether these are based on ethnic or religious groups, or on subgroup or transgroup identities.” 90 Notably, the three places under investigation display variation, including a mix of corporate and liberal rules for executive formation. In the fi rst Northern Ireland power-sharing executive (1973–74), three parties agreed to share power and to allocate portfolios via inter-party bargaining, overseen by the British government’s Secretary of State for Northern Ire- land, William Whitelaw. Based on party strength in the assembly following Institutional Design and External Actors 27 the 1973 elections, this was clearly a liberal form of consociational power sharing. Th e second power-sharing administration following the 1998 Agreement established a novel form of government formation. McGarry, O’Leary, and Simeon suggest the peace agreement “combines both corporate and liberal elements” because the fi rst minister and deputy fi rst minister are elected via cross- community support of national and unionist members, while the other ten ministers are selected according to the d’Hondt method “that is liberal or ‘diff erence- blind.’ It operates according to the strength of repre sen ta tion won by parties in the Assembly, not their national identity.” 91 D’Hondt was used again to form the executives that took offi ce in 2007 and 2011. Th e Northern Ireland (St. Andrews Agreement) Act 2006 revised the joint election of the two premiers to a straightforward nomination, rather than a cross-community vote. A procedural change meant that the fi rst minister would come from the largest party in the assembly and the deputy fi rst minister would be nominated by the second- largest party.92 Rick Wilford suggests the change meant that “future assembly elections would become de facto referendums on who would take the role of fi rst minister.” 93 Th e Northern Ireland power-sharing system is largely liberal as executive formation is subject to the parties’ respective strengths in the assembly. Th ere are, however, more corporate elements in the assembly voting procedures for key decisions (the “weighted majority” rule and “parallel consent”), which require support from members designated as “nationalist” and “unionist.” 94 Government formation in Bosnia requires the inclusion of representatives from the “constituent peoples.” Th e state three- member, directly elected presidency includes one Bosniak and one Croat from the Federation and one Serb from the RS. Th e presidency nominates the chair of the council of ministers who, in turn, nominates the ministers and deputy ministers. Th e chair and ministers are subject to a legitimizing vote in the House of Representatives.95 Following a landmark decision by the Bosnian constitutional court in July 2000, the constitutions of the two entities were revised to ensure repre sen ta tion of the three “constituent peoples” as well as some potential inclusion from the “others.” 96 Th e Federation’s government must include eight Bosniak, fi ve Croat, and three Serb ministers, and the prime minister may appoint an “other” in place of one of the Bosniaks. Th e president of the Federation (in agreement with the vice-presidents) appoints the government, subject to majority support in the House of Representatives. In the RS, the executive includes eight Serb, fi ve Bosniak, and three Croat 28 Chapter 1 ministers, and the prime minister may appoint an “other” in place of one of the Serbs. Th e Macedonian model is diff erent again. First, the president tasks the largest party to form a government that will require a majority vote of all representatives, thus ensuring the approval of the Albanian community without an explicit “cross- community” rule. Th e largest (Macedonian) party then ne- gotiates with Albanian parties and other smaller parties to form a government that will secure a majority vote. Th us, the Macedonian model allows more freedom regarding executive membership, subject to securing majority support in parliament. In contrast to the Bosnian model, government formation is not on the basis of predetermined groups or “constituent people.” Th is is clearly a much more liberal method of executive formation. But what impact do these various rules have on executive operation, and can we say some rules better promote inter- ethnic accommodation? For instance, what eff ect does the d’Hondt sequential portfolio allocation method have on power sharing? Is it appropriate to guarantee parties seats based on their electoral strength rather than require a pro cess of inter- party bargaining? Or, as McGarry and O’Leary argue, is d’Hondt an appropriate mechanism for post-confl ict societies because it is proportional, sequential, nonexclusionary, and incentivizes parties to take their executive seats?97 Does the Bosnian system of predesignation (strong corporate rules) have an eff ect on the system? Does it limit the potential of ethnic elites to work together? Conversely, does the more informal, fl exible system present in Macedonia based on post-electoral negotiations (without predesignation) produce improved relationships and more frequent government turnover? Or is there a need to formally institutionalize executive power sharing in Macedonia to guarantee proportionate inclusion of the Albanian community and smaller minorities?

Executive Decision Making: Veto Rights

How power-sharing executives make decisions is also pertinent to the potential for inter- ethnic cooperation. Are ministers in a power- sharing executive obliged to ensure consensus within the co ali tion on policy issues? Must ministers act in concert with their co ali tion colleagues, or do they have considerable room for maneuver on policy issues? Can a party with a greater number of ministerial portfolios determine the direction of policy Institutional Design and External Actors 29 and impose decisions? Or can ministers representing minority groups prevent this from happening? What impact do veto rules have on inter-ethnic cooperation? In Northern Ireland, the GFA encourages consensual decision making within the executive. Th e executive is charged to “provide a forum for the discussion of, and agreement on, issues which cut across the responsibilities of two or more Ministers, for prioritizing executive and legislative proposals and for recommending a common position where necessary (e.g. in dealing with external relationships).” 98 Th e executive’s responsibilities include agreeing on an annual program and budget, subject to cross-community approval in the Assembly. Under the Pledge of Offi ce in the 1998 Agreement, ministers are charged “to participate with colleagues in the preparation of a programme for government” and “to operate within the framework of that programme when agreed within the Executive Committee and endorsed by the Assem- bly.” As with the other two cases in the study, there are veto procedures that check majoritarian rule in executive decision making. In the assembly, thirty MLAs can petition a decision to be taken on a cross-community vote or refer a ministerial decision back to the executive for review. Within the executive, any three ministers can trigger a vote to be taken on a cross-community basis when a vote is required. At the state level in Bosnia, the three-member presidency must adopt all decisions by consensus. A member may, however, declare a decision to be harmful to a “vital interest” the entity by which he or she was elected; the decision is then referred to the legislature of that entity. Two- thirds of the relevant grouping in the entity can veto the presidency decision. Legislation imposed by the OHR in December 2002 made important changes to the process of decision making in the Council of Ministers (CoM). Th e Law stated that legislation to be voted on in parliament will need to fi rst secure majority support within the executive; on all other decisions the CoM needs to secure consensus. If consensus is not reached, the chair of the CoM con- sults with the dissenting member(s) to resolve the problem. If consensus is still not achieved within seven days, a majority vote is required within the CoM, including support of at least two members of each constituent people. In the Federation, ministers are responsible for “executing Federation Gov- ernment policies and enforcing Federation Government laws within the scope of his Ministry or assigned by the Prime Minister” and for “proposing and making recommendations concerning legislation within the scope of his Ministry or as assigned by the Prime Minister.” 99 In more general terms, 30 Chapter 1 the RS constitution charges the government to “direct and coordinate the work of the ministries” and to “supervise the work of the ministries.”100 In both entities there is a parliamentary procedure for the protection of the “vital national interests” of the three constituent peoples. Outside of the issues listed in the entities’ respective constitutions, two-thirds of a group’s delegates can trigger the veto procedure. In Macedonia, the constitution charges the government to “lay down principles on the internal or ga ni za tion and work of the Ministries and other administrative bodies, directing and supervising their work.”101 In the assembly, a veto mechanism exists for delegates of communities other than the majority to block decisions. Th e concurrent majority voting procedure applies to Article 69 of the constitution, which stipulates that with regard to legislation aff ecting culture, use of language, education, personal identity, and symbols, approval requires a majority of delegates present and voting, including a majority of delegates not from the majority community. Elsewhere I compare the use of veto powers and ask whether veto rights are an appropriate mechanism for minority group protection or whether they lead to executive gridlock.102 I suggest a conceptual framework to help institutional designers in power-sharing systems think about veto players, veto points, and veto procedures. I argue that Bosnia has too many veto rules and suggest limiting the areas and procedures in which parties can trigger a veto. Macedonia is considerably more fl exible than Bosnia, and Northern Ireland is somewhere in between the two. Th e Macedonian case confi rms the importance of veto rules for the minority community in power- sharing systems. Discussed in Chapter 8, Albanian elites refer to the concurrent majority voting procedure in parliament as a central feature of the peace agreement and rely on its application to constrain the Macedonian party in power. Although the focus here is on the potential of institutional rules for incentivizing inter-ethnic cooperation, some caution is required. Th e book does not consider institutional design to be a panacea for inter- ethnic confl ict. Th ere are limits to the potential of institutional design to foster inter- ethnic cooperation in deeply divided places. Institutions are not the whole story; context matters, too. Th ere may be considerable obstacles to compromise within the executive arising from the legacy of the confl ict, wider political problems (security issues, refugee returns), a sense of suspicion and mistrust, and diffi cult interpersonal relations. Florian Bieber notes one such obstacle: “Th e nature of a grand co ali tion poses a par tic u lar challenge that Institutional Design and External Actors 31 can be met only with diffi culty by institutional designers because it requires politi cal actors who are willing to cooperate.”103 Certainly, the presence of po liti cal elites “willing” to cooperate is the ideal scenario in a power- sharing executive. Th e hard reality of the situation, however, is that elites’ willingness to cooperate may not be forthcoming. To the view that the potential of institutional design may be limited I add that the executive is just one (though arguably the central) institution in a set of po liti cal institutions. But even when inter- ethnic cooperation is forthcoming within the coali tion, it might be hampered by the dysfunctionality of other institutions such as the electoral system, the legislature, and policing structures. And just because some institutional rules work in one situation does not mean they will work in another. I pay due attention to the importance of context and the partic u lar conditions in each case. Th at said, an emphasis on comparison, rigorous empirical analysis, and contribution to theory will go some way to identifying why power-sharing executives adopt par tic u lar rules and the impact of executive design on inter- ethnic cooperation. As Andrew Reynolds notes, “formal rules oft en provide the main incentives for divided groups to be conciliatory. Th e behavior of elites is paramount, but without power-sharing structures, accommodatory signals may never be encouraged.”104

External Actors: Incentivizing Cooperation?

Th e involvement of international actors in the implementation of peace agreements is an intriguing, and arguably central, aspect of power-sharing democracy. External actors’ state-building eff orts have involved the implementation of power-sharing settlements in Bosnia and Herzegovina, Iraq, Lebanon, Northern Ireland, and Macedonia and proposals for power sharing in Burundi, Cyprus, the Demo cratic Republic of Congo, Liberia, and Nepal. Rupert Taylor notes that we are embarking on a “new wave” of power-sharing theory.105 Compared with the classic cases of consociation, power-sharing arrangements have recently formed the basis of institutional arrangements in deeply divided places and have involved external actors in their implementation. Michael Kerr argues that external actors have the “ability to provide the motivation and incentives for internal elites to engage in consociation, to create the conditions for the implementation of power- sharing arrangements, and to engage in coercive consociational statebuilding 32 Chapter 1 itself.”106 Bieber discusses the role of external actors in forming and maintaining power sharing in the states formed following the disintegration of the Yugo slav federation. Assessing their contribution, he suggests that a principal problem of external intervention is the imposition of policy against the wishes of at least one of the contending groups.107 McGarry et al. note that “the international community, which usually preaches integration, has been prepared to back accommodation where that has been demanded, if only, and unfortunately, aft er rebellion has threatened order.”108 In the three cases under investigation, external actors have (on the whole) sought to uphold the respective peace deals, encourage peace implementation and wider policy reform, and facilitate inter-ethnic cooperation. Discussed in Chapter 2, however, the British and Irish governments failed to maintain Northern Ireland’s 1974 power- sharing executive in the face of considerable pressure. Yet does the involvement of external actors matter for the operation of power- sharing executives? To what extent are they involved in governance? Ultimately, does their involvement help or hinder inter-ethnic cooperation? I draw from the IR and EU studies literature on compliance, which helps explain the interaction among external actors (more oft en understood to be international organizations) and internal actors over policy reform. Th is literature has pitted rationalist against constructivist approaches to compliance. Although the rationalist approach focuses on coercion, cost-benefi t calculations, and material incentives, the constructivist approach highlights socialization and social norms. Importantly, there has been some eff ort to build bridges between these two approaches in order to arrive at a synthetic approach “encompassing both instrumental choice and social learning.”109 Research has focused on socialization, defi ned as “the pro cess of inducting actors into the norms and rules of a given community.”110 Th e outcome of socialization is understood to be “sustained compliance based on the internalization of these norms.” Th us, socialization implies domestic actors shift from a logic of consequences to a logic of appropriateness. Th is study seeks to determine whether politi cal parties representing the contending groups in deeply divided places have been subject to socialization eff orts on the part of external actors and have internalized power sharing as the most appropriate post-confl ict institutional arrangement. Suggested mechanisms of socialization include role playing, normative suasion, and strategic calculation.111 Ar- guably, strategic calculation is most applicable here, whereby actors “carefully calculate and seek to maximize given interests, adapting their behaviour to Institutional Design and External Actors 33 the norms and rules favored by the international community.”112 As Frank Schimmelfennig writes, socialization “works through reinforcement.” In this way, external actors “reward norm- conforming behavior and punish norm- violating behavior; target states conform with the norms and rules in order to avoid punishment and gain rewards.”113 In Northern Ireland, the British and Irish governments devoted an enormous amount of time to secure the agreement in April 1998 and to restore power sharing in 2007. Th e British Secretary of State has played an infl uential and, at times, controversial role. Successive secretaries of state opted to suspend the institutions under the Suspension Act (2000) in the face of diffi culties over IRA decommissioning and inter-party mistrust.114 Th e suspension legislation was criticized by the Irish government as well as by the SDLP and Sinn Féin as a breach of the British- Irish treaty. Th e British and Irish governments were, however, instrumental in pushing the parties toward restored power sharing following suspension in October 2002. At inter-party talks convened in Scotland in 2006, Prime Minister Tony Blair and Taoise- ach Bertie Ahern published the St. Andrews Agreement with a deadline for power sharing of 26 March 2007. To incentivize a deal between Sinn Féin and the DUP, Secretary of State Peter Hain warned of greater cooperation between London and Dublin (the two governments’ “Plan B”) should the parties fail to agree. Supported by Irish Foreign Minister Dermot Ahern, Hain warned of “devolution or dissolution” of the assembly. Th e parties ultimately accepted the St. Andrews provisions and agreed to form a power- sharing executive, which took eff ect from 8 May 2007. Th e British Secretary of State retained responsibility for policing and justice, with the expectation (as set out in the Act) that the new executive would agree on the transfer of policing and justice by May 2008. Th is deadline passed, however, as the DUP believed there was insuffi cient public confi dence in Sinn Féin’s support for the rule of law. Eff orts by the two governments eventually led to agreement on devolution of policing and justice in 2010. A number of international actors have been involved in Bosnia since the outbreak of war in 1992. Th e key external body, the OHR, has monitored the implementation of the DPA, encouraged the parties to comply, and facilitated discussions on policy reform. Yet the role of the High Representative (HR) in Bosnia has been likened to that of a “colonial governor.”115 At the Peace Implementation Council (PIC) meeting in Bonn in 1997, the OHR’s powers were strengthened to remove from offi ce offi cials who “violate legal commitments and the Dayton Peace Agreement, and to impose laws if he 34 Chapter 1 sees fi t should Bosnia and Herzegovina’s legislative bodies fail to do so.”116 Some commentators argue that the OHR’s frequent use of these powers has harmed the country’s transition. Th e Eu ro pe an Stability Initiative argues that because the HR has legislative power to impose decisions, this “has led many international offi cials to believe that they do not have to do the hard work of identifying and building support for their policy initiatives.”117 Moreover, if domestic actors believe that the HR will impose decisions, they are less incentivized to compromise and implement reform. As outlined in Chapters 5 and 6, successive High Representatives proposed policy reform to improve the functionality of power sharing, a trend that was increasingly opposed by Bosnian Serbs as an infringement on their rights and contravention of the integrity of the RS under the DPA. As the OHR moved toward closure, the EU became the principal external actor in Bosnia, working to ensure the implementation of the Stabilization and Association Agreement and to facilitate the country’s preparedness for accession. Chapter 6 explores the diffi culties in the transition from OHR to EU as the principal external actor and explains why EU conditionality has been ineff ective in bringing about policy reform. When violence erupted in Macedonia in February 2001, the international community played a crucial role in preventing confl ict from escalating into civil war and in brokering a ceasefi re. According to Robert Hislope, “what kept Macedonia from teetering off the edge was the persis tent activism from the international community” as the EU, NATO, and the U.S. persuaded both sides to arrive at a settlement.118 Th e EU delegation in Macedonia has since sought to ensure the implementation of the Ohrid Agreement. At times of politi cal stalemate, the EU has sought to persuade the parties to cooperate in the interests of the accession pro cess. For instance, diffi culties arose in the aft ermath of the 2006 parliamentary election when the largest ethnic Mace- donian party, VMRO-DPMNE, opted to form a coali tion with the ethnic Al- banian party, the Democratic Party of Albanians, even though the DUI had won a majority of that community’s support. Relegated to the opposition, the DUI later boycotted parliament in protest. Th e EU and the U.S. were instrumental in persuading the parties to negotiate, leading to the “May Agree- ment” in 2007, when the VMRO and DUI party leaders agreed on further implementation of the Ohrid Agreement. Th ese external actors are oft en criticized by domestic elites for their lack of accountability. Th e OHR in Bosnia is an essentially undemo cratic institution; the Peace Implementation Council selects the offi ce holder, who is Institutional Design and External Actors 35 unaccountable to the electorate. In Northern Ireland, nationalists and republicans criticized the Secretary of State for suspending the institutions, and his offi ce retained responsibility for policing and justice until 2010. Re- publicans viewed suspension as an attempt to “save” Ulster Unionist leader David Trimble, who faced increasing dissent within his own party over the implementation of the GFA and the slow decommissioning of paramilitary weapons. Historically, republicans viewed the British government as a cause of the confl ict and pro- unionist; however, Sinn Féin’s Martin McGuinness admitted that the peace process would have collapsed without the contribution of Blair and Ahern.119 Power sharing in deeply divided places wracked by violent confl ict is a diffi cult undertaking. Debates continue on whether power sharing is an appropriate institutional option for promoting peace and democracy. Th is book advances reasons for an optimistic outlook without denying the challenges. Moreover, it is reasonable to expect that power sharing will be a likely option considered by policymakers engaged in peace mediation and post- confl ict statebuilding. It is intended that the book’s refl ections on the three places under investigation, through rich description and explanation of power- sharing practice, will provide useful insights for such endeavors. This page intentionally left blank CHAPTER 2

The Sunningdale Executive: Lessons from Failed Power Sharing

I begin the analysis with a power-sharing failure. As the overall aim is to determine what works in power- sharing practice, this may seem a surprising point of departure. But perhaps there is no better way to begin my case analysis than explain why power sharing will not work. Indeed, the Sun- ningdale executive’s failure makes it potentially instructive. In assessing the creation, operation, and collapse of the Sunningdale executive, I consider several questions: To what extent was the executive doomed because it included the “moderate middle” facing pressure from intra- communal competition, some of which was engaged in violence? What role did the British and Irish governments play in its creation and demise? How eff ectively did the executive function when in operation? Does it off er any lessons in institutional rules for cross- community executive decision making? Arguably, Northern Ireland’s fi rst experience of power sharing (in terms of reaching an agreement, forming an executive, and ultimately its downfall) has much to off er us in trying to understand the dynamics of power sharing as a feasible and appropriate institutional mechanism in deeply divided places. Th e 1973 Sunningdale communiqué and executive formed on 1 January 1974 are subjects of existing research on Northern Ireland.1 Moreover, recent work explores previously unavailable archives and provides a detailed politi cal history of the Sunningdale negotiations and agreement, central elements of Northern Ireland’s so-called “lost peace process.” 2 Drawing on this existing research, I seek to provide an analysis of what the Sunningdale case tells us compared with contemporary cases of power sharing (Northern Ire- land since the Good Friday Agreement, Bosnia since the Dayton Peace Agreement, and Macedonia since the Ohrid Agreement). In partic u lar, this 40 Chapter 2 comparison focuses on investigating the impact of two key variables for the book: institutional rules (in this case, the focus on a “moderate middle” executive) and the role of external actors. I also outline how an imbalanced deal (perceived or actual) will have a negative impact on the power- sharing system. Rather than retell the full story of Sunningdale’s failure, I focus on the main lessons from this case and consider what insights they provide for the case studies to come.

A “Moderate Middle” Executive

Th e Sunningdale executive is an interesting instance of moderate middle power sharing. Th e peace negotiations leading up to the agreement focused on three moderate internal parties: the Offi cial Unionists, the SDLP, and the Alliance Party. Following the prorogation of the Stormont parliament in March 1972 against a backdrop of increasing violence, the British Conserva- tive government, led by Edward Heath, sought to restore a devolved administration following agreement by the moderate parties while excluding the more “extreme” politi cal forces.3 Th e ongoing violence stressed upon the British government the need to fi nd a politi cal settlement based on the minority community’s inclusion as well as a military solution to the confl ict. Th us, Heath and William Whitelaw, the fi rst Secretary of State for Northern Ireland, explored options for agreement among the nationalist and unionist parties on the way forward and, ultimately, a po liti cal settlement. Th e details of the politi cal process leading to the Sunningdale agreement are described in detail elsewhere.4 Of par tic u lar relevance here is the “rolling” process of proposals draft ed by the British government, which focused on reaching agreement among the moderate parties. Th is pro cess was established with the arrival of Whitelaw at Stormont Castle in Belfast with a team of Whitehall offi cials to work alongside se nior members of the Northern Ireland Civil Service. Th en–Northern Ireland Offi ce civil servant Kenneth Bloomfi eld recalls his fi rst impressions of working with the Secretary of State: “Amongst the rather colorless products of the modern po liti cal ma- chine Whitelaw was an exotic. Behind a façade of booming aff ability there lurked one of the shrewdest minds in British politics.” 5 In addition to White- law’s immediate team, se nior civil servants formed a “Future Policy Group,” with the task of drawing up options for a devolved administration.6 As early as July 1972, the group was ready to consider a number of options for a re- Sunningdale Executive 41 turn to devolution. Th e group considered the view that “entrenchment of the rights of any par tic u lar interest or interests to be represented in a new executive would be undesirable as it would tend to perpetuate existing divisions and militate against repre sen ta tion of the moderate ‘middle ground.’ ”7 Th e aim to provide po liti cal space for the “moderate” or “middle ground” was very much evident in the subsequent green and white papers setting out an agenda for the restoration of devolved government. Instigating a series of steps that ultimately led to a peace deal, the British government fi rst convened inter- party discussions at Darlington, En gland, in September 1972 where only three local parties (Offi cial Unionists, the Al- liance Party, and the Northern Ireland Labour Party) agreed to attend. Un- surprisingly, Whitelaw sought to create a visibly consultative process that would produce an inter- party agreement, rather than impose a settlement. But with little progress at Darlington, the Secretary of State published the government’s ideas in a green paper, Th e Future of Northern Ireland: A Paper for Discussion.8 Th e objective of the paper was to bring about a pro cess of inter-party talks that would achieve “the widest possible measure of agreement as to how Northern Ireland should in future be governed and how in par tic u lar the minority as well as the majority may be assured of an active, permanent and guaranteed role in its life and public aff airs.” A section of the paper considered possible institutional arrangements for a Northern Ireland assembly and executive. On the executive, it stated, “Here the crucial question is whether, in addition to any heightening of their infl uence, it is desirable and possible to secure the participation of the Northern Ireland minority in the actual exercise of executive powers.” 9 Inclusion of nationalists was recognized as a sine qua non of reformed po liti cal institutions. With regard to guaranteeing more inclusive represen ta tion and participation, the green paper stated that it may be “dangerous to make complex arrangements which can be manipulated to produce deadlock and frustration.” Despite this reticence, the paper was somewhat creative in acknowledging that other divided places employed institutional rules to ensure the participation of various groups in government. Th e paper accepted that “a number of these countries have had stable and successful coali tion governments over many years; and that there is no hope of binding the minority to the support of new politi cal arrangements in Northern Ireland unless they are admitted to active participation in any new structures.”10 Th ere was some acknow ledg ment, then, that the engagement of the nationalist community in politi cal life was imperative. In what the paper described as a 42 Chapter 2

“broadly based administration” rather than power-sharing government, various options were considered, with a warning that “any major politi cal element could choose, for its own purposes, to bring the system to a standstill.” Th us, the British government appeared to suggest some kind of compromise. Acknowledging nationalist objectives, the discussion paper also noted that future arrangements would have to take account of Northern Ireland’s relationship with the Republic of Ireland. Two guiding principles of British government policy were born: power sharing between the two communities and an agreed all- Ireland dimension. Th e 1972 paper for discussion formed the basis of the British government’s white paper, Northern Ireland Constitutional Proposals, published in March 1973. Th e British government concluded that “What ever their views on sovereignty and citizenship, most of those consulted (although not all) have fa- voured the restoration of some kind of devolved institutions of government in Northern Ireland.”11 On the issue of minority inclusion, the paper suggested this would, at a minimum, mean providing for its “eff ective voice” and “real infl uence.” Th e paper noted, however, “strong arguments that the objective of real participation should be achieved by giving minority interests a share in the exercise of executive power if this can be achieved by means which are not unduly complex or artifi cial, and which do not represent an obstacle to eff ective government.”12 Th e British government clearly preferred fl exible arrangements for inclusion rather than devising and establishing more formal rules. Th e document stated that ministers’ formal appointment “will be eff ected by the Secretary of State in accordance with any agreed understanding as to the formation of the Executive.” Th us, the makeup of the power-sharing executive would be subject to negotiations among parties following an election. Th e white paper also highlighted the potential “Irish dimension,” some form of institutionalized cooperation on an all-Ireland basis. Following the white paper, the Northern Ireland Constitution Act of 1973 required the Secretary of State to appoint ministers and ensure that “a Northern Ireland Executive can be formed, which, having regard to the support it commands in the Assembly and to the electorate on which that support is based, is likely to be widely accepted throughout the community.”13 Seeking to shore up support for the moderate parties, Whitelaw’s approach was to intensify divisions within unionism. Presumably this approach sought to deliver Faulkner as the preferred unionist leader with whom the two governments and other parties could do business and isolate more hard-line elements who threatened to hinder politi cal progress. Ulti- Sunningdale Executive 43 mately, this meant a split within the Offi cial Unionists over the 1973 white paper between “pledged” and “unpledged” unionists. Faulkner and his pledged unionist colleagues, as well as the SDLP and Alliance, were invited for talks at Stormont Castle in October. Th e discussions focused on fi nding a form of power sharing agreeable to these three parties. Setting a devolution date of 1 January 1974, Whitelaw maintained there should be a unionist majority on the executive and presented a formula whereby Faulkner’s pledged unionists would have six seats, the SDLP four seats, and the Alliance Party would have one member with voting rights and one member without. As Michael Kerr notes, however, problems arose because Faulkner did not have majority support in the assembly, and he had less support than the SDLP and Alliance Party numbers combined.14 Th e British government was aware of Faulkner’s precarious position as unionist leader and his diffi culty in bringing the wider unionist population with him. Reportedly, Whitelaw advised Heath that Faulkner needed “to be kept afl oat if an executive is to be formed.”15 But, arguably, not enough was done to bolster Faulkner’s position, leaving him vulnerable to criticism both inside and outside his party. Interestingly, Whitelaw also acknowledged the diffi culty for the SDLP that a unionist majority on the executive would bring. To assuage the party’s diffi culty with a unionist majority, the Secretary of State off ered nationalists an incentive by committing to the establishment of an all- Ireland institution. Th e details of the Council of Ireland were ironed out at a conference held in Sunningdale, En gland, 6– 9 December 1973. Th e delegations included the three main internal parties, British Prime Minister Edward Heath, and Irish Taoiseach Liam Cosgrave. With the pursuit of a settlement focused on Faulkner’s “pledged” unionists, the SDLP, and the Alliance Party, the British government chose to exclude the unionist opposition represented by Ian Paisley’s Democratic Unionist Party and William Craig’s Vanguard Unionist Party, a breakaway party formed aft er Craig failed to persuade the Ulster Unionist Council to reject the white paper. Kerr notes that Whitelaw’s successor, Secretary of State Francis Pym, invited both party leaders to Sun- ningdale “at the eleventh hour” to present their views at the opening plenary session.16 Paisley strongly rejected Pym’s invitation and threatened serious consequences for their exclusion. In the event, the Sunningdale communi- qué was agreed between the two governments and the three Northern Ire- land delegations. Following declarations on the status of Northern Ireland, the agreement stated that a Council of Ireland would be set up. Th e Council was to “comprise a Council of Ministers with executive and harmonising 44 Chapter 2 functions and a consultative role, and a Consultative Assembly with advisory and review functions.”17 Th e Council’s formal executive functions were yet to be agreed following studies on areas of common interest. Th e two governments also made commitments on security, including the possible creation of a common law enforcement area and the possible devolution of policing powers from Westminster to Stormont once security issues were resolved. Devolution day took place on 1 January 1974, and the power- sharing executive took offi ce on the basis of the formula created by Whitelaw. De- spite the challenges that lay ahead, the executive’s mood was optimistic. At the executive’s preparatory meeting on Monday 31 January 1973, the governing parties heralded 1974 as a “year of reconciliation.” Perhaps the parties shared a sense of purpose, of history making, and enthusiasm for the task ahead. Faulkner wrote that it was “hard to believe . . . how far we had come. Th ere was a feeling of comradeship and trust between those of us who had been through hundreds of hours of negotiations, whether on the same or diff erent sides of the negotiating table, and a sense of almost moral purpose.”18 Th e devolution of power from Westminster to Stormont was marked with a sense of occasion. Faulkner wrote about the “slightly absurd little ceremony” at Stormont Castle on 31 December, whereby the members of the new executive were given “warrants of appointment” by the Secretary of State Francis Pym before being sworn in by Lord Chief Justice Sir Robert Lowry.19 Lowry even wore his ermine and scarlet judicial robes as a way of indicating “the importance of the occasion.”20 Th e swearing in of the new executive was perhaps a useful exercise in demonstrating to the parties’ followers and the wider electorate that the governing parties were serious about sharing power and committed to governing the country together in a spirit of reconciliation and accommodation. Th e ceremony may, however, have added to Faulkner’s diffi culties. Th e under- pressure unionist leader thought the ceremony’s “school prize-giving atmosphere” gave “the damaging impression . . . to the public that the Executive was the child of the Secretary of State.”21 He suggested that the ceremony strengthened the loyalist argument “that the new Constitution and the power- sharing Executive had been forced on Northern Ireland . . . against the wishes of its population by British politicians and that the Executive therefore lacked any demo cratic status.” Although Faulkner was content that the executive “was a freely agreed coali tion of elected politicians, supported by a strong majority in the Assembly,” he later claimed that the ceremony made “the loyalist misrepresen ta tion that little bit easier.”22 Sunningdale Executive 45

Almost immediately the executive faced problems. On 4 January, the Ulster Unionist Council rejected the Council of Ireland proposals by 427 votes to 374. Faulkner then resigned as party leader on 7 January, continuing to serve as a somewhat belittled chief executive. According to Gillespie, “Faulkner’s resignation as leader of the Ulster Unionists ended the moral legitimacy of the Sunningdale deal and killed Sunningdale as an eff ective po liti cal package.”23 If this reading is correct, the executive was doomed from the beginning. It is certainly reasonable to argue that Faulkner’s increasingly precarious position, in no longer being unionist party leader, added further strain to the fl edgling co ali tion. Gillespie notes the signifi cance of Faulkner’s defeat for the power- sharing executive: “In practical terms it created a dangerous situation in which the views of the majority of unionists, let alone loyalists, were unrepresented in a supposedly ‘widely accepted’ administration. In this situation it was always likely that the agreement would either collapse through lack of support or be brought down by external pressure.”24 Discussed below, the executive was deprived of suffi cient support by the two governments, particularly in the aft ermath of the February 1974 Westminster election, when Heath suff ered defeat and was replaced by Har- old Wilson’s Labour minority government. Th e executive’s already shaky foundations suff ered a blow with the election results. Eleven of Northern Ireland’s twelve Westminster seats were won by the United Ulster Unionist Council, making diffi culties for Faulkner even more acute. In the aft ermath of this disastrous result for the power-sharing administration, it became even more diffi cult to say the executive was “widely accepted” throughout the community. Unionist misgivings toward the Sunningdale package were in stark relief. In a meeting with civil servants on 4 March, Faulkner reportedly confi ded that some of the UUP executive ministers were “shaky” at the election results and that “Everyone was shattered by the extent of the feeling against a Council of Ireland.”25 Th ough the chief executive acknowledged that the results made his party’s backbenchers “increasingly ner vous about implementing the Council of Ireland proposals,” he maintained the election was “a serious but not necessarily fatal blow” to the power-sharing government.26 Perhaps unsurprisingly, the executive parties campaigned on very diff erent politi cal agendas in the run up to the Westminster election. Argu- ably, however, this highlighted their respective po liti cal diffi culties and helped prevent the executive from forging a common sense of purpose. Despite these setbacks, the executive seemed determined to get down to the business of taking policy decisions. In terms of institutional rules for 46 Chapter 2 decision making, the Sunningdale executive is notable for power sharing in that decision making was to operate according to the principle of collective responsibility. At the executive meeting on 15 January, ministers approved a memo draft ed by Faulkner that executive decisions would be governed by the principle.27 Th is meant that all ministers and governing parties were obliged to support executive positions on policy. Faulkner’s memo stated as follows:

It has been agreed that the Northern Ireland Executive and Admin- istration should conduct the business of government in accordance with the principle of collective responsibility. It is self-evident that the standing and cohesiveness of an Administration could not long survive if its members were in open public disagreement with each other on public issues of any importance. In general, therefore, collective responsibility applies to action and comment over the whole fi eld of Government policy. In par tic u lar, decisions of the Executive bind all members of the Administration.28

Set out in the memo, if a minister opposes an executive decision, he may request the executive to discuss it again or resign, “but he may not remain a member of the Executive and publicly criticise it. Nor may a member of the Executive indicate publicly that a par tic u lar decision was taken against his advice. Th e position within the Executive must remain confi dential.”29 Th e memo did, however, recognize that ministers had autonomy to take day- to- day decisions: “Under their existing powers, Heads of Departments must take a series of decisions day in and day out; it would be impossible to obtain Executive sanctions for all of these, but unacceptable for them to be publicly criticised by other members of the Administration.” Much, then, remained at the discretion of individual ministers: “A balance must be struck by each member between seeking the approval of his colleagues for trivial proposals and committing them without consultation to a course with diffi cult po litical or other repercussions.” Th e principle of collective responsibility has been regarded as part of the broader Westminster convention of ministerial accountability, argued to be “one of the cornerstones of British constitutional practice.”30 Ministerial accountability is premised on two related conventions: (1) collective cabinet responsibility whereby ministers are collectively responsible for government policies and (2) individual ministerial responsibility whereby each minister Sunningdale Executive 47 is responsible to parliament for the work of his or her department. It is perhaps unsurprising that Faulkner was guided by Westminster conventions. He recalled that “many people believed that a government such as ours could not operate on a basis of collective responsibility.”31 For Faulkner, decision making on the basis of collective responsibility was about creating a “normal” government along Westminster lines. Yet arguably it was more than that. It was rational for Faulkner to want the executive to operate via collective responsibility because his party had a majority within the coalition. He controlled the majority within the executive, and the three governing parties were obliged to support executive policy. Yet Faulkner said, “Th ere was never a vote. Our decisions were taken by consensus, as we all knew we had to work together for the Executive to survive.”32 On 22 May, however, the SDLP, “with the deepest distress and regret,” withdrew from collective responsibility aft er a vote to approve an executive statement on the Sunningdale communiqué.33 Aft er some backroom discussions, the Deputy Chief Executive Gerry Fitt further reported that his party “were in a position to accept the proposal and that it could be regarded as carried unanimously.” Th e executive’s brief lifespan of just fi ve months did not allow suffi cient time to test the stability and eff ectiveness of decision making in power sharing via collective responsibility. As discussed in the next two chapters, the GFA and the St. Andrews Agreement established more sophisticated provisions requiring cross- communal voting in the assembly and, under certain circumstances, in the executive itself. It is worth contrasting this preference for the principle of unanimity via collective responsibility with the rules for veto rights in the current cases of power sharing in Northern Ireland, Bosnia, and Macedonia. In partic u lar, more formal rules for executive decision making have been in play in North- ern Ireland since the GFA. As I discuss in the next chapter, the power- sharing executive formed in 1999 was subject to intra-executive debates and open disagreements over policy decisions taken by ministers. Overall, the evidence shows that government ministers in a power- sharing government may well be prone to criticizing each other’s decisions. Th is was certainly the case in the Northern Ireland executive 1999–2002, when the two Sinn Féin ministers made policy decisions (on the site of a maternity hospital and the decision to abolish the transfer test for school pupils from primary to secondary education) that off ended their unionist colleagues. In Bosnia, members of the state presidency have oft en criticized each other and vetoed each other’s decisions. And in Macedonia, there have been intra- coalition 48 Chapter 2 disputes aplenty, including recent rows over Macedonian ministers’ plans to rejuvenate Skopje city center. So are these intra-executive spats, oft en taking on an inter-ethnic dimension, problematic for power-sharing democracy? Clearly it would be preferable for executive ministers to iron out their diff erences over policy behind closed doors. Empirical reality, however, suggests that departure from consensus is to be expected in power sharing, particularly when the issue at hand takes on an ethnic or cultural dimension. More- over, an important incentive for groups when entering power sharing is the right to protect their communal interests via veto powers.34 Th us, the Sun- ningdale case is something of an aberration with respect to its commitment to the principle of collective responsibility over veto rights. Despite this commitment to unanimity and consensus, the story of the Sunningdale executive is one of mounting inter-party tensions and pressure from politi cal forces opposed to power sharing and the creation of the Coun- cil of Ireland. Establishment of the Council was to prove a source of increasing diffi culty. For Faulkner, progress on ratifi cation of the Council could not take place without recognition of Northern Ireland’s constitutional status within the UK by the Irish government. Dublin’s reluctance to provide such a concession to aid Faulkner was further complicated by an Irish court case in which Kevin Boland (a former Fianna Fáil member of parliament) argued that the Irish government’s declaration in the Sunningdale communiqué contravened Articles 2 and 3 of the Irish Constitution. Th ough Boland lost the case, Faulkner remained intent on calling on Liam Cosgrave’s government to issue a formal declaration on Northern Ireland’s status. Faulkner also wanted to see the Irish government do more to tackle IRA violence before he would commit to the creation of an all- Ireland institution. Th ese tensions over security and status aff ected relations between Faulkner and the Irish government and among the executive parties. Faulkner and his unionist colleagues were increasingly at odds with their SDLP partners in government over the ratifi cation of the Council. While unionists wished to delay movement on the Council, nationalist ministers wanted to see immediate progress. Executive ministers, including the SDLP ministers, were reportedly reluctant to transfer considerable power from their respective departments to the Council of Ireland. Kerr notes that ministers believed that “delegating too much power too quickly to a Council of Ireland would destroy the power- sharing agreement.”35 Th e Sunningdale case has certainly much to say about the operation of a moderate middle power- sharing executive. Inter- communal cooperation Sunningdale Executive 49 within the executive may well be limited despite (or even because of ) the exclusion of more hard-line elements. Th e Sunningdale case clearly demonstrates that cooperation will likely be hampered by fl anking from outside and considerable intra-communal competition. Th is was particularly true for Faulkner, who sought to thwart pressure from Paisley and Craig and was ultimately defeated by the Ulster Workers’ Council strike. Kerr provides a full analysis of the strike with details of the industrial stoppage and the strikers’ objective of delivering a fatal blow to the Sunningdale executive.36 Of relevance here is how such opposition to power sharing and wider institutional arrangements (in this case the Council of Ireland) will certainly damage parties’ attempts to cooperate. Certainly, when one community appears to support such additional arrangements and the other fails to secure reciprocal gains, the agreement will be on rocky ground. By its very nature, a moderate middle may face pressure from politi cal forces who wish to bring the whole thing down. We might suggest, then, that a moderate middle coali tion is not a preferable arrangement and that an inclusive executive may have a greater chance of survival. Th e diffi culty here, however, is that a co ali tion of the middle might be the only show in town as other politi cal forces continue to support violence. In this situation, the best that moderate parties and third- party mediators can do is to anticipate the eff ect such pro-violence, anti- power-sharing forces may have. Extra vigilance is required in devising an institutional arrangement that can be sold to the wider community. In the following chapters, I consider these lessons from the Sunningdale case for the more inclusive power-sharing executives in Northern Ireland since the GFA, Bosnia since the DPA, and Macedonia since the OFA. Execu- tive formation rules under the GFA are more sophisticated than the Sunning- dale executive’s use of Whitelaw’s formula and his involvement in assigning parties seats and portfolios. Th e d’Hondt sequential portfolio allocation method, used to allocate ministerial seats under the GFA, allows for a more inclusive and proportional sharing of seats based on parties’ strength in the assembly. Bosnia has strong, corporate rules for executive formation under which power sharing requires the inclusion of representatives from the three constituent peoples and on an entity basis. Th e situation in Macedonia is much more fl exible; executive formation comes about as a result of inter- party bargaining. Of note, however, is that Bosnia and Macedonia have also had some experience of “moderate” power sharing. Explored in Chapter 5, the Alliance for Change Coali tion was formed in Bosnia in the aft ermath of the 2000 elections. Essentially patched together by the international 50 Chapter 2 community, the executive failed to produce the expected level of policy reform and compromise, disappointing its international sponsors and prompting the Offi ce of the High Representative to impose decisions. Similar to the Sunningdale executive, the Alliance for Change co ali tion was incoherent, lacking a common purpose, and ultimately unable to remain in power in the face of competition from more hard- line parties. Tensions among the parties in power led them to contest the 2002 elections on separate platforms, allowing more radical parties to top the polls. In Macedonia, there have been diff erent confi gurations of moderate and more radical parties in power. Post- electoral negotiations have been formed by a more moderate Macedonian party and a more radical Albanian party, and vice versa. Since 2008, power sharing has consisted of the more radical party on each side. As Chapters 7 and 8 demonstrate, there has been no marked diff erence in power sharing with the two more radical parties in power.

The Role of External Actors

In addition to the impact of institutional rules on power sharing, I consider whether and how external actors infl uence the practice of power-sharing democracy. In this case, the British and Irish governments, particularly the British government’s Secretary of State for Northern Ireland William Whitelaw, were important players in the peace negotiations. But can we really treat the two governments as external to the confl ict? To state an obvious point, the two governments hardly had a choice on whether to be involved. On the one hand, Northern Ireland was (and remains) constitutionally part of the UK; London was faced with the task of fi nding an alternative to direct rule following Heath’s prorogation of the Stormont parliament in 1972. And Dublin had a constitutional obligation to pursue the goal of Irish unity, and its collaboration with the SDLP was important for the outcome of the process. Th ough the Irish government did not want Irish unity to come about any time soon, the creation of an all-Ireland institution, the Council of Ireland, was seen as the best way to bring about a united Ireland in the future. Certainly, Whitelaw was instrumental in securing an agreement. Kerr refers to the Secretary of State’s “self-defi ned role as inde pen dent arbitrator,” in listening mode with the three internal parties.37 Following the publication of proposals in the 1972 green paper, Whitelaw sought to direct a “roll- Sunningdale Executive 51 ing” or “evolutionary” process toward a settlement on devolved government in Northern Ireland.38 Th is evolutionary pro cess would, he hoped, be more likely to secure support from nationalists and unionists than would an imposed settlement. Much, of course, remained within the control of the British government. Whitelaw reportedly remained “undecided on whether he would guarantee the minority community a number of seats on the executive,” favoring a committee system based on the strength of parties’ support in the assembly.39 If Whitelaw considered himself an “arbitrator,” should we consider the role of two governments as arbitration or mediation? Th e Merriam- Webster dictionary describes mediation as “intervention between confl icting parties to promote reconciliation, settlement, or compromise.” Arbitration is described as “the hearing and determination of a case in controversy by an arbiter.” An arbiter is either “(1) a person with power to decide a dispute” or “(2) a person or agency whose judgment or opinion is considered authorita- tive.” 40 Arguably, the two governments displayed elements of mediation as well as arbitration. Th e British government acted as an arbiter in Whitelaw’s imposition of his formula solution to the sharing of ministerial seats among the three parties. Indeed, the Secretary of State threatened that if the parties did not agree, he would invite an outsider to sit on the executive and he would allocate executive posts to the relevant parties.41 To break the deadlock over the “numbers game,” Whitelaw presented a new formula of fi ft een positions: eleven executive members (including chief executive and deputy chief executive), nine departments, and four nonvoting members outside the executive. Ultimately, the parties agreed to this approach to executive formation, including four members who would not have voting rights. Th e fi nal distribution among the parties was seven for Faulkner’s unionists, six for the SDLP, and two for the Alliance. As a result of Whitelaw’s direction, Faulkner had secured a majority on the executive but not across the administration as a whole. Th e mix of arbitration and mediation in this case is also evident in the other power-sharing cases under investigation. Th e following chapters highlight the importance of self-defi ned mediators or arbiters in power sharing. In the Northern Ireland peace pro cess leading to the GFA, the British and Irish governments adopted a binational intergovernmental approach to resolving the confl ict.42 In the context of the agreement, the two governments were themselves also key negotiators over the wider po liti cal settlement arrived at in the British- Irish Agreement, an annex to the multi- party agreement, 52 Chapter 2 dealing with the status of Northern Ireland and the implementation of north-south and east-west institutions. In discussions with the parties, Brit- ish Prime Minister Tony Blair and Irish Taoiseach Bertie Ahern were key mediators, seeking to “deliver” the commitment to a deal by unionists and nationalists, respectively. Th eir eff orts were supported by Bill Clinton’s administration, with former U.S. Senator George Mitchell acting as chairman of the talks. When a new deal was reached by the DUP and Sinn Féin following the St. Andrews Agreement of 2006 (a document produced by the two governments), London and Dublin sought to mediate discussions between the two sides but also threatened to impose their “Plan B” of greater British- Irish cooperation in the event of no agreement between the parties. In Bosnia, the DPA was largely imposed on the internal parties by the international community, particularly the U.S. team led by Richard Holbrooke. Discussed in Chapters 5 and 6, the issue of arbitration has been a sore point for the operation of power sharing in Bosnia, given the powers bestowed on the Offi ce of the High Representative to impose policy decisions over the heads of local politicians. In contrast, the situation in Macedonia is more akin to mediation than arbitration. In Chapters 7 and 8, I discuss the positive role played by the U.S. and the EU in securing the Ohrid Agreement in 2001 and their subsequent eff orts at persuasion, rather than imposition, regarding policy reform. Together, these instances of power sharing highlight how third parties or external actors in power sharing can shift from negotiator to mediator to arbiter. I expand upon these insights in the Conclusion chapter to consider how these self-defi ned roles can have negative consequences for the stability of power- sharing democracy.43 Th e Sunningdale case also highlights the commitment external actors may have regarding the implementation of the agreement and support for the power- sharing executive when it faces challenges. In the Sunningdale case, relations among the three co ali tion parties were under strain due to pressure from ongoing violence, intra- communal competition, and the risk of collapse. It was arguably a mistake that the British government did not seek to uphold the executive in the face of pressure from the Ulster Workers’ Coun- cil strike. Th e strike eff ectively brought Northern Ireland to a standstill and proved too much of an obstacle for the executive to manage. Of note is London’s attitude to the executive when it was facing collapse. Certainly, Wilson was not prepared to do much about it. Arguably, the lack of support on the part of the British agreement was crucial for the executive’s demise. In a televised address, Wilson accused the strikers of “viciously defy[ing] Sunningdale Executive 53

Westminster, purporting to act as though they were an elected government; people who spend their lives sponging on Westminster and British democracy and then systematically assault democratic methods.” 44 Unsurprisingly, Wilson’s attitude did not enamor the unionist population opposed to Sun- ningdale. Th e Irish government called on London to act decisively to save the executive, to no avail.45 Under these circumstances, the power-sharing executive felt increasingly bereft of support and isolated and had little faith in Wilson’s government to get power sharing back on track. Aft er just fi ve months in offi ce, the executive fell apart at the end of May 1974. Th e Sunningdale case speaks to the broader issue of externals’ commitment to upholding the peace agreement they helped to broker. We might presume that because external actors sought to secure a peace agreement in the fi rst place they will work to maintain the deal once the institutions are in operation and risk collapse. When power- sharing executives face a severe crisis, particularly due to pressure from anti-power- sharing forces, the external actors involved clearly need a contingency plan to deal with these challenges that threaten to bring the whole system crashing down. Will external actors have the po liti cal will, and authority, to ensure the implementation of a power-sharing settlement? Th e other cases studied here suggest some complexity regarding external actors’ commitments to the power- sharing deal. In post-GFA Northern Ireland, the British and Irish governments were (largely) committed to the deal, save some inconsistency on the part of the former. Christine Bell suggests that the British-Irish treaty section of the GFA “can be argued to provide an ‘external guarantee’ of the agreement’s commitments but also amounts to a self-guarantee of both states to each other, to support the peace process.” 46 Yet London’s commitment to upholding the power- sharing institutions was questioned following its legislation for authority to suspend the devolved assembly. Discussed in the next chapter, Secretary of State Peter Mandelson suspended the Northern Ire- land Assembly in February 2000, arguably to prevent the resignation of fi rst minister and unionist leader David Trimble in the absence of IRA decommissioning. With suspensions occurring several times before Novem- ber 2002, nationalist politicians and the Irish government all opposed these powers as an abuse of the British government’s obligations and commitment to the GFA. In Bosnia, external actors’ commitment to the power- sharing arrangements as set out in the DPA has been questionable. Th e Offi ce of the High Representative was created as a civilian actor to ensure the implementation of the agreement. In the early post-Dayton years, the High 54 Chapter 2

Representative imposed policy decisions to ensure the system worked according to both the letter and spirit of the DPA. We will see in Chapter 6, however, that decisions by the High Representative in more recent years have sought to amend the institutional framework in a manner that departs from the essence of group rights under Dayton. In Macedonia, the EU delegation has sought to uphold the Ohrid Agreement by encouraging both sides to cooperate and make policy reform.

An Imbalanced Peace Deal and Costly Commitments

Another potential lesson from the Sunningdale case relates to how an imbalanced peace agreement will have adverse consequences for the stability of the power-sharing institutions. Clearly, the Sunningdale agreement was imbalanced in the sense that Faulkner did not achieve as much for unionists as the SDLP gained for nationalists. As Kerr suggests, the SDLP and the Irish government over-negotiated, to the detriment of Faulkner’s already precarious position.47 For the SDLP and the Irish government, the overall aim was to achieve power sharing plus an all-Ireland dimension in the form of the Council of Ireland. For Dublin, the Council was crucial for the settlement as a whole, the objective being to “set up an institution which could evolve into an all- Ireland government, elected by and responsible to an all- Ireland electorate.” 48 At the Sunningdale negotiations there were several issues to be ironed out: whether the Council of Ireland should have executive functions, a policing role, a permanent secretariat, and the potential to evolve over time.49 Th e SDLP, Alliance, and the Irish government wanted a Council of Ireland with executive powers; Faulkner resisted going beyond a consultative model. Having been “terrorized” or “bullied” by Heath, Faulkner arguably accepted a format for the Council of Ireland he knew would be diffi cult to sell to the unionist population.50 Faulkner also lost out in the sense that he was unable to achieve revision of the Irish constitution as a precondition for executive formation in Northern Ireland. Resisting calls from Faulkner for constitutional amendments, the Irish government nevertheless indicated it would be prepared to recognize Northern Ireland once an agreement was in place. Its preparedness for such recognition was, of course, dependent on securing a robust Council of Ireland, an institutional outcome relentlessly pursued by Irish Minister for External Aff airs Garret FitzGerald.51 Sunningdale Executive 55

It would be naïve to hope for a peace agreement whereby the negotiating parties are content with everything in the deal. Th ey will likely have some reservations about some aspects that they will fi nd diffi cult to sell to their constituents. In promoting the deal as the best outcome (at least under present circumstances), parties will likely play down some aspects and emphasize others. Th is brings us to the notion of credible commitments in peace agreements. A subject of extensive scholarship within the wider IR literature, commitment problems occur when the negotiating parties feel unable to credibly commit to their obligations under the agreement in the future.52 Th e associated costs may simply be too much for a party to manage. Matthew Hoddie and Caroline Hartzell note that the costs to a group’s leader in signing a peace agreement are twofold: (1) the cost of conceding war aims and the need to compromise with rivals and (2) the “very act of agreement will create schisms within parties to the agreement between those who prove more and less amenable to the compromises that have been reached.” 53 Ar- guably, then, it is important that parties’ commitments are not “too costly,” or at least unsellable to their respective rank and fi le supporters. But how can negotiators know that the deal risks being too costly for one or more parties? For Barbara Walter, the engagement of external actors as a guarantor of the terms of the agreement can help increase cooperation among the contending groups in reaching a deal.54 But external actors engaged in the process will be reliant on local elites to determine what they can and cannot sell to their party faithful. External actors need to be cognizant of the potential problems arising from the issue of costly commitments. In the Sunning- dale case, Faulkner’s diffi culties were either not fully acknowledged by the British government or simply ignored. At the talks, Heath did not go out of his way to help the unionist leader. Moreover, Faulkner failed to secure side payments or concessions (certainly the Irish constitution remained an issue) that would have helped him sell the agreement to his supporters. His failings here contrasted with the achievements of the SDLP who, with the backing of the Irish government, achieved power sharing and the Council of Ireland. Th ough Faulkner later renegotiated aspects of the all- Ireland dimension, this was too late given the amassing of anti-Sunningdale support from within the wider unionist community. Th ere are clear lessons from the Sunningdale case regarding the challenges arising from a peace agreement that is imbalanced, when commitments may become too costly for a negotiating party and the opposing group appears to have gained much more. Th e evidence presented in later chapters 56 Chapter 2 shows that the objective of reaching a balanced deal is important, not just in securing the deal in the fi rst place, but in securing its implementation and continued support from the contending groups. Th e next chapter outlines how the outcome of the Northern Ireland peace talks in 1998 was a more balanced deal between nationalists and unionists— particularly in terms of the incentives set out by the two governments and the side payments or concessions off ered to the parties. Unionists were satisfi ed on the constitutional question of Northern Ireland’s status; nationalists secured power sharing and more sophisticated rules to guarantee their inclusion in government subject to parties’ electoral strength. Nationalists also secured north- south institutions, the all-Ireland dimension that was so controversial at Sunning- dale; east- west institutions were established as a balance and an incentive for unionists. Discussed in Chapter 4, there was an attempt by the two governments to achieve some balance between the DUP and Sinn Féin at the St. Andrews negotiations in 2006. Th e deal became a trade-off between the two internal parties: Th e DUP would have to agree to share power with Sinn Féin, and the latter would have to accept the legitimacy of the Police Service of Northern Ireland (PSNI). Ultimately, the DUP agreed to share power in the context of a changed security situation (following IRA decommissioning in 2005) and some revisions to institutional rules governing ministerial authority. Sinn Féin changed its policy to support the police ser vice, entered a power- sharing executive, and secured the devolution of policing and justice powers from Westminster. In Bosnia, the Dayton peace talks were about stopping the war and getting a deal, and not an attempt to reach a sophisticated constitutional and institutional package that could be sold to, never mind endorsed by, the three groups. Th e two entities (Republika Srpska and the Bosniak-Croat Federation) were taken as a given, much to the chagrin of the Bosniak community, who saw this accep tance of the entity structure as capitulation to the Serb aggressors. Yet the focus on securing group rights in the Bosnian constitution (a separate annex of the DPA) via strong corporate rules for power sharing was an attempt to ensure balance in decision making. Arguably, the Bosniak negotiators agreed to a costly commitment in accepting the two-entity structure; in subsequent years Bosniak elites have sought to gain support from the international community for radical revision of these structures, even the abolition of Republika Srpska. Costly commitments were not such a problem in reaching the Ohrid Agreement between Macedonians and Albanians in 2001. Th ough the deal Sunningdale Executive 57 was largely about addressing Albanian grievances that led to the confl ict, Macedonian elites were satisfi ed that the deal settled these issues and maintained the territorial integrity of the state. Th e Albanian community gained the right to use their own language in state institutions, police reform, increased repre sen ta tion in public administration, and a commitment to greater decentralization. Th e agreement also improved group rights in power sharing via a commitment to use a concurrent majority voting procedure in the assembly relating to legislation on culture, language, education, personal documentation, and symbols. Macedonian parties accepted these changes to secure Albanians’ accep tance of the unitary character of the state and prevent escalation of the confl ict. Moreover, the agreement did not go so far as to create a binational state, thus allowing Macedonians to maintain their privileged position as the majority group. Th ough Ohrid was not too costly for Macedonians, it is notable that the majority group elites have more recently sought to play down its importance, seeking to relegate it to the past. As a consequence of Faulkner’s “costly commitments,” pro- Sunningdale unionists were increasingly under pressure from anti-agreement forces who can arguably be included under Stephen Stedman’s concept of “spoilers.” 55 Certainly, a section of unionism, as well as loyalist paramilitaries, feared that the creation of the Council of Ireland would have the potential to bring about a united Ireland. Seeking to undermine the process, diff erent actors employed diff erent tactics, ranging from the violence orchestrated by loyalist paramilitaries (republican paramilitaries also hoped to derail po liti cal progress) to pressure from politicians Paisley and Craig and the UWC strike. Stefan Wolff suggests that “the mutual infl iction of pain legitimized the radicals in each community, thus further diminishing the ground on which moderates could seek and reach agreement.” 56 In 1998, the parties had greater space in which to reach an agreement owing to the paramilitary ceasefi res and the accep tance of the Mitchell principles of nonviolence.57 Th e problem of spoilers is, of course, a more general phenomenon in peace processes. In the later peace process of the 1990s, progress was hampered by ongoing paramilitary violence, and the early post-agreement period was threatened by the Real IRA bomb in Omagh in August 1998. Additional pressure came from the DUP who, though availing of the party’s two ministerial seats, continued to call for a new agreement. Th e chapters on Bosnia also highlight the problem of spoilers who are internal to the po liti cal system, including po liti cal elites who boycott parliamentary 58 Chapter 2 sessions and executive meetings in an eff ort to undermine the institutions agreed under the DPA. As Desirée Nilsson and Mimmi Söderberg Kovacs suggest, there is scope for new research on the variety of spoiler types and tactics used to derail a peace process. 58 Th e Sunningdale case demonstrated that when spoilers are strong the power-sharing executive will be subject to considerable pressure and ultimately risk collapse. Th ere is clearly a need to know more about how external actors engaged in a peace process can prevent the emergence of would-be spoilers or manage spoilers already manifest.59 Finally, the imbalanced nature of the Sunningdale agreement highlights the challenge of peace implementation when the deal lacks wider public approval. Notably, in the Sunningdale case, the election to the assembly took place in June 1973, several months before the negotiation of the deal. More- over, there was no referendum on the deal, though many unionists used their vote in the 1974 Westminster election to express their opposition. We can contrast this situation with the 1998 agreement, which was subject to (positive) referenda on an all-Ireland basis (in each jurisdiction) in May and a Northern Ireland assembly election in June. Th e challenge of devising a mechanism to seek wider public support is oft en ignored in peace processes. Th is was certainly the case with the largely imposed DPA that was not put to pop u lar scrutiny. It is unsurprising, then, that the DPA was considered illegitimate and increasingly problematic in its implementation. In Macedonia, parliament adopted constitutional amendments a few months aft er the signing of the peace agreement, approved by a majority of Macedonian and a majority of Albanian deputies. Th e cases raise the question of whether and how to secure public endorsement of a peace agreement. Should negotiators submit the deal to pro cesses of deliberative democracy, including referenda? As Stephen Tierney notes, the referendum in the republican tradition pre- supposes “a demos” and may act as a “homogenizing” mechanism in a deeply divided territory.60 Ian O’Flynn suggests that public deliberation may well create “a stronger sense of common national identity among members of a divided society.” 61 Yet Tierney suggests that it may possible to respect the reality of “multiple demoi” and devise a mechanism based on concurrent majorities. Indeed, as peace agreements oft en lack endorsement and wider legitimacy, these normative and prescriptive questions surrounding referendums in deeply divided places arguably warrant further research. In any case, what matters is whether “a referendum can meet the rigorous demo- Sunningdale Executive 59 cratic tests of both the substantive issue to be tested and the pro cess agreed to by actors across a deeply divided spectrum.” 62

Summary

In this chapter I have sought to draw several lessons from the Sunningdale case, an instance of failed power sharing. Despite the failure of power sharing in the 1970s, the basic ingredients of the deal (power sharing plus an Irish dimension) formed the basis of a rolling peace process in the 1990s, ultimately arriving at the GFA in 1998. Th e failure of the Sunningdale executive also provides important insights for power sharing more generally. First, the executive was an interesting experiment in “moderate middle” power sharing. Th ough the three main internal parties agreed to share power, the case demonstrates that the more moderate parties may, nevertheless, be unable to agree on executive formation and may require the engagement of an external arbiter to overcome deadlock. Once in power, the co ali tion may seek to get on with the business of making policy but may be hindered from doing so due to divisions within the executive and pressure from outside the institutions from spoiler forces opposed to power sharing and/or the wider peace deal. Building consensus among the executive parties may be diffi cult, particularly when there are residual diffi culties over aspects of the agreement, as per the Council of Ireland. We might be tempted to suggest that the experience of the Sunningdale executive makes clear that a fully inclusive co ali tion would be preferable over a moderate middle co ali tion. In the following chapters, I seek to tease out the nuances of fully inclusive power sharing over a preference for only moderate parties. Yet the Sunningdale case also off ers a word of caution in this debate. A fully inclusive co ali tion may simply not be possible as some actors refuse to relinquish their violent methods. Policymakers may well fi nd themselves in a situation whereby the only feasible option, given the ongoing engagement of some po liti cal forces in violence, is to court these moderate parties. Th ey should not assume, however, that a moderate middle executive will deliver moderation. Another signifi cant lesson from the Sunningdale case relates to the role of external actors, particularly in relation to their (sometimes questionable) commitment to uphold the power-sharing deal when the executive risks collapse. Th e case demonstrates that external actors may not be overly 60 Chapter 2 committed and may resist having to police or guarantee the agreement. Th e attitude of the British government to the executive in the context of the UWC strike shows that a coali tion under pressure from opposition forces will likely collapse in the absence of external actors’ support. Arguably, then, external actors will need to stay the course, be committed to upholding the deal, and be consistent in their dealings with the internal parties. To an extent, external actors’ responsibilities will have been set out in the agreement itself. But their role may well shift in practice from mediator to arbiter, guarantor, administrator, and norm promoter.63 As Christine Bell notes, a guarantor of the peace agreement “commits to ongoing interest in the implementation of the settlement, and possibly some form of po liti cal response to the viola- tions of the agreement.” 64 External actors need to consider whether they are prepared to police the deal, guarantee the institutions, and whether they have the (moral and legal) authority to do so. Discussed in Chapters 5 and 6 on Bosnia, the evidence shows that peace agreement guarantors obliged to uphold a peace agreement may in practice (and somewhat ironically) seek to reform the institutional arrangements previously agreed by the parties. Th e failure of the Sunningdale case also highlights the problem of an imbalanced agreement when one party secures more of its negotiating aims than the other(s). In this situation, the leader of the party with fewer “wins,” side payments, or concessions will be under pressure for failing to deliver the party’s goals, failing to protect group interests, and will be left vulnerable to accusations of treachery. Faulkner’s demise can be seen in this light. Th e unionist leader’s commitments to the Sunningdale package were simply too costly for the wider unionist community to bear. Th e case shows that external actors need to pay careful attention to when domestic elites stretch their room to maneuver and go beyond what their party members and wider community might accept. Th e role of internal and external spoilers is important here. Th e goal for policymakers is not just securing a deal but securing a deal that sticks. Finally, the Sunningdale case raises the more normative question of whether an imbalanced peace deal, where elites have made costly commitments, should be subject to public endorsement in a referendum. Th e Sunningdale case, and the case studies to come, point to the need for more research on whether and how procedures of deliberative democracy might handle the issue of peace agreement legitimacy in a contested state/territory with multiple demoi. CHAPTER 3

The Good Friday Agreement 1998: An Inclusive Co ali tion

Th e collapse of Northern Ireland’s power-sharing executive in May 1974 and the restoration of direct rule from London took place amidst violent confl ict that continued for three decades. More than 3,600 lives were claimed. In 1998, a peace agreement was signed among the politi cal parties that set up a power- sharing executive and institutions refl ecting relations between Northern Ireland and the Republic of Ireland and relations between the UK and Ireland. More than any preceding initiative, the power-sharing deal agreed between nationalists and unionists, loyalists and republicans in the Good Friday Agreement of April 1998 (GFA) promised the resolution of the Northern Ireland confl ict. As Joseph Ruane and Jennifer Todd note, the Agreement was a “momentous occasion” that seemed to be “a compromise not simply between unionism and nationalism but between loyalism and republicanism, and promising a permanent cessation of violence.”1 Aft er addressing the diff erences between the GFA and the Sunningdale agreement, I discuss how the GFA met the preferences of both nationalist and unionists. I then set out how the design of the power-sharing executive under the GFA was the result of an evolutionary process of institutional design. In addition to this evolutionary process, inter-party bargaining took place among negotiators in the fi nal days of the talks, with details of executive design reached as a compromise between the two largest parties, the UUP and the SDLP. Th e practice of power sharing illustrates the limitations of institutional rules when parties are themselves constrained by serious po liti cal and security issues relating to the confl ict. I argue that a more inclusive executive (facilitated by an improved security context and a constitutional compromise) provided for greater cooperation than the moderate 62 Chapter 3 middle Sunningdale coali tion. Although the eff orts of the British and Irish governments were crucial in setting up and maintaining power sharing, the role of the Blair government needs to be qualifi ed in light of some misjudged and inconsistent decisions along the way.

Addressing Party Preferences for a Peace Agreement

It took twenty- fi ve years of violent confl ict before the parties were able to agree on a peace settlement. Th ough the two governments eventually arrived at promoting power sharing as part of the rolling peace pro cess of the 1990s, there was some earlier variation in their approach, particularly on the part of successive British governments in the mid- to late 1970s and early 1980s. O’Leary and McGarry suggest that under Secretary of State Roy Mason, not much eff ort was expended to fi nd a politi cal solution “lest they heighten expectations and create ambiguity.”2 But once in power, the Conservatives pursued policies that considered power sharing, as per Secretary of State James Pri- or’s proposals of 1982 whereby devolution would take place incrementally. In this context, the Anglo-Irish Agreement of 1985 was an important development, given its commitment to power sharing and a formal legal role for the Republic of Ireland. By the 1990s, the two governments had arrived at an intergovernmental approach that included internal power sharing (along with north-south institutions) as a means to address the competing self- determination claims. Brendan O’Duff y argues that the similarities in the basic components of the Sunningdale communiqué and the GFA “mask vital diff erences in approaches to confl ict regulation that refl ect evolving and pluralizing concep- tions of both British and Irish nationalisms.”3 He notes a major diff erence between the approach of Labour Prime Minister Harold Wilson in 1974 and Labour Prime Minister Tony Blair in 1998, in that the latter “spoke of a ‘process’ which, while based exclusively on peaceful means, is explicitly open- ended with regard to the constitutional status of Northern Ireland, and the development of cross- border institutions between the Irish Republic and Northern Ireland.” 4 Th is key diff erence in approach on the part of the British government, coupled with an increasing intergovernmental approach with the Irish government, meant that an agreement would necessitate the accommodation of the two communities’ competing self- determination claims. O’Duff y describes the GFA as “the pinnacle of bi-national intergovernmental Good Friday Agreement 63 confl ict regulation.” 5 Th eir approach is encapsulated in an annex to the Agreement, which includes a statement by the two governments, who “recognise that it is for the people of the island of Ireland alone, by agreement between the two parts respectively and without external impediment, to exercise their right of self-determination on the basis of consent, freely and concurrently given, North and South, to bring about a united Ireland, if that is their wish, accepting that this right must be achieved and exercised with and subject to the agreement and consent of a majority of the people of Northern Ireland.” 6 Th e design of the po liti cal institutions under the GFA is notable for several reasons. Infamously described by SDLP politician Seamus Mallon as “Sunningdale for slow learners,” the 1998 Agreement is, in many ways, similar to the deal reached in 1973. Power sharing between the two communities and an all-Ireland institution were crucial elements of both Agreements and, following the collapse of Sunningdale, continued to be important elements of a potential settlement. Th ere are, however, important diff erences. Stefan Wolff notes the diff erences between the GFA and Sunningdale relating to “contextual factors and procedural regulations.”7 In relation to procedures, he notes that the GFA departed from Sunningdale with the inclusion of the d’Hondt procedure for executive formation; cross-community voting procedures in the assembly; and “the fact that the implementation of decisions taken by the North-South Ministerial Council has been made dependent upon their approval by both the Irish Parliament and the Northern Ireland Assembly.” As Wolff suggests, this latter provision “refl ects the commitment, in par tic u lar by the British and Irish Governments, to assure the unionist community that no decision can be made without their consent.”8 Th e GFA departs from the Sunningdale communiqué in a signifi cant way, given the inclusion of republicans in power sharing. Indeed, inclusivity is a key feature of the Agreement, which sets it apart from previous failed initiatives. As Horowitz writes, previous eff orts, particularly the Sunning- dale communiqué, “had been premised on the assumption that moderate unionists and moderate nationalists should join together against those who were unwilling to cooperate with the other side.” 9 Th is was not to be the case with the rolling peace pro cess of the 1990s, which culminated in the GFA. Horowitz explains that what transpired was a realization on the part of the politi cal parties of the “main ‘lesson’ of history: so many eff orts had failed to produce a politics of intergroup accommodation that it was time for a bold new departure. An inclusive regime, with air-tight minority guarantees, 64 Chapter 3 had not yet been tried.”10 Th e context allowing for inclusivity, of course, was very diff erent in the two periods. In the 1970s, republicans rejected power sharing as an ac cep tance of the partition of the island. Loyalists also rejected the Sunningdale deal in opposition to power sharing and any role for the Republic of Ireland in Northern Ireland aff airs, as promised by the Council of Ireland. Th e Sunningdale executive was hampered by ongoing violence by the IRA, loyalist paramilitaries, and the Ulster Workers’ Coun- cil strike. By the 1980s, a shift away from the armed struggle had begun in republican circles, ultimately leading to Sinn Féin’s engagement in electoral politics, the IRA ceasefi re of 1994, and Sinn Féin’s inclusion in inter-party talks in 1997.11 In terms of the type of institutions created to manage, or ultimately resolve, the confl ict, they clearly added up to a consociational settlement. Lij- phart’s four characteristics of consociational democracy are evident in the deal: cross-community executive power sharing; proportionality; mutual veto; and segmental autonomy, albeit the latter to a lesser extent. Interest- ingly, there is a view in the literature that the GFA is not wholly consociational. Paul Dixon argues that the power- sharing executive is not a grand coali tion; PR (STV) is used rather than the list system; the British government has an external veto in suspension powers; and the support for integrated education and housing measures and community relations confl ict with the idea of segmental autonomy.12 John Coakley suggests that group autonomy is absent from the politi cal framework in Northern Ireland, arguing that “the existence of de facto separate networks of Catholic and Protestant schools hardly amounts to a system of segmental autonomy.”13 He recommends dropping the element of segmental autonomy from consociational theory, thereby making it easier to apply the consociational label to empirical cases as well as clarifying the meaning of consociation.14 Yet there is considerable agreement among scholars that the GFA is, indeed, consociational. Horowitz says that while it is “not perfectly consociational . . . the consociational coherence of the document stands out.”15 Wilford refers to its “consociational (plus) template.”16 O’Leary argues that it is more than consociational due to its external dimensions: “It is one made with national and not just ethnic or religious communities, and it is one endorsed by both leaders and the led . . . the Agreement establishes an internal consociation built within overarching confederal and federal institutions.”17 Importantly, there was no predetermined plan to establish a consociational blueprint. Wilford and Wilson note that “the agreement was Good Friday Agreement 65 prepared with no regard to the wider intellectual debate between ‘consocia- tionalist’ and ‘integrationist’ approaches to constitutional engineering in divided societies.” Instead, the Agreement can be seen as following “common sense” consociationalism.18 Certainly, though the GFA was the outcome of inter- party talks, which began in June 1996, it was arguably rooted in a longer pro cess of eff orts to bring about an agreement.19 Th ese failed initiatives included the Constitutional Convention of 1975, the Anglo- Irish Agreement of 1985, the talks led by Secretary of State for Northern Ireland Peter Brooke in 1991 (carried on by his successor, Patrick Mayhew, from July 1992), the Downing Street Declaration of 1993, and the Framework Documents of 1995.20 Horowitz notes that talks leading to the GFA “were really a continuation of a series of meetings going back many years, from which an immanent consensus about what would work and what had not worked in Northern Ire- land had emerged.”21 But in terms of potential institutional designers of the deal reached in 1998, we need to consider how the settlement met the preferences of the main actors— the po liti cal parties in Northern Ireland and the British and Irish governments. O’Leary writes that the GFA “was the product of both tacit and explicit consociational thought, and of ‘pacting’ by most of the leaders of the key ethno-national groups and their respective patron-states.” 22 For nationalists, the GFA contained features that the SDLP had long campaigned for: internal power sharing and an “Irish dimension.” In relation to the Agreement’s Strand 2 (the cross- border relationship between Northern Ireland and the Republic of Ireland), the North-South Ministerial Council (NSMC) was intended to meet nationalist concerns for an institutionalized Irish dimension. Th is is not too far removed from the Council of Ireland structure, which had been such a thorny issue for the Sunningdale executive. It is worth stressing that the SDLP would not have been prepared to support an agreement solely on the basis of internal power sharing within Northern Ireland. Overall, the terms of the Agreement were held as a victory for the SDLP because they included aspects of party policy advocated since the early 1970s.23 Th e Agreement was arguably a momentous occasion, given support from the republican community whose politi cal aim has been Irish unity. So what explains the support of republicans led by Sinn Féin? O’Leary writes that republicans could “trade a long war that they could not win or lose for a long march through institutions in which they can reasonably claim that only their means have changed, not their end: the termination of partition.”24 To the republican base, moreover, it is the endgame that matters most. For 66 Chapter 3 as Sinn Féin president Gerry Adams writes in his memoirs, although while a united Ireland was not yet attainable, it remained the future prize: “Our view was that it was transitional . . . . We knew from the param e ters of the talks laid down by the two governments that Irish unity would not come out of this phase of the negotiations, but we set ourselves the task of weakening the British link while defending Irish national rights.”25 For Sinn Féin, the promise of a mechanism to bring about Irish unity was central in their eff orts to sell the deal to the wider republican community. Certainly, the Sinn Féin leadership would not have been able to support the GFA if the British government had failed to address Irish self- determination claims. Gerard Murray and Jonathan Tonge note that Sinn Féin “entered talks fully aware that they would lead to a three- stranded framework, which would be based upon Northern consent for change, hitherto derided as a unionist veto.”26 Th e party’s strategy was to encourage the republican community to support a po liti cal agreement that would generate wider nationalist support and then was sold as a “genuine advance toward full nationalist parity of esteem within the state, allied to a modest strengthening of the Irish dimension.”27 It is worth noting that the key players in arriving at an agreement were the two governments and the two largest parties, the UUP and SDLP. Th us, Sinn Féin was slightly apart from much of the negotiations. Th e party did not give explicit support for the Agreement for several weeks aft er 10 April. At a special party conference/Ard Fheis on the GFA, an overwhelming majority of delegates supported Sinn Féin representatives taking their seats in a future Northern Ireland Assembly. Among the many features of the Agreement, the two-year release deadline for prisoners clearly helped secure republican support. For unionists, the Agreement was about cementing Northern Ireland’s position within the UK and bringing an end to IRA violence. Arthur Aughey suggests that for unionists, nationalists and republicans had been making gains, particularly since the Anglo- Irish Agreement of 1985: “Th e convinc- ing moral was that unionists should become active participants in, rather than a passive victim of, politi cal change”; with David Trimble becoming leader of the UUP in 1995, he “promised to replace the ‘do nothing’ style of Unionist leadership.”28 Aughey likens Trimble’s position to that of the character Tancredi in the novel Th e Leopard by Guiseppe Tomasi di Lampedusa. Set in Sicily in 1860, the story chronicles the choice facing the aristocracy between resisting the emerging order and accepting some kind of accommodation. Aughey defi nes the Tancredi option thus: “Unless we ourselves Good Friday Agreement 67 take a hand now, they’ll foist a republic on us. If you want some things to stay the same things will have to change.”29 For Aughey, David Trimble’s “wager” on the Agreement “may be taken as the unionist Tancredi option, an attempt to make history rather than to dance to its fatalistic tune.” Given unionists’ opposition to the Republic of Ireland having a role in the aff airs of Northern Ireland, it is unsurprising that negotiations over Strand 2 were diffi cult (regarding the North- South Ministerial Council and the cross- border implementation bodies). Th e initial proposals were the result of negotiations between the British and Irish governments and were presented to the parties by an inde pen dent chairman, former U.S. Senator George Mitch- ell, in a draft agreement on 6 April 1998. Th e details included a wide range of all-Ireland functions with implementation bodies established by London and Dublin in advance of a Northern Ireland assembly election. Dean God- son writes of the strength of opposition from David Trimble and his UUP colleagues and reports how Mitchell was “the unwilling postman for a draft document produced by the two Governments that he knew would not work.”30 Refl ecting unionist concerns, Trimble’s deputy John Taylor memo- rably declared, “I wouldn’t touch this paper with a 40-foot barge pole.”31 Facing considerable opposition within his party and the wider unionist community, Trimble then engaged in discussions with Blair and with the Irish government. In the end, Trimble managed to limit the role for Strand 2 on the basis of a UUP text setting out six north-south bodies and areas of cooperation between the Republic of Ireland and Northern Ireland. Godson asserts that the renegotiation of north- south bodies “was Trimble’s greatest triumph during the talks.”32 Regarding Strand 3 (east- west relations), the British- Irish Council (BIC) was established to bring together representatives of the two governments and devolved institutions in Scotland, Wales, and Northern Ireland as well as representatives of the Isle of Man and the Channel Islands. Th e institution was intended as a countervailing balance to the NSMC under Strand 2. As Graham Walker notes, this institution was to “off er reassurance to Ulster Unionists” and “the inclusion of the BIC in terms of the Agreement appears to have been fundamental to the Ulster Unionists’ ac cep tance of the whole package.”33 Tony Blair’s then– Chief of Staff Jonathan Powell remarks that Strand 3 “was particularly important to the Unionists, because it helped them to present any agreement as being not about the Republic gaining a say in the aff airs of Northern Ireland but of a recalibration of relations between all the parts of the British Isles.”34 68 Chapter 3

Despite this fairly widespread support, not all po liti cal factions in Northern Ireland heralded the Agreement as an acceptable solution to the confl ict. Th ere was important opposition to the GFA from both unionists and republicans. For the DUP and the UK Unionist Party (UKUP), the Agreement could only mean a slippery slope toward a united Ireland. As Nigel Dodds (DUP) claimed, “Th e Northern Ireland recognised in this document is a diff erent one from the Northern Ireland I knew prior to this agreement. Th is is a Northern Ireland in transition to a united Ireland. . . . Once you establish the principle of an all-Ireland authority you have undermined Northern Ireland’s position as part of the United Kingdom in a very fundamental and radical way.”35 And though Sinn Féin ultimately supported the Agreement as a vehicle to realizing a united Ireland, other republicans rejected the deal. An alternative interpretation to that of Sinn Féin’s can be found in the positions of dis- sidents such as the Real IRA and views expressed by the Irish Republican Writers Group via “Th e Blanket” web site. Opposed to the Provisional IRA ceasefi res of 1994 and 1997, the Real IRA was responsible for the Omagh bomb of August 1998 in opposition to Sinn Féin’s support for the Agreement, which they viewed as a betrayal of the goal of a united Irish republic.36 Com- menting on intra-republican divisions, Anthony McIntyre suggests that by entering the Assembly, “republicans are included minus republicanism. Th e republican argument becomes marginalised anyway. Consequently, the Brit- ish state will have rendered ineff ectual the most potent challenge its rule has faced in Ireland since partition.”37

An Evolutionary Process of Executive Design

Th ough the GFA was based on the three interlinking strands of institutions, the focus for this research is Strand 1, which sets out the devolved institutions in Northern Ireland, including the power-sharing executive. Of note is that executive formation under the d’Hondt portfolio allocation procedure was the outcome of an evolutionary pro cess of institutional design refl ecting the parties’ concern for proportionality in power sharing and support for, or ac cep tance of, inclusivity. I also show that in addition to this evolutionary process, some bargaining took place on the details concerning rules for executive decision making. Good Friday Agreement 69

One of the most notable features of executive design agreed by the parties is the inclusion of the d’Hondt portfolio allocation procedure to allocate ministerial seats on the basis of parties’ assembly seat share. Th e inclusion of d’Hondt in the Agreement meant that the governing co ali tion would be made up of not just unionists and nationalists, but also republicans, subject to Sinn Féin’s strength in the assembly. So why did the parties agree to use this procedure to form a cross- community executive rather than simply engage in post- electoral bargaining, as per more “normal” co ali tion governments? Elsewhere I show that d’Hondt came to be included in the GFA as a pro cess of evolutionary and intentional institutional design.38 Th e inclusion of d’Hondt as the method for executive formation was not plucked from thin air but had a historical association with previous proposals and discussions on a settlement. Ultimately, the inclusion of d’Hondt as the preferred method evolved in line with proposals that would include the extremes rather than just the parties of the moderate middle. By the end of the 1980s and following the failure of the 1991–92 talks, the British and Irish Governments and the SDLP were of the view that an end to the confl ict and a politi cal settlement would only be possible with republicans on board. Th e Th atcher and Major Governments, despite public protestations to the contrary, were increasingly pursuing “back door” chan- nels of communication with the IRA to this end. British policy, therefore, was committed to devolved power sharing, and given Sinn Féin’s increasing success at the polls, successive initiatives appeared to be moving inextrica- bly toward the possibility of including republicans, subject to their renounc- ing violence. For instance, Article 4 of the Anglo-Irish Agreement (AIA) of 1985 stipulates a commitment to power sharing whereby “Both Governments recognise that devolution can be achieved only with the co-operation of constitutional representatives within Northern Ireland of both traditions there.”39 As Paul Mitchell, Brendan O’Leary, and Geoff rey Evans point out,“Aft er signing the 1985 Anglo- Irish Agreement it was the considered policy of the sovereign governments to isolate what they called the ‘politi cal extremes’ in Northern Ireland and build up what they called the ‘moderate center ground,’ from which a power-sharing government could be constructed.” 40 Although the AIA was an attempt by the two governments to shore up support for the moderate SDLP, the two governments gradually came to the position that a devolved settlement would require the inclusion of republicans. Secret talks took place between the British government and Sinn Féin 70 Chapter 3 from mid-1990, paving the way for the IRA ceasefi re of 1994 and inter-party talks.41 Over the course of the next decade, successive British and Irish governments came to embrace the peace process and opened the door to the extremes. Arguably an important genesis of the ideas on power sharing in North- ern Ireland comes from the academic literature. It is suggested that the parties were infl uenced by Richard Rose’s 1976 book, Northern Ireland: A Time of Choice.42 Rose groups a number of diff erent options for Northern Ireland under fi ve broad headings: self-government within the UK, direct rule, inde pen dence, a united Ireland, and repartition. In relation to the option of proportional power sharing, Rose writes,

Power sharing could also be realized through the creation of a popu- larly elected executive. An eight-man executive elected by proportional represen ta tion would result in a cabinet of fi ve Loyalists, two SDLP members, and one Alliance member, judging by the Conven- tion election results. . . . Specifi c ministerial offi ces could be chosen in order among the parties. Th e Loyalists, choosing fi rst, might take fi nance, the SDLP commerce, Alliance law reform, and so on. Junior ministers could be appointed by the parties in proportion to their strength in the Assembly— but from outside Assembly ranks.43

Coakley suggests that the move toward explicit proportionality in government formation originated in a paper sent to the Northern Ireland Offi ce in 1980 by two Queen’s University Belfast po liti cal scientists, Sydney Elliott and Jack Smith.44 By the 1990s, some academics were engaged in prescribing potential institutional solutions. In Northern Ireland: Sharing Authority, O’Leary et al. propose a model of shared authority for Northern Ireland between the British and Irish governments and local institutions.45 Inter- estingly, the former Labour Party spokesman on Northern Ireland, Kevin McNamara, claims that the thinking behind the IPPR document came from his offi ce.46 Attention to how new institutional arrangements should work is continued by McGarry and O’Leary, who stipulate that in any future internal settlement “proportionality principles will [also] have to be used by the assembly to establish its committee structures and chairs; and they must be used to elect the executive.” 47 McGarry and O’Leary’s preference was for Sainte- Laguë, which, “fair to small parties, will give both the Alliance Party and Sinn Féin a stake in any new executive or committee-structure for a Good Friday Agreement 71 new assembly.” 48 Th ey also point out that d’Hondt, benefi cial to larger parties, “has been supported by the Ulster Unionist Party and the Demo cratic Unionists in previous negotiations.” But why did the politi cal parties come to support a fully inclusive executive formed by the d’Hondt procedure? It is important that the SDLP came to believe that a solution to the confl ict had to include Sinn Féin. For instance, SDLP member Denis Haughey speaks of his party’s view that agreement among parties of the middle ground would not be suffi cient to sustain peace: “It became increasingly clear that a consensus of the middle ground was not going to be possible because of the tensions, the antagonisms, the bitterness, the problems created by campaigns of violence carried on by paramilitaries . . . by the end of the 1980s we became fi rmly convinced that we could only get an adequate working solution to the problem through . . . switching off the violence and the inclusion of the extreme parties of both traditions.” 49 Former UUP leader David Trimble claims that the d’Hondt procedure was proposed by his party in its submissions to the Brooke and Mayhew talks of 1991– 92. In Frank Millar’s book, the former Ulster Unionist leader wants to “put the record straight” and says, “If you look back at the Unionist Party proposals for the Strand One talks in 1992, about the purely internal Northern Ireland arrangements, it’s d’Hondt.” 50 Th e proposals on the part of the Ulster Unionists in 1992 were, of course, not based on executive power sharing but rather on administrative or legislative devolution in which departments would be controlled by a committee structure. But the unionist basis for power sharing, albeit limited, was nevertheless to be in proportion to electoral support and led to the UUP support for d’Hondt in 1998. As Trimble comments: “Go back to the talks unionist leaders had with Secre- taries of State Peter Brooke and then Sir Patrick Mayhew circa 1991–92 and the agreement on Strand One. Now that agreement wasn’t fi nalised but the broad shape of it was, and we went into the ’96 talks with our ’93 position, which was based on d’Hondt and with the principles of proportionality and automaticity at the heart of it.” 51 Ulster Unionist former minister Michael McGimpsey, who was part of the unionist delegation in the Brooke-Mayhew talks, confi rms that d’Hondt was fi rst suggested in the inter-party talks in the early 1990s: “At that stage, out of the blue, there was a proposal that departments would be shared around; in those days we were talking about committees and who would be chair and there was a suggestion that it would come out by way of d’Hondt.” 52 72 Chapter 3

Unionist support for d’Hondt appears to go back even further than the Brooke- Mayhew Talks and can be traced to a joint UUP- DUP document of 1987 that became the basis for the negotiating position in 1991– 92. Leaked to the Irish Times of 3 July 1991, the document, titled Administrative and Legislative Devolution, proposed a devolved assembly elected by proportional represen ta tion with either 85 or 102 members and a number of committees appointed to run the departments over which the assembly would have power.53 Th e chairs and deputy chairs of the committees would be allocated in proportion to party strength in the assembly under the d’Hondt rule. Furthermore, if a chair was to stand down, the party would nominate a successor, or if that party refused or failed to do so, the position would be fi lled by the party next in line under d’Hondt. Th e 1987 document is signifi cant in that the unionist parties were jointly proposing legislative power sharing on a proportional basis with d’Hondt as the par tic u lar method to ensure proportionality: “As the structure is based upon proportionality, majority and minority alike have a real say, infl uence, and part in the administration and decision making of the province.” 54 De- cision making, however, was not to extend to executive power sharing with nationalists: “As there is no executive, there is ipso facto no executive power sharing. Yet the SDLP can rightly say that they are represented at the highest level.” Th e overriding concern was to “control each committee . . . the larger the committee, the larger the unionist majority and the risk is consequently and proportionately reduced.” Th e important point here is that although the unionist parties were not prepared to countenance executive power sharing with nationalists in the early 1990s, they were fully supportive of the principle of proportionality, which was to become a central part of future initiatives. Former deputy First Minister/SDLP leader Mark Durkan claims that it was he, within the SDLP group negotiating the agreement, who advocated using d’Hondt for the appointment of ministers. For Durkan, the main reason for recommending the procedure was inclusivity: A “voluntary” co alition model would “not be good enough” as it would include the SDLP, the UUP, and Alliance but exclude Sinn Féin and the DUP.55 Th is exclusion would risk allowing parties outside the government to canvass for a “no” vote in the referendum campaign: “We had to make the power-sharing concept as ‘yes-able’ as we possibly could, and to show that, whether people who had already walked out of the negotiations, as the DUP had done, or what people like Sinn Féin had done, basically heckled a lot of the negotiations on Good Friday Agreement 73

Strand 1, the fact was that this was still a deal for them as well, that it wasn’t about certain parties trying to carve things out for themselves.” 56 Unionist support for d’Hondt in previous talks made it sensible for SDLP negotiators to take it up in 1998 because the procedure could attract cross-community consensus once the UUP agreed to ministerial portfolios: “We stuck to the executive power model, but given that people who were looking for a non-executive power model were talking about d’Hondt . . . it made sense to say d’Hondt.” 57 Certainly, the UUP proposed d’Hondt as the method to allocate posts in a committee system, upholding its opposition to executive power sharing. As Godson notes, however, the Ulster Unionist delegation at the talks leading to the Agreement realized how important executive power sharing was for the SDLP and that cabinet government would provide for more eff ective, centralized decision making. Given that the Northern Ireland Act 1998 confers considerable autonomy on individual ministries, Godson notes that the UUP “thus traded off their traditional politi cal stance for largely illusory administrative gains.” 58 Th e politi cal actors involved in the negotiations cite several reasons for agreeing to executive design under d’Hondt. First, they supported the use of d’Hondt to guarantee proportionality. Second, the procedure allowed for the inclusive nature of the executive and was supported by the pro- agreement parties, provided that Sinn Féin dealt with the decommissioning issue. Th ird, the automatic nature of d’Hondt limited executive formation transaction costs, avoiding inter- party wrangling over portfolio allocation. Th e automaticity of the method also spared the parties from having an assembly vote on executive formation, which arguably would have been politi cally impossible. Fourth, as the parties were familiar with the method from its usage in the Euro pe an Parliament, it appealed as a tried and tested procedure.59 Although some elements of institutional design under the GFA were the outcome of a pro cess of peace initiatives going back several de cades, the inter-party talks leading to the GFA also arrived at features of institutional design as the outcome of bargaining between parties. Th is bargaining took place between the UUP and the SDLP over the details of the power- sharing structures under Strand 1. A central issue of contention between the two parties related to whether the assembly would have limited administrative capacity or whether there could be agreement on full executive authority. Th e UUP was initially of the view that the assembly should set up executive committees rather than departments led by ministers, thereby “refl ecting the party’s historic opposition to institutionalised power sharing.” 60 An equal 74 Chapter 3 part in taking executive decisions was, however, crucial for nationalists. Th e SDLP had long campaigned for full executive power sharing in order to secure nationalists’ equal participation in government. What transpired was a compromise between the two parties. Following the resolution of diffi culties relating to Strand 2, it appears the UUP came around to accepting the SDLP model. Godson reports that Reg Empey, who was leading the Strand 1 negotiations for the Unionists, saw benefi ts in full executive power sharing and problems with a committee structure: “How can you get any coherence of decision making? You have got to have some way of centralising decision making and [making] trade- off s, rather than the proposals for local government which are rather like a hollow polo mint.” 61 Moreover, Godson writes that “Trimble appears to have been swayed in favour of a more maximalist form of government by po liti cal as well as administrative considerations,” including the need to give the SDLP something.62 Indeed, there were reports of John Hume becoming emotional upon hearing the news from Trimble that the UUP was ready to support a power-sharing executive. Th e SDLP was keen that the executive would have two positions at the center—what became the fi rst minister and deputy fi rst minister—to cement the binational nature of the politi cal system. Notably, the fi rst minister and deputy fi rst minister are joint heads of the executive with the same governmental authority. Th e institution of joint premiers was a creative way to accommodate the two groups: Th e majority community would be entitled to the position of fi rst minister, and the minority community had secured a top post of equal power.63 Th e SDLP was also keen to have voting procedures in the assembly based on parallel consent regarding important decisions. Again, it appears that the UUP came to see the merits of these arrangements. In a memo to Trimble on 5 April 1998, Empey wrote, “Perhaps going the ‘full Monty’ on the collegiate fi rst/deputy Secretaryship has advantages over a hybrid approach. People don’t understand parallel consensus (do you?). But they do understand ‘partnership government.’ Having a Unionist veto over the Assembly as a whole, including a negotiating position with any outside bodies, is not the worst thing that could happen. S[inn] F[éin] fears a unionist veto more than anything else.” 64 Th us, the UUP and SDLP agreed on full executive power sharing headed by a fi rst minister and deputy fi rst minister (with no diff erence between these two posts) and a form of parallel consent in the assembly. Th ough London and Dublin had made important contributions in sub- mitting proposals and convening talks for many years, the positive relations Good Friday Agreement 75 between British Prime Minister Tony Blair and Irish Taoiseach Bertie Ahern provided an impetus for achieving a deal by the late 1990s. Th ere has been some suggestion that Blair and Ahern and their respective negotiating teams employed certain tactics to encourage the parties to reach a deal. Re- calling the approach of the British government, Blair’s then–Chief of Staff Jonathan Powell recalls that they “wanted a dynamic approach to rush the parties into an agreement without allowing them to stop and think too much.” 65 Powell contends that Blair’s threat in January 1998 “that if the parties did not agree then the governments would put their proposals to a referendum North and South in May had been a useful way of galvanising the parties and it was a threat we kept coming back to even as late as the end of 2006, as a way of forcing progress. It was our perennial Plan B.” 66 Discussed below and in the next chapter, the two governments frequently used this strategy of publishing their own proposals as a means to make progress. Somewhat worryingly, Powell also comments that they sought to “surprise the parties into agreement.” 67 Although we might take planned “surprise” tactics with a pinch of salt, it is nevertheless clear that the two governments were ultimately successful in facilitating an agreement between the parties, thanks to their determination to reach a deal. As discussed below, the Brit- ish government played an important role in the operation of the Agreement, though on occasion it was criticized for badly judged decisions and incon- sistence in dealing with the respective parties. So what does this case teach us regarding institutional design in deeply divided places? Horowitz is right to direct us to the process that determines whether prescribed institutions will be adopted in deeply divided places.68 In terms of the coherence or hybridity of the institutional framework under the GFA, it is clear that the Agreement establishes a fairly coherent consociation. Referring back to Horowitz’s arguments, it is notable that this coherence came about as the result of an evolutionary process of institutional design and the outcome of strategic interaction among politi cal actors. Th e case points to the potential of institutional designers to arrive at an agreement with complementary elements. Institutional design in deeply divided places need not be planned, with actors adhering to a preordained blueprint. Certainly, the experience also points to the pattern of preferences on the part of the majority and minority communities that helps shape the nature of multi-party bargaining. For the unionist majority community, it was important to achieve reassurances relating to the consent principle (i.e., North- ern Ireland’s constitutional status within the UK would not change without 76 Chapter 3 majority support). Although self- government for the region was more than attractive to Unionists in terms of addressing the demo cratic defi cit of direct rule from Westminster, accep tance of full executive power sharing with nationalists was essentially a compromise with the SDLP. For the minority community, the institutional design pro cess as part of the peace pro cess was about securing guarantees: a guaranteed part in government and an institutionalized “Irish dimension.” Moreover, it is clear that the preferences of the bargainers or institutional designers narrowed over the years. Th e institutional options likely of mustering suffi cient cross- community support (power sharing + an “Irish dimension” + the consent principle) had evolved from a drawn- out pro cess, with proposals made during the confl ict with these key elements proving suffi ciently “sticky” to become the backbone of the GFA. Th e agreement reached between the parties on 10 April 1998 was subsequently endorsed at referendums in Northern Ireland and the Republic of Ireland.69 Th e referendums paved the way for the 1998 Assembly elections, with the results something more of a mixed bag. Out of 108 members, 42 nationalists (24 seats for SDLP and 18 for Sinn Féin) and 8 “others” (6 from the Alliance Party and 2 from the Women’s Co ali tion) were willing to support the Agreement. Th e 58 unionists, however, were narrowly divided. Th irty MLAs were members of parties who endorsed the Agreement— UUP (28 seats) and PUP (2 seats)—while 28 unionist MLAs were from parties who rejected it: DUP (20 seats), UKUP (5 seats), and dissident Ulster Unionists (3 seats). Having achieved an agreement that was suffi ciently supported by both communities, the parties were now faced with the challenge of operating the institutions they had heralded as promising politi cal stability.

Power- Sharing Practice: Limited Cooperation?

Before assessing the empirical evidence for the eff ects of institutional rules on elite cooperation, it is worth discussing the scholarly debates related to whether executive design under the GFA promised compromise and accommodation among the communal parties. A major criticism concerns the lack of opposition provided by a grand co ali tion. As Wilford notes, “the lack of a formal— let alone loyal!— opposition underlines the uniqueness of the Northern Ireland case.”70 Horowitz poses the question: “If all parties are in government and there is no longer opposition in the house, is opposition then to be institutionalised inside the cabinet?”71 He warns that these issues Good Friday Agreement 77 will “undoubtedly haunt cabinets constituted by the agreed methods” but suggests that two small steps were taken to counter the severity of these problems of ministerial accountability and inadequate opposition: “(1) an oath of offi ce . . . so that a minister could not blithely follow any rejectionist path that might be pursued by his or her own party; and (2) a fi rst minister and deputy fi rst minister elected by the house rather than merely nominated by other parties.” Th ere have also been some proposals for executive formation on the basis of support from 65 percent of MLAs present and voting, thereby removing the automatic right to be in government.72 Th e coali tion would necessarily be cross-communal to secure assembly ratifi cation, and parties would have to be conciliatory toward the other bloc in order to garner electoral support and attract coali tion partners. Similarly, Wolff suggests that scrapping d’Hondt would “greatly enhance the accountability of ministers to the As- sembly and also to the electorate, and it would mean that a sense of collective responsibility within the executive would be strengthened. It would also increase opportunities for smaller cross-community parties and at the same time create an eff ective opposition in the Assembly.”73 McGarry and O’Leary defend the use of d’Hondt on the basis that it is proportional, sequential, nonexclusionary, and incentivized parties to take their places in the executive.74 Interestingly, O’Leary recognizes that “the stipulation that Lijphart has sometimes made that consociation requires a grand coali tion of all the politi cal leaders of all signifi cant segments in a region or state creates diffi culties.”75 Refi nements have been made to consociational theory, distinguishing between complete, concurrent, and weak consociation based on the extent of inclusion.76 Rather than requiring a “grand coali tion,” O’Leary emphasizes the importance of “meaningful cross-community executive power sharing” whereby the main groups have at least plurality levels of support of their respective community. On the issue of opposition, O’Leary writes, “To the degree that [the consociation] is complete, opposition will indeed be weak; to the degree that it is either concurrent or weak, opposition will be more vigorously evident.”77 Following the Assembly elections in June and the election of First Min- ister (Designate) David Trimble and deputy First Minister (Designate) Sea- mus Mallon on 1 July 1998, the executive was not formed until 29 November 1999, almost eighteen months later.78 Despite the Assembly being up and running in shadow mode and the belief on the part of the two governments and the SDLP that this would signal the rapid creation of a shadow executive, Trimble refused to allow the operation of the d’Hondt procedure, thereby 78 Chapter 3 precipitating the fi rst of many crises in devolving power to the region. As Dean Godson notes, “the UUP leader’s desperately narrow margins meant that he was better able to insist that any transition period would now have to take place in a manner and at a pace much closer to his own needs; deadline aft er deadline was missed.”79 Elite cooperation within the executive was far from forthcoming. Indeed, the governing coali tion suff ered from a number of diffi culties and setbacks. First, the DUP adopted a semi-detached or “half in, half out” position vis-à- vis the executive, refusing to attend executive meetings and voting against executive policy including the Programme for Government. Th e DUP ministers felt entirely entitled to take this position and claim the Agreement af- forded them the opportunity to take their ministerial seats while remaining one step removed from any sense of cabinet coali tion government. Senior DUP politician Nigel Dodds says that because the Agreement did not require the executive to operate on the basis of collective responsibility and every minister would have full executive authority within his or her department, there was no need to attend the executive, and they were therefore simply doing what was available to them.80 But the non-attendance of the DUP at executive meetings was a politi cally diffi cult concern for the UUP. Th e potential damage done to the UUP in justifying their presence in government with Sinn Féin was certainly not lost on the DUP, as expressed by Dodds, who accused Trimble of “clasping the hand of Martin McGuinness and strik- ing out against fellow unionists.” 81 Th e DUP’s semi-detached position was bolstered by the rotation of its two ministers and opposition to executive decisions in the Assembly. On the matter of ministerial rotation, Trimble and Mallon denounced the DUP tactics: “How anyone can believe that rotating Ministers regularly will not harm the delivery of ser vices is a mystery.” Th ey also accused the party of abusing public interests for “narrow politi cal purposes.”82 Th e DUP’s Peter Robinson argued, however, that his party would ensure “no diminution in terms of the service provided to the people or the quality of those services” and that they were determined to “advance the cause” of their electorate and ultimately defeat David Trimble at the 2001 Westminster elections.83 Second, there were considerable tensions between the two unionist parties and Sinn Féin, largely over the slow progress of IRA decommissioning. Trimble was adamant that executive formation would not take place before decommissioning. On 4 July 2000, the assembly debated the fi rst exclusion motion proposed by DUP leader Ian Paisley, who argued that the IRA’s Good Friday Agreement 79

“continuing engagement in murder and other acts of violence, and its links to Sinn Féin, mean that Sinn Féin cannot enjoy the confi dence of this House.”84 Th e motion failed to garner support from the SDLP on the cross- community vote.85 A further attempt was made on 6 March 2002, but again the SDLP declined to support the DUP move. Indeed, intra-executive rows over decommissioning resulted in the suspension of the institutions on several occasions. In October 2000, the UUP leader took the step of banning Sinn Féin ministers from attending the North- South Ministerial Council as a sanction measure over the failure of the IRA to engage with the International Inde- pendent Commission on Decommissioning. Sinn Féin launched a legal challenge to the ban, and on 30 January 2001 a High Court judge ruled the banning was unlawful. Trimble nevertheless continued the ban and wrote to Sinn Féin ministers that in the absence of IRA decommissioning, they contravened Lord Justice Kerr’s judgment that “it would be open to the fi rst minister to conclude that a potential nominee was unsuitable for nomination because he had not made appropriate eff orts to implement the Agree- ment.”86 Trimble later lost his court appeal. Such instances demonstrate the degree of tension over IRA decommissioning. Following a somewhat ardu- ous pro cess, the IRA fi nally began the pro cess of putting its arms beyond use in October 2001. Th ird, it is well known that relations between the UUP First Minister David Trimble and SDLP deputy First Minister Seamus Mallon were extremely diffi cult on a personal as well as po liti cal level (relations within the Offi ce of First Minister and deputy First Minister [OFMDFM] improved somewhat aft er Mark Durkan replaced Mallon). Th e OFMDFM had been designed to display a mea sure of “jointery” at the center, with the two joint premiers working side by side. It has been well reported that relations between Trimble and Mallon were so strained that they had offi ces at opposite ends of the corridor and chose to travel separately.87 Martina Purdy writes that Trimble and Mallon “quickly discovered they had diff erent ideas about the nature of their politi cal marriage, their roles, and precisely what was meant by ‘equality.’ ”88 According to the SDLP’s Denis Haughey, “Trimble was inclined to behave as if he was Prime Minister and Seamus was John Prescott. Seamus was not having that. For Seamus, that was the big issue that superseded all others. We took the view that the ministers act in concert or not at all.”89 And although the Trimble/Durkan relationship proved to be less problematic than Trimble/Mallon, it was not without its diffi culties. Further evidence of the failure of what Trimble hoped would become a “voluntary 80 Chapter 3 co ali tion within a compulsory co ali tion,” 90 the UUP/SDLP axis was further thwarted by the tensions between the respective junior ministers, particularly between Dermot Nesbitt and Denis Haughey. In the context of these diffi cult inter- party relations, executive meetings were increasingly intermittent, with a lack of business coming forward from executive departments. Fourth, on the policy front, there were several bitter inter-party rows over ministerial decisions, particularly Sinn Féin Health Minister Bairbre de Brun’s decision on maternity ser vices and Sinn Féin Education Minis- ter Martin McGuinness’s announcement of the abolition of post- primary academic selection. Th ese episodes have been described as instances of ministers operating their “ministerial fi efdoms” or “solo runs.” As Wilson and Wilford write, “Any lingering assumption that there would be a joint sense of ‘own ership’ of the Executive was swift ly dismissed as ministers took up residence in their respective ‘fi efdoms,’ there to exercise their relative autonomy.” 91 Regarding McGuinness’s decision, Kenneth Bloomfi eld writes that “What was so appallingly wrong and democratically objectionable about the pro cess was that a single Minister could take and announce such a decision without any approval from Executive colleagues or the Assembly at large. Th is was the act of the commissar rather than a minister.” 92 Many of the politicians interviewed for this research agree that the 1998 Agreement and the Act bestow considerable autonomy on the ministers, with serious implications for government. Th e GFA does, of course, contain a Pledge of Offi ce and a Code of Conduct designed to encourage cooperation between ministers and introduce a measure of accountability.93 During the suspension period (2002– 7), the DUP made much of the argument that ministerial autonomy needed to be addressed. For instance, Nigel Dodds (DUP) maintained that under the GFA, ministers had a “free rein” and called for improved accountability in the future:

What we would be concerned about in any future administration is that you can’t have a situation where it’s basically every man or woman for themselves in government, there’s got to be accountability and we believe accountability should be exercised through the Assembly and if the majority of the Assembly, by whatever safeguards and mechanisms, have a partic u lar view to want something to happen, then that should be taken on board, not ridden rough- shod over, as it was in the old system.94 Good Friday Agreement 81

Th ere are, however, important arguments that suggest that these ministerial decisions were justifi able, in the sense that ministers must have executive autonomy. Commenting on de Brún’s decision to site maternity ser vices at the Royal Victoria Hospital (in the minister’s own constituency) rather than Belfast City Hospital, the SDLP’s Denis Haughey defends ministers’ right to take decisions as a crucial element of power sharing:

We would have upheld her right to take the decision even if our spokesperson on Health, Joe Hendron, reckoned she took the wrong decision. But our view is, was then, and is now that if executive power sharing is to be meaningful, then ministers must exercise power, must take decisions, and must have that responsibility and that right. Otherwise, the whole thing is a farce.95

To what extent, then, can we say that cooperation existed among the politi cal elites in the power- sharing executive? Some commentators take a rather pessimistic view. For instance, Bloomfi eld writes that the system under the GFA “was patently unstable and incoherent. It staggered into re- peated brief or prolonged suspensions. Any sense of common purpose was lacking. If the ‘moderate parties’ could not demonstrate some sense of common purpose and growing understanding, one could hardly expect the most bitter of po liti cal enemies to work constructively together.” 96 It is certainly true that power sharing suff ered from the acrimony surrounding the suspensions over IRA decommissioning. And the willingness of the parties to work together was far from evident on many occasions. How- ever, a more optimistic view would point to instances of joint decision making as evident of progress and cooperation. First, it was no mean feat that the parties negotiated and agreed on a Programme for Government. Other policy developments included free public transport for the el der ly, investment in students, the decision to appoint a children’s commissioner, the publication of a new regional development strategy, the launch of the Review of Public Ad- ministration, and the Reinvestment and Reform Initiative. Th e Needs and Ef- fectiveness Evaluations were seen as “an important part of the backdrop” to the Programme for Government and budget, initially relating to fi nancial assistance for industry and health and social care, later extended to evaluations of vocational education and training, education, and housing. It is also interesting to consider whether ministries worked together to resolve policy issues, as a marker of joined- up government. A visible instance 82 Chapter 3 of joined-up government was the executive’s response to the outbreak of the foot and mouth crisis (in spring 2001), managed by the SDLP Agriculture Minister Bríd Rodgers in cooperation with other departments. A cross- departmental strategy was coordinated via the Department of Agriculture and OFMDFM, with additional staff resources from other departments pledged to assist if required. Advice was provided by the Department of Health with regard to public health protection from the disposal of aff ected animals, and the Department of the Environment advised the executive on the risk of environmental pollution in relation to animal carcass disposal.97 Commenting on the executive’s response, Bríd Rodgers says that the four governing parties “normally at loggerheads with one another” were “working in perfect harmony.” 98 Her department had good relations with the Agricul- ture Committee, the Assembly, and the other departments on the issue. She contrasts the situation in Northern Ireland, where the executive was working as a team, with that in London, where they were at “sixes and sevens.” Rodg- ers maintains the executive’s response worked so well because there was a “common interest” and suggests this was an example of John Hume’s mantra that once parties “begin to work the common ground together, we would fi nd that we have more in common than divides us.” Although it was, of course, a sign of po liti cal progress that the executive was able to reach agreement on policy, it is not unreasonable to suggest that more could have been done, particularly in relation to social inclusion and community relations. Given the diffi culties with the executive, would a more moderate middle co ali tion have produced more cross- community cooperation? Tensions between the UUP and the SDLP within OFMDFM would suggest otherwise. Th e diffi culty here is that the GFA would not have been possible without including Sinn Féin. Arguably, without the inclusion of republicans in the peace process throughout the 1990s, and ultimately in the executive, the IRA would not have been incentivized to (albeit slowly) decommission their weapons and may even have continued their campaign of violence.

British-Irish Efforts: Incentivizing Cooperation?

Par tic u lar attention must be paid to the British and Irish governments to determine whether they helped promote cooperation among the governing parties. McGarry and O’Leary write that “the fi rm policy postures and in- Good Friday Agreement 83 ducements by the two governments were productive and confi rm the positive role that outside parties can play in facilitating power- sharing agreements.” 99 Certainly, the two governments played an important role in persuading and incentivizing the parties to make progress. For instance, London and Dublin sought to manage the gap between the UUP and Sinn Féin over IRA decommissioning and devolution. Although the UUP insisted that decommissioning take place before executive formation (“no guns, no government”), Sinn Féin insisted that such a precondition was outside the terms of the GFA. In the absence of progress by July 1999, Blair and Ahern published Th e Way Forward document, which called on all parties to reaffi rm three principles: an inclusive executive, decommissioning of all paramilitary weapons by May 2000, and decommissioning to be carried out in a manner determined by the International Commission on Decommissioning.100 Executive formation was to take place by 15 July. Decommissioning progress would be reported in September and December 1999 and May 2000. Beyond power sharing under Strand 1, Strand 2 of the GFA created the North-South Ministerial Council, designed to incentivize nationalists by meeting their preference for all-Ireland institutions. Strand 3, based on the British-Irish Council (BIC), would bring together representatives of the two governments and UK devolved institutions as well as representatives of the Isle of Man and the Channel Islands. Republicans, too, were incentivized to support power sharing. Sinn Féin appeared keen to do a deal, provided it could be sold to the republican community. In the run up to negotiations, Blair reportedly told Sinn Féin that “he would not be a persuader for a united Ireland but did want to create a situation in Northern Ireland that was fair.”101 In response to Sinn Féin’s “justice and equality agenda” and “parity of esteem,” the two governments gave assurances of progress relating to the equality agenda and prisoner releases. Blair even made a “private oral assurance” that the British government would bring forward prisoner release from two years to one.102 Even though the two governments sought to maintain their leverage in the implementation of the GFA, the British government took several coercive mea sures that were rejected by the parties. A key example of Downing Street’s misjudgments and incoherence arose with the Secretary of State for Northern Ireland Peter Mandelson’s suspension order in 2000. Recall- ing his decision to suspend the devolved administration, Mandelson refers to his “nuclear option” threat to Sinn Féin that in the absence of IRA decommissioning, he would “reinstate direct rule from London and would 84 Chapter 3 lay the responsibility for it on the IRA and Sinn Féin.”103 Clearly Mandelson had opted for suspension to prevent the collapse of power sharing should Trimble resign in the absence of IRA decommissioning. Th ough Mandelson concluded there was “no alternative to suspension if I was to retain the trust of unionists,” the decision provoked considerable protest from nationalists as well as discomfort for the Irish government.104 Ahern commented, “Each day [the institutions] are suspended is a day in which further damage is being done to the well-being of the agreement.”105 Th e Irish Taoiseach reminded that the terms of the British- Irish agreement “do not expressly include provision for suspension. In that context, suspension raises issues of concern for the Government and any signifi cant extension of it could make the situation more diffi cult.” Sinn Féin held the view that suspension had precipitated a huge crisis “because a government has unilaterally and illegally broken a commitment.”106 In the period following suspension, Sinn Féin’s Martin Mc- Guinness claimed that the British government had not “the foggiest notion” of how to get the process back on track and accused Mandelson of “fl ounder- ing.”107 Th e SDLP was also critical, warning that Mandelson had prioritized Trimble’s concerns and that suspension had “fundamentally weakened” the likelihood of IRA decommissioning.108 Other examples of the British government’s incoherence included the furor over the May 2000 Police Bill designed to implement the recommendations made by the Patten Commission.109 For Sinn Féin, the Police Bill “repre- sents a major departure” from the recommendations of the Patten Report,110 and the SDLP reiterated their “disappointment at the failure to implement unambiguously” some of the Patten recommendations.111 By August 2001, the SDLP was prepared to support the police ser vice having achieved “94 important gains” on police reform.112 Sinn Féin, however, continued to criticize the British government for reneging on its commitments. Th is controversy led to ongoing deadlock over policing, demilitarization, and decommissioning until the IRA “acts of completion” in 2005 and Sinn Féin’s support for the police in early 2007. Yet another instance of the British government’s misjudgments (and policies out of step with Dublin) was the postponement of the Northern Ireland assembly elections in 2003, viewed by Ahern as creating “more problems than it solves.”113 Th e executive faced mounting diffi culties throughout its period in offi ce. Th e ongoing decommissioning issue was a particularly controversial problem, and already fraught unionist- republican relations increased in the wake of the Colombia Th ree episode in August 2001 and the removal of Royal Ulster Good Friday Agreement 85

Constabulary Special Branch fi les from Castlereagh police barracks in March 2002.114 Th e divisions within unionism and Trimble’s delicate position as UUP leader led to the Ulster Unionist Council decision on 21 Sep- tember 2002 to threaten the withdrawal of UUP ministers in mid-January. By October 2002, the unveiling of an alleged republican spy ring at Stor- mont took over, with police raids on homes of republicans and Sinn Féin party offi ces at Stormont. On 8 October, the DUP issued a statement that the party’s two ministers were set to resign, and Trimble issued an ultimatum to Blair that unless Sinn Féin was expelled from the executive, the UUP ministers would also resign.115 Th e institutions were duly suspended on 14 October for the fourth time in less than three years.116 Powell recalls that the British government “had struggled for four years to implement the Good Friday Agreement, by giving a few concessions to one side and then a few to the other in the hope that we could build trust between the sides over time.”117 But he suggests that “the peace pro cess had become badly discredited and morally undermined.” Direct rule from Westminster was restored until the parties agreed to share power in 2007.

Summary

Th e Good Friday Agreement heralded a new departure for politics in North- ern Ireland. In contrast to Sunningdale, the GFA meant that devolved government would be fully inclusive and promised an end to violence. Th e two communities, led by their respective po liti cal parties, supported the agreement for diff erent reasons. For nationalists, the agreement meant a place in government and the all-important “Irish dimension” linking Northern Ire- land with the Republic of Ireland. For pro-Agreement republicans, the deal promised institutions that could be used as the vehicle to a united Ireland in the future and “parity of esteem” in the present. For pro-Agreement unionists, Northern Ireland’s constitutional position within the UK was further cemented, given the accep tance of the “consent principle” and the changes to the Irish Constitution. Power sharing was the result of an evolutionary process from the 1970s, with some further details (full executive power sharing and “jointery” between the fi rst minister and deputy fi rst minister) negotiated between the UUP and the SDLP in 1998. Th is case points to the need to consider how a pro cess of institutional design involving a multiplicity of actors can nevertheless lead to a fairly coherent institutional framework. 86 Chapter 3

Have the institutional rules under the GFA facilitated cooperation among ethnic elites within the executive? Arguably, the more fully inclusive executive under the GFA was a better option than the exclusive moderate middle under Sunningdale. It seems fair to suggest that one of the reasons why power sharing came about and muddled along (despite its many diffi - culties) was that Sinn Féin was included, the agreement having locked them into the system and facilitated eventual IRA decommissioning. Th e inclusion of republicans was also enabled by using the d’Hondt procedure to share out ministerial seats on the basis of parties’ strength in the assembly. Arguably, a fully inclusive executive would not have been possible if executive formation had been dependent on post-electoral negotiations, such was the ongoing animosity between parties, largely owing to the slow pace of decommissioning. Beyond the impact of institutional rules on cooperation, an additional factor relates to the role of the British and Irish governments. Both governments were engaged in the implementation of the Agreement and in managing the divergent expectations and preferences among the parties. Th e two governments relied on an intergovernmental approach to persuade the parties to make progress. Th is involved delivering concessions to each side and draft - ing proposals for progress. Th e British government also adopted negative, or even coercive, incentives in the form of suspension powers, arguably creating diffi culties for power sharing and threatening the cohesion of the intergovernmental approach. Blair himself acknowledged “there were fudges, things said and done that had little intellectual or politi cal consistency except that of seeing us through each set of obstacles.”118 Th e consistency of British- Irish eff orts would be tested in the ensuing suspension period, fi nally leading to the publication of the St. Andrews Agreement in 2006 and restoration of power sharing in 2007. CHAPTER 4

The 2007– 11 Executive: A New Era in Northern Ireland Politics?

Following suspension of the po liti cal institutions in October 2002, North- ern Ireland returned to direct rule from Westminster, leading the British and Irish governments to refuel their intergovernmentalist approach to restore power sharing. Th e suspension period was dominated by ongoing politi cal controversy over IRA decommissioning, Sinn Féin’s position on policing structures, and the DUP’s attitude to power sharing. Th ese po liti cal challenges endured until March 2007, when the DUP and Sinn Féin, the so- called extremes, fi nally agreed to share power in a co ali tion joined by the UUP and the SDLP. For DUP leader Ian Paisley, the decision to enter power sharing with Sinn Féin refl ected his party’s goal “to see devolution returned in a context where it can make a real, meaningful improvement in the lives of all the people of this part of the United Kingdom.” For Sinn Féin Presi- dent Gerry Adams, the agreement marked “the beginning of a new era of politics on this island.” Th e two governments also applauded the power- sharing deal as the potential culmination of the long-running peace process. Tony Blair declared that “Everything we have done over the last ten years has been a preparation for this moment.” And Bertie Ahern heralded the DUP- Sinn Féin agreement as having “the potential to transform the future of this island.”1 An important question for power sharing is why an Executive was fi nally agreed upon by DUP and Sinn Féin, two parties long considered the extremes and arch enemies in Northern Ireland politics.2 I track the eff orts of the two governments in incentivizing these two parties to share power. Th e St. Andrews Agreement of 2006 essentially involved a trade-off : Th e DUP would have to commit to power sharing, and Sinn Féin would need to 88 Chapter 4 sign up to the reformed policing structures. Th e new peace agreement included some revisions to the GFA, an instance of institutional change that allowed a hitherto rejectionist party, the DUP, to say that a new agreement had been achieved. Th e inclusion of these two parties in the top two positions is an important test for the fully inclusive versus moderate middle debate in power- sharing theory. Th e experience highlighted the serious challenges to power sharing from cultural/identity politics, but it signaled a degree of politi cal maturity, given the commitment by these two parties to maintain the power- sharing structures.

Incentivizing the Extremes to Share Power

Th e period of suspension lasted four and a half years before the nationalist and unionist parties fi nally agreed to enter a restored Northern Ireland Assembly in 2007. With the fallout from the alleged republican spy ring at Stormont, the ongoing problem of IRA decommissioning, and the strengthened position of the DUP as the leading unionist party, the two governments faced considerable challenges in persuading the parties to support the formation of a new power-sharing executive. In 2003, the British and Irish governments published their Joint Declaration with a number of conditions and incentives for agreement.3 Directed primarily at Sinn Féin and the IRA, the document stressed the need for a “transition from violence to exclusively peaceful means” that should be “brought to an unambiguous and defi nitive conclusion.” Such “acts of completion” would help ensure rewards in the form of the repeal of the British government’s suspension powers under the Northern Ireland Act 2000, security normalization including the “vacation and demo li tion of security installations, the closure of designated military bases,” as well as decreased military presence. Th e two governments also committed to facilitate inter-party dialogue on the devolution of policing and justice. Further aiming to address republican preferences, the document stated that all parties should play their part in the institutions and that the British government would “continue to discharge all its commitments under the Agreement in respect of the Irish language,” including funding for Irish language broadcasting. To incentivize the unionist parties, the Joint Declaration affi rmed the two governments’ position that progress would be conditional on IRA “acts of completion” and that the Review of the Agreement would include discussions on institutional rules under the The 2007– 11 Executive 89

GFA, namely the “question of designation and voting arrangements in the Assembly” as well as the eff ectiveness of the Pledge of Offi ce relating to ministerial responsibilities. Following the Assembly elections held in November 2003, it was clear that the two governments would need to focus their eff orts on the DUP and Sinn Féin, confi rmed as the largest parties of their respective blocs. Because these two parties would occupy the top two posts in the Assembly, many felt there was clearly little point in reconvening to elect a fi rst minister and deputy fi rst minister and restore the institutions.4 It certainly would not have been possible to secure the required cross- community support for the dual premiership because the DUP would not have been prepared to vote for a Sinn Féin deputy fi rst minister.5 Th e parties were, therefore, called to Stormont at the beginning of 2004 to undertake the review of the institutional arrangements provided for in the GFA.6 Th e Review of the Agreement illustrates the two governments’ aim of inducing the parties to share power in the changed politi cal context following the 2003 Assembly elections. In dealing with the DUP’s diffi culties with the GFA, they were careful to stress that the agenda would not stray from the fundamentals of the GFA. Irish Minister for Foreign Aff airs Brian Cowen outlined the principles of the agreement, which included power sharing on an inclusive basis:

Without being prescriptive, the fundamentals of the Agreement would, in our view, include the constitutional principle of consent, partnership government in Northern Ireland on an inclusive basis, the interlocking institutions of the Agreement, including its North/South and East/West dimensions, the entrenchment of human rights and equality for all, the removal of the use and threat of paramilitary violence, no matter what its origin, the normalisation of security arrangements on the ground and the consolidation of the new policing and criminal justice arrangements.7

Yet the DUP favored a diff erent form of power sharing to that set out in the GFA. Th e party’s Devolution Now document is notable due to the option of a co ali tion “formed following negotiations between the parties on what ever basis could attain a Key Vote majority.”8 Th e second option was for fully inclusive power sharing with the following important stipulation: “To ensure that Ministers were accountable, legislative change would be required to ensure that ultimate power would rest with the Executive as opposed to 90 Chapter 4 individual departments. In addition, decisions of the Executive could be challenged by the Assembly and would require a Key Vote of the Assembly to stand.” 9 Th e DUP maintained that this model would only be possible if the IRA met the “acts of completion”—termed the “Blair necessities”—on decommissioning and ended all paramilitary activity and criminality. If an Executive could not be formed, the DUP suggested, powers could be transferred to the Assembly in a corporate Assembly model whereby the Assembly would be responsible for the departments’ functions via key vote approval. Sinn Féin welcomed the DUP’s proposals relating to inclusive power sharing but remained suspicious of the detail. Martin McGuinness told the Irish Times that he saw a potential booby- trap in the DUP proposal for power to reside in the Executive as a whole rather than individual members: “Mr. McGuin- ness feared this was a ruse whereby unionist ministers, by having a majority of six against four in the cabinet, could eff ectively exercise unionist control.”10 McGuinness also said his party opposed the DUP proposal whereby 70 percent of the Assembly could pass key motions, which could eff ectively exclude Sinn Féin. Sinn Féin’s objectives for the Review were listed in the Agenda for Full Implementation of the Agreement, which centered on support for the GFA, the stability of the institutions, equality and human rights, and the expansion of north- south cooperation.11 Adams said his party would not accept the Agreement being “tweaked, twiddled or subverted.”12 Th e UUP did not advance proposals to the Review in relation to institutional reform. Speaking at the Foyle Unionist AGM, David Trimble argued there was little point in discussing possible changes to Executive formation and voting arrangements in the absence of decommissioning: “What is the point of reviewing the rules and regulations if the Assembly is never going to sit because paramilitaries will not wind up their criminal activities? Frankly it is as if the Northern Ireland Offi ce wishes to rearrange the deckchairs on the Titanic aft er it has hit the iceberg.”13 For Trimble, decommissioning was the principal issue to be resolved: “Th e underlying problem in the politi cal pro cess is the same as ever— the failure of paramilitaries, most notably the IRA, to engage in acts of completion.”14 Unsurprisingly, given its commitment to the GFA, the SDLP was opposed to what might be perceived as renegotiation and remained fully committed to inclusive power sharing. Th e party called for the devolution of policing and justice and argued that the DUP should not have a veto on progress: “Th e government also have to hold The 2007– 11 Executive 91 fi rm against the DUP’s half-baked proposals to turn the institutions of the agreement into something like Belfast City Council, and their craving for majority rule cannot be indulged.”15 Writing in the Irish Times on 2 Febru- ary 2004, SDLP politician Séan Farren responded to the DUP’s call for increased ministerial accountability: “But ministers clearly were accountable, every bit as much as ministers in Dublin or London are. Th eir policies were questioned, debated, and scrutinised closely. All their legislation had to be passed by the Assembly.”16 Following a break over the summer months, the two governments and the parties convened at Leeds Castle in Kent, En gland, in September 2004. In addition to the focus on decommissioning, the talks centered on reform of the Assembly and the Executive. Th e DUP sought tighter control on ministerial authority, while Sinn Féin resisted any such move.17 Aft er a few days of intense negotiations, the discussions failed to secure a deal given the stalemate over institutional arrangements. Th e Irish Times reported that there were two remaining institutional issues for the DUP: fi rst, the election of fi rst minister and deputy fi rst minister and, second, ministerial accountability.18 Th e paper reported that Alliance leader David Ford had proposed a potential compromise to elect the entire Executive but that the issue over ministerial accountability was hugely problematic, especially given the ap- prehension on the part of Sinn Féin and the SDLP that the DUP would be able to veto decisions of nationalist ministers as a “return to crude unionist majoritarianism.” As on previous occasions (and as would take place again during the St. Andrews negotiations in 2006), the two governments decided to take the matter into their own hands by publishing their proposals for the formation of a power-sharing executive. On 8 December 2004, the British and Irish governments published the joint Comprehensive Agreement, a set of proposals that were described by Taoiseach Bertie Ahern as “a hugely impressive, indeed a landmark, package” and “a dramatic surge toward fi nal closure.”19 Th is proposed “dramatic surge” is arguably characteristic of the two governments’ strategy, described by Jonathan Powell as the policy to “rush the parties into agreement.”20 Despite the failure to secure a deal between the DUP and Sinn Féin, reportedly over the need for photographs of IRA decommissioning, the two governments argued there had been considerable progress on ending paramilitary activity, decommissioning, and power sharing.21 De- spite the lack of agreement, it is interesting to consider parties’ responses to 92 Chapter 4 the proposed changes to Executive formation and ministerial accountability in the document, arguably an attempt to refi ne the 1998 consociational framework with important amendments. Wilford and Wilson suggest that the Comprehensive Agreement proposals were “the application of fi rst aid to the consociational formula arrived at in 1998.”22 Th e document states that proposals had been sent to the DUP and Sinn Féin and that “despite intensive eff orts over a number of months and very considerable progress, not all elements were agreed.”23 Th e document represented the two governments’ “compromise” package, including a number of annexes with anticipated statements from the DUP and the IRA.24 Th e relevant sections are set out in the Comprehensive Agreement Annex B, which deals with Executive formation and ministerial accountability. Th e fi rst proposal was for a statutory Ministerial Code requiring an amendment to the Northern Ireland Act 1998, so that ministers would have a duty to act in accordance with the provisions for ministerial accountability, notwithstanding their executive authority in their areas of responsibility as defi ned in the GFA. Th is proposal was clearly designed to address the DUP’s insistence that ministerial decisions would need approval of the Ex- ecutive as a whole. Th e Code—which would require cross-community support in the Assembly—would include “arrangements to ensure that, where a decision of the Executive could not be achieved by consensus and a vote was required, any three members of the Executive could require it to be taken on a cross-community basis.” What appears to have been an “informal” arrangement during devolution, the proposed amendment would give statutory eff ect to the provision for three ministers to trigger a cross- community vote. Another signifi cant proposal was for the Assembly to refer ministerial decisions to the Executive for review. Th e proposal suggested thirty MLAs could initiate a referral within seven days of a ministerial decision, subject to the presiding offi cer deeming it an issue of public importance. Th e Execu- tive would then consider the issue within seven days. It is important, however, that the proposal stipulates, “Only matters covered by the Ministerial Code . . . would require a collective decision by the Executive.”25 It would, therefore, seem likely that such an amendment in a future Executive would allow a de Brún– type maternity decision to be referred by the Assembly back to the Executive. But it is important that the Executive would simply have to “consider” the issue, a second referral could not be made on the same The 2007– 11 Executive 93 matter, and only matters covered by the Ministerial Code would require a collective decision. Th e document also proposed that the 1998 Act be amended to include a requirement in the Pledge of Offi ce that ministers would “participate fully in the Executive and NSMC/BIC and would observe the joint nature of the offi ce of First Minister and deputy First Minister.”26 Th is proposal was arguably designed to address nationalist criticisms that unionists felt they could cherry-pick their participation in the institutions. It sought to prevent the situation in the previous administration whereby the DUP boycotted the NSMC and was blocked from attending the British- Irish Council. DUP adviser Richard Bullick describes the proposed amendments to ministerial accountability as an important achievement for his party: “We were satisfi ed that no decisions could be taken by a minister which couldn’t at least have the capacity of requiring cross-community support. So we certainly took the view that there was no possibility that the maternity-type decision could arise again, which was obviously very important from our perspective, having made such an issue out of it and in the end, I think, Sinn Féin were certainly prepared to live with that.”27 Bullick asserts that the proposals “squared the circle” by enabling the DUP to maintain that they amounted to a new agreement while simultane- ously allowing Sinn Féin to claim that they were merely a modifi cation to the 1998 accord. He suggests that “if that was the nature of the debate, then nobody was going to lose too much out of it, which obviously was the case right up until the fi nal steps couldn’t be taken in relation to decommissioning in December.”28 Alliance claimed it had recommended that the Ministe- rial Code be placed on a statutory basis, but it argued that “the specifi cs of this remain ill-defi ned, and further negotiation among the parties may not achieve agreement.”29 However, Stephen Farry (Alliance) maintains that under the Comprehensive Agreement the “minutiae of things are improved” in relation to accountability because ministers would not be able to “go off and make decisions unilaterally.”30 Of further note is that the document also proposed a change to the rules governing the election of the fi rst minister and deputy fi rst minister from a cross-community vote in the Assembly to a straightforward nomination without a legitimizing vote. Th is proposed change to the GFA was advocated by the DUP, which preferred nomination by designation rather than election of the two top posts in the Assembly.31 Th is proposal constituted an important change in the St. Andrews Agreement of 94 Chapter 4

2006, ultimately allowing the DUP to enter a power-sharing executive without having to vote for a Sinn Féin deputy fi rst minister.

The St. Andrews Agreement: Power Sharing and Policing

Despite the two governments’ eff orts with the Joint Declaration of May 2003 and talks at Leeds Castle in September 2004 leading to the publication of the Comprehensive Agreement Proposals in December, progress on power sharing remained elusive. Th e potential for movement, however, increased with IRA “acts of completion” in 2005.32 But with little sign of the two largest parties agreeing to share power, the two governments concluded there needed to be a deadline to put pressure on the parties, particularly the DUP. Powell notes that they needed to convince the DUP they would “go ahead without them.”33 He recalls the British government’s plan: “We would establish an interim Assembly with a fi rm end date in the autumn and run d’Hondt in November aft er a clean IMC report. If that failed to result in a government, we would suspend the Assembly, cancel elections, and move to a Plan B of greater Anglo-Irish cooperation.”34 Th is was evidence of the Brit- ish government’s negative incentives (and threat of potential coercive measures) to the parties: “Agree to share power and form a government or face the potential consequences involving joint authority.” Around the same time, Northern Ireland Secretary of State Peter Hain announced some policy changes: the end of the eleven- plus selection exam from primary to secondary education, the reorga ni za tion of local government to reduce the number of local councils, and a move toward the introduction of water charges. Powell perhaps overstates the impact of these announcements in making clear to the DUP “that a devolved government might be the only way of avoiding a series of policies that they found distasteful and that their electorate hated. It was a very eff ective form of po liti cal blackmail.”35 By 2006, the potential for progress on power sharing came to be recognized as a trade-off between the two largest parties: Th e DUP would have to agree to share power with Sinn Féin and the latter would have to accept the legitimacy of the Police Ser vice of Northern Ireland (PSNI). 36 Although the British government may not have made this an explicit precondition for republicans, Sinn Féin’s support for the PSNI was certainly crucial for the DUP. Powell writes that in the autumn of 2006, “In a major move, Sinn Féin indicated that they would join the policing board in December if the DUP The 2007– 11 Executive 95 agreed to set up the Executive, and if they gave them a timetable for the devolution of policing and justice.”37 Th e latter was considered, according to Powell, to be Sinn Féin’s “talismanic demand on policing.” In order to get a deal, the two governments persisted in their eff orts to bring about a trade- off over policing and power sharing. Th e two governments made their next big push at the negotiations held at St. Andrews, Scotland in October 2006, where considerable diffi culties arose between the DUP and Sinn Féin over sequencing. For Sinn Féin, the party’s special conference (ard fheis) on policing would take place aft er the formation of an Executive, while the DUP was adamant that power sharing could not take place without real progress on policing. Due to pressures within the DUP, the party came to the view that there needed to be a period of testing republican support for the PSNI before devolution. Mary Alice Clancy writes that the outcome of the St. Andrews talks was, “like all other, ‘hot house’ negotiations, an agreement between the two governments to market St. Andrews as ‘an agreement,’ in the hope that this would put pressure upon Sinn Féin and the DUP to agree to do a deal.”38 Th e publication of the “agreement” refl ects the tendency of the two governments to suggest that the basis of a deal had been secured and that it was up to the parties to sign on the dotted line or risk being blamed for ongoing stalemate. O’Duff y writes that the “deal between the proximate antagonists was mainly a by- product of reinforced British- Irish intergovernmentalism.”39 In order to keep the process moving, the document produced by the two governments set out an element of sequencing: Th e parties were asked to confi rm their responses by 10 November; the fi rst minister and deputy fi rst minister would be nominated on 24 November; and power would devolve on 26 March 2007. Aft er considerable pressure from the two governments over the Christmas period, Sinn Féin called its ard fh eis on policing in January with a motion to support “civic policing through a police service which is representative of the community it serves, free from partisan politi cal control, and democratically accountable.” 40 With 90 percent of delegates in support of the motion, Adams declared the policy change another step toward the republican objective of achieving Irish unity: “You have created the opportunity to signifi cantly advance our struggle and you have seized the opportunity to further our primary objective of a united Ireland through the building of greater po liti cal strength.” 41 Yet Sinn Féin’s support for the PSNI was not the sole issue for the DUP in agreeing to share power. Th e changed security context following the IRA’s 96 Chapter 4 acts of completion in 2005 provided greater potential for the party to consider and justify power sharing with republicans.42 With regard to the institutions, the DUP was also determined to achieve changes to the GFA relating to ministerial authority. Th e party was adamant that the St. Andrews legislation include a provision that ministers are obliged to agree to all mea sures with the full Executive. Unsurprisingly, this change was opposed by Sinn Féin because it would potentially limit their ministers’ freedom to maneuver in taking executive decisions. In the end, an amendment was made to the Northern Ireland Act (1998) providing for a Ministerial Code to be agreed by the Assembly, subject to cross-community support. Th e Northern Ireland (St. Andrews) Bill stated that under such a code, ministerial decisions should be reached by consensus wherever possible or a vote should be taken.43 It also included the provision that if a decision could not be reached by consensus and a vote was required, any three members of the Executive could require it to be taken on a cross-community basis.44 Th is section of the Bill includes a provision for the referral of ministerial decisions back to the Executive— if thirty members petition the Assembly that a ministerial decision has been taken in contravention of the Ministerial Code or “relates to a matter of public importance.” Amendments were also made to the Pledge of Offi ce under the 1998 Act: Ministers were obliged to participate fully in the Executive, NSMC, and BIC; observe the joint nature of the offi ces of the fi rst minister and deputy fi rst minister; and to uphold the rule of law, including support for policing and the courts, as set out in the St. Andrews Agreement. Important changes were also made to procedures for the selection of fi rst minister and deputy fi rst minister (Section 16 of the 1998 Act). Th e new rules meant that the largest party of the largest po liti cal designation would nominate a member of the Assembly as fi rst minister, and the largest party of the second- largest po liti cal designation would nominate the deputy fi rst minister. In keeping with DUP preferences, the top two posts would now be fi lled via nomination rather than cross-community ratifi cation in the Assembly. Coakley notes the “remarkable development” that was “largely unnoticed and unreported” when the legislation was given eff ect at Westminster.45 Late 2006 and early 2007 saw ongoing speculation and deliberation on whether Sinn Féin would endorse the PSNI and whether the DUP would signal its readiness to enter a power-sharing executive. In the absence of agreement, Hain threatened, “It is either a functioning executive with powers devolved on 26 March or it’s dissolution and the politicians stop getting paid, Stormont closes down, and we pack up Northern Ireland politics, The 2007– 11 Executive 97 maybe for years to come.” 46 Once power sharing was agreed in March 2007, the DUP cited the threat of the two governments’ Plan B as crucial for their decision to reach a deal. In his fi rst interview aft er agreeing to power sharing with Sinn Féin, Paisley asserted, “We were told that if we didn’t do this then it was going to be curtains for our country. . . . How would I have faced my people if I had allowed this country to have the union destroyed and the setting up of a joint government by the south of Ireland?” 47 It is perhaps unsurprising the British government attested to the eff ectiveness of such threats. Powell contends that the threat of greater engagement by the two governments had, since 1998, “been a useful way of galvanising the parties and it was a threat we kept coming back to even as late as the end of 2006, as a way of forcing progress. It was our perennial Plan B.” 48

DUP- Sinn Féin: Cooperative Power Sharing?

Th e Northern Ireland (St. Andrews) Act 2006 provided for a transitional Assembly to prepare for the restoration of the devolved institutions. MLAs from the 2003 Assembly election convened at this new Assembly, which came into eff ect on 22 November 2006. On 24 November, nominations for the top two posts took place, with DUP leader Ian Paisley refusing to accept Sinn Féin’s nomination for him as fi rst minister alongside Martin McGuin- ness as deputy fi rst minister. Referring to the “lack of delivery by the republican movement” on support for the PSNI, courts, and the rule of law, as well as “a complete end to paramilitary and criminal activity and the removal of terrorist structures,” Paisley suggested that “as Sinn Fein is not yet ready to take the decisive step forward on policing, the DUP is not required to commit to any aspect of power sharing in advance of such certainty.” 49 Th ough the transitional Assembly lasted just a few months (up until its dissolution on 30 January 2007), it included a Programme for Government Committee with six subgroups designed to develop priorities for a restored power- sharing executive. Th ese priorities included the economy, the Review of Public Administration, schools admissions policy, the devolution of policing and justice powers, and water charges.50 Following Assembly elections in March 2007 and the agreement between the DUP and Sinn Féin to share power, the parties allocated ministerial positions on 2 April, in advance of devolution taking place on 8 May. At the fi rst sitting of the new Assembly, Paisley and McGuinness affi rmed the terms of the Pledge of Offi ce and took 98 Chapter 4 up their positions, followed by the formal nominations of ministerial portfolios under the d’Hondt mechanism.51 Unsurprisingly, the two largest parties have argued that their decision to share power has delivered great gains for Northern Ireland in a spirit of reconciliation. Media coverage has made much of the seemingly positive chem- istry between Paisley and McGuinness, dubbed the “Chuckle Brothers,” given the frequency they were seen smiling and laughing in each other’s company. McGuinness referred to the “transformed politi cal situation” and the “cordial and civilised atmosphere” between the two men.52 It might be more appropriate, if less imaginative, to suggest that both parties were committed to making the institutions work. Executive operation was not, of course, plain sailing. Indeed, given the historical (and even contemporary) animosity of these two parties, it is unsurprising the new Executive would encounter diffi culty over policy issues. Th e Executive faced several policy challenges, including schools admissions policy (the future of academic selection), Irish language legislation, the development of the Maze prison, and a Northern Ireland Bill of Rights. Th e issue of Irish language legislation refl ected the inter- ethnic tensions between the unionist and nationalist parties, an instance of “cultural contestation” in power sharing.53 Sinn Féin had hailed the commitment to an Irish Language Act in the St. Andrews agreement as a victory for the nationalist community. Under the section Human Rights, Equality, Victims and Other Issues, the British government made commitments to the Irish language and Ulster- Scots:

Th e Government will introduce an Irish Language Act refl ecting on the experience of Wales and Ireland and work with the incoming Ex- ecutive to enhance and protect the development of the Irish language. Th e Government fi rmly believes in the need to enhance and develop the Ulster Scots language, heritage, and culture and will support the incoming Executive in taking this forward.54

Yet unionist parties remained opposed to Irish language legislation and maintained they were not involved in any deal Sinn Féin may have made with Blair. Th e DUP’s Nelson McCausland remarked that the British government “seems intent on rewarding and encouraging the intransigence of republicans.” 55 The 2007– 11 Executive 99

In March 2007, the British government unveiled proposals for an Irish language bill and a further twelve-week consultation period, with clauses pitched as a “middle ground approach” (plans included the appointment of an Irish language commissioner, establishment of an Irish language scheme for public bodies, possible use of Irish in court proceedings, and statutory forms available in Irish).56 Th ese developments appear to have aggravated tensions within the Executive. In July 2007, a meeting took place between Gerry Adams and the DUP’s Minister for Culture, Arts, and Leisure, Edwin Poots. Briefi ng the media, Adams stated, “Th ere will be an Irish language Act, and it will be a strong Irish language Act. Th e only question, of course, is when we will have it.” 57 For Poots, however, the DUP did not negotiate the Irish language Act at St. Andrews, and there would be a need “to achieve cross-community consensus” on any such legislation.58 Poots had a warning for republicans: “We need people to be refl ective, and if individuals think they can use this as a politi cal weapon and ram it down other people’s throats, then that is not going to be an eff ective means of doing this.” Th e lack of progress on the Irish language issue led Sinn Féin to criticize openly their Executive partners. Sinn Féin MEP Bairbre de Brún accused the Depart- ment for Culture, Arts, and Leisure of having “missed the opportunity to reach out to Irish language speakers and to present linguistic and cultural diversity in an inclusive and positive way.” 59 Contention over Irish language legislation was one issue, in addition to the stalemate over policing and justice, that contributed to Executive diffi - culties over the summer of 2008. Sinn Féin refused to agree to Executive meetings, and there was even a fear that Adams was about to collapse the Assembly in the absence of movement on policing and justice and creation of Irish language legislation. At Hillsborough Castle in February 2010, in talks over policing and justice powers, the British government committed £20 million for the development of the Irish language. Adams announced that some of this investment would be allocated to the Irish Language Broadcast Fund, which had been “denied funding” by the DUP minister.60 In response to the additional resources committed by British Prime Minis- ter Gordon Brown, the DUP Culture Minister, Nelson McCausland, welcomed the investment as a boost for the local economy and announced that there would also be an Ulster-Scots Broadcasting Fund. At the time of writing, the issue of Irish language legislation remains unresolved. Yet the De- partment for Culture states the Executive has a duty “to adopt a strategy 100 Chapter 4 setting out how it proposes to enhance and protect the development of the Irish language and Ulster- Scots culture, heritage, and language” and promised to publish a Strategy for Indigenous or Regional Minority Languages.61 Another source of “cultural contestation” within the Executive concerned the redevelopment of the Maze prison site outside Lisburn, County Antrim. Sinn Féin sought to protect the buildings that had previously housed po liti cal prisoners, including the prison walls, one of the H-blocks, watch- towers, and the prison hospital where ten IRA and Irish National Liberation Army (INLA) hunger strikers died in 1981. Republicans advocated the transformation of the former prison buildings into an International Center for Confl ict Transformation. In the suspension period, a Maze Consultation Panel was set up, chaired by the UUP’s David Campbell and the SDLP’s Mi- chael McKernan. With the site covering 360 acres, one of the main proposals involved a new Northern Ireland sports stadium, an option supported by the Gaelic Athletic Association, Ulster Rugby, and the Irish Football Asso- ciation. For unionists, however, Sinn Féin’s proposals would create a “terror museum” and “memorial to the hunger strikers.” Hence some unionist politicians favored the construction of a new sports stadium in Belfast because of the proposed museum at the Maze. But Sinn Féin remained adamant on preserving the prison and building a museum. As put by Sinn Féin’s Paul Butler, “We repeatedly are on record as saying it should not be a shrine to any of the hunger strikers or any individual. Th e history of republicanism has to be told, their history, the history of imprisonment.” 62 As a decision on the required funds necessitated cross- community support, Sinn Féin made it clear that they would oppose plans for the stadium to be built anywhere other than the Maze site. Th e DUP’s Peter Robinson said the decision would depend on the available business case; he added, “Th ere’s a very strong opinion within the unionist community that they don’t want to have anything on that site which would glorify terrorism.” 63 Deputy First Minister Martin Mc- Guinness replied to these fears by stating that his party “was not arguing for any kind of shrine. If we want a confl ict transformation centre, then it has to concentrate on how we resolve confl ict.” 64 Following years of po liti cal wrangling, the Maze stadium plans were dropped in January 2009. In addition to inter- ethnic disagreement, there also appears to have been some disagreement within the DUP, with Edwin Poots (whose constituency included the site) in favor of the stadium at the Maze and some party colleagues favoring Belfast, largely due to the creation of the museum as part of the development plans. Following discussions be- The 2007– 11 Executive 101 tween the DUP and Sinn Féin, First Minister Peter Robinson and deputy First Minister Martin McGuinness issued a statement announcing that an agreement had been reached to submit an EU funding application for a Peace Building and Confl ict Resolution facility on the site of the former prison.65 Whereas McGuinness spoke of the center’s potential to be “a world-class facility of international importance designed to strengthen our peace-building expertise and to share our experiences with others throughout the world,” Robinson emphasized the potential benefi ts in developing the site to “revive our economic output in these diffi cult times.” Certainly, the Maze prison issue illustrates the diffi culties ethnic parties in power- sharing executives face in managing symbolic represen ta tions of the confl ict. In this instance, the two parties (or at least the joint premiers) appear to have reached a deal—one that can be claimed by Sinn Féin as a victory and one that can be promoted as economic opportunity by the DUP. Despite these challenges, there were important achievements made during this period, most signifi cantly the (eventual) agreement between the DUP and Sinn Féin over the devolution of policing and justice powers from Westminster. Th is agreement did not, however, come about without considerable contention and negotiation over the form of a justice department and the timing of devolution. In the period following the restoration of power sharing in May 2007, the DUP continued to insist that there was insuffi cient confi dence within the unionist community. Over the summer of 2008, ongoing negotiations over the devolution of these powers meant that other policy issues were put on the back burner. Th ere was even speculation that a frustrated Gerry Adams was prepared to “pull the plug” on the Assembly in the absence of progress.66 In November 2008, the Offi ce of the First Minister and deputy First Minister informed the Assembly and Executive Review Committee of an agreement on the structures for the devolution of policing and justice powers.67 A justice minister would require majority support in the Assembly on a cross-community basis (a majority of designated unionists and a majority of designated nationalists). Th e two parties had also signaled that they would not nominate candidates themselves. Th e ongoing stalemate, however, led to renewed eff orts by the two governments to press the parties to do a deal. Following talks at Hillsborough Castle in January 2010 chaired by British Prime Minister Gordon Brown and Irish Taoiseach Brian Cowen, the two parties fi nally agreed on the timing, with 12 April set as the target date for devolution. Th e parties also agreed to set up a working group on reforming the regulation of parades and committed to improving the 102 Chapter 4 functioning of the Executive.68 Following failed nominations by the UUP and SDLP, Naomi Long of the Alliance Party nominated her party leader, David Ford, as justice minister, a nomination that had, of course, been agreed previously between the DUP and Sinn Féin. Some measure of progress can be seen by the degree of cooperation within the Executive in the parties’ shared condemnation of violence carried out by dissident republicans, with Martin McGuinness describing the perpetrators as the “traitors” of Ireland.69 In response to the rioting in interface areas of Belfast in July 2011, McGuin- ness and Robinson issued a statement confi rming their “resolute” commitment to tackle sectarianism and described the situation as “outright thuggery and vandalism.”70 Th e reform of local government structures was an issue that initially promised some intra- Executive cooperation but fell victim to disagreement among the parties on how to proceed. Launched by the former Executive in 2002 to improve the effi ciency, eff ectiveness, and accountability of the delivery of public ser vices, proposals for the Review of Public Administration were developed by the Secretary of State for Northern Ireland Peter Hain during the suspension period. Proposals were put forward in November 2005, to include a reduction of twenty- six district councils to seven, criticized by the local parties as “super councils” that would lose touch with their geo graph i cal identity. Once the proposals were taken up by the new devolved administration, the Minister for the Environment, Arlene Foster (DUP), signaled they would be reviewed.71 Following inter-party agreement on eleven rather than seven councils, the Local Government (Boundaries) Act (Northern Ireland) 2008 gave eff ect to this decision. However, by May 2010 it had become clear that the creation of the new councils would be delayed, due to intra- Executive disagreement over the boundaries for district councils. In June the reform was abandoned following the failure of the Executive to obtain cross-community support. Th e SDLP, Sinn Féin, and Alliance supported the eleven- council model while the DUP voted against and the UUP abstained.72 Th e issue of a community relations strategy was also subject to delays and disagreements between the two largest parties. During the suspension period, the Offi ce of First Minister and deputy First Minister, headed by Minis- ter of State for Northern Ireland John Speller, launched a Good Relations Policy and Strategic Framework, to tackle sectarianism and create a more shared society. With the return of devolution, however, the DUP and Sinn Féin were unable to agree on how to advance the community relations strategy. Some progress was made in the context of discussions over policing and The 2007– 11 Executive 103 justice, with the publication of a strategy a crucial requirement of the Alli- ance party, whose leader, David Ford, became justice minister in April. In July 2010, the fi rst minister and deputy fi rst minister launched the consultation on their Programme for Cohesion, Sharing, and Integration, which promised to tackle the “physical and community division created by inter- faces” between the two communities. But with ongoing criticism of the document as “lacking in vision,” little progress was made in advancing the policy for the remainder of the Executive.73 Th ese challenges were heightened by the rioting in interface areas of Belfast during the July 2011 marching season. Divisions between the parties also continued over the issue of academic selection for the transfer of pupils from primary to secondary education.74 Th e Executive was clearly fairly modest in its achievements. On the co alition’s capacity to drive a clear agenda, Wilford refers to the “dearth of policy initiatives and something of a legislative famine for much of the period since 2007.”75 Th e lack of Executive meetings for several months in 2008 further hampered the Executive’s capacity to deliver. Th e smaller coali tion partners, the UUP and SDLP, felt somewhat excluded, with SDLP leader Margaret Ritchie criticizing the Executive’s functioning as a DUP–Sinn Féin “carve- up.”76 Yet the Executive was notable in achieving some mea sure of cooperation, allowing for the maintenance of the power- sharing system. Moreover, despite the challenges discussed above, there was less intra- Executive tension. As Wilford notes, despite the “undoubted and unresolved tensions among the four major parties, the vastly changed context” meant they evaded “the same intensity of in-fi ghting that hobbled its prede ces sor.” 77 He suggests that notwithstanding intra-Executive strains, the Executive “has found its organisational feet and has adopted a more mature, even trusting, approach to the management of its business.”78

Summary

Despite the challenges, the 2007– 11 Executive was notable for the maintenance of the power- sharing system and the improved relationship between the parties. Th e relative success of the Executive was helped by the commitment of the two main parties to making the institutions work. Th e issue of inclusion is important here: With these two parties having cemented their positions as the largest parties of their respective blocs, their engagement in and support for power sharing were crucial. Th is support, particularly on 104 Chapter 4 the part of the DUP, came about as a result of the changed security and political context since suspension in 2002.79 It is also clear that changes to the institutional rules governing the selection of fi rst minister and deputy fi rst minister, as well as to the rules on ministerial authority, allowed the DUP to point to revision of the GFA. Th e nomination of the top two posts (rather than being subject to cross- community ratifi cation in the Assembly) was, unsurprisingly, accepted by the DUP and Sinn Féin, but it was criticized by the other parties whose approval is not required. Arguably, the two governments’ incentives structure played an important role in restoring and helping maintain power sharing. External incentives continued to be important, helping to bring about power sharing via the St. Andrews Agreement, helping to manage sequencing leading to the agreement between the DUP and Sinn Féin to share power, and ensuring the operation of the Executive. Positive incentives (in the form of side payments) to the two largest parties were also important carrots in helping to ease reluctance to do a deal. Sinn Féin was able to secure the British government’s commitment to the Irish language, devolution of policing and justice, ongoing demilitarization, repeal of the British government’s suspension powers, and amnesty for “on the runs” (those wanted for terrorist off ences). DUP diffi culties were eased with IRA “acts of completion,” revisions to the institutions, republican support for the PSNI, and the promise of a generous fi nancial package. Given the apparent institutional lock-in of the system, there has arguably been some internalization and even habitualization of power sharing as norm-conforming behavior. On his fi nal day at the Assembly, Paisley declared that the institutions are “here to stay” and that despite communal divisions, “on the basis of democracy we can both work for the good of our own people and the country.”80 Arguably the relative success of power sharing in Northern Ireland has been the outcome of multilateral bargaining among the internal and external actors. On the basis of this bargaining over rewards (decommissioning, devolution of policing and justice, and all- Ireland institutions), compliance with external incentives for power sharing has ultimately helped parties secure a negotiated compromise over the constitutional issue. CHAPTER 5

Power- Sharing Stalemate in Post- Dayton Bosnia

Power sharing in Bosnia has been fraught with diffi culty. Fift een years since the Dayton Peace Agreement (DPA), the state remains far from functional, the extent of inter-ethnic cooperation is limited, and the international community continues to be heavily engaged. In this chapter, I analyze the fi rst ten years of power sharing in Bosnia following the negotiations at Dayton, Ohio in 1995. Th e less than positive experience of power sharing can be traced back to the context of the peace talks and the imposition of a fairly complex political system lacking in legitimacy. Th e practice of power sharing in the initial post-Dayton years points to the limitations of institutional rules in fostering cross-community cooperation when parties pursue politi cal preferences that regard institutions at the state level as weak and irrelevant. Th e role of external actors is crucial here in imposing a peace settlement, exercising robust executive powers in the implementation of the DPA, and incentivizing po liti cal parties to compromise on policy reform. Yet the eff ectiveness of these incentives is questionable, given that local elites were able to avoid inter- ethnic compromise in the knowledge that diffi cult decisions would be imposed anyway. Th e intervention of successive high representatives was increasingly problematic, ultimately leading to the announcement in 2006 that the Offi ce of the High Representative (OHR) would close, optimistically and prematurely signaling greater own ership on the part of local po liti cal elites.

Negotiating the Dayton Peace Agreement

Much attention has already been paid to the steps leading to the end of the Bosnian war and the peace talks in Dayton.1 Here I survey the incentives structure employed by the U.S. and EU negotiators in arriving at a deal at 108 Chapter 5

Dayton. It is well known that the external actors pressured the parties to accept an agreement, including power sharing, as the compromise to end the war.2 U.S. negotiator Richard Holbrooke recalls the objectives at Dayton: “to turn the sixty-day cease-fi re into a permanent peace” and “gain agreement for a multiethnic state.”3 At the talks, U.S. Secretary of State Warren Chris- topher set out conditions for a settlement: that Bosnia would remain a single state and that a deal would take into account “the special history and signifi cance” of Sarajevo; respect for human rights, including holding those responsible for atrocities to account; and the resolution of eastern Slavonia.4 Although the talks were essentially about ending the war (with less focus on governance than the division of territory), the compromise would be settled on the basis of “war time structures” with “two functioning levels of government” between Sarajevo and two entities, a Bosniak-Croat Federation (resulting from the Washington Agreement 1994) and a “Bosnian Serb entity with no claims to sovereignty.” 5 Th ough joined by Eu ro pe an and Rus sian delegations, the U.S. was the lead actor pushing for a settlement at Dayton. Recollections of the talks show that relations between the delegations were sometimes strained, given their diff erent emphases on the military versus civilian aspects of an agreement.6 Moreover, as EU negotiator Carl Bildt admitted, there was “no single Eu ro- pean negotiator or representative,” and his role was to coordinate with the German, French, and British delegations.7 Despite this multiplicity of external actors, the agreed approach was that “nothing is agreed until everything is agreed.” Th e incentive structure was based on both rewards for an agreement and threats of negative consequences in the absence of a deal. When Croat President Franjo Tudjman stalled over agreement on the Federation, the U.S. team reassured him that they would insist to Bosnian President Alija Izetbegović that “one of the top three posts in the central government be assigned to a Croat.”8 When faced with in- fi ghting in the Bosniak delegation, Holbrooke spelled out the achievements gained by the group, including a promised World Bank package following agreement.9 Christopher also warned that in the absence of their support for a settlement at Dayton, President Clin- ton would withdraw his support.10 Th e U.S. also threatened to end the talks unless the Bosniak delegation accepted Milošević’s proposal to put the contested city of Brčko under arbitration.11 Similar pressure was put on Milošević whereby, in the absence of a deal, the U.S. could “shut down without an agreement,” sanctions would continue, and confl ict could resume.12 Stalemate in Bosnia 109

Th e politi cal arrangements of postwar Bosnia were part and parcel of the overall deal. Of note, however, is Bildt’s frustration over the slow progress on the constitution, given the focus on territory. His preference was for “a state which could be accepted by all the parties, and in which the inter- entity boundary line on the map would gradually be of less and less importance over time.”13 On the details of the politi cal system, Bildt recalls that a “mass of diff erent draft s on constitutional issues were circulating, with positions oft en changing rapidly from one meeting to another.”14 He even suggests that potentially important constitutional issues “were handled with a nonchalance” that he considered “irresponsible.” Th at the external actors arrived at proposals for executive power sharing is unsurprising, given that they were working within the confi nes of the postwar po liti cal reality: the existence of the two entities and the desire of the groups to enjoy equal rights at the state level. Moreover, power sharing was clearly a feasible option with some credibility, given the groups’ familiarity with proportional repre sen tation under the Yugoslav administration, cultural autonomy, and decentralization, this historical legacy stretching back to Ottoman times.15 For the delegations representing the three groups at the talks, the territorial details were of greater importance than the institutional rules set out in the draft constitution. In response to external pressure, the three groups sought to “preserve the maximum degree of autonomy possible within the confi nes of a single state.”16 Th e Bosniak delegation sought a single state, with the entity structure ultimately accepted as a diffi cult compromise. Having held out on agreement over Brčko, they fi nally accepted a solution following pressure from the U.S. For the Croat delegation, the incentives for doing a deal included regaining eastern Slavonia from the Serbs, cementing the Federation Agreement, and securing equal repre sen ta tion in the state- level institutions. For the Bosnian Serbs represented by Milošević, the onus was on avoiding sanctions and retaining the 51– 49 percent split in territory between the Fed- eration and the Republika Srpska (RS).17 For all three groups, important incentives included not just an end to war but also the establishment of po liti cal institutions with a reward of resources and patronage for their respective supporters. Despite these incentives for each group, the nature of the Day- ton negotiations led to the “widespread feeling of imposition” rather than a process based on “local ownership.” 18 Th is is unsurprising given the belief by the external actors that “No-one thought it wise to submit the constitution to any sort of parliamentary or other similar proceeding. It was to be a 110 Chapter 5 constitution by international decree.”19 Arguably, this failure to secure legitimacy for the po liti cal structures has since haunted the practice of power sharing in Bosnia. As Ivo Daalder notes, because the agreement “did not resolve the underlying diff erences among the parties— namely, whether Bosnia should be de facto partitioned or reintegrated— implementation of- fered the parties a way to continue their basic confl ict.”20 Recognizing the territorial gains during the war, the Bosnian constitution under Annex 4 of the DPA confi rms the existence of the two entities: Republika Srpska and the Federation of Bosnia and Herzegovina.21 Th e Bos- nian state is notably complex; there are not just two entities with extensive power, but also cantons and municipal authorities within the Federation. Bosnia corresponds to a plurinational federation, albeit one “held together by international force” since the DPA.22 Th at much of the new state and its complex administrative structure were new heralded “an institution- building challenge on a daunting scale.”23 Article III of the constitution sets out the division of competences between the state institutions and the two entities.24 As is well known, the state level has been fairly weak, initially en- dowed with responsibility for a limited number of areas, including foreign policy, customs, refugee policy, international and inter- entity criminal law enforcement, and air traffi c control. All competences not assigned to the state institutions were assigned to the two entities. Th is division of competences has proven to be a source of considerable contention for the operation of the po liti cal structures. Th e local parties, particularly Bosnian Serbs and Bosnian Croats, sought to cement their power base in the RS and Federa- tion, respectively. Yet serious eff orts to shift the balance of power in the opposite direction—by transferring competences from the entities to the state level—have been undertaken by international actors. Th e transfer of competences to the state level is favored by the Bosniak community, mostly accepted by Bosnian Croats (provided their group retains equality with the other two), but opposed fi ercely by the Bosnian Serbs. In terms of the initial post- Dayton period, it is not hard to see why po liti cal elites preferred to focus their eff orts within the entities. Florian Bieber comments that the “vague division of competences between the layers of government and the fi scal dominance of the entities reduced any incentive in terms of commitment to the state.”25 Th is tension between the powers enjoyed by the entities and the international community’s eff orts to strengthen the state has proved to be an enduring challenge of the Bosnian power-sharing system. Impor- Stalemate in Bosnia 111 tantly, the dominance of the entities as the locus of power has been used to explain parties’ tendency to block decision making at the state level and the hardening of ethnic division. Bieber and Keil comment that the form of decentralization in Bosnia has “enforced ethnic dominance in the po liti cal system and facilitated obstruction of the po liti cal system.”26 Th ey suggest that “radical decentralization has prevented Bosnia from becoming a functional state” and has hampered the implementation of the DPA. Of note is their proposed lesson from the Bosnian case that the “problematic use of ethnic federalism” through a “combination of territorial and national autonomy” can reinforce ethnic separation and leave the state vulnerable to secessionist demands.27 It is recognized that plurinational federations are far from being a panacea for managing ethnic division. As McGarry and O’Leary write, “Federalism provides a territorially defi ned and concentrated nationality or ethnicity with politi cal and bureaucratic resources that it can use to launch a bid for inde pen dence.” 28 Yet it is also arguable that eff orts in the other direction— to centralize plurinational federations— may lead units to question the viability of the state and consider going alone. Moreover, it is unlikely that negotiators at Dayton could have arrived at any other structural outcome, particularly given the Washington Agreement of 1994 that provided for the Federation entity. At the state level, the parliamentary system is bicameral, composed of the House of Peoples and the House of Representatives. Because the cham- bers were provided with the same powers, there have been subsequent calls for reform to streamline their respective roles. Th e House of Peoples has fi ft een delegates, ten from the Federation (fi ve Bosniaks and fi ve Croats elected by the Federation House of Peoples) and fi ve from RS (elected by the RS National Assembly). Th e House of Representatives has forty-two members, two-thirds from the Federation and one-third from the RS, and members are directly elected from their respective entity. Of note is that delegates representing the three groups have recourse to veto powers in the operation of the parliamentary assembly. Under Article IV of the constitution, a proposed decision “may be declared to be destructive of a vital interest of the Bosniak, Croat, or Serb people” by a majority of a group’s delegates, and approval in the House of Peoples will require a majority of each of the three groups’ delegates present and voting. If a majority of any of the three groups objects, a Joint Commission will be set up, comprising a delegate from each group; if the Commission fails to resolve the issue within 112 Chapter 5 fi ve days, the matter will be referred to the Constitutional Court. Looking back to the negotiations on the constitution at Dayton, Bildt recalls the need for a “safety device” for the groups and claims that the Euro pe an delegation proposed a “mechanism for one constituent people to block a decision deemed contrary to its vital interests” and that a constitutional court should “determine whether the vital national interests of a population group were really involved.” He suggests that although this institutional rule “was certainly not perfect,” it was “the least bad of options.”29 Indeed, veto powers (within the parliamentary assembly and the operation of the presidency) have been an issue of considerable controversy in post-Dayton Bosnia. As discussed below, it is oft en argued that veto rules readily allow groups to obstruct decision making by using the “vital national interests” procedure (VNI). In addition to the VNI, decisions in the state parliament require that majority support include at least one- third of delegates from each entity. Th us, in addition to the “vital interests” veto, the three groups have an entity veto in parliament. Veto powers were extended to the entities’ constitutions, whereby two-thirds of a constituent people’s delegates can veto legislation in their respective assembly. Not specifi ed in the Bosnia constitution, the entity constitutions list areas of legislation subject to the VNI veto but also include a provision whereby two- thirds of a group’s delegates can trigger the veto on any issue.30 As head of state, the presidency comprises three members: one Bosniak and one Croat (directly elected from the Federation) and one Serb (directly elected from the RS). Annex 4 provides for a presidency chair rotating among the three members, and the presidency is to ensure consensus in decision making (decisions can, however, be adopted by two members when eff orts to achieve consensus have failed). Th e presidency is tasked to handle foreign policy, reporting on the state budget, executing decisions by parliament, and performing any other role assigned to it by parliament. Crucial for power sharing, each member has veto power in order to protect the “vital interests” of his or her entity. A member can declare a decision in violation of a vital interest within three days of its adoption. According to the member’s entity, the decision is then referred to either the RS National Assembly or the House of Peoples in the Federation. If the declaration is confi rmed by a two-thirds vote of a group’s delegates within ten days of the referral, the decision will be annulled. Th e presidency has a role in executive formation by nominating the Chair of the Council of Ministers. Within the Council of Ministers, no more than two-thirds of ministers are to be appointed from Stalemate in Bosnia 113 the Federation, and deputy ministers are not to be from the same constituent people as the minister. Th e Constitutional Court is also important for power sharing.31 Under Article VI of the Bosnian constitution, the Court has nine members: four selected by the House of Peoples of the Federation, two members selected by the RS National Assembly, and three members (not citizens of Bosnia or any neighboring state) to be selected by the President of the Euro pe an Court of Human Rights in consultation with the state presidency. Th e responsibility of the Constitutional Court is to uphold the constitution; it has jurisdiction to decide any dispute between the entities, between the Bosnian state authorities and any entity, or between the state institutions. Th e Constitutional Court has also played an important role relating to the use of the VNI. Fur- thermore, it set in motion a pro cess of constitutional reform following a signifi cant ruling in 2000, ultimately leading to revisions of the entity constitutions, as discussed below. Important for the operation of the power-sharing system, and a central focus for this study, is the role accorded to external actors. David Chandler notes that the DPA is signifi cant “because of the far-reaching powers given to the international community, which extended well beyond military matters to cover the most basic aspects of government and state.”32 Across eleven annexes of the DPA, the international community has considerable power over a range of issues, including the military and policing, refugees and displaced persons, and civilian implementation.33 Th e role of the OHR under Annex 10 is particularly signifi cant. Article II sets out the OHR’s mandate, including monitoring the implementation of the peace settlement and promoting the parties’ full compliance with the civilian aspects. Article V gives the high representative an extensive mandate as the “fi nal authority in theatre regarding interpretation of this Agreement on civilian implementation of the peace settlement.” Th e OHR’s authority was extended in 1997, allowing the High Representative to take executive decisions. Th e extent of the OHR’s intervention in the governance of the country has been the source of considerable debate and even antagonism with local po liti cal elites.

Critiques of the Bosnian Consociational System

Th e DPA is a clear instance of a corporate consociational settlement. Bieber describes the institutional set- up as “an extreme textbook example of 114 Chapter 5

consociationalism.” 34 Weller and Wolff suggest that the DPA might represent “two consociational settlements within the boundaries of a single state— the sovereign consociation of the state of Bosnia and Herzegovina and the regional consociation of the Federation.”35 O’Leary describes Bosnia as a “complex consociation” where, in addition to the four elements of consociation originally set out by Lijphart, the settlement includes an important role for international actors as well as “cross-border or confederal relationships (and sometimes institutions) for national minorities with their kin in other states.36 Th ere is also a tendency to consider the Bosnian constitution as “rigidly” consociational.37 Nina Caspersen suggests that although the DPA is predominantly consociational, it includes some integrative elements based on majority rule in a number of institutions (Constitutional Court, Central Bank, Joint Interim Commission).38 And although the institutional framework is certainly complex, it has also been shown to be amenable to evolution and fl uidity.39 Weller and Wolff write that the gradual evolution of the institutions, “rather than large- scale changes, emphasizes the transformative eff ects of the institutional arrangements agreed at Dayton.” 40 Such transformative eff ects are debatable, however, as suggested by the oft en robust critique that has been levied at the consociational elements of the Bosnian constitution. Much of this criticism is attributed to the array of veto procedures, the predetermination of positions for the three “constituent peoples,” and the tension between the weak institutions at the state level and the extensive autonomy in the entities. In relation to veto rights, Marie-Joëlle Zahar suggests the rules “tend to carry with them the risk of ‘immobilism’ or state pa- ralysis, which can only be countered by the concerted action of responsible leaders willing to compromise in order to maintain stability.” 41 Reviewing the use of vetoes in Bosnia, she argues that “elite intransigence or unwillingness to cooperate are more oft en than not the norm.” 42 Sofi a Sebastián notes that “the many provisions aimed at protecting ethnic interests have . . . slowed down the state-level decision-making process to the point of outright pa- ralysis at times.” 43 Similarly, Bieber refers to the blockage eff ect of an informal veto and suggests that even if veto powers have been used rarely, the threat of their use “pre-empts[s] decisions from being taken which might be vetoed.” 44 Not defi ned in the Bosnia constitution under Annex 4, the VNI veto was defi ned by the OHR in the constitutions of the two entities following the decision on constitutional amendments of 19 April 2002. Th e entity Stalemate in Bosnia 115 constitutions specify the application of the VNI to the right of constituent peoples to be adequately represented on legislative, executive, and judicial bodies; identity issues; constitutional amendments; or ga ni za tion of public authority; constituent peoples’ equal rights in decision making; education; religion; language; culture; territorial orga ni za tion; and the public information system. Th e constituent peoples can also trigger a procedure to deem an issue subject to the VNI rules. Th e Constitutional Court has the authority to determine whether an issue falls within the VNI. Bieber and Keil note that for this reason, “the main source of blockages” has been entity voting rather than the use of the VNI veto.45 Certainly, the record of power sharing in Bosnia suggests that the misuse of veto powers by intransigent political elites has been an obstacle to effi cient decision making. Importantly, an international offi cial concedes that the international community has been trying to reduce the extent of veto powers on the basis that they prevent progress on policy reform.46 An OHR offi cial suggested that the constitution should be “revamped” to limit and defi ne the VNI, which can otherwise be invoked “on absolutely anything.” 47 In addition to the formal powers set out in Annex 4 of the DPA and the entities’ constitutions, elites have exercised a more informal veto by simply not turning up to participate in the institutions, thereby delaying the passage of legislation or frustrating decision making in the Council of Ministers. Elsewhere, I argue that specifi c veto procedures can make a diff erence for cross- community cooperation in power sharing, and a more refi ned veto procedure may off er greater potential for accommodation via specifi cation of the issues groups may be able to veto and the limitation of veto rules.48 As discussed in the next chapter, veto rules have been a contentious issue in the ongoing process of constitutional reform. Th e degree of ethnic repre sen ta tion, whereby positions are predetermined for the three constituent peoples, is also critiqued in the academic literature and by the international community. Reviewing the DPA, Bieber refers to a weakness of consociationalism, “most notably the rigid enforcement of group identity.” 49 Marcus Cox notes that “the role of ethnicity in the constitution is the most controversial aspect of the peace settlement,” whereby “ethnicity is treated as a fi xed category and institutionalised the principal basis of repre sen ta tion.” 50 He suggests that this reliance on ethnic identifi cation leads to several problems.51 First, the specifi cation of par tic u- lar ethnic groups “assumes that the population is separated physically along 116 Chapter 5 ethnic lines.” Second, “there is no legal defi nition of the ethnic groups, and no objective criteria could ever be developed to distinguish among them.” He suggests that this becomes problematic when non- ethnic or multi- ethnic parties “pick and choose under which ethnic list to fi eld candidates for diff erent positions, causing the system to become incoherent.” Overall, Cox writes that the DPA is “strong in protecting the three ethnic groups, but weak on creating the po liti cal and institutional ties to bind them together.” 52 As discussed in the next chapter, it is also important that international actors have called for constitutional reform on the basis that predetermined positions in Bosnian po liti cal structures discriminate against individuals who do not as- cribe to any of the three constituent peoples. Beginning with the Council of Eu rope’s Venice Commission Opinion of 2005, this pro cess has led to the Eu ro pe an Court of Human Rights ruling in December 2009 that the constitutional provisions allowing only members of the three constituent peoples to hold the highest po liti cal offi ces are in contravention of the Eu ro pe an Convention on Human Rights. It is diffi cult, however, for the main po liti cal parties to agree on reform that would alter the predetermination of group repre sen ta tion because they have been protective of their rights under the Dayton framework and, unsurprisingly, resist any external tampering with those group rights.

Early Post- Dayton Difﬁ culties

Th e fi rst post- Dayton high representative, Carl Bildt, claimed that “What is necessary in order to make peace work is to have eff ective and true power sharing between the two entities and the three communities. . . . Power sharing is the essence of the Constitution that is at the core of the Peace Agreement.” 53 Despite Bildt’s call for eff ective power sharing, the early period of DPA implementation is known for a lack of cooperation among the politi cal parties and a high level of intervention by the OHR. Th e ongoing power of the nationalist parties is considered to be the main source of these diffi culties. For instance, Cox writes that the “refusal of the wartime regimes to support the new constitutional order left it weak and unstable.” 54 He suggests that the institutional set- up under the DPA remained “a paper creation” whereby the war time regimes were “unwilling to dismantle their parallel systems and transfer authority to new institutions under democratic control.” 55 Th is line of argument bolsters the idea that the parties had Stalemate in Bosnia 117 little incentive to participate in the state institutions. Because the state level was weak, the institutions, including the Parliamentary Assembly and the Council of Ministers, could be ignored by elites who focused on cementing their power in the entities. In 1996, the fi rst postwar elections at the state level produced victories for the three pre-war nationalist parties: the (Bosniak) Party for Democratic Action (SDA), the Serbian Demo cratic Party (SDS), and the Croatian Demo- cratic Union (HDZ). Having hoped that demo cratic elections would bring to power more moderate politi cal forces, the international community was disappointed by the electoral victories of the three nationalist parties. Ex- ploring this disjuncture between what international actors hoped for and what transpired, Carrie Manning refers to a set of assumptions, including the belief that “once elections are genuinely free and fair and supported by the necessary civil liberties, voter preferences for moderate parties will pre- vail.” 56 Yet the gains of the war time parties are perhaps unsurprising. Man- ning notes the manipulation by the wartime parties, “which ensured that the elections would serve merely to legitimize the results of wartime ethnic cleansing.” 57 She further questions the expectation that voters “would, if given the chance, remove from offi ce those responsible for the war and replace them with forward- looking leaders.” 58 Clearly this belief that postwar elections would see a rejection of the big three parties “underestimates the degree to which the cleavage lines created during the war might have come to refl ect diff erent interests, rather than simply being imposed by bad leaders.” 59 Moreover, it is also arguable that voters acted rationally in supporting the three main parties because they were the gatekeepers of resources. We know that the war provided the three main parties with the opportunity to take control of resources, including utilities and pension funds. In the postwar situation, access to employment in many areas depended “in part upon access to the politi cal interests that control them,” incentivizing people to vote for those they believed could protect them in the postwar setting.60 Th ough the international community may have been disappointed by the 1996 election results, there are clear structural and politi cal explanations for the big three parties’ grip on power. Existing research suggests that the obstruction of the local parties in the early post-Dayton years quickly put an end to any meaningful power sharing. Bieber writes that for the fi rst few years, the state institutions were “largely an empty shell.” 61 Th is is perhaps unsurprising as the constitution established new layers of governance or aimed to transform wartime 118 Chapter 5

institutions into functioning democratic post-confl ict structures. Yet progress in establishing eff ective institutions was unforthcoming. According to the International Crisis Group (ICG), “even minor progress requires mas- sive and sustained international pressure, and obstructionism continues to yield dividends.” 62 On the operation of the state institutions, the ICG refers to the “vetoes, squabbles, and boycotts in the House of Representatives,” which “forced” the HR to impose legislation.63 Cox describes the state power- sharing institutions as “venues for politi cal gamesmanship: boycotts and symbolic disputes to impress hard- line constituencies, followed by concessions to preserve relations with the international community.” 64 Seemingly straightforward issues, including the design of a state fl ag, encountered inter- ethnic disputes, leading the OHR to intervene.65 Th ough the power- sharing institutions did make some policy achievements— Central Bank legislation, a common currency, common license plates, state symbols, and customs reforms— much of the reform was imposed by the OHR. Under the DPA, the Council of Ministers could create new ministries as appropriate, a process that began with the formation of the Ministry for Civil Aff airs and Com- munications. But without entity support, the Council of Ministers had little power and room to maneuver in making policy. Certainly, it seems that there was a lack of investment by the Bosnian Croat and Bosnian Serb parties to embrace the new power- sharing structures. Given the po liti cal structures under Annex 4 and the legacy of the war, it is perhaps unsurprising that the nationalist parties were reluctant to engage in accommodation at the state level. During the initial post- Dayton years, there were clearly strains on peace implementation. As the victors of the 1996 elections, the nationalist parties sought to shore up their respective power bases in the entities and had little incentive to invest in power sharing at the state level. Th e ICG comments that the postwar rule by the big three parties had left the country “poor, dysfunctional, divided, corrupted, unreconstructed, and hopeless.” 66 To explain the diffi culties hampering cooperation, existing research goes beyond elite unwillingness and points to the brakes on eff ective power sharing arising from the Dayton structures. Weller and Wolff write that the institutions “appeared to be constantly unable to function. Elite cooperation simply did not set in for a long time. Excessive executive represen ta tion of all the main communities at all levels led to an in eff ec tive, but tremendously bloated, state apparatus.” 67 Th e ICG refers to “a sobering picture of non-compliance and non- implementation.” 68 On the implementation of Annex 4, the consti- Stalemate in Bosnia 119 tution has been described as “unwieldy, clumsy, unworkable, and bears no relation to the reality of Bosnian politi cal life.” 69 According to the ICG, the power-sharing institutions at the state level were “only as powerful as the entities allow them to be.”70 Indeed, the state institutions depended on fi nancial support from the two entities, giving “the separate ethnic politi cal structures an eff ective right of veto over the development of the state.”71 With little evidence of inter-ethnic cooperation, an increasingly frustrated international community responded by arming the OHR with robust executive powers.

The Ofﬁ ce of the High Representative: Assuming the Bonn Powers

Th e evolving role of the OHR is central to the practice of power sharing in the early post-Dayton period. Th e DPA tasked the OHR with coordinating and facilitating peace implementation. Th e OHR’s role, however, has since evolved considerably beyond facilitation to full executive powers. Chandler notes that the “demo cratic mandate attained by the three main nationalist parties in the elections counted for little against the mandate of the High Representative as the ‘fi nal arbiter’ of the demo cratisation pro cess.”72 Th e victors of the 1996 elections were seen by the international community “not as bearers of a substantial demo cratic mandate but as barriers to democratisation in Bosnia.”73 Cox writes that the initial postwar years were “a frustrating period for the international mission, which found itself babysit- ting new institutions that seemed to be achieving little.”74 With the OHR enjoying fi nal authority to interpret and implement the DPA, the Peace Implementation Council (PIC) took additional mea sures to overcome elite obstruction. At a meeting in Sintra, Portugal on 30 May 1997, the PIC Steer- ing Board implored local elites “to stop blaming each other, or the international community, for the problems they encounter, and to work together constructively and in a spirit of reconciliation for their common good.”75 In relation to the implementation of Annex 4 of the DPA, the PIC stated that the progress made was due to the eff orts of the OHR and called on parties in the power-sharing institutions to “put an end to mutual accusations.” Sev- eral deadlines were set relating to state symbols and citizenship legislation. Other areas covered by the declaration included cooperation on war crimes, economic reform, refugee return, human rights, local elections, and Brčko 120 Chapter 5 supervision. At the PIC meeting in Bonn on 10 December 1997, the PIC established that the OHR could take interim mea sures when local po liti cal parties are unable to agree, as well as actions against politicians/offi cials “in violation” of the DPA.76 Th e so-called Bonn powers have since been a source of considerable contention between the OHR and local elites. At the time, the new High Representative, Carlos Westendorp, welcomed PIC support “for a more robust interpretation of the mandate of the High Representative, in par tic u lar for his demands for a stronger interpretation of Article 5 of Annex 10 of the DPA, which outlines the ‘fi nal authority’ of the High Repre- sentative in the theatre of operations.”77 As Gerald Knaus and Felix Martin note, however, there was no thought given to “substantive or procedural checks on these new powers” that have since been used to impose reform and remove offi cials and politicians from offi ce.78 From the beginning, the Bonn powers were used fairly extensively to drive reform. For instance, on the issue of refugee return, High Representa- tive Westendorp cancelled all permanent occupancy rights in the two entities during and since the war in order to remove “the most substantial injustice in the fi eld of property rights” and pointed to the misallocation of property as an attempt “to legalise war profi teering.”79 Before leaving his post, Westendorp imposed a series of decisions, including the Decision on Restructuring of the Public Broadcasting Service, the Decision on Freedom of Information, the Law on Special Witness Identity Protection in Criminal Proceedings in the Federation, as well as amendments to legislation on the Supreme Court of the Federation and legislation on the Federation Prose- cutors’ Offi ce.80 A few days later, he imposed further decisions relating to identity cards, common insignia and symbols, the telephone network, administrative fees, and the removal of restrictions (previously exercised via the Bonn powers) on SDS Vice- President Dragan Čavić.81 Westendorp’s successor, Wolfgang Petritsch, similarly relied on the Bonn powers to drive reform. Interestingly, he recalls being under pressure to use the powers and says he “had to act as the most interventionist High Representative ever.”82 As discussed below, High Representative Paddy Ashdown eclipsed Petri- tsch’s use of the Bonn powers. Existing research is critical of the OHR’s reliance on the Bonn powers to implement the peace agreement over the heads of elected representatives. Chandler writes that “under the High Representative’s administrative guidance, consensus was obligatory and opposition seen as illegitimate obstruc- Stalemate in Bosnia 121 tion.”83 Certainly, opposition to reform was portrayed as a threat to eff ective state-building. For instance, in July 1998, when the House of Peoples failed to adopt the framework Law on the Privatization of Enterprises and Banks due to opposition from Bosnian Serb representatives, the OHR stated that such “obstruction can only be construed as an attempt to endanger the re- habilitation programme” and “to let discrimination and cronyism dominate over economic reform.”84 Moreover, it was the high representative, and not the Council of Ministers, who set the policy agenda.85 Th e OHR also resorted to the Bonn powers to remove offi cials and representatives judged to be blocking peace implementation. In March 1999, High Representative West- endorp dismissed the president of the RS, Nikola Poplasen, for seeking to unseat RS Prime Minister Milorad Dodik and for obstructing DPA implementation.86 In November 1999, High Representative Wolfgang Petritsch removed twenty-two individuals from offi ce for “pursuing anti-Dayton, anti- peace, anti- reconciliation, and extra- legal agendas” and for fostering the “poison of division.”87 As discussed below in relation to Paddy Ashdown’s tenure as High Representative, and in the next chapter, it is questionable how eff ective this coercive approach using the Bonn powers has been in incentivizing domestic elites to adopt policy reform. As Cox puts it, it has seemed that “statebuilding under the international protectorate was something done to Bosnians rather than by them.”88 According to the ICG, OHR reliance on the goodwill of the local parties “has led to an international community policy of pleading, cajoling, and begging in order to achieve DPA implementation.”89

Moderate Power Sharing: The Alliance for Change Co ali tion

Following elections in November 2000, the international community forged the ten- party Demo cratic Alliance for Change, which took power at the Federation and state levels in January 2001. Th e Alliance became possible following successes by the (largely Bosniak) Party for Bosnia and Herze- govina (SBiH) and the multi-ethnic Social Democratic Party of Bosnia and Herzegovina (SDP) over the main Bosniak party, the Party for Democratic Action (SDA). Th e ICG refers to the “energetic lobbying and arm- twisting” of the U.S. and UK ambassadors in inducing the parties to share power.90 With a weak majority in the state House of Representatives, the new Council of 122 Chapter 5

Ministers had to rely on a working arrangement with four parties from the RS. Th ough the Alliance spent just eighteen months in offi ce before the 2002 elections, it managed to achieve cross- community cooperation on a number of policy issues. Progress includes improvements in fi nancial discipline, fulfi lling the conditions for accession to the Council of Eu rope, the merger of Bosniak and Croat pension funds in the Federation, reforming the Federa- tion army, the negotiation of entity constitutional amendments, and eventual approval of election legislation.91 Th e Alliance’s operation of the state Council of Ministers (with six portfolios divided equally among the three constituent peoples) was notable for displaying some unity deriving from “a mea sure of pragmatism in the short term and a vague commitment to Eu rope an integration in the longer term.” 92 Arguably, the Alliance for Change made a positive impact compared with the record of power sharing led by the big three parties during the initial post- Dayton years. Bieber writes that it “provided for a decisive break and increased the functionality of the government.” 93 Yet progress made by the Alliance was limited, particularly on the economy. Th e Alliance was particularly slow in relation to privatization, unsurprisingly perhaps, given the parties’ desire to maintain control over public sector companies as a means to secure power and patronage. Lacking a sense of cohesion, the coali tion failed to forge a reform agenda on diffi cult issues, including refugee return and strengthening state structures. Th e high representative ended up imposing reform, including public service broadcasting legislation, amending legislation on identity cards, constitutional reform, the establishment of Judicial and Prosecutorial Councils at state and entity levels, and the creation of an in de pen dent civil ser vice agency. Th ere were also tensions between the SDP and the SBiH, with the two parties choosing to contest the 2002 elections separately. High Representative Petritsch recognized some of the challenges whereby the Alliance was “proving increasingly diffi cult to hold together” in the state parliamentary assembly as well as being “regularly challenged and obstructed by RS authorities.” 94 Aside from these tensions, exclusion of the main Croat party, HDZ, from the state and Federation executives created problems. Th ough the HDZ was excluded from power sharing, it retained control over four Croat- majority cantons and thus “enjoyed the double advantage of being able to inveigh against the failures of the Alliance in power while itself wielding power where it most counted.” 95 Overall, the Alliance has been described as a “weak and weird Stalemate in Bosnia 123 agglomeration of parties of varied strengths, national composition, and ide- ological descent.” 96 Th e parties’ responses to a Constitutional Court ruling in 2000 arguably largely infl uenced the coherence of the Alliance. In the fi rst major challenge to the po liti cal structures under Dayton, the court ruled that the provisions of the entities’ constitutions relating to the position of constituent peoples were unconstitutional. For the ICG, the ruling heralded an important opportunity to push the 1995 peace accords “to their limits and to permit BiH to become a functional multinational state.” 97 Th e background to the ruling is the case brought by former Bosniak member of the presidency Alija Izetbegović in 1998, on the basis that provisions of the entities’ constitutions violated the constitution of Bosnia-Herzegovina. Regarding the position of the constituent peoples in the entities, the case alleged that the RS discriminated against Bosniaks and Bosnian Croats and the Federation discriminated against Bosnian Serbs. In July 2000, the court ruled that “the constitutional principle of collective equality of constituent peoples following from the designation of Bosniaks, Croats, and Serbs as constituent peoples prohibits any special privilege for one or two of these peoples, any domination in governmental structures, or any ethnic homogenisation through segregation based on territorial separation.” 98 Th e entities were then tasked with implementing the decision by agreeing on and implementing a set of constitutional amendments. When the parties were slow to make progress, the OHR became involved by establishing constitutional commissions attached to the entities’ parliaments and facilitating inter- party dialogue. Th ough the parties agreed on a package of constitutional amendments on 27 March 2002 (dubbed the Sarajevo Agreement), the RS elites noted several “reservations” and later presided over diff erent changes that were contrary to those agreed. Th ese diffi culties were pertinent for the operation of power sharing among the parties that made up the Alliance. As the ICG notes, the RS amendments would leave the Alliance open to “accusations of treachery from Bosniak and Croat opposition parties for having signed up to a failed pact.” 99 Although the Federation commission reached agreement on a set of constitutional amendments, it sought assurance that “symmetrical” amendments would be passed in the RS. Th e OHR sought to incentivize the RS with warnings of possible sanctions should the RSNA fail to pass amendments in accordance with the Sarajevo Agreement.100 Yet the amendments draft ed by the RS constitutional commission were rejected and replaced with amendments 124 Chapter 5 prepared by sixty-eight Serb delegates. In the end, the OHR imposed decisions ensuring that the constitutional amendments complied with the Sara- jevo Agreement, “in order to overcome the obstructionist tactics of opposition parties in the Federation” and to correct “a small number of technical shortcomings” in the RS.101 With respect to the operation of the Alliance for Change coali tion, the OHR’s handling of the process is notable for allowing the coalition parties to become victims of criticisms of treachery and weakness, a problem that arguably contributed to the Alliance parties’ defeat in the 2002 elections.102 Additional challenges for the Alliance co ali tion came from the parties’ divergent commitments to state versus decentralized structures, evidenced in the attempt by some Bosnian Croat elites to declare self-government for their own group. In November 2000, the HDZ-led Croat National Congress held a referendum on whether Bosnian Croats should have their own political institutions. On 3 March 2001, Croat self-government was declared with an inter- cantonal council. A few days later, High Representative Petritsch removed the president of the HDZ, Ante Jelavić, from his position as the Croat member of the presidency. Th ough the HDZ boycott of the Federation parliament handed the Alliance an opportunity to push through reform, the coali tion’s illegitimacy in HDZ-dominated areas was all too clear and constrained its authority in that entity. Due to pressure from the international community and the fact that the HDZ was unable to pay its military, the Croat self- rule initiative ultimately failed. Since 2001, the position of Bos- nian Croats has been to secure equality at the canton, entity, and state levels. Th is policy has been punctuated, however, by occasional calls from Bosnian Croats for their own entity. Despite the international community’s hope that the Alliance for Change would herald more eff ective power sharing, the incoherence of the coali tion ultimately led to its downfall in the October 2002 elections. Th ough the Alli- ance made some progress on agreeing to policy, arguably too much was expected of a co ali tion that was fused together by international eff orts and in power for such a brief period. Reverting to pre- Alliance po liti cal dynamics, the 2002 elections produced wins for the three main nationalist parties.103 Th e (Bosniak) Party of Democratic Action (SDA), the Croatian Democratic Union (HDZ), and the Serb Demo cratic Party (SDS) each contributed a member to the tripartite presidency and were joined in the Council of Min- isters by the PDP and the SBiH, established on 13 January 2003. Th e 2002– 6 executive also coincided with Paddy Ashdown’s tenure as high representa- Stalemate in Bosnia 125 tive, a period that cemented a robust approach pursued by the international community.

The Drive for Reform: Ashdown’s Tenure

Paddy Ashdown’s tenure as high representative from May 2002 to January 2006 is signifi cant for a number of reasons. First, the new high representative appeared to adopt a more robust attitude to using the Bonn powers as a means of driving reform. Second, the period is important for the high representative’s attempts to shift the balance of power from the entities to the state level and the strengthening of the power- sharing institutions. Th ird, Ashdown placed a greater focus than his prede ces sors on the need to move beyond the Dayton structures in order to achieve integration in the EU and NATO. Ashdown is considered to have “wielded the Bonn powers, boldly, leaving behind perhaps the strongest legacy of any HR.”104 Th e ICG suggests that Ashdown moved away from his pre de ces sors’ focus on “partnership” to serving “notice on the nationalists that their room for manoeuvre— let alone for abuse of offi ce—will be slight.”105 Ashdown suggested there were two ways he could take decisions and was clear about the way forward: “One is with a tape measure, measur ing the equidistant position between three sides. Th e other is by doing what I think is right for the country as a whole. I prefer the second of these.”106 In terms of a policy agenda, Ash- down very actively imposed decisions in the fi eld of judicial reform, ranging from decisions on the High Judicial and Prosecutorial Councils in the two entities and the prosecutor offi ces in the Federation cantons, to a criminal code for Bosnia, to appointments of international judges to the state court. He was also particularly active in removing and suspending offi cials and elected representatives. Th ough some of his decisions lift ed previous suspensions on some offi cials, Ashdown removed others, including Nikola Grabovic, the Federation fi nance minister; the Croat member of the state presidency, Dragan Čović, who was under indictment by the state court for misuse of offi ce; and the RS minister for education, Milovan Pecelj, for non-cooperation with the OHR. He also removed a host of offi cials, particularly in the RS, for obstructing the work of the International Criminal Tribunal for the Former Yugoslavia (ICTY), with eighty-fi ve such decisions taken in 2004. 126 Chapter 5

Arguably, the high representative was compelled to impose decisions given that the Council of Ministers instigated very little in the way of reform. As the executive’s capacity continued to be limited vis-à- vis the entities, it was the high representative and not the governing parties who set the agenda. As the ICG notes, the reforms initiated by the OHR were “endorsed or tolerated by the nationalist parties,” the executive thereby acting in a largely “reactive rather than proactive” manner.107 Th ough Ashdown was prepared to use the Bonn powers to impose legislation and remove offi cials from offi ce, it also appears that the high representative facilitated a considerable amount of dialogue and consultation with the local parties. For instance, he did much to get the parties on board with his “Jobs and Justice” agenda and pushed six core objectives set out in the Mission Implementation Plan approved by the PIC Steering Board in January 2003.108 Assessing Ashdown’s record as high representative, an international offi cial has described the view that Ashdown adopted a heavy- handed approach to the Bonn powers, imposing measures where he saw fi t, as a “myth.” He maintains that Ashdown’s tenure involved considerable “consultation” with the local authorities that is oft en unacknowledged.109 Ashdown himself suggested that he took a judicious approach to the Bonn powers: “If I pass a decree that is refused, my authority is gone like the morning dew.”110 Nev- ertheless, it is clear that Ashdown resorted to the Bonn powers more oft en than his pre de ces sors and was robust in his approach to policy reform. Tracked by Bart Szewczyk, the use of the Bonn powers has “fl uctuated over time and across diff erent categories of decisions,” with the peak during Ashdown’s tenure and a reduction since the PIC’s announcement in 2006 of its intention to close the OHR.111 It has also been suggested that the “narrow view of legitimate politics” held by the OHR under Ashdown means that Bosnian institutions were “drained of their po liti cal role because Bosnian offi cials are judged to be representing and negotiating on behalf of their par tic u lar ethnic constituencies, interests which are defi ned as confl icting with the public interest.”112 Th e use of the Bonn powers to address ethnic politics is a central feature of the evolution and expansion of the OHR’s role from one of facilitation and coordination to “absolute powers over an open- ended spectrum of issues.”113 Th is does not mean, however, that there was a consistent approach on the part of successive high representatives. Notably, Christian Schwarz-Schilling, Ashdown’s successor, announced a shift in policy. He pledged to use the Bonn powers to “maintain peace and security or to further BiH’s cooperation with the ICTY. Stalemate in Bosnia 127

I have made it equally clear that I will not use the Bonn powers for anything else.”114 Th e reform imposed by Ashdown signaled his intention to improve the eff ectiveness of the power-sharing institutions. Th ese reforms included decisions to streamline decision making in the Federation and strengthen the power- sharing institutions at the state level. In October 2002, Ashdown amended the Federation laws on power sharing to bring them in line with the April entity constitutional amendments.115 Th e amended laws provided for sixteen ministers in the Federation, one prime minister, and two deputy prime ministers, and it abolished the position of deputy minister. Th e amended legislation established that the government would be composed of eight Bosniaks, fi ve Bosnian Croats, and three Bosnian Serbs. In December 2002, Ashdown imposed the law on the state Council of Ministers, increasing the number of portfolios from six to eight: Foreign Aff airs, Foreign Trade and Economic Relations, Finance and Trea sury, Communications and Trans- port, Civil Aff airs, Human Rights and Refugees, Justice, and Security.116 Th e Council of Ministers would have a nonrotating chair and two deputy chairs, to refl ect the equality of the constituent peoples. In terms of decision- making procedures, the Council of Ministers can adopt decisions, provided that at least half of its members are present, including at least two members from each constituent people. When the parliamentary assembly is to take the fi nal decision, the Council of Ministers can adopt an Act via majority support; on all other issues consensus is required. If a consensus is not attainable and the Council of Ministers is unable to resolve the issue, majority approval must include at least two members from each constituent people. In several other important areas, Ashdown had a similar agenda in shift ing the balance of power from the entities to the state level. In February 2003, he established the Indirect Tax Policy Commission, which was tasked with draft ing legislation on the merger of the separate customs administrations into a single customs administration; a single statewide value-added tax; and the creation of an Indirect Taxation Administration for the collection and administration of indirect taxes, including customs duties and VAT. 117 In May 2003, the high representative imposed decisions creating a defense reform commission and an expert commission on intelligence reform. Th e defense reform commission was tasked with draft ing legislation that would secure defense command and control at the state level; the intelligence commission was asked to draft legislation to establish a single 128 Chapter 5 intelligence agency.118 Ashdown’s reforms pointed to overall developments in addressing the autonomy of the entities and strengthening the state level. He believed the unifi cation of the three customs services and the introduction of VAT at the state level were among his most signifi cant achievements.119 Ashdown’s tenure is also notable for linking policy reform with the potential to move closer to Euro pe an integration. Progress would mean “moving out of the era of Dayton and into the era of Brussels.”120 Impor- tantly, police reform came to be an essential precondition for making progress toward eventual EU membership. In 2003, the OHR made police reform a priority, and the following year the EU made reform a prerequisite for the start of negotiations on a Stabilization and Accession Agreement (SAA). In June 2004, Ashdown announced the creation of a Police Restructuring Com- mission (PRC) to work on proposals for police reform that would strengthen police structures. Th e PRC’s main task was to propose a single structure of policing under a ministry or across several ministries of the Council of Ministers. Th e EU even specifi ed three principles that the Euro pe an Com- mission would set as a condition for signing a SAA. Police reform would have to (1) place exclusive competence for police legislation and budget at the state level; (2) reor ga nize police regions on the basis of functional police criteria; and (3) protect the police from po liti cal interference.121 Although the OHR was at pains to reassure Bosnian Serbs that police reform would not threaten the existence of the RS, Ashdown called on the parties to agree on the creation of a single state police structure, which would involve the RS giving up some power “to create a light- level functional state structure so this country can join Europe.” 122 Despite the eff orts made by the OHR and the Eu ro pe an Commission, cooperation among the parties proved elusive. Th e PRC’s fi nal report proposed that state- level institutions would have exclusive competence for policing. Th e details of ten police regions identifi ed by the OHR “were determined entirely on the basis of technical police criteria and ignored boundary lines between existing entities, cantons, and districts.”123 In eff ect, this would mean the abolition of entity forces. Unsurprisingly, the Bosnian Serb parties were not prepared to support the abolition of their entity police force and rejected proposals for police regions that would cross entity boundaries. Fol- lowing the failure to reach agreement among the parties on the basis of the PRC report, the OHR convened a series of talks with the local parties in April 2005. Although the parties signaled a potential agreement on a policing board Stalemate in Bosnia 129 that would provide for entity and cantonal consultation at the state level, the RS representatives continued to oppose police regions that would cross entity boundaries and any move to weaken entity control. Discussed in the next chapter, police reform was an ongoing objective for external actors in Bosnia and featured as a precondition for a SAA with the EU. Th e Peace Implemen- tation Council also stipulated that the conclusion of the SAA as well as constitutional reform would bring about a situation in which the OHR would close and the EU would become the lead external actor responsible for peace implementation and policy reform.124 Arguably, this announcement served as a marking point in the eff orts to increase “local ownership” and to move away from the contested politics of the initial post- Dayton years toward EU integration.

Summary

Th e Bosnian case shows how cross- community power sharing can be an outcome of “hothouse” peace negotiations as a compromise among the contending groups. Power sharing was, of course, only part of a wider settlement, including territorial and security elements and refugee return. Th e case is well known for the imposition of an agreement by external actors. Analysis of the Dayton talks shows there was no clear blueprint for what the po liti cal structures should look like. Even though a multiplicity of actors was involved, the delegations arrived at consociational elements in the draft constitution. Certainly, the proposed power-sharing institutions were the result of politi cal realities: the existence of the two entities and the determination of the three groups to have explicit group rights, resulting in veto rules and predetermined politi cal institutions. Th ere was also an important historical legacy at play, given that the Yugo slav system was based on decentralization and cultural autonomy, with proportional represen ta tion of the three groups in the former republic’s administration and Communist Party structures. Executive power sharing was thus the result of interaction among the internal and external actors and was ultimately a compromise among group preferences. Th is period also points to evolution of power- sharing rules whereby the OHR took the opportunity to extend the powers of the Council of Min- isters and revised the entity constitutions in 2002. Th e initial post-Dayton years highlight the limitations of institutional rules in fostering cross-community cooperation. On the basis of the Bosnian 130 Chapter 5 constitution, the main parties recognized that they could largely ignore power sharing at the state level because power was concentrated in the entities. Th is tension between the state level and the decentralized layers in a plurinational federation suggests some lessons for institutional design. Ar- guably, actors involved in designing pluralist federations in deeply divided places need to give further consideration to the division of powers. It seems that not nearly enough thinking about this balance of power was done at the Dayton talks. Th e concentration of power at the entity level failed to satisfy all groups. If state structures fail to meet group preferences, there is little potential for power- sharing rules to foster cross- community cooperation. Moreover, the rules themselves may be contested, further limiting the potential for eff ective power sharing. In partic u lar, veto rules have been problematic in terms of the issues that can be subject to veto procedures and the number of veto points throughout the power-sharing institutions. Ulti- mately, the fi rst ten years of postwar Bosnia highlight the limitations of power-sharing rules when a peace agreement fails to satisfy group preferences on state structures. At best, it can be said that the power-sharing institutions (bolstered by international engagement) have held things together. But given the complexity of the system and the seemingly ongoing unwillingness of some domestic actors to invest their politi cal eff orts at the state level, power sharing has been far from effi cient or eff ective, with limited evidence of cross- community cooperation. Finally, the Bosnian case points to the signifi cance of external actors in the formation, transformation, and operation of power- sharing structures. Th e Bosnian case underlines the crucial role for external actors and their eff orts to induce compromise and joint decision making among the local parties. Th e initial postwar period is notable for the imposition of politi cal structures by “international decree” and the evolution of external intervention, evidenced by the creation of the Bonn powers. Th e external actors have operated an incentives structure based on positive incentives (policy reform as a means to progress toward EU integration) as well as negative incentives and explicit coercive mea sures in the form of the Bonn powers. It is questionable, however, how eff ective these incentives were in fostering cross- community cooperation. We know that the OHR’s robust engagement proved to be an excuse for local elites to avoid inter-ethnic engagement and compromise, in the knowledge that diffi cult decisions would be imposed anyway. Increasingly frustrated with the failure of local elites to share power eff ectively, successive high representatives resorted to imposing decisions, Stalemate in Bosnia 131 albeit to a varying degree. At one level, we can point to the positive role of the international community (particularly the OHR) in holding the fragile state together. But the OHR’s role in peace implementation has been diffi - cult, given that the agreement it was tasked to uphold is fl awed and lacking in legitimacy. Moreover, the OHR has been increasingly of the view that the DPA is, to some extent, dysfunctional and has sought to revise it. It is in this context that robust criticism has been levied at the OHR. As Knaus and Martin write, the Bosnian case “shows the ease with which a state-building mission may start out with unlimited powers to meet extraordinary circumstances and end up as an uncomfortable caricature of a Utilitarian despot.”125 Th e Bosnian case, however, points to the diffi culties for external actors to revise po liti cal structures when domestic actors resist such intervention. Since 2005, the issue of constitutional reform has been particularly controversial, coupled with a growing debate over the strategy of the international community and the planned closure of the OHR. CHAPTER 6

From Dayton to Brussels?

Little more than ten years aft er the Dayton Peace Agreement, the Peace Im- plementation Council signaled that the Offi ce of the High Representative would close. Th is move promised an end to heavy external engagement and the beginning of greater own ership on the part of local elites to manage their own aff airs. Rather than a robust OHR armed with the Bonn powers, Bosnia would be guided by the EU delegation to manage the country’s preparations for eventual EU membership. Considerable faith was entrusted in the lure of EU integration to incentivize local parties to agree on policy reform, including reform of the Dayton structures. In this chapter, I explore how external actors have sought to incentivize constitutional reform and why this pro cess failed to produce agreement among the domestic actors. I also consider the impact of this pro cess on power- sharing practice, arguing that the focus on revising the po liti cal structures brings into stark relief the ongoing divergent positions among the parties on the kind of state Bosnia should be. Arguably the ongoing diffi culties refl ect the underlying self- determination dispute, now said to be “frozen.”1 Bosniaks see the entity structure as a reward for ethnic cleansing and want to create a new constitution based on a unitary state. Bosnian Serbs resist any attempt to erode the authority of the RS and have, on occasion, called for a referendum on inde pen dence. Bosnian Croats want to at least retain equal power to protect their group rights, and they occasionally make calls for the creation of their own entity. Th e Bosnian case shows that external incentives will be less eff ective when the peace agreement remains contested and external actors are perceived as pushing a structural reform agenda against the preferences of contending groups. I also highlight the limitations of the EU From Dayton to Brussels? 133 membership “carrot” in deeply divided places and the challenges for the integration process.

External Actors: Incentivizing Constitutional Reform

Th e practice of power sharing in Bosnia since 2005 has been marked by an ongoing pro cess of constitutional reform, encouraged by external actors including the OHR and the Euro pe an Union Special Representative (EUSR), the U.S., the Eu ro pe an Commission, and the Council of Eu rope’s Venice Commission. Th ese external actors have been committed to reforming the Dayton structures via agreement among the main po liti cal parties. Crucially, constitutional reform has been pitched by external actors as an essential condition for securing Bosnia’s integration into EU and NATO structures. In March 2005, the Council of Eu rope’s Venice Commission published its Opinion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the High Representative.2 Th e Venice Commission made an explicit link between the weak powers at the state level and the country’s lack of progress toward Euro pe an integration, suggesting that it would be “unthink- able” for the country to make progress on the basis of existing state structures and that the “EU will not countenance the kind of delay, indecision, and uncertainty that a multiplicity of government entails.” For the Venice Commission, the state institutions would need greater capacity to manage the negotiation of the Stabilization and Association Agreement (SAA) and would need to act as a “single interlocutor” with the Euro pe an Commis- sion.3 In order for Bosnia to progress toward eventual EU membership, the Venice Commission recommended the transfer of power from the entities to the state level; a clear defi nition of “vital national interest” veto rules in the Bosnian constitution; abolition of the entity veto in the state parliament; the VNI to be handled by the House of Representatives; and the abolition of the tripartite presidency, to be replaced by an indirectly elected president with limited powers. With respect to the compatibility of the Bosnian constitution with the Eu ro pe an Convention on Human Rights (ECHR), the Venice Commission focused on the provisions relating to the composition and election of the presidency and the House of Peoples and noted the tension between predetermined positions for the three constituent peoples and the principle of individual rights and equality of citizens, particularly the 134 Chapter 6 exclusion of “others” from po liti cal positions. Th e Commission was also critical of the overall territorial or ga ni za tion of the state, albeit accepting that abolition of the two entities “seems unrealistic in a medium- term perspective since this would not be accepted within the RS,” and it called for reform of the Federation, including the concentration of legislative authority at the entity rather than canton level.4 On the subject of the role of the OHR, the Commission accepted that the use of the Bonn powers may have been benefi cial for the country in the early post- Dayton period but stressed that “such an arrangement is fundamentally incompatible with the democratic character of the state and the sovereignty of BiH. Th e longer it stays in place the more questionable it becomes” and risks the creation of a “depen- den cy culture incompatible with the future development of BiH.” 5 Regarding the OHR’s use of the Bonn powers to remove public offi cials and politicians from offi ce, the Commission considered it “unacceptable that decisions directly aff ecting the rights of individuals taken by a politi cal body are not subject to a fair hearing or at least the minimum of due pro cess and scrutiny by an in de pen dent court.” 6 Th e Venice Commission’s Opinion is signifi cant for the extent of reforms proposed and for making an explicit link between institutional revision and the country’s progress on the path to EU integration. Th e Opinion is also important for having triggered a process of constitutional reform, characterized by ongoing eff orts by external actors to induce inter- party agreement. At a conference on post-Dayton statebuilding in October 2005, EU Enlargement Commissioner Olli Rehn stated that Bosnia “must now make the choice whether to maintain the current constitution with its functional limitations—or to opt for constitutional changes necessary to make herself a stable and functional country, ready to progress toward the EU.” He stressed that the po liti cal structures under the DPA are “too complex and fi scally unsustain- able” and implored the Bosnian authorities “to streamline this expensive and multilayered bureaucracy. It is also clear that this structure poses problems as to the ability of Bosnia and Herzegovina to negotiate and then to implement a Stabilization and Association Agreement with the EU, not to speak of accession.”7 Rehn also described the role of the OHR as “problematic,” signaling a need to move beyond the Bonn powers in order for local elites to have greater ownership of policy reform. At the offi cial opening of SAA negotiations in Sarajevo in November 2005, Rehn described constitutional revision as a “key reform” and argued that “it is crucial for this country to develop effi cient governance in a functional multi-ethnic state to serve better its citizens.”8 From Dayton to Brussels? 135

Th e U.S. was also involved in seeking to persuade the parties to agree on a package of constitutional reforms in order to help make power sharing more eff ective and to help ensure the integration of the state into Euro- Atlantic institutions. A working group was set up in March 2005 by former principal deputy high representative Donald Hays and Bruce Hitchner, di- rector of the Dayton Peace Accords Project at Tuft s University. Following a number of sessions over the summer months, several parties— SDA, SBiH, HNZ and SDP— were reportedly fairly close to agreeing on reform.9 Yet the HDZ insisted on the retention of safeguards for the three constituent peoples throughout Bosnia, and the Bosnian Serb parties remained opposed to any proposal that would weaken the authority of the RS. Th ere was, however, some mea sure of inter- party agreement on streamlining the House of Peoples with a limited set of powers and retaining responsibility for the VNI. It proved impossible, however, to make much progress on other issues, including further changes to the state parliament, the election and role of the president, and the responsibilities of the Council of Ministers. At a meeting with party leaders in Brussels in November 2005 sponsored by the Euro pe an Commission, there was some agreement that the House of Peo- ples would handle the VNI, with a defi ned list of areas that would fall under “vital national interests.” Yet the ongoing divergent positions on the strength of the entities prevented an agreement, with Bosniaks wanting to see ultimately the abolition of the entities and Bosnian Serbs wanting to secure the authority of the RS in any agreed constitutional reform. In the absence of agreement, a follow- up meeting took place in Wash- ington, coinciding with the tenth anniversary of the DPA. U.S. Secretary of State Condoleezza Rice said the parties’ agreement in “modernizing the Dayton accords” would be “essential” to “reach the full goal of integration into Euro-Atlantic institutions.”10 Despite this pressure, the parties failed to make progress, given their divergent positions over the power of the entities. In December, the U.S. and UK embassies in Sarajevo convened another series of meetings, with reportedly some progress on individual rights, greater authority for a prime minister, and making the Council of Ministers more eff ective, and discussion took place on a presidency with one president and two vice- presidents to replace the tripartite model.11 Th e central problem, however, related to the issue of entity voting in the state parliament, with Bosnian Serb parties making clear that entity voting would need to be maintained in order to reach agreement, while the SDA and SBiH sought its abolition. Th e proposed constitutional amendments agreed among the U.S.- led 136 Chapter 6 working group included a tripartite presidency elected by the House of Rep- resentatives, limiting the function of the House of Peoples to the VNI, increasing the House of Representatives to eighty-seven delegates (twenty-eight from each “constituent people” and three delegates representing “others”), adding agriculture and environment ministries to the Council of Ministers, and concentrating responsibility for fulfi lling conditions for EU accession at the state level. Arguably, these developments constituted some departure from the DPA and signaled the potential for ongoing discussion on further reform. It appears that the external actors involved believed the package would be the beginning of a series of constitutional reform initiatives, a step in the right direction toward more reform, including increased decentralization of the entities and cantons.12 Th e external actors involved in the process welcomed progress on reaching an agreement, adding that the reforms should constitute the fi rst steps in an ongoing pro cess to take place following elections in October.13 But the failure to agree on entity voting ultimately led the SBiH to withdraw its support for the pro cess; they were later joined by members of the HDZ in opposing the reform package. Th e “April package” presented to parliament failed to secure the required two- thirds majority in parliament by just two votes. Th e SBiH’s opposition to the package can be explained by the maintenance of entity voting in the state parliament and the party’s view that accep tance of reform “would imply ratifi cation” of the DPA, which had never been approved in parliament.14 As we have seen, the power of entities and the provision for entity voting are highly contentious issues. Given the parties’ divergent positions on these issues, it is perhaps unsurprising that talks would be diffi cult and prone to failure. Former UK Ambassador to Bosnia and Herzegovina Matthew Rycroft recalls that although the Bosnian Serb parties “always presented a united front,” the Bosniak side was less harmo- nious.15 Presumably, the SBiH felt that anything less than a complete deal on constitutional reform would leave the party wide open to criticism from the Bosniak community. Indeed, divisions on the package were evident on an intra- as well as inter-communal basis. Sebastián refers to the divisions between the Bosniak parties, the SDA, and SBiH, with the latter adopting a stricter approach, demanding change on the entity structure and abolition of entity voting.16 An SDA politician claims that the “propaganda” pursued by the SBiH on the limitations of the April package damaged his own party’s electoral chances in October 2006.17 And a split occurred among Bosnian Croats, leading a splinter party, HDZ 1990, to oppose the April package. From Dayton to Brussels? 137

Unsurprisingly, the external actors involved in pushing reform were disappointed the package was rejected, particularly in light of the positive signs from the working group over the previous few months. Hitchner laments the “opportunity lost” and suggests that the lesson from the April package is that “constitutional reform cannot be pursued lightly or without sustained and focused involvement by the politi cal parties and outside mediators.”18 Th e Peace Implementation Council (PIC) expressed its “deep regret” on the failure of the House of Representatives to pass the proposed reform and stated that constitutional reform “remains unavoidable and essential for a sustained and stable BiH in the future.”19 Off ering an insight into the negotiations, Rycroft recalls the “strange dynamic among the parties, who know each other pretty well” but whose discussions were nevertheless “very frustrating.”20 Much was made of the failure of the April package as a serious hurdle to EU integration. Rycroft described the rejection of the package as a “sad day” for the country and a “missed opportunity,” and claimed that “Th ose who have voted against this change have delayed the journey to Europe and normality.”21 High Rep- resentative Christian Schwarz-Schilling similarly implied that the package’s failure was a step back because agreement was necessary to “better equip BiH for EU integration and better equip government to serve its citizens.”22 Th e constitutional reform pro cess leading to the April package also highlighted a diff erence of approach between the U.S. and the EU. Th e U.S. actors—the working group initiative led by Donald Hays, Bruce Hitchner, and Paul Williams—were more proactive in seeking to bring about an inter- party agreement. Hitchner refers to their eff orts as “an initial and modest attempt to reduce the many structural impediments” in the state institutions; he says that the secretariat of the working group was “nothing more than a very modestly funded, virtually ad hoc second- track pro cess.”23 Neverthe- less, the initiative was clearly U.S.- led, with the EU taking a back seat. Sebas- tián refers to the EU’s “hands- off and passive stance” as contributing to the failure of the April package.24 She suggests that in the drive for agreement, the EU “became more of a spectator than an active engager” and was, to some extent, critical of the U.S.- led initiative.25 In contrast to the EU’s relative disengagement, she refers to the “hands- on” approach of the U.S., with Hays commenting that “Th e Euro pe ans wanted the Bosnians to actually do this, while the U.S. was much more interested in having an outcome.”26 Th e EU’s preference that reform should be a rolling process on the basis of inter- party dialogue has been a feature of the international community’s engagement in Bosnia, particularly since the PIC’s announcement in 2006 of OHR 138 Chapter 6 closure. As discussed below, the EU’s soft er emphasis on facilitating dialogue contrasted with the OHR’s reliance on the Bonn powers as a necessary mechanism to induce reform. It was, perhaps, unfortunate that the failure of the constitutional reforms to secure parliamentary approval took place so soon before the October 2006 elections. Before and aft er the elections, the external actors engaged in facilitating the April package continued to press for reform. For instance, Hays and Crosby warned that “if Bosnia is to successfully undertake the accession process, the Bosnian authorities and the politi cal parties that form the government must suffi ciently empower the state to meet the EU’s stringent requirements.”27 In the 2006 election, the SDA and SBiH won nine and eight seats, respectively, in the House of Representatives; the SNSD won seven seats, the SDS three seats, and a split in the Bosnian Croat vote resulted in HDZ BiH gaining three seats and HDZ 1990 two seats. For the state presidency, Haris Silajdžić (SBiH) secured the Bosniak vote, the SNSD’s Nebojša Radmanović won the Bosnian Serb vote, and Željko Komšić (SDP) took the Bosnian Croat seat.28 Th e election results, particularly the success of the SBiH, who rejected the April package, led to post- election uncertainty on the way forward. During the election campaign, the SBiH led by Silajdžić had called for the abolition of the entities, and the SNSD threatened a referendum on RS secession from the state, creating a fairly antagonistic po liti cal dynamic and little opportunity for compromise. It is also clear that the international community was unsure how to move the pro cess forward, with Hays reportedly saying that it should be up to the local parties to progress on constitutional reform “without being prodded from outside,” and that a renewed international eff ort was “unlikely to produce a stable constitutional settlement in the long term.”29 Politicians recall how the power and infl u- ence of the Bosniak and Bosnian Serb leaders, particularly the antagonism between Silajdžić and Milorad Dodik, removed any potential for progress on constitutional reform.30 Other parties believed that the strong position of Silajdžić and Dodik meant that constitutional reform was dead in the water. For the SDA, the overriding problem was that Dodik viewed Bosnia as simply the sum of the two entities, not a state with its own identity, a position irreconcilable with the Bosniak perspective.31 For several years following the failure of the April package, the U.S. and the EU were unable to do much to spark renewed focus on constitutional reform. Th e U.S. did, however, convene a meeting in Washington in late 2007, albeit without reaching agreement. Just a year following elections, the posi- From Dayton to Brussels? 139 tions of the parties, particularly the SBiH and the SNSD, remained maximalist. Dodik claimed that “Eu rope and the U.S. are wrong to try and make Bosnia a state with strong state- level bodies, since it is obvious this can’t work.”32 In seeking to cement the power of the entities, the SNSD had previously proposed a new territorial arrangement, including a Croat entity, Herceg- Bosna. Th ough Bosnian Croats may have welcomed the proposal, the SBiH rejected it, saying, “Th is federalization is practically the continuation of ethnic cleansing and . . . is completely unacceptable.”33 Th e pro cess of constitutional reform was given a boost in June 2006, when the PIC announced that the OHR would begin preparations for closure on 30 June 2007. It stated that a “reinforced EUSR Offi ce would constitute an important part of the EU’s comprehensive engagement aft er OHR.” Police reform, reform of the public broadcasting system, and full cooperation with the International Criminal Tribunal for the Former Yugoslavia (ICTY) were noted as preconditions for conclusion of the SAA. Th e PIC also stated that “BiH politicians must make constitutional reform a top priority in order to build a sustainable democracy and to make governmental and parliamentary institutions effi cient and eff ective.”34 Th us, the PIC indicated that constitutional reform as well as conditions for securing the SAA would need to be achieved in advance of OHR closure. With reform unforthcoming over the next few years, a number of conditions were put in place for OHR closure. Bosnian authorities were tasked to deliver on fi ve policy objectives: resolution of the state property issue, resolution of defense property, completion of the Brčko Final Award, fi scal sustainability, and entrenchment of the rule of law.35 Two further conditions needed to be fulfi lled: the signing of the SAA and a positive assessment of the politi cal situation “based on full compliance with the Dayton Peace Agreement.”36 In November 2008, three party leaders— Dodik (SNSD), Sulejman Tihić (SDA), and Dragan Čović (HDZ BiH)—met at Prud in northern Bosnia and announced an agreement on a number of issues relating to the PIC conditions for OHR closure and constitutional reform. In a joint statement, the parties stated they had reached agreement on the harmonization of the Bos- nian constitution with the ECHR; the competences and functioning of Bosnian state institutions; division of state property; and territorial or ga ni- za tion, particularly concerning the middle levels of government.37 However, the exact details of this agreement remained unclear, prompting the OHR to push for elaboration and implementation. High Representative Miroslav Lajčák (who replaced Schwarz-Schilling in July 2007) stated that the agreement 140 Chapter 6 must not be “a show” for the international community and encouraged the parties to present the details to parliament. Echoing previous high representatives, he threatened that lack of progress would risk Bosnia lagging behind its neighbors on the path to EU integration.38 Follow- up discussions took place in January 2009, with agreement that Bosnia would consist of three levels of government, and the middle level of government would be based on four territorial units.39 Silajdžić accused the parties of agreeing on the division of the Federation and the creation of Bosniak enclaves. By July 2009, the new high representative, Valentin Inzko (who replaced Lajčák in March 2009), stated that the negotiations relating to the Prud Agreement were at an end and that progress should involve all main parties.40 Ultimately, the so- called Prud Agreement did not come to much, except demonstrating that the parties were capable of negotiating some of the issues set out as conditions for OHR closure. Hitchner suggests that the lack of external “facilitation” was one contributory factor for the failure of the Prud Agreement.41 In contrast, the next stage of the constitutional reform process was led by the U.S. and the EU presidency, the so-called Butmir process, launched in October 2009. Th is new eff ort to search for an agreement on constitutional reform was a joint U.S.- EU initiative led by U.S. Deputy Secretary of State James Steinberg, the EU presidency (represented by Swed- ish Foreign Minister Carl Bildt), and the EU Enlargement Commissioner, Olli Rehn. At talks held over several days at the Camp Butmir military base outside Sarajevo, the external actors presented a package of reforms and suggested that progress had been made but that the parties would have to “demonstrate greater determination and fl exibility.” 42 Th e package included proposals to reform the state institutions in compliance with the ECHR, a shift from a bicameral to unicameral system, the creation of a more powerful Council of Ministers with a prime minister, and concentration of responsibility for preparing for EU integration at the state level.43 Refl ecting the desire to move to an EU-led process, Bildt stressed that the parties would need to meet the conditions for OHR closure in order to progress toward EU membership: “Th e Euro pe an Union is a union of sovereign democracies, not a union of semi-functioning semi-protectorates. . . . It is up to the Bos- nian politicians to take their country from a semi- protectorate to sovereign democracy.” Bildt and Steinberg reportedly tempted the parties with accel- erated integration into the EU and NATO in return for support of the whole package.44 Yet the talks at Butmir ultimately failed to produce a breakthrough. Hitchner suggests, “It would have been better had Butmir been the From Dayton to Brussels? 141 beginning of a well-thought- out and longer term new strategy that included, among other things, an in- depth and sustained constitutional reform initiative.” 45 Th e ICG criticizes the external actors’ ill-advised “all or nothing approach,” given that Bosniaks sought wholesale changes to the DPA, and Bosnian Serbs were opposed to anything more than minimal revision.46 Th e Bosniak parties were disappointed that the initiative did not go far enough and failed to abolish entity voting, and the Bosnian Serb elites rejected any reduction of entity power. Dodik claimed that the talks were an “unnecessary adventure” and that the Butmir pro cess “does not exist,” leading to renewed rumors that the SNSD would call a referendum on RS secession.47 Others have suggested that the EU and U.S. teams were ill prepared, that the timing was inadvisable given the run-up to elections in 2010, and that there was too much belief in the carrot of enlargement to overcome domestic elites’ po liti cal preferences.48 Th e plans appear to have changed from a “marathon conference” over several days to a “brief session” on 8– 9 October and pre- senta tion of the package on 20–21 October, followed by plans for an open- ended pro cess. With the failure of the Butmir initiative, the external actors were unsure how to progress the goal of constitutional reform. Complicat- ing the way forward, it might also be arguable that the Butmir process further weakened the OHR, given that it was confi ned to the sidelines by this EU-led process. On 22 December 2009, the Eu ro pe an Court of Human Rights (ECtHR) published its ruling in Sejdić and Finci v. Bosnia and Herzegovina. Th e plaintiff s, Dervo Sejdić and Jakob Finci (Roma and Jewish, respectively), contended that the Bosnian constitution’s provisions for the election of the presidency and House of Peoples based on the three constituent peoples are discriminatory and violate the ECHR. Th e ECtHR found Bosnia to be in violation of Protocol 12 of the ECHR on the rights to equal treatment and nondiscrimination by denying “others” the right to stand as candidates for the presidency. Bosnia was also found to be in violation of Article 14 of the ECHR (freedom from discrimination) and Article 3 of Protocol 1 (free elections to parliament), given the ineligibility of “others” to stand for the House of Peoples. Th e ruling was clearly signifi cant for the substance of potential constitutional reform and the EU integration process. Th e onus has since been on the politi cal parties to amend the Bosnian constitution to ensure equal treatment of all citizens, including “others,” in elections for the presidency and state parliament. An EU offi cial commented that Bosnia “will not have credible prospects for accession if there is noncompliance with the 142 Chapter 6 ruling” and that it is up to the Council of Europe to monitor progress and take measures. 49 Again, the parties were unable to agree on the way forward. In June 2010, the PIC expressed its concern over the delay in compliance with the ECtHR ruling and stated that although such constitutional reform is not a condition of OHR closure, it was nevertheless convinced that revisions are “necessary to fully comply with democratic principles as well as improve the effi ciency and functionality of BiH’s institutions so that they can be better equipped to serve the interests of BiH’s citizens and meet future requirements for Euro-Atlantic integration.” 50 Th e E u r o p e a n C o m m i s - sion regarded the delay in harmonizing the Bosnian constitution with the ECHR as a “fundamental issue of concern.” 51 Seeking to move the pro cess forward, German Chancellor Angela Merkel invited the leaders of the Bos- nian parties to Berlin in early 2011 to make a deal on executive formation and a commitment that EU integration would be handled at the state level. Refl ecting the ongoing divergent positions on state structures, Dodik reportedly rejected Merkel’s proposals, insisting that the entities must be in control of the EU integration pro cess.52 Since 2005, the eff orts of external actors in Bosnia have concentrated on bringing about agreement among the local parties on reforming the Dayton structures. Th e OHR/EUSR, U.S., Euro pe an Commission, and Venice Com- mission have all been involved in seeking to incentivize constitutional reform. Yet agreement among the local elites remained unforthcoming, despite the emphasis on DPA reform as a necessary development on the path to EU integration. As discussed below, there have been several problems with EU incentives that help explain why the lure of membership has failed to persuade elites to improve the functionality of the power- sharing institutions. Ultimately, the diffi culty refl ects the ongoing contestation over state structures. Strained inter-group relations are brought into stark relief by the external actors’ diffi culty in achieving reform, with Bosnian Serbs opposed to any tinkering of their rights under Dayton and Bosniaks hoping for greater intervention to create a centralized state.

Power- Sharing Practice 2006– 10: Increased Antagonism?

It is perhaps not surprising that following the failure of the April package, the ongoing inter-party divisions over constitutional reform impacted the operation of power sharing. Given the divergence of the parties’ positions, From Dayton to Brussels? 143 there was some delay to executive formation following the October 2006 elections. Th e new seven-party Council of Ministers was fi nally formed in February 2007, with the SNSD’s Nikola Špirić as chair. Th e record of this power-sharing executive is far from positive. One international offi cial described the Council of Ministers as “nasty,” claiming that the intra-executive antagonisms eclipsed that of the 2002– 6 Council of Ministers, which had managed to achieve at least some policy progress. Th e PIC expressed concern for a deteriorating po liti cal situation, the responsibility of leaders from both entities “who have blocked progress and undermined the po liti cal situation with their aggressive rhetoric.” Th e po liti cal situation was judged to be of “utmost concern to the international community.” 53 An OHR offi cial suggests that this period demonstrated a mea sure of “regression” and a “horri- ble about- turn” since 2006, with increased tensions and the RS wishing “to go back to 1995, 1996” in terms of retaining control.54 Another OHR offi cial speaks of a “downward spiral” and claims these years represent “the worst” period since the war, with RS elites engaging in a “clear attempt to undermine the state.” 55 Th e power- sharing institutions 2006– 10 have been shaped, and ultimately constrained, by tensions between the Bosniak and Bosnian Serb parties. In par tic u lar, the relationship between Dodik (SNSD leader) and Silajdžić (SBiH leader) has been extremely negative. Silajdžić has been referred to as “Dr No,” the “guy who blocks things.” 56 An ongoing critique of Silajdžić by international actors is that the Bosniak leader adopted an “all or nothing” approach in favor of a speedy abolition of the entity structure, rather than accepting that change will take place incrementally.57 In this way, the functioning of the power- sharing institutions was constrained by Silajdžić’s promise to the Bosniak community of delivering a state without entities. It is also suggested that owing to Dodik’s “aggressive nationalist rhetoric,” the Bosnian Serb and Bosniak leaders essentially “feed off each other.” Th e tensions between leaders and parties led to disputes over which layer of governance should progress policy making, with Bosnian Serbs pushing the entities and Bosniaks calling for increased transfer of powers to the state level. Th e ICG refers to this divergence as “the twin poles of radical DPA revision.” 58 Unsurprisingly, in this climate of inter- party hostility, there was little sign of compromise. Rather than advancing policy instigated by the Council of Ministers, the period was characterized by negotiations on police reform, a precondition for signing a SAA with the EU. Much of this period was taken up with external 144 Chapter 6 eff orts seeking to bring about an agreement on police reform. Noted in the previous chapter, police reform was a priority for Paddy Ashdown during his tenure as high representative, and the Euro pe an Commission had set three criteria on policing as a condition for realizing a SAA. International eff orts in bringing about an agreement on police reform were unsuccessful, particularly given that the RS rejected the abolition of its police force and opposed proposals for police regions that would cross entity boundaries. Inter- preting the proposals as a threat to their “sovereignty,” RS President Dragan Čavić declared, “Th e RS is a result of a four-year fi ght of the Serb people . . . not a gift of the international community.” 59 RS Prime Minister Milorad Do- dik announced that his position on police reform would remain unchanged “even if that means giving up the SAA negotiations.” 60 Following a protracted series of stop- start discussions, with considerable pressure from the international community and reluctance on the part of local parties, particularly elites in the RS, the parties fi nally made some progress in 2007. Much of this progress arose from the OHR’s insistence that a SAA would not be signed in the absence of an inter-party agreement on police reform. In September, High Representative Lajčák presented the parties with a compromise protocol on police reform to take account of the Euro pe an Commission’s three principles. He warned that without an agreement on police reform, Bosnia would not be able to progress negotiations on EU membership. While de- claring that entry into the EU should be “an important motivating factor” for the parties to cooperate, Lajčák acknowledged that “the integration factor is, for the time being, much weaker here than in Central Europe or in other countries of this region.” He suggested, “According to local po liti cal culture, compromise is not considered a victory but a defeat. But the public expects them to come to an agreement. Th e outcome of the police reform will determine whether integration or isolation are in store for Bosnia and Herzegovina.” 61 At the end of October 2007, the main po liti cal parties reached an agreement set out in the Mostar Declaration whereby the entity police forces would be reformed under a single state force from a single bud get and would be free of politi cal infl uence. Th e parties then agreed to an Action Plan for the implementation of the agreement. Th is included the establishment of a number of state- level bodies to coordinate policing in a single structure.62 In April 2008, the Bosnian state parliament adopted legislation for the implementation of police reform, fi nally realizing the goal of signing the SAA with the EU. Th e high representative confi rmed that this development opened the From Dayton to Brussels? 145 door to Europe: “Once embarked on this road, no country has failed to become a member of the EU. Th is decision sends a strong signal to the EU and to investors that BiH is making progress and is open for business.” 63 Bosnia signed the SAA with the EU on 16 June 2008. Of note is that elites had not met the EU’s criteria on police reform. Arguably, this has helped limit the legitimacy of EU strategy in Bosnia. For instance, the ICG warns that “Bos- nian politicians are likely to discount tough talk from Brussels in future because they anticipate similar retreats.” 64 Th e period is also notable for some imposition by the high representative designed to improve the functioning of the power- sharing institutions at the state level. On 19 October, Lajčák imposed a decision revising legislation on the Council of Ministers. In response to the ongoing blockages and lack of progress on policy, the high representative intended to improve power sharing by revising the decision- making procedures, making it more diffi - cult for a party or entity to block the Council of Ministers and parliament.65 Among the revisions to the 2002 Law on the Council of Ministers imposed by Ashdown, this new law provided for a majority decision to include at least one member of each constituent people (rather than two members of each constituent people). Lajčák’s decision also provided for the deputy chairs to convene a session of the Council of Ministers should the chair fail to convene two consecutive sessions. In response, the decision created uproar among Bos- nian Serb elites leading to the resignation of Nikola Špirić (SNSD), Chair of the Council of Ministers, and threats from Dodik that the party as a whole would withdraw from the power- sharing institutions. Th e SNSD’s reaction to the high representative’s decision clearly demonstrated the gap between the high representative’s robust style of state building and the diffi culties in creating power- sharing institutions that would develop of their own accord. To break the stalemate, the high representative published a follow- up decision on 3 December to clarify the eff ects of the October legislation. Th e “Au- thentic Interpretation” stressed that the legislation does not change the composition of the Council of Ministers and is designed to ensure that no member “can obstruct its work simply by an unjustifi ed and illicit absence from sessions and the need for the Council of Ministers to be able to take decisions at all times.” 66 Th is po liti cal crisis somewhat abated when the parties re- engaged in discussions on police reform, ultimately leading to the signing of the SAA in 2008. Since the PIC announced in 2006 its intention to close the OHR, the practice of power sharing has been hampered by increasing tension between 146 Chapter 6 the OHR/EUSR and the Bosnian Serb parties, particularly Dodik’s SNSD. Bosnian Serb elites became increasingly robust in their rejection of external statebuilding eff orts, evidenced by Dodik’s threat to bring 50,000 Serbs to Sarajevo to demonstrate against the high representative’s threat to remove him from power.67 Noted above, the imposition of reform by Lajčák in Oc- tober 2007 led to the resignation of the (Bosnian Serb) Chair of the Council of Ministers, Nikola Špirić. In June 2009, following OHR pressure on the RS to retract declarations opposing the transfer of competences, Dodik threatened a walk- out from institutions and a referendum on the Bonn powers.68 In September 2009, Dodik again threatened to remove Serb politicians from the state institutions should the high representative seek to impose further measures. 69 In the context of the Butmir pro cess, he made progress conditional “on agreement the RS has the right to secede aft er a three- year waiting period.”70 And more recently, the RS National Assembly passed a resolution approving a referendum on the state court and the OHR’s authority. It was only with a visit to Banja Luka in May 2011 by the EU’s High Representative for Foreign Aff airs Catherine Ashton that Dodik backed down and the situation was (at least temporarily) diff used. Clearly, Dodik has been in a strong position of power, both within the RS and at the state level. As the ICG comments, “No law can pass and no important government or diplomatic post be fi lled without his consent.”71 He resists any further transfer of competences from the RS to the state institutions and has sought transfer of competences in the opposite direction. Moreover, Bosnian Serbs see the statebuilding strategies of the international community as a threat to their interests and group rights. Dodik described the international policy thus: “Let’s pull the Serbs into the Dayton agreement, and then we will take away the right of Republika Srpska to exist, by talking about correcting Dayton.”72 It is in this context that Dodik has, on occasions, resorted to calling a referendum on inde pen dence. 73 In contrast, Bosniak politicians, particularly the SBiH and SDA, favor wholesale changes to the Dayton structures, a proposition that remains anathema to Bosnian Serbs. For instance, one SDA politician suggests that the state needs new territorial arrangements: new regions “based on socio- economic characteristics” rather than identity, which would necessitate a change in entity borders.74 Perhaps unsurprisingly, given their plurality status, Bosniak parties have been somewhat defensive of the power- sharing institutions at the state level. For instance, one politician acknowledged that although decision making can be extremely slow, it is important for power From Dayton to Brussels? 147 sharing to be inclusive and allow the parties from the contending groups to (eventually) reach a compromise.75 He stressed the need for mutual veto rules, though qualifying that the VNI needs to be refi ned further and concentrated in the House of Peoples and that entity voting in the state parliament must be abolished. Interviews for this research corroborate the view that eff ective power sharing is, indeed, hampered by the institutional rules under the DPA. Some external actors have been critical of the accumulated eff ect of rules designed to protect the constituent peoples, particularly the VNI, the tripartite presidency, and the bicameral system. For instance, the Venice Commission commented that there needs to be a balance “between the need to protect the interests of all constituent peoples on the one hand and the need for ef- fective government on the other,” but that the number of rules designed to do this “makes eff ective government extremely diffi cult, if not impossible.”76 According to Rycroft , the 2006–10 executive was hindered by both an unwillingness to accommodate on the part of domestic elites and their misuse of institutional rules, particularly the mutual veto, as “it is more important for parties to be able to say that they can stop others.”77 He suggests there is a need for both more eff ective power- sharing rules and greater po liti cal will. A key question, however, is why the EU integration process failed to incentivize po liti cal elites to compromise on decision making and constitutional reform.

The Limitations of EU Incentives

Th e EU’s Stabilization and Association process for the Western Balkans was established as “the framework for EU relations with the region” and the “longer term perspective within which the prospects for proposed closer relations with the countries of the region should be seen.”78 In Bosnia, the launch of the process in 2000 led the Euro pe an Commission to set a number of policy conditions for securing the country’s path to eventual membership. In March 2000, EU Commissioner Chris Patten published a “road map” outlining fourteen steps that would need to be implemented in advance of a feasibility study for a SAA. In November 2003, the Euro pe an Commission announced the successful completion of the feasibility study and the potential start of negotiations on the SAA. In its list of recommendations, the Euro pe an Com- mission called on Bosnia to deliver fundamental reform, including “more 148 Chapter 6 eff ective government at state level.”79 SAA negotiations were opened with Bosnia in November 2005. Discussed above, police reform was a central condition for the SAA, eventually signed in June 2008. Th e EU has also pushed for more eff ective po liti cal institutions, with the Eu ro pe an Commis- sion emphasizing the need for constitutional reform so that Bosnia can handle effi ciently its preparations for the accession pro cess.80 As we have seen, however, EU carrots on constitutional reform have not been particularly eff ective. Arguably, there are several reasons that help explain the limitations of EU incentives: an exaggerated faith in the potential for EU conditionality to overcome deep divisions in a post- confl ict situation, a lack of explicit requirements, and a gap between EU strategy and the ongoing divergent positions of the contending groups on state structures. First, some of the diffi culties arguably relate to the EU’s lack of experience in managing a deeply divided, post- confl ict territory such as Bosnia in the enlargement process. Th e EU has found itself trying to work through a set of challenges in the Western Balkans that are wholly diff erent from the challenges the or ga ni za tion faced in post- communist Central and Eastern Eu rope in the 1990s. Although both regions have had similar challenges moving from a post-socialist transition to a liberal market democracy, the Western Balkans face additional challenges from the collapse of the former Yugoslavia. And within the post-Yugoslavia successor states, Bosnia has had to face its own partic u lar challenges, given the deep divisions following a three-year war, during which around 100,000 people lost their lives and half the population became refugees or internally displaced. Arguably, the EU enlargement strategy applied to Central and Eastern Eu rope is not simply transferable to this post- confl ict situation. EU conditionality, as we have known it elsewhere, is seriously stretched when dealing with the legacy of violent ethnic confl ict. Th us, the EU needs a greater understanding of the post-confl ict dynamics and challenges in deeply divided places, and the enlargement strategy needs to refl ect this understanding. Second, the in eff ec tive ness of EU incentives in bringing about a reform of the Dayton structures can also be explained by the fact that the EU carrots have been somewhat vague, or at least not particularly explicit, and therefore open to misinterpretation by local politi cal parties. Th e EU has insisted that constitutional reform is needed to improve the “functionality” of the political institutions so that Bosnia can manage the integration process and the transposition of the acquis communautaire. But as negotiations on the acquis remain a long way off , parties are less convinced of a sense of urgency in From Dayton to Brussels? 149 agreeing to reform. Arguably, there has been no real impetus for constitutional reform because accession is far from pending. Moreover, domestic actors may see greater short- term electoral benefi ts in adopting a more hard- line ethnic position that resists EU pressure for reform. Sebastián suggests that the EU has lost credibility in Bosnia “primarily due to the lack of consistent strategy and a lengthened pro cess involving multiple stages— each lacking clear benefi ts.”81 Interestingly, some international offi cials agree that the EU has been insuffi ciently clear on what is needed, but they also maintain that EU conditions should work out in accession negotiations.82 Some domestic elites are also critical of the lack of guidance from the EU on the specifi cs of constitutional reform.83 As domestic elites are less than clear on what kind of reform is required, they can interpret freely EU actions and messages to suit their own politi cal agenda. Th e EU’s lack of specifi c guidance on constitutional reform can be explained by the lack of an acquis on constitutional arrangements. Th e lack of guidance on state structures is unsurprising, given that the EU comprises member states with varying degrees of decentralization. Th us, as one EU offi cial put it, the Euro pe an Commission “cannot dictate constitutional reform” and can at best just stress the need to improve the “functionality” of the power-sharing system.84 Th ough the Eu ro pe an Commission has stated that reform “to permit [BiH] institutions to function eff ectively is necessary before the Commission can recommend the granting of candidate status,” the EU delegation professes it can only insist that centralized and decentralized structures can complement one another, provided the system works ef- fectively.85 Nevertheless, divergent preferences among the groups make eff orts to “improve functionality,” never mind wholesale constitutional change, a serious challenge. Even in the aft ermath of the ECtHR ruling in December 2009, which found the constitutional restrictions upon “others” to be in violation of the ECHR, progress remained elusive. An EU offi cial suggests that the parties have dragged their heels on complying with the ECHR ruling because it would open a “Pandora’s box” necessitating major revision of the Dayton structures.86 Th ird, the eff ectiveness of EU incentives has been limited due to (some) domestic actors’ perception that the integration process will rob the entities of their power. Bosnian Serb parties have been opposed to strengthening the state institutions to the detriment of the RS. Indeed, evidence for this research suggests that Bosnian Serbs, particularly the SNSD, have sought to position the RS as the most effi cient level of governance to deal with EU 150 Chapter 6 integration. An OHR offi cial describes a “troubling” situation in which Dodik has claimed the RS could deal with the EU’s acquis directly and would be able to “ram through legislation,” whereas dealing with the state would hold up the process. 87 Th e EU’s insistence on capacity at the state level to manage the integration process is supported by the Bosniak parties, who would ultimately like to see the international community weaken the role of the entities. In turn, the EU’s position has played into fairly adverse relations among the groups, particularly between Bosniaks and Bosnian Serbs. As Sebastián comments, though EU integration is a shared goal among local actors, “the substance and pro cess of EU integration has become a divisive factor.”88 Gergana Noutcheva similarly argues that “the EU’s sovereignty-linked demands, in fact, reproduce divisions” among internal actors and “further entrench politi cal disagreements instead of bringing about a domestic consensus around an EU agenda which no politi cal actor can plausibly claim to be against.”89 Th ese diffi culties help explain why BiH reform has been, at best, “partially compliant” with EU conditionality.90 Overall, the in eff ec tive ness of EU incentives in Bosnia posed a considerable problem as the international community moved toward closing the OHR and handing over the reins to the EU as the lead external actor.

Shifts in External Engagement: From OHR to “Reinforced” EU Presence

Since the PIC’s announcement in June 2006, external actors have been working toward a politi cal situation in Bosnia judged appropriate for the closure of the OHR. On several occasions the PIC decided to postpone OHR closure, referring to the ongoing politi cal instability in the country and thereby necessitating sustained international engagement with executive powers. Th e extension of the OHR’s mandate led to an extended period of transition from the OHR to the EUSR as the lead international actor. Arguably, the intention to close the OHR derived from a belief that the offi ce was increasingly in eff ec tive in securing peace implementation and helping Bosnia function as a sustainable state. Discussed above, the operation of the Bonn powers has been subject to considerable critique and, in recent years, has been the source of increasing antagonism between the OHR and the local elites, particularly the Bosnian Serb parties. In light of these growing tensions between the OHR and local elites, the ICG refers to From Dayton to Brussels? 151 the former’s “growing in eff ec tive ness due to its loss of credibility.” 91 Th at elites threatened a backlash against OHR actions underlines this hemor- rhaging of the institution’s credibility over time. Th e ICG comments that the OHR “has become more a part of Bosnia’s politi cal disputes than a facilitator of solutions, and the High Representative’s executive (Bonn) powers are no longer eff ective.” 92 It is unsurprising that the PIC has deemed it time to change international policy in Bosnia, especially in the context of the EU integration pro cess. Th ere has also been variation on the part of successive high representatives in relation to the operation of the Bonn powers. Arguably, this shift or inconsistency of approach has helped fuel criticism of the OHR’s record in statebuilding. Th e previous chapter pointed to the increased reliance on the Bonn powers by the high representatives in the early post- Dayton years, from Westendorp, Petritsch, and Ashdown, the latter judged to have been the most interventionist. It has been noted, however, that Ashdown reduced his reliance on the Bonn powers toward the end of his tenure. Ashdown’s successor, Schwarz- Schilling, adopted a wholly diff erent approach by stating that he would not use the Bonn powers except in securing peace and security and Bosnia’s cooperation with the ICTY. According to the ICG, the “abrupt style change” that came with Schwarz-Schilling’s tenure “threw the OHR into turmoil.” 93 Rather than phasing out the use of the Bonn powers, this was to be an immediate shift , arguably leaving the OHR appearing somewhat rudderless in dealing with challenges to the DPA and the lack of progress on policy reform. Th e ICG is highly critical of Schwarz- Schilling’s tenure, commenting that his “calculated hands-off approach not only failed to stimulate Bosnian politicians to take responsibility for speeding up the process and giving it deeper roots; it also put at risk the credibility of the mechanism, probably beyond repair.” 94 Some international offi cials working for the OHR essentially concurred with this assessment and were critical of Schwarz- Schilling’s approach. One OHR offi cial suggested that Schwarz- Schilling “made big mistakes” in not using the Bonn powers too readily and eff ectively “let Dodik get away with it.” 95 With this lack of consistency in strategy among the high representatives, it is perhaps unsurprising that the Bonn powers became increasingly contentious and were called into question. Following Schwarz- Schilling’s rather brief tenure, Lajčák carved a path somewhere in between his two prede ces sors. On assuming offi ce, he declared that he would be ready to use the Bonn powers as a tool available under his mandate while also stressing that painful compromises would 152 Chapter 6 necessarily need to be achieved among the parties, a preferable option to “someone from the outside deciding for you.” 96 Lajčák did resort to using the Bonn powers, including the controversial decision discussed above, the 2007 Law on the Council of Ministers, which led to confrontation with the SNSD. Following this crisis, Lajčák avoided using the Bonn powers. Matthew Parish writes that the OHR then became “ever more irrelevant, its staff akin to media celebrities, giving interviews and making pronouncements, but with no real power.” 97 On leaving his post unexpectedly in January 2009, Lajčák said he did not want to be “the rider on a dead horse,” referring to the ineff ective ness of his offi ce.98 Lajčák was replaced by High Representative Valentin Inzko, who intimated that the Bonn powers would be used as a last resort. He summed up his strategy by saying that he would not hesitate to use the Bonn powers should eff orts at negotiation fail: “I would first use six steps to decision making. Th e fi rst is to listen, the second to have a talk, the third, to negotiate. Th e fourth is warning, that is ‘yellow card,’ and the fi ft h to try to convince in four eyes. Only at the end I would reach for ‘the Bonn Powers’ and remove a person.” 99 As per the challenges experienced by Lajčák, Inzko encountered considerable opposition from the SNSD following his use of the Bonn powers. When Inzko imposed a series of laws on 18 Septem- ber 2009, Dodik responded by threatening a Bosnian Serb boycott from state institutions should the OHR impose any further measures. Th e experiences of the successive high representatives demonstrate a balancing act between the facilitation of inter- party negotiation and their mandate to take executive decisions. Th e shift s from encouraging progress to imposing decisions points to some tension on whether the Bonn powers were an appropriate mechanism for externally led peace implementation. Th e increased in eff ec tive ness of the Bonn powers has helped inform the EU’s preparations for the transition to the EUSR as the lead international actor in Bosnia. In its intention to move away from international involvement with executive powers, the EU speaks instead of a “reinforced EU engagement” to give the orga ni za tion “unique leverage and responsibility.”100 In terms of the mechanisms for leverage, the EU envisages mobilizing “all the possible EU instruments with consistency and determination.” Th e Eu- rope an Partnership and SAA implementation are pitched as the “drivers of reform.” In an eff ort to clarify the EU’s position on reforming the Dayton structures, constitutional reform is considered “neither a requirement for OHR closure nor for BiH’s further journey toward the EU.” At the same time, and arguably adding to some confusion, the EU states that the “consti- From Dayton to Brussels? 153 tutional framework must evolve to ensure eff ective state structures capable of delivering on EU integration, including the requirement to speak with one voice.” As the transition from OHR to EUSR was extended, arguably a gap arose between the two organizations on the international community’s future role and authority in the country. Some OHR offi cials are critical of the EU’s faith in the carrot of membership as an incentive to persuade domestic elites to adopt policy reforms.101 One OHR offi cial commented that the EU needs to be more “prescriptive” and adopt more robust conditionality in securing reform.102 He claims that with the OHR being told it had to “go out of business” and the EU increasingly in charge since 2006, “the results are what they are.” Of note is Ashdown’s criticism of how the EU handled Dodik’s threat in May 2011 of a referendum the following month on the state court and the powers of the OHR. Following a visit of the EU’s High Repre- sentative for Foreign Aff airs Baroness Ashton to Banja Luka, Ashdown chose to contrast the achievements of the OHR in the fi rst de cade post- Dayton with the subsequent challenges faced by the EU. He stated, “For ten years Bosnia was the poster boy for international relations, making steady and sometimes miraculous progress toward statehood, capable of joining the EU and NATO.” Ashdown claimed that since 2006, the dynamic has been “moving in the wrong direction” and that if the situation is not reversed, there is real threat of renewed confl ict. He was reportedly “livid” at Ashton’s visit to Dodik, suggesting that securing the Bosnian Serbs’ about- turn on referendum plans was “bought at a price which we shouldn’t have paid.” 103 Interestingly, some politi cal elites (particularly from the Bosniak community) comment that maintaining the OHR with robust powers is necessary as BiH moves closer to EU integration. One SDA politician refers to OHR intervention as a guarantor against other parties’ intention “to take Bosnia back into the past.” He suggests that the OHR is “the only thing politicians are afraid of,” and should the OHR close, there could be nobody left to stop the transfer of powers from state to entity level.104 One SBiH politician warns that a “complete blockage” of the system would take place following closure of the OHR. He claims that because the OHR is the “only real authority in terms of the constitution,” it should postpone closure until constitutional reform has been agreed upon, including the abolition of entity voting.105 In contrast, the Bosnian Serb parties have rejected imposition by the OHR and called for its closure. Th e gap between the OHR and the EU on the way forward left the OHR somewhat in limbo. One OHR offi cial suggests that the transition left the 154 Chapter 6

OHR “holding in line,” unsure how to proceed.106 For the wider international community, the closure of the OHR is viewed as an essential development in securing Bosnia’s progress toward EU integration. For instance, EU member states of the PIC have inferred that “the EU would not be in a position to consider an application for membership by BiH until the transition of the OHR to a reinforced EU presence has been decided.”107 Th e challenges faced by the OHR are arguably compounded by a degree of incoherence among PIC members. Th e overall view is that it would be dangerous to close the OHR prematurely, hence the condition of a stable po liti cal environment. At the same time, it is well known that Rus sia has diffi culty with the Bonn powers and an open-ended transition. International actors on the ground claim that Russia would never allow the Bonn powers to be reinstated should the OHR close and there be future need for increased international engagement in Bosnia. An international offi cial claims that on the PIC Steering Board, there are two camps on the OHR- EUSR transition.108 First, calls for closure of the OHR by Russia are backed by Germany, France, and Italy. In contrast, the U.S. reportedly preferred to delay the OHR- EU transition (and retain the Bonn powers) for as long as possible, until the EU can demonstrate eff ective management of the accession pro cess.109 Th is preference to retain the Bonn powers corresponds to a view that if the international community is unable to respond to challenges to the DPA in a robust manner, the RS will be incentivized to move toward secession. Th e U.S.’s skeptical approach to the transition is reportedly backed up by the UK, Turkey, Japan, and Can- ada. Th ese diff ering viewpoints suggest a tension on the future of international engagement in Bosnia and a measure of incoherence among the PIC Steering Board on the details of the OHR- EUSR transition. Whenever the PIC decides the time is right for OHR closure, the “reinforced” EU presence will fi nd itself needing to demonstrate that it can guide the country’s preparations for accession. But what might a “reinforced” EUSR look like? An EU offi cial claims that the EUSR will be reinforced by merging with the EU delegation, with one contact point and one high-level offi cial who would be both head of the delegation and the EUSR.110 Th e ICG suggests that the EU presence “will need to make up for the lack of Bonn powers by using much more assertively the real powers of the purse with which Brussels should equip the offi ce.”111 Th e ICG has suggested the withholding of funds, visa bans, freezing of assets, and even the suspension of the SAA.112 Of note is the decision by the Council of the Eu ro pe an Union to adopt re- From Dayton to Brussels? 155 strictive measures in view of the situation in Bosnia. Th e Council agreed in March 2001 that mea sures should be imposed against individuals “whose activities undermine the sovereignty, territorial integrity, constitutional order, and international personality” of the country, “seriously threaten the security situation,” or undermine the DPA.113 Th ese measures include visa bans and the freezing of assets. Beyond these negative incentives, the EU clearly needs to convince the local parties that they have something to gain in the unfolding pro cess of accession. Th e danger is that should parties calculate it is more benefi cial to resist the EU carrot of membership, the pro cess will remain blocked.

Summary

Th e experience of power sharing in Bosnia highlights the limitations of institutional rules in deeply divided places. Such rules (in this case, a strongly corporate form of consociational power sharing) cannot do much to foster inter- ethnic cooperation when elites are unwilling to operate them eff ectively. Th e politi cal reality is that Bosnian Serb elites have preferred to obstruct power sharing at the state level to bolster the RS. In contrast, Bosniak leaders have sought to pull external actors in the opposite direction toward supporting greater centralization and, ultimately, the abolition of the entities. Th ese divergent positions have been compounded by international efforts to secure constitutional reform that would “move beyond Dayton.” In this highly sensitive context, cooperation has been in short supply. Argu- ably, no matter what the institutional rules are, power sharing is unlikely to be eff ective when leaders are antagonistic and po liti cal preferences remain unaccommodated by state structures established by a peace agreement lacking legitimacy. Po liti cal elites may choose antagonism over accommodation when the former is electorally advantageous, to cement their position of power as the rightful voice of their community, and will resist mea sures they view as a threat to their power and group rights. As argued in the previous chapter, power- sharing institutions (backed by external actors) managed to hold the Bosnian state together, but they have been unable to foster much inter- ethnic cooperation. Th e Bosnian case also points to how power- sharing rules are subject to change. Here, institutional change has been the result of imposition by the OHR in an eff ort to improve decision-making 156 Chapter 6 structures. And as discussed, external actors have called for comprehensive constitutional reform, perhaps somewhat optimistically described by the EU as “constitutional evolution.” Overall, the role of external actors has been a crucial variable in the operation of the post-Dayton system. External actors have sought to induce politi cal parties to agree to policy reform by using positive incentives (the carrot of EU membership), negative incentives (warnings of isolation), and even coercive incentives (removals from offi ce and imposed decisions) via the Bonn powers. Arguably, this incentives structure has been in eff ective. Th e Bonn powers have been increasingly contentious, and the carrot of EU membership has been insuffi ciently attractive to compel elites to compromise on policy reform. External incentives will likely be constrained when the peace agreement remains contested and international actors are perceived as pushing a structural reform agenda against the preferences of the contending groups. Moreover, the ineff ective ness of external incentives has been heightened by inconsistencies in the use of the Bonn powers and by the gap between the OHR and the EU over the transition and nature of international engagement in Bosnia. CHAPTER 7

Macedonia: From In de pen dence to the Ohrid Framework Agreement

Th e Macedonian case is notable for the practice of liberal executive power sharing following inde pen dence in 1991, without being mandated by the new state’s constitution. Despite the ac cep tance of power sharing as a central feature of the new politi cal system, cross-community government was arguably insuffi cient to stem growing ethnic tensions between the Macedo- nian and Albanian communities. By the end of the 1990s, Albanian grievances had become increasingly salient, and fuelled by the Kosovo crisis of 1999, led to a six-month confl ict in 2001. In this chapter, I provide an assessment of power sharing in the new Macedonian state, a discussion of the 2001 confl ict, analysis of the Ohrid Framework Agreement (OFA) in cementing power sharing, and evaluation of the eff orts on the part of coali tion parties to implement reform up to 2006. I also explore the role of international actors, particularly the EU and the U.S., in bringing about the power- sharing settlement and persuading elites to implement its mea sures.

Post- Independence Power Sharing

Post- independence Macedonia is interesting for power- sharing practice and theory, given elites’ preparedness to form cross- community co ali tions without being constitutionally or legally obliged to do so. Th e evidence suggests that Macedonian parties simply opted to form governments with Albanian partners as a rational decision to help improve the stability of the new state. Th e case shows that informal, liberal power sharing can become a convention in a multi- ethnic state, not necessarily as the outcome of a confl ict and 160 Chapter 7 subsequent peace settlement. Th e case also demonstrates, however, that a multi- ethnic or plurinational state may require further institutional design to accommodate groups and secure stability. As discussed below, the Alba- nian community felt increasingly frustrated by their representatives’ inability to secure additional group rights in important areas including language, education, and decentralization. So how important was executive power sharing for the operation of in depen dent Macedonia post- 1991? And why was power sharing unable to prevent growing inter-ethnic tensions and ultimately the outbreak of confl ict in 2001? Going back to the 1980s, pro-democracy movements representing each of the two main communities had gathered momentum. Importantly, however, Macedonians and Albanians had diff erent views on what a democratic Macedonian state might look like. As Jenny Engström writes, “Whereas the Macedonian movement was mainly oriented toward politi cal and economic transition, the Albanian movement regarded the integration of Alba- nians into Macedonia’s politi cal institutions as its main objective.”1 For Albanians, this integration would mean securing group rights and protecting their identity. Whereas Macedonians sought to emphasize individual rights, the Albanian community called for “consensus building and group rights” to redress the potentially adverse eff ects of majoritarianism in a multi-ethnic state.2 When a referendum on in de pen dence was held on 8 September 1991, Albanians largely boycotted the vote; around 95 percent of those who voted supported inde pen dence. Th e Albanian community’s boycott can be explained by a fear that they would be excluded from the new po liti cal system and denied group rights. Engström suggests that Albanians rejected the referendum question that asked for support for Macedonia to rejoin a federal Yugoslav state because of fear “that the Albanian people would become victims of a regime dominated by Serbia and Macedonia” and because of their dissatisfaction with the lack of clarifi cation of their status in an in de pen dent Macedonian state.3 Following the referendum on in de pen dence, a new constitution was draft ed quickly and adopted on 17 November 1991. Arguably, the 1991 constitution failed to lessen Albanian fears about their status in the new state and may even have heightened their misgivings. Notably, the constitution’s preamble stated that “Macedonia is established as a national state of the Macedonian people, in which full equality as citizens and permanent co-existence with the Macedonian people is provided for Albanians, Turks, Vlachs, Romanies, and other nationalities living in the Republic of Macedo- Macedonia 161 nia.” 4 By stating explicitly the privileged position of Macedonians as the “titular” nation, the constitution stoked existing Albanian reservations that they would be relegated to a second-class “nationality” alongside the other smaller minorities. Engström notes the Albanian view that classifying their community as a nationality “eff ectively downgraded them to the ranks of second-class citizens by putting them on a par with the much smaller Turkish, Roma, and Vlach communities.” 5 Given that the Albanian community was larger than the other nationalities, Albanian elites responded by calling for their community to be recognized as a constituent nation. Th e Albanian community also had diffi culty with Article 7 of the constitution, which states that the Macedonian language, written in the Cyrillic alphabet, is the offi cial language at the state level. Also under Article 7, in local self-government, a nationality composed of either a majority or a “considerable number” of the local population can use its language and alphabet in offi cial use, alongside the Macedonian language. Albanians sought to specify the threshold of the nationality’s population at the local level, with reform later achieved as part of power sharing. Article 48 of the constitution provided nationalities with “a right freely to express, foster, and develop their identity and national attributes.” Th e state would also guarantee “the protection of the ethnic, cultural, linguistic, and religious identity of the nationalities,” including the right to establish cultural institutions. Notably, Article 48 also provided members of the nationalities with the right to primary and secondary education in their language, in addition to the study of the Macedo- nian language. But despite such nods to minority rights, the Albanian community had reservations about the new constitution, given the absence of group rights explicitly for Albanians. Rather than being accorded rights for “members of the nationalities,” Albanian representatives increasingly called for the recognition of the Albanian community as a constituent nation, on equal terms with Macedonians. Despite the decision by Albanian representatives to abstain on the parliamentary vote, the constitution was passed with majority Macedonian support. Although the po liti cal parties were not constitutionally obliged to share power, the period since inde pen dence was marked by the accep tance among po liti cal elites, and presumably the electorate, that government would include both Macedonian and Albanian po liti cal parties. Th is development was a refl ection of the po liti cal reality and, arguably, some acquiescence on the part of Macedonian elites that power sharing was necessary in order to secure some stability for the new state. Th roughout this period, power sharing 162 Chapter 7 coali tions were formed by one of the two main Macedonian parties—the Social Demo cratic Alliance of Macedonia (SDSM; the successors of the League of Communists of Macedonia) or the Internal Macedonian Revolu- tionary Organization- Democratic Party for Macedonian National Unity (VMRO-DPMNE)—and one of the two main Albanian parties, the Party for Demo cratic Prosperity (PDP) or the Demo cratic Party of Albanians (DPA), though sometimes joined by smaller parties. As no party or bloc won a majority in the fi rst demo cratic elections in November 1990, an “expert government” was formed, headed by Nikola Kljusev, an economics professor, and guided by the new president, Kiro Gligorov, who was elected at the end of January 1991. Th ough the government of “experts” helped cement a relatively smooth transition, it was ultimately subject to a vote of no confi dence in parliament and forced to resign in 1992. Aft er consulting fi rst with VMRO- DPMNE, Gligorov asked the SDSM to form a new government, which it did with the Reformed Forces-Liberal Party (RS-LP); the Socialists; and two Al- banian parties, the PDP and the People’s Democratic Party (NDP).6 Argu- ably, the SDSM elites calculated the benefi ts of reaching out to their Albanian counterparts as an appropriate stabilizing mea sure in the country’s transition to democracy. According to Engström, the SDSM invited the PDP to share power “to alleviate some of the mistrust” between the two communities.7 Th us, from the beginnings of the new state, Albanian politi cal grievances were given a platform and became increasingly salient throughout the 1990s. Andrew Rossos suggests this new co ali tion “brought a greater semblance of security and stability to the country” and “worked together to resolve some of the issues that divided Macedonians and Albanians.”8 Despite these early positive signs in the country’s transition to democratic politics, the October 1994 parliamentary and presidential elections were marred by contention over voting lists, leading to the boycott of the second round of voting by VMRO-DPMNE and the Democratic Party. Elec- tion monitors reportedly recorded a situation of administrative incompe- tence, describing the voter lists as “incomplete and inaccurate” and polling offi cials as “inadequately trained.” 9 With VMRO- DPMNE and the Demo- cratic Party failing to win any seats in parliament due to their boycott, a power- sharing government was formed by an alliance of parties led by SDSM plus the Albanian party, the PDP. Th e latter was the main Albanian party in the period since in de pen dence and sought to pursue Albanian goals by participating in the politi cal system and negotiating executive formation with the SDSM.10 However, the record of the power- sharing Macedonia 163 government was not particularly impressive in managing Macedonian- Albanian relations. Th e International Crisis Group (ICG) refers to the PDP’s “frequent threats to leave the government” and notes the “uneasiness” and “delicate balancing act” within the coali tion. 11 Some progress was made, however, with the 1995 Law on Local Self- Government, whereby a minority language could be used at the municipal level when the minority constitutes at least 20 percent of the local population. For Albanians, this reform, to some extent, addressed one of the shortcomings of the 1991 constitution, which stated that a “majority” or “considerable number” of the local population could use their own language offi cially. Th e 1996 Law on Territorial Division increased the number of municipalities from 34 to 123, in order to improve access to local institutions in a highly centralized state. As Eben Friedman notes, however, the legislation was subject to criticism from Albanian leaders because the law “had the overall eff ect of making the urban municipalities with the largest Albanian populations less Albanian and more Macedonian.”12 Th e redis- tricting of municipal boundaries was to be a source of ongoing disagreement among Macedonian and Albanian parties. Addressing another issue of considerable sensitivity, the 1997 Law on the Use of Flags and Symbols gave minorities the right to use their fl ag in private celebrations as well as during cultural, sports, and other occasions.13 Yet restrictions applied prohibiting the use of the minority fl ag outside state institutions. Th e 1998 parliamentary elections produced the fi rst signifi cant change in power- sharing government since in de pen dence: On the Macedonian side, the SDSM lost to the alliance between the VMRO-DPMNE and the Demo- cratic Alliance (DA). Th is alliance then formed a power-sharing government with the DPA, the more nationalist Albanian party compared to the PDP, which had previously been in power with the SDSM. Th e new co ali tion enjoyed an overall majority of 73 seats and was approved by parliament on 30 November 1998. Th e leader of VMRO-DPMNE, Ljubco Georgievski, explained his party’s decision to include the DPA: to “prove [to] the international community that the Republic of Macedonia has the fi rm determination to [abandon] Balkan standards and become a creator of stability in the region and beyond, and thus to stop being a subject of concern [to] the international community” and to “obtain complete [internal] stability and security.”14 Perhaps unsurprisingly, given the reputation of these two parties as the most nationalist or extreme within their respective blocs, there was some doubt as to the “viability and ability” of the new power- sharing 164 Chapter 7 government “to reconcile interests that are oft en confl icting and even mutually exclusive.”15 Despite such reservations, the power- sharing executive between VMRO- DPMNE, DA, and DPA made some policy achievements that addressed explicitly several sensitive inter- ethnic issues. Th e participation of the DPA in government achieved certain successes for the Albanian community: prisoner releases with the Amnesty Act, improvement in the positions and vis- ibility of Albanian ministers in government, increased hours of broadcasting in the Albanian language, and potential progress on the use of the Albanian language higher education.16 On 25 July 2000, parliament approved the Law on Higher Education that tackled the contentious issue of Tetovo Uni- versity. Founded by the Albanian community in December 1994, the university had been declared illegal by the Macedonian government; its premises had been raided by the police; and the university president, Fadil Sulejmani, had been arrested and sentenced to two-and- a-half years in prison for “in- citing resis tance.” 17 A highly sensitive issue for inter- ethnic relations in the new state, compromise on the status of the university had been avoided by the previous power- sharing government. Six years aft er the university’s creation, the 2000 legislation was based on a proposal by the Or ga ni za tion for Security and Cooperation in Europe (OSCE) High Commissioner on National Minorities, Max van der Stoel, whereby a new private university would replace Tetovo University (the new university’s working language would be Albanian as well as one or two foreign languages—French or German—and the curriculum would include mandatory studies in the Macedonian language). Th e issue was not resolved fully, however, as the PDP rejected the 2000 law and campaigned for state recognition of Tetovo University. Th e university was fi nally accorded public status in February 2004. Th e practice of power sharing in the period following Macedonia’s inde pen dence is notable for the inclusion of Albanian parties in government following post- election negotiations with the leading Macedonian party. Arguably, the inclusion of the Albanian parties alongside their Macedonian co ali tion partners had a fairly positive impact on the po liti cal system in the immediate post-independence period. As Engström writes, their participation “helped maintain a balance that worked in favour of inter-ethnic stability in the country.”18 Macedonian parties (particularly SDSM) made signals that power sharing was important for the country’s stability in the transition to democracy. Some mea sure of cooperation or cross- community “jointness” was evident in policy reform, including the use of the Albanian Macedonia 165 language at local government level, on amnesty legislation, and on Albanian- language higher education. Moreover, progress was made in relation to the country’s integration into Euro-Atlantic structures. Following a year of negotiations with the EU, Macedonia signed a Stabilization and Association Agreement with the EU on 9 April 2001.19 As discussed in the next chapter, Macedonia’s progress toward EU membership has been a central feature of post-Ohrid politics. Power sharing in Macedonia was challenged by, and survived, the Kosovo crisis of 1999. In practical terms, the state had to deal with the infl ux of hundreds of thousands of refugees who were accommodated by Albanian fami- lies and eventually provided for in refugee camps set up by the state. In politi cal terms, the confl ict in Kosovo aff ected already delicate intra-state Macedonian- Albanian relations. Although Albanians supported Kosovo’s struggle for liberation from the Federal Republic of Yugo slavia, Macedo- nians feared the Kosovar Albanians’ movement for in de pen dence would spill over the border and threaten the stability of the state. Exploring the eff ects of the Kosovo crisis, Alex J. Bellamy notes the growth of the Kosovo Liberation Army (KLA) activity inside Macedonia, including fundraising and recruitment. He also notes the “complex relationship between po liti cal activism and criminality,” which ultimately led to the outbreak of violence in Macedonia in 2001.20 Th ough power sharing had become an important politi cal convention in post-independence Macedonia, this institutional success was clearly not suffi cient for accommodating group preferences. Indeed, the case points to the limitations of executive power sharing without additional group protection measures. Albanians were increasingly of the opinion that their identity and culture were not adequately protected in the new state. Such protection would require additional constitutional and institutional reform. Moreover, in the context of the Kosovar Albanians’ struggle across the border, the Al- banian community in Macedonia continued to highlight a number of grievances and claimed that the successive coali tion governments were slow to tackle Albanian problems. Engström writes that the “absence of signifi cant improvements” for the Albanian community, in turn, increased ethnic tension and the minority group’s “sense of exclusion.”21 Albanians continued to be underrepresented in public administration, and the police and army were “almost exclusively staff ed with ethnic Macedonians.”22 Th ey were more likely to have lower levels of education (the majority of whom left formal education following primary school) and were more likely to be unemployed. 166 Chapter 7

And the Albanian community continued to call for the right to use their own language in parliament and the court system. For many, the issue of Albanian- language higher education remained unresolved, as Tetovo Uni- versity had not been accorded recognition as a public university with state funding. Th roughout the 1990s, the grievances of the Albanian community became more salient. Despite the country’s apparent success story in avoiding the confl ict that befell other post-Yugoslav states, relations between the two communities had been deteriorating. Noting the ongoing segregation between the two groups, Farimah Daft ary and Eben Friedman write that a de cade aft er inde pen dence, “the country’s social and politi cal landscape was marked throughout by ethnic divisions, with ethnic Macedonians and ethnic Albanians living in parallel societies, such that there was little communication and interaction between ordinary people from diff erent ethnic backgrounds.”23 Ultimately, the moderate eff orts made in improving ethnic relations by the successive power- sharing co ali tions were overtaken by the security crisis in 2001, leading to six months of violent confl ict between the Macedonian security forces and the ethnic Albanian rebel group, the Na- tional Liberation Army (NLA).

From Violent Conﬂ ict to the Ohrid Framework Agreement

In February 2001, fi ghting between Macedonian border police and the NLA broke out on the border with Kosovo and soon spread to other parts of Mace- donia. As debates on the origins of the fi ghting grew, it was clear that the NLA originated from both Kosovo and Macedonia. Bellamy suggests that the NLA’s relationship with the KLA is diffi cult to defi ne, but he confi rms that the two organizations included some of the same people (e.g., Ali Ah- meti) and that the NLA can be considered a “re- badged” KLA.24 In terms of strategy, Bellamy suggests that Albanians from Macedonia who fought for the KLA “saw the effi cacy of force in limited and specifi c ways.”25 Owing to these diff erent infl uences, there was some confusion and debate over the NLA’s objectives. For the Macedonian state, the ultimate goal of the NLA was about claiming territory on behalf of Albanians and moving toward secession. Another viewpoint stressed the movement’s objectives in securing additional group rights for Albanians within the existing Macedonian state. As Daft ary and Friedman write, “While the NLA claimed to be fi ghting for the improvement of the rights of ethnic Albanians in Macedonia, the Macedonian Macedonia 167 government maintained that the NLA was working toward secession or a ‘Greater Albania.’ ”26 Polls at the time confi rmed the “strong sympathies” among Albanians for the NLA. Th e NLA’s objectives were perceived by supporters to be about “fi ghting for an improvement of ethnic Albanian rights as well as a greater say in the aff airs of the state, rather than any territorial claim.”27 Other scholars suggest NLA goals were about greater group rights for Albanians plus embarking on a journey toward integration with Kosovo. John Phillips suggests that because the DPA in government had failed to make much progress in securing additional rights for the Albanian community, rebels who had been active in the KLA “made plans for an ethnic Alba- nian insurrection in western Macedonia to achieve greater rights for the population and eventually join an inde pen dent Kosovo.”28 It appears, however, that reference to secession, a Greater Albania, or joining with an inde- pen dent Kosovo was progressively sidelined in preference to additional group rights within the existing state. Th e NLA’s goals were, arguably, not far removed from the objectives of Albanian po liti cal parties: the recognition of Albanians as an equal constituent nation, establishing Albanian as an offi cial language alongside Macedonian, and increasing Albanian participation in po liti cal institutions.29 As a result of the lessons learned from the Balkan wars of the early 1990s, the outbreak of confl ict in Macedonia was followed by a fairly prompt response from the international community.30 Politi cally, the EU’s Javier Solana helped form a Government of National Unity that took power in May 2001. Th e previous coali tion, led by VMRO-DPMNE and the DPA, was joined by the other main Macedonian and Albanian parties— the SDSM, the PDP, the Liberal Demo cratic Party, and the Liberal Party— to form a grand coali tion. External actors hoped this grand coali tion would help quell the crisis and negotiate a resolution of the escalating confl ict. As Daft ary and Friedman recount, however, the government “was anything but unifi ed.”31 Phillips notes that the government was soon “in deep disarray aft er the Al- banian party leaders reached a putative secret peace deal with the NLA.”32 An agreement among Albanian parties and the NLA was brokered by the OSCE representative, Robert Frowick, and focused on a common Albanian agenda, including revisions to the constitution and amnesty for NLA combatants.33 Internally, much of the progress toward a peace deal was led by the country’s president, Boris Trajkovski, while the leader of VMRO-DPMNE, Ljubco Georgievski, was considered by Western diplomats to be “a barrier to a peace settlement,” owing to his increasingly hawkish stance.34 As the 168 Chapter 7 confl ict continued, the EU, the U.S., and NATO intensifi ed their eff orts to achieve a permanent ceasefi re and a peace settlement acceptable to both sides. For several months, the international community sought to stop the violence and to persuade the two sides to reach a peace agreement that would include constitutional and legislative reforms to address the grievances of the Albanian community. EU and U.S. diplomatic eff orts were initially viewed as biased toward the position of the Albanian community but were more readily accepted by the Macedonian community as the peace talks process continued.35 Overall, the EU, the U.S., and NATO were crucial actors in bringing about an agreement between the two sides and an end to the confl ict. In June 2001, the U.S. appointed James Pardew from the State Department as a special envoy to Macedonia, and the EU appointed former French Defense Minister François Léotard as its representative. Pardew and Léotard were supported in their eff orts by the OSCE’s High Commissioner on National Minorities Max van der Stoel and NATO’s Peter Feith. Initially, talks led by President Trajkovski took place among the four main parties in the executive—VMRO-DPMNE, SDSM, DPA, and PDP.36 As the talks in Skopje failed to produce a deal, the EU and U.S. mediators decided to move the negotiations to the resort of Lake Ohrid in the southwest of the country. Th ough some progress had been made in the capital, diffi culties continued over some sensitive issues, including the use of the Albanian language and police reform. Th e negotiations at Ohrid lasted from 28 July to 9 August, and the offi cial signing of the peace deal took place at the residence of President Trajkovski in Skopje on 13 August. With regard to the external actors’ mediating role, it has been suggested Pardew delivered the Albanian parties in the negotiations while Léotard worked on persuading the Macedonians.37 It has also been noted that Léotard and Pardew were a “formidable double act,” playing a “soft cop, tough cop routine.”38 Th eir proposals for a peace settlement were called the “Framework Document,” which was drawn up with the help of French constitutional lawyer Robert Badinter.39 Th e peace deal arguably contained suffi cient incentives for each side and promised a straight victory for neither. Within the Macedonian community, there was some diff erence of opinion over the virtue of a military campaign versus po liti cal negotiations. As debates continued on the potential for the state to crush the rebels by military means, Georgievski was reportedly less amenable to compromise, while Brank Crvenkovski (SDSM) “believed that some modest changes to the constitution were inevitable if Macedonia 169 full-scale civil war was to be averted.” 40 Owing largely to pressure from the external actors involved, VMRO-DPMNE opted to engage in the peace talks as a means to protect the territorial integrity of the state. Once engaged in talks, the Macedonian parties insisted that the state should retain its unitary character and were, at best, prepared to discuss increased powers for decentralized structures. Th ough they were prepared to consider mea sures to improve minority rights, the Macedonian elites rejected Albanian demands that the state become “binational” with two recognized constituent nations (with equal recognition of their respective language and religion). By contrast, the two main Albanian parties called for wholesale changes to the constitution and the recognition of the Albanian community on an equal basis with Macedonians. In terms of incentives, Macedonian elites were prepared to support the OFA because it provided additional protection for the Albanian community without amending the state’s unitary character or af- fecting the privileged position of the majority group. Th e Albanian parties did not need much encouragement to support the deal because it promised considerable improvement in areas they had long campaigned for: the right to use their language in state institutions, police reform, and increased representa tion in public administration. Additional rights were secured for the Albanian community without making wholesale changes to the character or structure of the state. It was also important that the NLA, albeit not included at the talks, gave its support to the deal. NLA leader Ali Ahmeti reportedly suggested that without the deal, an extended confl ict could have had much more serious consequences: “If the Ohrid Agreement had not been signed, a new Bosnia would have unfolded with one or two thousand victims.” Th ough he referred to the deal’s residual “shortcomings,” Ahmeti welcomed the agreement as a new beginning “in which Macedonia has become a state for Albanians as well” and “a compromise in which Albanians are winners.” 41 Th e OFA set out several basic principles relating to a commitment to peaceful politi cal solutions, the inviolability of the state’s sovereignty and territorial integrity, the preservation of Macedonia’s multi-ethnic character, a commitment to international standards, and the development of local self- government.42 Under “cessation of hostilities,” the parties to the confl ict agreed on an unconditional and open-ended ceasefi re, complete and voluntary disarmament, and disbandment of the armed groups with NATO assistance. On decentralization, the parties committed to introduce legislation that would devolve enhanced competences to the local level as well as legislation on the fi nancing of local self-government. Th e boundaries of 170 Chapter 7 municipalities were also to be revised. Annex C stated that the assembly should adopt the revised Law on Local Self-Government within forty-fi ve days of signing the OFA and that a Law on Municipal Boundaries should be adopted by the end of 2002. Another central issue for the Albanian community included in the OFA was a commitment to reform in the area of “non-discrimination and equal represen ta tion.” Th e parties committed to adopt legislation on employment in public administration that would “include measures to assure equitable repre sen ta tion of communities in all central and local bodies,” with an obligation for authorities to “correct present imbalances in the composition of public administration, in par tic u lar through the recruitment of members of underrepresented communities.” Much of the discussion on equal represen ta tion during the talks centered on the issue of police reform, with the Albanian negotiators wanting much greater control at the local level. For the Macedonian parties, the Albanian preference that the police would refl ect the ethnic composition at the local level would mean increasing federalization of the state. A compromise was reached whereby the country’s police force would “generally refl ect the composition and distribution of the population of Macedonia” rather than of each municipality. To better refl ect the local population, local police chiefs were to be selected by municipal councils from a list prepared by the Ministry of Interior. Under the OFA’s Annex C “Implementation and Confi dence-Building Measures,” reform relating to the ethnic composition of the police force was to take place by 2004 and included an invitation to the U.S., EU, and OSCE to advise on the appropriate measures. On the security side, NATO conducted negotiations with Macedonian President Trajkovski and the NLA. Th e promised amnesty for NLA fi ghters who turned in their weapons was an important incentive for voluntary disarmament. Th e ICG reports that it was “NATO’s vigorous po liti cal intervention that proved decisive” in bringing the fi ghting to an end.43 Th ough NATO’s Secretary- General Lord Robertson had reportedly referred to the NLA as “thugs” at the start of the 2001 confl i c t , h i s o r g a n i z a t i o n d e v e l o p e d good relations with Ali Ahmeti, who had been keen to secure international intervention. NATO’s Operation Essential Harvest consisted of 3,500 troops to supervise the collections of arms from the NLA in August and September 2001. Led by the UK’s 2nd Battalion Parachute Regiment, the operation included a company of Gurkhas as well as troops from Eu ro pe an countries and logistics troops from the U.S. Phillips suggests that the NATO opera- Macedonia 171 tion was “a dramatic and highly unusual, if not unique, operation.” 44 Th e end of Operation Essential Harvest on September 26 was followed by a request from the government that some NATO peacekeepers remain in the country. As a follow- up, NATO’s Operation Amber Fox consisted of 750 personnel to support the Macedonian security ser vices.45

The OFA and Power Sharing: Cementing a Liberal Consociation

As discussed in Chapter 1, Macedonia has been described as a case of “complex consociation” due to the creation of institutional arrangements to settle a self- determination dispute; the development of a peace pro cess; an additional strategy such as territorial autonomy/decentralization; and the involvement of international actors in the implementation of the agreement.46 Th e analysis presented here also confi rms that Macedonia has been an instance of liberal, complex consociation. Th e liberal nature of power sharing is evident in the lack of predetermination in executive formation, both pre- and post- Ohrid.47 Further to the evolution of power sharing as a convention post- independence, the OFA served to cement power sharing as the most appropriate po liti cal system of government for the state. Without a formal power-sharing requirement (either by the predetermination of positions, as in Bosnia, or via sequential portfolio allocation, as in Northern Ireland), power sharing in Macedonia is considerably more fl exible and provides a greater opportunity for parties to form pre- and post-electoral alliances. In practice, parties have largely preferred to leave the cross- community confi guration of a new government to post- electoral negotiations. Th e introduction of the concurrent majority voting procedure in parliament was perhaps the most important reform relating to power sharing. Under Annex A “Constitutional Amendments,” laws that “directly aff ect culture, use of language, education, personal documentation, and use of symbols” must be adopted by a majority vote in the assembly, including majority support of representatives not in the majority community. Should a dispute occur in the assembly on the application of this procedure, that matter is to be resolved by the Committee on Inter- Community Relations.48 Known locally as the “Badinter principle,” owing to the input of Robert Bad- inter in draft ing the agreement, the rule is essentially the concurrent majority rule advocated by John C. Calhoun (U.S. vice-president 1824–32) to 172 Chapter 7 allow (pro- slavery) states to refuse to comply with federal government policy. Th e principle has since been recognized as an appropriate institutional rule to protect minority rights.49 Additional parliamentary reform provided for constitutional amendments and legislation on local self- government to be adopted by a two-thirds majority, including a majority of representatives not from the majority community.50 Th e concurrent majority voting procedure is a form of consociational veto. Notably, the rule applies to a limited set of policy areas. Bieber suggests that “the advantage of such a relatively restrictive regulation is that it can help prevent blockage of the entire decision-making process.” At the same time, he notes a potential drawback in that restricting the application of the veto rule to specifi ed policy areas means that other decisions (e.g., relating to the economy) are excluded.51 Th e introduction of veto rights in Macedo- nia has been described a “key tool of formal power sharing in Macedonia” and an instance of “indirect veto” via special majorities.52 Th is indirect veto is similar to provisions for a weighted/parallel majority in the Northern Ire- land assembly and can be contrasted with the more “direct” veto in Bosnia relating to constituent peoples’ “vital national interests.” Th e OFA provided for a measure of territorial segmental autonomy in the form of increased decentralization. In contrast to the situation in Bosnia, territorial autonomy did not fi gure as a central issue in the Macedonian confl ict, with the NLA and Albanian parties stressing additional group rights relating to language, higher education, and represen ta tion in the police service and public administration. According to Bieber, however, during the peace talks the Albanian elites sought the creation of larger Albanian municipalities in the west of the country, a move opposed by the Macedonian parties, who sought to limit territorial autonomy for the minority community.53 He suggests that the local self-government reforms “establish an opening for a weak form of territorial self- government for the Albanian community.” 54 Th e implementation of the OFA resulted in enhanced competences for local government and the revision of municipal boundaries, arguably reinforcing the ethnic makeup of sub- state levels. Th e OFA also provided for cultural segmental autonomy in the form of reforms relating to “education and use of languages,” hugely sensitive issue areas for both communities. Th e state was obliged to provide primary and secondary education in “students’ native language” as well as funding for university level education in languages spoken by at least 20 percent of the population of Macedonia. To address the small number of university stu- Macedonia 173 dents from minority backgrounds, the OFA stipulated that the “principle of positive discrimination” should be applied until such time as “enrolment refl ects equitably the composition of the population of Macedonia.” In relation to language issues, the OFA states that the Macedonian language is the offi cial language throughout the state and in international relations. Th e Macedonian parties were adamant that any additional recognition of minority languages should not compromise the state’s unitary character. A compromise was reached that allowed for a minority language to be considered an offi cial language in units of self- government if 20 percent of the population speaks a minority language, allowing individuals to communicate with regional authorities in their own language. Communication from regional authorities will then be in the minority language as well as Macedonian. In addition, individuals can communicate with central government in any offi - cial language and receive a reply in that language as well as Macedonian. Th ough the 20 percent threshold meant that only the Albanian community among the minority groups was eff ectively granted this right, it did not go as far as to recognize Albanian on equal footing with Macedonian. Th e OFA did not establish the Albanian language as an offi cial language throughout the state or establish the Albanian language as an offi cial language of the state in international relations. Since the OFA, Albanian elites have continued to call for additional reform to address these issues.55 Th e agreement also committed the parties to provide translations in court proceedings and to adopt legislation concerning the provision of personal documents in an offi - cial language other than Macedonian, in addition to Macedonian. Under An- nex B on legislative changes, the parties were committed to introduce new legislation providing that assembly representatives have the right to address plenary sessions and parliamentary committees in an offi cial language other than Macedonian. Th is reform was not extended to the work of government ministries, another issue that Albanian parties have since sought to remedy. Finally, the Macedonian po liti cal system provides for the proportionality element of consociation. Increasing the ratio of Albanian positions in public administration, policing, and university education was a central tenet of the Albanian community’s politi cal agenda. Th e OFA made explicit reference to reform in the area of “nondiscrimination and equal represen ta tion.” Th e deal included a commitment to adopt legislation ensuring “equitable represen tation of communities” in central and local authorities. As we shall see, the implementation of this commitment has been fairly contentious. Albanian politicians have called for greater eff orts to increase the numbers of people 174 Chapter 7 from their community employed by the state, and Macedonian politicians have countered that Albanians have misguidedly called for equal, rather than equitable or fair, repre sen ta tion.56 Th ere has been some suggestion that the Macedonian model is a mix of consociational and integrative elements. For instance, Zoran Ilievski writes that Macedonia is not yet a consociation.57 He points to consociational elements in the concurrent majority voting procedure in parliament and the provisions relating to equitable repre sen ta tion in public administration. Re- garding integrative elements, Ilievski refers to the incentives for inter- ethnic cooperation, whereby parties from each of the two main communities need to consider the interests of their respective “traditional” co ali tion partners.58 He maintains that such incentives “can be the fi rst step to creating pre- election multi- ethnic co ali tions, rather than post- election co ali tions, which make way to more radical agendas.” 59 It is perhaps more appropriate to note that the electorate votes largely for parties in their respective ethnic bloc. As we shall see more clearly in the next chapter, an expectation has evolved that coali tion government includes the largest Macedonian party and the largest Albanian party. In this sense, vote-pooling and pre-election coali tions become redundant. Parties campaign as the best guarantors of their community’s interests, and voters support the parties most likely to promote these interests as the largest party of their bloc. Th ere has also been some discussion on the ethnic versus civic dimension to the OFA.60 According to Daft ary, the OFA seeks to establish a civic state but also “perpetuates divisions between the majority and nonmajority communities,” particularly owing to the use of the concurrent majority voting procedure in parliament.61 Th is argument suggests that the presence of veto rules leads to heightened ethnic division. A counter-argument, however, would maintain that the provision of group rights in the form of (carefully designed, non- expansive) veto rules is important for accommodation and overall cooperation. When groups feel their cultural identity and interests are protected, there is less cause for division and antagonism. Bieber notes that “in shying away from explicitly referring to specifi c ethnic groups,” the OFA “reforms seek to enhance the civic nature of the state.” At the same time, he notes that the Albanian community has been aff orded rights “comparable to those of the Macedonian majority.” Overall, Bieber writes that the eff ect of the OFA has been about facilitating “the transformation of Macedo- nia into a bi-national state.” 62 In terms of these arguments, whether the OFA cements an ethnic or civic state, it is perhaps more useful to focus on the Macedonia 175 post- OFA system as a liberal accommodationist arrangement. Th e question remains whether the measures to accommodate the two main groups via liberal power-sharing rules help foster cooperation. Th e literature makes positive soundings on the potential for the Mace- donian po liti cal system to foster cooperation between the two main groups. Iliveski and Taleski note that in comparison with Bosnia, power sharing in Macedonia is more “inclusive, integrated, and fl exible.” Th ey write that “enhanced local self-governance and soft power-sharing mechanisms at the centre, without strict ethnoterritorial divisions and formalized grand co alition of the ‘classic’ consociational type,” means that the po liti cal system has “struck the right balance between the centripetal and centrifugal forces in its divided society.” 63 Arguably, there was no need to introduce ethnic quo- tas or institutional rules governing executive formation, given the tradition of power sharing in Macedonia since in de pen dence. Bieber writes that the informal nature or fl exibility of power sharing in Macedonia “off ers greater opportunities for reducing the signifi cance of ethnic belonging in the politi cal system and, perhaps in the long run, preventing the dominance of collective and individual identities.” 64 In the remainder of this chapter and the next, I investigate whether the liberal nature of consociation has made a diff erence for inter- ethnic cooperation in Macedonia and whether lessons can be drawn for power sharing more generally.

Power Sharing 2002– 6: Challenges of Peace Implementation

Just over a year aft er the signing of the OFA, parliamentary elections were held on 15 September 2002. Th e election results were notable for the vote against the governing parties, VMRO-DPMNE and the DPA. According to the ICG, the election results confi rmed polling results that indicated the electorate’s disdain for the governing parties, owing to allegations of corruption and criminality.65 Th e elections led to the formation of a power- sharing coali tion including the SDSM, the Macedonian party that had ruled for much of the period since inde pen dence, and the Democratic Union for Integration (DUI), the new Albanian party with links to the NLA. Electoral support for DUI (sixteen seats, compared to the DPA’s seven seats) arguably refl ected the Albanian community’s desire to shore up the position of NLA leader Ali Ahmeti in the post-confl ict stage. Th e alliance between the SDSM and the Liberal Democratic Party (LDP) won sixty seats. Th e SDSM- LDP 176 Chapter 7 alliance then entered negotiations with DUI, assisted by the EU’s Alain Le Roy and U.S. Ambassador Lawrence Butler. In forming a new power-sharing executive, the coali tion parties committed to an agenda to deliver the implementation of the OFA. Th e inclusion of DUI in government was arguably important for cementing the tradition of power sharing in Macedonia. Th e ICG refers to the “astonishingly smooth negotiation” conducted by the Mace- donian parties, “with an Albanian party previously labeled terrorists and with whom contact had been forbidden.” 66 Th e SDSM’s Branko Crvenkovski became prime minister, with DUI securing a post of deputy prime minister (Musa Xhaferi) and four portfolios ( justice, health, education, and transport and communication).67 Despite the fears that a power- sharing government including a po liti cal party born of the confl ict would be riddled with division, the SDSM- LDP alliance and DUI were able to agree on some mea sure of policy reform. Th e ICG describes Crvenkovski and Ahmeti as “models of transparency and de- corum” providing the media with “a consistent, remarkably moderate tone” and having “steadfastly refused to engage in cheap appeals to nationalism.” 68 Certainly, it appears that both party leaders were committed to the implementation of the OFA and were opposed to the partition of the country. Crvenkovski even acknowledged that legitimate Albanian grievances were behind the 2001 confl ict and that the minority community had never accepted the 1991 Constitution, thus necessitating constitutional revisions.69 Voicing DUI’s commitment to the politi cal system, Ahmeti struck a positive note by saying that the power-sharing government was characterized by “po liti cal will and good cooperation.”70 At the same time, the SDSM and DUI were subject to considerable criticism from their respective intra-bloc competition, VMRO-DPMNE and the DPA. In April 2003, these two main opposition parties declared the OFA “dead” and suggested that a better solution to the confl ict would be to partition the country.71 Th e situation was compounded by the fact that there were ongoing divisions about the Ohrid deal among the wider electorate because Albanians were more supportive than Macedonians. But to what extent did the parties succeed in implementing the peace deal? Despite a promising start, the power- sharing government faced considerable challenges in a number of areas, including decentralization legislation, police reform, the issue of Tetovo University, and the issue of national fl ags and symbols. One of the main challenges for the power-sharing coali tion was to agree on decentralization legislation, a principal demand of the Albanian com- Macedonia 177 munity and a central feature of the OFA. Th e work of the power-sharing government from 2002 to 2006 was largely dominated by this issue. Noted above, the OFA obliged the parties to adopt legislation on the devolution of power from the central state to the municipal level. Th e r e o r g a n i z a t i o n o f municipal boundaries was a particularly sensitive issue for legislation on local self- government. Following months of discussion, the government proposed the Law on Territorial Or ga ni za tion of Local Self- Government. In February 2004, parliament passed a package of reforms at fi rst reading. Th e new legislation would reduce the number of municipalities from 123 to 80 and devolve new powers to the municipalities in the areas of education, health, culture, urban planning, and local policing. Th e legislation also included provision for a minority language to be used alongside Macedonian in areas where the minority constitutes more than 20 percent of the population, thereby fulfi lling the obligations set out in Sections 6.5 and 6.6. of the OFA. In practical terms, this would mean that the Albanian language would acquire offi cial status in twenty- fi ve of the eighty municipalities.72 Th e reor- gani za tion of municipal boundaries was a particularly sensitive aspect of the package, given the likely change to the ethnic makeup of local government units. Under the new law, some hitherto Macedonian municipalities would become predominantly Albanian. Th ese issues shaped ongoing negotiations among the coali tion parties, who reached an agreement on municipal boundaries in July 2004. But the details of this agreement provoked a robust response from the opposition parties, with VMRO-DPMNE accusing the Macedonian governing parties of “treason,” of responding to Albanian demands to the detriment of the majority community, and acquiescing in gerrymandering. Among the revisions, Albanians would become a new majority in the city of Struga and would comprise more than 20 percent of the population in Skopje, thereby designating Albanian as the second offi cial language in the country’s capital. In an eff ort to prevent such developments, opposition parties supported an initiative by the World Macedonian Con- gress to collect 150,000 signatures required to trigger a referendum on the issue.73 Following parliamentary approval of the legislation on 11 August 2004, VMRO-DPMNE and the Congress acquired the required number of signatures, and a referendum on the law was set for early November. Th e opposition’s campaign to stage a referendum on legislation relating to the implementation of the OFA alerted the international community. Th e key external actors— the EU and the U.S.— were determined to shore up support for the governing coali tion and kill off the threat of the referendum. 178 Chapter 7

In partic u lar, the EU had been developing a greater hands-on approach to the implementation of the OFA, having established the post of EU special representative (EUSR) to Macedonia in 2001. In response to the opposition’s referendum campaign, the EU was quick to support the government’s decentralization legislation as an important mea sure of OFA implementation. EUSR Michael Sahlin warned that should the referendum be successful, “it would clearly mean postponing the pro cesses of Euro- Atlantic integration and decentralisation.”74 In a speech to the Macedonian parliament, Eu rope an Commission President Romano Prodi described decentralization as “a method of founding stable and deep roots of local democracy. Europe is here, at the reach of your hands. . . . However, the decision depends on you . . . to say whether you want Eu rope.”75 On 4 November 2004, the U.S. announced that it would recognize the state by the name “Republic of Macedonia” rather than the “Former Yugo slav Republic of Macedonia,” as insisted by Greece. Th is decision boosted the position of the power-sharing government and ultimately helped defeat the referendum. Th e referendum was largely boycotted by the Albanian community and turnout was fairly low, at around 26 percent. Th is victory for the power-sharing executive was tempered somewhat by the resignation of Prime Minister Hari Kostov, who stressed intra-executive tensions and accused the DUI of “nepotism and corruption.”76 Police reform was another central issue of the OFA. Th ough implementation was slow, some progress had been made relating to increased recruitment of Albanian offi cers and the transfer of border control from the army to the police ser vice. Slowly becoming more representative of the population at large, the police received training from the OSCE, as well as the EU Police Mission Proxima.77 Th e issue of police reform was highlighted by the EU’s Opinion on Macedonia’s membership application, published in November 2005. Acknowledging that improvements had been made since 2001, the Euro pe an Commission noted that additional eff ort was required in securing the rule of law throughout the country. Additional progress on police reform was deemed to be “crucial for more eff ective policing and the prevention of the possible escalation of incidents.”78 Importantly, the Eu ro pe an Commis- sion gave a positive assessment of the state’s adoption of a legislative program arising from the OFA and encouraged the Macedonian government to ensure eff ective implementation of the various reforms. Th e introduction of legislation governing the use of communities’ fl ags and symbols represented yet another important element of OFA implemen- Macedonia 179 tation. Under the section “Expression of Identity,” the OFA stipulated, “With respect to emblems, next to the emblem of the Republic of Macedonia, local authorities will be free to place on front of local public buildings emblems marking the identity of the community in the majority in the municipality, respecting international rules and usages.” Th e use of fl ags had been a source of inter-ethnic tension for some years.79 For ethnic Albanians, it is paramount to their equal status to have the right to fl y the Albanian fl ag, a double-headed black eagle on a red background. Th ey see the Albanian fl ag as their national fl ag and the Macedonian fl ag as the state fl ag. For Macedo- nians, however, the use of the Albanian fl ag is a threat to their identity and state sovereignty. As an outstanding element of implementation of the OFA, parliament passed the Law on the Use of Flags of the Communities in 2005, essentially a compromise between ethnic parties providing that majority communities in municipalities have the right to fl y their fl ag alongside the Macedonian state fl ag. As we shall see in the next chapter, debate continued over the issue of fl ags and national symbols, with a number of Macedonian politicians instigating proceedings before the Constitutional Court concerning the constitutionality of some provisions of the 2005 legislation. On language issues, the power- sharing government made some progress with the Law on Passports, adopted in March 2003, allowing Albanians to apply for the inclusion of their language on passport covers, in addition to Macedonian. Th e governing parties maintained considerable progress had been made since the signing of the OFA. At a roundtable or ga nized by the Project on Ethnic Relations, DUI highlighted progress relating to the use of the Alba- nian language, the implementation of amnesty legislation, and the increased number of Albanians in public positions, including the police force.80 Mem- bers of Macedonian governing parties suggested that progress had been made because people from their community accepted constitutional reform, the expansion of the use of the Albanian language, and greater decentralization.81 Yet criticism had been leveled at the coali tion’s perfor mance in implementing the OFA. One opposition Albanian politician suggested that the right to use the Albanian language had not been made “automatic” and that the SDSM- DUI co ali tion had lacked “po liti cal will” to treat Albanian as an offi cial language.82 For the Albanian opposition, the failure of the government to agree on a Law on Languages meant that the governing coali tion had failed to implement the peace agreement. Calling for a comprehensive law on languages, Albanian politicians continued to assert that their language was 180 Chapter 7 not treated as an offi cial language in parliament due to the position of Mace- donian representatives acting as chairs of parliamentary committees.83 On- going contention over OFA implementation, particularly in relation to the use of the Albanian language, set the scene for parliamentary elections in July 2006.

Summary

Th e Macedonian case points to the evolution of power sharing in the management of majority-minority relations. In contrast to Northern Ireland and Bosnia, the country had a tradition of power sharing as a feature of democratization and not the outcome of a peace deal. In the initial post- independence period, power sharing took place as an informal politi cal rule or convention, refl ecting the politi cal reality (and parties’ accep tance thereof ) that cross- community government would help secure stability. As discussed, however, executive power sharing alone was insuffi cient to address Albanian grievances in the absence of additional accommodationist mea sures (veto rules, segmental autonomy, and proportionality in public administration). In the aft ermath of the 2001 confl ict, the negotiators of the peace agreement sought to cement this tradition, not by introducing a formal requirement for power sharing, but by providing additional rights for Albanians and introducing procedures (e.g., the double majority rule in parliament) that supported the convention of inter- ethnic executives. Arguably, the cementing of power sharing in the OFA, particularly the concurrent majority voting procedure, had a positive impact on Macedonian- Albanian relations. With the main parties committed to the implementation of the OFA as the most feasible option for resolving the confl ict, they made some eff ort to cooperate, leading to a degree of power- sharing “jointness” on sensitive issues including decentralization and equitable repre sen ta tion. Th is is not to say that power sharing post-Ohrid was plain sailing. Intra- executive diffi culties were compounded by ongoing contention over the OFA, with Albanians more likely to support the deal. Unsurprisingly, the co ali tion partners were wary of these external constraints from the wider electorate and pressure from their respective intra- bloc opposition. More- over, there were instances of bad relations between the SDSM and DUI, with signs that the Macedonian party at best “tolerated” being in government with the Albanian party previously linked to the NLA.84 Macedonia 181

As with the other two places under investigation, external actors have played an important role in incentivizing the po liti cal parties to share power. In the Macedonian case, the EU and the U.S. were central actors in arriving at a deal that would address the groups’ grievances and end the confl ict. In contrast to the situation in Bosnia, the EU adopted a role of supporting and encouraging the implementation of the peace agreement. Simi- lar to the role of the EUSR in Bosnia (but not, of course, the executive powers of the Offi ce of the High Representative), the EUSR in Macedonia sought to incentivize policy reform as securing the country’s ticket to EU integration. Indeed, much of the power-sharing executive’s work in the post-Ohrid period was taken up with the task of preparing the country’s application for membership, formally submitted in March 2004. Th ese eff orts were re- warded late 2005 when the EU granted the state candidacy status. It has been suggested that the “prospect of EU integration gives politicians their main motivation for pursuing reform policies and helps guarantee peaceful coexistence of the main ethnic groups.”85 Th e next chapter investigates the signifi cance of the EU membership incentive in fostering power- sharing cooperation and po liti cal stability. CHAPTER 8

Toward a Binational Macedonia?

Th e granting of EU candidate status to Macedonia in December 2005 was a milestone for the country’s progress and international standing. Th e EU’s decision suggested the country was well on track to overcome any remaining ethnic tensions and ensure the full implementation of the Ohrid Frame- work Agreement. Th e experience of power sharing since then, however, has been hampered by a number of diffi culties over executive formation, parliamentary boycotts, tensions between the Macedonian and Albanian parties in power, controversy over cultural issues, and a slowing down of the country’s Euro-Atlantic integration. Here I review the impact of the various politi cal crises on power sharing and question the extent to which the parties have cooperated on full implementation of the OFA. I demonstrate that although Macedonia has become an increasingly binational state, power- sharing problems continued to exist due to contentious identity politics.

Elections, Exclusions, and Boycotts

Since 2006, the power-sharing executives have faced several diffi culties relating to free and fair elections, government formation procedures, and parties’ parliamentary boycotts. Th e management of free and fair elections has been an important concern for Macedonia in the country’s path to EU membership. Just six months aft er being granted candidate status, progress toward membership, as well as internal stability, were threatened by instances of violence during the election campaign.1 Th e international community was quick to stress that such events could only harm Macedonia’s progress toward Euro-Atlantic integration, with EU Special Representative Erwan Fouéré warning that violent incidents during elections must be con- Binational Macedonia? 183 signed to the past.2 Aft er pressure from the international community, the remainder of the campaign passed relatively smoothly. Th e election brought victory for the Macedonian co ali tion led by VMRO- DPMNE, who won forty-fi ve parliamentary seats (defeating the SDSM-led coali tion on thirty- fi ve seats). On the Albanian side, the DUI- PDP co ali tion won seventeen seats and the DPA eleven seats. As the largest party, VMRO-DPMNE then began a process of negotiations on government formation, fi rst consulting with DUI. But because the discussions failed to reach agreement, VMRO- DPMNE turned to the DPA as its Albanian partner in government. When these discussions produced agreement on coali tion formation, Macedonia faced a new challenge to the country’s power- sharing tradition. DUI leader Ali Ahmeti rejected the formation of a new power-sharing government without his party, which had secured the most votes from the Albanian community. In an interview with daily newspaper Dnevnik, Ah- meti criticized DUI’s exclusion from government and warned, “[Prime Minister] Gruevski will have a problem because he does not respect the will of the Albanians and because he humiliates them. We are the masters of our future and not the tenants in this country. I have already said that if we are not in the government, he will have to deal with us.”3 In the aft ermath of executive formation, DUI’s rejection of the new government led to considerable tension between the DPA and DUI and between the power-sharing executive and opposition.4 Perhaps unsurprisingly, DUI politicians felt much aggrieved that their party had been excluded from government despite win- ning more votes and seats than the DPA. Looking back on this episode, some leading DUI elites suggested there is a need for an extension of the Badinter principle to cover executive formation to prevent a similar situation in the future.5 Another DUI politician commented that even if there is no formal institutional rule on government formation, “respect for democratic rule” means that the largest Albanian party should be in power.6 Yet VMRO- DPMNE rejects the need for any formal rule on executive formation.7 Simi- larly, offi cials working for the EU delegation in Skopje viewed it unnecessary to formalize the arrangements for power sharing. Th ey claimed that VMRO- DPMNE, as a result of the furor over government formation in 2006, recognized that there was an issue with legitimacy in excluding the largest Albanian party. In addition to acknowledging the legitimacy of power- sharing coali tions, VMRO-DPMNE also came to understand that securing approval on legislation would be more straightforward with the largest Alba- nian party on board.8 An EU offi cial suggested that although power sharing 184 Chapter 8 in Macedonia is not new and is not a product of the OFA, “even the most hard-line nationalists recognize that they need a legitimate government” and that “this fragile state is based on this kind of power sharing.” 9 Th e exclusion of DUI was, however, to have serious consequences for the operation of the power-sharing government and politi cal stability more generally. At the end of January 2007, DUI launched a parliamentary boycott lasting four months. With ongoing tension related to the party’s exclusion from government aft er the 2006 election, the party complained that laws requiring the application of the Badinter principle had been passed without their support.10 As the months progressed, DUI set preconditions for returning to parliament, including an extension of the range of issue areas included under the concurrent majority voting procedure and a law on the status of former NLA fi ghters. Pressure from the EU and NATO for constructive dialogue between the government and opposition eventually led to negotiations between the VMRO- DPMNE and DUI leaderships. During a visit to Skopje in February, EU Enlargement Commissioner Olli Rehn referred to the “need to exert pressure on both the government and the opposition to respect the rules of the game and ensure the legitimacy of the po liti cal system.”11 He described his visit as a “strong wake- up call” for “a reinforcement of mutual trust, a consensus on the rights of minorities and reforms related to Euro pe an aspirations.” In what became known as the “May Agreement,” Ahmeti declared that a deal had been made on legislation securing benefi ts for former NLA fi ghters and the use of the Albanian language as an offi cial language throughout Macedonia. However, Prime Minister Nikola Gruevski and his party colleagues in VMRO-DPMNE denied that any such formal agreement had been made.12 Nevertheless, the EU and U.S. welcomed agreement between the leaders of the two parties, stressing that stable democracy and continuing politi cal dialogue among all politi cal parties would help ensure progress toward Euro- Atlantic integration.13 On the basis of Ahmeti’s statements, media reported that agreement had been reached on a number of issues: a list of forty-six laws to be included under the concurrent majority voting rule, changes to the makeup of the Committee for Inter-Community Relations (a VMRO-DPMNE member to be replaced by one from the SDSM), a pledge to provide support to the victims of the 2001 confl ict, agreement to draft a law on languages, and a commitment to discuss procedures for executive formation. As a result of this agreement, DUI called off its parliamentary boycott. Th e international community was quick to praise the eff orts of parties to engage in dialogue and get OFA implementation back on Binational Macedonia? 185 track. A joint statement from the EU delegation in Skopje and the U.S. Em- bassy described the deal as “an important step forward in relation to the integration pro cesses of the country toward the EU and NATO.”14 Th e controversies over executive formation and DUI’s position were resolved following early parliamentary elections held on 1 June 2008. Again, instances of violence took place during the election campaign and on election day.15 As in 2006, the international community, including the EU, OSCE, and U.S., were quick to condemn the conduct of the elections.16 Due to instances of intimidation and violence (including a shooting that left one person dead and around ten people injured) and voting “irregularities,” the vote was rerun a few weeks later in largely Albanian constituencies.17 Th e fi nal results saw the Macedonian co ali tion led by VMRO- DPMNE in sixty- four seats, the SDSM- led co ali tion in twenty- eight seats, DUI in eigh teen seats, and the DPA in eleven seats. Th is time, VMRO- DPMNE and DUI successfully negotiated a power- sharing co ali tion. As part of the power- sharing negotiations, VMRO and DUI reportedly agreed to continue discussions on a language law.18 In contrast to the situation following the 2006 parliamentary election, Gruevski and his party were more amenable to forming a government with DUI. Given the diffi culties relating to DUI’s former exclusion from power sharing, this was arguably an instrumental assessment of po litical reality. An offi cial at the EU delegation in Skopje suggests that “Gruevski has come to terms with the reality that he has a more stable government and can be much more eff ective in parliament if he has a broad majority, so he if selects the largest ethnic Albanian party, this is to his own advantage.”19 Apart from diffi culties relating to executive formation procedures, power- sharing practice was also challenged during this period as a result of parties’ parliamentary boycotts. Such boycotts have become a recurring feature of Macedonian politics in recent years.20 VMRO-DPMNE boycotted parliament before the 2006 elections over changes to the Electoral Law. In 2007, DUI boycotted parliament for four months, leading to inter- party discussions and the May Agreement between Ahmeti and Gruevski. Th e SDSM took its turn to boycott parliament in 2008 over the arrest of the party’s vice-president, Zoran Zaev.21 In August 2009, the DPA decided to boycott parliament, citing the government’s failure to secure the Albanian language as the country’s second offi cial language.22 Th e DPA’s boycott lasted a considerable time, until March 2011, in the run up to parliamentary elections.23 A few months earlier, in January 2011, the main opposition party, SDSM, announced another parliamentary boycott, with leader Crvenkovski claiming 186 Chapter 8 that “Macedonia suff ers at the hands of an antidemocratic, authoritarian regime that goes against the people.”24 Th e SDSM called for revisions to the electoral code, updated voting lists, and the need to unfreeze the bank accounts of A1 TV.25 Th e failure of SDSM and VMRO-DPMNE to agree on these issues exacerbated existing tensions and brought about early elections in June 2011. Th e frequency of parties’ decisions to boycott parliament led to some inconve nience for the successive power-sharing governments. In particu lar, DUI’s absence from parliament for four months in 2007 made it diffi cult for the government to push through legislation applicable to the Badinter rule. Th e tendency of parties to opt for a boycott is deemed to be bad parliamentary practice, irresponsible, and “juvenile” by the EU.26 In an eff ort to dissuade parties from boycotting, VMRO- DPMNE prepared draft legislation on the responsibilities of MPs, with penalties for parliamentary absences. According to the party, “Macedonia is the only country where boycotts are used to deliver street ultimatums rather than cause debate.”27 At the same time, however, DUI resisted the application of sanctions on a party for deciding to boycott parliament, defending it as “a po liti cal gesture [that] is the right of every party.” With the SDSM’s boycott triggering early elections, the party increased its vote from eighteen to forty-two seats, but VMRO- DPMNE and DUI cemented their positions as the largest parties in the respective blocs.28 As the largest Macedonian party won fewer seats than in the 2008 election (losing seven seats, from sixty-three to fi ft y-six), it was expected that Gruevski would have to make additional concessions to DUI in executive formation negotiations. DUI reportedly wanted agreement on a common policy platform focusing on economic development, Euro- Atlantic integration, corruption and orga nized crime, investment in education, and full OFA implementation.29

Toward Full Implementation of the OFA

As outlined in the previous chapter, considerable progress had been made on peace implementation in the early post-Ohrid years, leading to the country’s ac cep tance as an EU candidate country in 2005. For politicians interviewed for this study, EU candidate status was a direct reward for progress on OFA implementation. Th ough progress was slow, some improvements were achieved in decentralization legislation, increased represen ta tion of Albanians in the police, amnesty legislation, legislation on the use of fl ags, and legislation Binational Macedonia? 187 providing for the use of the Albanian language on passports. Despite this progress, however, the Albanian parties and the international community continued to call for the full implementation of the OFA. In partic u lar, DUI and the DPA focused on the need for a greater focus on three areas related to the peace agreement: equitable repre sen ta tion in public administration, decentralization, and the law on languages. In the aft ermath of the 2006 elections, the new power- sharing executive comprising VMRO- DPMNE and DUI appeared to give renewed focus to the issue of equitable repre sen ta tion in state institutions. Although VMRO- DPMNE pledged to increase the number of non-Macedonians in public administration, an Albanian representative of the new executive promised a comprehensive review of OFA implementation, including greater focus on state employment, in order to achieve “equitable and appropriate” representa tion of minorities.30 Yet, to some extent, progress remained contentious. Since 2006, progress on equitable repre sen ta tion in public administration has been described as the main problem relating to OFA implementation. According to one DUI politician, in many state institutions, Albanians have not been hired in proportion to their Macedonian counterparts; some institutions have had just 4 or 5 percent Albanians, and in some institutions, there have been no Albanians employed.31 He suggested that these diffi culties are owed to “blocking of the process from the Macedonian side.” Albanian politicians also stress that members of their community have been less likely to hold se nior positions in state institutions and public enterprises. A se nior DUI politician claimed there are not enough interpreters employed throughout the state institutions and that more funding needs to be allocated from the budget to remedy this problem.32 For the external actors engaged in monitoring OFA implementation, a problem has arisen in that public administration has become increasingly politicized. In its 2010 Prog- ress Report, the Eu ro pe an Commission referred to “reports of replacement of trained professionals with appointees of limited experience in several institutions.”33 Moreover, the Commission noted that the “recruitment of a large number of employees from the nonmajority communities is on a quantitative basis and without matching the needs of the institutions with the required training and qualifi cations.” Th e problem derives from both VMRO- DPMNE and DUI providing party members with jobs in the state institutions, oft en personnel without experience in public administration. In terms of addressing the numbers of Albanians in state employment, DUI representatives claim that VMRO- DPMNE has been blocking the fulfi llment 188 Chapter 8 of equitable represen ta tion and therefore denying the full implementation of the OFA.34 In return, VMRO-DPMNE suggests that the lack of training and expertise within the Albanian community explains any remaining problems in this area.35 Th e use of the Albanian language in state institutions has continued to be a source of controversy for the power-sharing governments since 2006. DUI continued to press for the right to use the Albanian language in the workings of government ministries and in representing Macedonia on the international scene. In August 2008, parliament adopted the Law on the Use of Languages regarding languages spoken by at least 20 percent of citizens. Th e Euro pe an Commission noted that the law “clarifi es and extends the scope for the use of nonmajority languages at all levels of state and local self-government within the limits set by the Ohrid Framework Agreement and the Constitution.”36 Th e law provides MPs and committee chairs with the right to use a minority language in their parliamentary duties. It extends the right to use the minority language in municipalities where the minority makes up less than 20 percent, if the council so decides.37 Yet it also specifi es that languages other than Macedonian may not be used in the police and the army. In its 2010 Progress Report, the Eu ro pe an Commission noted that progress had been made in the implementation of the Law on Languages, including the recruitment of additional interpreters in parliament and the use of Albanian in oral procedure in plenary and committee sessions, and that the parliamentary TV channel was being translated into Albanian.38 DUI has, therefore, gained incrementally on the use of the Albanian language in state institutions. But despite these developments, the issue continued to be important for the party who, following the 2011 elections, negotiated with VMRO- DPMNE to amend the law allowing government offi cials to address parliament in Albanian. DUI politicians continued to highlight that in the judicial system, Albanian judges have had to communicate with Albanian defendants in Macedonian, followed by a translation.39 Decentralization reform has also been crucial for OFA implementation, subsequently monitored by the Eu ro pe an Commission in its assessment of Macedonia’s progress in the EU integration process. As discussed in the previous chapter, the post-Ohrid power-sharing government made some initial progress on decentralization with the 2002 Law on Local Self- Government and the 2004 Law on Territorial Or ga ni za tion. Yet criticism continued that decentralization had not been implemented fully, as per the ICG’s assessment that fi nancial decentralization had been “proceeding slowly and at an Binational Macedonia? 189 uneven pace.” 40 Albanian politicians referred to the ongoing disparities between (largely Albanian) rural and urban municipalities and called for budgetary resources to address what they perceived to be discrimination against their community.41 DUI politicians have called for the transfer of more competences to the local level, proportional distribution of the bud get to address disparities across local government, and for approval on the budget to be included under the concurrent majority voting procedure in parliament.42 In assessing implementation, Euro pe an Commission progress reports acknowledged some eff orts by VMRO-DPMNE and DUI to progress on decentralization issues and called for greater focus to ensure full implementation of legislative measures. Th e 2 0 1 0 E u r o p e a n C o m m i s s i o n Progress Re- port noted that some additional municipalities had entered the fi nal stage of the fi scal decentralization pro cess and the execution of capacity- building programs on property tax administration, human resources, and fi nancial control. Th e Commission called for increased eff orts from central government, particularly the Ministry of Local Self- Government and the Minis- try of Finance, the latter needing to develop policy on local government budgetary issues. Th ere was also an identifi ed need to increase fi nancial and administrative capacity in some municipalities. Echoing some of the comments by Albanian politicians interviewed for this research, the Euro- pean Commission also noted the limited measures to address disparities in the provision of public ser vices at the local level, particularly in rural and small municipalities.43 Criticism was also levied at the power- sharing co- alitions for viewing decentralization as an electoral opportunity whereby capital investment and employment have become “dependent on politi cal affi liation.” 44 Th is somewhat mixed record on OFA implementation as a work in progress arguably points to a degree of “jointness” or cooperation within the power- sharing governments, while also signaling the challenges in handling policy reform in sensitive areas. A media commentator suggested that VMRO-DPMNE and DUI “cooperate on issues which they fi nd benefi cial for both sides or unavoidable.” 45 He explained this cooperation by each party’s commitment to Euro- Atlantic integration, a pro cess he viewed as important for stability. Th e Euro pe an Commission noted “constructive cooperation” in power sharing between VMRO-DPMNE and DUI and their commitment to the EU reform agenda, while also stating that “more dialogue is required on issues concerning inter- ethnic relations.” 46 A related point was made by an EU offi cial, who suggested that although full implementation of the OFA 190 Chapter 8 is important for the integration pro cess, “Ohrid is not a box that you can tick.” 47 As the peace agreement was ultimately a “framework agreement,” it has been deemed “much more about the spirit of inter- ethnic dialogue to give ownership to the local partners here to reach solutions.” Arguably, the two main parties have largely lived up to the ethos or spirit of the OFA by engaging in inter-ethnic dialogue on policy issues. A positive assessment of inter- ethnic cooperation in Macedonia points to the fact that VMRO- DPMNE and DUI have shared power on several occasions, thereby becoming the power-sharing “norm” and creating politi cal stability. Th e manner in which these two parties have shared power highlights the tendency of ethnic parties in power sharing to negotiate in backroom deals involving few people from each party. According to an EU offi cial, “the way power sharing works is that decisions are taken at the top level of each party and then an enormous amount of management is made by party leaders to keep party members on board.” 48 Even though progress on sensitive issues has more oft en been reached among party leaders in private discussions, with some pressure from the international community, it is clear that power sharing in Macedonia has had some success, given the progress on decentralization and equitable repre sen ta tion. Yet there is also a view that cooperation between the power-sharing parties has been rather limited and that the two main parties have had a strained relationship. Media reports refer to DUI’s position in the 2008– 11 executive as “little more than a sideshow.” 49 According to a Macedonian opposition politician, the power- sharing executive formed by VMRO- DPMNE and DUI lacked politi cal will to resolve ethnic issues.50 He suggested that the two parties were not concerned to work together and that the executive was characterized by “constant fi ghts” on policy issues. Rather realistically, one DUI politician commented that his party and VMRO- DPMNE are in co ali tion government “not because we like each other, but because together we have a stable parliamentary majority. It’s a marriage of interest.” 51 Th e sometimes- limited extent of cooperation is evident in that the Committee for Inter- Community Relations, set up under the OFA as a vehicle for parties to resolve sensitive ethnic issues, has not functioned eff ectively. In its 2009 Progress Report, the Eu ro pe an Commission noted that the committee was playing a limited role, restricting its activity to issues relating to the double- majority voting procedure.52 Th e Commission later noted that “six sessions out of a total of ten scheduled since its establishment in mid-2008 Binational Macedonia? 191 have been cancelled. Cooperation between the parliamentary committee for relations among communities and the corresponding committees in the municipalities remains insuffi cient.” 53 Meeting on rare occasions since 2008, DUI accuses VMRO- DPMNE of boycotting sessions and not taking inter- ethnic relations seriously.54 Commenting on the diffi culties faced by the Committee, another DUI politician suggested that the problem has arisen from the fact that fi ve of the seven Macedonian members are from VMRO- DPMNE, and the fi ve members representing the smaller minorities are in co ali tion with that party. Controlling ten members who fail to show up, out of a total of nineteen, VMRO- DPMNE was accused by DUI of essentially blocking the functioning of the Committee. Th e DUI member claimed that because VMRO- DPMNE resists debating any inter- ethnic “hot issues” in parliament, Committee sessions on topics ranging from Skopje 2014 to the usage of the Albanian language in public administration have not taken place. A de cade on from the signing of the OFA, wider criticisms relate to the success of the OFA in bringing about a stable multi-ethnic democracy and securing reconciliation between the Macedonian and Albanian communities. David L. Phillips refers to “two competing narratives about the OFA.” 55 On one level, the OFA has cemented inclusive power sharing and a plethora of legislation has been introduced to address the Albanian community’s grievances. On another level, however, the OFA has failed to address the interests of the smaller minorities and Macedonian society remains unequal and divided. Phillips suggests the reality lies somewhere in between, as VMRO- DPMNE has never been a strong supporter of the OFA and progress has been the result of an “expedient but uneasy relationship” between the two main power- sharing parties.56 Th ere is also a view that although the two main parties have, on the whole, cooperated to resolve po liti cal crises and ensure (albeit slow) implementation of the OFA, the agreement has been insuffi cient to overcome division between the two main communities. In- terestingly, an EU offi cial critiqued the OFA for not having promoted integration of the diff erent groups, considered necessary “for the consolidation of the state.” 57 Th e offi cial suggested that the agreement’s “weakness is that it does not provide guidance on integration” and allows the Albanian community to mistake integration for assimilation, which harks back to their experience in the former Yugoslavia. Another EU advisor suspected the Al- banians “want a kind of parity” with Macedonians, who “would never agree 192 Chapter 8 to that.” 58 Arguably, however, the OFA is ultimately an accommodationist rather than an integrationist agreement, and the po liti cal system has, as a result of OFA implementation, been moving toward a binational state. A related point is that opposition parties, perhaps unsurprisingly, criticized VMRO- DPMNE and DUI for failing to resolve inter- ethnic issues. Th e SDSM critiqued Gruevski’s style of governing and his party’s policies for limiting inter- ethnic reconciliation and halting the country’s accession to the EU and NATO. On the Albanian side, the DPA has, on occasions, called into question the potential of the OFA for making suffi cient changes to the status of their community. Th e DPA has been critical of DUI’s eff orts in government, claiming that the latter has been weak in its dealings with VMRO- DPMNE and has failed to secure adequate group rights for the Albanian community. It has been reported that before the 2011 parliamentary elections, the DPA’s electoral agenda called for the establishment of the Alba- nian community as a second constituent people and the extension of issue areas under the Badinter rule to include government formation, the budget and Judicial Council appointments.59 DPA’s leader, Menduh Th açi, reportedly claimed that the OFA was “practically dead” and suggested the need for a new agreement between Macedonians and Albanians.60 Similarly, a DPA politician suggested that a federal solution should remain the ultimate goal of the Albanian community.61 Although some of these criticisms refl ect normal party politics, they nevertheless stress the challenge of intra-bloc competition for the ruling coali tion and even some divergence between the two communities on what kind of state Macedonia should be: Macedonians favor a civic- based unitary state; Albanians favor decentralized and even (for some) federal arrangements.

Addressing Issues of Cultural Contestation

In addition to diffi culties of OFA implementation, power sharing in Mace- donia has been challenged by ad hoc controversies relating to issues of cultural contestation. Marc Howard Ross suggests that cultural contestation concerns “inclusion and exclusion from a society’s symbolic landscape . . . such inclusion or exclusion tells us about the politics of accep tance, rejection, and access to a society’s resources and opportunities.” 62 Arguably, issues of cultural contestation have served to challenge the stability of power shar- Binational Macedonia? 193 ing in Macedonia in several areas. I focus here on the publication of a controversial encyclopedia by the Macedonian Academy of Arts and Science in 2009 and VMRO-DPMNE’s urban regeneration project, “Skopje 2014.” In interviews for this research, Albanian politicians highlighted the furor created by the publication of an encyclopedia by the Macedonian Acad- emy of Arts and Science as an instance that challenged power- sharing stability. In September 2009, the state-funded Academy published an encyclopedia that referred to Albanians as “settlers” who arrived in Western Macedonia in the sixteenth century and as šiptari, a derogatory term. It also described Ali Ahmeti as a “war crimes suspect” and claimed that the U.S. and UK trained Albanian rebels during the 2001 confl ict.63 Its publication created outrage among the Albanian community and was also rejected ro- bustly by politi cal leaders in Albania and Kosovo. Local Albanian parties accused the Academy of publishing a “politi cal pamphlet” that “infl icted great evil upon the Albanian population.” 64 Aside from the contents of the encyclopedia, matters were aggravated by the attendance of Prime Minister Gruevski at the book’s launch, therefore seeming to endorse its contents. Amid street protests by the Albanian community and uproar by the Alba- nian media, there was even some speculation that DUI would quit the power-sharing executive in protest. DUI claimed the episode had a wholly negative eff ect on inter- ethnic relations within the government and called on Gruevski to condemn it.65 In an eff ort to play down the crisis, the prime minister ultimately distanced himself and VMRO- DPMNE from the of- fence caused, blaming “dark forces” for the crisis brought about by “individuals and forces that have not reformed and who do not comprehend the reality—that Macedonia belongs to all of its citizens.” 66 Th ere also appears to have been some pressure from external actors put on VMRO- DPMNE to issue a statement repudiating the encyclopedia.67 In response to the outrage, the Academy fi rst announced a revision of the controversial passages, the resignation of the editor, and fi nally the withdrawal of the book at the be- hest of the country’s president, Georgi Ivanov. Th e episode is a clear example of cultural contestation because it brought into sharp relief questions of identity, inclusion, and exclusion in Macedonian society. Evoking an image of Albanians as settlers or less- worthy citizens of the state unsurprisingly tapped into sensitive inter-ethnic relations. Th ough the furor did not origi- nate with the power- sharing parties, it clearly tested the mettle of the executive to move beyond such unhelpful and potentially explosive discourse. 194 Chapter 8

One of the most controversial issues since VMRO-DPMNE and DUI entered into power sharing in 2006 derives from the former’s plans to regenerate the capital city under the project “Skopje 2014.” Not a straightforward city development plan, the project has been described as “the prime minister’s state- sponsored nationalism.” 68 Th e controversial plans for Skopje’s city center included the construction of around seventeen statues and fi fteen buildings that depict classical architecture styles. One such statue is a mon- ument to Alexander the Great with lights and music. Supporters of the project say it will “endow the city with landmarks and buildings entwining the city with Macedonia’s national identity.” 69 Th e airport was renamed “Skopje Alexander the Great Airport,” the main roadway to the Greek border was renamed “Alexander of Macedon,” and the main sports stadium in Skopje was named aft er Alexander’s father, Philip. Th e government also funded a new edition of Th e History of the Macedonian People, which claims that Macedo- nians are descendants of the Macedonians of antiquity and are not of Slavic origin.70 In an increasingly binational state, however, it is no surprise that public celebration of one culture and identity creates tensions. For Macedonians, the project is about affi rming their nationhood. But for Albanians, Skopje 2014 is a one- sided project with little to off er their community. A DPA politician commented that VMRO- DPME has been pursuing since the country’s in de pen dence “a po liti cal project to construct, build a Macedonian identity” and has been “remaking history” by claiming that Macedonians are descendants of Alexander the Great.71 Similarly, an Albanian commentator suggested that questions of Macedonian identity, exemplifi ed by the linking of the Macedonian nation with Alexander the Great, cast a shadow over all other po liti cal issues, including the implementation of the OFA and solving the country’s economic problems.72 Moreover, it has been suggested that the project has been “marred by controversy” and “is indicative of the identity crisis” facing the country since the destruction of the former Yugo slavia.73 As an instance of cultural contestation, it is clear that debates over Skopje 2014 refl ect the complex social and politi cal meanings of Macedo- nia’s symbolic landscape. Arguably, VMRO’s plan to develop the city is an assertion of power in a region where Macedonian identity has been questioned in terms of the name issue, territory, the Macedonian Orthodox Church, and the Macedonian language.74 Debates over Skopje 2014 have also pointed to Binational Macedonia? 195 diffi culties within the power- sharing executive between VMRO- DPMNE and DUI. In par tic u lar, there have been allegations that the largest party in power has pursued certain policies that come to the executive table without prior consultation with their Albanian coali tion partner.75 It has even been suggested that Skopje 2014 has been Gruevski’s pet project, with a Macedo- nian Orthodox Church planned for the city center square and paid for by the state (subsequently withdrawn from plans). Faced with protestations from DUI on the nature of the project, it has been suggested that a compromise was reached between the two executive parties that a few additional monuments to represent Albanian heroes would be included.76 To some extent, DUI has gone along with the Skopje 2014 plan, or at least put up with it, in return for some statues depicting Albanian fi gures in Macedonia Square as well as a small project in the Old Bazaar in Čair, an Albanian majority municipality in the city. In terms of the eff ects of these plans, known locally as “antikvizacija” (antiquization), Skopje 2014 has arguably undermined the country’s progress toward EU accession and has increased inter-ethnic tension.77 Certainly, the project has alienated Albanians and other minorities and appears to be in contravention of the spirit or ethos of the OFA, which is based on inclusion. Moreover, it seems to have reignited feelings of discrimination among the ethnic Albanian community that hark back to the 1990s. A Macedonian opposition politician claimed that Gruevski’s antiquization project is an example of a “nonfunctional, constantly confrontational governmental co ali tion that has very negative consequences for the whole country.”78 He claimed the issue has led to a marked decrease in trust between the Macedonian and Al- banian communities, similar to the level of mistrust in the years prior to the 2001 confl ict. Th ese episodes of cultural contestation highlight the challenges of managing multiple public identities in a post- confl ict power- sharing democracy. Indeed, the management of such issues appears to be an important test for the operation of cross- community power- sharing executives. Th e case of Macedonia highlights the complexity of cultural identity issues in an increasingly binational state. With regard to Skopje 2014, the ethnic Macedo- nian community has sought to cement a coterminous link between nation and territory as part of a wider nation-building project, hence the antiquization strategy pursued by VMRO-DPMNE. Yet the ethnic Albanian community has sought to position itself on an equal footing with the majority and 196 Chapter 8 seeks state recognition of their language and symbols. Arguably, the pursuit of a project that celebrates one identity over others threatens the accommodationist basis of the OFA.

The EU Role: From Conditionality to Stalemate

Th e EU has been a central player in helping to resolve inter-ethnic tension in Macedonia, having facilitated the peace agreement in 2001 and since encouraged its full implementation. Th e Euro pe an Commission Delegation to the Former Yugoslav Republic of Macedonia (later becoming the Delegation of the EU) opened in March 2000 to help monitor the country’s compliance with EU criteria toward eventual membership. A Stabilization and Associa- tion Agreement was signed in April 2001.79 At the end of June, in an eff ort to negotiate an end to the confl ict, the EU appointed François Leotard as its special representative (EUSR), replaced by Alain Le Roy in September aft er the signing of the OFA. In March 2003, the EU’s Concordia mission took over responsibility for the international military presence from NATO, followed by a police mission, EUPOL Proxima. In November 2005, Erwan Fouéré was appointed the new EUSR and head of the Eu ro pe an Commis- sion Delegation in Skopje, a joint action by the Council of the EU and the E u r o p e a n C o m m i s s i o n . 80 Th e mandate of this double-hatted position was to coordinate eff orts of the international community to help ensure OFA implementation and to help guide the country on its path to EU integration. In March 2004, Macedonia presented its application for EU membership, and the Euro pe an Commission recommended granting the country candidate status in November 2005.81 In its Opinion, the Commission noted that Mace- donia “has successfully implemented the legislative agenda of the Ohrid Framework Agreement, which contributed to major po liti cal and security improvements in the country. Th is legislation now needs to be fully enforced.”82 Areas deemed in need of improvement included the electoral process, reform of the judiciary and the police, and the fi ght against corruption. A few weeks later, the Eu ro pe an Council approved Macedonia’s membership application, noting legislative progress relating to the OFA and the implementation of the SAA.83 In 2008, the Euro pe an Council adopted a new Accession Partnership with Macedonia and identifi ed eight key short- term priorities: full implementation of the SAA, constructive and inclusive dialogue between the politi cal parties, eff ective implementation of the law on Binational Macedonia? 197 police, implementation of judiciary reform, implementation of anticorrup- tion legislation, removal of po liti cal interference in recruitment of civil servants, increased job creation, and development of the general business environment.84 In March 2008, the Eu ro pe an Commission adopted a Com- munication on the Western Balkans, setting out benchmarks for each state to proceed in the accession process—the previous priorities set out in the Accession Partnership were to be the benchmarks for Macedonia’s progress.85 In October 2009, the Eu ro pe an Commission recommended opening negotiations for Macedonia’s accession into the EU. Th is overview demonstrates that fairly smooth, straightforward progress was made from the signing of the SAA in 2001 until the Euro pe an Commission’s decision in October 2009 to recommend the start of accession negotiations. Since then, however, progress has stalled over the country’s name dispute with Greece. Despite the Euro pe an Commission’s favorable recommendation, EU member states failed to take the unanimous decision to start negotiations, owing to opposition from Greece. As discussed below, the stalemate has had challenging and diffi cult consequences aff ecting the country’s reform agenda and inter- ethnic relations. With the deadlock ongoing, EU institutions have made an eff ort to revive the pro cess. President of the Euro pe an Commission José Manuel Barroso and EU Enlargement Com- missioner Stefan Fuele, on a tour of the Western Balkans in April 2011, urged the country to speed up its reforms and step up eff orts to resolve the name issue with Greece.86 Aside from the problems relating to the name issue, how eff ective has the EU been in facilitating OFA implementation? Under Annex C of the OFA, the “parties invite the international community to facilitate, monitor, and assist in implementation,” with the EU having an explicit coordination role. As an offi cial remarked, the EU has been acting as a “guarantor of the Ohrid Framework Agreement” alongside the other main international actors— the U.S. and NATO— and supported by the OSCE.87 Importantly, the EU has incorporated its monitoring of OFA implementation into Mace- donia’s accession procedures. Th e state’s progress in implementing the various legislative mea sures arising from the agreement has been highlighted by successive Euro pe an Commission progress reports. For instance, the 2006 progress report emphasizes that full implementation of the OFA “remains essential to foster a positive environment for further reforms.”88 In an eff ort to encourage further implementation, progress reports have included an explicit focus on the issue areas set out in the OFA, particularly 198 Chapter 8 decentralization and public administration. Th e 2007 progress report noted developments on decentralization implementation due to cooperation between central government and the municipalities, adding that much more needed to be done, including municipal tax collection and debt management. Regarding public administration, the Commission noted limited progress due to politi cal infl uence of state employment, which disrupted the government’s functioning, contributing to a legislative backlog and lack of expertise.89 In its key fi ndings of the 2008 progress report, the Commission stated that implementation of the OFA “continues to consolidate multi- ethnic democracy,” referring to the recent law on languages, as well as progress on judicial reform and police reform, while also highlighting an ongoing need to strengthen po liti cal dialogue in parliament.90 In the conclusions to its 2010 progress report, the Commission described the power-sharing coali tion as stable and noted the cooperation among the po liti cal parties. Almost a decade since the signing of the OFA, the Commission claimed the agreement remained “an essential element for democracy and rule of law in the country.” 91 Th ough noting the government’s progress in implementing the law on language, decentralization, and public administration, the Commission continued to call for sustained eff ort via politi cal dialogue to ensure the full implementation of the OFA. An advisor suggested that the EU is “an important driver” in Macedonian politics, particularly given the importance of EU fi nancial assistance in enabling OFA implementation.92 Th is view of widespread support for EU membership among the parties is unsurprisingly cor- roborated by politicians interviewed for this research, some of whom referred to the “lack of alternative” for Macedonia’s future or the “ultimate goal” of EU membership. Th at said, there was a reported souring of relations between Skopje and Brussels relating to the lack of progress made by Gruevski in resolving the name dispute.93 Th ere is also evidence that in its eff orts to ensure OFA implementation, the EU has responded to po liti cal crises that threaten policy reform or inter- ethnic relations. An EU offi cial contrasted their role in Macedonia with that in Bosnia, where the international community “is more muscular” in helping to bring about reform. At the same time, he suggested that the EU’s role in Macedonia is not “to sit back and watch” developments unfold that challenge ethnic relations or put the country’s accession process in jeopardy. It appears that the EU sought to restore cooperation among parties following parliamentary boycotts as well as to encourage progress on policy reform. Following DUI’s boycott of parliament from the end of January 2007, inter- Binational Macedonia? 199 national actors stressed that the party should return to parliament before any progress could be made on Euro- Atlantic integration. Th e problem of DUI’s absence from parliament was noted by the Euro pe an Commission in progress reports: “All politi cal forces should commit themselves to working constructively within the democratic institutions.” 94 It seems that the EU delegation and the U.S. Embassy in Skopje facilitated inter- party talks between VMRO- DPMNE and DUI, reported as “heavy behind- the- scenes involvement.” 95 Th ese talks led to the May Agreement, whereby DUI’s Ali Ahmeti declared that agreement had been reached on a number of issues relating to language legislation and the status of former NLA fi ghters. Ac- cording to a SDSM politician, the EU delegation has played a hugely positive role on a daily basis in its eff orts to maintain “a climate of trust and confi - dence” among the politi cal parties.96 It appears that the EU delegation and the U.S. Embassy have, on occasions, put pressure on domestic elites to resolve ad hoc politi cal crises. For instance, it seems that the EU encouraged Gruevski to distance himself from the controversial encyclopedia published by the Macedonian Academy of Arts and Science in 2009.97 In the end, the power- sharing parties issued a statement calling for inter- ethnic understanding and respect. As usual in the EU integration process, the Euro pe an Commission set a number of criteria for Macedonia to fulfi ll on its path to membership. Th ese criteria include the Copenhagen criteria (including respect for and protection of minorities) as well as country- specifi c benchmarks outlined in the Accession Partnership. Given the EU’s role as guarantor of the OFA, implementation of the agreement has been an important component in the country’s accession pro cess. Noted above, there is a view within the EU delegation offi ce that OFA implementation is less a “tick box exercise” and more about ensuring the parties engage in dialogue and cooperate on policy reform. Th e need for “constructive po liti cal dialogue” is a key priority under Macedonia’s EU Accession Partnership. George Vasilev suggests the EU’s focus on consensus politics and “politi cal dialogue” in Macedonia refl ects “a move away from substantive principles toward procedural ones as the centrepiece of conditionality.” 98 He suggests that parties in both Macedonia and Bosnia are encouraged to comply “in a manner that refl ects EU procedural standards.” 99 Th is emphasis on constructive po liti cal dialogue among po liti cal parties is perhaps unsurprising, given the frequency with which parties have boycotted parliament in recent years. Moreover, the emphasis on po liti cal dialogue arguably derives from the crisis following government formation 200 Chapter 8 negotiations in 2006 that excluded the largest Albanian party from power. In terms of the eff ect of this focus on consensus politics and EU conditionality more generally, Vasilev provides a positive assessment, maintaining that “membership conditionality has been facilitating rapid and relatively seamless change in Macedonia.”100 In comparison with the more diffi cult situation in Bosnia, he suggests that “the interests voiced by ethnic leaders are complementary and stand to be furthered, rather than obstructed, through EU membership.”101 For Macedonians, EU membership has been viewed as a vehicle to secure the territorial integrity of the state, further incentivizing Albanians to pursue their communal interests within the boundaries of the existing state. Given the sensitivity of Macedonian nationhood relating to controversy over the state’s name as well as religion and language, EU membership is also regarded as the best way to secure state recognition in the international arena. For Albanians, there is also much to be gained via EU membership, including fi nancial support, a focus on inter-ethnic relations and minority rights, and the imprimatur of the EU principle of subsidiarity in support of eff ective decentralization. Th ough Macedonians and Albanians share in their support of EU membership, progress toward that goal was blocked by controversy over the name issue. As is well known, Greece has objected to the name “Republic of Mace- donia” for implying territorial claims on the Greek province of the same name. With the name dispute ongoing since Macedonia proclaimed its in- depen dence in 1991, the United Nations has facilitated (as yet inconclusive) talks between the two sides. Greece remains opposed to the state’s use of “Republic of Macedonia” and wants to see a revision of the constitution, with possible compromise based on a geograph i cal qualifi er such as “North- ern” or “Upper” Macedonia. At NATO’s 2008 summit in Bucharest, Mace- donia’s membership in the orga ni za tion was vetoed by Greece on the basis that a solution on the name issue had not been reached. In the offi cial declaration, NATO members indicated that “an invitation to the former Yugo slav Republic of Macedonia will be extended as soon as a mutually acceptable solution to the name issue has been reached. We encourage the negotiations to be resumed without delay and expect them to be concluded as soon as possible.”102 And while the Euro pe an Commission recommended opening accession negotiations with Macedonia in 2009, progress was likewise blocked due to Greek opposition. Yet Macedonian elites continued to reject Greek demands when many other countries around the world, including the U.S., have accepted the name. Prime Minister Gruevski pledged to put to a refer- Binational Macedonia? 201 endum any compromise reached in negotiations between the two states.103 President Georgi Ivanov remarked, “Every solution that contributes to al- tering the reference to our language and our uniqueness is unacceptable for Macedonia. We cannot agree that Greek positions and proposals are promoted as mutually acceptable positions.”104 Arguably, the stalemate over the name issue led to a number of problems for Macedonia, including strained inter- ethnic relations. On one level, the stalemate weakened the EU’s leverage. Zoran Ilievski and Dane Taleski suggest that the name issue “removes a major instrument for securing compliance from the EU’s toolbox and it negates the major incentive for Macedonia’s po liti cal elites to comply.”105 Some Albanian politicians suggest that because of Greece’s decision to block Macedonia’s entry into the EU and NATO, and the response of VMRO-DPMNE to pursue its policy of Skopje 2014 and general antiquization, the impetus for fulfi lling EU conditions has faded.106 An EU offi cial claims that membership remains an important incentive for progress but acknowledges that there has been a “loss of momentum because of the delay” arising from the stalemate.107 Th e EU’s position is that resolution of the name issue is crucial for kick- starting accession negotiations. For instance, the Euro pe an Commission’s 2010 Progress Report stated, “Main- taining good neighbourly relations, including a negotiated and mutually acceptable solution to the name issue, under the auspices of the UN, remains essential.” In addition to the eff ects of the stalemate on the accession pro cess, it is also arguable that the dispute has had an adverse eff ect on relations between the power- sharing parties and the two groups more broadly. Vasilev notes the eff ects of delay on inter- ethnic relations, “given that each group’s commitment to cooperation and compromise is forged on the promise” of Euro- Atlantic integration.108 A SDSM politician commented on the emerging internal ethnic division over the name issue, suggesting that Albanians look to the progress made by Albania in joining NATO and the success of Kosovo inde pen dence and feel they are “lagging behind in Macedonia because of the behavior of the prime minister and his party” in failing to reach a compromise with Greece.109 A DUI politician commented that even though his community has been placed in a “delicate position” on the name issue, they have displayed “feelings of solidarity, tolerance, and cooperation with Mace- donians.” Nevertheless, he warned this sentiment could not last forever and that it is up to VMRO- DPMNE to resolve the problem.110 Increasingly negative inter- ethnic relations have also been acknowledged by external actors, 202 Chapter 8 with one international offi cial suggesting that the name issue has the potential to impact inter-ethnic relations “very seriously and may be extremely divisive.”111

Summary

On 12 August 2011, celebrating the tenth anniversary of the signing of the OFA, the principal external actors in Macedonia (the EU, the U.S., NATO, and the OSCE) issued a joint statement applauding the country’s progress in becoming “a stable, multi-ethnic, multi-cultural and inter-religious society.” Th ey pledged to continue to assist in the full implementation of the OFA in the context of Euro-Atlantic integration. Acknowledging the progress made since the country “stood at the brink of an all- out civil war” in 2001, the international actors nevertheless issued a warning to domestic elites “to hon- our the spirit of the agreement, to reinvigorate the pledges of the signatories toward tolerance and transparency, and to renew their commitment to building a unifi ed and secure state for all citizens—regardless of their ethnicity, gender, religion, or politi cal affi liation.” 112 Th ough much has been achieved in Macedonia, there is still work to be done in cementing the country as a stable multi- ethnic democracy. Overall, the experience of power sharing in Macedonia has been relatively successful, with evidence of cooperation in peace implementation. A note of caution is required, however, given the challenges arising from the slowing down of EU-related reform, the stalemate over the name issue, and the instances of cultural division, most notably brought about by Skopje 2014. Th is somewhat mixed record points to a number of lessons for power-sharing democracy in deeply divided places. Macedonia diff ers from the other two cases in the book given the country’s tradition of informal power sharing. Including Albanian parties in the post- independence governments was arguably important for some mea sure of internal stability throughout the 1990s. It was not enough, however, to stem the tide of increasing inter-ethnic division and the fallout of the Kosovo crisis that led to confl ict in 2001. Th e practice of power sharing reviewed in this chapter confi rms the importance of these informal, liberal structures of power sharing on the basis of post-electoral coali tion formation negotiations. Th e tradition of power sharing appears suffi ciently cemented as an accepted norm of the politi cal system. Th e Macedonian case also off ers an important Binational Macedonia? 203 lesson on the debates surrounding inclusion and exclusion in power sharing. Th e decision by VMRO- DPMNE in 2006 to form a government with the DPA, the second largest Albanian party, precipitated a po liti cal crisis. Al- though most Albanian politicians interviewed suggest there is no need to formalize rules for executive formation, they nevertheless remarked that it is important for the country’s democratic legitimacy to have the largest Alba- nian party in government. Th e case supports the view that veto rules are important for the minority community in power sharing. In Macedonia, Albanian politicians emphasize the signifi cance of the concurrent majority voting procedure in parliament, known locally as “the Badinter rule,” in protecting their group rights. Th e introduction of the rule is oft en judged by Albanian elites as a central feature of the peace agreement. It is also important that cooperation and compromise between the ethnic parties in power has come about via informal negotiations and backroom deals rather than the application of formal institutional rules. Th is cooperation is oft en facilitated by the main external actors, the EU delegation, and the U.S. Embassy. In a less robust manner than employed in Bosnia, external actors have, nevertheless, played an important role in encouraging policy reform. Overall, the case points to the evolution of an increasingly binational state based on power sharing between the two communities. Th e implementation of the OFA has gradually secured additional group rights for Albanians in the areas of language, local government, state employment, policing, and education. Macedonia’s ongoing challenges lie in the management of cultural contestation and contentious identity politics. Conclusion

So what does this research mean for our understanding of the operation of power-sharing democracy in deeply divided places? What can be derived from the study in terms of how cross-community cooperation might be fos- tered within power-sharing executives? In this chapter, I return to the main themes set out in Chapter 1: the process of institutional design leading to power- sharing executives; the institutional choice between a moderate middle and an inclusive executive; the use of formal rules versus negotiation in executive formation; the impact of veto rules; and the role of external actors in the design and implementation of power-sharing agreements. I provide a summary of the main fi ndings and consider their implications for power- sharing practice, power- sharing theory, and future research.

Forming and Transforming Power- Sharing Rules

Ostrom writes, “To understand institutions one needs to know what they are, how and why they are craft ed and sustained, and what consequences they generate in diverse settings.”1 Th e craft ing and consequences of po litical institutions in deeply divided places continues to preoccupy academics and policy makers seeking to uncover the most appropriate institutions to promote peace and stability. Apart from the type of institutions adopted, scholarship needs to consider the ways in which such institutions are created, so as to appreciate potential shortcomings in their operation. In post- confl ict situations, internal and external actors engaged in peace negotiations presumably have some room for maneuver in deciding the state’s confi guration (unitary or federal) and its institutional specifi cs. Some scholars have considered why actors arrive at a power- sharing deal. Caroline Hartzell and Matthew Hoddie remind us, “Former combatants require assurances that no single group will be able to use the power of the state to secure what they failed to win on the battlefi eld, and perhaps threaten the very survival of Conclusion 205 rivals.” Given the need for post-confl ict institutions that will protect the range of signatories to the peace agreement, “Power sharing serves as the mechanism that off ers this protection by guaranteeing all groups a share of state power.”2 Beyond the reasons why parties agree to share power in the post-confl ict environment, less focus has been paid to the process of institutional design involved in the creation of power-sharing democracy. As discussed in Chap- ter 1, the literature questions the reality of institutional design in deeply divided places. Horowitz makes the point that the proliferation of actors makes it unlikely that a design for politi cal institutions will be produced at the beginning of peace negotiations and then adopted by parties.3 Th e evidence presented here confi rms that power sharing is more likely to be the outcome of a multiplicity of actors and the interaction of their respective preferences. Th is interaction among actors in peace negotiations involves dealing with institutional preferences held by parties representing the contending groups. Th e accommodation of these preferences (through compromise, side payments, or concessions) arguably constitutes a pro cess of institutional design leading to cross-community power sharing. Th e pro cess of institutional design, however, may diff er depending on circumstances. Th e decision to establish executive power sharing might be largely imposed by external actors in peace negotiations. It may be infl uenced by the state’s historical experience. Or the idea of power sharing may have been circulating in parties’ proposals for some time. Th e Northern Ireland case shows that in terms of how a power-sharing deal came about in 1998, executive design was the result of an evolutionary process of institutional design from the 1970s, with some further details (e.g., the roles of the fi rst minister and deputy fi rst minister) negotiated between the two main parties at the time, the UUP and the SDLP. Th is case points to the need to consider how a process of institutional design involving a multiplicity of actors can nevertheless lead to a fairly coherent (in this case, consociational) framework. We are reminded of Horowitz’s recommendation for institutional coherence.4 Coherence can arrive as a result of actors’ pragmatism in terms of what would be accepted in the peace settlement. Notably, the election of the fi rst minister and deputy fi rst minister via parallel consent under the GFA is more an instance of integration than consociation. Th e UUP and SDLP agreed on the cross- community election of the two top posts with equal power.5 Th e rule was, however, revised in a consociational light in 2006 without major consequences. Arguably, then, executive 206 Conclusion power sharing in Northern Ireland came about not as the imposition of a fresh institutional blueprint, but as the outcome of pragmatic or “common sense” application of parties’ preferences. Moreover, though the GFA was ultimately the result of peace talks, there was a much longer process of discussions that set down the pa ram e ters of a future agreement. In contrast to this process of evolutionary design in Northern Ireland, executive power sharing came about in Bosnia as the result of “hothouse” peace negotiations led by external actors and a diffi cult compromise among the antagonists. Certainly, there was no blueprint for politi cal institutions prepared by negotiators for the onset of the Dayton talks. As noted in Chap- ter 5, EU representative Carl Bildt recalled his frustration at the circulation of numerous draft s of the constitution, including various institutional options.6 Yet the po liti cal reality dictated by the battlefi eld, and the U.S. team’s accep tance of the two entities, largely determined the creation of power- sharing institutions. Th e three groups insisted that within a single state they should maintain group rights and their respective share of politi cal positions. Hence, veto rules and predetermined positions were included in the constitution. Th e three groups’ emphasis on group rights also refl ected their historical experiences in the former Yugoslavia, given that the politi cal system was based on decentralization and cultural autonomy with proportional represen ta tion in public administration and Communist Party structures. On one level, the deal secured at Dayton was largely imposed by the external actors. On another level, for the respective groups, there was a need to secure nothing less than formal power sharing with veto rules and predetermined po liti cal positions guaranteed in the state constitution. Th e case also supports the idea that cross-community power sharing is the outcome of preferences on the part of the contending groups, interacting with external actors’ incentives, that help shape the nature of inter-party bargaining. In Macedo- nia, power sharing was a central feature of the post-independence politi cal system. Po liti cal parties were not constitutionally obliged to share power; nevertheless, power sharing refl ected the po liti cal reality of the new state (and was accepted by the largest Macedonian party at the time, the SDSM) in the need to secure Albanian participation for demo cratic stability. At the Ohrid peace talks, this tradition of power sharing was cemented by the inclusion of the concurrent majority voting procedure in parliament and the ongoing expectation that the Albanian community would be represented in government. Conclusion 207

Th is book also illuminates how power- sharing institutions may change once agreed upon and implemented. More generally, scholars have questioned the extent to which institutions are relatively fi xed or exiblefl and whether institutional change is a discrete event or the outcome of a process of incremental adjustment.7 B. Guy Peters summarizes the diff erent approaches to institutional analysis on these questions.8 On the one hand, institutions are treated as an equilibrium that will not change unless there is some “punc- tuation.” Th e “punctuated equilibrium” model suggests that institutions will change as a result of some exogenous shock or environmental shift .9 Alterna- tively, we can treat institutions as “human constructions” that are “subject to the whims of the very people (or at least analogous) that created them in the fi rst instance.”10 Recently there has been greater focus on the need to improve institutional theory to account for how and why institutions change. Th ese contributions help move us beyond exogenous shocks to the signifi cance of incremental, endogenous change. James Mahoney and Kathleen Th elen write that “institutions oft en change in subtle and gradual ways over time” and that change might be “slow and piecemeal.”11 Th ey argue that “institutional change oft en occurs precisely when problems of rule interpretation and enforcement open up space for actors to implement existing rules in new ways.”12 Th is focus on whether change occurs as an event or a longer process, the reasons for such change, and the politi cal space actors take advantage of, is relevant to our understanding of the practice of power-sharing democracy. Arguably, it is important to assess whether revisions to power- sharing structures come about to rectify “mistakes” in the initial agreement; to tempt previously rejectionist politi cal forces to sign up to power sharing; or whether changes come about as a matter of evolution, to iron out glitches in institutional rules. Th e evidence points to the capacity of power-sharing executives to change, whether this change is planned or takes place more gradually. In Northern Ireland, revisions to the power-sharing institutions occurred as the result of the St. Andrews Agreement of 2006. We can treat these changes as the outcome of an event, inter-party talks, and the proposals published by the British and Irish governments. Yet much of the content of the document had already been circulated in parties’ criticisms of the GFA; their respective positions during the Review of the Agreement in 2004; the two governments’ subsequent document, Proposals for a Comprehensive Agreement; and the parties’ responses to potential reform. Revision of power-sharing rules 208 Conclusion was, therefore, less the outcome of an event and more the result of an ongoing process. Th e reasons for institutional change refl ected the need to deal with changed po liti cal circumstances, particularly in light of the DUP becoming the largest party in 2003. Th e source of institutional change can be seen as partly deriving from a shift in the balance of power. Th e case demonstrates how a shift in the balance of power among parties following elections will likely change the dynamics of power sharing. Eff orts to incentivize the DUP to form a new power-sharing executive with Sinn Féin necessitated some accommodation of the former’s preferences, which called for revisions to the GFA. Th e experience in Bosnia also points to the evolution of power- sharing rules over time, in the context of a process of reform designed to improve the functionality of the system. As Paul Pierson suggests, institutional design will likely include some dysfunctional elements or mistakes that necessitate revision.13 In the Bosnian situation, however, institutional change to address shortcomings of the DPA has more oft en been the result of direct intervention by the high representative, as in the revisions to the entity constitutions in 2002 and changes to the operation of the Council of Ministers in 2002 and 2007. As discussed in Chapter 6, external actors have called for comprehensive constitutional reform to render power sharing more effi cient and eff ective. Th e pro cess of reform is described as “constitutional evolution” by the EU.14 In terms of politi cal actors’ responses to potential reform, the Bosnian case shows that politicians may resist institutional change they see as threatening to their group interests and pushed by outsiders. An interesting question arises here about who decides institutions are dysfunctional and in need of revision. It might be plain to see that the institutions do not work effi ciently or eff ectively. Yet the dysfunctional nature of the institutions may not be a problem for everyone. Indeed, a mea sure of stalemate or ineffi ciency may even suit some politicians. Th ey may even resist eff orts by others to improve the functioning of power sharing, especially when those calling for reform are external actors with questionable legitimacy and accountability. Diff erent from the notion of institutional persis tence, human agency can help explain why institutional change may not readily take place, despite the call from some (oft en external) quarters for institutional ineffi - ciencies to be addressed. In Macedonia, institutional change took the form of formalizing power sharing in the context of peace negotiations at Ohrid— more a consolidation than a transformation. Th ough power sharing had been taking place since Conclusion 209 in de pen dence, new institutional rules and a commitment to policy reform were agreed between the two groups. New rules included the Badinter principle, or concurrent majority voting in the assembly on specifi ed policy issues. Parties also committed to future institutional change in several areas, including decentralization, improved represen ta tion for Albanians in public administration, and the use of the Albanian language. Th e literature on post-confl ict institutions, including power-sharing democracy, can benefi t from the questions posed in the wider academic literature on institutionalism. Th ere are complementary concerns between the two literatures relating to the importance of structure versus agency in explaining politi cal behavior, how rules may be under constant renegotiation by those who designed them in the fi rst place, and how change can be incremental rather than revolutionary to address inadequacies or unintended consequences. Th ough power sharing may represent a fairly narrow fi eld of interest within the wider social science debates on the role of institutions, these fi ndings are arguably important for real- world politics, given the high stakes that come with institutional choice in post-confl ict, deeply divided places. Th ese insights are important for how we understand and think about peace pro cesses. Researchers and policy makers would do well to consider how institutional arrangements set out in a peace agreement may be constantly renegotiated by groups, either to improve their functioning or to meet po liti cal preferences that had not been secured in the original deal.

Institutional Rules and Cooperation

Th roughout the book, I have sought to determine the impact of institutional rules on the extent of inter- ethnic cooperation within power- sharing executives. Th is is a concern to add to existing knowledge of power-sharing practice and to say something about which institutional rules are feasible and “what works” in power- sharing democracy. As post- confl ict negotiations are preoccupied with agreeing on a constitution and politi cal institutions, this topic continues to be crucial for academics and policy makers. Th e case study chapters have presented evidence of coali tion parties working together ( joint decision making) as indicators of intra- executive cooperation. Overall, I argue that though inter-ethnic cooperation is far from guaranteed, power sharing can provide po liti cal space for an atmosphere of “jointness” or accommodation. Power sharing involves institutional choices: an 210 Conclusion inclusive executive or a moderate middle co ali tion, executive formation by negotiation or formal institutional rules, and the existence of group vetoes. A central question is whether and how these institutional choices matter for cooperation.

A Fully Inclusive or Moderate Middle Co ali tion?

One of the most challenging issues for policy makers relates to whether power sharing should be fully inclusive (providing room in government for hard-liners, the extremes, or politi cal forces with previous links to violence) or concentrated on a moderate middle. Having “everyone in the tent” may mean greater legitimacy among the population as a whole and reduce the threat of violence from outside, given the inclusion of hard-liners who have embraced purely democratic means to pursue their politi cal objectives. Th e argument for an exclusionary, moderate middle coali tion relates to the idea that these parties may fi nd it easier to compromise and that they will be able to shore up suffi cient support by forging ahead in a spirit of reconciliation. As discussed in Chapter 1, this institutional choice is a principal focus of the debates between consociationalists and scholars adopting a more incentives-based approach (the traditional Lijphart versus Horowitz debate). My research fi ndings support a more inclusive approach, albeit with qualifi cations. Th e evidence suggests that an inclusive executive may be preferable to a moderate middle, provided that parties in government agree to work within the politi cal institutions. Parties in government must have renounced violent means to pursue their po liti cal goals, and they must be prepared to operate the institutions in a stable manner (e.g., fulfi lling their ministerial obligations while respecting the right of other parties to do likewise). Th is requires an onus on inclusion in order to lock parties into a negotiated agreement and peace implementation. Th e intention here is for executive design to provide the politi cal space for moderation through an inclusive approach. My fi ndings also show that moderate middle co ali tions may well be lacking in moderation; certainly, they may be no more moderate than fully inclusive ones. To an extent, I agree with Ian O’Flynn, who seeks to debunk the either/or debate on inclusion and moderation.15 At the same time, however, I confess my aim is to off er some practical solutions for policy makers who want to know what works. In this sense, I advocate an onus on inclusion that may open up the space for cooperation and mod- Conclusion 211 eration, rather than a more limiting exclusionary focus on the moderate parties. Chapter 2 shows that Northern Ireland’s fi rst experience of power sharing was constrained by opposition from loyalists and republicans opposed to the Sunningdale communiqué. Yet the pa ram e ters of the new po liti cal system could not have been stretched to include these actors, who were opposed to the principle of power sharing. As McGarry and O’Leary note, had the 1974 executive included the “extremes” from both communities, “they probably would have destroyed it. Th is is because radicals at the time were virulently opposed to power sharing and committed to militancy.”16 In any case, the British government hoped to shore up support for the moderates, attempting to isolate the “extremes.” By the 1990s, the two governments had embarked on a more inclusive peace process following changes within the republican movement. Th e need for inclusion was largely resolved in 1998, when republicans and loyalists, nationalists and unionists were all prepared to support power sharing under the GFA. Moreover, the deal was endorsed by the electorate in referendums in Northern Ireland and the Republic of Ireland. Th e fully inclusive executive under the GFA has been more successful than the exclusive moderate middle executive that resulted from Sun- ningdale. Indeed, it seems fair to suggest that one of the reasons power sharing came about and muddled along (despite its many diffi culties) was that Sinn Féin was included, the agreement having locked them into the system and brought about eventual IRA decommissioning. Th e issue of inclusion continued to be central to the eff orts to restore power sharing following suspension of the institutions in 2002. With the DUP and Sinn Féin having cemented their positions as the largest parties of their respective blocs, their engagement and support for power sharing were crucial. As Rick Wilford notes, this support, particularly on the part of the DUP, came about as a result of the changed security and politi cal context since suspension.17 Changes to the institutional rules governing the selection of fi rst minister and deputy fi rst minister, as well as rules on ministerial authority, allowed the DUP to point to revision of the GFA. Th at these two parties have been willing to operate the power-sharing system in a fairly constructive manner provides further evidence that an inclusive executive can help deliver inter- ethnic cooperation and a measure of stability. Having been satisfi ed that their self-determination claims have been addressed, and having gained reciprocal side payments, the two main parties have been largely content to implement the agreement. 212 Conclusion

Similarly, the experience of power sharing in Bosnia points to the diffi - culties for a moderate middle co ali tion to secure cooperation and policy reform. Th e Alliance for Change coali tion (2000–2002), essentially patched together by the international community, did not produce the expected level of policy reform and compromise. Th e coali tion suff ered from having a weak majority in the state House of Representatives and had to rely on a working arrangement with parties from Republika Srpska. Th ough the coalition managed to make some policy progress, this was rather limited, prompting the high representative to impose decisions. Tensions among the parties in power led them to contest the 2002 elections on separate platforms. But to what extent have the more inclusive executives (including more “radical” parties) demonstrated inter- ethnic cooperation? Th e Bosnian case shows that the inclusion of more hard- line parties in a wider co ali tion is far from a magic solution for inter-ethnic cooperation. But neither is a more moderate, exclusionary coali tion. At best, we can argue that inter-ethnic cooperation will be more likely in power sharing when parties’ politi cal preferences (particularly relating to competing self-determination claims) have been satisfi ed. As Chapters 5 and 6 show, this has been far from the case in Bosnia. Th e experience of power sharing in Macedonia is interesting, given the diff erent confi gurations of moderate and more radical parties in power. Post- electoral negotiations have been formed by a more moderate Macedonian party and a more radical Albanian party and vice versa. Since 2008, power sharing has consisted of the more radical party on each side, VMRO-DPMNE and DUI. Th is variation allows us to consider whether the presence of a more radical or “nationalist” party of each group leads to more or less cooperation than when power sharing includes parties accepted as more moderate. Arguably, there has been no clear diff erence in power sharing with the two more radical parties in power. Power sharing between VMRO- DPMNE and DUI has delivered a mixed record in peace implementation. Th ese two parties have cooperated in an expedient manner when they have needed to tackle pressing policy issues. With both parties in support of the EU integration process, progress was made on decentralization and equitable repre sen ta tion. Th ere have, of course, been problems and strained relations over policy issues including language legislation, the operation of the Com- mittee on Inter-Ethnic Relations, the Skopje 2014 project, and stalemate over the name dispute with Greece. Conclusion 213

A related point to the debate on inclusion/exclusion relates to parties’ attempts to either exclude others from the po liti cal institutions or to boycott the institutions themselves, an act of self-exclusion. In the case studies, there have been boycotts and attempts to exclude parties that threatened the operation of power sharing. In terms of boycotts/self-exclusion, we can distinguish between those carried out by governing parties and boycotts by opposition parties. Governing parties’ boycotts have included absenteeism on the part of various Bosnian parties from the Council of Ministers, Sinn Féin’s decision not to attend executive meetings, the DUP’s boycott of the North- South Ministerial Council, and VMRO- DPMNE’s decision not to attend the Committee on Inter-Ethnic Relations. Th ough parties employed such boycotts as tactical maneuvers to apply some pressure on their coalition “partners,” such tactics highlight the need to cement politi cal will for power sharing to function eff ectively. Th e issue of boycotts/self-exclusion has been addressed by external actors engaged in peace implementation. In Bosnia, High Representative Miroslav Lajčák imposed a decision revising legislation on the Council of Ministers to make it more diffi cult for a party to block the Council and parliament through absenteeism.18 Somewhat ironically, the decision led to a stand- off between the high representative and the Bosnian Serb party SNSD. Nikola Špirić resigned as Chair of the Council of Ministers, and his party threatened to withdraw from the power- sharing institutions. Boycotts have also been a problematic feature of power sharing in Macedonia, particularly problematic during DUI’s parliamentary boycott in 2007, precipitated by ongoing frustration at being excluded from government following the 2006 elections. In terms of external actors’ responses to such tactics, international offi cials spoke of their eff orts to persuade the respective parties to operate the institutions fully.19 In Northern Ireland, self- exclusion and eff orts by some parties to exclude others have provoked controversy. Th e latter situation has been perhaps more controversial, evidenced by DUP’s attempts to exclude Sinn Féin ministers from the executive over slow progress on IRA decommissioning.20 In a similar eff ort to put pressure on republicans, fi rst minister David Trimble refused to nominate Sinn Féin ministers to attend meetings of the North-South Ministerial Council, an action that was referred to the High Court and ruled to be unlawful.21 Th e St. Andrews Agree- ment published by the British and Irish governments stipulated that all parties “need to be wholeheartedly and publicly committed, in good faith 214 Conclusion and in a spirit of genuine partnership, to the full operation of a stable power- sharing Government and the North-South and East-West arrangements.”22 Moreover, the statutory ministerial code obliges ministers to participate fully in the executive, the North-South Ministerial Council, and the British- Irish Council.23 Th ese examples show that power- sharing executives, whether inclusive or exclusive, may be impacted by parties’ decisions to opt out of politi cal institutions. Th e book’s fi ndings highlight that the dynamics of power sharing are about more than inclusive versus exclusionary executives— there may be elements of self- exclusion that may be problematic for the operation of a power- sharing co ali tion. If parties self- exclude from the institutions, power sharing will hardly have much chance to work eff ectively. Self- exclusion may make a mockery of power sharing if parties feel they can frequently leave the coali tion. Under such circumstances, how might it be possible to secure stable power sharing? Should parties have the right to self- exclude from the institutions? If they remain in the executive, should they have the right to opt out of specifi c policies? With regard to self- exclusion from power sharing, it would arguably be preferable for parties to negotiate with their coali tion partners over sensitive policy issues or employ veto rules to protect their group’s interests. Opting out of specifi c policies (leaving executive “partners” to pursue such a policy for their respective community) may be less controversial and may be agreed within the co ali tion. But to limit instances of self- exclusion, parties may well agree to obligations for each other in the form of a statutory ministerial code, as in Northern Ireland. Researchers and policy makers (especially external actors engaged in peace implementation) will need to keep a watchful eye on the variable eff ects of vetoes and boycotts/self-exclusion on the operation of the politi cal system.

Executive Formation: Via Rules or Negotiation?

Beyond the issue of inclusion, parties in peace negotiations will likely be faced with several key institutional choices. Executive formation rules are one such choice. Parties can opt to devise formal rules for executive formation or leave the makeup of the governing coali tion to the outcome of pre- or post- electoral bargaining. Executive formation rules may be corporate or liberal, corresponding to the modifi cations to consociational theory put for- Conclusion 215 ward by McGarry and O’Leary. Executive formation rules may be corporate, in that they secure executive positions for the contending groups through predetermined positions according to ascriptive identities such as religion or ethnicity. Alternatively, executive formation rules may be liberal, in that positions are accorded to whichever groups are victorious in democratic elections.24 I concur with the existing literature that liberal/self- determined power sharing is preferable to corporate rules that specify positions for certain groups. As discussed in Chapter 5, corporate executive formation rules apply in post- Dayton Bosnia, where positions are predetermined according to ethnicity. Power sharing requires the inclusion of representatives from the three “constituent peoples” and on an entity basis. Under the DPA, the state three-member, directly elected presidency includes one Bosniak and one Croat from the Federation and one Serb from the RS. Th e presidency nominates the chair of the Council of Ministers who, in turn, nominates the ministers and deputy ministers. Th e chair and ministers are then subject to a legitimizing vote in the House of Representatives.25 For the three groups, predetermination of executive positions off ers certain guarantees that would be denied under a single president model or if ministerial positions were settled solely as a result of bargaining. Despite (or even because of ) these guarantees, the use of predetermined positions in Bosnia has been subject to considerable controversy. Notably, international actors have called for reform of these rules. In 2005, the Council of Europe’s Venice Commis- sion reported on the compatibility of the Bosnian constitution with the Eu- ro pe an Convention on Human Rights (ECHR), focusing on the composition and election of the presidency and the House of Peoples. Th e Commission noted the tension between predetermined positions for the three constituent peoples and the principle of individual rights and equality of citizens, particularly the exclusion of “others” from politi cal positions.26 As discussed in Chapter 6, the Eu ro pe an Court of Human Rights found Bosnia to be in violation of the ECHR on the rights to equal treatment and nondiscrimination by denying “others” the right to stand as candidates for the presidency. Reform of these corporate rules remains unforthcoming, as parties unsurprisingly seek to retain group guarantees in executive formation. In Northern Ireland, the GFA provided for a novel form of executive formation via sequential portfolio allocation under the d’Hondt rule. Critics of the d’Hondt procedure have argued that the method creates a “mandatory coali tion” and a “carve-up” of positions between the parties representing the 216 Conclusion two communities.27 Th ough the d’Hondt method is the designated procedure for executive formation under the GFA, it is a stretch to describe the resultant executive as “mandatory” because parties are free to opt out. And allegations of the parties “carving up” ministerial seats would likewise be applicable under post-electoral negotiations. In terms of the benefi ts of using d’Hondt, McGarry and O’Leary emphasize its positive attributes, (i.e., the method is proportional, sequential, non-exclusionary, and incentivizes parties to take their places in the executive).28 Th e use of the sequential portfolio allocation method helped facilitate the fully inclusive nature of power sharing post- GFA. Arguably, a power- sharing executive including Sinn Féin would not have been possible if executive formation had been dependent on post-electoral negotiations, such was the ongoing animosity among parties, largely owing to the slow pace of IRA decommissioning. In this sense, d’Hondt was useful for securing inclusion and proportionality in an automatic way without having to engage in protracted negotiations. Th e St. An- drews Agreement changed the election of the fi rst minister and deputy fi rst minister via cross- community vote to an automatic nominations procedure. As John Coakley notes, this change “would absolve the DUP of responsibility for electing Sinn Féin’s Martin McGuinness” as deputy fi rst minister.29 On one level, we may lament the loss of a cross-community vote in the election of the top two posts. Th e election of Trimble and Mallon post- GFA signaled an element of inter- ethnic reconciliation and preparedness to jointly work the politi cal institutions. In practice, however, their strained personal relations soon limited the expected cooperation at the center. Moreover, the reelection of Trimble as fi rst minister alongside Mallon’s successor, Mark Durkan, in 2001 was also problematic, given the reluctance of two UUP members to give their support.30 On another level, then, the switch from election to nomination was a fairly pragmatic change, allowing the creation of a new power-sharing executive in 2007 with arch-enemies Ian Pais- ley and Martin McGuinness at the helm.31 Th e experience of power sharing in Macedonia points to the possibility of parties representing the contending groups to negotiate ministerial seats (liberal, informal executive formation) rather than resort to formal corporate rules. On the whole, the conduct of these post- electoral negotiations helped secure politi cal stability. In the post-independence period, the decision by the SDSM, and then VMRO- DPMNE, to invite an Albanian party into executive formation discussions arguably helped facilitate the country’s transition to democracy. Yet the situation following the 2006 elections, Conclusion 217 when VMRO- DPMNE formed an executive with the second- largest Alba- nian party, prompted a politi cal crisis. Unsurprisingly, DUI protested against its exclusion and accused the governing co ali tion of lacking legitimacy. Following DUI’s boycott of parliament for several months in 2007, the crisis dissipated as a result of discussions between Gruevski and Ahmeti on the implementation of the OFA. Th ese discussions paved the way for more positive relations between the two parties, who negotiated a power-sharing executive following elections in 2008 and 2011. Notably, some Albanian politicians called for institutionalization of executive formation procedures under the concurrent majority voting procedure in parliament. Others, however, argued there is no need to establish such formal institutional rules because the parties and electorate expect the largest party from each of the two main communities to share power.32 Despite the problems associated with corporate rules, I recognize that policy makers may opt for such rules for reasons of po liti cal expediency. Parties may simply make guaranteed positions for their respective group (regardless of electoral strength) a precondition of signing up to the agreement. Allison McCulloch notes that the corporate version of consociation is more likely to be adopted in negotiated settlements and recommends that peace agreements strike a balance between liberal and corporate rules.33 Arguably, it would be preferable for mediators to persuade parties to adopt liberal rules; provided that the parties have suffi cient electoral support, they will access positions of power. Th e evidence also suggests that beyond a preference for liberal power sharing, there is a choice between adopting formal rules for executive formation (e.g., a sequential portfolio allocation method) or allowing inter-party negotiations to determine the distribution of ministerial seats. Negotiations would be appropriate if parties are amenable to compromise over who gets what in terms of the number of ministerial positions and the distribution of portfolios. Th e case of Macedonia shows that such negotiations are possible (albeit diffi cult). Yet policy makers need to acknowledge that such negotiations may well break down, as in Bosnia. I suggest there are instances when it might be benefi cial to opt for formal rules, including a sequential portfolio allocation method for executive formation. Such methods can be useful when parties accept that positions will not be predetermined, but they nevertheless have diffi culty negotiating with their potential coali tion partners. In this situation, a sequential portfolio allocation method (e.g., the use of d’Hondt in Northern Ireland) can relieve parties of the pressure of negotiation over the distribution of ministerial seats. It is 218 Conclusion hoped that following executive formation, cooperation among the parties will develop as they get on with the business of governing. Future work in this area might look at the possibility of adopting d’Hondt or similar rules (Sainte-Laguë) in other deeply divided places as parties negotiate a peace settlement. It would also be important to determine how diff erent rules (d’Hondt versus Sainte-Laguë) could impact on the proportionality of parties in the coali tion. What matters is that parties feel a sequential portfolio method allows for a fair distribution based on their respective electoral strength.

Executive Decision Making: The Impact of Veto Rules

Th is book has also sought to uncover the signifi cance of veto rules in the functioning of power sharing. Veto is one of the four elements in Lijphart’s original formulation of consociational democracy, and we need to determine its eff ects on the operation of power sharing and the pursuit of inter- ethnic cooperation. We can assume that if veto rights are misused, they may lead to stalemate. If they are used appropriately, they may help foster cooperation among governing parties. I argue that although veto rights are crucial for groups to feel a sense of security in the new politi cal arrangements (and arguably necessary for them to sign up to the deal), specifi c veto procedures can make a diff erence in fostering cooperation among ethnic elites. Veto rules have been fairly contentious in Bosnia. At the state level, members of the tripartite presidency can veto a decision that contravenes their respective group’s vital national interests (VNI). A majority of delegates representing each constituent people can veto legislation in the state parliament on the basis of their VNI. If a majority of Bosniak, Croat, or Serb delegates object to the veto, a joint commission is convened to resolve the issue. If unsuccessful within fi ve days, the commission refers the matter to the constitutional court for review. In addition to the VNI procedure, decisions in the state parliament require that majority support includes at least one- third of delegates from each entity. Th us, in addition to the VNI veto, the power-sharing institutions in Bosnia have an entity veto. Similarly, in the entities, two-thirds of a constituent people’s delegates can veto legislation in their assembly. Th e entity constitutions list areas of legislation subject to the VNI veto but also include a provision whereby two-thirds of a group’s delegates can trigger the veto on any issue.34 Conclusion 219

Th e record of power sharing in Bosnia suggests that politicians have resorted to the application of veto powers to block the state institutions. More- over, in addition to formal veto rules under the DPA, elites have exercised a more informal veto by boycotting the institutions, thereby delaying the passage of legislation or frustrating decision making in the Council of Minis- ters. Th e high representative and other external actors have been critical of the use of veto powers as a tactic to prevent decision making.35 I suggest that problems with veto rules in Bosnia correspond to veto issues and veto points. As acknowledged by international actors and domestic elites, the list of legislation subject to the VNI procedure is too broad. Moreover, a two-thirds majority of delegates within a group can trigger a veto on any decision on the basis that it threatens their group’s vital interests. Th ere is a need to defi ne and clarify the areas of legislation that may be subject to VNI. Th ere is also a need to restrict the scope of these issues by removing the provision that two- thirds of a group’s delegates can trigger the VNI veto on any issue. In terms of veto points, there are arguably too many veto procedures at diff erent levels in Bosnia: the VNI procedure in the presidency, the VNI procedure in the entity parliaments, and both VNI and entity vetoes in the state parliament. Th e procedure for entity voting in the House of Represen- tatives has been particularly contentious. Th e rules are further problematic because the chamber has not had a mechanism to overcome gridlock, such as the Joint Commission in the House of Peoples. Arguably, without such a procedure, entity delegates in the House of Representatives are incentivized to block legislation rather than compromise to meet the recommendations from a parliamentary mediation body. Despite considerable challenges in Northern Ireland, the use of veto powers has not been particularly contentious for power sharing. Key decisions are taken on a cross- community basis: either via parallel consent (a majority of members present and voting, including a majority of nationalists and unionists present and voting) or a weighted majority (60 percent of members present and voting, including at least 40 percent of each of the nationalist and unionist designations).36 New veto mechanisms were introduced in the St. Andrews Agreement, including provision for a statutory ministerial code. Th e code (para. 2.12) includes provision for any three executive ministers to trigger a cross- community vote in the executive when a vote is required on an issue.37 In addition, thirty delegates can refer important ministerial decisions back to the executive for review, on the basis that the minister concerned has contravened his or her obligations.38 For the DUP, amendments 220 Conclusion relating to ministerial accountability were crucial for the party’s agreement to enter an executive with Sinn Féin. Th e party has since stressed that the changes secured at St. Andrews mean that unionists will have an ultimate veto over the decisions of nationalist ministers.39 When a vote has to be taken on a cross-community basis (triggered by three executive ministers), if a party constitutes a majority (under the parallel consent procedure) or more than 40 percent (under the weighted majority procedure) of either the designated nationalists or designated unionists on the executive, that party can veto the decision and the matter will fall. Arguably, however, the revisions do not signifi cantly alter control over ministerial authority. On one hand, the St. Andrews Agreement provides for the assembly to refer important decisions back to the executive, and it also extends the topics for discussion and agreement within the executive to include signifi cant or controversial issues, as agreed by the fi rst minister and deputy fi rst minister. On the other hand, the provision for three ministers to trigger a cross- community vote within the executive when a vote is required limits the potential for one group to veto a minister from the other group. In Macedonia, the double majority voting procedure in parliament has been a central feature of the post- Ohrid power- sharing system. Under this rule, legislation aff ecting culture, use of language, education, personal ID, and symbols requires a majority of delegates present and voting, including a majority of delegates not from the majority community. Th e case supports the view that the veto is enormously important for the minority community. Albanian politicians emphasize the signifi cance of the procedure in parliament, almost always referred to as “the Badinter rule,” for protecting their group rights. Some Albanian politicians have proposed the extension of the procedure to cover parliamentary approval on government formation and on bud get approval. On the whole, the application of the procedure has not been particularly contentious—an institutional victory for the Albanian community and a compromise largely accepted by Macedonians. With regard to veto rules, I argue that less is more.40 First, veto players should preferably be self-determined groups, though that might not always be possible, as in the rights of the three groups in Bosnia. Second, when groups want to ensure the protection of their fundamental interests, veto issues should be clearly defi ned in legislation and agreed among the political parties. Such issues should be limited to identity issues, symbols, language, culture, education, the bud get, and security. Th ird, veto points should also be limited: A complex set of veto rules at various points in the Conclusion 221 politi cal process without suffi cient clarifi cation and limitation will arguably not foster cooperation.

The Role of External Actors in Power Sharing

In addition to exploring the impact of institutional rules, I have sought to assess the infl uence of external actors on elite cooperation. Th ough existing literature suggests that external actors have a role in the implementation of power-sharing systems, we arguably need to understand better the interaction between external and internal actors in the formation and operation of power sharing.41 If we have embarked on a “new wave” of power-sharing theory, owing to the engagement of external actors, we need to take greater account of the ways in which they behave in power-sharing systems to manage inter-ethnic relations. Th ere is a need, therefore, to clarify mechanisms. Are the interests of domestic elites shaped by external actors engaged in socialization eff orts? Or is it simply down to domestic actors’ instrumental choices? Th e three countries’ diff erent experiences in power sharing demonstrate that externals help implement a power- sharing agreement through an incentives structure based on positive and negative incentives (and even coercive mea sures). I argue that external incentives will more likely be eff ective when they uphold an agreement that satisfi es groups’ constitutional preferences relating to competing self- determination claims. When groups feel these claims have been addressed satisfactorily, external incentives may even lead to internalization and the “habitualization [of power sharing] as norm-conforming behaviour.” 42 Incentives will be less eff ective when the peace agreement remains contested and external actors are perceived as pushing a reform agenda against the preferences of one or more of the contenders. In terms of the mechanisms employed by external actors to bring about policy change at the domestic level, the socialization mechanism in play in post-confl ict institution building is likely to be strategic calculation. Ex- ternal actors rely on incentives to induce domestic elites to share power, adopting politi cal conditionality linked to politi cal rewards. If compliance outweighs the costs of norm-violating behavior, elites will adapt their behavior to policies promoted by externals. Moreover, strictly rational behavior can result in some internalization, with post-confl ict institutions becoming “locked into” the system. In addition, the research fi ndings add to our 222 Conclusion understanding of mechanisms by highlighting that external actors’ engagement may be incoherent, with variation over time and among externals as well as (oft en misjudged) coercive measures that impact the incentives structure. Although external actors may employ “socialization through reinforcement,” 43 this depiction needs to account for the “muddling along” aspect of external policy whereby incoherence may aggravate existing intergroup tensions. In the Northern Ireland case, the Sunningdale communiqué points to the role of external actors, particularly Secretary of State for Northern Ire- land William Whitelaw, in bringing about power sharing between rival communities.44 Following the prorogation of Stormont in 1972, the British government made clear to the parties that majority rule was a thing of the past in Northern Ireland and that the minority community would need to be part of new politi cal institutions. Faulkner’s unionists acknowledged that power sharing (plus an “Irish dimension”) would be part of new arrangements and that devolution to the Stormont assembly would not take place without agreement with nationalists on these issues. Since the 1990s, the British and Irish governments have been crucial in arriving at a power- sharing settlement (the GFA and later the St. Andrews Agreement), con- scious of the need to incentivize the nationalist and unionist parties (on the constitutional question and by off ering side payments). Both governments were engaged in maintaining power sharing and managing the divergent expectations and preferences among the parties. Th eir intergovernmental approach involved off ering concessions to each side and draft ing proposals for progress. Negative incentives were also employed, by threatening to introduce reform should the parties fail to agree and threatening greater intergovernmentalism toward joint authority of Northern Ireland as the default “Plan B.” Th e British government also adopted coercive incentives, the key example being Secretary of State for Northern Ireland Peter Mandelson’s suspension order of 2000. As discussed in Chapter 3, this policy arguably created considerable diffi culties for power sharing and strained the Blair government’s relations with nationalists and with the Irish government. Th e two governments’ incentives structure was again much in evidence in their eff orts to restore devolution in 2007, incentivizing the DUP and Sinn Féin to reach agreement on power sharing and policing. Th e relatively positive experience in Northern Ireland points to the overall eff ectiveness of the British and Irish governments’ eff orts in upholding the GFA by rewarding parties with side payments and securing a constitutional compromise. Ar- Conclusion 223 guably, this case suggests that eff ective external incentives can lead to internalization and habitualization of power sharing over time. Th ough not without its diffi culties, the system is supported by the main parties. As Wilford notes, even though the DUP might favor a diff erent form of power sharing to that created under the GFA, party leader Peter Robinson “is pragmatic enough to realise that it is not on the horizon and that his colours and those of his party are therefore nailed indefi nitely to the fully inclusive power-sharing mast.” 45 Th e chapters on Bosnia show that external actors have employed an incentives structure to encourage the local parties to reform the power-sharing structures and move beyond the DPA as a means to realize the “dream of Eu rope.” Although the external actors initially sought to incentivize the eff ective workings of the DPA (oft en using the Bonn powers to do so), they have increasingly sought to bring about reform, transferring powers from the entities to the state, and seeking to modify the rules governing power sharing in the state- level institutions. Th us, their eff orts appear to be weakening the federal and consociational structures under the DPA (which secures group rights) to a more integrationist agenda, in which majority- rule democracy is seen as the Eu ro pe an norm. Th e Bonn powers have been used to varying degrees by successive high representatives and in the (ongoing) shift from the OHR to the EU Special Representative as the lead international actor. Th e evidence shows that engagement by external actors has not been fully coherent, infl uenced by fl uctuating politi cal demands, the preferred approach of successive high representatives, and the fl uid arrangements in moving toward OHR closure. With the EU striving to close the OHR and maintain a presence without executive powers, a degree of confusion or incoherence has arguably developed between the two main external actors. Th e dynamics of this shift have led some domestic actors, particularly Bosnian Serb elites, to question the legitimacy of the international community in Bosnia. Th e Bosnian experience highlights how external incentives can be ineff ective in bringing about behavioral change at the state level. Chapters 5 and 6 show that despite external actors’ eff orts to incentivize elites to reform the DPA structures (with EU integration the main carrot), these incentives have been slow to make a diff erence. Power sharing, far from becoming “habitualized,” has been destabilized because Bosniaks have felt incentivized to challenge the power- sharing arrangements, and Bosnian Serbs have gone back to proposing secession, partially in response to pressure from 224 Conclusion externals. Th e EU carrot of membership has failed to generate much progress on constitutional reform. Indeed, all signifi cant institutional reform has come about as a result of external imposition rather than inter- party agreement. Th is book’s fi ndings corroborate the view that internal actors have sought to “challenge the normative grounds of the EU’s policy and have responded with fake compliance, partial compliance, and noncompliance, respectively, with the latter provoking imposed compliance.” 46 Th e diffi culty in incentivizing reform refl ects the ongoing contestation over state structures and the “frozen” self- determination dispute. Strained intergroup relations are brought into sharp relief by the international community’s diffi culty in achieving reform, with Bosnian Serbs opposed to any tinkering with their rights under Dayton and Bosniaks hoping for greater intervention to centralize state structures. External actors have also played an important role in post- Ohrid Mace- donia. First, the EU and the U.S. (along with NATO and the OSCE) were key players in bringing about a peace settlement that ended the 2001 confl ict. Politicians interviewed for this research give these external actors much credit for mediating inter-party talks and draft ing proposals for an agreement. Since the signing of the OFA, the EU has been the main external actor engaged in persuading the parties to implement the agreement and coordinating wider international eff orts. Arguably, the EU’s eff orts in Macedonia have been less visible on the ground than in Bosnia, where the EUSR has been a double- hatted position with the OHR. Th is less- visible pressure has, however, been important in encouraging po liti cal parties to progress policy reform toward full OFA implementation and EU integration. Moreover, the EU has incorporated its monitoring of OFA implementation into Macedonia’s accession procedures. Chapter 8 shows that the engagement of the EU was fairly straightforward, from the signing of the SAA in 2001 and the granting of candidate status in 2005 until the Eu ro pe an Commission’s decision in October 2009 to recommend the start of accession negotiations. Since then, however, progress has stalled over the country’s name dispute with Greece. Although EU institutions call for a resolution of the dispute, much depends on the preparedness of the local parties, particularly the largest Macedonian party, VMRO-DPMNE, to do a deal with Greece. Th is book also highlights that in facilitating a power- sharing agreement, external actors may not just convene inter- party discussions; they will be heavily engaged in the process by drawing up proposals for a settlement. Th ese proposals are unlikely to be plucked out of thin air (refl ecting exter- Conclusion 225 nals’ awareness of internal parties’ preferences and prior engagement with politicians), but they may, nevertheless, contain an element of risk in terms of whether the parties will sign up to the content. If external actors are involved in facilitating an agreement in the fi rst place, it is likely the system will require their ongoing eff orts to maintain it. We can compare the disengagement of the British government aft er the Sunningdale communiqué with the commitment by the Blair and Brown administrations following the GFA and St. Andrews Agreement, respectively. In the Sunningdale case, the British government was arguably found wanting, particularly in light of the considerable pressures on the executive from oppositional forces, as well as intra-executive tensions. Th ere is, however, a delicate balancing act to be found between acting as a “guarantor” of a power- sharing settlement and heavy intervention that may constrain the parties’ inde pen dence. Bosnia is the showcase of heavy- handed engagement that allowed the parties to escape the pressures of compromise. Th e Bosnian experience of power sharing remains a key example of the diffi culties in promoting local ownership and responsible decision making when external actors enjoy considerable authority in the operation of power sharing. Th ese fi ndings on external actors have implications for power-sharing practice, future research on post- confl ict institutions, and power- sharing theory. First, the evidence highlights how external actors will likely face challenges relating to their powers of imposition, their lack of legitimacy and accountability, and their sometimes questionable commitment to sus- taining the power- sharing agreement. When parties continue to mistrust one another and the peace is a delicate one, external actors engaged in the process will need to consider how to manage their ongoing responsibilities. As O’Leary writes, “Arbitrators who become unaccountable and usurp sovereignty or authority deny the meaningful self- government that must un- derpin successful power sharing.” 47 But how can they seek to temper these diffi culties? Can external actors become more legitimate and accountable in their dealings with the local parties? Th ese are ongoing, pressing problems. Second, the fi ndings point to the need for future research on external actors’ role and responsibilities in power sharing. Christine Bell notes the multiple roles of external actors that “straddle any distinction between enforcement of the bargain, mediated development of the bargain, and administration of sub- elements of the bargain.” 48 In terms of external actors’ variable approaches to engagement, O’Leary presents a classifi cation from balanced actors, to partisans, to imposers, to destabilizers. He suggests that the “ideal 226 Conclusion role for outsiders is as voluntarily agreed balancers who play the welcome role of positive support for a power-sharing system, mediate diff erences that arise upon request, and arbitrate only when asked to do so.” 49 We need to know more about how these varying roles and approaches are executed by external actors working on the ground. Are international organizations cognizant of tensions arising from the shift s in their role (from mediator to arbiter to administrator) over time? And when external actors do make an eff ort to consult with local elites and civil society, what factors promote successful engagement? Much more remains to be investigated in terms of external actors’ governance role in deeply divided places, including power sharing. And given the multiplicity of international organizations involved in post- confl ict peacebuilding, future research will need to determine whether their interaction (cooperation or competition) matters for stable peace and democracy. NOTES

Chapter 1

1. “Union Flag Protests: Peter Robinson Says Politicians Not ‘Giving Up.’ ” 2013. BBC News Online, January 13. http:// www .bbc .co .uk /news /uk -northern -ireland -21003296 . 2. McAdam, Noel. 2013. “Martin McGuinness: I’ll Meet Flag Protesters but I Won’t Kowtow.” Belfast Telegraph, January 30. http:// www .belfasttelegraph .co .uk /. 3. Bosnia and Herzegovina comprises two entities: the Bosnian-Croat Federation and Republika Srpska (RS), the Serb Republic. 4. Th e Constitution of Bosnia and Herzegovina (hereaft er Bosnia) confers equal rights on the three constituent peoples: Bosniaks (Muslims), Bosnian Serbs, and Bos- nian Croats. 5. Ross, Marc Howard. 2009. “Cultural Contestation and the Symbolic Landscape: Politics by Other Means?” In Culture and Belonging in Divided Societies, edited by Marc Howard Ross. Philadelphia: University of Pennsylvania Press, 5. Th e book uses the term “deeply divided places” following its usage in McEvoy, Joanne, and Brendan O’Leary, eds. 2013. Power Sharing in Deeply Divided Places. Philadelphia: University of Pennsylvania Press. 6. Institutional rules are treated here as formal procedures set out in constitutions/ legislation that both shape actors’ preferences and constrain behavior. 7. Wolff , Stefan. 2010. “Building Democratic States aft er Confl ict: Institutional De- sign Revisited.” International Studies Review 12: 128– 41, 133. 8. Ibid., 134. 9. Ostrom, Elinor. 2005. Understanding Institutional Diversity. Prince ton, N.J.: Princeton University Press, 3. 10. Ibid. 11. Peters, B. Guy. 1999. Institutional Th eory in Politi cal Science: Th e “New Institu- tionalism.” London: Pinter, 150. 12. O’Leary, Brendan. 2005. “Debating Consociational Politics: Normative and Explanatory Arguments.” In From Power Sharing to Democracy: Post-Confl ict Institu- tions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s University Press, 33– 36. 228 Notes to Pages 5–12

13. Th e British and Irish governments are treated as external actors, given their role in bringing about the peace agreement in 1998 and the subsequent incentives employed in their eff orts to secure peace implementation and cooperation among the internal parties. Th is incentives structure is compared with that of the external actors in the other two cases under investigation. 14. “Iraqi Talks Possibly in Ireland.” 2008. BBC News Online, July 7. http:// www .bbc.co .uk . 15. Th e Albanian community in Macedonia constitutes around 25 percent of the population. Some Albanian leaders suggest the numbers may be higher. 16. Norris, Pippa. 2008. Driving Democracy: Do Power- Sharing Institutions Work? Cambridge: Cambridge University Press. Norris focuses on four formal institutional features: the type of electoral system, the horizontal concentration of power in the type of executive, the vertical centralization of power in unitary or federal states, and the structure and in de pen dence of the mass media. 17. Ibid., 37– 38. 18. Mattes, Michaela, and Burcu Savan. 2009. “Fostering Peace Aft er Civil War: Commitment Problems and Agreement Design.” International Studies Quarterly 53 (3): 737– 59. 19. For discussion on corporate and liberal consociation, see McCulloch, Allison. 2012. “Consociational Settlements in Deeply Divided Societies: Th e Liberal-Corporate Distinction.” Demo cratization iFirst, 1– 18, doi:10.1080/13510347.2012.748039. 20. See Risse, Th omas, Stephen C. Ropp, and Kathryn Sikkink eds. 1999. Th e Power of Human Rights: International Norms and Domestic Change. Cambridge: Cambridge University Press. 21. Bardon, Jonathan. 2001. A History of Ulster. Belfast: Blackstaff Press, 236. 22. Ibid., 350– 52. 23. Ibid., 494. 24. See Malcolm, Noel. 1996. Bosnia: A Short History. New York: New York Uni- versity Press, 132; Hoare, Marko Attila. 2007. Th e History of Bosnia: From the Middle Ages to the Present Day. London: Sagi Books, 61– 62. 25. Glenny, Misha. 1999. Th e Balkans 1804– 1999: Nationalism, War and the Great Powers London: Granta Books, 160– 62; Hoare, Th e History of Bosnia, 66– 70. 26. Friedman, Francine. 2004. Bosnia and Herzegovina: A Polity on the Brink. Lon- don and New York: Routledge, 13. 27. Ibid., 21; Redžić, Enver. 2005. Bosnia and Herzegovina in the Second World War. London and New York: Frank Cass. 28. Malcolm, Bosnia: A Short History, 174– 92. 29. Kaplan, Robert D. 2005. Balkan Ghosts: A Journey through History. New York: Picador, 56. 30. Glenny, Misha. 1999. Th e Balkans 1804–1999: Nationalism, War and the Great Powers. London: Granta Books, 228– 48. Notes to Pages 13–15 229

31. Poulton, Hugh. 2000. Who Are the Macedonians? Bloomington: Indiana Uni- versity Press, 100– 108. 32. Norris, Driving Democracy, 20. 33. March, James G., and Johan P. Olsen. 1984. “Th e New Institutionalism: Or- gani za tion al Factors in Politi cal Life.” American Po liti cal Science Review 78 (3): 734– 49, 735. A number of strands exist in the fabric of institutionalism, including rational choice, historical approaches, and normative approaches. See Peters, Institutional Th eory for a discussion of seven diff erent approaches to institutionalism: normative, historical, rational choice, empirical, socio log i cal, interest represen ta tion, and international. Although these distinct approaches share in the position that institutions are the central component of po liti cal life, they disagree on how agents and institutions interact. 34. Pierson, Paul. 2004. Politics in Time: History, Institutions and Social Analysis. Prince ton and Oxford: Prince ton University Press, 103. 35. D’Hondt is a sequential portfolio allocation rule and provides an algorithmic method for determining proportional allocation of co ali tion cabinet seats. See O’Leary, Brendan, Bernard Grofman, and Jorgen Elklit. 2005. “Divisor Methods for Sequential Portfolio Allocation in Multi- Party Executive Bodies: Evidence from Northern Ireland and Denmark.” American Journal of Po liti cal Science 49 (1): 198– 211; McEvoy, Joanne. 2006. “Th e Institutional Design of Executive Formation in Northern Ireland.” Regional and Federal Studies 16 (4): 447– 64. 36. Olsen, Johan P. 1997. “Institutional Design in Democratic Contexts.” Journal of Po liti cal Philosophy 5 (3): 203– 29, 205. 37. Ibid. 38. Horowitz, Donald L. 2000. “Constitutional Design: An Oxymoron?” In Design- ing Democratic Institutions, edited by Ian Shapiro and Stephen Macedo. New York: New York University Press, 16. 39. Ibid. 40. Goodin, Robert E. 1996. “Institutions and Th eir Design.” In Th e Th eory of Institutional Design, edited by Robert E. Goodin. Cambridge: Cambridge University Press, 28. 41. Horowitz, “Constitutional Design,” 270. 42. Van Parijs, Philippe. 2000. “Power- Sharing Versus Border- Crossing in Ethni- cally Divided Societies.” In Designing Demo cratic Institutions, edited by Ian Shapiro and Stephen Macedo. New York and London: New York University Press, 314. For discussion on centripetalism, see Reilly, Benjamin. 2001. Democracy in Divided Soci- eties: Electoral Engineering for Confl ict Management. Cambridge: Cambridge Univer- sity Press. Reilly defi nes centripetalism as a normative theory of institutional design that seeks to encourage (1) electoral incentives for politicians to win votes from groups other than their own; (2) arenas of bargaining whereby elites are incentivized to negotiate cross-ethnic vote-pooling deals; and (3) centrist parties or coali tions that seek 230 Notes to Pages 16–19 multi-ethnic support on the basis of negotiated policy commitments. Th e objective of this approach is to build a moderate center based on compromise, and it is associated with Horowitz’s work (see Horowitz, Donald L. 1985. Ethnic Groups in Confl ict. Berke- ley: University of California Press). 43. Goodin, “Institutions and Th eir Design,” 25– 26. 44. Ibid. 45. Ibid., 28. 46. Elster, Jon. 1983. Explaining Technical Change: A Case Study in the Philosophy of Science. Cambridge: Cambridge University Press, 61. 47. Ibid., 70. 48. Pierson, Paul. 2004. Politics in Time: History, Institutions and Social Analysis. Prince ton and Oxford: Prince ton University Press, 108. 49. Ibid., 103– 32. 50. Ibid., 115. 51. See Cortell, Andrew P., and Susan Peterson. 2001. “Limiting the Unintended Consequences of Institutional Change.” Comparative Politi cal Studies 34 (7): 768– 99. Th e authors show how the broader environment can aff ect the potential for actors to achieve their objectives in institutional reform. 52. Th e power- sharing executive formed on 29 November 1999 was suspended for the fourth time on 14 October 2002 amid allegations of a republican spy ring at Parliament Buildings, Stormont. Suspension led to a period of stalemate, with the British and Irish governments attempting to break the impasse on a number of occasions. At the end of the round of inter- party talks held at St. Andrews in Scotland in October 2006, the two governments produced the St. Andrews Agreement, which became the basis of agreement between the parties and restoration of power sharing in May 2007. 53. Offi ce of the High Representative. 2002. Decision Enacting the Law on the Council of Ministers of Bosnia and Herzegovina. h t t p : / / w w w . o h r . i n t . 54. Th e proceedings were initiated by Alija Izetbegovic, then (Bosniak) chairman of the state presidency, to determine whether the RS constitution was consistent with the state constitution. 55. Bieber, Florian. 2006. “Aft er Dayton, Dayton? Th e Evolution of an Unpopu lar Peace.” Ethnopolitics 5 (1): 15– 31, 27. 56. In Macedonia, the procedure is known as the “Badinter principle,” aft er French constitutional lawyer Robert Badinter, who was involved in draft ing the Ohrid Agreement. 57. Horowitz, Donald L. 1991. A Demo cratic South Africa? Constitutional Engi- neering in a Divided Society. Berkeley: University of California Press, 257. 58. See Roeder, Philip G. 2005. “Power Dividing as an Alternative to Ethnic Power Sharing.” In Sustainable Peace: Power and Democracy Aft er Civil Wars, edited by Philip G. Roeder and Donald Rothchild. Ithaca, N.Y.: Cornell University Press; Rothchild, Donald and Philip G. Roeder. 2005. “Power Sharing as an Impediment to Notes to Pages 19–21 231

Peace and Democracy.” In Sustainable Peace: Power and Democracy Aft er Civil Wars, edited by Philip G. Roeder and Donald Rothchild. Ithaca, N.Y.: Cornell University Press. 59. See O’Leary, Brendan. 2005. “Debating Consociational Politics: Normative and Explanatory Arguments.” In From Power Sharing to Democracy: Post- Confl ict Institu- tions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s University Press; McGarry, John and Brendan O’Leary. 2009. “Under Friendly and Less Friendly Fire.” In Consociational Th eory: McGarry and O’Leary and the Northern Ireland Confl ict, edited by Rupert Taylor. London: Routledge. 60. Choudhry, Sujit. 2008. “Bridging Comparative Politics and Comparative Con- stitutional Law: Constitutional Design in Divided Societies.” In Constitutional Design for Divided Societies: Integration or Accommodation? edited by Sujit Choudhry. Ox- ford: Oxford University Press. 61. McGarry, John, Brendan O’Leary, and Richard Simeon. 2008. “Integration or Accommodation? Th e Enduring Debate in Confl ict Regulation.” In Constitutional Design for Divided Societies: Integration or Accommodation? edited by Sujit Choudhry. Oxford: Oxford University Press, 52– 53. 62. Choudhry, “Bridging Comparative Politics,” 27. 63. See McGarry, O’Leary, and Simeon, “Integration or Accommodation?” 64. Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Explora- tion. New Haven, Conn.: Yale University Press, 25. 65. See O’Leary, “Debating Consociational Politics,” 12– 13. In a complete consociation, representatives of all main segments are included in government; in a concurrent consociation, each signifi cant segment has represen ta tion with at least majority support respectively; and in a weak consociation, at least one segment may have only plurality support. Th is distinction is expanded in McGarry, John, and Brendan O’Leary. 2006. “Consociational Th eory, Northern Ireland’s Confl ict, and Its Agreement. Part 1: What Consociationalists Can Learn from Northern Ireland.” Government and Opposition 41 (1): 43– 63; and O’Leary, Brendan. 2013. “Power Sharing in Deeply Divided Places: An Advocate’s Introduction.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: Univer- sity of Pennsylvania Press. 66. McGarry, John, and Brendan O’Leary. 2009. “Under Friendly and Less- Friendly Fire.” In Consociational Th eory: McGarry & O’Leary and the Northern Ireland Confl ict, edited by Rupert Taylor. London: Routledge, 350. 67. See McGarry, John, and Brendan O’Leary. 2007. “Iraq’s Constitution of 2005: Liberal Consociation as a Politi cal Prescription.” International Journal of Constitutional Law 5 (4): 670– 98. 68. Horowitz, Donald L. 2002. “Constitutional Design: Proposals Versus Pro- cesses.” In Th e Architecture of Democracy: Constitutional Design, Confl ict Manage- ment and Democracy, edited by Andrew Reynolds. Oxford: Oxford University Press, 23. 232 Notes to Pages 21–25

69. See Fraenkel, Jon, and Bernard Grofman. 2006. “Does the Alternative Vote Foster Moderation in Ethnically Divided Societies?” Comparative Politi cal Studies 39 (5): 623–51; McCulloch, Allison. 2013. “Th e Track Record of Centripetalism in Deeply Divided Places.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 70. McGarry, John, and Brendan O’Leary. 2004. Th e Northern Ireland Confl ict: Consociational Engagements. Oxford: Oxford University Press, 25– 26. 71. O’Flynn, Ian. 2007. “Divided Societies and Deliberative Democracy.” British Journal of Po liti cal Science, 37 (4): 731–51, 743. 72. Ibid., 750– 51. 73. In forging the “Alliance for Change,” the international community sought to marginalize the three nationalist parties that had been instrumental in the Bosnian war and in power since Dayton. 74. Th ough the Bosnian Croats voted for a presidency member from the non- nationalist Social Democratic Party (SDP), the electorate continues to vote for nationalist parties and “moderate” candidates fared badly. Th e former ruling parties (SDA, HDZ, and SDS) were defeated by newer parties such as the SNSD (Bosnian Serb) and SBiH (Bosniak). 75. O’Leary, “Debating Consociational Politics,” 19. 76. Horowitz, “Constitutional Design: An Oxymoron?”: 259. 77. Taylor, Rupert. 2001. “Consociation or Social Transformation?” In Northern Ire- land and the Divided World, edited by John McGarry. London: Oxford University Press. 78. See International Crisis Group. 2011. Bosnia: Europe’s Time to Act. Eu rope Briefi ng No. 59, January 11, http:// www .crisisgroup .org; International Crisis Group. 2011. Bosnia: State Institutions under Attack. Europe Briefi ng No. 62, 6 May 2011, h t t p : / / w w w . c r i s i s g r o u p . o r g . 79. See “Bosnian Elections: Stephen Sackur Talks to Haris Silajdzic, Co-President of Bosnia.” 2006. BBC News, October 13. http:// news .bbc .co .uk /2 /hi /programmes /hardtalk /6049454 .stm . 80. Venice Commission. 2006. Draft Opinion on Diff erent Proposals for the Elec- tion of the Presidency of Bosnia and Herzegovina. h t t p : / / w w w . v e n i c e . c o e . i n t . 81. Venice Commission. 2006. Preliminary Opinion on the Draft Amendments to the Constitution of Bosnia and Herzegovina. h t t p : / / w w w . v e n i c e . c o e . i n t . 82. Bieber, Florian. 2005. “Partial Implementation, Partial Success: Th e Case of Macedonia.” In Power-Sharing: New Challenges for Divided Societies, edited by Ian O’Flynn and David Russell. London: Pluto Press, 109. 83. Lijphart, Arend. 2002. “Th e Wave of Power- Sharing Democracy.” In Th e Archi- tecture of Democracy: Constitutional Design, Confl ict Management, and Democracy, edited by Andrew Reynolds. Oxford: Oxford University Press, 46. 84. Andeweg, Rudy B. 2000. “Consociational Democracy.” Annual Review of Po liti cal Science 3: 509–36, 519–20. See also Halpern, Susan. 1986. “Th e Disorderly Notes to Pages 25–29 233

Universe of Consociational Democracy.” W e s t E ur op ea n P o l i t i c s 9: 181– 97; Lustick, Ian. 1997. “Lijphart, Lakatos, and Consociationalism.” World Politics 50: 88– 117. 85. McGarry and O’Leary, “Consociational Th eory,” 43– 63, 62– 63. 86. O’Leary, “An Advocate”s Introduction,” 11. 87. See Ross, Marc Howard. 2007. Cultural Contestation in Ethnic Confl ict. Cam- bridge: Cambridge University Press; Ross, Marc Howard, ed. 2009. Culture and Be- longing in Divided Societies: Contestation and Symbolic Landscapes. Philadelphia: University of Pennsylvania Press. 88. Lijphart, Arend. 1995. “Self-Determination versus Pre-Determination of Eth- nic Minorities in Power-Sharing Systems.” In Th e Rights of Minority Cultures, edited by Will Kymlicka. Oxford: Oxford University Press, 275– 76. 89. Ibid., 285. 90. McGarry and O’Leary, “Iraq’s Constitution of 2005,” 670– 98, 675. 91. McGarry, O’Leary, and Simeon, “Integration or Accommodation?” 62– 63. 92. Th e Northern Ireland Act 1998 was amended by the Northern Ireland (St. Andrews Agreement) Act 2006, substituting new Sections 16A, 16B, and 16C for the original Section 16. Th e pro cess set out in §16A provides for the nomination of the fi rst minister and deputy fi rst minister by the largest party of the largest politi cal designation and the largest party of the second-largest designation, respectively. Section 16C(6) provides that if the largest politi cal party of the largest politi cal designation is not the largest politi cal party, the post of fi rst minister shall be nominated by the largest po liti cal party. Th e deputy fi rst minister is then nominated by the largest party of the largest po liti cal designation. See Northern Ireland Act 1998 (revised) at http:// www .legislation .gov .uk /ukpga /1998 /47 /contents . 93. Wilford, Rick. 2010. “Northern Ireland: Th e Politics of Constraint.” Parlia- mentary Aff airs 63 (1): 134– 55. 94. Th e 108 Members of the Legislative Assembly (MLAs) must designate as “nationalist,” “unionist,” or “other.” “Parallel consent” requires a majority of those present and voting, including a majority of designated nationalists and a majority of designated unionists. A “weighted majority” requires 60 percent of those present and voting, including 40 percent of both nationalists and unionists. 95. No more than two- thirds of ministers are to be from the Federation, and deputy ministers are not to be of the same constituent people as their respective minister. 96. See Constitutional Court of Bosnia and Herzegovina. 2000. Request for Evaluation of Certain Provisions of the Constitution of Republika Srpska and the Constitution of the Federation of Bosnia and Herzegovina, Case No. U5/98–111, July 1, 2000. 97. McGarry and O’Leary, “Under Friendly and Less- Friendly Fire.” 98. Agreement Reached in the Multi- Party Negotiations. London: HMSO, 1998, Strand One, para. 19 234 Notes to Pages 29–34

99. Constitution of the Federation of Bosnia and Herzegovina, available at http:// www.ohr .int . 100. Constitution of Republika Srpska, Article 90, available at http:// www .ohr .int . 101. Constitution of Th e Republic of Macedonia, Article 91, available at http:// www.sobranie .mk /en /default -en .asp . 102. McEvoy, Joanne. 2013. “We Forbid! Th e Mutual Veto and Power- Sharing Democracy.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 103. Bieber, Florian. 2005. “Power- Sharing Aft er Yugo slavia: Functionality and Dysfunctionality of Power- Sharing Institutions in Post- War Bosnia, Macedonia, and Kosovo.” In From Power- sharing to Democracy: Post- Confl ict Institutions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill- Queen’s University Press, 95. 104. Reynolds, Andrew. 2000. “Majoritarian or Power-Sharing Government.” In Democracy and Institutions: Th e Life Work of Arend Lijphart, edited by Markus M. L. Crepez, Th omas A. Koeble, and David Wilsford. Ann Arbor: University of Michigan Press, 168. 105. Taylor, Rupert. 2009. “Introduction: Th e Promise of Consociational Th eory.” In Consociational Th eory: McGarry and O’Leary and the Northern Ireland Confl ict, edited by Rupert Taylor. New York: Routledge, 7. 106. Kerr, Michael. 2005. Imposing Power-Sharing: Confl ict and Coexistence in Northern Ireland and Lebanon. Dublin: Irish Academic Press, 7. 107. Bieber, Florian. 2013. “Balkans: Promotion of Power-Sharing by Outsiders.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 108. McGarry, O’Leary, and Simeon, “Integration or Accommodation?” 87. 109. Checkel, Jeff rey T. 2001. “Why Comply? Social Learning and Euro pe an Iden- tity Change.” I n t e r n a t i o n a l O rg an iz at i o n 55 (3): 553– 88, 581. 110. Checkel, Jeff rey T. 2005. “International Institutions and Socialization in Eu- rope: Introduction and Framework.” I n t e r n a t i o n a l O rg an iz at i o n . 59: 801– 26, 804. 111. Ibid. 112. Ibid., 809. 113. Schimmelfennig, Frank. 2005. “Strategic Calculation and International So- cialization.” I n t e r n a t i o n a l O rg an iz at i o n , 59: 827– 60, 830. 114. Kerr, Imposing Power- Sharing, 180– 81. 115. Chandler, David. 2002. “Bosnia’s New Colonial Governor?” Guardian, July 9. 116. Offi ce of the High Representative. 2012. “Th e Mandate of the OHR.” Available a t h t t p : / / w w w . o h r . i n t / o h r - i n f o / g e n - i n f o / d e f a u l t . a s p ? c o n t e n t _ i d =38612. 117. Euro pe an Stability Initiative. 2001. In Search of Politics: Th e Evolving Inter- national Role in Bosnia and Herzegovina. h t t p : / / w w w . e s i w e b . o r g / p d f / e s i _ d o c u m e n t _id _22 .pdf, 4. 118. Hislope, Robert. 2003. “Between a Bad Peace and a Good War: Insights and Les- sons from the Almost- War in Macedonia.” Ethnic and Racial Studies 26 (1): 129– 51, 130. Notes to Pages 35–44 235

119. Watt, Nicholas and Owen Bowcott. 2007. “We had no fi le on him but it was clear he was up for the business.” Th e Guardian, 14 March.

Chapter 2

1. See Gillespie, Gordon. 1998. “Th e Sunningdale Agreement: Lost Opportunity or an Agreement Too Far?” Irish Politi cal Studies 13 (1): 100–114; Wolff , Stefan. 2001. “Context and Content: Sunningdale and Belfast Compared.” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press; McGrattan, Cil- lian. 2009. “Dublin, the SDLP and the Sunningdale Agreement: Maximalist National- ism and Path Depen den cy.” Contemporary British History 23 (1): 61–78; Kerr, Michael. 2011. Th e Destructors: Th e Story of Northern Ireland’s Lost Peace Process. Dublin: Irish Academic Press. 2. Kerr, Th e Destructors. 3. For background reading on the confl ict, see Bew, Paul, Peter Gibbon, and Henry Patterson. 2002. Northern Ireland 1921– 2001: Po liti cal Forces and Social Classes. London: Serif; Dixon, Paul. 2001. Northern Ireland: Th e Politics of War and Peace. New York: Palgrave; McGarry, John, and Brendan O’Leary. 1995. Explaining Northern Ireland. Oxford: Blackwell; O’Leary, Brendan, and John McGarry. 1996. Th e Politics of Antago- nism: Understanding Northern Ireland (2nd ed.). London: Athlone Press; Hennessey, Th omas. 1997. A History of Northern Ireland, 1920– 1996. Dublin: Gill and Macmillan. 4. Kerr, Th e Destructors. 5. Bloomfi eld, Kenneth. 2007. A Tragedy of Errors: Th e Government and Misgov- ernment of Northern Ireland. Liverpool: Liverpool University Press, 30. 6. Ibid., 30– 31. 7. Future Policy Group. 1972. Minutes of Meeting, 17 July 1972. Public Records Offi ce of Northern Ireland, Reference CAB/J90/6. 8. Northern Ireland Offi ce. 1972. Th e Future of Northern Ireland: A Paper for Discussion. London: HMSO. http:// cain .ulst .ac .uk /hmso /nio1972 .htm . 9. Ibid., para. 57. 10. Ibid., para. 58. 11. Northern Ireland Offi ce. 1973. Northern Ireland Constitutional Proposals (Cmnd. 5259). London: HMSO. http:// cain .ulst .ac .uk /hmso /cmd5259 .htm, Part 1, Article 5a. 12. Ibid., Part 1(2). 13. Northern Ireland Offi ce. 1973. Northern Ireland Constitution Act. London: HMSO, Part 1, 2(1). 14. Kerr, Th e Destructors, 96– 97. 15. Cited in ibid., 98. 16. Ibid., 113. 17. See the text of the Sunningdale Agreement (December 1973) at http:// cain .ulst .ac.uk /events /sunningdale /agreement .htm . 236 Notes to Pages 44–53

18. Faulkner, Brian. 1978. Memoirs of a Statesman. London: Weidenfeld and Nicolson, 239. 19. Ibid. 20. Lowry, Robert. 1973. “Telegram to Frank Cooper, 20 December 1973.” Public Rec ords Offi ce of Northern Ireland, Reference CJ4 /487. 21. Faulkner, Memoirs of a Statesman, 239. 22. Ibid. 23. Gillespie, “Th e Sunningdale Agreement,” 108. 24. Ibid., 109. 25. Cooper, Frank. 1974. “Note for the Record of a Meeting with Brian Faulkner and Ken Bloomfi eld, 4 March 1974.” Public Records Offi ce of Northern Ireland, Refer- ence FCO 87/334. 26. Faulkner, Memoirs of a Statesman, 252. 27. Northern Ireland Executive. 1974. Report of Executive Meeting, 15 January 1974, Belfast. PRONI Reference OE/2/2. 28. Ibid. 29. Ibid. 30. Silkin, A. 1989. “Th e ‘Agreement to Diff er’ of 1975.” In Ministerial Responsibil- ity, edited by G. Marshall. Oxford: Oxford University Press, 57. 31. Faulkner, Memoirs of a Statesman, 240. 32. Ibid., 242. 33. Northern Ireland Executive. 1974. Report of Executive Meeting, 22 May 1974, Belfast. PRONI Reference OE/2/26. 34. See McEvoy, Joanne. 2013. “We Forbid! Th e Mutual Veto and Power- Sharing Democracy.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 35. Kerr, Th e Destructors, 153. 36. Ibid. 37. Ibid., 51. 38. Ibid., 59. 39. Ibid. 40. Merriam- Webster dictionary online: http:// www .merriam -webster .com /. 41. Kerr, Th e Destructors, 99. 42. See O’Duff y, Brendan. 2007. British-Irish Relations and Northern Ireland: From Violent Politics to Confl ict Regulation. Dublin: Irish Academic Press. 43. See also O’Leary, Brendan. 2013. “Power Sharing: An Advocate’s Conclusion.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 44. Cited in Kerr, Th e Destructors, 238. 45. Ibid., 227. 46. Bell, Christine. 2008. On the Law of Peace: Peace Agreements and the Lex Pa- cifi catoria. Oxford: Oxford University Press, 179. Notes to Pages 54–61 237

47. Kerr, Th e Destructors. 48. Cited in ibid., 88. 49. Ibid., 92– 93. 50. Ibid., 132– 34. 51. Ibid., 128. 52. See Walter, Barbara F. 1997. “Th e Critical Barrier to Civil War Settlement.” Inter- n a t i o n a l O rg an iz at i o n 51 (3): 335–64; Walter, Barbara F. 2002. Committing to Peace: Th e Successful Settlement of Civil Wars. Princeton, N.J.: Princeton University Press; Walter, Barbara F. 2009. “Bargaining Failures and Civil War.” Annual Review of Po liti cal Science 12: 243– 61; Hoddie, Matthew and Caroline Hartzell. 2003. “Civil War Settlements and the Implementation of Military Power Sharing.” Journal of Peace Research 40 (3): 303– 20; Mattes, Michaela and Burcu Savun. 2009. “Fostering Peace Aft er Civil War: Commit- ment Problems and Agreement Design.” International Studies Quarterly 53 (3): 737– 59. 53. Hoddie and Hartzell, “Civil War Settlements,” 306. 54. Walter, “Th e Critical Barrier.” 55. Stedman, Stephen John. 1997. “Spoiler Problems in Peace Processes.” Interna- tional Security 22 (2): 5– 53. 56. Wolff , “Context and Content,” 18. 57. Th e Mitchell principles of democracy and nonviolence were established in a report by the International Body on Arms Decommissioning set up by the British and Irish governments in 1995. Th e body was chaired by former U.S. Senator George Mitchell along with former Finnish Prime Minister Harri Holkeri and Canadian Chief of Defence Staff General John de Chastelain. See International Body on Arms Decommissioning. 1996. Report of the International Body on Arms Decommissioning. h t t p : / / w w w . n i o . g o v . u k / i i c d _ r e p o r t _ 2 2 j a n 9 6 . p d f . 58. See Nilsson, Desirée and Mimmi Söderberg Kovacs. 2011. “Revisiting an Elu- sive Concept: A Review of the Debate on Spoilers in Peace Processes.” International Studies Review 16: 606– 26. 59. Ibid. 60. Tierney, Stephen. 2009. “Constitutional Referendums: A Th eoretical Enquiry.” Modern Law Review 72 (3): 360– 83, 379. 61. O’Flynn, Ian. 2006. Deliberate Democracy and Divided Societies. Edinburgh: Edinburgh University Press, 67. 62. Tierney, “Constitutional Referendums,” 381. 63. See Bell, On the Law of Peace, Chapter 9. 64. Ibid., 177.

Chapter 3

1. Ruane, Joseph, and Jennifer Todd. 1999. “Preface.” In Aft er the Good Friday Agreement: Analysing Politi cal Change in Northern Ireland, edited by Joseph Ruane and Jennifer Todd. Dublin: University College Dublin Press, vii. 238 Notes to Pages 62–65

2. O’Leary, Brendan, and John McGarry. 1993. Th e Politics of Antagonism: Under- standing Northern Ireland. London: Athlone Press, 202. 3. O’Duff y, Brendan. 2007. British- Irish Relations and Northern Ireland: From Vi- olent Politics to Confl ict Regulation. Dublin: Irish Academic Press, 167. 4. Ibid. 5. Ibid., 175. 6. Th e Agreement: Agreement reached in the multi-party negotiations, 1998. An- nex: Agreement between the Government of the United Kingdom of Great Britain and Northern Ireland and the Government of Ireland. 7. Wolff , Stefan. 2001. “Context and Content: Sunningdale and Belfast Compared.” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press, 13. 8. Ibid. 9. Horowitz, Donald L. 2002. “Explaining the Northern Ireland Agreement: Th e Sources of an Unlikely Constitutional Consensus.” British Journal of Po liti cal Science 32 (2): 193– 220, 194. 10. Ibid., 215. 11. See English, Richard. 2003. Armed Struggle: Th e History of the IRA. London: Pan Macmillan; Moloney, Ed. 2007. A Secret History of the IRA. London: Penguin. 12. Dixon, Paul. 2005. “Why the Good Friday Agreement in Northern Ireland Is Not Consociational.” Po liti cal Quarterly 76 (3): 357– 67, 362. 13. Coakley, John. 2009. “Implementing Consociation in Northern Ireland.” In Consociational Th eory: McGarry and O’Leary and the Northern Ireland Confl ict, edited by Rupert Taylor. London: Routledge, 124. 14. Ibid., 143. 15. Horowitz, Donald L. 2002. “Constitutional Design: Proposals Versus Pro- cesses.” In Th e Architecture of Democracy: Constitutional Design, Confl ict Manage- ment and Democracy, edited by Andrew Reynolds. Oxford: Oxford University Press, 30. 16. Wilford, Rick. 2001. “Th e Assembly and the Executive.” In Wilford, Rick (ed) Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press, 107. 17. O’Leary, Brendan. 1999. “Th e Nature of the Agreement.” Fordham International Law Journal 22: 1628– 67, 1631. 18. Wilford, Rick, and Robin Wilson. 2003. “Northern Ireland: Valedictory?” In Th e State of the Nations 2003: Th e Th ird Year of Devolution in the United Kingdom, edited by Robert Hazell. Exeter: Imprint Academic, 99– 100. 19. See Hennessey, Th omas. 2000. Th e Northern Ireland Peace Process: Ending the Troubles? Dublin: Gill and Macmillan. 20. See ibid.; Bew, Paul, Petter Gibbon, and Henry Patterson. 2002. Northern Ire- land 1921– 2001: Po liti cal Forces and Social Classes. London: Serif. Notes to Pages 65–70 239

21. Horowitz, Donald L. 2004. “Some Realism about Constitutional Engineering.” In Facing Ethnic Confl icts: Towards a New Realism, edited by Andreas Wimmer et al. Lanham, Md.: Rowan and Littlefi eld, 255. 22. O’Leary, Brendan. 2001. “Th e Character of the 1998 Agreement: Results and Prospects.” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Ox- ford University Press, 49. 23. See Murray, Gerard, and Jonathan Tonge. 2005. Sinn Féin and the SDLP: From Alienation to Participation. Dublin: O’Brien Press, 200– 203. 24. Ibid. 25. Adams, Gerry. 2003. Hope and History: Making Peace in Ireland. Dingle: Bran- don Books, 367– 68. 26. Murray and Tonge, Sinn Féin and the SDLP, 262. 27. Ibid. 28. Aughey, Arthur. 2001. “Learning from Th e Leopard.” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press, 191. 29. Ibid., 189. 30. Godson, Dean. 2004. Himself Alone: David Trimble and the Ordeal of Unionism . London: HarperCollins, 328. 31. Cited in ibid., 329. 32. Ibid., 333; Powell, Jonathan. 2008. Great Hatred, Little Room: Making Peace in Northern Ireland. London: Bodley Head, 90– 97. 33. Walker, Graham. 2001. “Th e British-Irish Council.” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press,130. 34. Powell, Great Hatred, Little Room, 22. 35. Dodds, Nigel. 1998. “Accept, and We Are on the Road to a United Ireland.” Parliamentary Brief 5 (6): 21. 36. Twenty-nine people were killed and around 200 people injured, the worse single atrocity committed during the Northern Ireland “Troubles.” 37. McIntyre, Anthony. 1998. “Why Stormont Reminded Me of Animal Farm.” Sunday Tribune, 12 April. 38. McEvoy, Joanne. 2006. “Th e Institutional Design of Executive Formation in Northern Ireland.” Regional and Federal Studies 16 (4): 447– 64. 39. British and Irish Governments. 1985. Anglo-Irish Agreement. http:// cain .ulst .ac .uk /events /aia /aiadoc .htm . 40. Mitchell, Paul, Brendan O’Leary, and Geoff rey Evans. 2001. “Northern Ireland: Flanking Extremists Bite the Moderates and Emerge in Th eir Clothes.” Parliamentary Aff airs 54 (4): 725– 42, 725. 41. English, Armed Struggle, 267. En glish notes the Sinn Féin 1994 document Set- ting the Record Straight: A Record of Communications Between Sinn Féin and the Brit- ish Government October 1990–November 1993; see also Maloney, Ed. 2007. A Secret History of the IRA. London: Penguin, 247 240 Notes to Pages 70–74

42. Th e infl uence of Rose’s book was suggested to the author by Professor Bren- dan O’Leary. 43. Rose, Richard. 1976. Northern Ireland: A Time of Choice. London: Macmil- lan, 151. 44. Coakley, “Implementing Consociation,” in Northern Ireland”: 137– 38. He notes that the use of the d’Hondt procedure was then taken up by a group close to the paramilitary orga ni za tion, the Ulster Defence Association. 45. O’Leary, Brendan, Tom Lyne, Jim Marshall, and Bob Rowthorn. 1991. North- ern Ireland: Sharing Authority. London: Institute of Public Policy Research. 46. Interview with Kevin McNamara, 22 July 2005 (by telephone). Th e authors of the IPPR document were all advisers/staff to Mr. McNamara in the early 1990s and were part of a committee set up to produce the policy document in response to the inter- party talks of 1991– 92. Although McNamara says he was “one degree removed” from the document, the people involved were his “very close associates.” 47. McGarry, John, and O’Leary, Brendan. 1995. Explaining Northern Ireland: Broken Images. Oxford: Blackwell, 373– 74. 48. Ibid., 374. 49. Interview with Denis Haughey, 9 June 2005. 50. Cited in Millar, Frank. 2004. David Trimble: Th e Price of Peace. Dublin: Liff ey Press, 59. 51. Ibid., 62– 63. 52. Interview with Michael McGimpsey MLA, 13 June 2005. 53. Unionist Working Group. 1991. “Administrative and Legislative Devolution,” Irish Times, 3 July. 54. Ibid. 55. Interview with Mark Durkan MP MLA, 20 May 2005. 56. Ibid. 57. Ibid. 58. Godson, Himself Alone, 338. In Millar’s book on Trimble, the former fi rst minister says that his decision to agree to cabinet government refl ected a desire to pursue a “genuine partnership” between his party and the SDLP. See Millar, David Trimble, 57– 58. 59. McEvoy, Joanne. “Th e Institutional Design of Executive Formation in North- ern Ireland.” Regional and Federal Studies 16 (4): 455– 59. 60. Godson, Himself Alone, 337. 61. Cited in ibid., 338. 62. Ibid., 339. 63. Th e semantics of the fi rst minister and deputy fi rst minister titles incurred some debate in later years. Following the nomination of Sinn Féin’s Martin McGuin- ness to the post of deputy fi rst minister in 2007, some discussion ensued over the use of upper- or lowercase d with deputy First Minister becoming the offi cial version, in keeping with the Northern Ireland Agreement Act (1998). See “Martin’s D-lemma: Notes to Pages 74–77 241

Lowering the Case of the Minister’s Title Took Top Aides Weeks.” 2008. Belfast Tele- graph, 21 March; “Are All Th ings Created Equal?” 2009. BBC News Online, 17 Febru- a r y . h t t p : / / w w w . b b c . c o m / n e w s / . 64. Cited in Hennessey, Th e Northern Ireland Peace Pro cess, 168. 65. Powell, Great Hatred, Little Room, 31. 66. Ibid., 31– 32. 67. Ibid., 32. 68. Horowitz, Donald L. 2000. “Constitutional Design: An Oxymoron?” In Design- ing Demo cratic Institutions, edited by Ian Shapiro and Stephan Macedo. New York: New York University Press, 253–84 69. Th e referendums saw a large majority in the North (71 percent), on a record turnout of 81 percent in a domestic UK jurisdiction, and an overwhelming majority in the South (94 percent), on an average turnout of 56 percent. However, a substantial number of unionists rejected it; an RTE/Lansdowne exit poll reported that 55 percent of those who described themselves as unionist supported the Agreement at the referendum. See Mitchell, Paul. 2001. “Transcending an Ethnic Party System?” In Aspects of the Belfast Agreement, edited by Rick Wilford. Oxford: Oxford University Press, 45, fn.7. 70. Wilford, “Th e Assembly and the Executive,” 116. 71. Horowitz, “Explaining the Northern Ireland Agreement,” 210. 72. See Wilford, Rick, and Robin Wilson. 2006. Th e Trouble with Northern Ireland: Th e Belfast Agreement and Democratic Governance, Dublin: Tasc at New Island, 39. 73. Wolff , Stefan. 2003. “Between Stability and Collapse: Internal and External Dynamics of Post-Agreement Institution Building in Northern Ireland.” Global Review of Ethnopolitics 3 (1): 8– 21, 20. 74. McGarry, John, and Brendan O’Leary. 2009. “Under Friendly and Less-Friendly Fire.” In Consociational Th eory: McGarry and O’Leary and the Northern Ireland Con- fl ict, edited by Rupert Taylor. London: Routledge, 360. 75. O’Leary, Brendan. 2005. “Debating Consociational Politics: Normative and Explanatory Arguments.” In From Power Sharing to Democracy: Post-Confl ict Institu- tions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s University Press, 12. 76. A “complete” consociational executive requires all signifi cant segments of the divided society to be represented and corresponds to Lijphart’s “grand co ali tion”; under a “concurrent” consociation, each segment has repre sen ta tion with at least majority support; and a “weak” consociation has repre sen ta tion of each signifi cant segment, but at least one segment has only plurality support. 77. O’Leary, “Debating Consociational Politics,” 17. 78. Th e Shadow Assembly sat nineteen times, and eleven committee reports were passed before the devolution of powers. See http:// www .niassembly .gov .uk /io /archive dreports .htm . 242 Notes to Pages 78–84

79. Godson, Himself Alone, 378. 80. Interview with Nigel Dodds MP MLA, 6 September 2005. 81. “DUP Attacks ‘Childish’ Behaviour of Its Rivals.” 2000. Irish Times, 10 June. 82. Northern Ireland Executive Information Ser vice, 5 July 2000. 83. “Cut in Ministers’ Powers Urged.” 2000. BBC News Online, 22 July. http:// www .bbc.com /news /. 84. Northern Ireland Assembly Report, July 4, 2000, 1999–2000 Session Offi cial Reports. 85. Th e DUP motion was backed by UUP members Pauline Armitage, Roy Beggs Jr., Derek Hussey, and Peter Weir; Weir had lost the UUP party whip and later joined the DUP in May 2002. 86. “Trimble: ‘Sinn Fein veto to stay.’ ” 2001. BBC News Online, 9 February. http:// www.bbc .com /news /. 87. See Purdy, Martina. 2005. Room 21: Stormont—Behind Closed Doors. Belfast: Brehon Press, 216– 18. 88. Ibid., 202. 89. Cited in ibid., 203. 90. Millar, David Trimble, 57. 91. Wilson, Robin, and Rick Wilford. 2001. “Northern Ireland: Endgame.” In Th e State of the Nations 2001, edited by Alan Trench. Exeter: Imprint Academic, 84– 85. 92. Bloomfi eld, Kenneth. 2007. A Tragedy of Errors: Th e Government and Misgov- ernment of Northern Ireland. Liverpool: Liverpool University Press, 256. 93. Agreement Reached in Multi- Party Negotiations, Annex A. 94. Interview with Nigel Dodds MLA MP, 6 September 2005. 95. Interview with Denis Haughey, 9 June 2005. Haughey stresses that his party colleague and chair of the Health Committee, Dr Joe Hendron, supported the minister’s decision to site the maternity ser vices at the Royal Victoria Hospital. 96. Bloomfi eld, A Tragedy of Errors, 255. 97. Executive Information Ser vice, 26 April 2001. 98. Interview with Bríd Rodgers, 17 October 2005. 99. McGarry, John, and Brendan O’Leary. 2009. “Power Shared Aft er the Deaths of Th ousands.” In Consociational Th eory: McGarry and O’Leary and the Northern Ire- land Confl ict, edited by Rupert Taylor. London: Routledge, 44. 100. British and Irish Governments. 1999. Th e Way Forward: Joint Statement by the Irish and British Governments at Stormont on 2 July 1999. http:// cain .ulst .ac .uk /events /peace /docs /bi2799 .htm . 101. Ibid., 37. 102. Blair, Tony. 2010. A Journey. Hutchinson: London, 172; Powell, Great Hatred, Little Room, 101. 103. Mandelson, Peter. 2010. Th e Th ird Man: Life at the Heart of New Labour. Lon- don: Harper, 291. 104. Ibid., 297. Notes to Pages 84–88 243

105. Ahern, Bertie. 2000. “Arms Body Now Believes IRA Will Decommission.” Irish Times, 14 February. http:// cain .ulst .ac .uk /events /peace /docs /ba14200 .htm . 106. “NI Talks Deadlocked.” 2000. BBC News Online. 17 February. http:// news .bbc . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 4 5 1 0 8 . s t m . 107. “Government ‘No Clue’ About Stalled NI Politics.” 2000. BBC News Online, 2 8 M a r c h . h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 9 3 3 7 0 . s t m . 108. “NI Assembly Suspension ‘Wrong.’ ” 2000. BBC News Online, 13 February. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 4 1 5 8 7 . s t m . 109. Th e Patten Commission, provided for in the GFA as the Inde pen dent Com- mission on Policing, had responsibility for making recommendations for police reform. For the details of the Police (Northern Ireland) Bill 2000, see http:// cain .ulst .ac .uk/issues /police /policeact /bill2000 .htm . 110. Sinn Féin. 2000. “Statement in Response to the Police (NI) Bill, 16 May 2000.” h t t p : / / c a i n . u l s t . a c . u k / i s s u e s / p o l i c e / p o l i c e a c t / g a 1 6 5 0 0 . h t m . 111. Ibid. 112. SDLP. 2001. “Response to the Proposals for the New Police Ser vice, 20 August 2001.” http:// cain .ulst .ac .uk /issues /police /docs /sdlp200801 .htm . 113. “PMs Look to Kick-Start Peace Process.” 2003. BBC News Online, 1 May. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 3 0 0 4 5 6 5 . s t m . 114. In August 2001, three Irish republicans were arrested in Colombia for travel- ling on false passports. In 2004, they were sentenced to seventeen years in jail for training FARC rebels. 115. “Stormontgate: Here Today and Gone Tomorrow Outlook for MLAs.” 2002. News Letter, October, 8; “Trimble Gives Blair Seven Days to Expel Sinn Féin from Executive.” 2002. Irish Times, 9 October. 116. Interestingly, Durkan later revealed that Blair had asked the SDLP to back an assembly motion to exclude Sinn Féin. See Irish Times, 2 November 2002. 117. Powell, Great Hatred, Little Room, 210. 118. Blair, A Journey, 189.

Chapter 4

1. “Historic N Ireland Deal Reached.” 2007. 26 March. http:// news .bbc .co .uk /2 /hi /europe/6497389 .stm . 2. But see Mitchell, Paul, Geoff rey Evans, and Brendan O’Leary. 2009. “Ex- tremist Outbidding in Ethnic Party Systems Is Not Inevitable: Tribune Parties in Northern Ireland.” Politi cal Studies 57 (2): 397– 421. Th e authors argue that power- sharing institutions incentivize ethnic parties to compete on a more moderate platform while also campaigning as the “strongest voice” to represent their respective community. 3. British and Irish Governments. 2003. Joint Declaration. http:// www.nio.gov.uk /joint_declaration_between_the_british_and_irish_governments.pdf . 244 Notes to Pages 89–91

4. Note the 2003 Assembly election results compared to 1998: DUP 30 seats (+10), UUP 27 seats (−1), Sinn Féin 24 seats (+6), and SDLP 18 seats (−6). See http:// www .arc .ac.uk . 5. Of note here is the nomination by Sinn Féin leader Gerry Adams of Paisley as fi rst minister and Martin McGuinness as deputy fi rst minister to the recalled Assem- bly on 22 May 2006. Unsurprisingly, given the DUP’s diffi culty sharing power with republicans, Dr. Paisley replied, “Certainly not, Madam Speaker. It goes without saying that my reasons are well known across this Province, and they have been endorsed by the majority of unionist voters.” Northern Ireland Assembly Report, 22 May 2006, 2005– 6 Session Offi cial Reports. 6. Section 11, paragraph 8 of the 1998 Agreement stipulates that “the two governments and the parties in the Assembly will convene a conference four years aft er the Agreement comes into eff ect, to review and report on its operation.” Th e Review convened on 4 February 2004, a slight delay of two months as per the text of the Agreement. 7. Cowen, Brian. 2004. Remarks by Minister Cowen at the Plenary Meeting of Re- view of the Operation of the Good Friday Agreement, February 3, 2004. http:// www .dfa .ie/home /index .aspx ?id =25847. 8. Demo cratic Unionist Party. 2004. Devolution Now: Th e DUP’s Concept for De- volution. http:// cain .ulst .ac .uk /issues /politics /docs /dup /dup050204text .htm . 9. Ibid. 10. “SF Concerns over Unionist Majority Rule in DUP Plan.” 2004. Irish Times, 11 February. 11. Sinn Féin. 2004. Agenda for Full Implementation of the Agreement. 28 January. http:// cain .ulst .ac .uk /issues /politics /docs /sf /sf280104 .htm . 12. “Scene Is Set but Noone Can Say for What Drama.” 2004. News Letter, 4 Feb- ruary. 13. “Trimble to Reject Change in Voting Procedures.” 2004. Irish Times, 13 January. 14. “Scene Is Set.” 15. “Focus ‘Is Vital’ for the Review.” 2004. Irish News, 22 January. 16. “Impetus Must Now Come from Dublin and London.” 2004. Irish Times, 2 February. 17. “Reform of Stormont Tops Agenda at New Talks Table.” 2004. Belfast Telegraph, 16 September. 18. “Monumental Failure of Po liti cal Will If Deal Slips Away Now.” 2004. Irish Times, 20 September. 19. “Imagination Required as Final Hurdle Looms Closer.” 2004. Irish News, 9 December. 20. Powell, Jonathan. 2008. Great Hatred, Little Room: Making Peace in Northern Ireland. London: Bodley Head. 21. According to Blair, the issue of photographs had been the “outstanding question.” He confi rmed that the two governments had submitted a compromise proposal Notes to Pages 92–94 245 for photographs to be taken at the time of decommissioning and published once the Executive was set up; the parties rejected the proposed compromise. See “Ahern, Blair Put Positive Gloss on Progress Made.” 2004. Irish Times, 9 December. 22. Wilford, Rick, and Robin Wilson. 2006. From the Belfast Agreement to Stable Power Sharing.” Paper presented at Po liti cal Studies Association Territorial Politics Conference, Queen’s University Belfast, January 2006, 10. 23. British and Irish Governments. 2004. Proposals for a Comprehensive Agree- ment. http:// cain .ulst .ac .uk /issues /politics /docs /nio /bi081204proposals .pdf . 24. It was clear at the two governments’ press conference, however, that these statements were a best guess, particularly with reference to the IRA. Blair conceded that the compromise of photographs taken at the time of decommissioning and published when the Executive was set up was not acceptable to republicans. In response, the Sinn Féin leader, Gerry Adams, stressed that they had dismissed the issue of photographs at Leeds Castle in September and were thus “surprised on November 17th when we received their joint statement when this demand was contained in a paragraph of a draft IICD report. . . . We asked the two governments to take it out.” His request was denied as Adams said the two governments responded that there was no other way of getting the DUP to agree to the package. See “Governments’ Document Is a Good Deal, Says Adams.” 2004. Irish Times, 9 December. 25. British and Irish Governments, Proposals for a Comprehensive Agreement, Annex B, para 6. 26. Ibid., para 8. 27. Interview with Richard Bullick, 29 June 2005. 28. Ibid. 29. Alliance Party. 2004. “Response to Proposals by the British and Irish Govern- ments for a Comprehensive Agreement.” Alliance News, November/December. 30. Interview with Stephen Farry, 18 June 2005. 31. Th e DUP’s Devolution Now document included the option of a “mandatory coali tion” model whereby arrangements could be made to allow “each designation” ( unionist and nationalist) to “elect to the positions of First Minister and Deputy First Minister as appropriate.” See http:// www .dup .org .uk /pdf /DUPDevolution _lo -res.pdf . 32. See Inde pen dent International Commission on Decommissioning. 2005. Re- port of the In de pen dent International Commission on Decommissioning, 26 September 2005. http:// cain .ulst .ac .uk /events /peace /decommission /iicd260905 .pdf. Th e DUP was, however, critical of the lack of photographic evidence and absence of an inventory, with Paisley referring to the “duplicity and dishonesty of the two governments and the IRA.” See “IRA Weapons: Statements in Full.” 2005. BBC News Online, September 26. http:// www .bbc .co .uk /. 33. Powell, Great Hatred, Little Room, 276. 34. Ibid. 35. Ibid., 279. 246 Notes to Pages 94–99

36. See Clancy, Mary-Alice C. 2010. Peace Without Consensus: Power Sharing Politics in Northern Ireland. Burlington, Vt.: Ashgate, 156– 57, 159– 66. 37. Powell, Great Hatred, Little Room, 280. 38. Clancy, Peace Without Consensus, 163. 39. O’Duff y, Brendan. 2007. British-Irish Relations and Northern Ireland: From Violent Politics to Confl ict Regulation. Dublin: Irish Academic Press, 189. 40. “Sinn Fein Policing Motion in Full.” 2007. BBC News Online, 28 January. http:// www .bbc .co .uk /. 41. “Sinn Fein Votes to Support Police.” 2007. BBC News Online, 28 January. http:// www .bbc .co .uk /. 42. Wilford, Rick. 2010. “Northern Ireland: Th e Politics of Constraint.” Parliamen- tary Aff airs 63 (1): 134–55. 43. Northern Ireland (St. Andrews Agreement) Bill (2006). http://cain.ulst.ac .uk/hmso/nistandrewsbill161106.pdf . 44. Ibid., para. 28A. 45. Coakley, John. 2009. “Implementing Consociation in Northern Ireland.” In Consociational Th eory: McGarry and O’Leary and the Northern Ireland Confl ict, edited by Rupert Taylor. London: Routledge, 142. 46. “Hain in Paisley and Adams Talks.” 2007. BBC News Online, 9 March. http:// www.bbc .co .uk /. 47. “No Alternative to Deal—Paisley.” 2007. BBC News Online, 4 April. http:// www .bbc.co .uk /. 48. Powell, Great Hatred, Little Room, 31– 32. 49. “Paisley Will Accept Nomination.” 2006. BBC News Online, 24 November. http:// www .bbc .co .uk / 50. Committee on the Programme for Government. 2006. Press Release, 4 Dec 2006, available at: http:// www .niAssembly .gov .uk /transitional /pfg /press /pn _02 .htm . 51. Th e transitional Assembly approved a draft ministerial code on 20 March 2007. See http:// www .niAssembly .gov .uk /transitional /info _offi ce/draft _ministerial _code .pdf. 52. “McGuinness Confi rms Chuckle Image.” 2007. BBC News Online, 7 November. http:// www .bbc .co .uk /. 53. Ross, Marc Howard. 2007. Cultural Contestation in Ethnic Confl ict. Cambridge: Cambridge University Press; Ross, Marc Howard, ed. 2009. Culture and Belonging in Divided Societies. Philadelphia: University of Pennsylvania Press. 54. British and Irish Governments. 2006. Agreement at St. Andrews. http:// www . n i o . g o v . u k / s t _ a n d r e w s _ a g r e e m e n t . p d f . 55. “Irish Language Future Is Raised.” 2006. BBC News Online, 13 December. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 1 7 7 3 2 1 . s t m . 56. “Irish Language Delay ‘Bad Faith.’ ” 2007. BBC News Online, 15 March. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 4 5 4 7 5 5 . s t m . Notes to Pages 99–104 247

57. “Implement Language Clause, Adams.” 2007. BBC News Online, 26 July. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 9 1 7 0 4 0 . s t m . 58. Ibid. 59. Sinn Féin. 2007. “Bairbre de Brún Addresses MEPs on Situation of Irish Lan- guage.” Sinn Féin News, 14 November. http:// www .sinnfein .ie /contents /10628 . 60. “£20m ‘Pledge for Irish Language.’ ” 2010. BBC News Online, 15 February. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 8 5 1 6 4 8 1 . s t m . 61. Department of Culture, Arts and Leisure. n.d. Ulster-Scots. http:// www .dcalni .gov.uk . 62. “Maze Site ‘Should Feature Museum.’ ” 2006. BBC News Online, 16 August. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 4 7 9 9 0 5 1 . s t m . 63. “Robinson Open- Minded on Maze Site.” 2007. BBC News Online, 21 July. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 9 0 9 2 4 3 . s t m . 64. “Maze ‘Will Not Be Terror Shrine.’ ” 2007. BBC News Online, 25 July. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 6 9 1 6 0 4 8 . s t m . 65. “Maze Site Set to Be Redeveloped.” 2010. BBC News Online, 29 July. http:// www.bbc .co .uk /news /uk -northern -ireland -10809186 . 66. “Is Stormont Heading for Stalemate?” 2008. BBC News Online, 19 July. http:// news .bbc .co .uk /2 /hi /uk _news /northern _ireland /northern _ireland _politics /7500762.stm . 67. See http:// cain .ulst .ac .uk /issues /politics /ofmdfm /ofmdfm181108letter .pdf . 68. “NI Parties Agree Power- Sharing Deal.” 2010. BBC News Online, 5 February. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 8 4 9 9 9 5 2 . s t m . 69. “McGuinness Speaks out over Th reat.” 2009. BBC News Online, 24 April. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 8 0 1 5 8 8 1 . s t m . 70. “Robinson and McGuinness Condemn NI Rioting.” 2011. BBC News Online, 13 July. http:// www .bbc .co .uk /news /10617267 . 71. “Councils Won’t Be Tied by Boundary Commission.” 2007. Belfast Telegraph, 12 June. 72. “Th e Executive Fails to Agree on Council Reform.” 2010. BBC News Online, 15 June. http:// www .bbc .co .uk /news . 73. “NI Anti-Sectarian Proposals Criticised.” 2010. BBC News Online, 29 October. http:// www .bbc .co .uk /news /uk -northern -ireland -11652370 . 74. “MLAs Get 11-Plus Stalemate Plan.” 2010. BBC News Online, 20 May. http:// www .bbc .co .uk /news . 75. Wilford, “Th e Politics of Constraint,” 147. 76. “Executive ‘Carved-Up’—Ritchie.” 2007. BBC News Online, 19 October. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 7 0 5 2 1 1 8 . s t m . 77. Wilford, “Th e Politics of Constraint,” 152. 78. Ibid., 151. 79. Ibid. 248 Notes to Pages 104–109

80. “Ian Paisley’s Final Day at Stormont as MLA.” 2011. BBC News Online, 24 March. http:// www .bbc .co .uk /news /uk -northern -ireland -12832095 .

Chapter 5

1. See Daalder, Ivo H. 2000. Getting to Dayton: Th e Making of America’s Bosnia Policy. Washington, D.C.: Brookings Institution; Dyker, David A., and Ivan Vejvoda, eds. 1996. Yugo slavia and Aft er: A Study in Fragmentation, Despair and Rebirth. London and New York: Longman; Glenny, Misha. 1996. Th e Fall of Yugoslavia . London: Pen- guin; Silber, Laura, and Allan Little. 1996. Yugo slavia: Death of a Nation. London: Penguin; Woodward, Susan L. 1995. Balkan Tragedy: Chaos and Dissolution Aft er the Cold War. Washington, D.C.: Brookings Institution. 2. Note that Croatian President Franjo Tudjman negotiated on behalf of the Bos- nian Croats, and Serbian President Slobodan Milošević negotiated on behalf of the Bosnian Serbs, sidelining the Bosnian Serbs present, who were apparently treated with “open contempt” by Milošević. See Holbrooke, Richard. 1998. To End a War. New York: Random House, 243. 3. Ibid., 232. Holbrooke comments that these objectives required agreement on several issues: “eastern Slavonia, the Federation, a constitutional framework, elections, a three-person presidency, a national assembly, freedom of movement and the right of refugees to return to their homes, compliance with the International War crimes Tribu- nal, and an international police force.” 4. Ibid., 237. 5. Ibid., 241. 6. See Bildt, Carl. 1998. Peace Journey: Th e Struggle for Peace in Bosnia. London: Weidenfeld and Nicolson, 130– 31. 7. Ibid., 124– 25. Former Swedish prime minister Bildt was the EU’s Special Envoy to the Former Yugoslavia and co-chair of the peace talks in Dayton. He was then appointed as the fi rst high representative for Bosnia and Herzegovina, serving from December 1995 to June 1997. 8. Holbrooke, To End a War, 262– 63. 9. Ibid., 274– 75. 10. Ibid., 275. 11. Ibid., 308. 12. Ibid., 280. 13. Bildt, Peace Journey, 136. 14. Ibid., 145. 15. See Bieber, Florian, and Sören Keil. 2009. “Power Sharing Revisited: Les- sons Learned in the Balkans?” Review of Central and East Eu ro pe an Law 34 (4): 337–60. 16. Cox, Marcus. 2003. “Building Democracy from the Outside: Th e Dayton Agreement in Bosnia and Herzegovina.” In Can Democracy Be Designed? Th e Politics Notes to Pages 109–114 249 of Institutional Choice in Confl ict- torn Societies, edited by Sunil Bastian and Robin Luckham. London & New York: Zed Books, 258. 17. See Bildt, Peace Journey, 157. 18. Weller, Marc, and Stefan Wolff . 2006. “Bosnia and Herzegovina Ten Years Aft er Dayton: Lessons for Internationalized State Building.” Ethnopolitics 5 (1): 1– 13, 2. 19. Bildt, Peace Journey, 139. 20. Daalder, Getting to Dayton, 161. 21. Offi ce of the High Representative. 1995. Th e General Framework Agreement for Peace in Bosnia and Herzegovina. h t t p : / / w w w . o h r . i n t / d p a / d e f a u l t . a s p ? c o n t e n t _ i d =380 . 22. McGarry, John, and Brendan O’Leary. 2009. “Must Plurinational Federations Fail?” Ethnopolitics 8 (1): 5– 25, 6. 23. Cox, “Building Democracy,” 261. 24. Th e RS Constitution was adopted in 1992 and the Constitution of the Federation of Bosnia and Herzegovina was adopted in 1994 as part of the Washington Agreement. 25. Bieber, Florian. 2006. “Aft er Dayton, Dayton? Th e Evolution of an Unpopu lar Peace.” Ethnopolitics 5 (1): 15– 31. 26. Bieber and Keil, “Power Sharing Revisited,” 356. 27. Ibid., 357. 28. McGarry and O’Leary, “Must Plurinational Federations Fail,” 6. 29. Bildt, Peace Journey, 138. 30. Th e Dayton Peace Agreement accorded veto rights to the constituent peoples and the two entities in the state parliament; the VNI veto in the entities was introduced following a Constitutional Court decision in 2000 and the subsequent imposition of constitutional amendments by the high representative in 2002; see the OHR web site: http:// www .ohr .int . 31. See McCrudden, Christopher, and Brendan O’Leary. 2013. Courts and Conso- ciations: Human Rights versus Power Sharing. Oxford: Oxford University Press. 32. Chandler, David. 2000. Bosnia: Faking Democracy Aft er Dayton. London: Pluto Press, 43. 33. Ibid., 43– 51. 34. Bieber, “Aft er Dayton, Dayton,” 27. 35. Weller and Wolff , “Bosnia and Herzegovina,” 4. 36. O’Leary, Brendan. 2005. “Debating Consociational Politics: Normative and Explanatory Arguments.” In From Power Sharing to Democracy: Post-Confl ict Institu- tions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s University Press, 34. 37. Belloni, Robert. 2004. “Peacebuilding and Consociational Engineering in Bos- nia and Herzegovina.” International Peacekeeping 11 (2): 334– 53; Bieber and Keil, “Power Sharing Revisited.” 38. Caspersen, Nina. 2004. “Good Fences Make Good Neighbours? A Compari- son of Confl ict-Regulation Strategies in Postwar Bosnia.” Journal of Peace Research 41 (5): 569– 88. 250 Notes to Pages 114–118

39. Bieber, “Aft er Dayton, Dayton,” 27; Chandler, David. 2006. “From Dayton to Europe?” In Peace Without Politics? Ten Years of International State-Building in Bos- nia, edited by David Chandler. London: Routledge. 40. Weller and Wolff . “Bosnia and Herzegovina,” 4. 41. Zahar, Marie- Joëlle. 2005. “Th e Dichotomy of International Mediation and Leader Intransigence: Th e Case of Bosnia and Herzegovina.” In Power Sharing: New Challenges for Divided Societies, edited by Ian O’Flynn and David Russell. London: Pluto Press, 127– 28. 42. Ibid., 128. 43. Sebastián, Sofi a. 2007, November. Leaving Dayton Behind: Constitutional Re- form in Bosnia and Herzegovina. FRIDE working paper. 3. 44. Bieber, “Aft er Dayton, Dayton,” 21. 45. Bieber and Keil, “Power Sharing Revisited,” 354. 46. Interview with international offi cial, Sarajevo, May 2008. 47. Interview with OHR offi cial, Sarajevo, May 2008. 48. McEvoy, Joanne. 2013. “We Forbid! Th e Mutual Veto and Power-Sharing De- mocracy.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 49. Bieber, “Aft er Dayton, Dayton,” 25. 50. Cox, “Building Democracy from the Outside,” 261– 62. 51. Ibid., 262. 52. Ibid., 263. 53. Cited in Chandler, David. 2000. Bosnia: Faking Democracy Aft er Dayton. Lon- don: Pluto Press, 67. 54. Cox, “Building Democracy from the Outside,” 253. 55. Ibid., 268. 56. Manning, Carrie. 2004. “Elections and Po liti cal Change in Post- War Bosnia and Herzegovina.” Demo cratization 11 (2): 60– 86, 63. 57. Ibid., 64. 58. Ibid., 67. 59. Ibid., 68. 60. Ibid. 61. Bieber, “Aft er Dayton, Dayton,” 18. 62. International Crisis Group. 1997. Dayton: Two Years On: A Review of Progress in Implementing the Dayton Peace Accords in Bosnia (Report No. 27). Sarajevo: ICG Bosnia Project, 12. 63. International Crisis Group. 1999. Is Dayton Failing? Bosnia Four Years Aft er the Peace Agreement (Balkans Report No. 80). Sarajevo: ICG Bosnia Project, 23. 64. Cox, Marcus. 2008. “Bosnia and Herzegovina: Th e Limits of Liberal Imperial- ism.” In Building States to Build Peace, edited by Charles T. Call. Boulder, Colo.: Lynne Rienner, 259. Notes to Pages 118–121 251

65. See Offi ce of the High Representative. 1998, 12 January. “Th e High Represen- tative Declares the Establishment of an In de pen dent Commission to Propose Alterna- tives for the Flag of Bosnia.” Press Release. http:// www .ohr .int /ohr -dept /presso /pressr /archive .asp ?m= & yr=1998 . 66. International Crisis Group. 2002. Bosnia’s Alliance for (Smallish) Change (Eu- rope Report No. 132). Sarajevo: ICG Bosnia Project, 2. 67. Weller and Wolff , “Bosnia and Herzegovina,” 5. 68. International Crisis Group, Is Dayton Failing, 2. 69. Ibid., 3. 70. Ibid. 71. Cox, “Th e Limits of Liberal Imperialism,” 259. 72. Chandler, Bosnia: Faking Democracy Aft er Dayton, 70. 73. Ibid., 71. 74. Cox, “Th e Limits of Liberal Imperialism,” 259. 75. Peace Implementation Council. 1997. PIC Sintra Declaration, May 30, 1997. h t t p : / / w w w . o h r . i n t / p i c / d e f a u l t . a s p ? c o n t e n t _ i d =5180 . 76. Peace Implementation Council. 1997. Bonn Conclusions, December 10, 1997. h t t p : / / w w w . o h r . i n t / p i c / d e f a u l t . a s p ? c o n t e n t _ i d =5182 #11 . 77. OHR. 1997. “Statement to the Media: PIC Proceedings 9 December 1997.” h t t p : / / w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =7432 . 78. Knaus, Gerald, and Felix Martin. 2003. “Th e Travails of the Euro pe an Raj: Les- sons from Bosnia and Herzegovina.” Journal of Democracy 14 (3): 60– 74, 64. 79. OHR. 1999, 14 April. “Decisions on Property Laws.” Press Release. http:// www . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4723. 80. OHR. 1999, 31 July. “Decisions Furthering the Implementation of the Dayton Peace Agreement in Bosnia and Herzegovina.” Press Release. http:// www .ohr .int /ohr -dept/presso /pressr /default .asp ?content _id =4769 . 81. OHR. 1999, 3 August. “Additional Decisions Furthering the Implementation of the Dayton Peace Agreement in Bosnia and Herzegovina.” Press Release. http:// w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4770. 82. Cited in Bose, Sumantra. 2006. “Th e Bosnian State a De cade Aft er Dayton.” In Peace Without Politics? Ten Years of International State-Building in Bosnia, edited by David Chandler. London: Routledge, 16. 83. Chandler, Bosnia: Faking Democracy, 70. 84. See OHR. 1998, 22 July. “High Representative to Impose Framework Privatisa- tion Law Following Failure of House of Peoples to Do Th is.” Press Release. http:// www . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4637 . 85. See OHR. 1999, 5 February. “Council of Ministers.” Press Release. http:// www . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4695 . 86. OHR. 1999, 5 March. “President Poplasen Dismissed.” Press Release. http:// www . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4705. 252 Notes to Pages 121–126

87. OHR. 1999, 29 November. “High Representative and OSCE Head of Mission Announce the Removal from Offi ce of Various Public Offi cials.” Press Release. http:// w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _ i d =4825 . 88. Cox, “Th e Limits of Liberal Imperialism,” 262. 89. International Crisis Group, Is Dayton Failing, 50. 90. International Crisis Group, Bosnia’s Alliance, 1. 91. Ibid. 92. Ibid., 16. 93. Bieber, “Aft er Dayton, Dayton,” 19. 94. OHR. 2001, 18 July. 19th Report by the High Representative for Implementation of the Peace Agreement to the Secretary-General of the United Nations. http:// www .ohr . i n t / o t h e r - d o c / h r - r e p o r t s / d e f a u l t . a s p ? c o n t e n t _ i d =5126. 95. International Crisis Group. 2003. Bosnia’s Nationalist Governments: Paddy Ashdown and the Paradoxes of State Building (Europe Report No. 146). Sarajevo: ICG Bosnia Project, 6. 96. Ibid., 5. 97. International Crisis Group. 2002, 16 April 16. Implementing Equality: Th e “Constituent Peoples” Decision in Bosnia & Herzegovina. Sarajevo: ICG Bosnia Proj- ect, i. 98. Constitutional Court of Bosnia and Herzegovina, Case No. U5/98 (Partial Decision Part Th ree), para. 54. 99. International Crisis Group, Implementing Equality, ii. 100. Ibid., 8. 101. OHR. 2002, 19 April. “Process of Constitutional Change in Bosnia and Her- z e g o v i n a ’ s E n t i t i e s I s C o m p l e t e d . ” P r e s s R e l e a s e . h t t p : / / w w w . o h r . i n t / o h r - d e p t / p r e s s o /pressr/default .asp ?content _id =7478 . 102. See International Crisis Group, Bosnia’s Nationalist Governments, 3– 5. 103. Th e 2002 elections were run for the fi rst time by the Bosnia authorities rather than the OSCE and elected governments for four rather than two years. 104. International Crisis Group. 2007, 15 February. Ensuring Bosnia’s Future: A New International Engagement Strategy (Eu rope Report No. 180). Sarajevo: ICG Bosnia Project 6. Carlos Westendorp (HR 1997–99) made 76 decisions, Wolfgang Petritsch (HR 1999– 2002) 250 decisions, and Paddy Ashdown (HR 2002– 6) 447 decisions. Inter- view with OHR offi cial, Sarajevo, May 2011. 105. International Crisis Group, Bosnia’s Nationalist Governments, 1. 106. Cited in Chandler, David. 2002. “Bosnia’s New Colonial Governor.” Guardian, 9 July. 107. International Crisis Group, Bosnia’s Nationalist Governments, 38. 108. Peace Implementation Council. 2003, 30 January. Declaration of the Politi cal Directors of the Peace Implementation Steering Board. h t t p : / / w w w . o h r . i n t / p i c / d e f a u l t .asp ?content _id=29144 . Notes to Pages 126–132 253

109. Interview with OHR offi cial, Sarajevo, May 2011. 110. “Farewell Sarajevo.” 2005. Guardian, 2 November, cited in Szewczyk, Bart M. J. 2010. Th e EU in Bosnia and Herzegovina: Powers, Decisions and Legitimacy. Eu ro- pean Union Institute of Security Studies, 19. http:// www .iss .europa .eu /uploads /media /OccasionalPaper83.pdf . 111. Szewczyk, Th e EU in Bosnia, 35– 39. 112. Chandler, “Bosnia’s New Colonial Governor.” 113. Knaus and Martin, “Lessons from Bosnia and Herzegovina,” 69. 114. Cited in International Crisis Group, Ensuring Bosnia’s Future, 6. 115. OHR. 2002, 21 October. “High Representative Changes Law on Ministries i n t h e F e d e r a t i o n B i H . ” P r e s s R e l e a s e . h t t p : / / w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t .asp ?content _id=28196. 116. See OHR. 2002, 3 December. “Law on the Council of Ministers.” http:// www . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d =28609 . 117. See OHR. 2003, 12 February. “Decision Establishing the Indirect Tax Policy C o m m i s s i o n . ” h t t p : / / w w w . o h r . i n t / d e c i s i o n s / e c o n d e c / d e f a u l t . a s p ? c o n t e n t _ i d =29240 . 118. See OHR. 2003, 9 May. “Decision Establishing the Defence Reform Com- m i s s i o n . ” h t t p : / / w w w . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d =29840; OHR. 2003, 30 May. “Decision Establishing the Expert Commission on Intelli- g e n c e R e f o r m . ” h t t p : / / w w w . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d = 29988 . 119. International Crisis Group, Bosnia’s Nationalist Governments, 37. 120. OHR. 2003, 3 December. “Hard Work and Confi dence Will Take BiH from D a y t o n t o B r u s s e l s . ” P r e s s R e l e a s e . h t t p : / / w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t .asp ?content _id=31296 . 121. See International Crisis Group. 2005, 6 September. Bosnia’s Stalled Police Re- form: No Progress, No EU (Eu rope Report No.164). Sarajevo: ICG Bosnia Project. 122. “Bosnian Police Reform Mapped Out.” 2005. BBC News Online, 1 February. h t t p : / / n e w s . b b c . c o . u k / 2 / h i / e u r o p e / 4 2 2 5 4 4 9 . s t m . 123. International Crisis Group, Bosnia’s Stalled Police Reform, 8. 124. Peace Implementation Council. 2006. “Towards Ownership: From Peace Im- p l e m e n t a t i o n t o E u r o - A t l a n t i c I n t e g r a t i o n . ” h t t p : / / w w w . o h r . i n t / p i c / d e f a u l t . a s p ? c o n t e n t _id=37503. 125. Knaus and Martin, “Lessons from Bosnia and Herzegovina,” 73.

Chapter 6

1. Bieber, Florian. 2008. “Power- Sharing and International Intervention: Overcom- ing the Post- Confl ict Legacy in Bosnia and Herzegovina.” In Settling Self- Determination Disputes: Complex Power-Sharing in Th eory and Practice, edited by Marc Weller and Barbara Metzger. Leiden: Martinus Nijhoff , 194. 254 Notes to Pages 133–138

2. Venice Commission. 2005. Opinion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the High Representative, CDL-AD (2005) 004, Ven- ice, 11 March 2005. http:// venice .coe .int . 3. Ibid., 8. 4. Ibid., 25. 5. Ibid., 22. 6. Ibid., 26. 7. Rehn, Olli. 2005. “From Peace-Building to State-Building.” Ten Years of Dayton and Beyond, Conference, Geneva, 20 October 2005. http:// europa .eu /rapid /press ReleasesAction .do ?reference=SPEECH /05 /629 . 8. Rehn, Olli. 2005, 25 November. “Bosnia and Herzegovina: Moving Closer to Europe.” http:// www .eu -un .europa .eu /articles /en /article _5361 _en .htm . 9. Hays, Don, and Jason Crosby. 2006. From Dayton to Brussels: Constitutional Pre- parations for Bosnia’s EU Accession. Washington, D.C.: United States Institute of Peace, 7. 10. U.S. State Department. 2005, 22 November. “Rice Hails Bosnian Pledge to Em- bark on Constitutional Reform.” Press Release. http:// usinfo .state .gov . 11. Hays and Crosby, “From Dayton to Brussels,” 8. 12. Sebastián, Sofi a. 2007. Leaving Dayton Behind: Constitutional Reform in Bos- nia and Herzegovina. Madrid: FRIDE Working Paper, 6. 13. See U.S. State Department. 2006, 21 March. “United States Welcomes Bosnian Constitutional Reform Agreement.” Press Release. htt://usinfo.state.gov. 14. Sebastián, Leaving Dayton Behind, 6. 15. Interview with Matthew Rycroft , UK Embassy, Sarajevo, May 2008. 16. Sebastián, Leaving Dayton Behind: 12. 17. Interview with SDA politician, Sarajevo, May 2008. 18. TransConfl ict. 2010, 16 November. “Bosnia—Th e Challenges of Constitutional Reform.” Interview with Professor Bruce Hitchner. http:// www .transconfl ict .com . 19. Peace Implementation Council. 2006, 28 April. “Steering Board Regrets Reform F a i l u r e . ” h t t p : / / w w w . o h r . i n t / p i c / d e f a u l t . a s p ? c o n t e n t _ i d =37082. 20. Interview with Matthew Rycroft , UK Embassy, Sarajevo, May 2008. 21. McLaughlin, Daniel. 2006. “ ‘Sad Day’ for Bosnia as Reform Plan Is Rejected.” Irish Times, 28 April. 22. OHR. 2006, 17 April. “HR/EUSR Comments on Failure to Adopt Constitu- t i o n a l A m e n d m e n t s . ” h t t p : / / w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s r / d e f a u l t . a s p ? c o n t e n t _id=37070. 23. TransConfl ict, “Bosnia.” 24. Sebastián, Leaving Dayton Behind, 1. 25. Ibid., 5. 26. Cited in ibid., 7. 27. Hays and Crosby, From Dayton to Brussels, 3. 28. Th ere was some controversy over Komšić’s victory. Some Bosnian Croat politicians protested that though Komšić is an ethnic Croat, his party, SDP, attracts votes Notes to Pages 138–142 255 mainly from the Bosniak community, and he did not secure majority support in any majority Croat canton in the Federation. 29. Balkan Insight. 2007, 25 January. “Bosnia: Constitution Reform Setback.” http:// birn.eu .com /en /67 /10 /2141 /. 30. Interview with SDA politician, Sarajevo, May 2008. 31. Interview with SDA politician, Sarajevo, May 2008. 32. Balkan Insight. 2007, 6 November. “Washington Summit Aims to Revive Bos- nia’s Reforms.” http:// birn .eu .com /en /84 /10 /3070 /?tpl=30 . 33. Ibid. 34. Peace Implementation Council. 2006, 23 June. “Towards Own ership: From P e a c e I m p l e m e n t a t i o n t o E u r o - A t l a n t i c I n t e g r a t i o n . ” h t t p : / / w w w . o h r . i n t / o h r - d e p t / r u l e - o f - l a w - p i l l a r / p r c / p r c - p i c / d e f a u l t . a s p ? c o n t e n t _ i d =37631. 35. In 1999, the Arbitral Tribunal for Brčko issued the Final Award establishing the Brčko District under BiH sovereignty with local self- government. Th e completion of the Brčko Final Award implementation was set as a condition for OHR closure. See OHR. 2001, 28 August. “History and Mandate of the OHR North/Brcko.” http:// www .ohr.int /ohr -offi c e s / b r c k o / h i s t o r y / d e f a u l t . a s p ? c o n t e n t _ i d =5531. 36. Peace Implementation Council. 2008, 27 February. “Declaration by the Steer- ing Board of the Peace Implementation Council.” http:// www .ohr .int /pic /default .asp ?content _id=41352 . 37. SETimes .com. 2008, 11 November. “Brussels Welcomes BiH Agreement, Calls for Concrete Action.” http:// setimes .com . 38. OHR. 2008, 12 November. “Lajčák: Prud Agreement Must Not Be a Show for the International Community.” http:// www .ohr .int /ohr -dept /presso /pressr /default .asp ?content_id =42609 . 39. SETimes .com. 2009, 27 January. “BiH Main Parties Announce New Decen- tralisation Agreement.” http:// setimes .com . 40. SETimes.com. 2009, 20 July. “Inzko Says BiH’s Prud Process Basically Dead.” http:// setimes .com . 41. TransConfl ict, “Bosnia.” 42. SETimes .com. 2009, 22 October. “BiH’s Butmir Pro cess to Continue.” http:// setimes .com . 43. International Crisis Group. 2009, 12 November. Bosnia’s Dual Crisis. Europe Briefi ng No. 57, 4. 44. Ibid. 45. TransConfl ict, “Bosnia.” 46. International Crisis Group, Bosnia’s Dual Crisis, 5. 47. Cited in Balkan Insight. 2009, 23 October. “Dodik Slams Butmir Package, P r o p o s e s N e w T a l k s . ” h t t p : / / w w w . b a l k a n i n s i g h t . c o m / e n / a r t i c l e / d o d i k - s l a m s - b u t m i r -package -proposes -new -talks . 48. Interview with international NGO, Sarajevo, May 2011. 49. Interview with EU Delegation to BiH, Sarajevo, May 2011. 256 Notes to Pages 142–145

50. Peace Implementation Council. 2010, 30 June. “Communiqué of the Steering Board of the Peace Implementation Council.” http:// www .ohr .int /pic /default .asp ? content _id=45102 . 51. Eu ro pe an Commission. 2010. Communication from the Commission to the Council and the Eu ro pe an Parliament “Enlargement Strategy and Main Challenges 2010– 2011”: Conclusions on Bosnia and Herzegovina. COM(2010) 66 fi nal. http:// ec .europa . e u / e n l a r g e m e n t / p d f / k e y _ d o c u m e n t s / 2 0 1 0 / p a c k a g e / c o n c l u s i o n s _ b o s n i a _ e n . p d f . 52. Interview with international NGO, Sarajevo, May 2011. Following the general election in October 2010, the parties failed to agree on co ali tion formation. Much of the delay was caused by the insistence of two Croat parties, HDZ and HDZ 1990, that only a person nominated by them should become chair of the Council of Ministers. Th e deadlock was fi nally broken in late December 2011. See Balkan Insight. 2011, 29 December. “Vjekoslav Bevanda to Be Named Bosnian PM.” http:// www .balkaninsight . c o m / e n / a r t i c l e / v j e k o s l a v - b e v a n d a - t o - b e - t h e - n e w - b o s n i a n - p m . Th e new Council of Ministers took eff ect in February 2012, sixteen months aft er the election. 53. Peace Implementation Council. 2007, 31 October. “Declaration by the Steering Board of the Peace Implementation Council.” http:// www .ohr .int /pic /default .asp ? content _id=40758 . 54. Interview with OHR offi cial, Sarajevo, May 2011. 55. Ibid. 56. Interview with OHR offi cial, Sarajevo, May 2008. 57. Ibid. 58. International Crisis Group. 2009, 9 March. Bosnia’s Incomplete Transition: Be- tween Dayton and Eu rope. Eu rope Report No. 198, 2. 59. Cited in International Crisis Group. 2005, 6 September. Bosnia’s Stalled Police Reform: No Progress, No EU. Eu rope Report 164, 8. 60. Cited in International Crisis Group. 2007, 15 February. Ensuring Bosnia’s Fu- ture: A New International Engagement Strategy. Eu rope Report No. 180, 19. 61. OHR. 2007, 21 September 21. “Lajčák: Without Police Reform the Door to the E U W i l l B e C l o s e d . ” h t t p : / / w w w . o h r . i n t / o h r - d e p t / r u l e - o f - l a w - p i l l a r / p r c / p r c - a r t i c l e s / d e f a u l t . a s p ? c o n t e n t _ i d =40560 . 62. See OHR. 2007, 22 November. Action Plan for Implementation of the Mostar Declaration. h t t p : / / w w w . o h r . i n t / o h r - d e p t / r u l e - o f - l a w - p i l l a r / p r c / p r c - o t h e r / d e f a u l t . a s p ?content _id=40959 . 63. OHR. 2008, 19 April. Breakthrough for BiH on the Road to Europe . http:// www . o h r . i n t / o h r - d e p t / r u l e - o f - l a w - p i l l a r / p r c / p r c - p r / d e f a u l t . a s p ? c o n t e n t _ i d =41628 . 64. International Crisis Group, Bosnia’s Incomplete Transition, 20. 65. OHR. 2007, 19 October. “Decision Enacting the Law on Changes and Amend- ments to the Law on the Council of Ministers of Bosnia and Herzegovina.” http:// www . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d =40687. 66. OHR. 2007, 3 December. “Decision Enacting the Authentic Interpretation of the Law on Changes and Amendments to the Law on the Council of Ministers of Bos- Notes to Pages 146–151 257 nia and Herzegovina Enacted by the Decision of the High Representative of 19 October 2 0 0 7 . ” h t t p : / / w w w . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d =40931. 67. Ibid., 7. 68. International Crisis Group, Bosnia’s Dual Crisis, 3. 69. Ibid., 2. 70. International Crisis Group, Bosnia’s Incomplete Transition, 5. 71. Ibid., 8. 72. Ibid., 8, n. 68. 73. Cited in ibid., 9. 74. Interview with SDA politician, Sarajevo, May 2008. 75. Interview with SBiH politician, Sarajevo, May 2008. 76. Venice Commission. 2005, 11 March. Opinion on the Constitutional Situation in Bosnia and Herzegovina and the Powers of the High Representative. CDL- AD (2005) 004, 9 (para. 29). http:// venice .coe .intpara . 77. Interview with Matthew Rycroft , UK Embassy, Sarajevo, May 2008. 78. Eu ro pe an Commission. 1999, 26 May. Communication from the Commission to the Council and the Euro pe an Parliament on the Stabilisation and Association Process for Countries of South-Eastern Europe . COM(1999) 235 fi nal. http:// europa .eu /legislation _ s u m m a r i e s / e n l a r g e m e n t / w e s t e r n _ b a l k a n s / r 1 8 0 0 3 _ e n . h t m . 79. Eu ro pe an Commission. 2003, 18 November. “Bosnia and Herzegovina: Com- mission Approves Feasibility Study.” Press Release. http:// europa .eu . 80. Eu ro pe an Commission. 2011. Progress Report Conclusions on Bosnia and Her- zegovina. h t t p : / / e c . e u r o p a . e u / e n l a r g e m e n t / p d f / k e y _ d o c u m e n t s / 2 0 0 9 / c o n c l u s i o n s _ o n _bih_en .pdf . 81. Sebastián, Sofi a. 2010, September. Seizing Momentum in Bosnia. FRIDE policy brief No. 54, 2. 82. Interview with UK Embassy, Sarajevo, May 2008. 83. Interview with SDA politician, Sarajevo, May 2008. 84. Interview with EU Delegation to BiH, Sarajevo, May 2011. 85. Euro pe an Commission, Progress Report Conclusions on Bosnia and Herzegov- ina; Interview with EU Delegation to Bosnia and Herzegovina, Sarajevo, May 2011. 86. Interview with EU Delegation to Bosnia and Herzegovina, Sarajevo, May 2011. 87. Interview with OHR offi cial, Sarajevo, May 2008. 88. Sebastian, Seizing Momentum in Bosnia, 2. 89. Noutcheva, Gergana. 2009. “Fake, Partial and Imposed Compliance: Th e Lim- its of the EU’s Normative Power in the Western Balkans.” J o u r n a l o f E ur op ea n P u b l i c Policy 16: (7): 1065– 84, 1075. 90. Ibid., 1077– 79. 91. International Crisis Group, Bosnia’s Incomplete Transition, 15. 92. International Crisis Group, Bosnia’s Dual Crisis, 1. 93. International Crisis Group, Ensuring Bosnia’s Future, 6. 94. Ibid., 26. 258 Notes to Pages 151–160

95. Interview with OHR offi cial, Sarajevo, May 2008. 96. Compare Lajčák’s media interviews in July 2007: Prlenda, Antonio. 2007. “In- terview: Miroslav Lajčák, ‘Action Starts from September.’ ” Oslobodenje, 5 July. http:// w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s i / d e f a u l t . a s p ? c o n t e n t _ i d =40142; and Stojaković, Dušan. 2007. “Interview: Miroslav Lajčák, High Representative/EU Special Represen- tative for BiH, ‘I Have Not Come Here to Divide.’ ” Vecernje Novosti, 17 July. http:// w w w . o h r . i n t / o h r - d e p t / p r e s s o / p r e s s i / d e f a u l t . a s p ? c o n t e n t _ i d =40221 . 97. Parish, Matthew. 2009. “Lajčák Resigns: What Next for Bosnia?” Oslobodjenje, 31 January. http:// www.matthewparish.com/pdf/Lajčákresigns.pdf . 98. Balkan Insight. 2009, 2 February. “Western Bodies in Bosnia ‘Dead Horse’— L a j č á k . ” h t t p : / / w w w . b a l k a n i n s i g h t . c o m / e n / p a g e / a l l - b a l k a n s - h o m e . 99. “Interview: Valentin Inzko, EU Special Representative and High Representa- tive in BiH.” 2009. Vecernje Novosti, 8 April. http:// www .ohr .int /ohr -dept /presso /pressi /default .asp ?content _id=43382. 100. Eu ro pe an Commission. 2008, 10 November. Summary Note on the Joint Re- port “EU’s Policy in Bosnia and Herzegovina: Th e Way Ahead” by Javier Solana, EU High Representative for the CESP and Olli Rehn, EU Commissioner for Enlargement. h t t p : / / e u r o p a . e u / r a p i d / p r e s s - r e l e a s e _ I P - 0 8 - 1 6 7 2 _ e n . h t m . 101. Interviews with OHR offi cials, Sarajevo, May 2008 and May 2011. 102. Interview with OHR offi cial, Sarajevo, May 2011. 103. Waterfi eld, Bruno. 2011. “Bloodshed to Return to Bosnia, Paddy Ashdown Fears.” Daily Telegraph, 27 May. http:// www .telegraph .co .uk /news /worldnews /europe / s e r b i a / 8 5 4 1 5 7 8 / B l o o d s h e d - t o - r e t u r n - t o - B o s n i a - P a d d y - A s h d o w n - f e a r s . h t m l . 104. Interview with SDA politician, Sarajevo, May 2008. 105. Interview with SBiH politician, Sarajevo, May 2008. 106. Interview with OHR offi cial, Sarajevo, May 2011. 107. Peace Implementation Council, “Communiqué of the Steering Board.” 108. Interview with international NGO, Sarajevo, May 2011. 109. Ibid. 110. Interview with EU Delegation to Bosnia and Herzegovina, Sarajevo, May 2011. 111. International Crisis Group, Ensuring Bosnia’s Future, 28. 112. International Crisis Group, Bosnia’s Incomplete Transition, 22. 113. Council of the Euro pe an Union. 2011, 22 March. “Council Decision 2011/173/ CFSP of 21 March 2011 Concerning Restrictive Measures in View of the Situation in Bosnia and Herzegovina.” Offi cial Journal of the Eu ro pe an Union, L 76/68. http:// eur -lex.europa .eu /LexUriServ /LexUriServ .do ?uri =OJ:L:2011:076:0068:0071:EN:PDF .

Chapter 7

1. Engström, Jenny. 2009. Democratisation and the Prevention of Violent Confl ict: Lessons Learned from Bulgaria and Macedonia. Surrey: Ashgate, 110. 2. Ibid. Notes to Pages 160–165 259

3. Ibid., 112. 4. Tschentscher, Axel, ed. and trans. 1991. Republic of Macedonia Constitution. http:// www .servat .unibe .ch /icl /mk00000 _ .html. Th e reference to “nationality” (narod- nosti) derives from the 1974 constitution under the former Socialist Federal Republic of Yugo slavia, denoting groups not accorded the same rights as constituent peoples (narod). 5. Engström, Democratisation, 113. 6. Rossos, Andrew. 2009. Macedonia and the Macedonians: A History. Stanford: Hoover Institution Press, 273. 7. Engström, Democratisation, 115. 8. Rossos, Macedonia and the Macedonians, 273. 9. International Crisis Group. 1998, 9 October. 1998 Elections in Macedonia. Bal- kans Report No. 45. 10. In 1994, a split occurred within the PDP, leading to the formation of the Party for Demo cratic Prosperity of Albanians (PDPA). In 1997, the PDPA and the small People’s Demo cratic Party (NDP) merged to form the Demo cratic Party of Albanians (DPA). 11. International Crisis Group, 1998 Elections in Macedonia, 10. 12. Friedman, Eben. 2009. “Th e Ethnopolitics of Territorial Division in the Repub- lic of Macedonia.” Ethnopolitics 8 (2): 209– 21, 215. 13. See Republic of Macedonia. 1997. Law on the Use of Flags through Which the National Minorities in the Republic of Macedonia Are Expressing Th eir Identity and National Characteristics. http:// www .minorityrights .org /?lid=6401& tmpl =printpage. 14. Cited in International Crisis Group. 1999, 8 January. Macedonia: “New Faces in Skopje.” Balkans Report No. 51. 15. Ibid., 12. 16. Th e amnesty legislation related to Albanian politicians arrested and impris- oned in 1997 for encouraging their community to fl y the Albanian fl ag in front of public buildings. See International Crisis Group. 2000, 2 August. Macedonia’s Ethnic Albanians: Bridging the Gulf. Balkans Report No. 98, 16. 17. International Crisis Group, New Faces in Skopje, 24. 18. Engström, Democratisation, 115. 19. Euro pe an Commission. 2004. “Council and Commission Decision of 23 Febru- ary 2004 Concerning the Conclusion of the Stabilisation and Association Agreement between the Euro pe an Communities and Th eir Member States, of the One Part, and the Former Yugoslav Republic of Macedonia, of the Other Part.” Offi cial Journal of the E ur op ea n U n i o n 47: 1– 2. http:// eur -lex .europa .eu /JOHtml .do ?uri=OJ:L:2004:084:SOM: EN:HTML . 20. Bellamy, Alex J. 2002. “Th e New Wolves at the Door: Confl ict in Macedonia.” Civil Wars 5 (1): 117– 44, 129. 21. Engström, Democratisation, 116. 22. International Crisis Group, New Faces in Skopje, 26. 260 Notes to Pages 166–169

23. Daft ary, Farimah, and Eben Friedman. 2008. “Power-Sharing in Macedonia?” In Settling Self-Determination Disputes: Complex Power-Sharing in Th eory and Practice, edited by Marc Weller and Barbara Metzger. Leiden and Boston: Martinus Nijhoff , 272. 24. Bellamy, “Th e New Wolves at the Door,” 131. See also Hislope, Robert. 2003. “Between a Bad Peace and a Good War: Insights and Lessons from the Almost-War in Macedonia.” Ethnic and Racial Studies 26 (1): 129– 51. 25. Ibid. 26. Daft ary and Friedman, “Power- Sharing in Macedonia?” 266. 27. Ibid., 268. 28. Phillips, John. 2004. Macedonia: Warlords and Rebels in the Balkans. London: I. B. Tauris, 80– 81. 29. See Chivvis, Christopher S. 2008. “Th e Making of Macedonia.” Survival 50 (2): 141–62, 143–44. 30. Th ere was a UN preventive deployment in Macedonia for much of the 1990s. Th e operation was terminated in 1999 following China’s veto against extending the mandate. See Stamnes, Eli. 2004. “Critical Security Studies and the United Nations Preventive Deployment in Macedonia.” International Peacekeeping 11(1): 161– 81. 31. Daft ary and Friedman, “Power-Sharing in Macedonia?” 275. See also Chivvis, “Th e Making of Macedonia.” 32. Phillips, Warlords and Rebels in the Balkans, 118. 33. Frowick was criticized by the Macedonian parties and by international organizations for facilitating direct contact between the NLA and the Albanian parties. See Chivvis, “Th e Making of Macedonia,” 146. 34. Phillips, Warlords and Rebels in the Balkans, 122– 23. 35. See ibid., 117– 36. 36. Th e ICG notes that although the NLA was not represented in the peace talks, “their viewpoints were assumed to infl uence the positions taken by the ethnic Alba- nian party leaders.” See International Crisis Group. 2001, 15 August. Macedonia: War on Hold. Balkans Briefi ng Paper, 4 fn.15. 37. International Crisis Group. 2002, 15 November. Moving Macedonia Toward Self- Suffi ciency: A New Security Approach for NATO and the EU. Balkans Report No. 135. 38. Cited in Phillips, Warlords and Rebels in the Balkans, 131. 39. Badinter had previously led the EC Arbitration Commission of the Peace Con- ference on the former Yugo slavia. 40. International Crisis Group, Macedonia: War on Hold, 4. 41. Cited in Popetrevski, Vasko, and Veton Latifi . 2004. Th e Ohrid Framework Agreement Negotiations. In Th e 2001 Confl ict in FYROM—Refl ections, 29–36. London: Confl ict Studies Research Centre. http:// www .da .mod .uk /colleges /arag /document -listings /balkan /csrc _mpf -2004 -07 -22 /04(15) -Chap3 -JP .pdf . 42. Ohrid Framework Agreement. 2001, 13 August. http:// www.ucd.ie/ibis/fi lestore /Ohrid%20Framework%20Agreement.pdfhttp://faq.macedonia.org/politics/frame work_agreement.pdf . Notes to Pages 170–174 261

43. International Crisis Group, Moving Macedonia Toward Self- Suffi ciency. 44. Phillips, Warlords and Rebels in the Balkans, 144. 45. In March 2003, the EU took over from NATO with the launch of its military operation Concordia, which ended in December 2003. Its mandate was to help secure an environment allowing the implementation of the OFA. See Mace, Catriona. 2002. “Operation Concordia: Developing a ‘Eu ro pe an’ Approach to Crisis Management?” International Peacekeeping 11 (3): 474– 90; Rodt, Annemarie Peen, and Stefan Wolff . 2012. “Euro pe an Confl ict Management in the Western Balkans”Civil Wars 14 (3): 414– 30. 46. O’Leary, Brendan. 2005. “Debating Consociational Politics: Normative and Ex- planatory Arguments.” In From Power Sharing to Democracy: Post- Confl ict Institutions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s Univer- sity Press. 47. Allison McCulloch, however, points out the existence of a corporate consociational rule in Macedonian power sharing: Th e Committee on Inter- Ethnic Relations, which stipulates which groups have members and in what proportion (seven Macedo- nians, seven Albanians, and fi ve members from among the Turks, Vlachs, Romanies, and other minorities). See McCulloch, Allison. 2012. “Consociational Settlements in Deeply Divided Societies: Th e Liberal- Corporate Distinction.” Demo cratization, 1– 18. doi: 10.1080/13510347.2012.748039. 48. Under the OFA, the parties committed to establishing the Committee on Inter-Community Relations, consisting of seven members each from the Macedonian and Albanian parliamentary groups and fi ve members from among the Turks, Vlachs, Romanies, and two other communities. 49. For discussion of Calhoun’s ideas, see Baskin, Darryl. 1969. “Th e Pluralist Vi- sion of John C. Calhoun.” Polity 2 (1): 49, 65; Kateb, George. 1969. “Th e Majority Prin- ciple: Calhoun and His Antecedents.” Po liti cal Science Quarterly 84 (4): 583– 605. 50. Ohrid Framework Agreement, Section 5.1. 51. Bieber, Florian. 2005. “Power Sharing Aft er Yugoslavia: Functionality and Dysfunctionality of Power-Sharing Institutions in Post-War Bosnia, Macedonia, and Kosovo.” In From Power Sharing to Democracy: Post- Confl ict Institutions in Ethnically Divided Societies, edited by Sid Noel. Montreal: McGill-Queen’s Univer- sity Press, 97. 52. Bieber, Florian, and Sören Keil. 2009. “Power- Sharing Revisited: Lessons Learned in the Balkans?” Review of Central and East Eu ro pe an Law 34 (4): 337– 60. 53. Bieber, Florian. 2005. “Partial Implementation, Partial Success: Th e Case of Macedonia.” In Power Sharing: New Challenges for Divided Societies, edited by Ian O’Flynn and David Russell. London: Pluto Press, 118. 54. Ibid., 119. 55. Interviews with DUI politicians, Skopje, June 2010. 56. Interviews with politicians from VMRO-DPMNE and DUI, Skopje, Macedo- nia, June 2010. 262 Notes to Pages 174–179

57. Ilievski, Zoran. 2008. Between Consociational and Integrative Power-Sharing: Th e Case of Macedonia. Paper presented at Association for the Study of Nationalities Conference, New York, 30. 58. Th e literature treats such vote pooling as centripetalist, following the work of politi cal scientist Donald Horowitz. See O’Leary, Brendan. 2013. “Power Sharing in Deeply Divided Places: An Advocate’s Introduction.” In Power Sharing in Deeply Di- vided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 59. Ilievski, “Between Consociational,” 34. 60. Daft ary, Farimah. 2001. “Confl ict Resolution in FYR Macedonia: Power- Sharing or the ‘Civic Approach’?” Helsinki Monitor 12 (4): 291–312; Bieber, “Power Sharing Aft er Yugo slavia.” 61. Daft ary, “Confl ict Resolution in FYR Macedonia,” 304. 62. Bieber, “Partial Implementation,” 109. 63. Ilievski, Zoran, and Dane Taleski. 2009. “Was the EU’s Role in Confl ict Man- agement in Macedonia a Success?” Ethnopolitics 8 (3): 355– 67, 359. 64. Bieber, “Partial Implementation,” 120. 65. International Crisis Group, Moving Macedonia Toward Self- Suffi ciency, 1. 66. Ibid., i. 67. Following President Boris Trajkovski’s death in an airplane crash in 2004, Crvenkosvki was elected president in April 2004 and Hari Kostov became prime minister. 68. International Crisis Group. 2003, 23 October. Macedonia: No Room for Com- placency. Eu rope Report No. 149, 26. 69. Cited in ibid., 26 n.162. 70. Cited in ibid., 27. 71. Ibid., 29. 72. International Crisis Group. 2004, 3 August. Macedonia: Make or Break. Europe Briefi ng No. 33, 5. 73. Ibid., 6. 74. Cited in International Crisis Group. 2005, 25 February. Macedonia: Not Out of the Woods Yet. Eu rope Briefi ng No. 37, 3. 75. Cited in ibid., 6. 76. Ibid. Kostov was replaced by the SDSM’s Vlado Buckovski. 77. Th e EU Police Mission Proxima expired in December 2005 and was replaced by the smaller EU Police Advisory Team (EUPAT). 78. Euro pe an Commission. 2005, 9 November. Communication from the Com- mission: Commission Opinion on the Application from the Former Yugo slav Republic of Macedonia for Membership of the Eu ro pe an Union. COM(2005)562 fi nal. http:// eur -lex .europa.eu /LexUriServ /LexUriServ .do ?uri =COM:2005:0562:FIN:EN:PDF. 79. In 1997, four ethnic Albanian politicians were arrested for raising the Albanian fl ag in front of town halls in Tetovo and Gostivar, leading to outbreaks of violence. Notes to Pages 179–185 263

80. Project on Ethnic Relations. 2003. Macedonia’s Interethnic Coali tion: Th e First Six Months, 6 . h t t p : / / w w w . p e r - u s a . o r g . 81. Ibid., 6– 7. 82. Ibid., 7. 83. Ibid., 10. 84. To some extent, bad relations within the SDSM- LDP- DUI executive derived from personality clashes. A frequently cited example relates to the lack of communication within the Ministry of Interior owing to bad relations and a lack of commu nication between the Macedonian minister (the SDSM’s Ljubomir Mihajlovski) and Albanian deputy minister (DUI’s Fatmir Dehari). See Project on Ethnic Relations. 2005. Mace- donia: On the Road to Brussels, 13. http:// www .per -usa .org; International Crisis Group. 2006, 12 January. Macedonia: Wobbling toward Eu rope. Eu rope Briefi ng No. 41, 7. 85. International Crisis Group, Macedonia: Make or Break.

Chapter 8

1. “Violence Mars Macedonia Campaign.” 2006. BBC News Online, 26 June. http:// news .bbc .co .uk /2 /hi /europe /5118886 .stm . 2. “Polls Close in Macedonian Vote.” 2006. BBC News Online, 5 July. http:// news .bbc.co .uk /2 /hi /europe /5148972 .stm . 3. Nikolovski, Zoran. 2006. “Macedonia Gets New Government, DUI Uses Militant Rhetoric.” S o u t h e a s t E ur op ea n T i m e s , 28 August. http://www .setimes .com . 4. See Project on Inter- Ethnic Relations. 2006, November. Macedonia’s Euro- Atlantic Integration: Advancing Common Interests. h t t p : / / w w w . p e r - u s a . o r g . 5. Interviews with DUI politicians, Skopje, June 2010. 6. Interview with DUI politician, Skopje, June 2010. 7. Interview with VMRO- DPMNE politicians, Skopje, June 2010. 8. Interview with EU delegation to FYR Macedonia, Skopje, June 2010. 9. Ibid. 10. Nikolovski, Zoran. 2007. “Politi cal Crisis in Macedonia as DUI Leaves Parlia- ment.” Southeast Eu ro pe an Times, 30 January. http:// www .setimes .com . 11. Lungescu, Oana. 2007. “EU Warns Macedonia over Entry.” BBC News Online, 9 F e b r u a r y . h t t p : / / n e w s . b b c . c o . u k / 2 / h i / e u r o p e / 6 3 4 4 9 0 7 . s t m . 12. Ilievski, Zoran, and Dane Taleski. 2009. “Was the EU’s Role in Confl ict Man- agement in Macedonia a Success?” Ethnopolitics 8 (3): 355– 67, 363. 13. Macedonian Information Centre. 2007, 29 May. “EU and U.S. Welcome Po litical Dialogue Agreement.” http:// micnews .com .mk /node /9215 . 14. Stojanvska, Marina. 2007. “DUI Returns to Macedonian Parliament, DPA Remains in Government.” S o u t h e a s t E ur op ea n T i m e s , 3 0 M a y . h t t p : / / w w w . s e t i m e s . c o m . 15. “One Dead in Macedonia Pre- Election Violence.” 2008. Balkan Insight, 12 May; “Macedonian Poll Marred by Unrest.” BBC News Online, 1 June. http:// news .bbc .co.uk /2 /hi /europe /7429712 .stm . 264 Notes to Pages 185–188

16. Stojanvska, Marina. 2008. “OSCE, ODIHR Say Elections in Macedonia Were Substandard.” S o u t h e a s t E ur op ea n T i m e s , 4 June. http:// www .setimes .com . 17. “Polls Close in Macedonia Poll Rerun.” 2008. Balkan Insight, 15 June. 18. Trajkov, Goran. 2008. “VMRO-DPMNE and DUI Form a Ruling Coali tion in Macedonia.” S o u t h e a s t E ur op ea n T i m e s , 8 July. http:// www .setimes .com . 19. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. Th e offi - cial also stressed that although the EU has emphasized the need for politi cal stability, they have not dictated which parties should be in a ruling co ali tion. 20. Taleski, Misko. 2011. “Macedonia No Stranger to Politi cal Th eatre.” Southeast E ur op ea n T i m e s , 25 February. http:// www .setimes .com . 21. Zoran Zaev, Mayor of Strumica, and three offi cials were arrested on allegations of abuse of offi ce relating to the tender for a local shopping center. See “Macedonia Op- position in Boycott amid Arrest.” 2008. Balkan Insight, 17 July. 22. Southeast Euro pe an Times, “Macedonia’s DPA to Boycott Parliament.” 2009. S o u t h e a s t E ur op ea n T i m e s , 11 August. http:// www .setimes .com . 23. “Macedonia’s DPA to Return to Parliament.” 2011. S o u t h e a s t E ur op ea n T i m e s , 2 7 M a r c h . h t t p : / / w w w . s e t i m e s . c o m . 24. “Macedonia Opposition Announces Parliament Boycott.” 2011. Balkan In- sight, 28 January. SDSM was joined in its boycott by the smaller parties, the New Social Demo crats, the New Alternative Party, and the Liberal Party. Th e day before, the TV station A1 and three newspapers from the same media group accused the government of orchestrating the decision to freeze their bank accounts. Th e owner of A1, Velij Aramkovski, and sixteen employees were arrested in December for tax evasion. 25. “Macedonia’s Gruevski, Crvenkovski Fail to Reach Agreement.” 2011. South- e a s t E ur op ea n T i m e s , 1 6 M a r c h . h t t p : / / w w w . s e t i m e s . c o m . 26. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. 27. “Macedonian MPs to Pay Dear for Boycotts.” 2011. Balkan Insight, 7 April. 28. Th e VMRO- DPMNE co ali tion won fi ft y-six seats, the SDSM-led coali tion forty- two seats; DUI won fi ft een seats and DPA eight. 29. International Crisis Group. 2011, 11 August. Macedonia: Ten Years Aft er the Confl ict. Eu rope Report No. 212, 6. 30. Burg, Steven. 2007. Macedonia’s Euro- Atlantic Integration: Advancing Common Interests. P r i n c e t o n , N . J . : P r o j e c t o n I n t e r - E t h n i c R e l a t i o n s , 6 . h t t p : / / w w w . p e r - u s a . o r g /Reports/Mavrovo %20Final %20Report %20Nov %202006 .pdf . 31. Interview with DUI politician, Skopje, June 2010. 32. Interview with DUI politician, Skopje, June 2010. 33. Euro pe an Commission. 2010, 9 November. Th e Former Yugo slav Republic of Macedonia 2010 Progress Report. COM(2010)660, 10– 11. http:// ec .europa .eu /enlarge m e n t / p d f / k e y _ d o c u m e n t s / 2 0 1 0 / p a c k a g e / m k _ r a p p o r t _ 2 0 1 0 _ e n . p d f . 34. Interviews with DUI politicians, Skopje, June 2010. Notes to Pages 188–193 265

35. Interview with VMRO- DPMNE politician, Skopje, June 2010. 36. Euro pe an Commission. 2008, 5 November. Th e Former Yugoslav Republic of Macedonia 2008 Progress Report. COM(2008)674, 19. http:// ec .europa .eu /enlargement /pdf /press _corner /key -documents /reports _nov _2008 /the _former _yugoslav _republic _of_macedonia _progress _report _en .pdf . 37. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, 16. 38. Eu ro pe an Commission, Macedonia 2010 Progress Report. 39. Interviews with DUI politicians, Skopje, June 2010. 40. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict 19. 41. Interviews with DUI politicians, Skopje, June 2010. 42. Interview with DUI politician, Skopje, June 2010. 43. Euro pe an Commission, Macedonia 2010 Progress Report. 44. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, 19– 20. 45. Interview with Sasho Ordanovski, ALSAT TV, Skopje, June 2010. 46. Euro pe an Commission, Macedonia 2010 Progress Report, 9. 47. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. 48. Ibid. 49. “Albanians Flex More Muscle in Macedonia’s New Government.” 2011. Balkan Insight, 4 July. 50. Interview with SDSM politician, Skopje, June 2010. 51. Interview with DUI politician, Skopje, June 2010. 52. Euro pe an Commission. 2009, 14 October. Th e Former Yugoslav Republic of Macedonia 2009 Progress Report. COM(2009)553, 7– 8. 53. Euro pe an Commission, Macedonia 2010 Progress Report, 7. 54. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, 14 n.146. 55. Phillips, David L. 2011. “Macedonia Must Recover the Spirit of Ohrid.” Balkan Insight, 6 July. 56. Ibid. 57. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. 58. Ibid. 59. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, 20. 60. Ibid. 61. Interview with DPA politician, Skopje, June 2010. 62. Ross, Marc Howard. 2009. “Cultural Contestation and the Symbolic Land- scape: Politics by Other Means?” In Culture and Belonging in Divided Societies, edited by Marc Howard Ross. Philadelphia: University of Pennsylvania Press, 1. 63. “Macedonian Encyclopaedia to Be Changed.” 2009. Balkan Insight, 23 Sep- tember. 64. “Macedonian Albanians Prepare Protests.” 2009. Balkan Insight, 28 September. 65. “Macedonia’s DUI Considers Leaving Government.” 2009. Balkan Insight, 29 September. 266 Notes to Pages 193–197

66. “Macedonia Blames ‘Dark Forces’ for Encyclopaedia.” 2009. Balkan Insight, 28 September. 67. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. 68. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, i. 69. “Critics Lash ‘Dated’ Aesthetics of Skopje 2014.” 2010. Balkan Insight, 24 June. 70. “Ghosts of the Past Endanger Macedonia’s Future.” 2009. Balkan Insight, 27 October. 71. Interview with DPA politician, Skopje, June 2010. 72. Interview with journalist, ethnic Albanian daily newspaper Koha, Skopje, June 2010. 73. “Skopje 2014—A Recipe for Urban Disintegration.” 2010. Balkan Insight, 16 February. 74. Th e name dispute with Greece is an important backdrop and context to these cultural and identity issues, making them ever more complicated. Certainly, Macedo- nians feel that Greece’s opposition to the “Republic of Macedonia” threatens their national identity. Macedonian elites seek to cement their nationhood, which in turn aff ects relations with the Albanian community. Th e state also sees its identity threatened by Bulgaria, who considers Macedonians to be ethnic Bulgarians and the Mace- donian language to be a dialect of Bulgarian. 75. Interviews with international offi cials, Skopje, June 2010. 76. Interview with international offi cial, Skopje, June 2010. 77. International Crisis Group, Macedonia: Ten Years Aft er the Confl ict, 2. 78. Interview with SDSM politician, Skopje, June 2010. 79. Th e SAA entered into force in April 2004. Macedonia was the fi rst country in which SAA negotiations were concluded. 80. Council of the Eu ro pe an Union. 2007, 15 February. Council Joint Action 2007/109/CFSP Amending and Extending the Mandate of the Eu ro pe an Union Special Representative in the Former Yugoslav Republic of Macedonia. http:// eur -lex .europa .eu /LexUriServ /site /en /oj /2007 /l _046 /l _04620070216en00680070 .pdf . 81. Euro pe an Commission. 2005, 9 November. Communication from the Commis- sion: Commission Opinion on the Application from the Former Yugo slav Republic of Macedonia for Membership of the Euro pe an Union . COM(2005)562 fi nal. http:// eur -lex.europa .eu /LexUriServ /LexUriServ .do ?uri =COM:2005:0562:FIN:EN:PDF. 82. Ibid., 6. 83. Council of the Eu ro pe an Union. 2006. Brussels Euro pe an Council 15/16 De- cember 2005 Presidency Conclusions. http:// www .consilium .europa .eu /ueDocs /cms _Data/docs /pressData /en /ec /87642 .pdf . 84. Council of the Euro pe an Union. 2008. 2008/212/EC: Council Decision of 18 February 2008 on the Principles, Priorities and Conditions Contained in the Accession Partnership with the Former Yugo slav Republic of Macedonia and Repealing Decision 2006/57/EC. http:// eur -lex .europa .eu /LexUriServ /LexUriServ .do ?uri =OJ:L:2008:080: 0032:01:EN:HTML . Notes to Pages 197–202 267

85. Euro pe an Commission. 2008, 5 March. Communication from the Commission to the Euro pe an Parliament and the Council, Western Balkans: Enhancing the Euro pe an Perspective. COM(2007)127 fi nal. http:// ec .europa .eu /enlargement /balkans _communi cation /index _en .htm . 86. “Barroso, Fuele Urge Macedonia to Boost Reforms.” 2011. Balkan Insight, 11 April. 87. Interview with po liti cal adviser to the EUSR in Macedonia, Skopje, June 2010. 88. Euro pe an Commission. 2006, 8 November. Th e Former Yugoslav Republic of Macedonia 2006 Progress Report. COM(2006)649 fi nal, 7. 89. Euro pe an Commission. 2007, 6 November. Th e Former Yugo slav Republic of Macedonia 2007 Progress Report. COM(2007)663 fi nal, 7– 10. 90. Eu ro pe an Commission. 2008. Key Findings in Progress Report 2008: Th e For- mer Yugoslav Republic of Macedonia. http:// ec .europa .eu /enlargement /press _corner /key-documents /reports _nov _2008 _en .htm . 91. Euro pe an Commission, Macedonia 2010 Progress Report. 92. Interview with EU Delegation to Former Yugoslav Republic of Macedonia, Skopje, June 2010. 93. “Brussels Losing Patience with Macedonia’s Hardline Leaders.” 2010. Balkan Insight, 17 December. 94. Euro pe an Commission, Macedonia 2007 Progress Report, 7. 95. “Albanians Humble Macedonia’s Ruling Party.” 2007. Balkan Insight, 6 No- vember. 96. Interview with SDSM politician, Skopje, June 2010. 97. Interview with EU Delegation to Former Yugoslav Republic of Macedonia, Skopje, June 2010. 98. Vasilev, George. 2011. “EU Conditionality and Ethnic Coexistence in the Bal- kans: Macedonia and Bosnia in a Comparative Perspective.” Ethnopolitics 10 (1): 51– 76, 54. 99. Ibid., 72 (italics added). 100. Ibid., 58. 101. Ibid., 62. 102. NATO. 2008, 3 April. Bucharest Summit Declaration. http:// www .nato .int /cps /en /natolive /offi c i a l _ t e x t s _ 8 4 4 3 . h t m . 103. “Macedonian PM Vows Name Referendum.” 2009. Balkan Insight, 1 June. 104. “Macedonia President Addresses Name Issue.” 2010. Balkan Insight, 28 April. 105. Ilievski and Taleski, “Was the EU’s Role in Confl ict Management,” 365. 106. Interviews with DUI politicians, Skopje, June 2010. 107. Interview with EU Delegation to FYR Macedonia, Skopje, June 2010. 108. Vasilev, “EU Conditionality and Ethnic Coexistence in the Balkans,” 73. 109. Interview with SDSM politician, Skopje, June 2010. 110. Interview with DUI politician, Skopje, June 2010. 111. Interview with international offi cial, Skopje, June 2010. 268 Notes to Pages 202–207

112. Delegation of the Euro pe an Union to the Former Yugoslav Republic of Mace- donia. 2011, 12 August. “Joint Statement by Heads of Mission of EU, NATO, OSCE, and U.S. in Skopje on Tenth Anniversary of Ohrid Framework Agreement.” Press Re- lease. http:// eeas .europa .eu /delegations /the _former _yugoslav _republic _of _macedonia /press _corner /all _news /news /2011 /20110812 _01 _en .htm .

Conclusion

1. Ostrom, Elinor. 2005. Understanding Institutional Diversity. Princeton, N.J.: Princeton University Press, 3. 2. Hartzell, Caroline, and Matthew Hoddie. 2003. “Institutionalizing Peace: Power Sharing and Post– Civil War Confl ict Management.” American Journal of Politi cal Sci- ence 47 (2): 318– 32, 319. 3. Horowitz, Donald L. 2000. “Constitutional Design: An Oxymoron?” In Design- ing Demo cratic Institutions, edited by Ian Shapiro and Stephen Macedo. New York and London: New York University Press. 4. Ibid. 5. See Godson, Dean. 2004. Himself Alone: David Trimble and the Ordeal of Unionism . London: HarperCollins, 337–39. 6. Bildt, Carl. 1998. Peace Journey: Th e Struggle for Peace in Bosnia. London: Wei- denfeld and Nicolson, 145. 7. Th ere is a vast literature on institutional per sis tence and change. For useful overviews, see Peters, B. Guy. 2005. Institutional Th eory in Po liti cal Science: Th e New Institutionalism. London and New York: Continuum; Th elen, Kathleen. 2002. “How Institutions Evolve: Insights from Comparative-Historical Analysis.” In Comparative Historical Analysis in the Social Sciences, edited by James Mahoney and Dietrich Rue- schemeyer. New York: Cambridge University Press; Hall, Peter A., and Rosemary C. R. Taylor. 1996. “Politi cal Science and the Th ree New Institutionalisms.” Politi cal Studies 44: 936–57. 8. Peters, Institutional Th eory in Po liti cal Science. 9. See Krasner, Stephen D. 1984. “Approaches to the State: Alternative Concep- tions and Historical Dynamics.” Comparative Politics 16: 223– 46; Krasner, Stephen D. 1988. “Sovereignty: An Institutional Perspective.” Comparative Politi cal Studies 21 (1): 66– 94. 10. Peters, Institutional Th eory in Po liti cal Science, 162; For this non- equilibrium argument, see Riker, William H. 1980. “Implications from the Disequilibrium of Major- ity Rule for the Study of Institutions.” American Po liti cal Science Review 74 (2): 432– 46. 11. Mahoney, James, and Kathleen Th elen. 2010. “A Th eory of Gradual Institu- tional Change.” In Explaining Institutional Change: Ambiguity, Agency, and Power, edited by James Mahoney and Kathleen Th elen. Cambridge: Cambridge University Press, 1. 12. Ibid., 4. Notes to Pages 208–216 269

13. Pierson, Paul. 2000. “Th e Limits of Design: Explaining Institutional Origins and Change.” Governance 13 (4): 475– 99. 14. Interviews with EU offi cials, Sarajevo, May 2011. 15. O’Flynn, Ian. 2007. “Divided Societies and Deliberative Democracy.” British Journal of Po liti cal Science 37 (4): 731–51. 16. McGarry, John, and Brendan O’Leary. 2006. “Consociational Th eory, North- ern Ireland’s Confl ict, and Its Agreement. Part 2: What Critics of Consociation Can Learn from Northern Ireland.” Government and Opposition 41 (2): 249– 77, 262. 17. Wilford, Rick. 2010. “Northern Ireland: Th e Politics of Constraint.” Parlia- mentary Aff airs 63 (1): 134– 55. 18. OHR. 2007, 19 October. Decision Enacting the Law on Changes and Amend- ments to the Law on the Council of Ministers of Bosnia and Herzegovina. http:// www . o h r . i n t / d e c i s i o n s / s t a t e m a t t e r s d e c / d e f a u l t . a s p ? c o n t e n t _ i d =40687. 19. Interviews with offi cials from the EU Delegation, U.S. Embassy, and UK Em- bassy, Skopje, June 2010. 20. Attempts to exclude Sinn Féin were unsuccessful because the motions failed to secure the necessary cross- community support. 21. Murphy, Sandra. 2001. “Trimble Suff ers Fresh Blow as Court Rules SF Ban Un- lawful.” Irish In de pen dent Online, 31 January. http:// www .independent .ie /irish -news /trimble -suff ers -fresh -blow -as -court -rules -sf -ban -unlawful -26101018 .html . 22. British and Irish Governments. 2006. Agreement at St. Andrews, para. 3. http:// w w w . n i o . g o v . u k / s t _ a n d r e w s _ a g r e e m e n t . p d f . 23. See Northern Ireland Executive. 2007. Ministerial Code. http:// www.northern ireland.gov.uk/pc1952_ni_exec_min_code.pdf . 24. McGarry, John, and Brendan O’Leary. 2007. “Iraq’s Constitution of 2005: Lib- eral Consociation as Politi cal Prescription.” International Journal of Constitutional Law 5 (4): 670– 98. 25. No more than two-thirds of ministers are to be from the Federation, and deputy ministers are not to be of the same constituent people as their respective minister. 26. Venice Commission. 2006, 7 March. Draft Opinion on Diff erent Proposals for the Election of the Presidency of Bosnia and Herzegovina. h t t p : / / w w w . v e n i c e . c o e . i n t . 27. Interview with Robert McCartney, MLA, Belfast, March 2005. 28. McGarry, John and Brendan O’Leary. 2009. “Under Friendly and Less- Friendly Fire.” In Consociational Th eory: McGarry & O’Leary and the Northern Ire- land Confl ict, edited by Rupert Taylor. London: Routledge, 360. 29. Coakley, John. 2011. “Th e Challenge of Consociation in Northern Ireland.” Parliamentary Aff airs 64 (3): 473– 93, 485. 30. Th e executive was restored when members of the bi-communal Alliance Party redesignated from “other” to “unionist” and secured the vote. 31. Arguably a sign of Sinn Féin’s investment in the power-sharing system, McGuinness later referred to himself as “Joint First Minister.” Although the minister was correct from a legal viewpoint, his omission of “deputy” bothered the DUP, who 270 Notes to Pages 217–221 dismissed McGuinness’s choice of language as “republican speak.” See Gordon, Ga- reth. 2009. “Are All Th ings Created Equal?” BBC News Online, 17 February. http:// n e w s . b b c . c o . u k / 2 / h i / u k _ n e w s / n o r t h e r n _ i r e l a n d / 7 8 9 5 4 7 0 . s t m . 32. Interviews with DUI politicians, Skopje, June 2010. 33. McCulloch, Allison. 2012. “Consociational Settlements in Deeply Divided Societies: Th e Liberal-Corporate Distinction.” Demo cratization. doi: 10.1080/13510347 .2012.748039. 34. Th e DPA accorded veto rights to the constituent peoples and the two entities in the state parliament; the VNI veto in the entities was introduced following a Constitutional Court decision in 2000 and the subsequent imposition of constitutional amendments by the high representative in April 2002. Th e defi nitions of VNI in the constitutions of the two entities include repre sen ta tion of constituent peoples; identity of a constituent people; constitutional amendments; orga ni za tion of public authorities; equal rights of the constituent peoples in decision making; education, religion, language, culture; territorial or ga ni za tion; and the public information system. 35. Venice Commission. Draft Opinion on Diff erent Proposals. 36. Key decisions including the election of the chair of the Assembly, the fi rst minister, and deputy fi rst minister (since revised); standing orders; and the bud get. Otherwise, thirty Assembly delegates can trigger a “petition of concern” for a vote on any other issue, to be taken on a cross- community basis. 37. According to an offi cial working for the Northern Ireland Executive, this provision has been invoked. Unfortunately, though perhaps unsurprisingly, the offi cial was unable to provide additional information on the relevant issues. 38. In June 2008, an Assembly petition referred a decision by Minister of the En- vironment Sammy Wilson on environmental governance back to the Executive. See Northern Ireland Assembly Reports 23 May 2008, 9 June 2008, and 17 June 2008. http:// archive .niassembly .gov .uk /record /hansard _session2007 .htm. In response to the As- sembly petition, the Executive issued a statement that the minister’s decision did not contravene his obligations under Section 28A(1) of the Northern Ireland Act. Th e Ex- ecutive noted the minister’s intention to bring to the Executive for discussion and agreement any matters relating to his decision that require approval under the North- ern Ireland Act and the ministerial code. 39. McDonald, Henry. 2007. “Unionists Will Hold Veto.” Th e Observer, 6 May. 40. See McEvoy, Joanne. 2013. “We Forbid! Th e Mutual Veto and Power-Sharing Democracy.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 41. On the role of external actors in power sharing, see Kerr, Michael. 2005. Im- posing Power- Sharing: Confl ict and Coexistence in Northern Ireland and Lebanon. Dublin: Irish Academic Press; Bieber, Florian. 2013. “Th e Balkans: Th e Promotion of Power-Sharing by Outsiders.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press; Notes to Pages 221–226 271

McEvoy, Joanne. 2014. “Th e Role of External Actors in Incentivizing Post-Confl ict Power- Sharing.” Government and Opposition 49 (1): 47– 69. doi:10.1017/gov.2013.19. 42. Schimmelfennig, Frank. 2005. “Strategic Calculation and International So- cialization.” I n t e r n a t i o n a l O rg an iz at i o n 59: 827– 60. 43. Ibid. 44. Se Kerr, Michael. 2011. Th e Destructors: Th e Story of Northern Ireland’s Lost Peace Pro cess. Dublin: Irish Academic Press. 45. Wilford, “Northern Ireland: Th e Politics of Constraint,” 153. 46. Noutcheva, Gergana. 2009. “Fake, Partial and Imposed Compliance: Th e Lim- its of the EU’s Normative Power in the Western Balkans.” J o u r n a l o f E ur op ea n P u b l i c Policy 16 (7): 1065– 84, 1067. 47. O’Leary, Brendan. 2013. “Power Sharing: An Advocate’s Conclusion.” In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O’Leary. Philadelphia: University of Pennsylvania Press. 48. Bell, Christine. 2008. On the Law of Peace: Peace Agreements and the Lex Pa- cifi catoria. Oxford: Oxford University Press, 194. 49. O’Leary, “Power Sharing: An Advocate’s Conclusion,” 399. This page intentionally left blank INDEX

Accommodation, 19, 180 Calhoun, John C., 171– 72 Adams, Gerry, 66, 87, 99, 101, 245 n.24 Caspersen, Nina, 114 Ahern, Bertie, 33, 52, 75, 83, 84, 87, 91 Čavić, Dragan, 120, 144 Ahern, Dermot, 33 Chandler, David, 113 Ahmeti, Ali, 22, 169–70, 175– 76, 183, 185, Choudhry, Sujit, 19 199, 217 Christopher, Warren, 108 Alliance for Change co ali tion, 22, 49– 50, Coakley, John, 64, 70 121–25, 212 Collective responsibility, 46–47 Alliance Party, 21, 24, 40, 43, 54, 70, 91, Comprehensive Agreement document, 103 91– 92 Alternative Vote, 20 Consociationalism: complete, concurrent, Ashdown, Paddy, 120–21, 124, 125– 29, 144, and weak, 231 n.65, 241 n.76; corporate 151, 153 and liberal consociational rules, 20, 26–27, Ashton, Catherine, 153 171, 217; four principles of, 64; predeter- Aughey, Arthur, 66– 67 mination and self-determination, 26 Constitutional Court of Bosnia and Badinter principle, 19, 171– 72, 184, 186, 192, Herzegovina, 18, 27, 113, 123 203, 209, 220 Cosgrave, Liam, 43 Badinter, Robert, 168 Council of Eu rope: Venice Commission, 24, Barroso, José Manuel, 197 116, 133– 34, 142, 147, 215 Belfast, 1 Council of Ireland, 43, 45, 48–49, 54, Bell, Christine, 53, 60, 225 56, 59 Bellamy, Alex, J., 165 Čović, Dragan, 125, 139 Bieber, Florian, 18, 30– 32, 110– 11, 117, Cowen Brian, 89, 101 172, 174 Cox, Marcus, 115– 16, 118– 19 Bildt, Carl, 108– 9, 116, 140, 206, 248 n.7 Craig, William, 43, 49, 57 Blair, Tony: DUP- Sinn Féin deal, 87; Good Croatian Demo cratic Union (HDZ), 22, 117, Friday Agreement, 52, 62, 75, 83; IRA 124, 136 decommissioning, 244– 45 n.21, 245 n.24; Crvenkovski, Branko, 168, 176, 185 St. Andrews Agreement, 33, 98 Bloomfi eld, Kenneth, 40, 80–81 Daalder, Ivo, 110 Boland, Kevin, 48 Daft ary, Farimah, 166–67, 174 British-Irish Agreement, 51 Dayton Peace Agreement (DPA): a British- Irish Council (BIC), 67, 83, 214 consociational settlement, 113– 15, 215; Brooke, Peter, 65, 71 negotiations, 107– 13; operation of, Brown, Gordon, 101, 225 116– 31, 142– 56; reform of, 18, 133– 142, Bullick, Richard, 93 223; two entities, 56 Butler, Lawrence, 176 De Brun, Bairbre, 80– 81, 99 Butler, Paul, 100 Demo cratic Alliance (DA), 163 274 Index

Demo cratic Party of Albanians (DPA), 24, Fouéré, Erwan, 182, 196 34, 162, 175, 185, 192, 194, 203 Friedman, Eben, 163, 166– 67 Demo cratic Union for Integration (DUI): Frowick, Robert, 167 agreement with VMRO- DPMNE, 34; Fuele, Stefan, 197 co ali tion with SDSM, 22, 175– 76, 180; co ali tion with VMRO- DPMNE; 212; Georgievski, Ljubco, 163, 167– 68 excluded from government, 183– 94, Gillespie, Gordon, 45 198– 99, 217; name issue, 201 Gligorov, Kiro, 162 Democratic Unionist Party (DUP): changes Godson, Dean, 67, 74, 78 to the GFA, 96, 208, 219–20; development Good Friday Agreement (GFA): commit- of the Maze prison, 100–101; operation of ment of British and Irish governments, 53; executive 1999–2002, 78; opposition to the executive rules, 29; operation of, 61– 86; Good Friday Agreement, 68, 213; power Review of the Agreement, 88–94, 207, sharing with Sinn Féin, 6, 17, 33, 56, 87, 245 n.6 104, 211, 222; Review of the Agreement, Goodin, Robert, 14, 16 89–90, 245 n.31; Sunningdale, 43 Greece, 200– 201, 212, 266 n.74 D’Hondt: evolution of power- sharing Gruevski, Nikola, 25, 183– 86, 193, 195, proposals, 13, 68– 69, 71– 73; executive 199– 201, 217 formation, 20, 27– 28, 49, 63, 77, 86, 97–98, 215–18 Hain, Peter, 33, 94, 96– 97 Dixon, Paul, 64 Hartzell, Caroline, 55, 204– 5 Dodds, Nigel, 68, 78, 80 Haughey, Denis, 71, 80, 81 Dodik, Milorad, 121, 138– 39, 141– 46, Hays, Donald, 135, 137 151–52 Heath, Edward, 40, 43, 50, 55 Durkan, Mark, 72, 79, 216 Hislope, Robert, 34 Hitchner, Bruce, 135, 137, 140 Elliott, Sydney, 70 Hoddie, Matthew, 55, 204– 5 Elster, Jon, 16 Holbrooke, Richard, 52, 108, Empey, Reg, 74– 75 248 n.3 Engström, Jenny, 160, 164– 65 Home Rule, 10 Eu ro pe an Commission, 142, 144, 147– 48, Horowitz, Donald, 14– 15, 19, 21, 63– 65, 178, 187– 91, 196– 201 75– 77, 205, 210 Eu ro pe an Court of Human Rights, 116, Hume, John, 74, 82 141– 42, 149, 215 Eu ro pe an Union (EU): in Bosnia, 107, 125, Ilievski, Zoran, 174, 175, 201 128– 33, 137– 38, 140, 148– 49, 152, 155– 56; Integration, 19 in Macedonia, 5, 177–78, 181– 86, Internal Macedonian Revolutionary 196– 203, 224 Movement- Democratic Party for Euro pe an Union Special Representative Macedonian National Unity (VMRO- (EUSR), 133, 152– 54 DPMNE): co ali tion with DPA and Evans, Geoff rey, 69 exclusion of DUI, 34, 183– 84, 216– 17; co ali tion with DUI, 185– 95, 213– 13; Farren, Sean, 91 Government of National Unity, 167; in Farry, Stephen, 93 opposition, 175– 77; Ohrid negotiations, Faulkner, Brian, 42– 48, 51, 54– 55, 60, 222 169; parliamentary boycott, 162 Fitt, Gerry, 47 International Commission on Decommis- FitzGerald, Garret, 54 sioning, 83 Foot and mouth crisis, 82 International Criminal Tribunal for the Ford, David, 91, 103 Former Yugo slavia (ICTY), 125– 26, Foster, Arlene, 102 139, 151 Index 275

International Crisis Group (ICG): on McNamara, Kevin, 70, 240 n.46 Bosnia, 118– 19, 125– 26, 141, 143, 146, Merkel, Angela, 142 150– 51, 154; on Macedonia, 163, 188 Millar, Frank, 71 Inzko, Valentin, 140, 152 Milošević, Slobodan, 108– 9, 248 n.2 Irish Republican Army (IRA): ceasefi re, 70; Mitchell, George, 16, 52, 67 decommissioning, 53, 78, 83– 84, 86, 88, Mitchell, Paul, 69 91, 104, 211; Sunningdale, 48, 64 Murray, Gerard, 66 Ivanov, Georgi, 193, 201 Izetbegović, Alija, 108 National Liberation Army (NLA), 166– 67, 199 Jelavić, Ante, 124 NATO, 34, 125, 133, 140, 168– 71, 192, Kaplan, Robert D., 12 200– 201 Keil, Sören, 111 Nesbitt, Dermot, 80 Kerr, Michael, 31, 43, 48– 50, 54 Nillson, Desirée, 58 Kljusev, Nikola, 162 Norris, Pippa, 7, 13 Knaus, Gerald, 120, 131 Northern Ireland Assembly, 24, 27, 76, 233 Komšić, Željko, 138, 254– 55 n.28 n.94 Kosovo, 165, 167, 202 Northern Ireland Suspension Act (2000), 33, Kosovo Liberation Army (KLA), 166 53, 83, 222 North- South Ministerial Council (NSMC), Lajćák, Miroslav, 139– 40, 144– 45, 151– 52 67, 83, 213– 14 Léotard, Francois, 168, 196 Noutcheva, Gergana, 150 Le Roy, Alain, 176 Liberal Demo cratic Party, 175 O’Duff y, Brendan, 62– 63, 95 Lijphart, Arend, 20, 27, 64, 210 Offi ce of the High Representative (OHR): Long, Naomi, 102 Bonn Powers, 119–20, 126, 132, 150– 52, 156, 223; closure, 107, 153– 54, 223; Macedonian Academy of Arts and Science, critiques of, 33– 34; decision- making, 18, 193, 199 29, 53– 54 Mahoney, James, 207 Offi cial Unionists, 40 Mallon, Seamus, 63, 77– 79, 216 O’Flynn, Ian, 21, 58, 210 Mandelson, Peter, 53, 83– 84, 222 Ohrid Framework Agreement (OFA), 2, 5, 6, Manning, Carrie, 117 18, 34, 54, 159– 203, 208–9 Martin, Felix, 120, 131 O’Leary, Brendan: British and Irish Mason, Roy, 62 governments, 82– 82; complete, concur- Mattes, Michaela, 7 rent and weak consociation, 20; complex Mayhew, Patrick, 65, 71 consociation, 4, 114; corporate and liberal McCausland, Nelson, 98 consociation, 26– 27, 214– 15; d’Hondt, 28, McCulloch, Allison, 217, 261 n.47 70– 71, 77, 216; external actors, 225– 26; McGarry, John: British and Irish govern- Good Friday Agreement, 64– 65; ments, 82– 82; complete, concurrent, and jointness, 25; plurinational federations, weak consociation, 20; corporate and 111; Sunningdale, 211 liberal consociation, 26– 27, 214– 15; Orga ni za tion for Security and Cooperation d’Hondt, 28, 70– 71, 77, 216; jointness, 25; in Eu rope (OSCE) High Commissioner plurinational federations, 111; Sunning- on National Minorities, 164 dale, 211 Ostrom, Elinor, 4, 204 McGimpsey, Michael, 71 Ottomans, 11– 12 McGuinness, Martin, 1, 35, 78, 80, 84, 97– 98, 100– 102, 216, 240– 41 n.63 Paisley, Ian, 43, 49, 56, 78, 97– 98, 104, 216, McIntyre, Anthony, 68 244 n.5 276 Index

Pardew, James, 168 17, 33, 56, 104, 211, 222; support for Party of Demo cratic Action (SDA), 22, 117, policing, 94– 95 124, 146, 153 Skopje 2014, 194– 95, 201– 2, 212 Party for Bosnia and Herzegovina (SBiH), Smith, Jack, 70 121, 136, 146, 153 Social Demo cratic Alliance of Macedonia Party for Demo cratic Prosperity, 162 (SDSM), 22, 162, 167, 180, 185– 86, 199, Patten, Chris, 147 201, 216 Patten Commission, 84 Social Demo cratic and Labour Party Peace Implementation Council (PIC), (SDLP): d’Hondt, 71– 72; Good Friday 33– 34, 119– 20, 129, 137, 139, 142– 43, Agreement, 13, 61, 73, 79, 205; Sunning- 145– 50, 154 dale, 40, 43, 47, 54– 55 Pecelj, Milovan, 125 Socialization, 32– 33 Peters, B. Guy, 207 Söderberg Kovacs, Mimmi, 58 Petritsch, Wolfgang, 120, 124, 151 Solana, Javier, 167 Pierson, Paul, 13, 16– 17, 208 Speller, John, 102 Police Ser vice of Northern Ireland (PSNI), Špirić, Nikola, 22, 143, 145, 213 94– 96 Stabilization and Association Agreement Poots, Edwin, 99– 100 (SAA), 128– 29, 133– 34, 143– 45, 147– 48, Poplasen, Nikola, 121 152, 165, 196– 97, 224 Powell, Jonathan, 67, 75, 85, 91, 94 St. Andrews Agreement, 27, 33, 52, 86– 87, Prior, James, 62 94– 98, 213– 14, 219– 20, 233 n.92 Purdy, Martina, 79 Stedman, Stephen, 57 Steinberg, James, 140 Radmanović, Nebojša, 138 Sunningdale: communiqué, 5, 21, 39, 48, 63, Real IRA, 57, 68 211, 222, 225, 235 n.17; executive, 22, Rehn, Olli, 134, 140, 184 39–60, 64 Reynolds, Andrew, 31 Szewczyk, Bart, 126 Rice, Condoleeza, 135 Robinson, Peter, 1, 78, 100– 102 Taleski, Dane, 175, 201 Rodgers, Brid, 82 Taylor, John, 67 Roeder, Philip, 19 Taylor, Rupert, 31 Rose, Richard, 70 Tetovo University, 164, 176 Ross, Marc Howard, 2, 192 Th açi, Menduh, 192 Rossos, Andrew, 162 Th elen, Kathleen, 207 Rothchild, Donald, 19 Tierney, Stephen, 58– 59 Ruane, Joseph, 61 Tihić, Sulejman, 1, 139 Rycroft , Matthew, 136– 37, 147 Tito, 11– 12 Todd, Jennifer, 61 Savun, Burcu, 7 Tonge, Jonathan, 66 Schimmelfennig, Frank, 33 Trajkovski, Boris, 167– 68, 170 Schwartz- Schilling, Christian, 126, 137, 151 Trimble, David, 35, 53, 66– 67, 71, 77– 78, Sebastián, Sofi a, 114, 137 213– 16 Serb Demo cratic Party (SDS), 22, 117, 124 Tudjman, Franjo, 108, 248 n.2 Silajdžić, Haris, 24, 138, 140, 143 Simeon, Richard, 27 Ulster Unionist Council (UUC), 45, 85 Sinn Féin: development of the Maze prison, Ulster Unionist Party (UUP), 13, 61, 66– 67, 100– 101; eff orts to exclude the party from 205; Review of the Agreement, the Assembly, 78– 79, 213; Good Friday 90–91 Agreement, 65– 66, 68, 83; Irish language Ulster Workers’ Council (UWC) strike, 49, legislation, 99; power sharing with DUP, 52, 57, 60, 64 Index 277

United States, 5, 14, 15, 34, 107– 9, 133, 135, Westendorp, Carlos, 120, 151 137, 168, 177, 181, 184– 85, 203, 206 Whitelaw, William, 26, 40– 44, 50– 51, 222 Van Der Stoel, Max, 164 Wilford, Rick, 27, 64, 103, 211, 223 Vanguard Unionist Party, 43 Williams, Paul, 137 Van Parijs, Philippe, 15 Wilson, Harold, 45, 52– 53, 62 Vasilev, George, 199 Wilson, Robin, 64 Veto rules, 8, 30, 114– 15, 133, 135, 218– 21, Wolff , Stephan, 57, 63, 77, 114 249 n.30, 270 n.34 Xhaferi, Musa, 176 Walker, Graham, 67 Walter, Barbara, 55 Zaev, Zoran, 185, 264 n.21 Weller, Marc, 114 Zahar, Marie- Joelle, 114 This page intentionally left blank A CK N O WL E D GM E N T S

Writing this book took longer than planned. I found it diffi cult to stop writing, seeking to capture the ongoing, contentious nature of power- sharing politics in deeply divided places. Research for the project took several years: at Queen’s University Belfast, as a Sawyer Mellon fellow at the Penn Pro- gram in Ethnic Confl ict at the University of Pennsylvania, and as a lecturer in the Department of Politics and International Relations at the University of Aberdeen. I owe much gratitude to mentors, friends, and colleagues. I wish to thank the University of Pennsylvania Press, especially Peter Agree, for his professionalism and for bearing with me as I carried out additional comparative research in Northern Ireland, Bosnia, and Macedonia. My pro- found thanks go to Professor Rick Wilford at QUB and Professor Brendan O’Leary at the University of Pennsylvania, who gave me the opportunity to spend an intellectually rewarding year in Philadelphia. I gratefully acknowledge the fi nancial assistance from the Andrew W. Mellon Foundation, which supported my research and travel. I also wish to express my gratitude to the many politicians, embassy personnel, and other international offi cials who gave their time for interviews in Belfast, Sarajevo, and Skopje. As a qualitative researcher in the social sciences, I feel privileged to have access to key political actors prepared to share their thoughts about the challenges of building peace in confl ict- torn territories. Closer to home, I am deeply grateful to my parents for their endless support as I continue to return to their loving home in Northern Ireland. Finally, my heartfelt appreciation goes to my husband, Giancarlo, for his unfailing love and encouragement. I dedicate this book to him and to our son, Luca.