Policy Case Selection
The two countries compared here are post-conflict Bosnia and Macedonia. The main selection criterion, as outlined in the above sections, is based on an empirical puzzle – the variable pattern of elite accommodation across ethnic lines in different policy areas. Despite contextual and political factors suggesting that the opposite should hold true: that political elites in Bosnia should accommodate less than elites in Macedonia, this is not the case. In this sense, Mill's method of agreement as
13 Chekel, 'It's the Process, Stupid!'
14 Peter Hall, 'Policy Paradigms, Social Learning and the State', Comparative Politics, 25 (1993): 275. 'most different' cases is applied to determine the comparison between the cases.15 At the policy level, Mill's method is applied within each country case. In both countries, two highly sensitive and contentious policy issues for ethnic relations are compared. However, despite their equally sensitive and problematic nature, the outcomes in terms of political elite accommodation differ – one case sees compromise and the other sees resistance. The policy cases analysed are 'most similar' cases where the variation in outcomes is explained.
The policy cases in both countries are also selected as 'most difficult' cases for ethnic accommodation. In order to explain which factors lead to greater or lesser ethnic accommodation in divided societies, the policies analysed need to be difficult cases for accommodation, such as policy areas that have caused much ethnic tensions and elite disagreements in the past.16 If a mutually acceptable solution was within easy reach, then accommodation would have been the most likely outcome, and the explanatory variables could not have been expected to have had a major impact. Moreover, the in-country and cross-country comparisons between similarly sensitive policies that have led to accommodation and to resistance, further helps isolate the effect of those explanatory factors present in the former cases. This facilitates the analysis of the links between the explanatory and dependent variables and enhances the validity of the findings.
In Bosnia and Herzegovina, the policy cases included are: police reform and military reform. Both the police and the army in Bosnia were heavily involved in the conflict as well as in committing violence, war crimes and ethnic cleansing. Their continued functioning in each of the entities had been a difficult issue in domestic politics since the end of the war. Reforming the ethnically divided police and military forces and placing them under common state control was bound to raise nationalist passions and accusations based on war-time actions. However, while police reform saw little elite accommodation and the political elites from the three ethnic groups could not come to an agreement, military reforms proceeded with less resistance across ethnic lines, despite the sensitive nature of security sector reforms. The country now has a centralised country-wide army and military structure.
In Macedonia, minority education policy has been contested continuously between the Macedonian and Albanian elites since 1991. In 1994 the establishment of the unrecognised Tetovo University led to police violence and one casualty. Decentralisation has been similarly problematic, after ethnic Macedonian politicians, fearing attempted secession, centralised the state in 1991 and rejected subsequent demands for decentralisation. When the municipal offices in Albanianmajority towns of Gostivar and Tetovo displayed Albanian flags in 1997 the police intervened and there were several casualties, as well as the imprisonment of the
15 Todd Landman, Issues and Methods in Comparative Politics: An Introduction
(Abingdon: Routledge, 2003), pp. 70–71.
16 George King, Robert Keohane and Sydney Verba, Designing Social Inquiry:
Scientific Inference in Qualitative Research (Princeton: Princeton University Press, 1994). mayor of Gostivar. Yet after 2001, reforms promoting decentralisation have in general proceeded less contested and resisted by political elites of all ethnic groups, while minority education remains an intractable problem despite numerous efforts to find a mutually acceptable and sustainable solution.
Table 1.2 Policy cases in Bosnia and Macedonia
Policy cases Towards ethnic accommodation Towards ethnic resistance
Bosnia Military reform Police reform
Macedonia Decentralisation reforms Minority education
The four policies shown in Table 1.2 are the main focus of the empirical analysis in the book and the subject of in-depth process-tracing analysis in the following chapters. However, examples from other policies and policy issues are at times also used, although not analysed as thoroughly, wherever there is need to illustrate a point and further support a claim that arises from the analysis of the four main policy cases.
Data Collection and Availability
To conduct valid analysis of ethnic accommodation among political elites, empirical data were collected and analysed for each of the stages of the policymaking process. This enables tracing the mechanisms leading from one stage of the policy-making processes to the next. Given the design of the dependent variable presented in the above sections, no significant problems with data availability and collection were encountered. The documents necessary for answering the above questions are public and official documents (legislation, legislation proposals, reform strategies, records of government and parliamentary debate, voting records, policy evaluations and reports) and are available to the wider public, either online or from archives upon request. Additional data were collected through interviews with representatives from the political leadership in each of the two states. The interviews were used to collect data about the personal experience and perceptions of the participants in the policy process that are not captured in official policy documents and records, but which help one to better understand the behaviour of a certain actor in the policy process. The questions used during the interviews were open-ended, as the interviews were aimed at the individual and their personal experience and narrative about the policy processes in which they had participated.17
The interviewees are mostly still active politicians and civil servants, therefore widely accessible and available for interviews. In addition, interviews were also
17 Joel D. Abermach and Bert A. Rockman, 'Conducting and Coding Elite Interviews',
Political Science and Politics, 35 (Dec 2002): 673–6. conducted with representatives from the international community (EU and NATO officials) and with local ambassadors. The selection of interviewees was on the basis of their relevance to the policy in question, through the snowballing technique and reputation referral among their colleagues.18 In total, 40 interviews were conducted over a period of six months between April and October 2010, evenly distributed between Bosnia and Macedonia.19 Interviewees included politicians, journalists and diplomats, all interviewed in their native language. In addition, speeches, statements, interviews and other public remarks available either through media archives or through government's and ministries' press archives were used, while media archives (most of which are online for the period studied) were used for accessing commentaries, opinion pieces, features and news articles relating to public debate and discourse on the policies that are analysed. Finally, the EU documents and statements, also available from EU's online archives, were used.
The next chapter of the book provides an overview of the available academic literature related to the question addressed. It also maps and reviews the existing knowledge on this topic and evaluates its usefulness and applicability for this study. Chapter 2 further distils the available knowledge, arguments and hypotheses about the effects of the explanatory variables, elaborating on the definitions and theoretical approach used throughout the analysis.
In Part II, Chapters 3 and 4 provide a brief overview of the political and institutional history of both states, which illustrates the institutional and ideologico-political legacies that shape the current political environment in which Bosnian and Macedonian political actors operate. Though the two states share part of their political and institutional history as federal units in the Socialist Yugoslav federation, they took separate trajectories after independence. These trajectories and the effects they had on the interactions between political elites in the policy process are traced for both states from the early 1970s, when the last Yugoslav constitution was adopted, until the end of the ethnic conflicts, in 1995 in Bosnia and Herzegovina and in 2001 in Macedonia.
In the third part of the book, Chapters 5 and 6 investigate the success stories of post-conflict politics in the two countries. The focus is on two policy issues that were highly sensitive for inter-ethnic relations in the conflict aftermath, but were successfully resolved and ceased to cause ethnic tensions. In Bosnia the success of military reforms between 2002 and 2007 is examined. In Macedonia the focus
18 Kenneth Goldstein, 'Getting in the Door: Sampling and Completing Elite Interviews', Political Science and Politics, 35 (Dec 2002): 669–72.
19 The number of interview fits most social science guidelines about optimal 'sample size' for elite interviews, see: Steinar Kvale, Interviews: An Introduction to Qualitative Research Interviewing (Thousand Oaks, CA: Sage, 1996). is on the successful decentralisation reform between 2002 and 2009. Both policies provide an insight into the drivers of successful ethnic accommodation.
Part IV of the book looks at the failed attempts to resolve ethnic problems in post-conflict politics in Bosnia and Macedonia. Chapter 7 focuses on police reform in Bosnia between 2004 and 2008, and Chapter 8 examines minority education policy in Macedonia between 2002 and 2010. Despite repeated efforts at reform and finding a mutually acceptable compromise, these issues remain unresolved and continue to strain inter-ethnic relations. The two chapters look at the factors behind the failed efforts of domestic politicians to find a compromise over these issues, contrasting them to the two policy issues examined in Part III of the book. The final chapter of the book provides an overview of the main findings of the earlier chapters and summarises them with reference to the main arguments of the book. The findings of the book highlight the importance of domestic factors in inter-ethnic cooperation in post-conflict Bosnia and Macedonia. The chapter further discusses the relevance of the findings in relation to other scholarly work on the region and offers some thought about the prospects and nature of the
democratic regimes in the two countries.