Abstract
To end the civil war in Sierra Leone the government and the Revolutionary United Front (RUF) signed a peace agreement guaranteeing power-sharing in July 1999. Such power-sharing is a widely used, often recommended political arrangement to overcome deep divisions between groups. However, scholars disagree on whether power-sharing causes peace, or, on the contrary, causes continuing violence. One reason for this is the literature's tendency to neglect how power-sharing is actually put into place. But post-agreement implementation is essential if we are to judge the performance of power-sharing. Therefore, we investigate the role played by power-sharing in terminating the civil war in Sierra Leone. We argue that the government was able to use the peace agreement to pursue its goal of ending the war through marginalising the RUF.
Sierra Leone experienced an eleven-year civil war during the 1990's between the Revolutionary United Front (RUF) and the Government of Sierra Leone. By the time the war was declared to be over in 2002, between 30,000 and 75,000 people had died and thousands of others had suffered atrocities such as gang rape, sexual slavery, torture, forced enlistment, forced labor, starvation, dismemberment, mutilations, and looting (Lord 2000). Atrocities were carried out by both parties, but especially the RUF were infamous for maiming and mutilating civilians. Still, the peace agreement signed in Lomé on 7 July 1999 called for an amnesty and a power-sharing government including the incumbent SLPP (Sierra Leone People's Party) and the RUF.
Power-sharing is a widely used, often recommended political solution to overcome deep divisions between (violent) groups. Its main characteristic is a joint government. Even though power-sharing has been used to ease post-election tensions in Kenya and Zimbabwe and to terminate civil wars in, for example, Burundi, Sudan, Nepal and Lebanon, the relationship between sharing power and stability is still debated. Scholars disagree as to whether it causes peace (e.g. Hartzell and Hoddie 2007), or, on the contrary, causes continuing violence (e.g. Tull and Mehler 2005).
The civil war in Sierra Leone is a crucial case when it comes to understanding the relation between power-sharing and peace. It took less than a year from when the Lomé Agreement was signed until power-sharing broke down, and United Nations (UN) and British troops were needed to secure a fragile peace. Unfortunately, because it ended rather quickly, researchers have concluded that power-sharing failed and have not shown interest in how it actually played out in the months before its breakdown. 1 However, analysing the post-agreement implementation is essential if we are to judge the performance of power-sharing.
To a large extent this mirrors a shortcoming in the general power-sharing literature as well. Quantitative studies, in particular, have studied power-sharing's effect on post-conflict stability by looking at the text of peace agreements (Walter 2002; Hartzell and Hoddie 2007). Less attention has been given to how stipulations in the agreements have been implemented (Jarstad and Nilsson 2008).
Therefore, we investigate the role played by power-sharing in terminating the civil war in Sierra Leone. Even though the peace agreement prescribed a number of power-sharing provisions, the SLPP was able to use the agreement to pursue its goal of ending the war at the expense of the RUF. Disagreements over the agreement splintered and weakened the rebels. Additionally, the government managed to tie the agreement to the Constitution and thus keep the reigning government in place, with the RUF sitting in the cabinet as a political minority. This further marginalised the RUF, and, with the rebels split and rebel leaders settled in Freetown, it became an easy task to arrest rebel leader Foday Sankoh and eliminate the threat from the RUF when a rebel faction kidnapped UN peacekeepers in May 2000.
This article is organised as follows. The first section gives a brief introduction to the background and developments during the civil war. Thereafter we introduce the concept of power-sharing and explain why it is viewed as especially helpful in creating postwar peace, but also why others argue the opposite. The third section describes the power-sharing agreement signed in Lomé in July 1999 and the extent to which the provisions were implemented. In the fourth section we discuss how the SLPP government, although militarily weak, was able to use power-sharing to defeat the RUF through political marginalisation. We conclude by summing up the core argument and point out how lessons from Sierra Leone contribute to the understanding of power-sharing. 2
We are grateful for funding from the Norwegian Ministry of Foreign Affairs and the Norwegian Research Council. We would also like to thank two anonymous reviewers as well as participants at various seminars and conferences for comments.
Civil War in Sierra Leone
The Sierra Leonean civil war officially began when the RUF, a small rebel group headed by Foday Sankoh, invaded eastern Sierra Leone from Liberia on 23 March 1991. The invasion force numbered approximately one hundred guerilla fighters primarily from Sierra Leone who had been recruited and trained in Libya and/or Liberia. Accompanying the RUF were Liberian soldiers from the National Patriotic Front of Liberia (NPFL) and mercenaries from Burkina Faso. Nigerian and Guinean troops assisted the Government of Sierra Leone in the counter-offensive (Keen 2005; Richards 1996). 3
The RUF's stated aim was to overthrow the APC (All People's Congress) government, accusing it of being centralised, corrupt, and neglecting socio-economic development outside the capital. The gradual withdrawal of the state from rural areas since independence in 1961 and the collapse of the patrimonial system of governance left many feeling disenfranchised and excluded (Richards 1996; Bøås 2001; Keen 2005).
By the end of 1991, the RUF controlled much of the diamond-rich southern and eastern regions of Sierra Leone, having quickly overpowered a weak national army that pre-war president Joseph Momoh had disarmed and stopped funding (Gershoni 1997). Government soldiers, unhappy with conditions on the war front, ousted Momoh in a coup in 1992 and installed the National Provisional Ruling Council (NPRC) as the new government. The NPRC contracted mercenaries from Executive Outcomes (a private South African security group) to fight the RUF in return for diamond mining concessions. Despite the successful intervention of Executive Outcomes on behalf of the government, the RUF controlled vast swathes of the country's territory by 1995. A military coup in 1996 removed the head of the NPRC from power (Clapham 1996; Conteh-Morgan and Dixon-Fyle 1999; Hirsch 2001). In February 1996, democratic elections were held, but the RUF chose not to participate. Ahmad Tejan Kabbah, representing the SLPP, won the presidential election and threw his support behind local pro-government Civil Defense Forces (CDF) comprised of ethnic/hunter militias (the largest of which are the Kamajors) that had been active since the early days of the war.
The government of Sierra Leone and the RUF met in Abidjan for talks in February 1996. A cease-fire was signed in March, and discussions continued through July, culminating in the signing of a peace agreement between the government and the RUF on 30 November 1996. However, another military coup in 1997 forced President Kabbah to flee to Guinea. The new ruling junta, called the Armed Forces Revolutionary Council (AFRC), was sympathetic to the RUF and invited the RUF to join the government, resulting in stiff opposition at home and abroad (Keen 2005).
In retaliation for the coup, the Economic Community of West African States Monitoring Group (ECOMOG), the cease-fire and military monitoring wing of the Economic Community of West African States (ECOWAS), began attacking the AFRC, finally ousting the junta in February 1998 and restoring Kabbah and his government to power on 10 March 1998. But the AFRC and the RUF continued to fight against ECOMOG and the reinstalled government. Despite the presence of ECOMOG and the reinstatement of the elected government, the RUF and AFRC controlled much of the country. In January 1999 RUF and AFRC rebels launched the particularly brutal “Operation No Living Thing” on Freetown (Olonisakin 2008: 32f). 4 ECOMOG successfully repelled the attack in conjunction with the Kamajors (Conteh-Morgan and Dixon-Fyle 1999; Fithen and Richards 2005; Funke and Solomon 2002; Hirsch 2001; Keen 2005).
Which groups were responsible for the attack on Freetown remains unclear. RUF veterans interviewed in Sierra Leone in September 2008 reject that the RUF was involved. Most believe the RUF was at least involved, although some argue the West Side Boys (AFRC hardliners) were the main drivers (Keen 2005: 222).
Renewed peace negotiations took place in Togo in May 1999, and a power-sharing peace agreement was signed in the Togolese capital of Lomé on 7 July the same year. 5 However, fighting continued between the RUF, AFRC, ECOMOG, CDF and United Nations peacekeepers (UNAMSIL, United Nations Mission in Sierra Leone). The agreement collapsed completely when the RUF kidnapped 500 UNAMSIL peacekeepers, and this was followed by the RUF's shooting of protestors outside of Sankoh's Freetown home just days after the kidnapping. As a result of the shooting, Sankoh and other members of the RUF were arrested and stripped of their positions in government. A final cease-fire and demobilisation agreement was signed in Abuja, Nigeria in 2001 and UNAMSIL's strength was expanded to 17,500 troops to assist with the demobilisation of fighters and in the training of the Sierra Leonean armed forces (Fithen and Richards 2005; Funke and Solomon 2002; Keen 2005; Peters 2004).
The events following the signing of the peace agreement will be thoroughly discussed later.
President Kabbah declared “di war don don”– the war is at an end – on 18 January 2002. Presidential and parliamentary elections were held in May 2002; Kabbah was re-elected as president and his party, the SLPP, won 83 of 112 legislative seats. The RUF's political party, RUFP, did not win a single seat in parliament. 24,000 RUF and 37,000 CDF combatants had completed the disarmament programme before the elections (Keen 2005: 268).
Power-Sharing as a Peacebuilding Tool
To evaluate the role played by power-sharing in Sierra Leone we need to know more about the theoretically expected relationship between power-sharing and peace. 6 The aim of power-sharing as a conflict resolution tool is to overcome combatants' security concerns by including them in decision-making bodies. 7 As such, power-sharing is elite-centred, aiming at formal and informal cooperation among top leaders, without necessarily taking mass-level reconciliation concerns into account (Mehler 2009). Although the intention is to create peace by inclusion, there is little consensus whether power-sharing actually fulfils this goal. While much has been written about power-sharing and peace, we limit ourselves to discuss only the central proponents and their critics. 8
Researchers define power-sharing differently, but most scholars agree that the Lomé Agreement in Sierra Leone is a power-sharing agreement (Hartzell and Hoddie 2007; Jarstad 2008; Mehler 2009; Sriram 2008). We therefore do not attach ourselves to a specific definition nor discuss the differences between various definitions.
Power-sharing was initially introduced as a governing system aimed at democratic governance in deeply divided societies (Lijphart 1977). See Jarstad (2008) for a discussion of how scholars of democracy theory and conflict management view power-sharing differently.
Power-sharing is also evaluated according to its effect on post-conflict democratisation (Jarstad 2008), however, this is beyond the scope of this paper.
According to scholars optimistic about the effect of power-sharing, parties to a civil war will lay down their weapons and participate in peaceful post-conflict government only if they are guaranteed positions with decision-making power (Walter 2002; Hartzell and Hoddie 2007). Parties to civil war fear a post-conflict situation where their adversary gets sole access to power and uses this power to threaten the interests of others (Hartzell and Hoddie 2007). The fear of being overruled hinders combatants from seeking peace unless they can be absolutely sure that their enemies will comply with the terms of agreement (Walter 2002). Hartzell and Hoddie (2007) argue that these security concerns can be reduced if access to power is limited and shared between conflict parties. 9 On the other hand, Tull and Mehler (2005) claim that power-sharing creates new insurgency. They argue instead that such agreements reward violent behavior and provide recognition and legitimacy to warmongers. Because of power-sharing, the road to state power is “closed to non-violent political actors” (Tull and Mehler 2005: 391). Therefore, power-sharing between rebel groups and the government provides an incentive for other power seekers to arm and fight to gain access to power through expanding the power-sharing arrangements. In addition, rebel hardliners might disagree with concessions at the negotiating table and refuse to disarm (Jarstad 2008).
Walter (2002) additionally argues that in order for power-sharing to function as a security provider, civil war settlements depend on third party security guarantees. In a post-conflict situation, where belligerents are increasingly vulnerable because of demobilisation and disarmament, parties to war will seek outside protection to avoid marginalisation if opponents attempt to break the agreement.
Thus, on the one hand, there is the claim that including rebels in decision-making positions through power-sharing is the best alternative in terms of establishing peace. This proposition is supported mainly by large-N statistical analyses, which find correlations between various dimensions of power-sharing and post-conflict peace (Hartzell and Hoddie 2007; Mattes and Savun 2009; Walter 2002). On the other side, a number of in-depth case studies reveal a less rosy trajectory, where power-sharing agreements fail to bring peace and stability (Spears 2002; Sriram 2008; Tull and Mehler 2005). One reason for these diverging views can be different interpretations of what power-sharing is, however, a more likely explanation is differing expectations of what power-sharing should achieve. Whereas quantitative studies require only the absence of large-scale civil war to define stable peace, qualitiative studies often aim for improved human rights protection and good governance as well.
The positive effect of power-sharing on post-conflict peace can be linked to two conditions. First, combatants' security concerns are reduced when they gain increased influence through power-sharing. Inclusion overcomes the fear of marginalisation as well as addresses (legitimate) claims to participation. Second, combatants' security concerns are reduced when their former foes signal credible commitment to peace by accepting to share power. Because the costs of signing power-sharing agreements are high (Hartzell and Hoddie 2007), former combatants' commitment to the agreement will be credible. When security concerns recede, the incentive to resume violence recedes as well. The negative effect of power-sharing, on the other hand, works mainly through the creation of spoilers (Stedman 1997). This comes either in the form of new or splinter insurgency groups, or through the negative, but unintended consequence of freezing or strengthening divisions between groups (Rothchild and Roeder 2005; Jarstad 2008). 10
See, however, Gilbert (2009) on the difference between permanent and transitional power-sharing agreements in Africa.
Nonetheless, both proponents and those more critical of power-sharing agree that inclusion of warring parties in government is a common peacebuilding strategy, one that is in particular favoured by outside actors pushing for peace negotiations (Hartzell and Hoddie 2007; Jarstad 2008; Mehler 2009). While focusing on positive or negative consequences of power-sharing, neither of these opposing hypotheses say much about the actual implementation of power-sharing agreements. Overall, how former antagonists carry into effect co-operative measures in peace agreements is seldom studied. Although Stedman, Rothchild and Cousens (2002) discuss challenges of implementing agreements by outside actors, in particular the United Nations, they overlook the challenges to – and responsibilities of – the warring parties themselves. 11 Obviously, challenges emerge after a peace agreement is signed. Our argument is that the effect of power-sharing on peace should not be accredited to the provisions written on paper, but to how these provisions are realised. Therefore, in the next sections we describe the power-sharing agreement in Sierra Leone and the implementation of its terms, and in particular discuss how the SLPP government was able to out-manoeuvre and marginalise the RUF.
Jarstad and Nilsson (2008) provide one of the few quantitative analyses of implementations of power-sharing and durable peace. However, they do not discuss the implementation process itself, only argue that power-sharing concessions entailing high costs are more likely to result in stable peace than less costly concessions.
Power-Sharing in Sierra Leone
Both during the Abidjan negotiations in 1996 and the Lomé negotiations in 1999, the RUF's main demands were a power-sharing government, the vice-presidency for rebel leader Sankoh and the removal of foreign troops from Sierra Leone. The government initially rejected these demands. During the 1996 negotiations, the government had the upper hand militarily and was able to force Sankoh to sign the Abidjan accord even though it included neither power-sharing nor Sankoh's vice-presidency. By 1999, however, the situation had changed and the government had to accept both power-sharing with the rebels and the status (but not the authority) of vice president for Sankoh.
Kabbah began negotiating again with the RUF to end the war when it became clear that Nigerian troops were likely to be pulled out of Sierra Leone after the impending elections in Nigeria in February 1999 (Reno 2001). Nigerian troops comprised the bulk (approximately 90%) of peacekeeping troops serving under the ECOMOG banner in Sierra Leone, and their departure would leave the Sierra Leonean government vulnerable (Keen 2005). Withdrawal of Nigerian troops, coupled with the government's loss of territorial control and control over the army, 12 put the government in a very weak position to win the war militarily.
In April 1999 70% of Sierra Leone's territory was outside government control and the main highway leading from Freetown to the rest of the country was under RUF control (Bangura 2000).
The RUF's superior position by 1999 gave Kabbah little choice but to negotiate a peace agreement (Abraham 2004). Both parties realised that the conflict was not winnable. Kabbah was still viewed as the legitimate president of Sierra Leone. However, rejecting negotiations would have led to losing international and regional sympathy. Moreover, it meant continuing violence and instability (Rashid 2000). The situation for the RUF/AFRC was equally difficult. According to Rashid, “the alliance could either transform its control over these areas to freedom for its leaders, amnesty for its war crimes and legitimate political power through negotiations, or continue to fight an unwinnable war and be treated as pariahs” (2000: 28).
During the first months of 1999 Kabbah and Sankoh met regularly. 13 In addition, both the RUF/AFRC alliance and the government held consultations separately to prepare for negotiations. 14 Among the requirements for peace from the RUF/AFRC were: the establishment of a power-sharing transitional government that would remain in power for four years; the withdrawal of all foreign forces fighting in Sierra Leone; participation of RUF/AFRC members in a new Sierra Leonean army; and recognition of RUF/AFRC control over certain areas (RUFSL 1999; see also Rashid 2000). Although the report from the government's preparatory consultations recommended a broad-based Commission for the Consolidation of Peace (CCP), the report “strongly opposed any form of power-sharing with the AFRC-RUF” (Rashid 2000: 30). Kabbah himself also rejected any form of power-sharing in his opening speech at the preparatory conference, and Kabbah's cabinet further refused such a proposal, threatening a cabinet revolt in June 1999 if power-sharing was implemented (Hirsch 2001).
Sankoh had been sentenced to death for his role in the 1997 coup and was appealing against the sentence from prison in Freetown. Kabbah met him in prison, and also allowed others to meet him. Among those were military leaders, foreign diplomats etc. (Rashid 2000).
Although the preparatory meeting was between the RUF and the AFRC, tensions between the two allies were deep. AFRC leader Koroma was not present at the discussions, in fact, some argue he was held captive by the RUF (Keen 2005: 255; Rashid 2000: 29).
Provisions to address military and humanitarian issues were adopted in early June, while questions of Sankoh's freedom and status and the RUF's demand for a transitional power-sharing remained difficult issues to agree upon. Talks and positioning continued during the summer, but the agreement was finally signed on 7 July 1999, including a variety of power-sharing arrangements.
Provisions to Share Power within the Lomé Agreement
One of the main power-sharing arrangements was a grand coalition cabinet with four ministerial and four deputy ministerial posts allocated to the RUF in an 18 member cabinet. Furthermore, the agreement called for the establishment of a Commission for the Consolidation of Peace, to “implement a post-conflict programme that ensures reconciliation and the welfare of all parties to the conflict, especially the victims of war” (Lomé Agreement 1999: Article VI). The Commission was to be composed of two members from civil society, and “one representative each named by the Government, the RUF/SL [The Revolutionary United Front of Sierra Leone] and the Parliament”. The peace agreement established a number of such broad and inclusive commissions, strengthening the power-sharing character of the agreement.
Another power-sharing characteristic of the Lomé Agreement was the inclusion of the new Sierra Leonean Army (SLA). Whereas the Abidjan Accord (1996) mentioned only the RUF in relation to this military power-sharing, the 1999 agreement stated that “those ex-combatants of the RUF/SL, CDF and SLA who wish to be integrated into the new restructured national armed forces may do so provided they meet established criteria.” (Lomé Agreement 1999: Article XVII, 2).
Perhaps the most crucial power-sharing mechanism was granting RUF leader Sankoh the chairmanship over the newly established Commission for the Management of Strategic Resources, National Reconstruction and Development (CMRRD) – a position that had the status of vice-president. While the RUF/AFRC's initial demand for recognition of their control over certain areas was not explicitly addressed in the Lomé Agreement, Sankoh's chairmanship was de facto control over the diamond rich areas in the east and north of the country. Sankoh's position was probably the single most important element in convincing the RUF to sign the peace agreement. 15
The agreement also guaranteed amnesty for all combatants and collaborators. This was so controversial that the UN representative signed the agreement with the stipulation that “amnesty and pardon shall not apply to international crimes of genocide, crimes against humanity, war crimes and other serious violations of international humanitarian law” (Francis 2000: 366).
The many power-sharing provisions illustrate well how the agreement reflected the situation on the ground, with a militarily weak government and international pressure to strike a deal, and Sankoh's power to influence that deal. Nonetheless, the peace agreement stated that power-sharing was restricted to last only until the next elections, scheduled in the Constitution to take place in just two years time (Lomé Agreement 1999: Article VI, 10). Thus, even though the provisions rewarded the RUF/AFRC, it was only for a limited time period.
Implementing the Power-Sharing Provisions
The power-sharing deal was signed, but were the stipulations carried out? To some extent, they were. On 21 October 1999 three RUF and one AFRC representative were included in president Kabbah's reshuffled cabinet, and two days later RUF leader Sankoh was appointed as chairman of the CMRRD. Although not stated in the agreement, on the same day, AFRC leader Johnny Paul Koroma was appointed as chairman for the Commission for the Consolidation of Peace. By including Koroma, Kabbah showed flexibility and seeming willingness to meet the AFRC rebels' concern about being excluded from the power-sharing arrangements. In the months that followed, Sankoh and Koroma travelled throughout the country, often together, explaining the peace agreement to their followers and urging them to participate in the Disarmament, Demobilisation and Reintegration (DDR) programme (IRIN News, 10 November 1999). 16 The two also joined president Kabbah in various meetings abroad. Four weeks after his inclusion in the cabinet, Minister of Trade and Industry Mike Lamin (RUF) represented Sierra Leone at a World Trade Organisation meeting in the USA; the other rebel ministers represented the government on various other occasions.
Gberie (2005: 164) claims that when Sankoh spoke in English he encouraged the rebels to disarm while in the Mende and Temne languages he told them not to.
However, the ministries assigned to the RUF and AFRC were not the ones mentioned in the agreement. Article IV of the agreement states that
“the Government of Sierra Leone shall give ministerial positions to the RUF/SL in a moderately expanded cabinet of 18 … as follows: (i) one of the senior cabinet appointments such as finance, foreign affairs and justice; (ii) three other cabinet positions” (Lomé Agreement 1999: Article IV).
In fact, the RUF was allocated the Ministry of Trade and Industry; the Ministry of Land, Housing and Central Planning; the Ministry of Energy and Power; and the Ministry of Tourism and Culture. The RUF complained that it had not received one of the ministries promised it in the Lomé Agreement, while the government took the position that
“the expression ‘such as’ should not mean ‘that is’ and it could therefore designate any posts considered to be of the same standing as those indicated” (Bright 2000: 38).
The RUF also argued that it had not been allocated the diplomatic and parastatal positions promised to it. Sankoh expressed his complaints both in speeches and letters. He claimed that the RUF ministers in the power-sharing Cabinet were marginalised by the government, arguing that neither the ministries nor his own Commission (CMRRD) were provided with suitable offices and facilities. 17 The government rebutted this by maintaining that the delays and imperfect implementation of the Lomé Agreement were the results of RUF's foot-dragging (Sierra Leone Web, 31 December 1999). Nonetheless, according to RUF and AFRC veterans, the RUF and AFRC representatives in the Cabinet were hindered in influencing high politics and were even excluded from Cabinet meetings. 18 Kabbah also expanded the Cabinet from 18 to 21 ministers, diluting the power of the RUF and AFRC within the government.
Various letters and statements from the RUF and AFRC are available from the Sierra Leone Web: http://www.sierra-leone.org/AFRC-RUF.html. See also IRIN News for Sankoh's criticism of cabinet appointments (22 October 1999).
Interview with one of the AFRC/RUF ministers, Freetown, September 2008.
Clashes continued in the provinces between the RUF and the AFRC, ECOMOG troops, CDF, the SLA and UNAMSIL, and hindered efficient implementation of the DDR process. The initial deadline set for the disarmament of the estimated 45,000 combatants in the country was 15 December 1999, but by late January only 13,000 had surrendered their arms (IRIN News, 31 January 2000). The DDR process proceeded slowly and efforts to reintegrate rebels in a new national army were poor. RUF combatants complained that they were not offered positions in the new army. During winter and early spring 2000 there were recurring fights between the different groups, and, especially the RUF, seized weapons, food and equipment from ECOMOG and UNAMSIL forces, and abducted UN peacekeepers. 19
Sankoh vehemently disagreed with the deployment of UNAMSIL, viewing it as illegal and inconsistent with the Lomé Agreement, despite the fact that Article XVI calls for the establishment and deployment of a peacekeeping force (Gberie 2005). UN peacekeeping forces were nicknamed “beachkeepers”, and Sankoh claimed the soldiers were only interested in the Sierra Leonean women and argued the money spent on UNAMSIL should rather be spent on his own followers (Sierra Leone Web, 14 February 2000).
Things came to a head in May 2000, when the RUF kidnapped and took hostage 500 Zambian UNAMSIL peacekeepers in Makeni. Civil society groups organised a large demonstration outside Sankoh's home in Freetown, calling on him to release the peacekeepers and stop a reported (but false) RUF advance on Freetown (Richards and Vincent 2008). RUF soldiers guarding the house opened fire and 21 civilians were shot dead (Gberie 2005: 167). This caused Sankoh to flee, but he was captured ten days later and imprisoned. The other RUF representatives in the government were also arrested, effectively putting an end to power-sharing less than a year after the peace accord was signed.
The Relationship between Power-Sharing and Peace
The above section shows that power-sharing was partly implemented in Sierra Leone. However, it also reveals that power-sharing was not put into effect exactly as the Lomé Agreement prescribed. Additionally, the agreement contained statements in fine print that gave the SLPP government more room to manoeuvre than the RUF/AFRC probably understood. Even though the SLPP was in the weaker position during the Lomé negotiations, at least in terms of military capacity, it was able to obtain approval for an agreement that empowered and strengthened the Kabbah government as the legitimate political authority. As such, power-sharing was used by one party to the conflict to pursue its goal: defeat of its adversary through political marginalisation and eventual elimination.
To what extent this was a constant deliberate strategy, as claimed by a government representative at the negotiations in Lomé, 20 or a stroke of luck for the SLPP is difficult to discern today, ten years after the peace agreement was signed. In hindsight, there is ample evidence that then-President Kabbah and the SLPP managed to out-manoeuvre the RUF/AFRC, both during the negotiations and afterwards. Consistent with Jarstad's (2008: 118) claim, the agreement and its implementation strengthened the divisions between the RUF and AFRC, and created splits between the top leadership and rank-and-file soldiers. Furthermore tying the agreement to the Constitution enabled Kabbah to keep both hands on the steering wheel when the agreement's provisions were to be implemented. Finally, the “sunset clause” for the sharing of power shortened the RUF's available time to transform into a political party and compete in the elections (Richards and Vincent 2008).
Interview in Freetown, August 2008.
The Lomé Agreement was written and signed as a bilateral agreement between the government and the RUF. Because of this, it failed to accommodate the interests of, or treat as parties to the agreement, the AFRC as well as ex-SLA-personnel, and the CDF (including the Kamajors) (Abraham 2004; Alao and Ero 2001; Francis 2000). Although the leaders of RUF and AFRC constantly insisted they were united, deep divisions between the two groups had grown, particularly after the 1999 invasion of Freetown (Rashid 2000). A number of clashes between the two and the RUF's ousting of AFRC troops from Makeni in October 1999 confirm the split. Sankoh also faced increasing problems with Sam “Maskita” Bockarie, one of the RUF's top commanders. He refused to surrender weapons, in particular to Nigerian soldiers, regardless of whether they were ECOMOG or UNAMSIL troops. This made him a worrying spoiler threat (Stedman 1997), but, fearing that Sankoh planned to assassinate him, Bockarie fled to Liberia in December 1999. He was, however, not the only one dissatisfied with how the peace agreement was implemented (IRIN News, 22 December 1999). Many of the RUF rebels did not trust President Kabbah and wanted Sankoh to live in Kailahun until the government had fulfilled the terms of the agreement. 21 Sankoh's decision to live in Freetown rather than among his men in the provinces disappointed his followers.
Interviews with RUF veterans (Makeni, September 2008).
The fact that Sankoh established his residence in Freetown, where he could easily be captured, and contrary to his followers' advice, can be interpreted as signalling a strong commitment to the peace process. 22 Such costly concessions are expected to increase the prospects of stable post-conflict peace (Hartzell & Hoddie 2007). Besides his harsh words when pointing out violations of the Lomé Agreement, Sankoh also repeatedly insisted that he was sincerely supporting the peace agreement. Abraham (2004: 218; see also Bright 2000), on the other hand, argues that the RUF was never sincere in the peace negotiations and that Sankoh only participated in the peace process to buy time so he could seize power through violence. Gberie (2005: 166) claims Sankoh was preparing a coup when he was arrested in May 2000. While it is a complicated task to assess sincerity, it seems likely that Sankoh and the other rebel leaders were pleased with their new status. Being located in the capital, they enjoyed power and benefits previously inaccessible to them. This inclusion in government probably appeased the rebel leadership to such an extent that they considered power-sharing a better option than continued war. From an elite perspective the RUF and AFRC's security concerns were addressed, as anticipated by researchers favourable to power-sharing (e.g. Hartzell and Hoddie 2007; Mattes and Savun 2009). In contrast, lower rank rebels saw few benefits from power-sharing. Reintegration of ordinary rebels both into society and the new national army was slow and, to a large extent, failed, at least during the first year after the peace agreement was signed. Both Sankoh and Koroma encouraged their followers to disarm, but continuing clashes between the warring factions made combatants reluctant to do so.
Interview with one of the AFRC/RUF ministers, who was convinced that Sankoh was sincere in his commitment to the peace agreement (interview in Freetown, September 2008).
Drawing on the “spoiler problem” (Stedman 1997), Tull and Mehler (2005) and Jarstad (2008) discuss how power-sharing agreements can encourage splinter groups and new insurgents to continue with violence and obstruct peace processes. While this can certainly be the case, in Sierra Leone such fragmentation weakened the rebel side instead and enabled the SLPP to take advantage of these splits, strengthening its own image as the responsible and legitimate government.
While the Lomé Agreement called for a power-sharing transitional government, it also stated that the incorporation of the RUF into governance would be “within the spirit and letter of the Constitution” (1999: Part Two). This gave the Constitution priority over the peace agreement. 23 The agreement recognised the government as the legitimate political authority whose rule the RUF would be subjected to and where they sit in the minority. Richards and Vincent see the UN as having a strong hand in this:
Whether the government's eagerness to abide by the constitution reflected a sincere commitment to maintain a democratic policy, or a patrimonial logic of retaining power at all costs, remains to be settled.
“In Sierra Leone the UN interest lay in supporting the Kabbah government … The UN was determined, at the end of the 1990's, to improve its peacekeeping image by bolstering the Sierra Leone government's chosen strategy for ending the war” (2008: 91).
Thus, the SLPP retained the structure and powers of the ruling government, giving it a significant tactical advantage. Instead of accepting the RUF's demand for a transitional government with equal participation from the SLPP and the RUF, the newly empowered government could control the appointment of ministries and the structure of the post-accord government. Because of this, the SLPP was able to under-compensate the RUF and AFRC during the implementation of the agreement by allocating weak ministries, a powerless position to Sankoh, 24 and diluting cabinet power through expanding the number of ministers.
Sankoh's position had the status but not the authority of a vice-president; it was not a cabinet-level position, and control over the country's diamond mines remained with the Ministry of Mines (Interviews with various individuals in Sierra Leone, September 2008; see also Richards and Vincent 2008).
In addition, connecting the agreement to the Constitution ensured that the transitional period would be as short as possible. In order to adhere to the constitutional arrangements requiring elections every fifth year, the next elections to be held in the country were to take place in 2001. However, the limitation of the time period when power-sharing was guaranteed did not seem to worry the RUF. This can partly be explained by Lomé providing a (perceived) lucrative position for Sankoh, 25 and by the RUF's misconception of their support among the Sierra Leonean people. The RUF was convinced it would win the coming election. Sankoh might have gained significantly more votes in the 2002 elections if he had not been imprisoned and stood as candidate for the presidential elections instead. He was apparently very good at distributing patronage, and for example funded education and health services in RUF-controlled areas. But the RUF probably lacked any genuine public support due to its brutal tactics. 26
One of the government officials who attended the Lomé talks stated in an interview that once Sankoh was granted the chairmanship of the CMRRD, the negotiations went very easily. “The real sticking point was Sankoh's own position” (Interview by authors, September 2008).
Interviews with RUF elites and civil society representatives (Freetown, September 2008). See also Richards and Vincent (2008) on the reasons behind RUF's political disintegration and dismal performance in the 2002 elections.
The provision for a major international peacekeeping force in the Lomé Agreement was of tremendous importance (Richards and Vincent 2008). It provided the government with a security guarantee to essentially do what it wanted in the aftermath of the agreement. The UN mission “was authorized to use deadly force if needed” and to “tackle potential spoilers” (Olonisakin 2008: 42), this is exactly the kind of third party security guarantee Walter (2002) argues is necessary in order to ensure that power-sharing agreements are successfully implemented. 27 Even though only factions of the RUF and AFRC were largely responsible for sporadic violence, it was quite clear that the RUF as a whole embodied the greatest spoiler to the peace process. Agreeing to share power with the RUF and AFRC had one important consequence for the government: it brought key RUF players into town, where they could be kept under constant watch by the government. Because Sankoh and much of the RUF leadership were in Freetown, it was possible to immediately arrest and imprison RUF members following the May 2000 events. Once Sankoh was removed from the picture, the RUF was effectively dead in the water and collapsed as its fighters were tired and unwilling to continue the fight, and as disagreements and factions within the RUF undermined its prior unity and strength (Keen 2005).
UN peacekeeping was crucial in terminating the civil war. However, elaborating on its role is beyond the scope of this article. See for example Olonisakin (2008) and Williams (2001) for analyses of UN and British troops in Sierra Leone.
Conclusion
For Sankoh and the RUF, power-sharing was a main issue and an implacable demand during negotiations. For them, “power-sharing and transitional government meant substantial control over the state apparatus” (Rashid 2000), which was what they had been fighting for all along. It was probably not a realistic alternative to have Sankoh sign a peace agreement if it did not include the RUF in post-conflict decision-making processes. To terminate the devastating civil war was a global concern, and if power-sharing was needed to end it, the SLPP government, neigbouring countries and the UN were willing to accept this measure. However, even though the Lomé Agreement over-compensated the RUF on paper, with amnesty provisions, participation in a transitional Cabinet, Sankoh's chairmanship over CMRRD with the status as a vice-president, and military power-sharing through inclusion of former rebels in a new national army, the government was able to under-compensate the RUF/AFRC during the implementation of the agreement. The power-sharing agreement enabled the SLPP government to take advantage of splits in the RUF/AFRC, and to pacify key RUF members and keep them in Freetown, making them an easy target when rebels, unhappy with the Lomé Agreement, broke its terms. In this regard, power-sharing played a significant role in terminating the civil war in Sierra Leone. Even though excluding the RUF during implementation of the agreement was probably not a main purpose, it was not completely accidental either. President Kabbah was extremely reluctant to share power, but was “in a box” (Gberie, 2005: 157) and had to give in to world-wide pressures to strike a deal. It was not surprising that Kabbah used the possibilities that emerged to marginalise the RUF. Although power-sharing ended, a fragile peace slowly emerged. British troops, initially intervening to evacuate foreign citizens, defended Freetown from rebels (Alao and Ero 2001: 128). By early autumn 2000 the RUF controlled half of Sierra Leone's territory and 90% of the diamond areas (Keen 2005: 265). But the rebel forces gradually softened. 28
Many factors caused this. In the border areas the RUF was attacked by Guinean forces, causing heavy casualties and the RUF to become more positively inclined towards UN peacekeeping. The UNAMSIL also improved its performance and in March 2002 numbered 17,544 peacekeeping troops. Because of international pressure on the diamond trade banning “blood diamonds” and sanctions against Liberia, the RUF's finances and supplies were reduced. Finally, the British intervention and their retraining of the SLA convinced many RUF rebels that the war was not winnable, and a general feeling of war-weariness among the RUF made the rebels soften and prepared for something other than continued war (Keen 2005: 267ff).
Developments in Sierra Leone after the signing of the peace agreement in Lomé in July 1999 show that implementation is crucial. How power-sharing provisions are put into place is essential to understand the relationship between power-sharing and peace. Looking only at the terms of an agreement is not enough to fully comprehend the challenges of peacebuilding transitions, researchers must also know more about the implementation process. Besides, mediators and policy makers must be aware of the problems associated with effectuating power-sharing agreements. In addition to the post-agreement trajectory in Sierra Leone, recent experience in Sudan and Zimbabwe also suggests that, although incumbent governments sign power-sharing agreements, the preferable strategy is marginalisation rather than inclusion. Unfortunately, this strategy creates a new challenge. Why would rebels sign peace agreements if their opponent's aim is to use power-sharing to win the war by other means? While we are reluctant to make generalisations based on one case, it seems that mediators and third parties face two alternatives: either to provide enough financial, diplomatic and military support to the preferred party, or to make sure concessions are significant enough and sincerely implemented so rebels lay down their weapons permanently.