Power Sharing in the Western Balkans: A Critical View on the Relationship between Consociationalism, Good Governance, and Political Stability

(1)

Power Sharing in the Western Balkans:

A Critical View on the Relationship between Consociationalism, Good Governance, and Political Stability

FINAL THESIS

---

Faculty of Governance and Global Affairs

A thesis submitted in fulfilment of the requirements for the degree of: MSc in Public Administration: International and European Governance

--- Leila Cornips

S1720821

Date of completion: 28-02-2018

Supervisor: V. Karakasis MSc

(2)

Abstract

The issues regarding ethnic representation and minority rights have been topics of discussion within many political systems, international organisations, and academic debates. Fast changing demographic compositions have led to the increased political mobilisation of minority groups in plural societies, often causing ethnic strife or even civil conflict. The implementation of consociational power sharing as described by Lijphart (1969) has been a commonly used approach to tackle these issues. However, the reality shows that consociationalism as a peacebuilding mechanism often merely provides a short-term solution. Eventually, a recurrence of ethnic tensions, often followed by violent conflict, can be observed in many post-conflict states. This phenomenon has large implications for the international community, as civil conflict in one state can often bring about massive influxes of refugees and spill-overs into another. While the theory as put forward by Lijphart has been extensively criticised for the past decades, a relatively new academic debate has appeared on the role that good governance plays in the success or failure of consociationalism. This research attempts to build on this line of reasoning by empirically examining and exploring the causal links between the concepts of consociationalism, good governance, and political stability. The initial results have been derived from the WGI index, a quantitative tool provided by the World Bank. As this tool proved to be insufficient to adequately answer the research question, an alternative qualitative analysis has been added to the thesis. The aim of this research is twofold; on the one hand, its goal is to highlight and test a new direction of research as an addition to the academic literature on consociationalism and good governance, while, on the other hand it is expected to provide information that could be beneficial for the work of national as well as international institutions and monitoring bodies that are involved in the processes of democratic development in the Western Balkans.

Based on the findings of this research, the conclusion can be drawn that this thesis provides basic evidence that good governance does indeed play an important role in the success or failure of consociationalism as a peacebuilding mechanism. Additionally, the findings also indicate that institutionalised ethnic power sharing can affect good governance, as the consociational design appears to reinforce corruption. Moreover, the intricate structure of consociational systems makes it challenging to monitor and sanction corrupt practices. Further research is recommended to study the potential influence of international actors and developments, as well as path dependency, as an alternative explanation for the political status-quo in both countries. Additionally, the potential causal relationships that need further consideration are the effect of corruption on legitimacy, government effectiveness, and accountability, and the direction of causality between political stability and good governance.

(3)

Acknowledgements

After a period of intense learning, full of academic and personal development, I would like to express my gratitude to the people that stood by me and supported me throughout this process. First and foremost, I would like to thank my family for endlessly and untiringly bearing with me in everything I do. Furthermore, my sincere gratitude goes to Vasileios Karakasis for his guidance, encouragement, and insightful comments during this research process, and for providing me with the feedback that I needed to successfully complete this thesis. Additionally, I would like to thank Professor Doctor Joris Voorhoeve for his highly appreciated input on my initial ideas for this research. Last but not least, I want to thank my friends and colleagues, for their patience, sociability, critical feedback, and dedication during this valuable and rewarding experience.

Declaration of Authorship

I hereby certify that the research as put forward in this thesis has been composed and conducted by me, and is based on my own work, unless referred to as otherwise. I have authored this report independently, and I confirm that all work that has been based on external sources has been duly accredited. All passages which are generally or literally retrieved from these sources have been denoted as such. This thesis, with the title ‘Power Sharing in the Western Balkans: A Critical View on the Relationship between Consociationalism, Good Governance, and Political Stability’, has been developed and submitted in order to obtain the Master of Science degree in Public Administration, with a specialisation in International and European Governance, at Leiden University. The research process of this thesis has been supervised by V. Karakasis.

(4)

List of Tables and Figures

Table 1: An introduction to consociationalism ... 9

Table 2: Operationalisation of indicators ... 23

Table 3: Data on BiH in percentile ranks 1996 - 2016 (World Bank, 2017) ... 28

Table 4: Data on FYROM in percentile ranks 2000 - 2016 (World Bank, 2017) ... 29

Table 5: Qualitative sources used per focusing event ... 36

Figure 1: Overview indicators BiH based on percentile ranks 1996 - 2016 (World Bank, 2017) ... 28

Figure 2: Overview indicators FYROM based on percentile ranks 2000 - 2016 (World Bank, 2017) ... 30

Figure 3: Chain of events in post-conflict BiH ... 37

Figure 4: Public opinion in BiH on the main challenges preventing stability (UN Resident Coordinator Office in Bosnia and Herzegovina, 2015) ... 45

Figure 5: Public opinion in BiH on the elements of society preventing stability (UN Resident Coordinator Office in Bosnia and Herzegovina, 2015) ... 45

Figure 6: Chain of events in post-conflict FYROM... 47 Figure 7: Public opinion on trust in FYROM's public institutions (National Democratic Institute, 2017) 53

(7)

Abbreviations

BiH : Bosnia and Herzegovina

CC : Control of Corruption

DUI : Democratic Union for Integration

EBRD : European Bank for Reconstruction and Development

EFF : Extended Fund Facility

EIU : Economist Intelligence Unit

EU : European Union

FYROM : Former Yugoslav Republic of Macedonia

GE : Government Effectiveness

IMF : International Monetary Fund

MP : Member(s) of Parliament

MSSD : Most Similar Systems Design

NGO : Non-governmental Organisation

OECD : Organisation for Economic Cooperation and Development

PS : Political Stability

RL : Rule of Law

RQ : Regulatory Quality

SDSM : Social Democratic Union of Macedonia

UK : United Kingdom

USA : United States of America

VA : Voice and Accountability

VMRO – DPMNE : The Internal Macedonian Revolutionary Organisation – Democratic Party for Macedonian National Unity

(8)

1

1 Introduction

“The trouble in the Balkans today is not Russian meddling, though there is some of that, but a special case of the malaise afflicting Eastern Europe: unchecked executive power, erosion of the rule of law, xenophobia directed at neighbours and migrants and pervasive economic insecurity.” – International Crisis Group (2017)

1.1 Thematic Overview

Issues regarding minority accommodation and ethnic representation have been a topic of debate in many political systems and international organisations. Fast changing demographic compositions, be they a result of changes in birth rates or widespread international migration, lead to increased political mobilisation of minority groups in plural societies. While this often happens in a peaceful manner and through well-established democratic processes, in many other cases it leads to ethnic strife or even civil conflict. A popular approach to tackle these issues is the implementation of power sharing. Examples of cases that successfully operate or used to operate on a power sharing basis are the Netherlands, Belgium, Switzerland, Lebanon and South Africa, and some would even argue that the European Union (EU) is a system based on the foundations of power sharing (Bogaards & Crepaz, 2002).

In many countries that have faced civil war as a result of escalated ethnic tensions, power sharing has been employed as a peacebuilding tool. The necessary conditions for successful power sharing have extensively been debated upon by many scholars. The most popular approach has been propounded by Lijphart (1969). His theory on consociational democracy has set forth four essential conditions for power sharing to work in ethnically divided societies. Specifically, these are the concepts of a grand coalition, mutual veto, proportionality, and segmental autonomy. According to Lijphart’s hypothesis, when these four conditions are implemented, power sharing will lead to a stable democracy. However, while power sharing - established by means of these four conditions - proves to be an effective mechanism to bring an end to violence and bloodshed, it often merely provides a short-term solution (Cheng & Zaum, 2011). Eventually, it may be after several months or years, ethnic tensions tend to rise again. This presents a problem to the safety and security of the country in question, as well as to the international community as a whole. To give an example, it is in the interest of the EU to maintain stability within its neighbouring countries, to prevent large influxes of migration, speed up processes for accession, and to prevent conflict spill-overs into its own territory.

To illustrate this point, this research will focus on the cases of two of those neighbouring countries, specifically Bosnia and Herzegovina (BiH), and the Former Yugoslav Republic of Macedonia (FYROM).

(9)

2 These two EU candidates share a common history, as both were part of the former Socialist Federal Republic of Yugoslavia before declaring independence in respectively 1992 and 1991 (Bieber, 2013: 133 – 134). Being culturally heterogeneous - BiH consisting of Bosniaks as well as Bosnian Serbs and Croats, and to a smaller extent Jewish and Roma communities, and FYROM having a majority of ethnic Macedonians, yet at least 25% of the population is Albanian - both states have experienced some degree of ethnic conflict in the wake of the Yugoslavian era (ibid.). In both cases, a system of minority inclusion through constitutional power sharing has been implemented as a peace arrangement (ibid.).

In the case of BiH, negotiations took place on the Wright-Patterson Air Force Base in Dayton, Ohio, between the presidents of Serbia, Croatia and BiH, and the Bosnian minister of foreign affairs (Borger, 2015). External parties that were present at the rounds of negotiations were the presidents of France and the United States of America (USA), the British and Russian prime ministers, and the German chancellor (ibid.). The process was coordinated by the USA chief peace negotiator and secretary of state (ibid.). The so called ‘Dayton Peace Accords’ included the partition of the country into two separate entities, specifically the Federation of Bosnia and Herzegovina (largely consisting of Bosniaks and Croats) and the Republika Srpska (mainly Serbs) (ibid.). Annex 4 of the agreement included a reformed constitution in which political rights were specifically allocated to the three ethnic communities (ibid.). The accords were concluded on the 21st_{of November 1995 (ibid.). Even though the new constitutional system brought an end}

to the civil war, political unrest has been commonplace. Poverty, high (youth) unemployment rates, and a seemingly disinterested political elite, have led to increased discontent among the local population. In recent years, many demonstrations and uprisings have shown that BiH is far from being a stable and successful state (Spaic, 2017; Ozturk, 2017; Latal, 2017).

FYROM has followed a fairly similar path. At first, after gaining independence from the Social Federal Republic of Yugoslavia in 1991, the country managed to establish a sovereign and relatively peaceful state, and it was not until 1999 that tensions started to rise between the ethnic Macedonian and Albanian communities (International Crisis Group, 2002). Before the conflict could escalate into a full-blown civil war, a peace deal was signed between the government of FYROM and ethnic Albanian representatives in 2001. The ‘Ohrid Framework Agreement’ included extended rights for the Albanian minority and addressed issues of ethnic representation and the use of the Albanian language within the public administration (International Crisis Group, 2001). Additionally, the Constitution of Macedonia was established to provide the Albanian community with fifteen basic rights (Bajrani, 2012). After this, the state experienced a relatively stable co-existence of ethnic communities for about twenty years. However, in the past few years, political tensions have begun to rise again, and several purported violations of good governance principles

(10)

3 within the public administration have caused riots, violent clashes, and overall, political instability (International Crisis Group, 2002).

1.2 Problem Definition and Research Question

Although the main objective of implementing consociationalism as a peacebuilding mechanism is to create a stable and peaceful state, it has become clear that the cases of BiH and FYROM present a rather different outcome. Consociationalism as a peacebuilding tool tends to provide a short-term solution, in other words a quick end to violent conflict, yet after the implementation of minority accommodation, ethnic tensions often reoccur. This has large implications for security and stability within the countries themselves, as well as for the region and the international community as a whole. Political instability and violence within one state often bring about streams of mass migration, regional spill-overs, and international conflict through proxy actors. The two cases presented in the previous paragraph demonstrate the need to dive further into the necessary conditions that are required to establish political stability within post-conflict consociational systems. The concepts as put forward by Lijphart (1969), a grand coalition, mutual veto, proportionality, and segmental autonomy, appear to be insufficient to reach this intended result. This raises the question: ‘what concepts, conditions, or factors are missing in order to establish political stability?’. Many scholars have attempted to answer this question and a range of criticism of the theory has been put forward in the past decades. However, in the cases of BiH and FYROM, one could recognise a specific phenomenon that seemingly fuels political unrest after the implementation of consociationalism: the absence of good governance. This observation sets forth the following research question:

‘How does the concept of good governance mediate the relationship between consociational power sharing and political stability in post-conflict societies?’

This thesis is devoted to examining the way in which good governance facilitates the relationship between consociationalism and political stability, as well as how one of these concepts affects the other. The academic literature on the role of good governance in consociational systems, as a critique on Lijphart’s consociational democracy theory, will be presented, discussed, and subsequently tested in this thesis.

1.3 Methodology

As set forth in the previous paragraph, the theory that will be tested is the assertion that, in order to establish political stability, it is essential to complement the implementation of consociationalism in post-conflict societies with the foundations for good governance. Therefore, the initial goal of the research, explanatory in its core, is deductive theory testing, employed by means of a small-N comparative research design. To partial out any external effects that can cover up the possible causality, a most similar systems design strategy (MSSD) is applied. The selection of a small number of ‘most similar’ cases, BiH and FYROM,

(11)

4 facilitates the process of drawing solid conclusions. Taking into consideration the limited number of cases, the aim of this research is not to prescribe a causal law that is applicable to all consociational democracies, but rather to open a new line of research in this particular field.

Initially, the decision was made to purely conduct the research through quantitative analysis. The statistical data is derived from the World Bank’s Worldwide Governance Indicators (WGI), through observation of the interactive dataset as provided by the World Bank. The World Bank defines countries’ percentile ranks based on statistical data stemming from several sources. It hereby distinguishes between representative and unrepresentative sources, which is why its rendering of data is expected to be reliable. The purpose of the quantitative analysis is to examine if correlations between the independent variables (voice and accountability, regulatory quality, control of corruption, government effectiveness, and rule of law) and the dependent variable (political stability) can be observed in the statistical data. The reason for adopting a quantitative approach is the density and contextualisation of qualitative data, and the simplicity and efficiency of observation through the World Bank’s WGI.

During the process of data collection and subsequent analysis hereof, it has become clear that the quantitative results were not sufficient to give a good understanding of the situation. Therefore, a supplementary chapter has been added in order to explore and highlight the underlying factors that function as the causal mechanisms between consociationalism and political stability. As an alternative to the quantitative analysis, a qualitative approach was adopted in order to examine the cases in more depth. The qualitative research is conducted through the process tracing method, by analysing specific and carefully selected focusing events in the aftermath of the implementation of consociationalism in both countries.

1.4 Academic and Societal Relevance

The goal of conducting this specific research project is twofold. First, it seeks to contribute to the academic literature on consociationalism, good governance, post-conflict democratic development, peacebuilding, and public administration. Since the introduction of the consociational concept by Lijphart (1969), a broad spectrum of criticism has occurred on the alleged missing links and the factors within the theory which have been downplayed. Examples of these hypothesised links and factors are the external power game, the role of foreign influence, and disproportionate ethnic representation. These concepts have extensively, one could even say exhaustively, been debated upon for over three decades. However, the notion of good governance as a missing link in the theory is relatively new to the consociational debate. Various scholars have hypothesised the causal relationship between consociationalism, good governance, and political stability, yet the empirical research on this supposition remains very limited. By applying this line of reasoning to the comparative case study of BiH and FYROM, and by using a combination of quantitative and qualitative research methods, a unique contribution is made to the academic literature. The fact that an

(12)

5 in-depth and within-case analysis has not yet been conducted on this particular topic, is what sets this thesis apart from other research. Hence, this research adds to consociational theory by setting forth a new perspective on the necessary conditions under which power sharing will lead to political stability, opening a new line of research on this particular causal relationship. As this research is interdisciplinary in its core, it has the potential to be relevant to researchers in public administration, as well as to scholars in the fields of political science, peacebuilding, and (ethnic) conflict research.

Secondly, the research aims to create a better understanding of the political situation in the Western Balkans, a region of foremost significance to the EU, as well as to the broader international community. As mentioned in the previous paragraphs, both BiH and FYROM are candidates for EU accession. It is therefore in the Union’s interest that these states meet the Copenhagen Criteria (most relevant in these countries is the development and implementation of stable institutions, democratic principles, minority-rights, and a functioning market-economy), in order to speed up accession processes. It is of importance for the Union to maintain stability within its neighbouring countries. If the states would regress into violent conflict, this could have dramatic implications for stability within the Union as well. The refugee crisis resulting from the conflict in Syria has already underlined the EU’s struggle to create a firm and unanimous approach to the reception and intake of refugees. Therefore, new influxes of refugees from the Western Balkans could significantly destabilise the political situation within the Union as well (think about increased feelings of nationalism and rising tensions between member states), as its ability to absorb more migration is questionable. Additionally, a return to civil war in the neighbouring states could also present large security concerns for the Union. Ethnic conflict in these countries could spark the mobilisation of proxy actors and therefore further destabilise international affairs (i.e. EU – Russia relations). This highlights the importance for the EU to maintain a political dialogue with and continue their effort to support democratic development within BiH and FYROM.

Moving on from the EU, the International Monetary Fund (IMF) has been a vital partner for the Western Balkan states when it comes to financial support and loan programs (IMF, 2016, 2017). However, in the past years, the policy changes and implementations as required by the IMF have, in several cases, in fact destabilised the political situation within the countries (Spaic, 2017). The main reason for this is that the international institutions often have little power to monitor the political processes within a national public administration, and therefore have little access to the information they need to tailor their policy towards these specific countries. This research could contribute to this by creating a better understanding of the underlying factors that block policy reform in certain areas and that prevent the states from creating stable and transparent government institutions. Therefore, this thesis is expected to provide beneficial information

(13)

6 to international institutions, as well as national and international monitoring bodies that are involved in the processes of democratic development in BiH and FYROM.

So far, the international community has largely focused on the constitutional design in BiH and FYROM in order to establish political stability. It has, however, become clear that the constitutional design that has been put forward by Lijphart (1969), and implemented in the Dayton Peace Accords as well as the Ohrid Framework Agreement, is no blueprint for success and stability in these states. Therefore, this research attempts to contribute to the field of knowledge regarding the shortcomings within the consociational systems in BiH and FYROM. Finally, the relatively unexplored field that is focus of this thesis can serve as a first step towards anticipating on, and tackling the recurring obstacles facing the political spectrum, in order to create long-term political stability in BiH and FYROM.

1.5 A Reader’s Guide

The structure of this thesis starts with an overview of the main principles of consociational democracy theory as set forth by Lijphart (1969). These are applied to the cases of BiH and FYROM in chapter 2. Subsequently, chapter 3 presents the academic debate focused on the role of good governance in power sharing systems as being a missing link within consociational theory, followed by the conceptualisation and operationalisation of the concepts of good governance and political stability. Chapter 4 sets forth the research design and data collection, including the key concepts, measurement, data sources, research approach, the rationale behind the case selection, and possible threats to inference. The findings from the research as conducted based on the initial research design, the quantitative approach, are put forward in chapter 5. As the quantitative method has proven to be insufficient to adequately answer the research question, an alternative approach is adopted in chapter 6, by means of a qualitative method. The final chapter, chapter 7, concludes the research, analyses the limitations to it, and embarks upon the opportunities and implications for further research.

(14)

7

2 Institutionalised Peace: Power Sharing in the Western Balkans

The implementation of power sharing has been one of the most predominant strategies for conflict resolution in ethnically divided states (Binningsbø, 2013). Power sharing has been applied in many post-conflict1_{societies, including the Western Balkan states BiH and FYROM. The mechanism is often}

employed in the form of a peace agreement in societies with histories of civil war and violent ethnic strife. However, opinions vary widely on what the concept of power sharing specifically entails. In this chapter, a constitutional perspective on the power sharing institutions in BiH and FYROM will be presented, by analysing the cases through the lens of consociational theory.

2.1 Introduction to Consociationalism

Power sharing is composed of two components, unambiguously ‘power’ and ‘sharing’. The first component points out that, in order to implement the notion of power sharing, a state, institution or entity needs to exhibit some system of power. This condition arises from the idea that in order to implement power sharing, an institution simply needs to have power to share. Worth mentioning here is that sharing of power should not just be interpreted as dividing the power, tasks, and responsibilities among segments. On the contrary, power sharing indicates close cooperation and consultation among different (ethnic) groups, and it “requires the consent of a broad representation of [factions] in a given system” (Bieber, 2005: 86).

A wide variety of theoretical and practical implications of power sharing can be found in the literature. Some of the authors that have attempted to tackle the concept are McGarry & O’Leary (1993, 1995, 2006), Barry (1975), Lustick (1975, 1979), Kerr (2006) and Horowitz (1985, 2014). Nevertheless, the works on ‘consociational democracy’ by Lijphart (1969, 1977, 1985, 1999) are central. Lijphart was one of the first scholars to introduce the concept of consociational democracy and has often been described as the “father of power sharing” (Binningsbø, 2013: 95). Although he conceded that he was not the first to introduce or describe the concept – several authors preceded him with terms as Proporzdemokratie2 and Konkordanzdemokratie3_{– Lijphart certainly initiated a renewal of interest in models of power sharing in}

divided societies (Andeweg, 2000).

The concept of consociational democracy, as described by Lijphart, has on the one hand a political dimension, signifying the importance of political cooperation among elites, and, on the other hand, a socio-political dimension, indicating a “pluralistic character of the society” (Binningsbø, 2013: 95). More

1_{Definition of post-conflict by Oxford Dictionaries: ‘following or occurring after military conflict’.} 2_{Literal translation: proportional democracy. Described by Lehmbruch, 1967.}

(15)

8 specifically, consociational democracy theory is based on a form of power sharing in states where society is divided in different groups along the lines of ethnicity, linguistics, religion or other racial or ideological affiliations. In these societies, power sharing is established based on proportional representation of different sects, often described in consociational theory as “segmental cleavages” (Sileikaite, 2016; Schöpfer, 2015; Lijphart, 1977: 3). Lijphart (1969) describes how fragmented societies can be stable democracies by pointing to the behaviour of political elites as a crucial variable. These fragmented elites can “engage in competitive behaviour” that would lead to instability and subsequently, civil conflict (Lijphart, 1969: 212). However, Lijphart uses a game-theory approach to explain how these elites often decide to turn away from their competitive inclination, and on the contrary, make efforts to establish some form of overarching cooperation, in order to prevent conflict that would transform into a disadvantageous situation for all stakeholders (ibid.). With the introduction of consociational democracy theory, a first attempt had been made to step away from the belief that political stability is inherent to a homogeneous society.

As Lijphart acknowledges, special attention should be given to the implementation of consociationalism in post-conflict countries. Characteristic for the consociational model of power sharing is that it is often implemented as part of a peace agreement, as in the cases of BiH and FYROM. In BiH, formal power sharing has been implemented under the Dayton Peace Accords in 1995 while FYROM has had a system of power sharing since 1992. Albanian minorities were included in the public administration, although other conditions for consociationalism had not yet been met. Due to the underrepresentation of the Albanians, armed conflict broke out between state security forces and the Albanian National Liberation Army in January 2001. The conflict was concluded in August 2001 by means of the Ohrid Framework Agreement that provided extended rights for the Albanian minority. The Ohrid Framework Agreement transformed FYROM’s institutions into a formal system of power sharing (Bieber, 2013: 133 – 134).

Lijphart underlines four interlinked factors that are essential for the employment of consociational democracy as a power sharing arrangement. Specifically, these are a grand coalition, a mutual veto, proportionality, and segmental autonomy. Though Lijphart “has given these aspects different weights throughout his writings on consociationalism, we can consider all four to be key factors of a power sharing system” (Bieber, 2013: 134; Lijphart, 2004, 2008; McGarry & O’Leary, 2006).

Key Concept Definition Conditions Data source

Consociational Democracy A regime of agreed guarantees, including proportional group participation in government ▪ Grand coalition ▪ Mutual veto ▪ Proportionality ▪ Segmental autonomy ▪ Lijphart (1969, 1977, 1985, 1999)

(16)

9 and minority vetoes of

ethnically sensitive policies.

Table 1: An introduction to consociationalism

2.2 Grand Coalition

A grand coalition is formed when all substantial segments are represented by a leadership figure in the political elite, a contrast to types of democracy with a significant opposition and little majority support (Lijphart, 1977: 25). The proposition “violates” the principle that parliamentary systems should be supported by a majority, but not necessarily an overwhelming one (ibid.: 25 – 26). Lijphart explains the emergence of grand coalitions by pointing out the propositions of Riker’s (1962) ‘size principle’, an argument closely related to game-theory. The argument states that “where players are rational, and where they have perfect information, only minimum winning coalitions occur” (Lijphart, 1977: 62). These conditions for minimum winning coalitions occur solely in homogenous societies where sectarian strife is practically trivial. These claims directly lead to understanding for the other extreme, the grand coalitions, that will emerge in deeply divided societies as players are motivated by divergent ideologies and have ‘little information’. The grand coalition can take many institutional forms, as it can be discovered in parliamentary settings, councils or presidential offices, yet it is not necessarily institutionalised.

The BiH case shows an explicit inclusion of segmental groups in the public administration although the formal requirements for this are limited. It is noteworthy that the BiH central government is relatively small and its powers are limited, as the latter are strongly concentrated in the two constitutional and largely autonomous entities (Republika Srpska and the Federation of BiH). Currently, the Council of Ministers consists of ten members (Bieber, 2013: 135). This number has grown steadily from the three ministers that the country had immediately after the Dayton Agreement (ibid.). Since 2003, the council has been composed of three Serbs, three Bosniaks and three Croats. One deputy minister, the tenth member, is appointed from either one of the three dominant groups or a different national minority community (ibid.). Moreover, both autonomous entities must be represented. The Federation can hold a maximum of two-thirds of all ministerial posts (ibid.). These explicit requirements often lead to a difficult and time-consuming process of government formation. As with the central government, the position of head of state is also designed as a power sharing institution. The presidency is held by three members. One of them is a Serb appointed in the Republic. The other positions are filled by one Bosniak and one Croat, both elected in the Federation. The chair position rotates among the three members on an eight-month basis (ibid.: 135 – 136).

Since the Ohrid Framework of 2001, the inclusion of minorities in the central government has proven to be a developing practice in FYROM. This is mainly a result of the recognition by the national majority that

(17)

10 excluding the Albanian minority would lead to an increase of ethnic tensions and possible relapse into conflict. Additionally, polarisation emerged between the two majority parties, something that made it more likely for these to form a coalition with the Albanian minority than with each other (ibid.: 136). However, the majority parties do cooperate during elections in order to prevent the Albanian political parties from taking over (Bieber, 2013:136, Marusic, 2013). Here it is important to note that the Ohrid Framework Agreement did not include formal requirements regarding the implementation of the grand coalition principle, yet the state “de facto institutionalized this practice” (Bieber, 2013: 136 - 137). FYROM symbolically transformed itself into a ‘bi-national state’ by ceding substantial rights to the Albanian minority. For vital laws to pass, the majority of the MPs, as well as the majority of those that identify as a minority, is needed (ibid.: 137). For this reason, the efficiency of government depends largely on the conformation of a coalition, as through excluding an Albanian coalition partner the government will most probably lack the support of the minority MPs.

2.3 Mutual Veto

A second crucial factor for the establishment of consociational democracy “represents negative minority rule”, the mutual veto (Lijphart, 1977: 36). The mutual veto serves to offer protection to minority segments so that they are not outvoted by their majority counterparts. A major pitfall of the mutual veto is that it can lead to “minority tyranny”, a situation that will challenge cooperation in the coalition to a similar extent as outvoting by a majority (ibid.: 37). In addition, extensive use of the veto can lead to policy deadlock within the consociational administration. However, according to Lijphart, this “danger is not as serious as it appears” as extensive use of the veto will likely be turned against the vetoing party when other parties start to deliberately use the veto themselves (ibid.). The mutual veto can be exercised through an “informal and unwritten understanding” or through formally agreed upon procedures (ibid.: 38).

BiH has two distinct forms of veto rights that can be employed in parliament. These are the entity veto and the veto by constituent people. For all laws to pass, support is needed of “at least one-third of the MPs from each entity in both chambers of parliament (…), as well as the support of a majority of all three constituent peoples in the upper chamber” (Bieber, 2013: 140). This veto by constituent people, or in other words, ‘vital interest clause’ has not been formalised in the constitution, though it is defined in the Constitutional Court’s jurisprudence. When invoked, an ‘ad hoc commission’ is established to resolve the issues. When the commission fails to do so, the Constitutional Court will decide upon the matter (Bosnia and Herzegovina Const. art. IV, 1995). In practice however, the use of the vital interest clause has been rather limited (Bieber, 2013: 140). On the other hand, the entity veto has proven to be a somewhat obstructive mechanism as it has often led to policy deadlock. The veto permits two-thirds per entity to directly block any law from coming into practice. In the past years, representatives from both entities have employed the entity veto to “hijack

(18)

11 the parliament for their exclusionary ethnic interests and this discourages cooperation and compromise” between the ethnic parties (Bahtic-Kunrath, 2011: 899). Examples of this are mainly legislative proposals on cultural matters, the Proposed Law on Decorations, the Proposed Law on Holidays and the Proposed Law on State Anthem (ibid.: 907). Another example is the Proposed Law on Census in BiH that was delayed in parliament for over two years. After 80 sessions, the bill was declared to be failed (ibid.: 911 - 912). Since the implementation of the Dayton constitution, more than half of all decisions were blocked by the entity veto (Bieber, 2013: 140).

In FYROM, laws on issues related to Albanian and other minority communities need the support of a majority of MPs, as well as the majority of these that identify with minority communities. The same principle is in place for laws in areas of far-reaching impact, i.e. constitutional modifications or municipal regulations. These veto arrangements are called double majority (Bieber, 2013: 140; Former Yugoslav Republic of Macedonia Const. art. 69 amend. X, 2001). These mechanisms have proven to be effective in safeguarding the minority approval of policy decisions of wide significance (Bieber, 2013: 141).

2.4 Proportionality

Thirdly, proportionality is presented as an essential principle for the successful employment of consociationalism. The concept of proportionality in consociationalism provides a way to allocate government positions and resources among the political elites and largely prevents discordance issues in the decision-making process (Lijphart, 1977: 39). Through proportionality, “all groups influence a decision in proportion to their numerical strength” (Steiner, 1971). However, Lijphart points out that for decisions with a dichotomous nature, there will remain both winners and losers. The effects hereof can be minimised in two ways, i.e. through reciprocal concessions and delegation of most heavy weighted decisions to a segment summit. In the latter case, efficiency of the negotiation process is maximised and the possibility of deadlock through veto is minimised (Lijphart, 1977: 40).

In the post-war period, the BiH constitution required that “officials appointed to positions in the institutions of Bosnia and Herzegovina shall be generally representative of the peoples of Bosnia and Herzegovina” (Bosnia and Herzegovina Const. art. IX3, 1995). The implementation of the Civil Service Law in 2002 sharpened this regulation by stating that the government should “generally reflect the ethnic structure” (Bieber, 2013: 139). This mechanism is somewhat more flexible than the explicit proportionality in the civil services of the entities as the national system does not specify particular quotas for recruitment. Rather, discrimination of any kind is strongly prohibited in the recruitment process. In practice, the percentage of positions filled by Bosniaks, Serbs, Croats and minorities fairly reflects the national census created in 1991 (ibid.: 139).

(19)

12 The Ohrid Framework Agreement included the principle of “just and equitable representation” in FYROM and implemented specific procedures to incorporate more minorities in its police force (Ohrid Framework Agreement art. 4, 2001). Before these arrangements were established, only seven percent of jobs in the civil service were held by Albanians, while the latter comprised approximately a quarter of the population. With the enactment of the Ohrid Framework Agreement, this number grew rapidly. One enduring issue here is that the access to jobs in the civil service remains limited for the other non-Albanian minorities (Bieber, 2013: 138 – 139).

2.5 Segmental Autonomy

A final condition for consociational democracy is segmental autonomy. Segmental autonomy implies that a certain (ethnic) segment within the political elite has its own exclusive area of interest to rule over. This area will largely consist of members of the segment in case, in other words: divisions are made according to the distinctive subcultural lines. In systems with segmental autonomy, matters of general national interest are decided upon by the political elite, while all other issues and their policy-implementation within a certain autonomous area are responsibility of the segments ruling these. Delegation in this sense, “together with the proportional allocation of government funds to each segment” is presented by Lijphart to be a “powerful stimulus to the various segmental organizations” (1977: 41 – 42). Through segmental authority, the existing subcultural cleavages are overtly recognised and acknowledged, making plural societies more plural in order to turn “segments into constructive elements of stable democracy” (ibid.: 42). Here it is worth noting that Lijphart describes federalism as a special type of consociationalism. Through federalism, the subcultural cleavages accord with regional ones. However, there are some pitfalls to the notion of federalism within consociational democracies. When territorially divided segments are given the autonomy to rule over their designated area, it can lead to demands for higher proportionality and greater autonomy, eventually leading to civil conflict and possibly secession (ibid.: 44). In addition, conflict can occur when “segments are geographically intermingled”, fuelling disputes regarding ‘who has the right to rule’ (ibid.: 45). In these arguments, Lijphart includes the propositions of Nordlinger, who states that “the combination of territorially distinctive segments and federalism's grant of partial autonomy sometimes provides additional impetus to demands for greater autonomy” (Nordlinger, 1965: 32). Nevertheless, both authors leave practical examples out of the scope of their writing.

In fact, BiH exemplifies how federalism can fuel demands for greater autonomy and the domestic tensions this can bring about. In both entities, measures have been taken to prevent the development of ‘a state within a state’. In the sense of territorial self-government, BiH has a great degree of segmental autonomy. The two entities have explicitly formalised borders, legislation, competences and power. Additionally, the entities

(20)

13 are governed under their own constitutions. To uphold the status of ethnoterritorial4_{units, in 2000 the}

Constitutional Court ruled that “the three constituent peoples would have equal status throughout the country” (Bieber, 2013: 142). Specifically, this meant that both the Federation and the Republic needed to diminish all arrangements that advantaged one community (ibid.: 142). Consequently, Serbs and other communities were represented in the Federation, while in the Republic, seats in parliament and government needed to be reserved for Bosniaks, Croats and other communities. This is to weaken the features of a nation-state and to reduce the danger of secession (Bieber, 2006: 123 - 136).

FYROM has a lower degree of segmental autonomy than BiH. In FYROM, the power to rule and legislate remains in the hands of the central government, yet the absence of entities, as in BiH, has given the municipalities greater power to rule. To give an example, municipalities in FYROM can name the local police commander and they have a say in the implementation of health and education policies (Bieber, 2013: 142). However, some protection mechanisms are in place. Decisions regarding unilateral changes of borders, as well as changes in competences and funding of municipalities, cannot simply be implemented by the central government. These decisions and policy implementations need the support of the minorities (ibid.: 142).

2.6 Conclusion from the Application of Consociationalism

It has become clear how consociational theory applies to the case of BiH as well as FYROM. All four conditions of consociational power sharing are to some extent present in the states. However, it can be argued that the theory is highly focused on constitutional design. A clear link between constitutional power sharing as a peace arrangement, and actual stability on the societal level, is missing. The next chapter will therefore highlight new ideas and approaches focused on the conditions under which power sharing can be successful.

4_{English Dictionary definition: ‘relating to ethnoterritory’ – researcher’s own explanation: territories that are} determined based on ethnic features.

(21)

14

3 The Role of Governance

This third chapter will present a critical view on consociationalism, followed by the introduction of a new approach towards successful outcomes of power sharing. Subsequently, the chapter will set forth the main concepts and variables used in the research. Specifically, the concepts of good governance and political stability will be defined, conceptualised, and operationalised in measurable variables by means of the prevailing literature and several widely used measurement tools. The tools that are used to determine these variables are the World Bank’s WGI index and the Economist Intelligence Unit (EIU). During the conceptualisation, operationalisation, and throughout the research, political stability will be deliberately treated as an outcome of good governance, rather than one of its dimensions. The idea behind this decision is that practices of good governance are embedded in the decision-making process, rather than the outcomes of these. Similar approaches have been used by The Millennium Challenge Corporation and the OECD’s ‘Governance at Glance’ index (Mazzarella, 2009; Erkkilä & Piironen, 2014). This theoretical overview will build the framework for the quantitative as well as qualitative data collection and analysis.

3.1 Consociationalism Reviewed: Power Sharing and Good Governance

Since its first appearance in the academic literature, the concept of consociationalism has been subjected to extensive criticism and debate. Norris (2008), a prominent supporter of power sharing, describes some of the unresolved issues surrounding consociationalism. She emphasises that although consociational arrangements have aided peace-settlements, “as exemplified by the Dayton Agreement governing [BiH]”, strong controversy continues to exist around the theory (ibid.: 2008). Norris further states that the widespread criticism of the theory highlights the importance of exploring the underlying driving factors of democratisation processes (ibid.: 2008). Additionally, Norris describes how the theory undermines the importance of the governance of human development as a route to stable democracy. She explains how much emphasis has been put on constitutional design, and that “power sharing arrangements [alone] are not automatically better for encouraging political stability” (Norris, 2005: 20). Norris points out how conditions that are conducive to social development will indirectly lead to political stability and a stable democracy. She also explains how investing in social development by focussing on better and more accessible education, health care and fair trade could stabilise political tensions, as it will reduce the incentives for unrest (Norris, 2005: 20).

A recurring critique on consociationalism is the strong focus on elites and cooperation among these. Andeweg (2000) argues that even if elite cooperation is established, the relationship between cooperation at the elite level and societal collaboration is downplayed in Lijphart’s works. He discusses how little

(22)

15 attention has been given to the exact causal factors that link elite cooperation to political stability, and therefore, the theory proves to be incomplete (Andeweg, 2000: 520). McAuliffe describes this elite-focus as “sheer instrumentalism”, as it presents a rather narrow definition of ‘politics and alliances’ (McAuliffe, 2017: 209; Bendaña, 1996: 76). Hartzell & Hoddie (2003) redefine the concept of power sharing by discussing how, rather than distribution of political power alone, a wider range of governance dimensions should be included in consociational peace-arrangements. This is especially the case when specific segments have been “economically marginalized” during periods of war (ibid.: 320). The authors therefore emphasise the importance of attention for “access to economic resources and the identity of those controlling the levers of economic power” (ibid.).

McAuliffe (2017) further adds to the argument presented by Hartzell & Hoddie by stating that the parties often do not sufficiently address issues regarding the division of wealth or welfare expenditure, and that political players in young democracies implement measures to “reassure security concerns or to distribute political power” rather than focus on socioeconomic rights (McAuliffe, 2017: 199). Interestingly, the latter has often been one of the root-causes of civil conflict in the first place. According to McAuliffe, a breakdown of power sharing may follow when tensions occur in the governance system and the political system cannot “respond to pressing public needs” (ibid.: 205). In such situations, power sharing may lead to a “cold peace” where violence no longer prevails, but at the same time, reconciliation does not ensue (ibid.: 205). Another problem may occur in post-conflict societies when the new ‘political elites’ tend to distribute physical and intellectual resources to “satisfy the major demands of the individuals and groups who can undermine the settlement” rather than the mass of society (ibid.: 212). This problem presents a trade-off between responding to cheaper, though often corrupt, elites or the masses that have too little power to change the status-quo. Though this strategy might tackle the direct threat of a recurrence to violent conflict, it will not result in political stability in the long run.

Cheng & Zaum (2011) present a more in-depth analysis of the latter phenomenon. The authors describe how “divisions within societies affected by conflict weaken shared conceptions of the public good”, causing them to be more prone to poor public governance performance (Cheng & Zaum, 2011: 1). Additionally, these often-weak administrative structures have a limited capacity to monitor the corrupt behaviour of political elites (ibid.). According to the authors, negotiated power sharing settlements actually fuel opportunities for corruption and patronage. As international actors are often pressured to end violent conflict as soon as possible, they often accept that “the opportunities offered by a settlement could be exploited by corrupt actors” who would misuse “economic opportunities provided by government positions” (ibid.: 8 – 9). One example of this is the negotiated deal by Western diplomats to bring an end to the Liberian civil war. This “resulted in the divvying up of cabinet positions between the three warring

(23)

16 factions which effectively gave them a two-year period to reap the full economic benefits of holding official positions” (ibid.: 9). When ending violence is the foremost priority, actors tend to shove ambitious ideas about governance reforms aside (ibid.). One could call this the ‘institutionalisation of corruption’ (ibid.). Consociationalism is established, though the lack of mechanisms to safeguard transparency and good governance presents a challenge to political stability. In these cases, “corruption becomes a rational response to the context”, meaning that corruption grows into being part of the political culture. As corruption prevails, citizens lose their trust in public institutions and often “turn to non-state mechanisms to meet their needs” (ibid.). Such situations often develop into an intrastate security dilemma5_{, with political}

instability as a result. According to Cheng & Zaum (2011), post-conflict settings often present principal-agent problems as the appointed political elites “are likely to have interests that diverge from those of external peacebuilding actors” and “it is rarely in the interests of local elites to establish independent and well-resourced institutions that threaten the networks that sustain their power” (ibid.: 14 – 20). Institutionalising societal divisions under the ‘consociationalist umbrella’ further limits opportunities to monitor governance.

Another interesting view on the causal relationship between consociationalism and political stability has been presented by Cammett & Malesky (2012). The authors argue that a mere focus on procedural democracy and constitutional design overlooks something essential, specifically “whether the system actually delivers for citizens by providing effective governance” (ibid.: 984). Patterns of governance tend to be fundamentally important in transitional states. The underlying idea for this is that when governments succeed in delivering effective governance, motivations for the feasibility of conflict and violence are suppressed (ibid.: 1009). More explicitly, “when citizens believe that their governments deliver basic services more effectively and respect the rule of law, they are less likely to favour a return to arms” (ibid.: 1009). Additionally, the authors explain that good governance can aid human development and fuel economic interactions between ethnic groups. (ibid.: 984). Through this, public trust in the government is restored and any motivations for conflict are alleviated. Cammett & Malesky (2012) introduce the idea that governance is a ‘key mechanism’ connecting constitutional power sharing to peace and stability (ibid.: 1009). Based on empirical analysis, they describe how post-conflict states with improved governance “are likely to have more durable (…) peace agreements” (ibid.: 1009). These results show how the selection of institutions that improve governance is of substantial importance for achieving political stability. In addition

5_{As a variation on the security dilemma theory as described by Herz in his 1951 book ‘Political Realism and} Political Idealism’, Roe (1999) initiated the idea of the intrastate security dilemma in which ethnic parties are so anxious to avoid conflict or any form of personal loss that they mobilise within their communities, which eventually results in conflict. Thus, the dilemma occurs.

(24)

17 to a grand coalition, institutionalised veto rights, proportional representation and segmental autonomy, good governance presents a fifth condition for a stable and successful consociational democracy.

The aim of the research explicated in this thesis is to continue to examine the conditions under which power sharing leads to political stability. Specifically, this thesis assesses how various dimensions of governance mediate the relationship between constitutional power sharing (read: consociationalism), and political stability in post-conflict societies.

3.2 The Principles of Good Governance

To understand the role that good governance plays in the causal relationship between consociational power sharing and political stability, it is important to shed a light on the concept itself. The previous paragraph emphasised the importance of good governance. In this section, more attention will be given to the question of what good governance specifically entails, as there has been a lot of confusion about the meaning of the concept. Up to the present day, no dominant definition prevails in the literature on good governance. Additionally, the concepts of governance and good governance are often confused or equated with one another. For this reason, a clear distinction has to be made between the two terms.

The concept of governance can be defined as “institutionalized modes of coordination through which collectively binding decisions are adopted and implemented” (Börzel et al., 2008: 6). An important difference is that the concept of good governance is carried by a more normative view of governance, specifically through the emphasis on how states ‘should’ operate. This can be observed in the definition of good governance presented by Voorhoeve (2007: 33): “the fair and efficient management of a country’s public resources in a sustainable fashion, as well as the upholding of laws and institutions that encourage optimal private management of people’s lives and private resources without trespassing basic human rights and freedoms”. Nevertheless, there continue to be differing views and opinions on the core features of good governance. A clear distinction can be made between the in- and output, in other words the process of decision-making by the administration and the results hereof (Börzel et al., 2008: 6). While some scholars include output indicators into their measurements of good governance, it is worth mentioning that the wide variety of definitions largely applies to the decision-making process, thus the administrative input, rather than its outcome (Van Doeveren, 2011: 303). The latter conception of good governance will be employed throughout this thesis.

3.3 Governance Indicators

A second question to be answered is how one can measure values of good governance over time and across countries. Several international organisations and institutions have quantified these values in governance indices. Kaufmann, Kraay & Mastruzzi (2004), the World Bank’s scholars on measures of governance,

(25)

18 have put forward six indicators in order to rank states on their performance regarding good governance. The indicators set forth by the scholars are voice and accountability, political stability and the absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption. The establishment of these indicators was the first step towards the World Bank’s interactive dataset on good governance. Nowadays, the dataset encompasses information from over two hundred countries, derived from more than thirty sources. The sources “reflect the views on governance of public sector, private sector and NGO experts, as well as thousands of citizens and firm survey respondents worldwide” (Kaufmann et al., 2007: 1). As the World Bank aggregates data from a wide range of indices, combined with numerous qualitative sources, the WGI dataset will be used to provide a framework for the quantitative research as carried out in this thesis. The World Bank’s scholars have argued that the WGI index is a useful tool for causal research, and its data has been widely used for academic purposes, as well as by national governments. The latter largely use the index as a basis for foreign policy decisions (Thomas, 2010: 31). The use of the indicators is particularly attractive to researchers and governments due to their broad coverage and ‘maximum precision’. Additionally, the fact that the WGI project has been promoted and financed by the World Bank “lends [it] the authority of that powerful institution” (ibid.: 33).

The World Bank has defined good governance as a matter of administrative decision-making, specifically as “the manner in which power is exercised in the management of a country’s economic and social resources” (Kaufmann & Kraay, 2007: 5). Two issues can be discovered by applying this definition to empirical research, especially with regard to the World Bank’s own governance indicators. 1) The high level of abstraction in this definition poses a challenge in identifying the underlying dimensions of the concept. The concepts of ‘economic and social resources’ provide a broad idea of possible governance dimensions, yet they seem somewhat vague and insufficient to fully grasp the comprehensive notion of good governance. 2) The definition does not quite conform to the governance indicators as presented by the World Bank itself. Taking the broad definition into account, one could state that it falls short in consistency regarding the indicator of political stability and the absence of violence, which is an outcome of good governance rather than part of the decision-making process. As it encompasses all indicators of good governance as hypothesised in this thesis, good governance as defined by Voorhoeve (2007) will serve as the starting point for further use of the concept6_.

A clear notion of the positioning of political stability and the absence of violence (in this report referred to as political stability) as a variable is essential in this research. Following the line of thought that practices of good governance are embedded in the decision-making process, the outcome of this particular process will be left out of consideration when measuring governance. Therefore, political stability will be treated

(26)

19 as a dependent outcome of good governance, rather than an indicator. The other governance indicators as provided by the World Bank, voice and accountability, government effectiveness, regulatory quality, rule of law, and control of corruption, will serve as independent variables. To support the justification for the choice in hypothesised independent variables, as well as the choice in repositioning the concept of political stability, the use of these particular indicators by other international institutions can be highlighted to exemplify an approach as such. The Millennium Challenge Corporation uses voice and accountability, government effectiveness, regulatory quality, rule of law, and control of corruption to determine a country’s performance on dimensions of good governance, leaving political stability out of its scope7_(Mazzarella,

2009). Additionally, the OECD’s ‘Governance at Glance’ index follows a similar method (Erkkilä & Piironen, 2014).

To create a better understanding of the dimensions of governance (read: independent variables), the thesis emphasises the features that these indicators encompass. The underlying sub-indicators as hypothesised by Kaufmann et al. (2007) are employed to clarify this matter. The indicator of voice and accountability measures freedom of expression, freedom of association, free media and free elections. Government effectiveness directs to the quality of public services, quality and independence of the civil service, quality of policy formulation and implementation and government’s commitment to these policies. Regulatory quality measures to what extent the government formulates and implements policies or regulations aimed at the development of the private sector. Rule of law indicates if agents abide by the rules of society, the quality of contract enforcement, the quality of police and courts and the quantity of crimes. At last, control of corruption is evaluated by looking at the extent to which public power is used for private gain (ibid.). These sub-indicators will be utilised during the qualitative analysis.

3.4 Defining Political Stability

The concept of political stability is a challenging one to define and operationalise. Through its elusive nature, a broad or abstract definition can easily be misleading. The most widely used and specific definition of political stability is probably the one as set forth by Ake (1975). The author defines the concept as “the extent that members of society restrict themselves to the behaviour patterns that fall within the limits imposed by political role expectations” (ibid.: 273). Ake describes how all acts that go over these set limits indicate political instability. However, it must be noted that the concept cannot be treated in a dichotomous manner, a political system is not simply characterised by stability or instability (Seligson, 2011). Rather, “political stability (…) covers a very broad continuum” (ibid.: 2477). Weber describes political stability as a direct result of legitimacy. The type of legitimacy that would lead to political stability would be

7_{By the corporation itself described as ‘Ruling Justly’. These indicators are used to determine which countries are} entitled to receive financial development aid (Mazzarella, 2009).

(27)

20 legal legitimacy (Weber, 1921). Lipset added to this idea by arguing that such stability is established “when political systems are seen by their citizens as being effective at delivering public services and are held accountable for their actions” (Seligson, 2011: 2478; Lipset, 1960).

The main idea behind this is strongly related to the argument by Cammett & Malesky as presented in paragraph 3.1, i.e. that when states achieve this government effectiveness and accountability, citizens would have no motivation for violence or unrest (Seligson, 2011: 2478). One issue with the notion that political stability is a direct result of legitimacy is that it does not allow for the individual measurement of political stability as an independent construct. Therefore, it is necessary to move beyond legitimacy, towards more specific measures of political stability. To exemplify this, Gilley (2009) includes indicators as public opinion and number of protests to measure the concept. Similarly, Vallier (2017) describes indicators of social trust, cooperative behaviour, low variability and immunity to destabilisation, and Blanco & Grier (2009) use measurements as demonstrations, riots, strikes, and government crises.

The interpretation of political stability in this research is a combination of these indicators. Political stability will be treated in this thesis as a system in which armed conflict, violent demonstrations, international tensions, social unrest and orderly transfers are absent or limited to a minimum. These sub-indicators have been introduced by the EIU and are employed by the WGI project. As the indicators are embedded in the data from the WGI index, they will similarly be used throughout the qualitative analysis.

3.5 Conclusion from the Conceptualisation

Continuing on the consociational framework, this third chapter has provided a critical view of the shortcomings of Lijphart’s consociational theory. To sum up, scholars have contended that the theory places too much emphasis on elite cooperation and neglects some additional dimensions that fuel cooperation on the societal level. Specifically, the dimensions of good governance are argued to have a positive effect on societal cooperation, and thus, on political stability. It is important to note here that the dimensions of good governance are treated as administration input, or decisions made on the political elite level, rather than the observable results of these. The indicators of good governance as provided by the World Bank, specifically voice and accountability, government effectiveness, regulatory quality, rule of law, and control of corruption, will serve as independent variables for the quantitative analysis. The concept of political stability will serve as the dependent (outcome) variable. This presents the following hypotheses:

H1: Voice and accountability affects political stability; H2: Government effectiveness affects political stability; H3: Regulatory quality affects political stability; H4: Rule of law affects political stability;

(28)

21 H5: Control of corruption affects political stability.

The data from the World Bank’s WGI index will be employed for the quantitative analysis. The next chapter will further elaborate on the methodology applied in the research.

Power Sharing in the Western Balkans: A Critical View on the Relationship between Consociationalism, Good Governance, and Political Stability