Abstract
Both Bosnia in 1995 and Northern Ireland in 1998 were extremely fragile in the immediate aftermath of brokered peace negotiations. Each instituted a form of consociationalism—a government that institutionalizes a voice for each ethnic group—as an element of brokered peace. In this article, I examine Bosnian postwar governance with comparative insights from Northern Ireland. Bosnia was the recipient of a large amount of international aid. While this aid was crucial to the initial state-building effort, the problems Bosnia now faces are due to its consociational governance structure. Some of the group-based aspects of consociationalism are in tension with individual rights, a problem that cannot be addressed by aid alone.
Keywords
The small country of Bosnia-Herzegovina, with a population of less than 4 million, has received a substantial amount of attention and aid from the international community since the war there between 1991 and 1995. Foreign aid was received in two ways: (1) direct intervention by internationals in brokering and then supporting the Dayton Peace Agreement of 1995 and (2) funding for a variety of projects to aid Bosnia’s democracy, infrastructure, and economy. Both types of aid were crucial to making peace possible, and the need for some support remains. However, aid cannot reconcile all of the problems that can emerge in the governance of states with mixed ethnic or religious populations. In essence, the effort to guarantee representation of different groups requires the institutionalization of group differences. 1 The Dayton Agreement was intended to end a fierce and bloody war, but it also became the default setting for consociational governance institutions. Consociationalism is a form of government that is intended to institutionalize a voice for each group, and it relies heavily on cooperation among elites. Many critics argue that the agreement is not suited to long-term governance structures, but changes to these institutions that could be accepted by all parties have been elusive (Chivvis 2010, 66–7).
A similar set of circumstances can be found in Northern Ireland, where the Good Friday Agreement of 1998 set the foundation for its own consociational structures. It serves as another example of postconflict consociational institutions within a European setting. 2 A “country” within the United Kingdom’s complex devolved power structure, Northern Ireland also received direct intervention to calm its conflict, both internationally and from the UK government. 3 Funding was provided in a variety of forms from the UK government, as well as from international, UK-based, and Irish NGOs (nongovernment organizations). These interventions were crucial in setting the foundations for peace. As in Bosnia, the long-term logical problems of its consociational structures remain and cannot be addressed by aid alone. But at the same time, the institutional tensions in Northern Ireland appear to be less debilitating to the workings of governance than they have in Bosnia.
This article focuses on Bosnia with some comparative insights from Northern Ireland. Both places were extremely fragile in the immediate aftermath of their brokered peace negotiations, in Bosnia in 1995 and Northern Ireland in 1998. Both have moved to a far more robust status over the past two decades. While all-out war in each is now unlikely, both share traits of high segregation among groups, low-level conflict, frequent government crises, and divisive rhetoric from both sides. I first give an overview of some of the extensive aid that was distributed in Bosnia, highlighting how it contributed to the foundation and persistence of the Dayton Agreement. I then examine how aid was related to the state-building and transition effort; I follow with an outline of the premises of consociational democracy, with an in-depth examination of how these structures have created a paradoxical situation for Bosnia. In addition, I sketch some of the aspects of consociationalism in Northern Ireland, noting how it differs from Bosnian governance. I then evaluate whether some of these differences might account for a more stable and workable means of governance in Northern Ireland, and whether some of these considerations might be applicable to Bosnia.
Aid and Influence
Actors in the aid trajectory
Bosnia-Herzegovina has received both focused intervention and an extremely large amount of international aid. The intervention began during the country’s civil war and involved a NATO-backed force that included thirty-six countries, with strong involvement from the United States (McMahon and Western 2009, 71). Bosnia became the recipient of the most aid per capita since the Marshall Plan (Sebastian 2010, 2; Pasic 2011). The first stage of international aid consisted of direct intervention in the form of the Dayton Peace Agreement, brokered in November 1995. The consociational structures that emerged within the Dayton framework were the end result only after several alternative institutions proposed by internationals failed to reach agreement among all parties. While much criticized for its complexity, Dayton offered the only balance of concessions to each side that succeeded in producing a lasting agreement.
With the signing of the Dayton Agreement in November 1995, the second stage of international intervention began—taking the form of aid packages for the reconstruction effort. Not all groups in the conflict were included in these discussions. Because the Serbs were widely perceived by many potential donors as having had a strong role in the onset of the conflict, they were initially excluded from aid packages. 4 Conditions for future Serb inclusion in aid were imposed, including the capture of war criminals who were believed to be in the Bosnian Serb entity (Los Angeles Times 1996). After the end of the war, international aid began to focus increasingly on governance, infrastructure, and the economy, although security still remained a part of these budgets. United States involvement has decreased incrementally, and by 2011 was allocated for the following priorities: 40 percent for peace and security, 39 percent for “governing justly and democratically,” and 21 percent for economic growth (United States Department of State 2012; Woehrel 2013, 9). 5 USAID, the European Union (EU), and the UN also focused efforts on minority returns—the resettlement of individuals who had been displaced from their homes by the war, but who returned as local ethnic/religious minorities (USAID 2013a; Sebastian 2010, 5–7).
While the U.S. involvement has been significant (just short of $2 billion), the largest donor to Bosnia overall has been the EU. Between 1991 and 2010, the EU donated €6.8 billion, or $8.85 billion. It has focused especially on reconstruction, governance, economic and social development, regional cooperation, and harmonization with EU norms (Dimireva 2010). 6 By 2003 the EU had taken control of the police mission there (previously run by the UN), and in 2004 NATO transferred its Bosnia mission to the EU (Chivvis 2010, 57–8). The EU has been working with Bosnia on fulfilling the criteria for EU membership, although these discussions have hit a recent impasse over constitutional matters (International Crisis Group [ICG] 2011a) that cannot be addressed by aid. As discussed more below, these obstructions relate to the logical problem of guaranteed representation for specific identity groups within a democracy.
In addition to the United States (USAID) and the EU, other significant donors have included the World Bank, Sweden, the Netherlands, and Norway (World Bank 2004; Sebastian 2010, 2). The United Nations Development Program (UNDP) has also had a significant presence (UNDP 2009). Aid began to decrease significantly in the wake of the crisis in Kosovo in 1999. A calculation of the overall amount of funding received by Bosnia is a difficult one, due to an absence of comprehensive and coherent data from a variety of sources, which range from governments to international organizations to NGOs and religious groups (Sebastian 2010, 2–3). An influx of Islamic groups from Saudi Arabia and other countries contributed aid to build mosques and for social services. But not all of these groups had the same goals, and some represented more extremist elements that contributed to the creation and support of local groups. This international Islamic presence is a well-known fact on the ground. However, it is quite difficult to document the extent and exact nature of this kind of involvement. Moreover, because contributions from religious groups overall can contribute to increased group divisiveness, a question remains with regard to how to establish general categories for international involvement when different actors support different goals (ICG 2013).
Aid strategies
There has been a great deal of discussion regarding the strategies that have worked and not worked in Bosnia in terms of the logistics of distributing aid. Some of the more common modalities of aid distribution have historically been partnership agreements, in which an international donor agrees to fund a project that is headed by several local actors. This strategy has been used by several international NGOs in Bosnia, and some detailed discussions have emerged from some of the experiences of USAID, which operates according to these practices. Those granting aid also engage in partnerships with other donors (USAID 2013b). Another strategy is to offer loans, a method that has been more visible recently in the EU’s provision of €50 million in the form of loans to Bosnia (European Commission 2013). While perhaps understandable in light of the EU’s financial difficulties, the effectiveness of this strategy remains questionable, especially in the long run. Bosnia’s economy is in terrible shape (McMahon and Western 2009, 74), and unemployment is currently around 25 percent (Rohde 2012). The means by which the state might be able to pay back these loans remains unknown, and could do some damage to hopes for Bosnia’s economic recovery in the future.
Aid distribution has appeared to work best when donors take a view that they are committed to their programs for a long period, rather than searching for a “quick fix.” Choosing a few programs and sticking with them is usually better than funding several different initiatives in the hope that some might bear fruit. Programs should also be designed according to needs on the ground and with the input of locals, rather than being designed exclusively in office boardrooms far away. In addition, the project approach itself may create problems, as it encourages short-term thinking. More long-term funding of organizations to conduct work is likely to be a more effective strategy (Rohde 2012; Sebastian 2010, 9–10, 14). It is crucial to first observe the local context and listen to what those with local knowledge think must be done, as well as pay attention to potential pitfalls as they outline them. In addition, it is important to know how potential obstacles are understood by local partners and to design an incremental strategy that adapts to the connected nature of different problems (Huddleston 1999, 149–50; Stroschein 2002). Adaptability and flexibility are crucial. For example, an adherence to a Western-based management style can be a disadvantage, as it may prove a hindrance within the local culture. In addition, sometimes organizations must adapt to the skill sets of the personnel who are actually available to take on the project, because many otherwise skilled individuals cannot afford to leave their full-time work for a short-term project commitment (Huddleston 1999, 154–5; Sebastian 2010, 11–2).
Perceptions of aid
Interviews and surveys of Bosnians demonstrate mixed opinions on aid. For example, in the area of democracy assistance, respondents understand that many civil society organizations would not exist without international aid. However, with regard to actual progress on democracy itself, respondents tend to reflect more pessimism due to the perceived difficulties of achieving actual change (Sebastian 2010, 7–8). According to 2010 poll results from the Gallup Balkan Monitor, 41 percent of respondents in Bosnia said they were unsure if the role of the international community over 15 to 20 years has been harmful or helpful. Only 32 percent thought that it had been helpful, and 21 percent thought that it had been harmful (Gallup Balkan Monitor 2010, 6). Bosniaks and Croats were more likely to favor an international role (59 percent for Bosniaks and 50 percent for Croats), while Serbs tended not to favor international oversight (77 percent) (Gallup Balkan Monitor 2010, 7).
Some of these survey results regarding aid may reflect a pessimism by Bosnian citizens regarding the overall state of their country. A 2010 National Democratic Institute (NDI) poll of Bosnians indicated that a staggering 87 percent thought their country was moving in the “wrong direction” (NDI 2010, 5). A total of 84 percent said that their economic situation had either “stagnated” or “deteriorated,” and 41 percent cited unemployment as the top problem for the country, with corruption as the second, trailing distantly at 13 percent (NDI 2010, 6, 9). It is clear that while aid has helped to move the economy forward, it cannot solve deep structural problems in the economy under the Dayton Agreement, including high levels of long-term unemployment. In February 2014, large and violent protests erupted throughout Bosnia, as individuals attacked symbols of government to protest their economic impoverishment. As a result of the Dayton Agreement, leaders of each ethnic group have been able to enrich themselves in economic reform processes such as privatization. Few benefits of the new market economy have in fact trickled down to average individuals (Štiks 2014; Hemon and Mujanovic 2014; Vachudova 2014). As Vachudova has noted, the economic reforms that intertwined with the Dayton Agreement made it a particularly “predatory arrangement” (Vachudova 2014). International aid has provided some direct employment for individuals who work with internationals, but this provision is also one aspect of dependency on aid that may be problematic in the long term. Individuals who might have otherwise become civil society actors of their own accord became part of this international aid machine, potentially depriving the country of some grassroots innovations (Bieber 2002, 27–28).
These problems demonstrate that large amounts of aid and international commitment are not enough to build long-term stability in Bosnia. At the same time, conflict has not resumed, despite the dire economic circumstances in the country. There could be some nonaid reasons for Bosnia’s relative quiescence. One such factor might be war fatigue among the population and a desire to simply lead “a normal life” (Pickering 2007, 1). Alternatively, it could be that the carrot of EU and NATO membership has provided an incentive to avoid a resumption of intergroup conflict. As of 2010, 69 percent of the population considered EU membership as a good thing (Gallup Balkan Monitor 2010, 3). In June 2008 a stabilization agreement was signed with the EU (Chivvis 2010, 61) to formalize increased cooperation.
However, the notion that the EU has been the driver of Bosnian stability faces several logical challenges based on other events that began to unfold in the same year that the cooperation agreement was formalized. First, Kosovo also declared independence in February 2008, an act that prompted the Bosnian Serb leader Milorad Dodik to declare that the Republika Srpska should also be able to gain independence from Bosnia. Over the next few years Dodik continued to push this line, contributing to increasingly fragmentary politics in Bosnia despite EU efforts to keep the system together (ICG 2009, 2011b). In addition, a constitutional crisis that was sparked by the European Court of Human Rights is currently preventing Bosnia from moving further in the accession process, a situation considered more fully below. In addition to these problems, the EU’s economic crises over the past few years have rendered it not only less powerful but also less of a shining light to Bosnians. One of the main reasons for wishing to join the EU is its potential role in improving the Bosnian economy (Gallup Balkan Monitor 2010, 3)—and the EU’s credibility in this role has suffered some setbacks over the past few years.
As this overview shows, Bosnia presents a case study of high levels of international involvement and aid, both of which it has received since 1991. The country has certainly benefitted from these interventions in terms of economic, political, and social development, as well as minority refugee returns (United Nations High Commissioner for Refugees [UNHCR] 1998). These contributions have enabled the ongoing relative stability of the Dayton Peace Agreement. At the same time, a logical paradox in its governance structures presents serious problems that aid alone cannot overcome (in addition to its serious economic problems).
Aiding States by Building Consociations
The Dayton Peace Agreement was reached after years of devastating conflict (1991–1995). In addition to the staggering number of dead and injured the conflict produced, the conflict displaced “more than half of Bosnia’s population, 2.5 million people” (Pickering 2007, 29). 7 Peacekeeping forces also came under attack, which led to NATO airstrikes against Serb forces in 1994 (Randal 1994). During the conflict years, various failed attempts at agreements were made. From the vantage point of several years later, the Bosnian state structures that resulted from the Dayton Agreement are easy targets of criticism, as they encourage frequent stalemates by giving voice to each of the three constituent groups (McMahon and Western 2009; Hayden 1993). But it is important to keep in mind that the agreement was intended to end the fighting (Chivvis 2010, 48–9), and it has succeeded in preserving peace for nearly 20 years.
Only with international intervention could an agreement like Dayton have been brokered. When a country with severe identity conflicts reaches a point at which leaders of each group have stopped talking to one another, progress can only be made via third-party brokerage. The logic of this dynamic is as follows. In divided societies, a crucial mechanism for preserving peace requires that elites of each group maintain lines of communication with one another to negotiate a means out of potential political conflicts (Fearon and Laitin 1996). If this cross-group elite communication were to break down, conflict would inevitably increase. Without elite communication to broker solutions to group grievances, populations have no choice but to mobilize to express those grievances (Stroschein 2012, 21–23). It is at this point that internationals and international aid can intervene most effectively, to restart elite communication and to establish institutions to preserve the machinery of group communication via elites (with support of and thus legitimation from group populations).
The support of group populations is crucial to a successful elite dialogue. Elites in the negotiation must have a relationship of resonance with “their” masses, meaning that the aims of these elites resonate among their publics. Self-styled elites without this resonance or legitimacy from their populations will be powerless to end disputes, as they will not address the actual grievances that drive popular mobilizations. The requirement of both elite negotiation and popular resonance provides a caveat to the international community to exercise care when choosing elites for participation in cross-group dialogues, “training,” or other programs intended to reduce conflict. In addition, internationals should keep in mind that in conflict-ridden situations, those elites with whom they work may become discredited within their own populations as being “co-opted,” or for receiving benefits from internationals that do not reach their own communities. As one example, an effort to broker talks among elites of different ethnic groups in Romania in 1993 by an international NGO resulted in a backlash against those elites who participated (Stroschein 2013). In some cases, extremist elites may be those who have the most popular resonance, thus requiring internationals to bring even these more extremist elites into negotiations (Stroschein 2012, 22).
If a peace agreement is to hold, all parties must feel that it will provide them with some concessions. Peace agreements easily become the foundation for new political institutions, especially in democracies, as conflicting groups must have a common starting point for the creation of rules for governance. Negotiations for a successful or lasting peace agreement often identify the one equilibrium point on which the different groups can agree. Minorities will want protections against majority tyranny, and majorities will tend to wish to exercise democracy as their “birthright,” often preferring majoritarian democratic institutions. Balancing out these stances is not easy, but consociational institutions provide a means to give some concessions to different sides.
Consociational institutions operate on the principle that to give each group voice in politics, they must be separated into segments of representation. In addition, elites of each group are given strong responsibilities to negotiate with elites of the other group(s) to resolve the most controversial political matters. Rather than decision-making on a majoritarian principle, elites of the different groups must agree on political decisions (Andeweg 2000, 509).
There are four primary aspects of consociational governance structures, as outlined by Arend Lijphart (1977) in his seminal discussion of consociations as an institutional category. First, government must include a powerful “grand coalition” of political leaders of the different segments of society, placing a strong focus on these elites to resolve potential group disputes. Second, there is a “minority veto” mechanism, by which minorities can block decisions that they find harmful to their identity, without needing power in numbers to do so. Third, there should be proportionality among groups in representation across several areas. For elections, the recommended electoral system is proportional representation (PR), so that groups are represented in proportion to their demographic percentages in the population. Divided societies that are just exiting violent conflict tend to exhibit party cleavages that automatically divide along group lines. Voting for these identity-based parties in proportion guarantees voice for each group in a way that majoritarian electoral systems (as in the United States and the UK) cannot. In addition, appointments to government posts, the civil service, and the police should use a proportional principle. In this way, members of each group can see individuals who are “like” them in posts across the state apparatus. In addition, there should be proportionality in the distribution of financial resources, for example with schools of each group funded proportionally, rather than only those of the majority receiving support. Fourth, there should be segmental autonomy, in which a form of federalism or decentralization preserves minority control over some local government structures in the country. This decentralization should include some financial autonomy for the segments even if this requires central subsidies, reflecting that there should be some proportionality in the distribution of financial resources (Lijphart 1977). 8
These insights were useful to internationals in brokering the agreements in Bosnia and Northern Ireland, partly because the strong emphasis on representation provided a clear means to grant concessions to each side and to move toward durable final settlements. This set of institutions provides a means for group elites to negotiate their differences and for each group to have some representation in the structures of the state. It can thus be very effective in preserving minority aims and identities. However, majorities may chafe at the fact that some provisions can lead to stalemates, such as the minority veto. In addition, over time and as the memory of conflict wanes, publics can begin to lose some satisfaction with this model of governance due to the primary role given to elites and the strong potential for stalemates. In addition, their group-based premises may begin to be viewed as reinforcing group identities and extremist elements, as well as hampering the expression of individualism and nongroup concerns (Andeweg 2000).
Many critiques of Dayton relate to some of its consociational features. Indeed, as noted in the discussion in the next section, there is room for considering where some of these aspects might be modified. A primary critique is the prioritization of ethnicity and territory in terms of how rights are allocated throughout the state for the Bosniaks, Serbs, and Croats. This prioritization does two negative things: (1) it omits those who are not part of these three groups, such as the Roma, Jews, and others; and (2) it requires an ongoing identification with the groups that fought against each other in the war, which does not facilitate building interethnic ties that can promote peace in the long term. Serious attempts to change these institutions have been ongoing since 2006, and have run aground on the same problems as the pre-Dayton attempts at negotiations. Up to this point, the two sides cannot agree on a workable solution outside of the Dayton parameters.
A similar case to Bosnia is that of Northern Ireland, where the consociational institutions of the Good Friday Agreement of 1999 have preserved relative peace for 14 years. Its codification of groups as political forces has led to some of the same problems as in Bosnia, with an inability for the country to establish widespread inter-religious ties between Catholics and Protestants. But Northern Ireland is not constrained by the same territorial considerations as is Bosnia—group membership is based on a personal principle and does not have an official territorial basis in the way that Dayton has demarcated for Bosnia. The next section outlines a paired comparison of Bosnia and Northern Ireland, to examine how the consociational aspects of both agreements have provided both a help and a hindrance to these societies. This comparison also notes some of the differences between the particular incarnations of consociationalism in each place, to explore which specific institutions might present a potential obstacle to long-term stability. Many of the debates on consociationalism have focused on the pros and cons of its general application, rather than considering the implications of its specific institutions (see Lijphart 1977; Lustick 1979). This inductive examination is a first step to producing general statements (or hypotheses) about the effects of these institutions that can be examined in a variety of other places.
Consociational Democracy in Practice
Bosnia did not have a history of its own statehood. It was historically a province of the Ottoman Empire that was later occupied by the Austro-Hungarian Empire in 1878 with the weakening of the Ottomans (Magocsi 1993/1995, 80–82; Jelavich 1983a/1995a, 360). Serbia had never been fully satisfied with this transfer, as it wanted more of its own influence over the province (Jelavich 1983b/1995b, 59). The assassination of Austro-Hungary’s Archduke Franz Ferdinand by a Serbian conspirator in Bosnia in 1914 was the spark that ignited World War I (Magocsi 1993/1995, 121). After the war, Bosnia-Herzegovina became part of the Serbian-dominated Kingdom of Serbs, Croats and Slovenes, or Yugoslavia (Magocsi 1993/1995, 128). During the Second World War, Bosnia experienced high levels of bloodshed due to guerilla fighting (Magocsi 1993/1995, 140), and after the war remained part of Yugoslavia until its referendum on independence in 1992, which was boycotted by the Serbs (Glenny 1992/1994, 165–66). Bosnia’s democratic constitution, established by the Dayton Agreement, thus was not built on a historical foundation of independent statehood; Bosnia, as part of Yugoslavia, had very little experience with democracy during the interwar period. Northern Ireland provides an interesting contrast as it is also a province of a larger entity, the United Kingdom, but it has had a longer history of democracy. Northern Ireland’s conflict with the United Kingdom could be framed as a perpetual tension that received heightened clarity with the independence of “the rest of” Ireland from the United Kingdom that took place between 1919 and 1922.
The consociational structures created for Bosnia and for Northern Ireland have broad structural similarities but also some specific institutional differences, as outlined below. Both the similarities and differences reflect some of the general problems that emerge from the fact that the consociational structures were intended to end conflict, rather than designed with decades of governance over the long term in mind. It is important to highlight these contradictions. Bosnia’s current challenges and governance crises emerge not from a lack of aid, but from competing imperatives with regard to its institutional structure. Some of the specific elements of the Northern Ireland institutions circumvent similar problems, making it instructive to examine both examples together.
Bosnia’s consociational structures and their implications
One of the main obstacles to previous peace agreements during the war had been disputes over territory. The Dayton Agreement established separate territories and governance institutions for groups, with the Serbs receiving 49 percent of the territory and the Muslims and Croats sharing a federated unit with 51 percent of the territory (Erlanger 1996). 9 According to figures from 2000, Bosniaks (Muslims) are 48 percent of the population, with Serbs at 37 percent and Croats at 14 percent (CIA World Factbook 2013). Despite the asymmetry of population groups and territory, it is not the territorial designations that have been challenged with the governmental crises. Instead, it is the structure of the governance institutions that has been the focal point of controversy, namely, the fact that only members of the three ethnic constituent groups might contest for presidential office. The Republika Srpska (RS) and the Muslim-Croat Federation (Federation) have very different internal governing structures. The RS remains quite centralized, while the Federation is divided into two further units, one with a Croat majority and one with a Bosniak majority. Reflecting the Serb desire to be as independent from overall Bosnian structures as possible, the RS and the Federation retain many governing powers, and the central government has historically had a quite weak center. Each of the two territorial entities has their own parliament and executive, but with an asymmetric design. The RS has a president and a unicameral parliament, and the Federation has a bicameral parliament as well as a president and vice president. The president and vice president of the Federation must be from the two different groups; if one is Croat, the other must be a Muslim.
The overall central level of governance involves a bicameral parliament, and the leadership of the parliament must represent each ethnic group. The executive is a three-member rotating presidency, one of the most unique features of Bosnia. Each of the three members must be selected from among the three ethnic groups, such that each group has a voice in the presidency. Presidency members are elected for four-year terms, during which they rotate through the main presidential office twice in eight-month stints. While in the main presidential seat, the member in office is expected to actively consult with the other two members on governmental matters (Bose 2002, 60–89).
Stalemates within the government have occurred, as would be expected in this type of consociational structure. However, the main sticking point of the constitutional design lies in the notion that to give voice to each group, the population must remain separated into different groups for political activities. One practice included in these institutions has been the selection of each group’s member of the presidency from within “its” territorial unit. However, this practice has meant that minority returns, for example a Bosniak who might have returned to her former home in the Republika Srpska, are not able to vote for the presidency. In 2000, the Bosnian Constitutional Court issued a decision that such restrictions that tied ethnicity to a specific territory were a violation of rights for individuals. The decision also implied more recognition for those who might be ethnic “others,” such as Roma, outside the three constituent ethnic groups (ICG 2002, 2–3).
This issue has remained unresolved, due to an inability to negotiate a compromise among the groups to address it. The only institutional foundation from which the groups can operate is a consociational structure with an “emphasis on state-building rather than human rights” (ICG 2012). The United States engaged in an effort to try to reform the constitution by proposing a weak and indirectly elected president, a move that failed in 2006, and the United States has since decreased its involvement in Bosnian political structures (Woehrel 2013, 9; ICG 2012, 2). Objections had been raised to the U.S. proposal by Serbs and Croats, both of which saw the collective presidency as preserving their interests. Serbs prefer to remain as independent from central institutions as possible. For Croats, the guarantees of voice that are part of the consociational structures are important because they are the smallest demographic group (ICG 2012, 4, 7, 13).
An in-depth analysis reveals that changing the presidency is a highly complex matter. Not only do groups have their own reasons for objecting, but the collective presidency also remains one of the most successful political institutions of the Dayton structure. As outlined by the International Crisis Group, “It is the only directly elected, high-level institution that Bosnian voters can identify with. It is not a broken institution and has not been the cause of any of Bosnia’s recent crises” (ICG 2012, 4). It is important to note that Bosnia’s inability to form a governing coalition for over a year after the 2010 elections was not a feature of the collective presidency, but rather was due to the dynamics of its central parliament (ICG 2012, 10). In addition, a restructure of the presidency would require a renovation of all of the Dayton institutions (ICG 2012, 1), due to the fact that it would be an attack on the consociational premises that are the agreement’s bedrock.
The stakes were raised in 2009, when the European Court of Human Rights issued a ruling that the structure of the presidency discriminates against ethnic others, or those who are not Bosniaks, Serbs, or Croats (Remikovic 2012; ICG 2012, 6). The ruling was based on a case brought by Sejdic, of Roma ethnicity, and Finci, of Jewish ethnicity. Several attempts to resolve this issue have failed.
The EU has tried to raise the stakes even further by making the resolution of this issue a condition for Bosnia’s EU membership, stating that a “credible effort” must be made before a “credible application” is made (ICG 2012, 8). The EU has made attempts to set deadlines for a resolution, but these have been missed repeatedly (Remikovic 2012; Alic 2012). The EU continues to express disappointment over the lack of progress in regard to the Sejdic-Finci ruling (European Commission 2012, 8–9, 16, 19). Some analysts have proposed that the EU should abandon these criteria, given the serious difficulties of making adjustments to the Dayton institutions. A new challenge lies on the horizon with a case being brought to the European Court by Pilav, a Bosniak in the Republika Srpska, who was banned from running for the presidency (ICG 2012, 9, 15).
Even without these human rights obstacles, Bosnia would face a difficult road to EU accession due to its internal political struggles. One of the most visible sources of difficulty is Milorad Dodik, president of the Republika Srpska. Since Kosovo’s declaration of independence in 2008, Dodik has been making the case that the RS should also be considered for independence from Bosnian structures. Since then, he has continued his threats to secede ( SETimes.com 2008; Radio Free Europe/Radio Liberty 2011; Gallup Balkan Monitor 2010, 4). 10 Moreover, government crises and stalemates over the past few years have produced a population that is frustrated with the ability of their representatives to conduct business. By June 2013, large protests took place against politicians’ inability to make decisions. A baby in need of stem cell treatment in Germany, who was denied a passport due to a stalemate on a law for issuing travel documents, was the spark that set off these protests. The baby subsequently died, and protestors mobilized to prevent representatives from leaving the parliament building until they had addressed the issue (Štiks 2013).
These problems for Bosnia do not appear to bode well for the future of the consociational model of governance. First, the intervention on the part of courts interested in human rights may signal a serious problem for the model’s requirement of the representation of constituent groups. Second, the stalemates inherent in the model can bring on public frustration with the government. Consideration of Northern Ireland as an example of consociationalism helps to explore how some of the problems in Bosnia might be mitigated.
Consociationalism in Northern Ireland
The Good Friday Agreement signed in Belfast between Protestant and Catholic leaders in 1998 established a solid constitutional framework that was intended to give each group a voice in politics. The brokerage of this agreement required not only the involvement of local actors, but also agreements between the British and Irish governments. In addition, the United States as an external actor played a role in bringing the parties to negotiations, with President Bill Clinton’s visit in 1995 perceived as a step toward dialogue between the parties. The discussions included some of the insights of consociational scholars, and, as in Bosnia, the representative aspects of consociational institutions provided a means to grant concessions to each side, fostering a move to agreement. Interestingly, while the mechanisms of the agreement in Northern Ireland strongly favored politics by group, it did not involve a strict formal requirement that a voting individual adhere to a group (Mac Ginty 2003, 7–8). The Northern Ireland agreement thus differs in this regard from the group-based requirements in Bosnia’s Dayton Agreement. But it remains the case that representatives in the Northern Irish Assembly must make a declaration regarding whether they are “nationalist, unionist, or other,” thus preserving consociational group-based decision-making within the parliament (Horowitz 2002, 194).
With regard to leadership by grand coalition, there is a dual premiership of the Deputy and First Minister, who must represent each group (Mac Ginty 2003, 8; Horowitz 2002, 194). There is also a rule that both groups can veto provisions they find threatening to their identity. However, the veto means that both groups must formally give their “parallel consent” to a set of institutions that guarantee the possibility of a veto for each side. According to these rules, at least 60 percent of all assembly members must participate in a vote and 40 percent of each of the identity-based group representatives must also vote—both nationalist and unionist (Horowitz 2002, 194). In a manner similar to the Bosnian institutions, “others” who do not identify with these opposing groups fall outside of these institutions. Because of the bifurcated nature of the group veto, it can be difficult for those who are part of the “other” category to gain leverage in the assembly (Horowitz 2002, 195).
In addition, there are frequent stalemates due to a structure that requires some consensus between groups. This problem is made less obvious by the fact that Northern Ireland is a unit within the United Kingdom’s structure, but it remains clear to locals. The Northern Irish parliament at Stormont, a unicameral body and one of the features provided by the 1998 agreement, has been “shut down” by the Westminster government for much of its existence, due to an inability for parliamentarians to work together (Taylor 2006). During such periods of shutdown, legislation on Northern Ireland comes out of the British government in Westminster.
There are also some inconsistencies in the Northern Ireland structure. The Good Friday Agreement provides that a referendum can be held in the future on the status of Northern Ireland. The results of this referendum will depend on group demographics. According to 2011 census figures, Northern Ireland is 48 percent Protestant and 45 percent Catholic. The figures for the 2001 census had registered 53 percent for Protestants, demonstrating a trend for Protestant decline over time (Devenport 2012). This shift is likely to mean that there could be a Catholic challenge to Northern Ireland’s presence within the United Kingdom once Catholics gain a majority. However, as noted by Roger Mac Ginty (2003), the provision of a referendum, a majoritarian device, is in tension with the otherwise strongly consociational and representative features of the agreement’s institutions. Even before the agreement, Protestants registered declining support for a potential agreement in relation to Catholics (Hughes and Donnelley 2003). The referendum provision is unlikely to improve Protestant views.
A problem that the Good Friday Agreement shares with the Dayton Agreement is that it prioritizes groups over individuals. While the focus on Bosnia has been greater than the focus on Northern Ireland in this regard, observers of Northern Ireland have commented that the consociational structures of government run counter to the premises of liberal democracy and individual rights (Taylor 2006, 220). One of the practical manifestations of this issue has been the matter of a Bill of Rights for Northern Ireland (Taylor 2006, 220–1; McCrudden 2001, 377–9). As in Bosnia, the matter of human rights for individuals appears to be in tension with the collective or group premises of the consociational structures. At the same time, the group-based provisions for Northern Ireland are less stringent than for Bosnia. In Northern Ireland it is only in the assembly that representatives must declare an alignment, and then may also make use of the “others” category. Voting can take place for ordinary citizens without such a declaration, unlike in Bosnia.
Conclusion: Some General Patterns
A comparison of the Bosnian and Northern Irish consociations illustrates a few common tensions between a consociational structure and liberal democracy. The group-based tenets of consociational institutions are in tension with liberal democracy’s endorsement of individual agency and the ability of individuals to maneuver in politics. At the same time, there are some specific institutional differences between Bosnia and Northern Ireland that appear to demonstrate a mitigation of these effects in Northern Ireland. One example of this difference is the fact that the grand coalition of executives elected on the basis of the separate geographic entities in Bosnia can hinder the participation of voters in this selection process, if they live as minorities in the “wrong” entity. Some attempts have been made to mitigate these effects, but it can remain difficult for minorities in the different entities to be fully engaged in this part of the political process. The Northern Irish executives are not selected on this type of identity-based geography.
Bosnia’s structures also contain more stringent provisions that prioritize the constituent peoples at the expense of ethnic others, which has resulted in the Sejdic-Finci ruling. A candidate for president must come from the constituent groups. Northern Ireland shares this provision to the degree that the Deputy and First Minister must come from the different groups. This provision has not been challenged on human rights grounds, and Northern Irish institutions maintain an uneasy relationship with a bill of rights. When human rights are defined as individual rights, they will be in tension with the group-based structures of consociational arrangements. Within the legislative bodies in both Bosnia and Northern Ireland, it has been the case that nonaligned individuals tend to be marginalized in politics. In Bosnia, this marginalization of others is a product of both the consociational structures and the proportional representation electoral system. In Northern Ireland, the preference ranking Single Transferrable Vote system, a proportional system in which voters rank their preferences for a list of candidates in a multimember district, can mitigate some of these effects. But within the parliament the structure of parallel consent has a polarizing effect, even though assembly members can declare themselves to be “other” to work informally with other “others,” as in the case of a self-identified women’s group.
Bosnia and Northern Ireland might also be at tension with liberalism due to the nature of their collectivities. Ethnicity and religion are identity categories that are difficult to alter, although in theory, one could change religion more plausibly than one could change ethnicity. What about language? In theory, an individual could switch languages more easily than ethnicity or perhaps religion, depending on the degree to which their personal identity is a strong component of their language of choice. These considerations make Belgium an interesting example to note. Belgium has consociational institutions that are language-based. It also has a highly decentralized federal structure that allows these linguistic communities ample control over matters relating to language. These communities are not based on territory—education in Brussels is legislated by the linguistic community linked to the language of the schools in question and not by a Brussels territorial authority (Jacobs and Swyngedouw 2003). Is it this nonterritorial aspect or the linguistic aspect that renders the Belgian consociational institutions relatively free from criticism? These types of questions can aid in a consideration of potential options for a redesign of the Bosnian state structures to resolve the dilemma raised by the ethnic “others” who fall outside of the Dayton institutions.
Some scholars and policy-makers have focused on nonterritorial structures as a way out of these territorial limitations (Coakley 1994; Nimni 2005; Bowring 2002; ECMI 2012). Provisions for Kosovo’s state structures incorporate some of these nonterritorial ideas, with provisions for a loose association of Serbian settlements and cross-border arrangements with Serbia for Kosovo’s Serbs (Stroschein 2008). Nonterritorial structures provide a means to preserve group-based premises for decision-making. However, they are de-linked from territory, which allows room for more individual choice regarding the group in which an individual prefers to take part. In this way, a French and a Flemish speaker might be neighbors in Brussels, but would each vote and participate in different language communities that make decisions regarding the schools that their children attend.
Some constraints remain, nevertheless; the French, Flemish, and German communities in Belgium limit language provisions to these groups and not to others. Group-based systems will always have the problem that individuals who do not identify with one of the available represented groups will find themselves outside of the decision-making structures of the state. Such is the problem with the Sejdic-Finci case in Bosnia. Faced with these dilemmas, it may be the case that consociationalism provides an initial set of arrangements for such societies, but that more long-term solutions lie in governance structures that mix group-based decision-making with mechanisms to provide for individual participation. One area of promise might lie in preferential voting systems, such as that used in Australia. Such systems allow for group-based parties, but also require that parties try to collect votes from individuals from other groups, fostering moderation. Indeed, Northern Ireland’s elections demonstrate some of the advantages of such a ranked system. A full consideration of how preferential voting might work in Bosnia is a fruitful area for future research and discussion (Reilly 2001).
In this contribution, I argue that Bosnia provides an example of how aid and intervention were crucial to setting the country on a firm foundation for governance. Despite much criticism of the consociational structures established by the 1995 Dayton Agreement, the state has not collapsed again into violence after nearly 20 years. In the short term, international intervention and aid have been crucial to this success. However, in the long term, some of the governance tensions inherent in the consociational structures of the Dayton Agreement have emerged, such as establishing democracy on the basis of particular groups. The fact that the Northern Ireland Good Friday Agreement has some similar (though not as many) tensions illustrates that the consociational contradictions are not simply Bosnian problems, but rather reflect some of the fundamental institutional tensions between group-based decision-making structures and individual human rights.
In the long term, successful revisions to the Dayton Agreement cannot be imposed by internationals, but instead must be negotiated on the ground among the groups. International involvement could be useful in playing a facilitator role in these discussions, but only a locally negotiated arrangement would be a lasting replacement to the complex Dayton institutions. Interestingly, one result of the February 2014 antigovernment economic protests has been the spontaneous emergence of popular assemblies or plenums that have begun to hold government officials to account, particularly in Tuzla and Sarajevo (Štiks 2014; Hemon and Mujanovic 2014). Such stirrings may be the beginning signs of how lasting and productive change might come to the Bosnian institutions.
Footnotes
Notes
Sherrill Stroschein is a senior lecturer (associate professor) in Politics at University College London, where she directs the master’s program in democracy and comparative politics. Her research examines the politics of ethnicity in democratic and democratizing settings, and she has published in a variety of journals including Perspectives on Politics, Party Politics, and Nations and Nationalism. She is author of Ethnic Struggle, Coexistence, and Democratization in Eastern Europe (Cambridge 2012), and editor of the collection Governance in Ethnically Mixed Cities (Routledge 2007). She also serves as associate editor of the journal Problems of Post-Communism.