Power-Sharing
The principal means of conflict termination since the end of the Cold War has been negotiated settlement rather than military victory. This is true not only globally (Hartzell and Hoddie 2007: 10), but particularly in Africa, where few wars over the past two decades have ended with an outright military victory. This is remarkable in itself and reflects a new trend: today only between one-fifth and one-third of all wars end in military victory. Yet not all negotiated settlements to violent conflicts are sustainable. Many stakeholders seek to avoid ‘losing the peace’ even at the cost of starting a new war. In response to this challenge, another trend has emerged: in negotiated conflict settlements, power-sharing devices have become more prominent (Mehler 2009).
Power-sharing may be defined as an elite pact between representatives of political or military parties on the division of responsibility in different fields of political and economic life. It has become an attractive solution to severe political crises because it aims to avoid future armed confrontations by minimizing the security concerns of rival groups. It is in this context that classic debates regarding the appropriate division of power – or more specifically the necessary curtailing of presidential power – have been revived. The spill-over of crisis-solving approaches into fundamental debates over the appropriate systems of government in post-conflict countries hints at the multi-dimensionality of power-sharing. Accordingly, one might distinguish between: short-term and medium- to long-term intentions and effects on one axis; and the spheres in which power is shared, most notably the different branches of government and the military, economy, and territory of a given polity.
Consociational Democracy and Post-Conflict Power-Sharing
The academic debate on power-sharing is strongly influenced by the problem of how to make democracy work in divided societies. Given that the main characteristic of democracy is majority rule, minority groups may find themselves constantly dominated by the majority and unable to defend their vital interests. They may become alienated from the general polity and seek violent solutions to accommodate their grievances. Classical power-sharing theory therefore proposes strategies to make democracy work in plural societies. Most prominent among this group of scholars is Arend Lijphart (1977), whose concept of ‘consociational democracy’ may be summarized as follows:
An influential school of thought represented by Donald Horowitz and others opposes Lijphart’s view with its own ‘integrative approach’ to power-sharing. Indeed Horowitz (1985: 566) has warned that Lijphart’s standard recipe could easily cement ethnic identities and consequently provide disincentives for elites to moderate their discourse and behaviour. While both consociationalists and integrationists address the same core problem, the remedy they identify is radically different: whereas integrationists advocate institutional mechanisms that provide incentives for cooperation across the main dividing lines of plural societies, consociationalists follow a strategy of minority empowerment. This debate has evolved and gained new prominence as a result of the aforementioned trends. In contrast to the classic debate between Lijphart and Horowitz, the contemporary post-conflict power-sharing literature focuses less on the quality of democracy than on the role played by power-sharing in the preservation of fragile peace.
Again, there are both staunch supporters and opponents to power-sharing as a key element in peacebuilding strategies. Supporters base their argument on the ‘security dilemma’ evident at the point of conflict termination. Disarmament of the various parties to a conflict is rarely the key to resolving a crisis of mutual trust, particularly when a government retains – or regains over time – control of the state’s security forces (Curtis, this volume). Power-sharing arrangements typically seek to resolve this dilemma in the immediate pre- and post-conflict termination periods by providing guarantees regarding the uses of state power, thus rendering commitments to peace more credible. The main arguments in favour of post-conflict power-sharing are therefore that they reduce uncertainty in the immediate post-conflict era, and accommodate entrenched grievances through the introduction of appropriate power-sharing institutions. The main arguments against post-conflict power-sharing resemble the integrationists’ critique of consociational democracy – namely, that post-conflict power-sharing offers rewards to rebels, thereby providing incentives to take up arms, and deals with identities as primordial and thus incentivises individuals to portray themselves as group members, in the process cementing societal cleavages.
A second discussion touches on the nexus between post-conflict power-sharing and democracy. Again, the views on the matter diverge substantially. Proponents of power-sharing expect a transitional power-sharing phase to be beneficial in facilitating the transition to democracy (Hartzell and Hoddie 2007; for a critical view see Roeder and Rothchild 2005). Former rivals learn to live and work with each other, develop mutual trust and esteem; a government of national unity may work as a ‘school of democracy’. Moreover, former adversaries take the ‘risk of democracy’, as they have guarantees of at least minimal representation in power-sharing institutions. The opposite view argues that even established civilian parties are sidelined in power-sharing deals, which are likely to impact negatively on democratic transitions. Power-sharing could also be argued to result in the effective abandonment of democracy, given that the rule of the majority has lost significance. Furthermore, critics suggest that once in government, rebel movements are likely to be less disposed to relinquish power at elections than civilian movements, given that they have paid a ‘higher price’, or to cooperate with other armed movements for the same reason (Spears 2000).
Important refinements to these arguments have been put forward over the last decade or so. Walter (2002) has proposed a so-called ‘credible commitment theory’ to explain the success or failure of peace agreements. One central element of this argument is that by making ‘costly signals’ that demonstrate their genuine commitment to the process, parties can enhance their credibility, which may result in the incumbent government being willing to cede important positions of power to other actors. This thinking suggests that we need to include further dimensions of power-sharing in our analysis. Whereas Lijphart focused on the political dimension only, Hartzell and Hoddie (2007; also Hoddie and Hartzell 2003) have proposed an influential distinction between political, territorial, military, and economic forms of power-sharing. A political dimension is nearly always part of a power-sharing deal; the other dimensions are only variably included, but can be greatly relevant in practice.
Alongside these debates, an expanding quantitative literature on post-conflict power-sharing has focused predominantly on the inclusion of power-sharing provisions in the text of a peace agreement (e.g. Pearson et al. 2006; Derouen et al. 2009), and has only rarely addressed their actual implementation (e.g. Jarstad and Nilsson 2008). The findings of this literature are often contradictory and inconclusive, perhaps unsurprisingly, given the many problems that one faces when coding complex treaties and their haphazard implementation. By contrast, qualitative studies often focus on single cases and lead to largely positive or largely negative results. A recent comprehensive effort by Sriram (2008), for instance, covers 25 cases in Europe, Latin America, Asia, and Africa and comes to a rather sceptical general conclusion with regard to the value of power-sharing. However, qualitative approaches always face the risk of selection bias, which may support the preconceived opinions of their authors.
Africa’s Experience with Consociational Democracy
African states have often employed some, but rarely all, elements of the consociational model. Governments of national unity, very familiar to Africa from independence onward, come immediately to mind when thinking about grand coalitions (Rothchild and Foley 1988). The African experience with this form of government is ambivalent; it may have avoided severe conflicts in some states but also precluded democracy in others. Oversized governments have been common but problematic, most notably the forced inclusion of smaller parties within unified parties in the 1960s. Significantly, coalition agreements were rarely enacted by party congresses, but were instead signed by party leaders with little attempt to include supporters. More recent attempts at governments of national unity have explicitly followed the power-sharing model, albeit in countries such as Kenya and Zimbabwe, which did not experience full-fledged civil war (see Cheeseman and Tendi 2010).
By contrast, the minority veto is rarely an established feature in African constitutions. Yet this does not necessarily dictate against its de facto employment. Lijphart (1977: 38) himself writes that the minority veto can be an informal and unwritten understanding. In the case of Burundi, the Tutsi minority had for decades enjoyed the ‘ultimate’ veto in the form of a monopoly on military power (Sullivan 2005: 88). As Sullivan notes, it was ‘the threat of losing that ultimate veto which caused it to be used, thus ending the attempt at peace’ (ibid.). This could be an argument for more explicit, formal veto rights. Transparency on the option and procedures of a veto can only be achieved by formally acknowledging those veto rights in peace agreements or (interim) constitutions.
Proportional representation has also rarely made its entry into an African constitution as the guiding principle in the distribution of senior positions within the government and bureaucracy. However, some ethnically polarized countries have recurrently experienced near-to-proportional representation in higher offices. One country in which the media constantly counts and compares the ethnic proportions of newly appointed governments, or a new higher military command, is Cameroon. However, this did not make the country any more of a consociational democracy (or any other form of democracy for that matter). The record is more balanced with regard to proportional representation at national elections. A good number of African countries have experienced proportional representation (PR) in legislative elections, such as South Africa and Namibia (pure PR systems), Burkina Faso (since 2002) and Mozambique (PR in medium and large multi-member constituencies), and Burkina Faso (1992–97) and Guinea-Bissau (PR in small constituencies).1 However, PR has not always prevented the concentration of power within one party as the comparative literature would suggest: even in democratic South Africa and Namibia proportional voting systems have proved compatible with one-party dominance rule.
Finally, the continent has not experienced many federal experiments and those that have been introduced have not met with great success (Suberu, this volume). Currently, Comoros (since 2002), Ethiopia (since 1994; earlier period 1952–62), Nigeria (since 1954/1960), South Africa (since 1996), Sudan (since 2005–11; earlier period 1972–83) and Tanzania (since 1964) exhibit federal systems. In Ethiopia (Ethiopia and Eritrea in the earlier period), Sudan (Northern and Southern part), and Tanzania (Tanganyika and Zanzibar), only two states constituted the federation, so these systems were far from ideal cases of how to accommodate group interests in a multi-ethnic setting. Nevertheless, in all cases it was hoped that the introduction of a federal system would mitigate conflict.
There are persuasive arguments for and against what Bunce and Watts (2005: 136) call ‘ethnofederalism’. On the one hand, it may counter two typical temptations in multi-ethnic settings: of minorities to defect and of majorities to dominate. It may also legitimize difference and empower minorities, thereby creating trust. Conversely, ethnofederalism may unnecessarily cement differences and identities, undermining commonality. Such a system may also undermine cooperation and give minorities the institutional prerequisites for later secession. The South African case is usually seen as a success (as are India and Canada), while all other African cases are much more contested. The Comoros case exemplifies the pattern of lengthy negotiations with many setbacks, facilitated by external help. In recent years there has been a strong tendency to establish a degree of decentralization, sometimes in the name of conflict prevention and power-sharing (Spears 2000: 115; Crawford and Hartmann 2008). Only a few of the many decentralization reforms that have been introduced across the continent were comprehensively conceived and implemented by ruling elites. Taken together, we can hardly say that the four main elements of consociational democracy have had uniformly positive effects on African countries. Indeed, most have a decidedly mixed track record.
Africa’s (Recent) Experience with Post-Conflict Power-Sharing
One of the best-known peace agreements in Africa is the Arusha agreement meant to end the civil war in Rwanda in 1993. It contained a good number of power-sharing mechanisms designed to include both the Hutu-dominated government and the Tutsi-dominated rebels. One of the problematic effects of the agreement was the refusal of regime hardliners to accept the outcome (i.e. the loss of substantial power); as a consequence of the peace deal and a number of other processes there followed a period of political radicalization among extreme Hutu factions which ultimately led to the genocide in which over 800,000 Tutsi lost their lives. After the spectacular failure of the Arusha agreement, more complex negotiations have taken place in Africa’s severest crises. If one concentrates on the time span between 1999 and 2010, 18 of 19 peace agreements concluded in Africa had at least one power-sharing element. Not all were inclusive in nature. At least in name, the peace agreement between the central government and southern rebels in Sudan is a comprehensive one. However, in reality the agreement did nothing to stop the new bloody conflict in Darfur, and some would argue that it even contributed to the violence (Woodward 2006: 177). The apparent irrelevance of the agreement to other conflict zones in Sudan (and other local rebel movements) proved problematic as the ‘national cake’ (that is, oil from South Sudan and top positions in the state apparatus) was shared between only two partners.2 Equally problematic was the fact that the very title ‘Comprehensive Peace Agreement’ masked the continuing heterogeneity of the South Sudanese rebels. Yet the ‘paradox of inclusion’ is not specific to Sudan. In numerous cases in Africa, the inclusion of one rebel group in a peace agreement and subsequently in a power-sharing government has left others excluded, who then face incentives to strengthen their war efforts in order to demonstrate that they merit inclusion at a later stage. Limited inclusion may therefore undermine the sustainability of peace.
Burundi
Burundi is rightly presented as the most complex and complete case ever of power-sharing in Africa. Different reasons are invoked to explain this. One is the Belgian experience with consociational democracy, given Burundi’s history as a UN mandate territory under Belgian rule after the First World War. Another is the role of South African mediators, who drew on their own experiences with elements of consociational democracy. A third argument, advanced by Vandeginste (2009), is that Burundi’s complete form of power-sharing is best explained with reference to the country’s trial-and-error approach to finding lasting solutions to a recurrent and extremely violent conflict pattern that has claimed around 200,000 lives between 1993 and 2005, and has accounted for maybe double that number of deaths since independence. Both the war and the peace processes impacted on the main actors: the Hutu rebels were weakened militarily as a result of combat, while the Tutsi establishment was politically weakened through years of inconclusive negotiations. These factors certainly carried as much weight as the substance of the agreement shaping the outcome of the negotiations, which was a complex and sophisticated architecture of institutions and rules.
Based on the assumption that societal cleavages, and in particular the main Hutu–Tutsi divide, is a constant in public life, the main provisions for peace were fixed within the country’s constitution. Significantly, a strong emphasis on minority rights for the Tutsi (approximately 14 per cent of the population) was already encapsulated in the Burundian Constitution of 1992 and the preceding Charter of National Unity (1991). This had major effects on the behaviour of political actors, but it did not have positive consequences for any of the political parties that participated in the process. The winner of the first multi-party elections in 1993, the Front pour la Démocratie au Burundi (FRODEBU, the Front for Democracy in Burundi), lost its leader with the assassination of President Melchior Ndadaye (the first Hutu president) in a bloody coup attempt in 1993 and was entirely removed from power in a second military coup d’état in 1996. Although FRODEBU subsequently re-entered the peace process, it never regained a dominant position. FRODEBU’s old rival, the Union pour le Progrès National (UPRONA, the Union for National Progress), fared little better, losing political weight (and ultimately ministerial positions and parliamentary seats) to a number of more extremist parties and (ex-)rebel organizations.
In the wake of the civil war that raged following Ndadaye’s assassination until 2005, progress at the negotiation table was at first extremely slow. This can be explained by the tactical moves and material motives of some stakeholders, but also by the antagonistic positions that they held. The power-sharing deal reached on 8 October 2003 in Pretoria contained provisions on territorial power-sharing and reinforced the significance of the provincial level of government through the introduction of ‘political’ governors. Significantly, rebels were given a share of the posts of governor and municipal administrator positions in the 2003 peace negotiations. Additionally, the Burundian peace agreement, which was later codified in the constitution, fixed the ethnic quota for the distribution of governmental seats. A Constitutional Referendum was held in February 2005, and the resulting document enshrined a new set of power-sharing rules regulating access to positions of power. Accordingly, parliament must be 60 per cent Hutu and 40 per cent Tutsi. In the Senate, each of the 17 provinces must be represented by a Hutu and a Tutsi. Moreover, the cabinet must also feature a 60:40 Hutu/Tutsi split. Similar provisions were fixed for all other public positions.
Burundi’s first post-war parliamentary elections, in July 2005, were won by the former rebel movement Conseil National pour la Défense de la Démocratie-Forces pour la Défense de la Démocratie (CNDD-FDD, National Council for the Defence of Democracy – Forces for the Defence of Democracy), but both UPRONA and FRODEBU obtained substantial parts of the vote and thereby gained their share of power. Pierre Nkurunziza, the leader of the most important rebel movement CNDD-FDD, was indirectly elected president thereafter and all three parties formed a coalition government. The second rebel movement, Forces Nationales de Libération (FNL, the National Forces of Liberation), joined the peace process belatedly in 2006, and those four parties/(ex-)rebel movements were expected to become the main building blocks of the party system, with smaller parties excluded from parliament as a result of the adoption of a 5 per cent threshold. However, this was not what materialized.
Elections are obviously critical to the viability of this architecture. In 2010, a new series of elections was held, and both the electoral process and outcome risked jeopardizing the peaceful transition. The party of President Nkurunziza (CNDD-FDD) won municipal elections (held prior to national elections) by a large margin and other parties cried foul.3 However, while reports of widespread intimidation against opponents of the regime are indisputable, the complaints of opposition parties were not supported by local and international observers. Subsequently, most parties opted to boycott both presidential and parliamentary elections.4 This had major consequences, as parties not contesting the polls were barred from joining a broad-based coalition government. Following the election, no major party other than the former single-party UPRONA (associated with Tutsi minority rule in the 1970s and 1980s) and the ruling CNDD-FDD was represented in parliament. A split-off from FRODEBU (FRODEBU Nyakuri) under former party president Jean Minani received five seats and one ministerial position in government, but neither FRODEBU nor the FNL – both credited with substantial Hutu support – gained seats. UPRONA obtained three positions in the cabinet. Indirect senatorial elections were even more problematic with UPRONA being able to win only the two Tutsi seats in the capital Bujumbura and Bururi province. The boycott and the warped outcome of the elections created a crisis of Burundi’s post-conflict democracy, and at the time of writing it was difficult to predict whether or not also it would lead to a new civil war. However, some features of Burundi’s consociational system may mitigate this risk: ethnic power-sharing in Burundi is organized not only between parties, but also within. If CNDDFDD and UPRONA can represent within their structures – and thus within the different layers of government – both Hutu and Tutsi communities, it may help to prevent the inequalities underpinning party competition in the country from exploding into fresh violence.
Côte D’ivoire
On 19 September 2002 a rebel attack on different sites in the metropolis of Abidjan by former government soldiers of northern origin started a low-intensity civil war. The rebellion did not achieve its aim to take over power from President Laurent Gbagbo, but the government also failed to roll back the rebels beyond the southern part of the country. Former student leader Guillaume Soro became the leader of the rebels, claiming to represent the marginalized north of the country. After lukewarm support for the government in the first weeks of the confrontation, French troops stationed in Côte d’Ivoire stabilized the situation. This was the starting point for a drawn-out power-sharing process.
The list of negotiations and negotiators is long. After a first round under the aegis of the Economic Community of West African States (ECOWAS), a complex attempt was made by Côte d’Ivoire’s former colonial power France, leading to the much-cited agreement of Linas Marcoussis, signed on 23 January 2003. Participants of the negotiation were all political parties represented in parliament or government plus the three rebel organizations of the time. The agreement addresses the majority of the most salient political problems of the country, including questions of citizenship, which have implications for electoral eligibility, human and civil rights and land ownership. The deal ultimately proved unsuccessful, in part because the proposed reforms would have required a change of constitution and this was not popular with the ruling Front Populaire Ivoirien (FPI, Ivorian Popular Front). However, just as important to the failure of the agreement was the role of France as mediator, and the power-sharing formula subsequently imposed (during a second Paris summit immediately after the Linas Marcoussis agreement). The government, and particularly the army leadership, found it simply unacceptable that the rebels should get the defence and interior portfolios, ministries that would give them preponderance in all security issues.
As a result, some concrete details regarding the formation of the government of national unity were left to a new summit in Accra (Ghana) on 7 March 2003 when the rebel groups, now united under the label Forces Nouvelles (FN, New Forces), obtained two senior ministries (territorial administration and communication). Additionally, a 15 member-strong National Security Council was established, in which all parties were represented. The Accra developments could be interpreted as a second layer of power-sharing (within the military), building on the first layer that was put forward in Paris (the sharing of government positions). Subsequently, the consensually elected Prime Minister Seydou Diarra was able to build a grand coalition that included 10 ministries for the FPI, seven each for the two other main political parties – Rassemblement des Républicains (RDR, Rally of the Republicans) led by former Prime Minister Alassane Ouattara, and Parti Démocratique de la Côte d’Ivoire (PDCI, the Democratic Party of Côte d’Ivoire) led by former President Henri Konan Bédié – while nine went to the FN and six to smaller parties. The earlier impasse regarding the interior and defence ministries was resolved via a compromise that saw them assigned to technocrats.
However, the country remained geographically divided and this division was entrenched by the presence of ECOWAS and UN peacekeepers. The territorial division was in no way the result of the power-sharing agreement, but nonetheless had some of the same effects as a formal system of decentralization or regional autonomy, as the northern part of the country was no longer under the command of the central administration. Even essential commercial flows were controlled by rebel leaders for at least four years.
In the period that followed one camp refused to disarm; the other refused to change the constitution to allow for fair elections. This led to two Pretoria summits in 2005, which were attended by Gbgabo, Soro, Ouattara, Bédié and Diarra: minor political parties were no longer part of the game. The most important aspect of these meetings concerned details of article 35 of the Ivorian Constitution concerning eligibility for political office: the refusal of incumbents to allow rival leaders to contest the presidency had been an important source of political tension. South Africa’s Thabo Mbeki identified a way to avoid a constitutional referendum by interpreting article 48 of the Ivorian Constitution (on the exceptional rights of the president) as an instrument through which President Gbagbo, after consultation with the president of the National Assembly and the Constitutional Court, could authorize the candidatures of personalities that would be presented by those individuals who had signed the agreement of Linas-Marcoussis. This looked like an apt juridical way out of the deadlock that had been caused by Gbagbo’s insistence that the constitution must be followed.
However, new confrontations of a smaller scale continued to take place and neither the disarmament of the rebels and of pro-government militias, nor other preconditions for elections, were fulfilled in time to allow them to take place in 2005. In turn, the South African mediation effort lost credibility when the calendar of the peace process could not be kept. At the same time, a new prime minister was imposed on Gbagbo. Konan Banny, the ambitious former director of the regional central bank and PDCI member, was believed to be a true challenge to Gbagbo’s rule, but Gbabgo succeeded in curtailing Banny’s influence and 2006 proved to be a year of deadlock. Things changed in 2007. After a month of intense negotiations in what was called ‘direct dialogue’ between Gbagbo and Soro, the Agreement of Ouagadougou was signed on 4 March 2007. The agreement represented a much more viable power-sharing arrangement. The mediator and third signatory, President Blaise Compaoré of Burkina Faso, was an ‘insider’ in contrast to the mediators of all preceding agreements. He was very close to the rebels, but gradually became aware of the nefarious effects of the Ivorian crisis on his country and so had good reason to seek a permanent settlement to the dispute. Under his guidance a so-called ‘permanent consultation framework’ was established which consisted of Gbagbo, Soro, Compaoré, Bédié and Ouattara, and was significant because it gave all of the main players a function during times of crisis. The content of the agreement was rather specific and gave an indication of how two of the major bones of contention should be resolved: the issuing of identity papers was to occur after the holding of special decentralized registration sessions (which then formed the basis of the electoral registration), and the creation of a unified army was to be advanced via the creation of an integrated military command centre in which the government’s and the rebels’ chiefs of staff held joint authority. This represented an important step in building trust between the rival parties. However, the elite power-sharing deal between President Gbagbo and rebel leader – and by this point Prime Minister – Soro was not made part of the agreement, although it was clearly related and a pre-condition for securing Soro’s signature. This created a situation whereby Soro was explicitly rewarded for taking up arms.
The permanent consultation framework proved useful during a series of political crises in 2010. Gbagbo first dissolved the government and the Electoral Commission in February, provoking an outcry from the opposition. Thanks to the established consultative procedure, however, it was possible quickly to form a new broad coalition government and a new Electoral Commission.5 The consultation framework also proved instrumental to resolving disputes when, after lengthy disagreement, discussions between the four key leaders mediated by Compaoré led to an agreement to accept a final voters’ list on 6 September 2010. As the identification of voters and citizens were parallel processes, and the marginalization of certain communities within both economic and political life was one of the main reasons for the rebellion of 2002, this touched at the heart of the Ivorian crisis.
Despite the progress made after 2007, the power-sharing arrangement was not sufficiently resilient to withstand the controversy that followed the contested presidential elections on 28 November 2010. According to the Electoral Commission, opposition candidate Ouattara won in the second round, securing 54.5 per cent of the votes against Gbagbo’s 45.5 per cent (first round Gbagbo 38 per cent, Ouattara 32 per cent, Bédié 25 per cent). Subsequently, the Constitutional Council (the highest court in the land, which consisted mainly of Gbagbo supporters) invalidated the results in a couple of northern constituencies so that the president would retain power with 51.5 per cent of the vote.
Almost all international actors and bodies responded by siding with Ouattara and demanding Gbagbo’s resignation. The UN even went so far as to ‘certify’ the figures provided by the Electoral Commission, conferring legitimacy on Ouattara’s claim to be the rightful president of Côte d’Ivoire. The standoff between Gbagbo and Ouattara sparked armed incidents leading to several dozen casualties, mainly in Abidjan. Ouattara responded to Gbagbo’s intransigence by naming a shadow cabinet under Soro and seeking to take control of the various government institutions. Violence escalated in Abidjan and western Côte d’Ivoire (which probably cost more lives than the original civil war). Ultimately, with the support of French troops, Ouattara’s forces were able to surround and then capture the former president.
The Ivorian power-sharing process shows some interesting characteristics: all governments since the first major peace agreement in Linas-Marcoussis in January 2003 were broad based and included a strong majority of non-FPI ministers. However, Gbagbo’s party managed to secure some of the key portfolios and the share of FPI ministers in the cabinet rose from about one-quarter to one-third between 2003 and 2010, to the detriment of the RDR and PDCI. This may have given the impression that Gbagbo was less and less willing to share power with others. However, consecutive prime ministers (Diarra, Banny and Soro) were distinctively non-FPI. Moreover, non-Gbagbo factions secured substantial concessions: Ouattara was ultimately allowed to contest elections and his supposed supporters were registered to vote and given the necessary identity papers, while the Election Commission was placed under non-FPI rule. This did not prevent the post-electoral crisis, however, perhaps because too few provisions to secure the fate of electoral losers and ensure them an ongoing role in the political process were included in the relevant agreements.
Liberia
Liberia’s first experience with power-sharing was meant to end the country’s first civil war, which had begun in 1989. In August 1995 the main warring factions signed an agreement brokered by Ghana’s then-President Jerry Rawlings, in Abuja. Charles Taylor, leader of the strongest military organization, agreed to a ceasefire. A ruling council of six members was formed, including three main warlords (Taylor, Alhaji Kromah, and George Boley). Other factions were included in the government, which could not, however, effectively run the affairs of the country as they continued to work against each other. Hostilities continued and only ceased after a further agreement in Abuja, Nigeria later that year. The viability of the agreement was not tested over a long period. Instead, Liberia rushed to hold presidential and parliamentary elections in 1997, leading (after widespread intimidation) to the clear victory of Charles Taylor.
Amos Sawyer (2004: 451), a prominent Liberian intellectual who led a governance reform coalition in 2010, criticized the peace settlement and in particular its power-sharing content. Sawyer argues that the power-sharing government was ‘substantially, if not totally, controlled by armed groups whose leaders could hardly find in such arrangements sufficient incentive to blunt their greed and ambition’. Indeed, the peace thus brokered proved unsustainable. Taylor’s election enabled him to continue his warlord politics, now in the guise of a elected president. In 1999 war broke out again after the emergence of a Guinean-backed rebel group, the Liberians United for Reconciliation and Democracy (LURD). This rebellion was later joined by a second military group, the Movement for Democracy in Liberia (MODEL), operating from Côte d’Ivoire.
The renewed civil war ended in 2003, with the LURD rebels close to a decisive military victory; Charles Taylor handed power to his vice-president, Moses Blah, and left the country. The three warring parties and 18 civilian party representatives met in Akosombo and Accra, Ghana, from 4 June 2003 to 18 August 2003 within the framework of the ECOWAS Peace Process for Liberia and under the auspices of ECOWAS Chairman and Ghanaian President John Kufour. The talks were mediated by General Abdulsalami Abubakar, former head of state of Nigeria. The Accra Agreement contained some clear power-sharing elements. At the military level, all of the warring parties contributed representatives to the newly established National Commission for Disarmament, Demobilization, Rehabilitation and Reintegration. The Accra Agreement determined that state forces could be drawn from the ranks of the present government of Liberia forces, the LURD and the MODEL, ‘as well as from civilians with appropriate background and experience’ (Article VII, 1 (b) of the Accra Agreement). No territorial power-sharing was provided for in the Accra Agreement, although economic power-sharing was regulated in annex 4 to the agreement, with each of the three warring parties being given four positions at state corporations and two positions in national agencies (and 10 state corporations/ 14 national agencies attributed to the collectivity of ‘political parties and civil society’).
At the political level, while Taylor’s vice-president was allowed to run government affairs for a short period until October, the transitional institutions that were developed were clearly designed according to standard power-sharing prerequisites. Significantly, it was agreed that no representative of a warring faction could hold the position of chairman or vice-chairman in the transitional government. Thus, independent businessman Gyude Bryant became head of the executive branch of government. However, 15 out of 21 cabinet posts were allocated between the two rebel movements and the former Taylor government – in other words, once again the warring factions were given a strong majority. Taylor’s faction retained internal affairs, defence, planning and economic affairs, health and social welfare, and post and telecommunications, which meant that Taylor’s defence minister was thereby allowed to continue in his job. LURD received transport, justice, labour and finance, and the ministry of state, while MODEL secured agriculture, commerce, foreign affairs, public works, and land, mines and energy.
The problem of rebel leaders with dubious intentions gaining control over civilian posts was replicated in the legislature. The National Transitional Legislative Assembly was composed of the three warring parties (with 12 seats each), political parties (18 in total), civil society and special interest groups (7), plus one representative each of the different counties (15 in total) (Article XXIX, 4). Moreover, a controversial LURD leader was made the interim speaker of the Legislative Assembly. This was meant to be an interim arrangement but nonetheless reveals what motivated those who signed the agreement: jobs, cars, and money. The heavy involvement of peacekeepers subsequently enabled early national elections, held at the end of 2005 under a new constitution. However, although these polls allowed for the transition to a civilian government, they also permitted several key figures of Taylor’s regime and warlords of the type of the notorious Prince Johnson to become elected legislators and thereby acquire immunity. Taylor, by contrast, was indicted by the UN-sponsored Special Court for Sierra Leone, and finally ended up behind bars in The Hague.
The Accra Agreement has been heavily criticized. In the view of Amos Sawyer (2004: 454), ‘fixing the central state is important but insufficient … Authority must be constitutionally shared at other levels of government and local people must become empowered participants.’ The alternative security architecture envisaged by Sawyer would have been organized across national borders and would have involved, where appropriate, religious bodies and community militia units. This is a useful suggestion, given that numerous locally based ethnic disputes have fuelled the civil war at the national level. Furthermore, Sawyer’s insistence on the need for international involvement in maintaining the peace makes sense given that for all of the externally brokered power-sharing agreements in Liberia’s history only the strong and active presence of international peacekeepers achieved a respectable degree of stability and security.
The experiences of Burundi, Côte d’Ivoire, and Liberia with power-sharing arrangements exhibit both similarities and differences. Liberia is an example of an agreement driven by material interests that worked to some extent because of the strong presence of peacekeepers. Côte d’Ivoire is an example of an iterative approach that led to a partially viable agreement and dealt with some of the important causes of conflict, but failed to produce a set of rules that bound actors in the long term. Burundi is the only example of a case in which actors sought to use constitutional devices and the over-representation of a minority group to regulate a long-enduring confrontation.
On the basis of this summary a general lesson may be drawn. In all three examples, power-sharing provisions operated in tension with established electoral practice; in other words, we need to think about the implications of power-sharing both on (short-term) conflict mitigation and (long-term) democratic consolidation in plural societies. The choice of institutions can have an impact on whether actors choose peace or conflict, but may also shape formation, competitiveness and survival of political parties. Rewarding or banning parties for being dominantly mono-ethnic, choosing an electoral system that guarantees former militia leaders will remain within the political elite (and that others will not), and representing regional interests via a newly introduced Senate may undermine the position of existing civilian political movements with significant – although often unintended – consequences for democracy.
Conclusion
There are a number of issues that warrant closer inspection: the question of the ‘amount’ of power-sharing, the local-national nexus, the appropriate negotiation of power-sharing, and the idea of power-sharing as a process constituted by a series of episodes. Walter (2002) has claimed that the more power-sharing, the better, but what is more? Should we understand this to mean that more dimensions are included (political, military, territorial and economic power-sharing if we follow Hoddie and Hartzell’s distinction)? That more (relevant) actors are included? Or that more substantive power is shared between the government and its contenders? All three aspects are likely to be important.
How inclusive an agreement is depends on the structure of the formal process of peace negotiations. We have seen that the mediation practice is rather heterogeneous. ‘More inclusive’ does not always mean ‘more effective’. Yet to exclude some of the main players is always problematic. Quite frequently, part of the leadership of a rebel movement agrees to the terms of a peace agreement while others, who do not feel accommodated, initiate a split within the organization, in many cases sparking new violence. Now leading rebel movements in their own right, such ‘spoilers’ often emerge from renewed conflict with a stronger claim to a place at the table in the next round of negotiations.
Measuring and evaluating other aspects of power-sharing agreements is also complex. Scholars could better define what it means for an agreement to be successful; it seems somewhat arbitrary to simply state that a deal ‘worked’ because it was followed by X months of peace. Similarly, determining the extent or degree of power-sharing is a complicated exercise. A first step would be to identify the ‘real’ positions that confer power on actors in a given society, as this may not be the same across countries and cultures. Revisiting the classical literature on power in Africa may therefore be appropriate to find out whether the most fundamental power relations are touched at all by power-sharing deals. To be more explicit, as long as the president of the republic remains the one that hands out the sinecures and makes the most important decisions with regard to appointments to ‘juicy’ positions, redistributing ministries or administrative posts will not result in a fundamental change in the rules of the neopatrimonial game.
The dominant assumption in the existing literature is that the sharing of national power leads to a territorially uniform and locally meaningful peace process – the level of analysis on which scholars operate is more or less systematically the nation-state. However, this assumption is not grounded empirically. Studies that look more specifically at sub-national arenas particularly affected by violence would permit us to advance more far-reaching conclusions with regard to the contribution of power-sharing agreements to peace on the ground. National elites cannot simply determine how politics is played at the local level. A more realistic assumption is that national peace and power-sharing accords are unlikely to act as the foundation for countrywide peace if they ignore local constellations of actors and their interests.
The relative prominence of negotiators may also merit greater scrutiny. In some cases, mediators are more active than in others: at times, they sign the document themselves and represent a ‘power guarantor’ capable of lending legitimacy to the agreement, but may also be discreet to the point of not being mentioned. One potential explanation may lie in the different post-agreement roles that mediators are assigned to play. However, we must then ask who assigns them a specific role? The more informal aspect of such processes also needs to be considered. The realm of African peace negotiators is growing. Some of the continent’s leaders are – or were – frequently invited to engage in peace talks, such as Gabon’s Omar Bongo Ondimba until his death in 2009. When mediating negotiations, many drew on their own country’s experience with power-sharing (such as Presidents Mandela and Mbeki in South Africa), while others enjoyed a close relationship to some participants, putting them in a unique position in relation to the conflict concerned (see the above discussion of Compaoré in Côte d’Ivoire).
Path-dependency is an equally important concept to consider. Initial power-sharing agreements rarely resolve all problems; further agreements are frequently needed after new outbursts of violence. It is often the case that the development of effective power-sharing deals comes only after many years of ‘trial and error’. Historical institutionalism – the search for ‘critical junctures’ in peace processes – thus merits more attention. By working within this framework scholars may be able to develop an account of power-sharing that would enable key actors to value past, better understand apparently ‘failed’ peace agreements, and develop best practices.
Notes
1For more on these variations see Erdmann and Basedau (2008: 248). Note that the level of disproportionality between the distribution of voters and the distribution of seats in the legislature grows, the smaller the constituencies.
2The failure of peace negotiations for Darfur is well described by an insider in Alex de Waal’s ‘I Will Not Sign’, London Review of Books, www.lrb.co.uk/v28/n23/waal01_html (accessed 22 December 2010).
3CNDD-FDD received 62 per cent, FNL 15 per cent, UPRONA 8 per cent, and FRODEBU (the clear winner of the first free elections in 1993) only received 6.4 per cent.
4Nkurunziza as the sole candidate obtained 91 per cent of the votes. In parliamentary elections CNDD-FDD obtained 81 seats, UPRONA 17 and the FRODEBU Nyakuri splinter group five seats. Indirect senatorial elections produced 32 CNDD-FDD senators and two UPRONA senators.
5Soro was again named prime minister and could present a full government on 4 March. Most instrumental was the nomination of a PDCI member at the helm of the Electoral Commission, replacing another PDCI representative accused by the presidential camp of rigging the registration process.
Bibliography
Bunce, V. and Watts, S. (2005) ‘Managing Diversity and Sustaining Democracy: Ethnofederal versus Unitary States in the Postcommunist World’, in P. Roeder and D. Rothchild (eds) Sustainable Peace: Power and Democracy after Civil Wars, Ithaca, NY: Cornell University Press.
Cheeseman, N. and Tendi, M. (2010) ‘Power-Sharing in Comparative Perspective: The Dynamics of “Unity Government” in Kenya and Zimbabwe’, Journal of Modern African Studies 48(2): 203–229.
Crawford, G. and Hartmann, C. (eds) (2008) Decentralisation in Africa: A Pathway out of Poverty and Conflict, Amsterdam: Amsterdam University Press.
Derouen, K., Lea, J. and Wallensteen, P. (2009) ‘The Duration of Civil War Peace Agreements’, Conflict Management and Peace Science 26(4): 367–387.
Erdmann, G. and Basedau, M. (2008) ‘Party Systems in Africa: Problems of Categorising and Explaining Party Systems’, Journal of Contemporary African Studies 26(3): 241–258.
Hartzell, C. and Hoddie, M. (2007) Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars, Pennsylvania: Pennsylvania State University Press.
Hoddie, M. and Hartzell, C. (2003) ‘Civil War Settlements and the Implementation of Military Power-Sharing Arrangements’, Journal of Peace Research 40(3): 303–320.
Horowitz, D. (1985) Ethnic Groups in Conflict, Berkeley, CA: University of California Press.
Jarstad, A.K. and Nilsson, D. (2008) ‘From Words to Deeds: The Implementation of Power-Sharing Pacts in Peace Accords’, Conflict Management and Peace Science 25: 206–223.
Lijphart, A. (1977) Democracy in Plural Societies: A Comparative Exploration, New Haven, CT: Yale University.
Mehler, A. (2009) ‘Peace and Power Sharing in Africa: A Not So Obvious Relationship’, African Affairs 108(432): 453–473.
Pearson, F.S., Lounsbery, M.O., Walker, S. and Mann, S. (2006) ‘Rethinking Models of Civil War Settlement’, International Interactions 32: 109–128.
Roeder, P.G. and Rothchild, D. (2005) ‘Power Sharing as an Impediment to Peace and Democracy’, in P. Roeder and D. Rothchild (eds) Sustainable Peace: Power and Democracy after Civil Wars, Ithaca, NY: Cornell University Press.
Rothchild, D. and Foley, M. (1988) ‘African States and the Politics of Inclusive Coalitions’, in D. Roth-child and N. Chazan (eds) The Precarious Balance: State and Society in Africa, Boulder, CO, and London: Westview.
Sawyer, A. (2004) ‘Violent Conflicts and Governance Challenges in West Africa: The Case of the Mano River Basin Area’, Journal of Modern African Studies 42(3): 437–463.
Spears, I.S. (2000) ‘Understanding Inclusive Peace Agreements in Africa: The Problems of Sharing Power’, Third World Quarterly 21(1): 105–118.
Sriram, C.L. (2008) Peace as Governance: Power-sharing, Armed Groups and Contemporary Peace Negotiations, New York, NY: Palgrave Macmillan.
Sullivan, D. (2005) ‘The Missing Pillars: A Look at the Failure of Peace in Burundi through the Lens of Arend Lijphart’s Consociational Theory’, Journal of Modern African Studies 43(1): 75–95.
Vandeginste, S. (2009) ‘Twenty Years of Trial and Error’, Africa Spectrum 46(3): 63–86.
Walter, B. (2002) Committing to Peace: The Successful Settlement of Civil Wars, Princeton, NJ: Princeton University Press.
Woodward, P. (2006) ‘Peacemaking in Sudan’, in O. Furley and R. May (eds) Ending Africa’s Wars: Progressing to Peace, Aldershot: Ashgate.