Increased representation of women in Serbia’s public leadership roles

Serbia’s public leadership roles

A qualitative study on the

mechanisms that have stimulated

Serbia’s gender equality in the public sector, in a period of EU

influence

Teodora Simovic | s2280175

Master Thesis

Faculty of Public Administration International and European Governance

Supervisor | Prof. dr. Sandra Groeneveld Co-assessor | Dr. Maarja Beerkens Date | 20 December 2019

2

List of contents

1.1 Aim and research question ... 8

1.2 Relevance ... 10

1.3 Outline of the thesis ... 11

2. Theoretical framework ... 13

2.1 Representative bureaucracy ... 13

2.1.1 Weber’s ideal type of bureaucracy ... 13

2.1.2 The importance of representation ... 14

2.1.3 Power, equal opportunity & diversity management ... 16

2.2 Institutional isomorphism ... 19

2.3 Expectations ... 21

3. Methodology ... 23

3.1 Research Design ... 23

3.1.1 What is a Case Study? ... 23

3.1.2 Serbia’s public sector as a case study and context ... 23

3.2 Data collection ... 25

3.2.1 Documents ... 25

3.2.2 Interviews ... 27

3.2.3 Operationalization ... 28

3.3 Process tracing ... 29

3.3.1 What is process tracing? ... 29

3.3.2 Conducting process tracing ... 31

3.3.3. The four tests of process tracing ... 33

3.4 Methodological limitations ... 34

3.4.1 Limitations of case study research ... 34

3.4.2 Limitations of process tracing ... 36

4. Representation of women in Serbia’s public sector 2008–2019 ... 37

4.1 General increase in women in public sector ... 37

4.2 A more nuanced perspective on the general increase ... 43

4.3 Conclusion ... 45

3

5.1 Institutional changes in the EU ... 47

5.1.1 Treaty of Amsterdam ... 47

5.1.2 Treaty of Lisbon ... 48

5.1.3 Other initiatives ... 49

5.2 Institutional changes in Serbia ... 50

5.3 Gender legislation in Serbia ... 53

5.3.1 Gender Quota of 30% - 2004 ... 53

5.3.2 Law on Gender Equality - 2009 ... 53

5.3.3 Gender Equality Index (‘GEI’) - 2013 ... 55

5.3.4 The Coordination Body for Gender Equality (‘CBGE’) - 2014 ... 55

5.3.5 Gender-responsive Budgeting Plan - 2015 ... 56

5.3.6 National Strategy for Gender Equality – 2016-2020 ... 56

5.4 Conclusion ... 57

6. Analysis ... 58

6.1 Public administration reforms in times of political uncertainty 2002–2008 ... 60

6.2 The creation of a multilateral partnership 2008–2013 ... 63

6.3 Increased promotion of gender representation 2014–2018 ... 65

6.4 Contemporary debates on gender equality ... 70

6.6 Conclusion ... 71

7. Discussion & Conclusion ... 73

7.1. Conclusion ... 73

7.2. Limitations of this study ... 75

7.3 Theoretical and societal implications of this study ... 77

References ... 78

Appendix 1 ... 84

4

Abstract

Over the past two decades there has been a significant increase in the representation of women in public leadership roles on the European continent. In Serbia, almost every public sector has experienced increased passive representation of women in decision-making bodies. In this thesis, the increase in female representation in Serbia’s public sectors is explained by using the theoretical framework of institutional isomorphism. The rebuilding of the country after the wars in the nineties and the gradual accession process in the EU has forced Serbia to change its institutional design. Mimetic, coercive and normative mechanisms explain how Serbia has adopted changes according to the EU framework. The various laws, strategies, policies and quotas in Serbia are a direct product emanating from the EU (Ignjatovic & Boskovic, 2013). The subsequent increased representation of women in public leadership roles is a consequence of this active promotion of gender equality. Firstly, the pressure, guidelines and coercion of the EU accession process has forced Serbia to include more women in public leadership roles. Secondly, Serbia has mimicked the institutional design of EU member states and EU institutions to achieve this goal. Domestic institutional design initially did not allow for policies and strategies that promote equality. Thirdly, by bringing Serbian officials that have been educated in the EU into the government and public sectors, Serbia included western norms in the ministries and policies. These norms, including gender equality, shaped the changes in Serbia’s institutional design. In conclusion, Serbia's fragile position and EU accession process caused changes in the institutional design. It is thanks to these changes that the ratio of women increased in Serbia’s public sector.

Keywords: institutional isomorphism, representative bureaucracy, process tracing,

women in public leadership roles, Serbia, EU accession, EU enlargement, normative process

5

Abbreviations

ABBREVIATION EXPLANATION

CBGE Coordination Body for Gender Equality DM Diversity Management

ECHR European Convention on Human Rights EEAS European External Action Service

EIGE European Institute for Gender Equality EO Equal opportunities

EU European Union GEI Gender Equality Index

GTZ German Technical Cooperation Agency

ICTY International Criminal Tribunal for Yugoslavia NSGE National Strategy for Gender Equality

OPCW Organization for the Prohibition of Chemical Weapons PAR Public Administration Reform

RTS Radio and Television of Serbia

SAA Stabilization and Association Agreement SAP Stabilization and Association Process

TFEU Treaty on the Functioning of the European Union UN United Nations

6

“We expect now a full, ambitious political strategy to tackle equality between men and women, which has been in the EU’s values since the very beginning.”

7

Acknowledgment

First and foremost I would like to thank prof. dr. Sandra Groeneveld for her guidance, knowledge, expertise, time and patience. This product is a direct result of her support, help, and insights, and for that, I am truly grateful.

Secondly, I want to thank my friend Eva Boumans for peer-reviewing my work and offering suggestions that have helped me finish this research. I also want to thank Monica O’Brien for correcting my language and grammar, and pointing out any inconsistencies. You are amazing. I extend my gratitude to my mother, who has supported me mentally and emotionally during this process. Thank you for encouraging me. Along with the previous three strong women, she has been a pillar during this process.

Lastly, I would like to thank my father. This thesis would not have been possible without his insights, knowledge, suggestions, and support. I thank him for believing in me and pushing me to achieve this academic milestone.

8

1. Introduction

Equality and diversity are taking the main stage in the 21st century. Movements like the ‘Black Lives Matter’ campaign or the global ‘Women’s March’ are examples of public reactions to racial and gender inequality. It is not only the public that is fighting the battle for equality. The public sphere has also seen developments and progress regarding gender equality. Gender equality includes the representation of women at all levels of public life, including leadership roles. In the first decade of the 21st century, the European Union (‘EU’) introduced new regulations and norms on gender equality that member states are expected to implement in their own public spheres. Some are mandatory, others just recommendations. According to the World Bank, the proportion of seats held by women in national parliaments in the EU almost doubled in the last twenty years. In 1997 women accounted for 17.114 seats, whereas this number increased to over 30.500 in 2018 (World Bank 1997 & 2018). With that, we see much progress in the EU. The first decade of the 21st century was a more turbulent time for the Western Balkans. The breakup of Former Socialist Federal Republic of Yugoslavia (Yugoslavia). caused tumultuous politics and stood in the way of the progress that the EU has enjoyed. Although gender disparity in leadership roles has long haunted the Western Balkans, significant steps have been taken towards creating a more inclusive society. In the last decade, many countries in the region have adopted legislation and strategies to empower every social, cultural, economic and political group. We see the positive results of these efforts with women being appointed to significant functions. Most recently, Kolinda Grabar-Kitarovic was elected President of Croatia, and Ana Brnabic, who is openly homosexual, was appointed as Prime Minister of Serbia (Berendt & Orovic 2015; Surk 2017). The appointment of Ana Brnabic defied all expectations, because of her sexual orientation and her gender. Yet it seems as though this is not an exception but part of an overall increase in the relative percentage of women serving in public leadership jobs.

1.1 Aim and research question

This Master thesis aims to contribute to a better academic understanding of the causal mechanisms behind this increased representation of women in public leadership roles in Serbia. Representation, in this thesis, refers to the passive form in which bureaucrats share the demographic trade of gender, as the population (Andersen, 21017). When speaking about representation of women, this thesis refers to the fact that the bureaucrats themselves are

9 women. Serbia is a middle-income country in the Western Balkans. Serbia is specifically interesting because of its fairly conservative patriarchal history that is in conflict with its accession into the EU. The EU accession process today, depends increasingly on the compatibility between EU norms and the norms of member states. Over the last decade and a half this accession process has marked Serbia’s political activity and the slow move towards western liberal norms, such as gender equality. This makes Serbia an interesting topic of study.

The main question is: How did institutional changes, emanating from the EU, contribute

to the representation of women in public leadership roles in Serbia since 2002? The level to

which women are participating in the positions of power in Serbia is the dependent variable of this research, whereas the institutional changes identify the independent variables. The goal of this research is to contribute to a better academic understanding of institutional mechanisms that influence gender representation in public institutions in Serbia.

The theoretical framework is built on two pillars. The first pillar is the concept of representative bureaucracy and the second pillar is the theory of institutional isomorphism. Representative bureaucracy is a notion that explores the benefits of a demographically diverse public sector workforce (Bradbury & Kellough 2008). The theory of representative bureaucracy, as further presented by prof. dr. Sandra Groeneveld and Steven Van de Walle (2010) is important in understanding the role that representation plays in the public sector. Representation, as a theory, takes into account the developments that have influenced national governments towards more gender equality. This theory is an important tool for understanding why there is an increased representation of gender legislation and the subsequent increased representation of women in public leadership roles.

The second pillar, institutional isomorphism, will determine whether and how Serbia has reformed its public administration according to the EU model. Institutional Isomorphism is a theory by DiMaggio and Powell (1983) that focuses on the relationship between internal structures of institutions and the impact of other institutions on these structures. More concretely, institutional isomorphism is a theory that elaborates on mechanisms that ‘force one organization to resemble others that face the same set of environmental conditions’ (Hawley 1968). The three main types of isomorphism at play are coercive, mimetic and normative. The hypothesis of this research is that there is a close link between the two pillars and they are complementary to each other. This research will analyze similarities in institutional design between Serbia and EU and determine to what extent EU norms and policies have contributed to increasing representation of women in the leadership roles within Serbian bureaucracy. It

10 sets out to identify which isomorphic processes have played a role in increasing gender representation in Serbia’s public sector bodies.

1.2 Relevance

Both representative bureaucracy and institutional isomorphism are important concepts within their fields. Yet, bureaucratic theory and institutional theory have rarely been used in parallel to analyze a phenomenon. This is unfortunate because the two concepts complement one another. No research has been conducted on the impact of isomorphic mechanisms on representation in democracies. Existing research is limited to education (Croucher & Woelert 2016; Feruahason, Harley & Natalier 2013), or banking (Park 2014; Hirsh & Morris 2010). There is a lack of introducing institutional isomorphism to understand the accession process of the EU. This also applies to research on representative bureaucracy. Previous research is mostly limited to national cases within the framework of national developments. Representative bureaucracy emphasizes internal employment structures as a political issue, touching upon topics like democracy, efficiency, and legitimacy (Van der Wall & Groeneveld 2010). This thesis embeds this bureaucratic concept into an institutional theory, by introducing the mechanisms of institutional isomorphism. In that respect, this thesis aims to expand both theories.

The concept of representative bureaucracy is crucial in understanding why representation is an important aspect in public sectors. The theory explores the various benefits and necessities of increasing the diversity and representation of marginalized groups in public sectors. Institutional isomorphism, on the other hand, explains how representation was actually adopted in the public sector by looking at the way in which Serbia has shaped its public administration according to the EU model. In the case of Serbia, both the ‘why’ and the ‘how’ are of great essence to understand the development towards a higher representation of women in public leadership roles, within the Serbian public sector. Although representation encompasses both active representation (advocacy) and passive representation (diversity), this research will limit its scope to passive representation. The aim is to discover the relative increase in women holding leadership positions and how this increase in passive representation can be explained. Actual power and active representation are not variables that will be included.

In the 21st _{century, where globalization is prevailing, it is important to reconsider}

traditional theories regarding public administration. This is especially the case when international institutions, such as the EU, exercise a lot of power and dominate political

11 administrations. DiMaggio and Powell's theory has never been applied to the normative aspects of EU integration and enlargement. The normative aspect is, in this specific research, the representation of women in Serbian public leadership roles. This research focusses specifically on analyzing how institutional changes have come about and why these changes are important to understand this socio-political progress towards gender equality in decision-making. This research aims to contribute to the understanding of how gender equality may come about in institutional design. It explains why it has become such an important issue and how the concept of representation has been implemented in legislation and policies. Much research analyzing EU accession and enlargement overlooks the role of representation of women in public leadership roles. Although this deduction is merely the independent variable of the research, it is valuable for understanding how the accession procedures adopt normative issues – like gender equality – into the process. In that respect, this research provides new insights into the EU as a powerful institution. This research will demonstrate the balance of power between international institutions and domestic institutions, and their impact on gender representation.

Representation is mostly a domestic issue. However, as will become clear during this thesis, representation also concerns international politics. The link between bureaucratic theory and institutional theory, more specifically: the link between representative bureaucracy and institutional isomorphism, may serve as a merged pillar for future research and a stepping stone for implementation of more gender-equal norms and regulations.

1.3 Outline of the thesis

The subsequent chapter, chapter 2, provides the theoretical framework of this research. This chapter describes the interplay between representative bureaucracy and institutional isomorphism. This thesis will elaborate on the theory, as it stands now, and demonstrate how these two concepts are linked. Chapter 3 is the methodological chapter of this thesis. This chapter is important because it discusses how this research will be conducted. Central to this chapter is the elaboration of ´process tracing´ as the central method to determine the possible causal mechanisms that have introduced serious progress in gender equality in leadership. In addition to explaining the actual method of analysis, chapter 3 elaborates on what the variables are, how evidence will be found and what sources play an important role for the data-collection. Chapter 4 outlines the empirical evidence on the dependent variable. The dependent variable is the actual increase in percentage of women in public leadership roles in Serbia. Statistical data provides the necessary information to understand the extent of this increase, relative to the EU

12 average. The research in Chapter 5 offers evidence on the independent variables. These are the institutional changes in the EU and the institutional changes in Serbia regarding gender equality policies and the accession process. Chapter 6 ties the previous chapters together by applying theory to the evidence. This chapter analyses the dependent variable by looking into the independent variables and applying the theory of institutional isomorphism to understand the mechanisms behind the increase in women. Chapter 7 is the conclusion and discussion, which will answer the research question. The main argument will be that isomorphic mechanisms have been at play in increasing the representation of women in Serbian public sectors. The coercive, mimetic and normative mechanisms are unequally relevant in understanding the legislative and regulative measures that Serbia has taken to promote gender equality in decision-making. The conclusion explains this from the point of view of the attempts to satisfy the EU demands (coercive isomorphism), to copy the institutional design of successful institutions (mimetic isomorphism) and to incorporate professional standards introduced by experts educated in various EU countries and previously involved in projects for European institutions (normative isomorphism). The mechanism of coercive isomorphism has eminently been dominant in the increased representation of women in public leadership roles, due to Serbia's accession process to the EU. In the following chapters, it will be clear how this development has come about. The discussion offers the social and scientific implications of this study.

13

2. Theoretical framework

This chapter elaborates on the state of the art academic theory that will serve as a framework for answering the research question. The two main theoretical pillars are representative bureaucracy and institutional isomorphism. The first theory helps explain ‘why’ representation has become an important tool for contemporary institutional design. The second theory is essential in understanding ‘how’ gender representation has been implemented in public sectors in Serbia. This chapter starts with explaining the ideal type of bureaucracy, introduced by sociologist Max Weber. His notion is a predecessor of many bureaucratic theories (Serpa & Ferreira, 2019). Although his proposition has been debunked by several academics (Bartels, 2009), the fundament of his argument remains the reference for the argument central to representative bureaucracy. This aspect of the theoretical framework will look at the link between legitimacy and bureaucracy and why their understanding is crucial for further analytical analysis. After establishing the importance of representation in modern bureaucracies, institutional isomorphism is used to explain how new ideas on legitimacy and institutional design is adopted in modern bureaucracies. This chapter explains the different mechanisms that are at play. Both academic theories illustrate how changes in institutional design have caused an increased representation of women in public leadership roles in Serbia.

2.1 Representative bureaucracy

2.1.1 Weber’s ideal type of bureaucracy

Sociologist Max Weber is considered to be the father of bureaucratic theory (Serpa & Ferreira, 2019). Through the lens of sociology he provided political studies with a new perspective on organizational theory. His aim was to demonstrate how rationality is a crucial component for a well-functioning bureaucracy and institution. In his vision, ideal bureaucracies are technical and hierarchical rather than traditional (Kalberg, 1980). Traditional bureaucracies were elitist and bureaucrats were not trained, but rather by personal connection placed in those positions. According to him, the most efficient and legitimate organizational form is that where nepotism and charisma are excluded (Allen, 2004). In his type of bureaucracy, ‘legitimacy’ and ‘efficiency’ are acquired through bureaucrats, as neutral actors, that act under a clear lawful framework that is organized in hierarchical structures. Strict training, hierarchy, and rules are

14 objective structures and every bureaucrat is recruited on the basis of merit for the purpose of increasing efficiency (Allen, 2004; Kalberg 180). For Weber, bureaucracy is a rational tool in reaching goals in the most effective and efficient way.

This traditional version of bureaucracy was met with much controversy and was disputed in the academic world during the evolution of post-modernistic ideas on human nature (Bartles, 2009). Post-modernism philosophy contested the strict rationality aspect of the traditional bureaucratic theory (Gasché, 1988). Instead, they promoted the idea that individuals (i.e. bureaucrats) are driven by preferences, interests and individual desires (Davis, 2007). These new insights force us to approach bureaucratic theory from a different angle. This thesis offers the alternative of ‘representative bureaucracy’ as a more efficient and legitimate concept. Donald Kingsley used the term ‘representative bureaucracy’ for the first time in 1944 (Gravier, 2013). According to this modern approach to bureaucracy, it is impossible to have well-trained and objective bureaucrats that are resistant to favoritism and individuality, challenging the ‘efficiency’ and ‘legitimacy’ of Weber’s bureaucracy. In this reformed concept, bureaucrats let their interests and preferences influence the decision that they make, consequently becoming responsive to the demographic they are part of (Gravier, 2013). This means automatically that these bureaucrats should reflect as much of a diverse demographic as possible, because their actions are shaped by personal interest. The interests of bureaucrats derive from their demographic identity. This means that a homogenous demographic within a bureaucracy does not represent the population.

Both Weberian bureaucracy and representative bureaucracy acknowledge, to some extent, rationality as the predominant driver of actions. In Weber’s ideal type of bureaucracy, bureaucracy is an actor in itself. Representative bureaucracy, on the other hand, looks at individual bureaucrats as driven by rational acting, yet subjected to their own beings (Bailey, 2004). Maximizing personal interests, preferences and prejudices drive this type of rational acting. Emphasizing individual rationality, as opposed to rationality of bureaucracy, is where the two theories split.

2.1.2 The importance of representation

Central to the theory of representative bureaucracy, is the concept of representation. The goal is to answer how institutional changes, emanating from the EU, contributed to the

representation of women in public leadership roles in Serbia, since 2002. It is essential to

15 According to representative bureaucracy “bureaucracies need to be representative of the population for them to have legitimacy and public credibility” (Groeneveld & Van de Walle, 2010, 244). Within this context, legitimacy is based on representation, whereas the Weberian bureaucracy bases legitimacy on objectivity and neutrality of bureaucrats. According to representative bureaucracy, Weberian bureaucracies cannot be legitimate because they fail to represent the population equally, creating a democratic deficit (Seibel, 2010). Constituents and citizens do not democratically elect bureaucrats. Instead, bureaucrats hold influential positions without having any form of direct representation (Meier, 1975). Traditional scholarship revolved around the assumption that electoral representation and legitimacy of the decision-making process are necessarily linked (Murdoch, Connoly & Kassim, 2017). Precisely because of this democratic deficit caused by a lack of direct elections, bureaucracies need to make way for higher levels of representation that reflect the society. Bureaucrats make decisions that have an impact on society. The society consists of people from different races, sexual orientations, gender, and religions. The bureaucrats therefore need to reflect the population they serve by passively representing them.

There are various ways of determining the legitimacy of an organization. Governing institutions lose their authority and risk ceasing to exist if they are considered illegitimate. In their book, Antonis Ellinas and Ezra Suleiman (2012) paraphrase political scientist Robert Dahl and his ‘reservoir theory’. Here bureaucratic legitimacy is of great importance to prevailing institution because it guarantees a certain level of stability (Ellinas and Suleiman, 2012). There are two main elements of legitimacy: input legitimacy and output legitimacy (Schmidt, 2012, 4). Input legitimacy is concerned with the responsiveness, participation, representation, accountability and processes that establish rules. The main objective of input legitimacy is to uncover the underlying mechanisms of decision-making processes. Output legitimacy, on the other hand, focuses on the quality of the outcomes of these processes (Mena & Palazzo, 2012). Legitimacy is weighed against the performance of the processes from which the decisions have been established (Schmidt, 2012,). In terms of the outcomes of their decisions, institutions and bureaucracies are considered legitimate if their decisions entail collective benefits, which the member states cannot provide independently (Murdoch, Connolly & Kassim, 2017, 3). The main focus of ‘input’ is on the mechanisms behind decision-making, whereas the main focus of ‘output’ is on policy outcomes that derive from these processes (Schmidt, 2012, 19).

16

2.1.3 Power, equal opportunity & diversity management

It is clear that representation in the workforce is essential to the legitimacy of institutions. Yet, representation is not a clear-cut concept because it acquires meaning through context and academic discipline. Initial associations in political sciences include: serving interests on behalf of a group and selecting a spokesman for the job (Rogowski, 1981; Mansbridge, 2011). Representation in this context often entails a principle-agent morale in which constituents are the principles and politicians the agents (Shapiro, 2005). Here the politicians are spokesmen elected for the purpose of serving the interests of their public. In this increasingly complex political climate, however, the principle-agent relationship is being challenged by a fragmentation of the decision-making model and increase in bureaucracy (Kassim & Menon, 2003). The delegation of decision-making, which has an impact on the principles, is not ruled by elected agents but also by bureaucrats. The concept of representative bureaucracy tries to solve this issue.

In their article, Groeneveld and Van de Walle, elaborate on three different elements of representative bureaucracy: power, equal opportunities (‘EO’) and diversity management (‘DM’) (Groeneveld & Van de Walle, 2010). Although all three elements of representative bureaucracy are equally important, the concept of ‘power’ is not relevant for understanding the increased role of representation in contemporary institutions. This theory on ‘power’ is outdated because it is focused on the representation of elites instead of variety in demographic. Instead, this research will focus on EO and DM for understanding the elements of representative bureaucracy.

Their second element is that of equal opportunities, applying representation to marginalized groups (Groeneveld & Van de Walle, 2010, 244). The main argument revolves around the moral conception that more representation equals more responsiveness. This means that bureaucrats more easily respond to the needs of the demographic that they represent. This approach differentiates passive and active representation, where the former “collectively mirrors the population” (Hindera, 1993) and the latter actively takes into account the desires of their marginalized groups (Riccucci & Meyers, 2004). The main difference between passive and active representation lies in its purpose. The purpose of passive representation is not to act on behalf of a group but rather serve as a symbolic figure of the group (Gravier, 2013). The purpose of active representation, on the other hand, is to act on behalf of groups (Bell & Rosenthal, 2003). Groeneveld and Van de Walle problematize active representation. In practice, it is hard to achieve active representation because of the scarcity of positions in which it is possible, and favorable to promote these interests (Groeneveld & Van de Walle, 2010, 246).

17 The concept of equal opportunity is limited to passive representation. The essence of equal opportunity is symbolic because it has no functional purpose. It is a normative dimension that emphasizes the need for representation for improving the position of the institutions in the eyes of the public (Groeneveld & Van de Walle, 2010) According to this element of representative bureaucracy, representation is necessary because it is the right thing to do to increase responsiveness and participation. The concept of representative bureaucracy puts individual preferences central to the workforce of a bureaucracy (Bailey, 2004). This means that passive representation benefits the demographic groups of the bureaucrats by being responsive to their needs. Additionally, institutions can be held accountable if they obstruct inclusion. One of the consequences of this inclusion-trend is that marginalized groups step up to demand greater inclusion in the workforce (Özbilgin and Tatli, 2011, 7). Consequently, the public may exercise their power by scrutinizing the public service if their recruitment procedures are unfair, which automatically challenges the legitimacy of their administration (Ellinas and Suleiman, 2012). In that respect, the theory of EO emphasized the role of input legitimacy because organizational structures and recruitment procedures of public service should represent the diverse population. The elements of EO are similar to those of input legitimacy: responsiveness, participation, representation and accountability.

The approach of diversity management (‘DM’) emphasizes the importance of diverse representation for the benefit of increasing performance quality and effectiveness (Groeneveld & Van de Walle, 2010, 247). According to this approach, increased representation is motivated by strategical considerations rather than moral obligations. DM is an instrumental tool because representation increases the quality of job performance, productivity and success, rather than symbolic gestures (Groeneveld & Van de Walle 249). Enhancing the decision-making process with a focus on successful outcomes is closely linked to output legitimacy. Every aspect of human identity, such as gender, age, sexual orientation, religious beliefs and ethnical background, influences the way people act and their perception of best possible outcomes (Lawrence, 2000). Diverse workforces base outcomes on a variety of perspectives that derive from the people who work there; ultimately creating a policy that enjoys variation, creativity, and collaboration. In this respect, representation in bureaucracies is meant for the purpose of increasing success, because a diverse workforce brings additional perspectives, information, and creativity to those of homogenous institutions. The output legitimacy of an institution, thus, depends on the diversity of the workforce.

In sum, the legitimacy of an institution depends on representation. Contemporary bureaucracies gain input legitimacy through passive representation, which is put forward by the

18 concept of EO. Equally, institutions use representation to increase their output legitimacy, because diversity increases the performance of outcomes. These concepts will help answer ‘why’ representation is increasingly playing a role in institutional design. Both symbolic motivations and instrumental motivations give two different perspectives on this issue as shown in the diagrams below.

Figure 1

The interplay between representation, legitimacy and equal opportunity

Figure 2

The interplay between representation, legitimacy and diversity management

Equal Opportunity Input legitimacy Symbolic Representation Diversity Management Output legitimacy Instrumental Representation

19

2.2 Institutional isomorphism

Institutional isomorphism is an interesting theory to explain how representation and representativeness have become important to bureaucracies. Institutional isomorphism is a theory by DiMaggio and Powell that explains the trend of homogeneity between institutions and the design of their administration. In their article, “The Iron Cage Revisited”, they elaborate on bureaucracy as a primary institutional form. Contemporary bureaucracy is becoming increasingly similar and institutional structure, culture and output become homogenous within certain fields without necessarily increasing their efficiency (DiMaggio & Powell, 1983). This theory lays the groundwork for chapter 5 in which evidence will be analyzed in the context of this theoretical framework. It will demonstrate what mechanisms have been at play in Serbian public sectors. The analysis will apply these various mechanisms to the evidence to see what the causal mechanisms are that have influenced the increased representation of women in public leadership roles. The aim of this chapter is to uncover the extent to which representation has become crucial to institutional legitimacy. Central to the theory of DiMaggio and Powell are three mechanisms that influence this homogeneity in institutions within certain fields. These include coercive isomorphism, mimetic isomorphism and normative isomorphism (DiMaggio & Powell, 1983). All three mechanisms of isomorphism can be applied to understanding an increasing drive toward representation. They are not mutually exclusive, although their origins and mechanisms differ significantly.

Coercive Isomorphism is determined by both formal and informal pressure of other

institutions, cultural expectations, and authority regulation. There is a hierarchical dimension to this type of isomorphism, in that bigger institutions exercise pressure, use force, persuasion or other methods to ensure compliance by smaller or newer institutions (DiMaggio & Powell, 1983). Similarly, governmental entities may impose guidelines, mandates, and laws, which require institutions to abide by these rules. These rules apply to all institutions that operate equality and equal opportunity. These norms about representation are similar to what is called symbolic representation in the previous section. Informal pressures cause expectations of what is ‘the right thing to do’ in regard to increasing equal opportunity. However, authorities may also exercise formal pressures in the form of framework, which is dominated by guidelines, rules, and laws. In the context of representation, this may be expressed in legislation on gender equality, quotas, conditionality and pressure (Weaver, 2011). These norms and rule-based frameworks then apply to all relevant bureaucracies. The international community and the legitimacy of organizations may be negatively impacted if they do not comply with these

20 impositions. In this respect, coercive isomorphism is a tool to increase the input legitimacy of an organization.

Mimetic Isomorphism, on the other hand focuses on conformity in times of uncertainty.

This type of isomorphism copies from institutions intentionally when problems are ambiguous and strange. DiMaggio and Powell use the term ‘modeling’ to refer to the actions that institutions take during these periods (DiMaggio & Powell, 1983). Institutions that have enjoyed success, or organizations that are ‘big’, may become prestigious and, therefore, become a ‘model’ for less successful or new institutions (Haveman, 1993). This is an informal mechanism because there is no official agreement for modeling between the successful institution and the institutions that is doing the copying (Glazier & Hall, 1996). Successful institutions unconsciously serve as prototypes for reform or establishment of other organizations. Mimetic isomorphism explains the instrumental aspect of homogeneity between institutions. In the context of this thesis, if high levels of representation are proved to be essential in successful institutions, less successful institutions will increase the diversity of their staff to enjoy the same success. The purpose then is to increase the efficiency of the institution by including representation in their staff policy. The emphasis here lies in the output legitimacy because lesser institutions aim at increasing their performance by stealing policies from successful examples. As a consequence, various small institutions may copy from the same successful institution. This would cause a homogenous adoption of similar policies for the same purpose.

Normative Isomorphism looks at more micro-contextual processes that determine

homogeneity. Organizational isomorphism originates from professionalization produced by formal education and informal socialization/networking between actors within specific fields. The staff of organizations becomes increasingly homogenous because universities produce similar cognitive structures (DiMaggio & Powell, 1983). Therefore, students from certain disciplines, who aim at a specialized career, derive their knowledge from the same sources. The other factor that contributes to the homogeneity of the workforce is the fact that staff members from certain fields or institutions socialize among each other, exchanging ideas, methods, and knowledge (DiMaggio & Powell, 1983). This socialization is enabled by conferences and networking events, as well as personal relationships and professional collaborations. In this respect, one could argue that a form of micro-level modeling may occur among peer-professionals. The recruitment procedure of staff is based on filtering from the same industries, networks and educational institutions (DiMaggio & Powell, 1983). Normative isomorphism contributes to increasing representation, in that the socialization between bureaucrats enforces

21 gender equality in the workforce. An organizational field may organize networking events and campaigns to advocate for equality and diversity of staff. An example is the OPCW (Organization for the Prohibition of Chemical Weapons), that organized an event on 11th January 2019, promoting diversity and inclusion in the workplace (Organization for the Prohibition of Chemical Weapons, 2019). Such awareness campaigns aim at changing the dominant discourse of the organizational field. Different institutions are attracted by these events, which should ultimately cause this input to be adopted by other institutions. Specialists from other institutions may bring new ideas on representation or equality.

All three mechanisms provide us with various examples of how gender representation has infiltrated Serbia’s public sector. The extent to which all these mechanisms have been at play will be clear, and why representation has become such an important aspect of the public workforce. Both from the perspectives of institutional isomorphism and representative bureaucracy, the legitimacy of institutions depend on the level of representation.

2.3 Expectations

This theoretical framework provides us with the necessary tools to understand the underlying forces that may have caused an increasing proportion of women in public leadership roles in Serbia. This research will measure the increase in percentage of women in Serbian public sectors against the isomorphic processes that may have been at play in recent decades. The highly globalized world has forced Serbia to comply with international legal frameworks, such as the UN Declaration of Human Rights, EU quotas and domestic equality advocacy networks. As such, it is expected that coercive mechanisms have influenced female representation. Equally, it is anticipated that findings will show that Serbia copies institutional design from EU institutions to increase its performance and efficiency. It is not unthinkable that Serbia has used other EU member states as authoritative examples. The last element of isomorphism, which is the normative dimension, will also play a crucial part in the empirical analysis of this research. The hypothesis is that due to the constant interaction between officials from EU countries and Serbia, there is much EU influence on the cultural and normative dimension of Serbia’s public administration. Both bottom-up as well as top-down forces are at work in the promotion of women in public leadership roles, which would include domestic campaigns and the abovementioned international socialization.

This study addresses the issue of representation in Serbia’s public sector from an institutional perspective, rather than instrumental. This research tries to elaborate on the forces

22 that have created more representation and diversity in institutional structures. Action for increasing representation in leadership roles is expected to derive from all three mechanisms. The hypothesis is that the EU is the successful institution and that Serbian public sector is the subordinate institution. Instrumental convictions may have played a role for the EU to adopt legislation on gender or to increase the representation of women in their public institutions. Serbia is in the process of becoming a member of the EU. Therefore, the expectation is that complying with EU guidelines and norms, modeling against other EU member states and the constant interaction between the EU and Serbia has been an incentive for Serbia to increase the percentage of women in public leadership roles in Serbian bureaucracies.

23

3. Methodology

This chapter will outline the structure of this study by elaborating on the research design, the data collection and the method of analysis. This study is inherently a small-N case study research, conducted with qualitative data. The first part of this chapter distinguishes the independent and dependent variables. This is essential for understanding what this research is about and what the various elements of analysis are. In summary, this first part of the chapter will explain what type of study this is and which variables will be studied. Subsequently, this chapter indicates where the evidence is collected. The second section will elaborate on the motivations for using documents and interviews as the main sources of research. This part of the chapter explains how the research will be conducted. The last section is on process tracing, which is a method for analyzing evidence. Process tracing is used for tying the theoretical framework with collected data. The concluding remarks of this chapter discuss this method, and provide nuance for other types.

3.1 Research Design

3.1.1 What is a Case Study?

This research is a ‘single-case study’. Blatter and Haverland (2012) provide several characteristics of single-case study research. They state that case study is suited for research that is based on empirical data regarding a single phenomenon, which aims at discovering causal mechanisms rather than causal effects or the effect of a cause (Blatter & Haverland 2012 p.20). Causal mechanisms are processes that lead to an outcome and not a direct cause of an outcome (Falleti & Lynch 2009). The contexts in which causal mechanisms operate are essential to understanding their outcome (Falletti & Lynch 2009). In comparison to large-N design, there are no concrete lines between the actual phenomenon and the context. The focus of a single case study is on the context, the process and the specific situation of one object.

3.1.2 Serbia’s public sector as a case study and context

This study is focused on uncovering underlying mechanisms that have caused the increased representation of women in public leadership roles in Serbia. Specifically, it will look into the socio-political changes that have occurred in the past decade and how these have impacted the percentage of women in public sectors. The focus of this research is on Serbia specifically

24 making it inherently a single case study with a small-N design. Within a small-N research the object of analysis is one specific subject rather than a population of subjects. Therefore, this research is limited to the situation in Serbia and will, therefore, refrain from producing general statements about the underlying mechanism of the increased representation of women in public leadership roles in general.

Figure 3: The dependent variable (Y) and independent variable (X)

2.a Regulatory framework (X)

EU accession (X) 1. Gender equality strategy (X)

2.b Increased representation (Y)

The research question is divided by a dependent variable (Y) and independent variables (X) as shown in the figure above. The dependent variable of this research is the representation of women in public leadership roles in Serbia. These sectors consist of political institutions, like governmental branches, but also include for example national banks and public broadcasting stations (like RTS). The independent variables are EU accession, interventions, legislations, policies and features that have instigated changes in Serbia’s institutional design. This specific research is not set out to analyze the cause of the increased representation of women in public leadership roles or the causes of the variables X and Y. In this context, there is not one independent variable as the cause of the dependent variable. Instead, this research is focused on understanding the causal mechanisms that are at play. The goal is to uncover these causal mechanisms, rather than pinpointing the exact cause of the outcome Y. The various confounding causal factors force us to look at X1 →X2 →X3 →X4 → Y. This study is inherently

an explanatory research with a focus on causal mechanisms represented as '→'. The analysis will apply the theory to explain the →, and draw a conclusion about the dependent variable Y.

25

3.2 Data collection

There are various ways to gather evidence for qualitative research. The subsequent section will elaborate more on the type of sources that are used for this research. This section of the chapter establishes the reliability of the study. It offers insight into the credibility of information and the operationalization. In order to understand the study's outcome, it is important to know what type of sources are used, where they are found and how they relate to answering the research question. The two main sources of information include documents and interviews.

3.2.1 Documents

The most important source of data is derived from primary sources in the form of documents. This research considers online databases with statistical data on gender ratio, official policies, governmental speeches, and statements from EU officials and Serbian politicians. However, primary sources are not necessarily reliable and accurate. It is therefore vital to assess all relevant sources to determine the extent to which they offer a valuable contribution to the study. The four main criteria for the quality of documents are its authenticity, credibility,

representativeness, and meaning (Scott, 1990).

Authenticity refers to authorship, origin, integrity and sincerity of the source. The authenticity of a source depends on the extent to which it is genuine in reporting on what it claims to report. The evidence must be tested to avoid using forgeries or falsification. It is especially important to be alert if documents have obvious mistakes, inconsistencies, different versions, unreliable secondary sources, unreliable authors or are from dubious platforms (such as Wikipedia) (Platt, 1981). It is very possible to determine the authenticity of many sources available to this research. However, in the case of government documents – even if signed by a Minister – the initial author may be a civil servant. Governmental documents are trivial, in the sense that it is hard to determine the origin of the source and the author of the document. When using evidence from government websites, policies and/or statements the author has to be taken for granted (Mogalakwe, 2006).

Credibility refers to the veracity of the content of the document. The credibility of the documents is based on the distortion of insincerity that might be related to them (Scott, 1990). There are various errors that may taint the credibility of documents. These include the preferences and mistakes of the author, coming from personal preferences and points of view. It should be taken into account that errors may occur, even if the author did not intend them. The primary documents in this research are for the most part honest and accurate, considering

26 they do not depict an author’s opinion or experience. Statistical databases and policies are not subject to the personal input of the author. However, this research will also use statements from officials and policy reports that are, in essence, ‘subjective’ materials. The veracity of the documents is hard to determine, in that respect. This will be especially interesting because the ‘subjective’ aspect of these sources is evidence in itself and can be used to debate the information available.

Representativeness of the documents is hard to determine for most of the sources. Representative evidence ‘is typical of its kind’ (Mogalakwe, 2006). This means that the evidence reflects on the general position of all other documents in the field. It is hard to determine the representativeness of documents that go beyond borders and are on an institutional level. Therefore, this research will have trouble determining whether the EU's progress reports on Serbia represent the position of the EU as an institution. Yet, this should be taken for granted and assumed so. On the other hand, statements that speak on behalf of their institution (such as Serbian politicians or EU officials) are not an accurate representation of the position of the institution.

Meaning is a criterion that depends on the readability of the document and the interpretation of the information. The readability of the document depends on whether the language is clear and accurate and whether it is possible to literally understand the meaning of the text (Scott, 1990). The interpretation of the meaning is another obstacle for the researcher. First, the researcher should interpret the information within the context in which it is written and then the researcher should be aware of the context from which his interpretations derive.

Especially the latter two criteria may be subject to bias and interpretation. This is something to be considered in the process. I, as the researcher of this study, am conditioned by my gender (female), my origin (Dutch and Serbian) and my personal views and biases. It is of great importance to aim at limiting these influences as much as possible. Yet to a certain extent, these aspects have control over the researcher and the researcher should therefore, be aware of them beforehand and acknowledge them in the research (Rich, 1984).

In Appendix 1 there is a list of all the documents that were consulted during the thesis. This list comprises statistical databases, government websites, governmental press releases, online articles, EU progress reports, treaties and similar policies. All these documents are non-academic documents, some of which are primary sources.

27

3.2.2 Interviews

To complement to the abovementioned documents, the second source of information is derived from interviews. This means that the interviews will serve to provide more in-depth insights into the practices of the Serbian government from first-hand officials or other types of experts (Wilson, 2016). Interviews provide similar threats to the quality of data that documents do. It is hard to determine the accuracy of the information, the representativeness, and the meaning. Both the interviewer and interviewee are subject to interpretation (Seidman, 2006). This means that the interviewer may be biased in the selection of interviewees, in the information that the interviewer chooses to use and the interpretation of the wording (Lawrence, 2000). However, the positive aspect of interviews is that the interviewer can ask follow-up questions if he does not understand something or if the researcher is afraid that the information is unclear, insufficient or inaccurate.

This research makes use of four interviews: two are with an academic expert and two are with participants that have been employed in leadership roles in Serbia. Respondent 1 is an expert in the field of social and political philosophy. His insights are valuable in understanding the motivations of high-ranking politicians in Serbia and the response of political and social justice movements. The interview with Respondent 1 is of great value in understanding how the political structure of Serbia is shaped today.

The second academic interviewed, Respondent 2, is a Croatian Sociologist. She has become an expert in the field of EU enlargement in the Balkans from a sociological point of view. Her perspective is especially interesting for two reasons. Firstly, the personal background provides her with a subjective understanding of the normative and political climate in Serbia. Secondly, she has worked for international institutions and has lived in the EU for over twenty-five years.

The second two interviewees are women who have been active in different fields of Serbia’s public administration. Respondent 3, who is an active member of the Democratic Party, was also one of the major journalists working for Radio and Television of Serbia (‘RTS’). The interview with her provides insights into the practices of a public institution and her political affiliations may serve as an interesting tool to uncover the underlying forces of their regime.

Ultimately, I have interviewed a prominent Serbian lawyer, Respondent 4. The Supreme Court is a public institution. Apart from analyzing public political institutions and public broadcasting, it is interesting to take into account the female ratio within the jurisdiction. The role of women in the Supreme Court is an interesting dimension of gender equality. Her experiences as a legal scholar are important in understanding the feminization of legal studies.

28

3.2.3 Operationalization

This research is focused on analyzing which isomorphic processes have influenced the increased representation of women in public leadership roles. This means that the evidence is essential in trying to prove or debunk the hypothesis that coercive, mimetic and normative mechanisms have been at play in the female-male ratio in Serbia’s public sector. The dependent variable of this study is the increase in the percentage of women in Serbia’s public leadership roles. This is quantitative data that is derived from statistical databases and/or secondary sources that use similar information. The European Institute for Gender Equality (‘EIGE’) has accumulated data on gender statistics in the EU and several countries outside the EU, including Serbia (Lazic, 2018). This has been the main source of data for determining the ratio of men and women in the public sectors in Serbia from 2008 until today (Appendix 1, Source 19). The database has a wide variety of data that is not all necessary for this research. The main criteria are the available data from ‘women and men in decision-making’. For this research, the only three relevant sectors of decision-making are ‘politics’, ‘public administration’ and ‘judiciary’. In this database, it is possible to compare data from different countries and from the EU average. The comparison between Serbia and the EU average is especially interesting because it puts the information from Serbia into the EU perspective.

The independent variables refer to the isomorphic mechanisms. This is the qualitative aspect of the research. Coercion, pressure, hierarchical power play and informal pressures determine coercive mechanisms. Sources that support the increased of women in public leadership roles in Serbia need to consider an ‘authoritative’ figure that pressured Serbia’s public sector to increase the ratio of women. The authoritative figure can be a domestic legislative body such as the leading party, coalition or a judge. The authoritative figure can also be an international institution. In Serbia’s situation, the EU is an authoritative institution that can express pressure and conditionality to enforce compliance (Weaver, 2011). Therefore it is essential to consider domestic policies and legislation on gender equality in decision-making, enlargement policies of the EU in terms of Serbia’s accession and the EU's progress reports on Serbia. Legislation and pressure are tools that institutions can use to impose institutional changes onto smaller institutions.

Mimetic isomorphism is more difficult to analyze because it is difficult to understand intrinsic motivations through secondary sources or from statements. The mimetic mechanism implies that smaller institutions copy from more successful institutions in times of uncertainty (DiMaggio & Powell, 1983). It is first and foremost important to determine when and if Serbia’s public administration has experiences uncertain times. If this seems to be the case, it is

29 important to understand how Serbia has responded to these uncertainties. There are public strategies available that imply that Serbia has changed its institutional design because it experienced turbulent times. An example is the Public Administration Reform strategy (PAR strategy) (Dzinic, 2011). Analyzing this strategy, statements of officials that have worked on this strategy and the public administration of successful institutions will then determine whether mimetic mechanisms are at play. Statements report on underlying motivations of politicians (Mihajlovic, 2018). Several important actors have explicitly stated that they want to copy from the EU model.

To most difficult mechanism for this research is that of normative isomorphism. Normative isomorphism is concerned with the professionalization and socialization of staff. DiMaggio and Powell (1982) consider the normative mechanism to be most prevalent because institutions hire from the same set fields, causing homogeneity in the workforce. Education (and other) institutions produce like-minded people. These people then interact with other people from their field, exchanging ideas and values. To determine whether we can speak of normative isomorphism it is important to consider the educational and ideological background of the officials that have caused reforms in institutional design, by promoting ideas, producing policies and legislations, appointing more women and setting up strategies (DiMaggio & Powell, 1983). Additionally, it is valuable to look into events in which EU member states and Serbia have participated, and that may have opened up dialogue on the topic of gender equality or where officials may have shared norms and values. However, many interactions between EU officials and Serbian politicians have happened behind closed doors. Therefore it is very hard to analyze the causal relationship between interaction and action.

3.3 Process tracing

3.3.1 What is process tracing?

Process tracing is a popular method in historical and social sciences for analyzing qualitative data in within-case research. The main objective is to outline the causal mechanisms between X and Y variables. Causal process tracing is a useful method for looking at ways in which certain phenomena develop overtime. This temporal aspect of the method is what makes it an attractive tool to determine how time and space impact phenomena. Drawing causal inference allows researchers to pinpoint crucial moments in time that influence a specific case (Blatter and Haverland 2012). There are four main types of causality, which include; counterfactual, regularity, configurational and generative causality.

30 Counterfactual causality assumes that there is a strong cause-effect (X-Y) relationship and that without that specific cause, the effect would not happen. In other words, there is no Y without that specific X. One way of testing the counterfactual claim is to intervene with the X to test whether the outcome would still be that Y. In that respect, counterfactual causality differentiates two identical subjects, with a minor intervention in one, to determine the impact on cause and effect. For this research it is impossible to test the counterfactual because this needs to be in a controlled environment where there is a claim for a specific cause and effect, to test whether this cause really results in that effect. This research is not in a controlled environment and the aim is not to test a specific cause for the increased representation of women in public leadership roles in Serbia.

The regularity perspective on causality, on the other hand, looks at the “frequency of

association between two observations to assess an effect” (Punton & Welle, 2015, 2). The main

assumption is that two phenomena are associated with each other because in more than one occasion they have been linked. However, it is hard to determine which one is the cause and which one is the effect. There is no clear X-Y and therefore it is hard to understand how, why and what the causal relationship is. This research does not analyze the frequency of association between women in public leadership roles and certain causes. That would entail researching this trend in other institutions and countries as well, instead of limiting the topic to merely Serbia. The regularity perspective is therefore not relevant for this research.

The third perspective, called configurational causation, is more concerned with a multitude of causal origins. Central in this way of thinking is the assumption that all phenomena derive from a plurality of interplaying and independent causes, meaning that there are several X’s for one Y. Although this type of configurational thinking is central to the method of process tracing, as Blatter and Haverland explain, it is insufficient to fully grasp the complexities behind single case analysis. This type of causality is limited to understanding what factors work together to create an outcome. But configurational causation does not include the deep-rooted mechanisms that allow for this causality to happen.

According to some, the configurational perspective is mainly relevant in the context of comparative analysis, whereas the fourth approach, that of generative causation, is a more based on the causal mechanisms of a concrete effect. Here causality is explained in the temporal state, emphasizing ‘why’, ‘when’ and ‘how’ a certain outcome came into effect. Process tracing is a synthesis of the latter two approaches. Configurational thinking is central to the assumption that no outcome derives from only one cause. The configurational aspect is important because there is not just one cause for the increased representation of women in public leadership roles in

31 Serbia. Yet, this study aims at uncovering the causal mechanisms of a specific outcome, as provided by the generative perspective. Therefore, both versions provide to some extent relevant insight. For this research causal mechanism is:

“… a mechanism is a set of interacting parts – an assembly of elements producing an

effect not inherent in any of them. A mechanism is not so much about ‘nuts and bolts’ as about ‘cogs and wheels’ – the wheelwork or agency by which an effect is produced.”

(Hernes, 1998, p.78).

These interacting parts are similar to the configurational interaction of certain causes leading to an outcome. This quote suggests that the process and development of a certain outcome are far more important for this type of research than the actual outcome itself. Process tracing is essential in understanding what mechanisms were at play that have introduced a serious progress of gender equality in leadership roles in Serbia. The contextual assumptions play a far more important role than pinpointing the exact moment women became increasingly important in leadership roles or how they are fulfilling their jobs.

3.3.2 Conducting process tracing

Process tracing can be used for several purposes. Derek Beach and Rasmus Brun Pedersen developed three different versions of process tracing. Two of them are theory-centric and one is case-centric. The theory-centric versions are ‘theory-testing’ and ‘theory-building’ and the case-centric version is ‘explaining outcome’. The first is ‘theory-testing’, where evidence is collected to support a correlation between X and Y. This variant is deductive in nature, as it tests existing theory on observations regarding causal mechanisms and whether they were present in a specific case. Theory building, on the contrary, aims at building a theory supported by observations from a specific case, which can be generalized in its context. Theory-centric process tracing is inductive and deductive in practice, whereas case-centric process tracing is abductive (Toshkov 2016). This variant aims at ‘explaining outcome’ by uncovering the causal mechanisms of a specific case that cannot be generalized. In this case, there is no X-Y relation that is subject to analysis. Instead, there are various causes that interplay and correspond in creating the outcome (Y). Many other researchers claim that process-tracing and research that is focused on uncovering specific and contextual causalities often correspond with abductive reasoning. Some scholars claim “process tracing most closely resembles abduction” (Beach & Paterson 2013, p.28). The emphasis on the case study of this research requires abductive analysis because this method seeks case-specific evidence (Kay & Baker 2014). The goal of

32 this research is not to test the theoretical framework of isomorphism on the exemplary case of Serbia’s bureaucracy, instead the goal is to explain this specific case, its context and causal processes by means of using existing theory. Therefore, this research is abductive in its nature. On the subsequent page, you will find table 1, which outlines the types of research of process-tracing, as presented by professor Derek Bach (2012).

Table 1

Types of research for process tracing