This chapter compares and synthesises the findings of the policy cases analysed in the book. It compares the factors that led political elites towards more accommodation over some ethnically sensitive issues to those that led towards ethnic resistance and tensions. It also discusses factors that enable and facilitate the work of post-conflict institutions and policy tools in Bosnia and Macedonia. Based on these findings, this chapter argues that in both countries formal powersharing mechanisms are necessary for accommodation. However, although some such mechanisms, such as veto powers, do tend to facilitate accommodation, others such as functional and territorial autonomy can lead to more resistance. Informal practices tend to support accommodation and supplement the work of formal institutions, especially when they create spaces for building crosscutting group identities, or when they cause ethnic confrontation to take place in the informal domain rather than in the public. Finally, within the institutional constraints of post-conflict political systems, external actors' influences tend to enable ethnic accommodation, but on some occasions can incline domestic actors towards resistance and continued ethnicisation.
The chapter contextualises the cases by discussing the wider political and social environment in the two countries and examining the post-conflict trajectories they have followed from conflict towards EU and NATO membership. The final sections of the chapter discuss the wider implications of the findings of this book, relating them to Bosnia and Macedonia's prospects for consolidating democracy and managing ethnic diversity in a peaceful and efficient manner.
Post-conflict Bosnia and Macedonia
Upon the dissolution of the Yugoslav federation in 1990–91, both Bosnia and Macedonia were faced with the challenge of managing inter-ethnic relations in their ethnically diverse societies. After initial failures to reduce ethnic tensions and fully address ethnic problems in society, both states went through violent ethnic conflict: immediately after independence in 1992 in Bosnia; some years later, in 2001, in Macedonia. The conflicts led to the restructuring of the institutional and political architecture of each state, with the intention of achieving greater inclusion and access to political power for all ethnic groups. The restructuring was initiated by the Dayton and Ohrid Agreements, which put an end to the violence in each state and contained provisions for the new post-conflict constitutional structures. Since the end of the ethnic conflicts, Bosnia and Macedonia have not lapsed into violence and conflict has not recurred. However, ethnic tensions have remained. As previous chapters discussed, some issues and policies remain highly sensitive and problematic for ethnic relations, despite efforts to come up with mutually acceptable solutions. In seeking an explanation for the continued ethnic contestation of some policies, this book investigated the factors that led to greater ethnic accommodation over some policies and those that led to continued ethnic resistance over others.
In each country a set of power-sharing arrangements was introduced by its postconflict agreement in order to facilitate democratic and accommodating politics. The policy cases in this book analysed two in particular: the effects of veto powers; and territorial and functional autonomy.
In line with the theoretical claims examined in Chapter Two, the findings in the empirical chapters confirm that veto mechanisms had a significant impact on the policy process.1 Veto mechanisms prevented each side from being outvoted and, in particular, gave the numerically inferior side the guarantee that its interests would not be neglected. Veto mechanisms affected the outcome of the policy process in two ways: directly, through the invoking and using of vetoes, and indirectly, through implicit threats to use the veto during negotiations of policy proposals. Direct and indirect uses of the veto mechanism appear to have had different effects on ethnic accommodation. Direct veto use led towards more ethnic resistance, while its indirect use led towards more accommodation.
During police reform efforts in Bosnia in 2005–06, the Serb side directly used its veto to prevent the adoption of a policy proposal it had considered harmful to its interests. This blocked police reform, but allowed Serbs to protect their interests in the policy process. However, this made accommodation more difficult later on. Even when, after the initial rejection, High Representative Paddy Ashdown amended his reform proposal, having voted against police reform only weeks earlier made it more difficult for the RS parliament to adopt. In Macedonia, the abandoning of a proposal to make Macedonian language compulsory in schools was also an implicit recognition of a direct veto. After the Albanian party in the government coalition had rejected the proposal it would have been futile to introduce it in parliament, where it would have failed to get enough votes for a double majority. The government has not attempted to re-introduce the measure in parliament since then, despite sporadic talks about renewing efforts on the education strategy.
1 F. Bieber, 'The Challenge of Institutionalizing Ethnicity in the Western Balkans: Managing Change in Deeply Divided Societies', European Yearbook of Minority Issues, Vol.3 (2003–4), p. 93. Indirect veto use was more frequent in the policy cases analysed. The groupveto entitlement usually encouraged all sides to negotiate and compromise until an acceptable solution had been found. In Macedonia this happened through the intra-coalition agreement between the Macedonian and Albanian coalition partners. Prior to proposing legislation to recognise the legality of Tetovo University and introduce decentralisation, the two coalition partners had already agreed to the contents of reforms. In Bosnia the availability of both group and entity vetoes facilitated negotiations in the Defence Reform Commission, where consensus solutions over military reform were sought. Thus, the indirect use of the veto encouraged compromise and concessions, enabling political elites to reach accommodating outcomes. While most academic work refers to direct vetoes and their harmful effects,2 widening the focus of analysis to include indirect uses of veto mechanisms shows that the opposite is often true. When used indirectly, veto mechanisms can result in more accommodating policy proposals. These findings are true in both countries, despite the differences in the nature and extent of veto mechanisms in Bosnia and Macedonia.
Unlike veto powers, the effect of territorial and functional autonomy on ethnic accommodation at the policy level is less clear. The findings from the Macedonian case studies suggest that territorial rather than functional autonomy is more conducive to accommodation. In the case of minority-education policy, functional autonomy was established in 2001, allowing Albanians to run their own education system in their mother tongue. This created an ethnically exclusive policy domain where decisions made by Albanian leaders affected only Albanians. Such policy design, although agreed at the elite level, did not reduce ethnic tensions and animosity in education. Instead it led to greater segregation and tensions between Albanian and Macedonian youths, as encapsulating each group in its own exclusive education system meant it had no exposure to the other. These divisions further prevented the rise of cross-cutting alternative identities and cleavages in this area, maintaining ethnicity as the main line of political contestation.
Increased territorial autonomy in decentralisation was not accompanied by ethnically exclusive rights and benefits, but thanks to the multi-ethnic composition of the municipalities both groups benefited equally from increased local self-government. Gradually, this led to the relaxing of ethnic relations at the local level and the de-ethnicisation of decentralisation policy. It has commonly been argued that territorial autonomy and decentralisation tend to pave the way to secession.3 However, when such autonomy was implemented based on
2 See discussion in Chapter 2; Also see a more nuanced discussion in J. McElvoy, 'We Forbid! The Mutual Veto in Power-Sharing Democracy', in Power-Sharing in Deeply Divided Societies, ed. J. McElvoy and B. O'Leary (Philadelphia: University of Pennsylvania Press, 2013).
3 D. Rothchild and P. Roeder, 'Power Sharing as an Impediment to Peace and Democracy', in Dilemmas of State-building in Divided Societies, ed. D. Rothchild and
P. Roeder (Ithaca: Cornell University Press, 2005). non-ethnic criteria, it led to greater accommodation than did the creation of ethnically exclusive policy domains, which tended instead to pull populations and leaders further apart and fuel further ethnic tensions. Moreover, as regions are rarely fully ethnically homogeneous, ethnic diversity at regional or local levels would allow more communities to feel the benefits from increased local or regional authority. However, territorial autonomy in Macedonia is limited to decentralisation and some increased powers for local governments, which falls short of the extensive territorial autonomy that groups enjoy in federal states such as Bosnia. Therefore these findings cannot easily be generalised for other cases of more extensive territorial autonomy.
In Bosnia, territorial and functional autonomy are combined in the extensive powers and authority of entities and cantons. Both the police reform and the military reform were attempts to reduce this autonomy and transfer their functional control to the central level. Both cases showed the difficulty of reducing functional autonomy, since removing exclusive policy domains from groups' control tends to result in defensive behaviour by politicians, who see it as an attempt to limit their power. Ethnically exclusive policy competencies tend to turn policy-making into a zero-sum game for ethnic political elites. While for the average citizen the services that are available might not differ much in each case, for the political elites having exclusive control over a certain policy is more desirable than having to share it with other political actors. Shared control over policy areas implies more compromises and more concessions than exclusive control. This further elucidates the reasons behind the success of military reform in Bosnia. Although the military was territorially divided into two entity forces, it was not under the exclusive control of ethnic political elites, but under international control. This made it easier for entity politicians to centralise it, as the transfer did not diminish their personal and group power, unlike giving up their exclusive control over the police, which would have tangibly reduced their political power.
Seeing politics in the zero-sum terms of personal and group political power disinclines political actors from accommodation. The findings from previous chapters suggest that functional autonomy often tends to result in zero-sum perceptions of politics among local political elites and leads to greater resistance across ethnic lines. Thus, there is some evidence that functional autonomy provides local ethnic elites with 'institutional weapons' to resist the majority. Contrary to Rothchild and Roeder's claims though,4 there is almost no support for the argument that those 'weapons' are used to further increase the power of local elites, since there were no efforts to expand functional autonomy to additional areas.
Overall, the findings suggest that power-sharing was important for ethnic accommodation, but that, in and of itself, it was not a sufficient condition for successful ethnic accommodation over contested issues in the two states. Instead of looking at the existence of power-sharing as an explanation, this book suggests that it is rather the nature and specific elements of power-sharing arrangements
4 Rothchild and Roeder, 'Power Sharing as an Impediment to Peace and Democracy'. that can explain the outcome of particular policy processes. Veto mechanisms and inclusive executive coalitions, for instance, tend to facilitate compromise solutions, while functional autonomy and policy domains under exclusive group control tend to hinder accommodation. Similarly, whether a specific power-sharing arrangement is of a temporary or permanent nature can affect the perceptions and behaviour of political actors towards more or less accommodation. Although some power-sharing scholars advocate temporary power-sharing,5 the findings in this book suggest that the threat of revoking power-sharing and related fears of marginalisation in the policy process are more likely to result in resistance than in accommodation. Therefore discussions about abolishing power-sharing institutions in Bosnia and Macedonia do not substantially contribute to debates about reforming the post-conflict political system. Rather, discussing how to adapt specific mechanisms to better work in the national and social context in each country seem to offer a better solution to improving the accommodating potential and efficiency of those countries' policy processes.