Consociationalism After Civil War

IPSA 22nd World Congress of Political Science, 8-12 July 2012, Madrid

Consociationalism After Civil War: What is Different? Paper to be presented in the panel “Is Expansive Isolationism More Efficient than Multiculturalism? Justice of Individuals vs. Justice of Democracies”.

Matthijs Bogaards Professor of Political Science School of Humanities and Social Sciences Jacobs University Bremen Campus Ring 1, D-28759 Bremen, Germany Phone: 0049 (0)421 200 3462 Fax: 0049 (0)421 200 3303 E-mail: [email protected] Home page: https://www.jacobs-university.de/directory/mbogaards/

Work in progress. Please do not quote without prior permission of the author.

Abstract Whereas originally power sharing was thought of as a way to make democracy possible in a country divided along linguistic, religious, or ethnic lines, in the new millennium power sharing is increasingly seen as a means to bring peace to post-conflict societies. Almost all peace agreements by now contain some form of power sharing and all new consociational democracies were born out of civil war. So far, this trend has not received due attention. There are two reasons for this omission. First, most studies of power sharing in post-conflict societies are case studies. Second, the consociational literature has developed separately from peace and conflict studies. My paper brings together these two bodies of literature in a sketch of a comparative analysis of four contemporary cases of consociationalism after civil war: Bosnia-Herzegovina, Burundi, Lebanon, and Northern Ireland. The aim is to determine what is different about consociationalism after civil war, both to generate new insights on these cases, and to contribute to the further development of consociational theory.

2

Introduction Consociationalism was invented by the political scientist Arend Lijphart (1975) to describe a particular way of sharing power in an attempt to explain democracy and political stability in his native Holland, a country characterized for most of the early part of the 20th century by deep religious and ideological divisions. Almost immediately, the empirical scope was extended to Austria, Belgium, and Switzerland; all peaceful, prosperous, industrialized countries in Western Europe with consolidated democracies, Austria being an exception due to the experience with Austrofascism in the 1930s (Lijphart 1977). Although the boundaries of the consociational universe are not entirely clear, Lebanon (1989-), BosniaHerzegovina (1995-), Northern Ireland (1998-), and Burundi (2000-), have all been claimed and analyzed as contemporary consociational democracies.1 Lebanon returned to consociationalism after the Ta’if Accord ended a prolonged civil war; Bosnia-Herzegovina emerged as a new country out of the violent breakup of Yugoslavia, endowed with consociational institutions thanks to the Dayton Agreement; in Northern Ireland, the Good Friday Agreement ended “The Troubles” and restored regional autonomy under consociational rules; and the Arusha Peace and Reconciliation Agreement was an important step in the peace and democratization process in Burundi, paving the way for the 2005 Constitution, which stipulated power sharing. As can be gleaned from this quick overview, all countries that turned consociational after the end of the Cold War did so after going through a civil war. This raises the question, paraphrasing McGarry and O’Leary (2009), what is different about consociationalism “after the deaths of thousands”? Strangely, even though Taylor (2009b) for example notices a “major regeneration of the consociational research programme” (p.7) and a “new wave” of consociations (p.7), involving “conflicted societies moving from war to peace” (p.9), so far the special challenges of consociationalism after civil war have not received due attention in the rich consociational literature. This is problematic for at least two reasons. Empirically, we would want to know whether the institutions and behavioral patterns that characterize consociationalism work in the same way when consociationalism is established not to avoid conflict, but after violent ethnic conflict has broken out and cleavages have turned into battle lines. There are good reasons to think that civil war makes a difference. First, the starting position: maintaining peace and stabilizing democracy versus restoring peace and (re)democratization, often coupled with state-building. Second, the genesis of consociationalism: amicable agreement versus peace agreements. Third, the internal players: political parties versus armies and militias. Fourth, the role of 1

The list is probably longer, but a proper classification of contemporary consociational regimes should go beyond an “Advanced Google Scholar” search as provided by Taylor (2009b: 6).

3

external and international players: largely absent in classic consociational democracies, very prominent in all the new cases of consociationalism after civil war. While the notion of “complex consociationalism” (Wolff 2007) has helped to highlight the international dimension and the importance of territorial selfgovernance, it fails to capture the intricate relationship between consociationalism, state-building, peace-building, and (re)democratization that we observe in cases of consociationalism after civil war. We would be on a more solid footing in stating that “there is considerable support, analytically and empirically, for the proposition that consociational institutions offer a viable strategy to build peace, states and democracy” (Wolff 2011: 1796) if we had a better understanding of how consociationalism after civil war actually works and how it comes about. Instead, the literature on institutional choice in post-conflict societies has an unfortunate habit to focus on the post and less on the conflict phase and its ramifications, thereby obliterating much of the distinction between saving and (re)building democracy.2 This is all the more important as consociational theory “is not only an empirical but also a prescriptive theory” (Lijphart 2008c: 269). Or, in the words of Taylor (2009b), “a theory that matters”. As Wippman (1998: 216) already noted: “In promoting settlements for contemporary ethnic conflicts, the international community has strongly urged adoption of various consociational or quasiconsociational practices”. Dixon (2011: 313) even writes about “’academic entrepreneurs’ who are ‘selling consociational theory”. However, those who think consociationalism is the best choice when faced with ethnic diversity still have to make the argument why consociationalism is also the best choice in the context of violent ethnic conflict and we would want them to reflect on how consociationalism might have to be adapted when adopted after civil war. This paper only begins to fill that gap by sketching the contours of a research project on consociationalism after civil war. To clear the conceptual ground, the relationship between the concepts of power sharing and consociational democracy is examined. I argue that consociationalism is best viewed as a particular form of power sharing and that the two terms should not be used interchangeably. The paper then reviews the quantitative research on consociationalism, first, and power sharing, next, and their record in securing the objectives of democracy and peace. While these studies have yielded many important insights, they do not address directly the cases of consociationalism after civil war. That is the topic of the subsequent section. The conclusion formulates the next steps in my research project on consociationalism after civil war. An example would be Barnes (2001), whose analysis, despite the title, is much broader than postconflict societies and indeed has little specific to say about the unique challenges in the latter context. 2

4

Power sharing and consociational democracy What is power sharing and how is it related to consociationalism? Often, power sharing is used as an adjective. It is common to talk about power-sharing institutions, power-sharing democracy, and power-sharing regimes. Norris (2008) uses all three terms interchangeably, but there are important differences. In 2000, I criticized the equation of power sharing with consociational democracy as can be found in the later work of Lijphart (Bogaards 2000), citing three problems. First, to the extent that the concepts overlap, it is a waste of terms. Second, to the extent that they do not overlap, their interchangeable use invites confusion and imprecision. Third, power sharing is usually understood in a much broader sense, referring to any type of consensual or non-majoritarian practice.3 In line with the original broad sense of the term, I proposed to locate the concept of power sharing higher on the ladder of abstraction and to conceive of consensus and consociational democracy as specific forms of power sharing. In his response, Lijphart (2000) defends the equation of power sharing democracy and consociational democracy. O’Leary (2005: 37) sides with Bogaards, writing that ‘’’power sharing’ is not a synonym for consociation because there are other than consociational ways to share power....”. Unfortunately, confusion has only increased (See Lijphart 2002, 2008b,c). For example, for Schneckener (2002: 204-205) power sharing is synonymous with consociational democracy which in turn is presented as the particular form that consensus democracy assumes in multi-national polities. He then adds arbitration as a fifth feature to consociational democracy/power sharing.4 Narrowing the concept of power sharing to consociational democracy is one conceptual trap. Another pitfall is to loosen power sharing so much that it not only includes consociational democracy but also its opposite. This is what Sisk (2003) does when he treats integrative majoritarianism as a form of power sharing democracy. A third problem occurs when the relationship among types remains unspecified. This problem plagues McRae’s (1997) otherwise insightful account. McRae (1997: 285) criticizes Lijphart’s models of consociational and consensus democracy as “descriptive and static” and proceeds to identify fourteen different consensual elements in six selected democracies. While highlighting the variety of consensual devices, McRae does not offer a way to integrate them, beyond stating that they all present alternatives to the practices of majoritarian, 3

For example, Esman (2004) uses power sharing as an overarching concept, distinguishing between federalism and consociationalism as two patterns of power sharing. 4

To the extent that arbitration refers to informal meetings among group leaders, this is already part of what Lijphart (1975) called the “politics of accommodation” and which involved summit diplomacy. To the extent that it refers to formal arbitration forums including special courts, this is a new element.

5

Westminster democracy. As we will see below, the definition and operationalization of power sharing in the literature on peace agreements makes reference to Lijphart’s work, but does not restrict itself to either consensus or consociational democracy. Wolff ‘s (2011: 1791) phrase “consociational power sharing” is fortuitous in that it signals that consociationalism is a particular form of power sharing and builds an explicit bridge between the literatures on consociationalism and power sharing, broadly conceived, but is too cumbersome for daily use. Quantitative research on consociationalism Most of the consociational literature consists of case studies. Even small- N, qualitative comparisons are rare. Lijphart (1977) provided an overview and review of the entire consociational universe at the time, while Bogaards (2013) examines seven consociational parties, parties that allegedly practice(d) consociationalism internally, from across the world. Interestingly, there are by now a couple of quantitative studies that take consociationalism as the independent variable and democracy and/or peace as the dependent variable. Linder and Bächtiger (2005) operationalize power sharing through a “power sharing index (PSI)” that consists of two dimensions: vertical and horizontal power sharing. Vertical power sharing captures the territorial and functional delegation of authority and distinguishes between four levels. Horizontal power sharing looks at representation, rules of decision making, and the inclusiveness of the executive. In other words, while vertical power sharing aims to capture segmental autonomy, horizontal power sharing tries to get at the other three consociational principles of a grand coalition, proportionality (although only in formal representation) and the mutual veto. There are five degrees and constellations of horizontal power sharing. The power sharing index adds the scores for the two dimensions, even though they are only weakly related (Linder and Bächtiger 2005: 877). One of the main claims of Linder and Bächtiger (2005) is that power sharing has a positive effect on democratization, measured as the average level of democracy over a thirty year period in 62 African and Asian countries. This finding is quoted approvingly by Lijphart (2008c: 274). Linder and Bächtiger find that only horizontal power sharing is statistically significant. Surprisingly, cultural heterogeneity, which according to consociational theory constitutes the very reason for adopting power sharing in the first place, seems to have no impact (p.873). From this, Linder and Bächtiger (ibid.) conclude that it is power sharing arrangements as such, irrespective of social conditions, that promote democratization. Their study does not single out post-conflict societies and, to the best of my knowledge, their Power Sharing Index has not been used in other work.

6

The leading question for Norris (2008: 4) is: “Do power-sharing regimes generally serve to dampen armed conflicts in deeply divided multiethnic societies and thereby produce a durable peace settlement, political stability, and the conditions under which sustainable democracy can flourish?”. Unfortunately, her book does not provide an answer. First, it is not clear what the dependent variable is. Sometimes, the book is about democratization and democratic transitions. At other times, the focus is on democratic consolidation and peace. For practical purposes, the dependent variable throughout the book is the level of democracy as measured through three quantitative measures of democracy (Freedom House, Polity IV, and Vanhanen).5 In the chapter on type of government, an indicator of political crisis is used in addition. Although Norris (2008: 214) makes claims about the relationship between power-sharing arrangements and communal violence, her book does not look systematically at post-conflict societies. Moreover, different from Linder and Bächtiger, Norris examines consociational institutions separately, not in combination. Thus, while Linder and Bächtiger and Norris are to be applauded for initiating quantitative research on consociationalism, this line of research is still in its infancy, hampered by problems of operationalization and so far failing to show light on the performance of consociationalism where it is supposed to matter most: in post-conflict societies. Quantitative research on power sharing, peace, and democracy Studies that examine the comparative performance of different institutional choices in post-conflict societies do exist, of course, but these are not grounded in the consociational literature, although they frequently make reference to it. The literature on power sharing after civil war is at least a decade old by now. Walter (2002) pioneered research on power sharing as a means to promote peace after civil war. She focuses on three types of power sharing: political, territorial, and military. A political part is said to exist when “a settlement offered the combatants guaranteed positions in the new government at the level of cabinet or above, or a specific quota of power in at least one of the main branches of government” (Walter 2002: 62-63). Hartzell and Hoddie (2007), in their search for institutional determinants of peace, focus on the content of negotiated settlements and truces. In particular, they are interested in the way in which state power is reconfigured after civil war through four types of “power-sharing and power-dividing institutions” In their first publications, they still write about “power-sharing institutions”, defined as “those rules that, in addition to defining how decisions will be made by groups within the polity, allocate decision-making rights, including access to state resources, among collectivities competing for power” (Hartzell and Hoddie 2003: 320). In 5

For a discussion of measures of democracy, see Bogaards (2007a/b).

7

their 2007 book, they use the more cumbersome phrase “power-sharing and power-dividing institutions”, without making clear what precisely the distinction is and which is which.6 Political power sharing is operationalized as any of the following provisions: a change to electoral proportional representation; administrative proportional representation; executive proportional representation. Territorial power sharing is operationalized as either federalism/confederalism or regional autonomy. Military power sharing is operationalized as integration of armed forces on the basis of size or parity; the appointment of members of weaker factions to key leadership positions; permission for antagonists to remain armed; or permission for antagonists to retain their own armed forces. Finally, economic power sharing is operationalized as either the distribution of resources to disadvantaged groups specified as a percentage of resources or as a precise amount or policies to direct economic assets towards groups based on group membership or geographic location.7 They hypothesize that “the greater the overall number of power-sharing dimensions specified, the more likely that peaceful relations among collectivities will endure” (Hartzell and Hoddie 2003: 321).8 Power-sharing institutions are expected to have a cumulative, mutually reinforcing effect while the inclusion of multiple dimensions of power sharing protects against the failure to implement any single provision of the agreement. In light of this logic of the more powersharing the better, one would have expected the building of indexes also for the individual power-sharing institutions. This is especially true for political and military power sharing which consists of three and five items respectively. As it stands, a positive score on any of these by itself is enough to be coded positively for power sharing in the respective dimension. In a series of publications (Hartzell and Hoddie 2003, 2007; Hoddie and Hartzell 2005) find corroborating evidence for their proposition. However, instead of a steadily increasing positive effect of power-sharing institutions, it turns out that adopting only one or even two power-sharing institutions actually decreases the duration of peace after a negotiated settlement or truce. Only when at least three out of the four power-sharing institutions are adopted is there a positive impact. Compared to the strongly negative impact of adopting less than three powersharing institutions, the positive impact of adopting the (almost) complete package is rather low. Hartzell and Hoddie (2007: 82) acknowledge the existence 6

For the importance of this distinction, see Roeder (2005).

7

The operationalization of power sharing is described in Hartzell and Hoddie (2007: 159160). 8

As this quote already indicates, Hartzell and Hoddie (2003, 2007) appear to conceive of civil war as violent conflict between “collectivities”, “communities” and “groups” (Hartzell and Hoddie 2007: 23). This understanding of the nature of civil war has implications for the analysis of the resolution of war that should be made more explicit.

8

of a “breaking point” but do not explain why the adoption of a limited number of power-sharing institutions should actually harm peace.9 No attempt is made to look more closely whether type of power-sharing institution matters and whether there are perhaps particular combinations that work better than others. Hartzell and Hoddie (2007) are also interested in the favorable factors for the adoption of power-sharing institutions. They find that the duration (positive) and the intensity of the conflict (negative) are predictors of the extent of power sharing, as is the presence of a peacekeeping operation (p.59). Previous level of democracy and level of development, on the other hand, are not statistically significant. One notices how the favorable factors for adopting power sharing after civil war are different from the favorable factors that have been linked to the adoption of consociational democracy (See Bogaards 1998). Different from Hartzell and Hoddie (2003), Mukherjee (2006) examines powersharing agreements after military victories as well. In 111 civil wars fought between 1944 and 1999, 61 power-sharing agreements were offered. Of these, 34 resulted in peace, meaning that the civil war did not resume within 24 months (Mukherjee 2006: 480). A power-sharing agreement should consist of at least one of four concessions: 1) appointment of members/and or leaders of insurgent groups to ministerial, sub-ministerial, and cabinet positions; 2) appointment of members and/or leaders of insurgent groups to the central bureaucracy, civil service, courts, foreign service, and commissions; 3) use of proportional representation electoral system; 4) allowance of members of insurgent groups to organize and form parties to participate in the political process (Mukherjee 2006: 494). If any of these four concessions is made, a power-sharing agreement is said to have been offered. As with Hartzell and Hoddie (2003, 2007), it is not clear why instead of a dummy variable, no index was formed for political power sharing, especially as there is a big difference between merely allowing former rebels to turn into politicians and guaranteeing them a place in the cabinet.10 Likewise, the adoption of PR as electoral system by itself in most accounts does not constitute power sharing but rather provides the basis for it. Mukherjee (2006) argues that power-sharing agreements are likely to result in peace only when offered by the military winner of the civil war, be it the government or the insurgents.11 In case a power-sharing agreement is offered after a military stale mate, the durability of peace suffers. 9

One explanation is that almost all negotiated settlements and truces, their unit of analysis, have some form of power sharing. This guarantees that any variation in outcome will be attributable to the extent of power sharing. 10

An index of political power sharing would also have allowed Mukherjee (2006) to better test her theoretical argument, which makes a distinction between maximal and minimal concessions. 11

The result for power-sharing agreements after a victory by insurgents, however, is not statistically significant at conventional levels (Mukherjee 2006: 498).

9

Jarstad’s research (Jarstad 2008, 2009; Jarstad and Sundberg 2007; Jarstad and Nilsson 2008) looks at the implementation of peace agreements, which she argues, is the “missing link” between pacts and peace (Jarstad and Nilsson 2008: 209). The expectation is that the implementation costs of power sharing agreements differ by the type of power sharing and that the higher the implementation costs, the more likely that implementation will contribute to peace. Political power sharing, defined as guaranteed positions in the government or a specific quota in at least one of the main branches of government, is said to be relatively cheap and easy. Parties can accept positions and decline or withdraw at any time. Military power sharing, defined as a provision that offers rebels integration into the national forces and/or command structures or a provision that allows rebels to main their own separate armed forces, is much more costly, as it involves issues of security and vulnerability. Territorial power sharing, defined as a provision for some form of regional autonomy or a provision that allows the rebels to continue to administer an area under their de facto control is argued to be time-consuming and complicated.12 Jarstad focuses her analysis on the post-cold war period, counting 83 peace agreements in internal armed conflicts between 1989 and 2004 (Jarstad and Sundberg 2007: 78-79). When peace duration is taken as the dependent variable, a Cox proportional hazards model shows that the adoption of power sharing pacts themselves has no effect on peace, with the exception of territorial pacts. What matters is that the pacts are implemented and even more so, what type of pact is implemented. The rate of implementation differs between the various type of facts, with political power-sharing arrangements (27 out of 36 fully implemented) being much more likely to be implemented than territorial pacts (18 of out 33 fully implemented) and especially military power-sharing agreements (10 out of 29 fully implemented) (Jarstad and Nilsson 2008: 215). The implementation of political power sharing has no significant effect on peace, whereas the implementation of military and territorial power sharing does (Jarstad and Nilsson 2008: 216). Still, it goes too far to conclude that “the implementation of a political pact in general does not serve to mitigate or end violent conflict” (Jarstad and Sundberg 2007: 88). First, more than half the cases in which a political power-sharing agreement was implemented saw peace endure for at least five years. Second, this success rate is higher than for those cases of non or partial implementation (ibid.).13 Finally, a more detailed analysis

12

The definitions of the three forms of power sharing are taken from Jarstad and Sundberg (2007: 79-80). 13

In a new line of research, Jarstad (2009) looks at post-accord elections and armed conflict. She finds that 17 of the 38 legislative elections stipulated by the peace agreement, were followed by armed conflict. She therefore concludes that “the holding of elections have been a poor tool for promoting peace” (p.161). However, Jarstad also finds that in 9 of these cases, the violence came from parties that were not a signatory to

10

should be done of the particular form(s) of political power sharing. Jarstad bases her research on data collected by the Uppsala Conflict Data Program (UCDP). These data distinguish between six political provisions: election, interim government, integration into government/civil service, the right to become a political party, and “power sharing” described in the codebook as ‘provisions for extensive power-sharing in the new government’.14 Seventy of the 101 peace agreements counted by the UCDP between 1989 and 2005 have one or more political provisions, but only 15 feature power sharing (Harbom, Högbladh and Wallersteen 2006: 624). One lesson from the studies summarized above seems to be that political power sharing by itself does little to promote peace and democracy in post-conflict societies. Other types of power sharing - military, territorial, economic - appear to matter much more. At first blush, this finding goes against much of the consociational literature. That this discrepancy has not been noticed so far is most likely because the quantitative literature on power sharing, though paying lip service to the older scholarship on consociational democracy - which has been relabelled as power-sharing democracy by Lijphart – has actually very little in common with the latter. There are two reasons for this disjuncture. First, the independent variable in the analysis differs. The literature on power sharing in peace agreements goes much beyond the narrow concept of consociational democracy and looks at a wide variety of power-sharing devices, including military and economic power sharing. Second, the dependent variable is different. Whereas consociational theory is concerned with the prospects of democracy in divided societies, the literature on power sharing in peace agreements wants to know whether power sharing contributes to peace. The problems with the operationalization of political power sharing noted above would caution against an overinpretation of the results of the quantitative literature on power sharing after civil war, which in any case does not provide a test of consociationalism. Again, what is needed is a comparative analysis of cases of consociationalism after civil war. Consociationalism after civil war In consociational theory, the adoption of the four principles of a grand coalition, proportionality, mutual veto, and segmental autonomy is the outcome of a deliberate decision by elites to change their style of behaviour from competition to collusion. In a self-negating prediction, “elites cooperate in spite of the segmental differences dividing them because to do otherwise would mean to call forth the prophesied consequences of the plural character of the society” (Lijphart the original agreement and that in the few cases were no elections were organized, violence always broke out (p.158). This suggests that a more detailed analysis is in order and sweeping conclusions about the role of elections are best avoided. 14

UCDP Peace Agreement Dataset Codebook Version 1.0, October 2006.

11

1977: 100). Lijphart and others after him identified a range of favorable factors that are thought to help create and maintain a consociational democracy (See Bogaards 1998), but in the end consociational theory is voluntary. What is different in cases of consociationalism after civil war? Looking first at process, Horowitz (2002) and Shapiro (2003) already have focused attention on the “how” question. Horowitz (2002: 20) argues that the “consociational approach is motivationally inadequate”, relying on statesmanship. In his own work, Horowitz examines more carefully the incentive structure of ethnic elites and how this can be influenced through political engineering. In turn, Horowitz can be criticized for failing to pay due attention to the special nature of post-conflict societies. Some of the favorable factors from consociational theory seem relevant to post-conflict politics, for example the demographic balance between the various groups, but there is a need to go beyond these background variables and to include the variables from research on conflict onset, duration, and termination. After all, reforming an existing democracy in the direction of consociationalism to better accommodate the challenge of ethnic diversity (e.g. Belgium) is different from adopting consociationalism as part of a peace agreement after civil war (e.g. Burundi). Different from the classic cases of consociationalism, consociationalism after civil war is not left to politicians, nor are domestic elites necessarily in the driving seat. Indeed, one encounters frequent criticism of decisions being made elsewhere: the role of the USA and the EU in Bosnia, Syria’s intervention in Lebanon, South Africa leading the peace process in Burundi, the crucial role of Dublin and London in the Belfast Agreement. Ironically, where the local population was asked to endorse the resulting agreements, as happened in Northern Ireland, caution has been expressed about this “majoritarian device at the heart of a largely consociational agreement” (Mac Ginty 2003: 17; see also Mac Ginty et al. 2001). There is some evidence that consociational democracies learn from each other (see Guelke 1997). Party as a consequence, many contemporary cases of consociationalism after civil war exhibit what might be called compromised sovereignty. Bieber (2002: 330) observes how “Bosnia has gained some attributes of a protectorate since 1997” and Hayden (2005: 242) even talks about a “condominium with absentee landlord”. O’Leary (2001: 53) notes how the British-Irish Agreement of 1998 created a consociational settlement in Northern Ireland while “externally it established confederal relationships”. The role of Syria, both formal and informal, has been noted by almost all scholars working on Lebanon (See, f.e., Zahar 2002). The notion of “complex power sharing”, described by Wolff (2007: 389) as “a practice of conflict settlement that has a form of self-governance regime at its heart, but whose overall institutional design includes a range of further mechanisms for the accommodation of ethnic diversity in divided societies, including those recommended by advocates of consociationalism....”, partially captures these phenomena.

12

Moving from process to practice, is there anything particular about consociational institutions adopted after civil war? Different from the often informal “politics of accommodation” one finds or found in the Netherlands and Switzerland, consociational features in the contemporary cases tend to be written down in peace agreements and subsequently codified and constitutionalized. Temporally, the emergence of consociationalism after civil war coincides with the introduction of a distinction between “liberal” and “corporate” consociations in the literature. According to McGarry and O’Leary (2007: 675) a corporate consociation relies on “predetermination” and “accommodates groups according to ascriptive criteria, such as ethnicity or religion, on the assumption that groups identities are fixed and that groups are both internally homogeneous and externally bounded”. In contrast, a liberal or “self-determined” consociation rewards whatever salient political identities emerge in democratic elections, whether these are based on ethnic or religious groups, or on subgroup or transgroup identities. The new constitution of Iraq and the new dispensation in Northern Ireland conform to McGarry and O’Leary’s favoured liberal consociation whereas Lebanon approximates more closely the corporate type of consociation (See also McGarry and O’Leary 2004).15 However useful, the juxtaposition of liberal and corporate consociationalism has obscured the fact that three of the four contemporary cases of consociationalism after civil war have adopted electoral systems that encourage or even mandate cross-ethnic voting. Burundi has ethnic quota in parliament while at the same time insisting that all parties are internally multi-ethnic, thereby killing two birds with one stone (Sullivan 2004; Reyntjes 2005; Lemarchand 2007; Vandeginste 2009, 2011). Lebanon has a complicated electoral system that has undergone many changes in the last two decades, but it has served the purpose of ethnic vote-pooling (Hudson 1997; Jabbra and Jabbra 2001; Horowitz 2002; Salamey and Payne 2008; Salamey 2009; Haddad 2010). Northern Ireland has adopted the Single Transferable Vote, although opinions diverge as to how much this electoral system has facilitated, as it is expected to do theoretically, voting across ethnic lines. Without a systematic, comparative analysis of consociationalism after civil war nothing more can be said at this point. This analysis should center on BosniaHerzegovina, Burundi, Lebanon, and Northern Ireland. Other cases of consociationalism after civil war may include Kosovo and Macedonia (Bieber 2005).16 If the divided island of Cyprus, a failed consociation (Lijphart 1977; Kliot and Mansfield 1997), is to have a common future, this will most likely entail consociational features, as with the Annan Plan, which was rejected in a

15

For a critique of liberal consociationalism in Northern Ireland, see Horowitz (2001).

16

Ivanisevic (1998) already in 1998 proposed consociationalism for Kosovo.

13

referendum in 2004 (Lijphart 2002; Vural and Peristianis 2008).17 The analysis should also look critically at failed consociations, most prominently Lebanon (Lijphart 1977; Hudson 1976; Seaver 2000; Dekmejian 1978) and Cyprus. Finally, a comparative study of consociationalism after civil war should go beyond the here and now and revisit historical experiences of consociationalism after civil war. Two (sets) of cases come to mind. First, Colombia in the 1950s and 1960s, when the elites of the two main political parties agreed on sharing power after their members fought a very bloody civil war. Even though Lijphart (1977) described Colombia as a “semi-consociational democracy”, the country never attracted much attention from consociational scholars, probably because the large scale violence and the political rather than societal origins of the conflict were at odds with the logic and experience of consociationalism in Western Europe. These very same reasons make the case of Colombia so interesting for a comparative study of consociationalism after civil war. Then, although not many people realize this, the classic cases of consociationalism in Western Europe were not entirely devoid of political violence. Austria experienced a civil war in the 1930s, the Swiss (con)federation was established after a religiously motivated civil war and in the Netherlands, some claim – although few believe – the “Pacification” of 1917 helped prevent a civil war. The number of casualties of the civil wars in Austria and Switzerland is counted in the hundreds, not the thousands or hundreds of thousands, but a study of consociatinalism after civil war would not be complete without an analysis of the impact of these civil wars on the emergence of consociationalism in these West European countries. Conclusion Whereas originally consociationalism was conceived of as way to make democracy possible in a divided society, in the new millennium consociationalism and power sharing more broadly are increasingly regarded as means to bring peace to post-conflict societies. Almost all peace agreements by now contain some form of power sharing and all new consociational (semi-)democracies were born out of civil war. Scholarly interest in power sharing has broadened from the consociational literature in comparative politics to peace and conflict studies. Unfortunately, these two bodies of work rarely speak to each other. The result has been conceptual confusion and a fragmentation of empirical knowledge. This paper has sought to bring them together, preparing the ground for a full-fledged study of consociationalism after civil. The aim is to answer the leading question whether consociationalism in post-conflict societies is a recipe for democracy and peace, to identify the factors that contribute to success or failure, to evaluate possible trade-offs, and to draw lessons for other cases. 17

Consociationalism has also been proposed for post-genocide Rwanda (Njoku 2005; Vandeginste and Huyse 2005).

14

Although Paris (2004: 193), based on the Bosnian case, suggests that “consociationalism is ill suited to the domestic conditions of war-shattered states”, the fact is that the “wave of power-sharing democracy” (Lijphart 2002) has swept ashore mostly countries and regions coming out of a civil war. Even critiques of consociationalism have acknowledged this, normally adding that consociationalism at best is a temporary solution to end the violence, not a setup that is viable or desireable in the long run (see, f.e., Roeder 2005). To better understand the relationship between consociationalism and civil war, the next steps in my project will be: •

• •

•

A comparative analysis of all contemporary cases of consociationalism after civil war (most prominently: Bosnia-Herzegovina, Burundi, Lebanon, and Northern Ireland); A re-examination of Colombia, a semi-consociational democracy established after a civil war between two political parties and their militia; A search for the “bloody origins of classic consociations”, concretely the civil wars in Austria and Switzerland that preceded consociationalism there; An analysis of cases of “consociationalism before civil war”, i.e. those consociations that broke down and ended in civil war (most prominently: Lebanon and Cyprus).

Empirically and theoretically, this should result in a better understanding of the character of consociationalism after civil war, the process leading up to it, its functioning, and outcomes. For the purposes of prediction and prescription, a study of the relationship between consociationalism and civil war should help us in two ways. First, in making more precise recommendations about the process of building consociations. At the moment, advocates of consociationalism have little more to offer than the advice to elites in war-ravaged countries to adopt consociational arrangements, without telling those coming out of a civil war how this outcome can be reached. There has always been something mystical about the ways in which consociations come about, something about wise segmental leaders who stepped over their own shadow to save their country. That is a lot to ask of warlords, rebels, and oppressive governments. Second, at the moment, recommendations for consociationalism lack specificity, very much in line with Lijphart’s idea about consociational principles whose precise embodiment depends on the local circumstances, without however specifying under which circumstances which institutions or arrangements work best. The distinction between liberal and corporate consociations and the notion of complex consociation are important innovations, but they still fail to do justice to the empirical variation in consociational institutions one finds around the world. There is much to discover and learn with respect to consociationalism after civil war.

15

References Barnes, Samuel (2001) The Contribution of Democracy to Rebuilding Postconflict Societies, American Journal of International Law 95(1), 86-101. Bieber, Florian (2002) Governing Post-War Bosnia-Herzegovina, in Gál (2002), pp.321-337. Bieber, Florian (2005) Partial Implementation, Partial Success: The Case of Macedonia, in O’Flynn and Russell (2005), 107-122. Bogaards, Matthijs (1998) The Favorable Factors for Consociational Democracy: A Review, European Journal of Political Research 33(4): 475-496. Bogaards, Matthijs (2000) The Uneasy Relationship between Empirical and Normative Types in Consociational Theory, Journal of Theoretical Politics 12(4): 395-424. Bogaards, Matthijs (2007a) Elections, Election Outcomes and Democracy in Southern Africa, Democratization 14(1): 73-91. Bogaards, Matthijs (2007b) Measuring Democracy through Election Outcomes: A Critique with African Data, Comparative Political Studies 40(10): 1211-1237. Bogaards, Matthijs (2013) Consociational Parties: Power Sharing in Divided Societies. Basingstoke: Palgrave Macmillan. Dekmejian (1978) Consociational Democracy in Crisis: The Case of Lebanon, Comparative Politics 10(2), 251-265. Dixon, Paul (2011) Is Consociational Theory the Answer to Global Conflict? From the Netherlands to Northern Ireland and Iraq, Political Studies Review 9(3), 309322. Esman, Milton (2004) An Introduction to Ethnic Conflict. Cambridge: Polity. Gál, Kinga (2002) (ed.) Minority Governance in Europe. Budapest: Open Society Institute. Guelke, Adrian (1997) Comparative Peaceful: The Role of Analogy in Northern Ireland’s Peace Process, Cambridge Review of International Affairs, 11(1), 2845.

16

Haddad, Simon (2010) The Political Consequences of Electoral Laws in Fragmented Societies: Lebanon’s 2009 Elections, The Journal of Social, Political and Economic Studies 35(1), 45-78. Harbom, Lotta, Stina Högbladh and Peter Wallensteen (2006) Armed Conflict and Peace Agreements, Journal of Peace Research 43(5), 617-631. Hartzell, Caroline and Matthew Hoddie (2003) Institutionalizing Peace: Power Sharing and Post-Civil War Conflict Management, American Journal of Political Science 47(2): 318-332. Hartzell, Caroline and Matthew Hoddie (2007) Crafting Peace: Power-Sharing Institutions and the Negotiated Settlements of Civil Wars. University Park, PN: The Pennsylvania State University Press. Hoddie, Matthew and Caroline Hartzell (2005) Power Sharing in Peace Settlements: Initiating the Transition from Civil War, in Roeder and Rothchild (2005), pp.83-106. Hayden, Robert (2005) “Democracy” Without a Demos? The Bosnian Constitutional Experiment and the Intentional Construction of Nonfunctioning States, East European Politics and Societies 19(2), 226-259. Horowitz, Donald (2001) The Northern Ireland Agreement: Clear, Consociational, and Risky, in McGarry (2001), pp.89-108. Horowitz, Donald (2002) Constitutional Design: Proposals Versus Processes, in Reynolds (2002), pp.15-36. Hudson, Michael (1976) The Lebanese Crisis: The Limits of Consociational Democracy, Journal of Palestine Studies 5(3-4), 109-122. Hudson, Michael (1997) Trying Again: Power-Sharing in Post-Civil War Lebanon, International Negotiation 2(1), 103-122. Ivanisevic, Bogdan (1998) Power-Sharing Proposal for Kosovo, Eurobalkans 32(autumn), 34-39. Jabbra, Joseph and Nancy Jabbra (2001) Consociational Democracy in Lebanon: A Flawed System of Governance, Perspectives on Global Development and Technology 17(2), 71-89. Jarstad, Anna (2008) Power Sharing: Former Enemies in Joint Government, in Anna Jarstad and Timothy Sisk (2008) From War to Democracy: Dilemmas of Peacebuilding. Cambridge: Cambridge University Press, pp.105-133.

17

Jarstad, Anna (2009) Post-Accord Elections and Armed Conflict, in Ashok Swain, Ramses Amer, and Joakim Öjendal (eds.) The Democratization Project: Opportunities and Challenges. London: Anthem Press, pp.153-163. Jarstad, Anna (2010) The Prevalence of Power-Sharing: Exploring the Patterns of Post-Election Peace, Africa Spectrum 3, 41-62. Jarstad, Anna and Desirée Nilsson (2008) From Words to Deeds: The Implementation of Power-Sharing Pacts in Peace Accords, Conflict Management and Peace Science 25, 206-223. Jarstad, Anna and Ralph Sundberg (2007) Peace by Pact: Data on the Implementation of Peace Agreements, in Ashok Swain, Ramses Amer, and Joakim Öjendal (eds.) Globalization and Challenges to Building Peace. London: Anthem Press, pp.72-89. Kliot, N. and Mansfield, Y. (1997) The Political Landscape of Partition: The Case of Cyprus, Political Geography 16(6), 495-521. Lemarchand, René (2007) Consociationalism and Power Sharing in Africa: Rwanda, Burundi, and the Democratic Republic of Congo, African Affairs 106(422), 1-20. Lijphart, Arend (1975) The Politics of Accommodation: Pluralism and Democracy in the Netherlands, Second Revised Edition. Berkeley: University of California Press. Lijphart (1977) Democracy in Plural Societies: A Comparative Exploration. New Haven, CN: Yale University Press. Lijphart, Arend (2000) Definitions, Evidence and Policy: A Response to Matthijs Bogaards’ Critique, Journal of Theoretical Politics 12(4): 425-432. Lijphart, Arend (2002) The Wave of Power-Sharing Democracy, in Reynolds (2002), pp. 37-54. Lijphart, Arend (2008a) Thinking About Democracy: Power Sharing and Majority Rule in Theory and Practice. London: Routledge. Lijphart, Arend (2008b) Introduction: Developments in Power Sharing Theory, in Lijphart (2008a), pp.3-22. Lijphart, Arend (2008c) Conclusion: Power Sharing, Evidence, and Logic, in Lijphart (2008a), pp.269-281.

18

Linder, Wolf and André Bächtiger (2005) What Drives Democratization in Asia and Africa, European Journal of Political Research 44(6), 861-880. Mac Ginty, Roger, Rick Wilford, Lizanne Dowds, and Gillian Robinson (2001) Government and Opposition 36(4), 472-492. Mac Ginty, Roger (2003) Constitutional Referendums and Ethnonational Conflict: The Case of Northern Ireland, Nationalism and Ethnic Politics 9(2), 1-22. McGarry, John (2001) (ed.) Northern Ireland and the Divided World: PostAgreement Northern Ireland in Comparative Perspective. Oxford: Oxford University Press. McGarry, John and Brendan O’Leary (2004) The Northern Ireland Conflict: Consociational Engagements. Oxford: Oxford University Press. McGarry, John and Brendan O’Leary (2007) Iraq’s Constitution of 2005: Liberal Consociation as Political Prescription, International Journal on Constitutional Law 5(4): 670-698. McRae, Kenneth (1997) Contrasting Styles of Democratic Decision-Making: Adversarial Versus Consensual Politics, International Political Science Review 18(3), 279-295. Mukherjee, Bumba (2006) Why Political Power-Sharing Agreements Lead to Enduring Peace Resolution of Some Civil Wars, But Not Others? International Studies Quarterly 50(2), 479-504. Norris, Pippa (2008) Driving Democracy: Do Power-Sharing Institutions Work? Cambridge: Cambridge University Press. McGarry, John and Brendan O’Leary (2009) Power Shared After the Deaths of Thousands, in Taylor (2009a), pp.15-84. Njoku, Raphael Chijioke (2005) Deadly Ethnic Conflict and the Imperative of Power Sharing: Could a Consociational Federalism Hold in Rwanda? Commonwealth & Comparative Politics 43(1), 82-101. O’Flynn, Ian and David Russell (2005) (eds.) Power Sharing: New Challenges for Divided Societies. London: Pluto Press. O’Leary, Brendan (2001) Comparative Political Science and the British-Irish Agreement, in McGarry (2001), pp.53-88. O’Leary, Brendan (2005) Debating Consociational Politics, in Sid Noel (ed.) From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies. Montreal:McGill-Queen’s University Press, pp.3-43.

19

Paris, Roland (2004) At War’s End: Building Peace after Civil Conflict. Cambridge: Cambridge University Press. Reynolds, Andrew (2002) (ed.) The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy. Oxford: Oxford University Press. Reyntjes, Filip (2005) Burundi: A Peaceful Transition After a Decade of War? African Affairs 105(418), 117-135. Roeder, Philip (2005) Power Dividing as an Alternative to Ethnic Power Sharing, in Roeder and Rothchild (2005), pp.51-82. Roeder, Philip and Donald Rothchild (2005) (eds.) Sustainable Peace: Power and Democracy after Civil Wars. Ithaca: Cornell University Press. Rothchild, Donald (2005) Reassuring Weaker Parties after Civil Wars: The Benefits and Costs of Executive Power-Sharing Systems in Africa, Ethnopolitics 4(3), 247-267. Salamey, Imad (2009) Failing Consociationalism in Lebanon and Integrative Options, International Journal of Peace Studies 14(2), 83-105. Salamey, Imad and Rhys Payne (2008) Parliamentary Consociationalism in Lebanon: Equal Citizenry vs. Quotated Confessionalism, The Journal of Legislative Studies 14(4), 451-473. Schneckener, Ulrich (2002) Making Power-Sharing Work: Lessons from Successes and Failures in Ethnic Conflict Regulation, Journal of Peace Research 39(2), 203-228. Seaver, Brenda (2000) The Regional Sources of Power-Sharing Failure: The Case of Lebanon, Political Science Quarterly 115(2), 247-271. Shapiro, Ian (2003) The State of Democratic Theory. Princeton, NJ: Princeton Univerisity Press. Sisk, Timothy (2003) Power-sharing after Civil Wars: Matching Problems to Solutions, in John Darby and Roger Mac Ginty (eds.) Contemporary Peacemaking: Conflict, Violence and Peace Process. Basingstoke: Palgrave Macmillan, pp.139-150. Sullivan, Daniel (2005) The Missing Pillars: A Look at the Failure of Peace in Burundi Through the Lens of Arend Lijphart’s Theory of Consociational Democracy, Journal of Modern African Studies 43(1), 75-95.

20

Taylor, Rupert (2009a) (ed.) Consociational Theory: McGarry and O’Leary and the Northern Ireland Conflict. London: Routledge. Taylor, Rupert (2009b) Introduction: The Promise of Consociational Theory, in Taylor (2009a), pp.1-11. Vandeginste, Stef (2009) Power-Sharing, Conflict and Transition in Burundi: Twenty Years of Trial and Error, Africa Spectrum 44(3), 63-86. Vandeginste, Stef (2011) Power-Sharing as a Fragile Safety Valve in Times of Electoral Turmoil: The Costs and Benefits of Burundi’s 2010 Elections, Journal of Modern African Studies 49(2), 315-335. Vandeginste, Stef and Luc Huyse (2005) Consociational Democracy for Rwanda?, in Stefaan Marysse and Filip Reyntjes (eds.) The Political Economy of the Great Lakes Region in Africa: The Pitfalls of Enforced Democracy and Globalization. Basingstoke: Palgrave Macmillan, pp.101-122. Vural, Yücel and Nicos Peristianis (2008) Beyond Ethno-Nationalism: Emerging Trends in Cypriot Politics after the Annan Plan, Nations and Nationalism 14(1), 39-60. Walter, Barbara (2002) Committing to Peace: The Successful Settlement of Civil Wars. Princeton: Princeton University Press. Wippman, David (1998) International Law and Ethnic Conflict. Ithaca, NY: Cornell University Press. Wolff , Stefan (2007) Conflict Resolution Between Power Sharing and Power Dividing, or Beyond, Political Studies Review 5(3): 377-393. Wolff, Stefan (2011) Post-Conflict State Building: The Debate on Institutional Choice, Third World Quarterly 32(10), 1777-1802. Zahar, Marie-Joëlle (2002) Peace by Unconventional Means: Lebanon’s Ta’if Agreement, in Stephen John Stedman, Daniel Rothchild, and Elizabeth Cousens (eds.) Ending Civil Wars: The Implementation of Peace Agreements. Boulder, CO: Lynne Rienner, pp.567-597.

21