The challenges of power-sharing and transitional justice in post-civil war African countries

Comparing Burundi, Mozambique and Sierra Leone

Dr Sadiki Koko is a Research Associate with the Department of Politics and International Relations, University of Johannesburg. He holds a doctorate degree in Political Studies from the same university. His areas of specialisation include identity politics, and conflict and peace processes in Central Africa and the Great Lakes region.


This article uses the cases of Burundi, Mozambique and Sierra Leone to analyse transitional justice processes in African societies where power-sharing was used as a key tool to end very protracted and violent civil wars. It is argued that, by affording warring parties a prominent role in the post-settlement political environment, power-sharing inadvertently impeded the pursuit of both restorative and criminal justice in all three countries. As an instance of ‘warriors’ justice’, power-sharing was used by such actors as an opportunity to avoid facing retributive justice. Indeed, due to the central position they held within the power-sharing dispensations, former warriors emphasised amnesty while paying lip service to reparations for victims. In all three countries, the decision to revert to the international judicial system or not was mainly motivated by political calculations rather than any genuine concern for justice. However, notwithstanding the shortcomings above, the consensus brought about by the power-sharing dispensations enabled the three countries to effect meaningful institutional reforms, albeit with limited and different levels of success.


The post-1990 world environment has witnessed a radical shift in the nature of wars around the world, characterised by a decrease of interstate wars and an increase in the number of internal or civil wars. Civil wars take place in individual states and involve civilians as both fighters and principal victims of atrocities. The fact that civil wars result in the commission of large scale atrocities against civilians brings about the need for the pursuit of post-conflict or transitional justice as one of the conditions for achieving sustainable peace.

Transitional justice thus poses a major challenge to societies emerging from civil war. The situation becomes even more complex when dealing with post-conflict societies where the resolution of the conflict involved negotiations and the establishment of an all-inclusive power-sharing transitional mechanism. As former fighters become new rulers and are directly involved in state management under the transitional power-sharing dispensation, several questions arise. Can those responsible for committing atrocities during the fighting be expected to champion the cause of justice once in power? Can those responsible for wartime atrocities be dragged into transitional justice mechanisms without endangering the fragile post-settlement peace? Does power-sharing as a conflict resolution strategy promote impunity and undermine justice in post-conflict societies?

This article seeks to answer the questions raised above (and others), relating to the challenge of pursuing transitional justice in post-violence societies where the resolution of the conflict involved power-sharing.

It discusses the cases of three African countries that have in recent years reverted to power-sharing as the main mode of war termination, namely Burundi, Mozambique and Sierra Leone.1 Two criteria have dictated the selection of the case studies. Firstly, all cases involved herein relate to civil wars that ended in negotiations (as opposed to military victory by one of the fighting parties). Secondly, the chosen cases are spread evenly across three different African regions, namely central Africa (Burundi), southern Africa (Mozambique) and western Africa (Sierra Leone) as a way to ensure geographical balance.

The main argument of this article is that, while each of the three countries analysed herein resorted to some form of transitional justice mechanisms to deal with civil war-related atrocities, they only applied justice aspects that were not meant to threaten their elites’ political survival, freedom and personal security. This situation was the result of the war termination strategy applied in these countries, namely negotiations followed by power-sharing, which had suddenly turned former warring parties into
key stakeholders in the transitional dispensations of these countries.

Before discussing the specific country cases, it appears important to clarify the concepts of power-sharing and transitional justice, attempting to highlight philosophical tensions between the two.

Conceptual clarification: Power-sharing and transitional justice

Debating power-sharing

As concepts, power-sharing and the different mechanisms designed for its operationalisation (government of national unity, inclusive government or coalition government) are prone to semantic confusion. In fact, two strands of research use the term power-sharing, often without recognising the differences between them (Jarstad 2008:108). The first strand, labelled the ‘consociational democracy’ school, is concerned with practical strategies of distributing power among socio-political stakeholders in divided societies as a means to guarantee adequate group representation and foster democratic participation. One of its renowned proponents, Arend Lijphart (1999), was concerned with addressing the exclusion of minorities brought about by a rigorous application of liberal democratic principles and rules in ethnically and/or religiously divided societies and where political allegiances may align with ethnic and/or religious identities.

In such societies, Lijphart argues, majority rule is not only undemocratic, but also dangerous; it spells majority dictatorship and civil strife rather than democracy. He thus proposed consociational democracy, a group-based form of democracy, grounded on a grand coalition of the political leaders of all significant segments of the plural society; mutual veto rule (which serves as an additional protection of vital minority interests); proportionality (as the principal standard of political representation, civil service appointments and public funds allocation); as well as a high degree of autonomy of each segment of society to run its own internal affairs (Lijphart 1977).

The second strand of research relating to power-sharing is rooted in the field of conflict management. ‘In this discourse the main function of power sharing is to end violence’ (Jarstad 2008:108). This practice of power-sharing generally emerges in the context of stalemated civil wars. It is based on the understanding that ‘[b]y dividing power among rival groups during the transition, power sharing reduces the danger that one party will become dominant and threaten the security of others’ (Papagianni 2008:1). This is insofar relevant as exclusion is one of the main drivers of conflicts on the African continent (Lemarchand 2006). In this regard, it is no surprise that, with its emphasis on the inclusion of non-state stakeholders (armed groups, political parties and civil society organisations) in transitional mechanisms, power-sharing has been commended as ‘a recipe for peaceful cohabitation’ (Lemarchand 2006:2).

This article focuses on the conflict management dimension of power-sharing. It examines the practice of power-sharing as an incentive towards ending internal armed conflict. In contrast to the consociational approach that is both preventive and built on a long-term perspective, the conflict management dimension of power-sharing is reactive and temporary. It seeks to address the problem of power illegitimacy through accommodative transitional mechanisms entrusted with conducting free and inclusive elections for institutional renewal in ‘post-settlement’ societies. Provisions of power-sharing in this approach are generally derived from peace (or political) agreements signed by parties and, depending on a specific conflict situation, guarantee ‘the participation of representatives of significant groups… in the executive, but also in the legislature, judiciary, police and army’ (Papagianni 2008:42).

Understanding transitional justice

Transitional justice can be defined as the full range of ‘processes, strategies and institutions that assist post-conflict or post-authoritarian societies in accounting for histories of mass abuse as they build peaceful and just states…’ (Armstrong and Ntegeye 2006:3). It seeks to

…halt ongoing human rights abuses; investigate past crimes; identify those responsible for human rights violations; impose sanctions for some of those responsible for serious human rights violations; provide reparations to victims; prevent future abuses; preserve and enhance sustainable peace and promote [community] and national reconciliation (Fombad 2008).

Overall, there is a need to distinguish between three different sets of transitional justice mechanisms. Firstly, there are those targeting perpetrators, namely trials, exiles, vetting or purges and amnesties. Secondly, there are ‘victim-oriented restorative justice processes’ (Olsen, Payne and Reiter 2010:803) that include reparations or compensations and victims’ empowerment programmes. Thirdly, there are mechanisms targeting the wider society, including institutional reforms, truth commissions, and
other public memory projects whose aim is ‘to officially recognize but pardon past acts’ (Olsen, Payne and Reiter 2010:803). However, as Bhargava (2000:54) argues, in spite of their institutional variation, the shared overarching goal of all transitional justice mechanisms is to restore the dignity of individuals and communities victimised by atrocities, deter future violations and prevent a repeat of past horrors.

Power-sharing and transitional justice: Purposes and challenges

Power-sharing and transitional justice tend to come to prominence in countries emerging from civil war ended through negotiations. Whereas power-sharing is used in such societies to enable former warriors to partake in post-war politics, the pursuit of transitional justice is generally regarded as a guarantee to address victims and society’s wartime losses as well as to prevent the recurrence of violence.

Power-sharing and transitional justice can both be very effective, but in the implementation of each of them, seriously challenging problems may arise – which have already attracted waves of criticisms. As far as power-sharing is concerned, it has been argued that the ‘[i]nclusion of warring parties in a power-sharing arrangement does not always end violence’ (Jarstad 2008:117). Power-sharing is also said to be elite-driven and to exclude the general public from matters directly affecting national life and, in so doing, to undermine democratic processes. Lastly, by providing rebels with a share of state power, the practice of power-sharing tends to ‘contribute to the reproduction of insurgent violence [as it creates] an incentive structure would-be leaders can seize upon by embarking on the insurgent path’ (Mehler 2009:455).

On its part, transitional justice poses a problem of timing as the ‘post-conflict’ context to which it is designed to apply varies from one conflict situation to another. Although only processes ‘initiated within five years following an armed conflict’ (Binningsbø et al. 2012:733) should be regarded as transitional justice, experiences in post-conflict societies are filled with instances of transitional justice mechanisms being established even over a decade after the signing of a peace agreement. In this latter situation, the decision by government to set up transitional justice mechanisms can be seen as a ploy for pursuing political goals, including settling scores with political opponents.2

Transitional justice also faces a challenge with regard to its content and/or scope. Two scenarios ought to be distinguished in this regard. In cases of internal conflicts that ended in the military victory of one of the parties, the transitional justice mechanism is designed by the winners. Hence, it is criticised as victors’ justice (Reydams 2013). In cases where the war ended through negotiations and power-sharing, transitional justice mechanisms have to be agreed upon by all parties involved, especially warring parties, resulting in what can be termed ‘warriors’ justice’ (Sadiki 2013:254). The fact that alleged human rights violators are directly involved in determining the timing and content of transitional justice leads to the criticism that transitional justice is first and foremost a political matter and justice is simply its second, sometimes far-distant, aim.

If anything, the potential tensions between power-sharing and transitional justice in post-settlement societies as highlighted above, bring about the debate regarding the operationalisation of transitional justice mechanisms or processes in scenarios involving power-sharing as a mode of civil war termination, as discussed below.

Dialectics of power-sharing and transitional justice in post-civil war Burundi, Mozambique and Sierra Leone

According to Järvinen (2004:36), ‘[o]ne of the greatest challenges to any post-conflict society is how to deal with past crimes … and other human rights abuses’. Firstly, internal wars generally occur in states with weak or dysfunctional institutions, including the justice system, which the armed conflict contributes to weakening further. In this context, institutional inefficacy is a critical factor in the inability of post-conflict societies to effectively prosecute cases of wartime crimes and human rights abuses (Souaré 2008).

Secondly, civil wars are periods of utter anarchy when thousands or even millions of crimes are committed. The sheer number of such crimes and their alleged perpetrators would make it virtually impossible for even the most willing states to prosecute – given the amount of time and resources required, as well as the overall impact of such prosecutions on the society and its peace prospects.

This section explores the manner in which Burundi, Mozambique and Sierra Leone dealt with the issue of transitional justice following the conclusion of their respective civil wars.

Background to the Burundian, Mozambican and Sierra Leonean civil wars

Since the killings of 1972 and the subsequent establishment of the Hutu People’s Liberation Party – National Liberation Front (Palipehutu-FNL), Burundi had been experiencing a low-intensity civil war as exiled Hutus (mainly based in Tanzania and Rwanda) explored ways to challenge militarily the Tutsi-controlled regimes ruling Burundi after the 1966 military coup. But, the crisis intensified following the assassination of the country’s first democratically elected and first Hutu President, Melchior Ndadaye, and other senior government officials from the ruling Front for Democracy in Burundi (FRODEBU) party, in October 1993. The assassination was the work of ‘radical’ elements within the Burundian Tutsi-dominated security and political circles who were concerned with the prospect of losing their long-held privileged positions.

The Burundian civil war was fought along ethnic lines, pitting the Tutsi-controlled national government against an array of Hutu-led rebel groups, including the Palipehutu-FNL and the National Council for the Defence of Democracy – Forces for the Defence of Democracy (CNDD-FDD).

The war resulted in the death of over 300 000 people and the displacement of 1.8 million more within the country and in neighbouring states (Tanzania, Democratic Republic of the Congo [DRC] and Rwanda) (Falch 2008:1). The war officially ended in 2008 following the signing of the Magaliesberg Peace Agreement between the Burundian government and the country’s last standing rebel group, the Palipehutu-FNL. However, the peace process had started as early as August 2000 when the Arusha Agreement for Peace and Reconciliation in Burundi (henceforth Arusha Agreement) was signed between the Burundian government and 19 political parties.

In contrast to the Burundian civil war that was essentially driven by internal dynamics, the Mozambican civil war (1976–1992) was a complex emergency triggered by several internal and external factors. On the one hand, it was a product of internal disagreements within FRELIMO3 and discontent within the Mozambican society. On the other hand, it was the result of interferences from the Rhodesian and Apartheid South African governments in their design to destabilise and, if possible, get rid of the Mozambican regime under President Samora Machel with its policy of support for liberation struggles in Zimbabwe, Namibia and South Africa.

The Mozambican civil war equally embodied the characteristics of a proxy war within the Cold War context, pitting the Marxist-oriented FRELIMO (supported by the Soviet Union) against the Mozambican National Resistance (RENAMO) backed by the USA. The war resulted in the death of between 600 000 and 1 million people (Hirsch 2009) and the displacement of many more both within the country and in neighbouring countries. Lastly, as was the case in Burundi, the Mozambican civil war ended in a stalemate, leading the protagonists to explore peaceful means to end the conflict.

On its part, the Sierra Leonean civil war erupted in March 1991. It formed part (and was the product) of the country’s troubled post-colonial evolution characterised by political repression and violence, military coups, the politicisation of (ethnic) identities, economic collapse and the uninterrupted weakening of state institutions, including the security forces. The war was waged by the Revolutionary United Front (RUF) under the leadership of Corporal Foday Sankoh. The RUF’s official goal consisted of ridding Sierra Leone of the corrupt and unjust regime under President Joseph Mommoh (1985–1992).

However, the actual situation in areas controlled by the RUF contrasted significantly with the group’s public claims. The regime of fear and extreme violence imposed by the RUF on civilian populations contributed to alienating the rebel group not only from the Sierra Leonean people but also from the international community.

By January 2002 when the civil war officially ended, an estimated 100 000 people had been killed while thousands had had their arms or limbs amputated and approximately 2 million more had been displaced within and outside the country (Tejan-Cole 2009). In an effort to find a peaceful solution to the crisis, the Economic Community of West African States (ECOWAS) dispatched a peacekeeping mission, the ECOWAS Monitoring Group (ECOMOG), to Sierra Leone in 1997. The ECOMOG was credited with, among other things, the signing of the Lomé Agreement in July 1999, later complemented by the Abuja Agreements I and II signed in November 2000 and May 2001 respectively.

Peace agreements and power-sharing in Burundi, Mozambique and Sierra Leone

The Arusha Agreement was grounded on two main principles, namely the choice of negotiation as the preferred mode to end the civil war, and the establishment of a transitional power-sharing mechanism involving all of the country’s main political stakeholders.

In spite of its significant contribution to paving the way to peace, the Arusha Agreement only involved the Burundian government and 19 political parties as all armed groups remained sceptical about the government’s commitment to a peaceful settlement of the conflict. However, in 2002, a ceasefire agreement was signed between the government and the CNDD-FDD’s wing led by Jean-Bosco Ndayikengurukiye and the Palipehutu-FNL’s wing led by Alain Mugabarabona. This was followed, in November 2003, by another agreement between the transitional government and the CNDD-FDD’s main wing led by Pierre Nkurunziza.

As per the Arusha Agreement, the transition was comprised of two phases designed to last 18 months each. In November 2001, the transition was inaugurated with the swearing in of sitting President Pierre Buyoya (Tutsi). On 30 April 2003, Buyoya handed the presidency over to his deputy, Domitien Ndayizeye (Hutu), for the second phase of the transition which was extended to August 2005 as the government failed to abide by the November 2004 deadline to organise the much anticipated general elections.

Power in transitional Burundi was shared between Hutus and Tutsis in national government, national parliament, provincial administration as well as the security forces. However, it ought to be observed that, in contrast to the civilian sector where the Hutus’ demographic superiority was recognised and implemented, power in the security sector was shared on the equalitarian ratio of 50:50 between Hutus and Tutsis. Lastly, there is need to recall that – in conforming with the recommendation contained in the Arusha Agreement for the work of the country’s transitional justice mechanisms to commence during the transition – President Ndayizeye enacted Law No 01/021 on 27 December 2004 establishing the country’s (first) truth and reconciliation commission (TRC). However, this initial TRC never got off the ground for several reasons – including the fast-approaching end of the transition, the attempt by UPRONA4 and FRODEBU to control the TRC’s working and appointments as well as protests from human rights organisations regarding the political affiliation of the TRC’s nominated officials (International Centre for Transitional Justice 2011:10). As a consequence, Law No 01/021 was abrogated by Law No 1/18 discussed in the next section.

As far as Mozambique is concerned, tentative peace talks between the government and RENAMO started as early as August 1989 in Nairobi, through the facilitation of the Mozambican Christian Council (CCM). The talks involved a direct meeting between President Joachim Chissano (who had succeeded Samora Machel following the latter’s death in 1986) and Afonso Dlakhama, RENAMO’s leader. They resulted in the issuing of three statements that established the basis for a peace accord in Mozambique, namely the ‘Twelve principles for peace’ from government, RENAMO’s ‘Six-point declaration’, as well as the ‘Seven-point proposal’ from the USA government (Almeida, Sanches and Raimundo 2010:7).

Following two years of continued direct negotiations between the government and RENAMO under the auspices of the Community of Sant’Egidio, the agreement to end the civil war in Mozambique was signed on 4 October 1992 in Rome (Italy). The General Peace Agreement, as the agreement was called, began with a statement of seven protocols, dealing with basic principles including the establishment and recognition of political parties, elections, military issues, guarantees, ceasefire and donors. The Agreement also contained a declaration on humanitarian assistance, a joint declaration on the conclusion of the peace process and other joint communiqués and declarations signed by the parties during the two-year negotiation process.

The Agreement did not provide for a power-sharing transitional dispensation at national level as was the case in Burundi (see above) and Sierra Leone (see below). Instead, it formalised the de facto ‘territorial power-sharing’ reality brought about by the armed conflict. This means that RENAMO and the FRELIMO-led government maintained administrative management over the areas under their respective control at the time of the signing of the agreement. In this regard, public administration services in RENAMO-controlled areas could employ only citizens residing in those areas (including RENAMO’s members).

At the same time, the government was requested to afford these ‘public servants’ and the institutions under their care the respect, treatment and support required for the discharge of their duties – on a basis of strict equality and without any discrimination in relation to others performing similar duties at the same level in areas under FRELIMO’s control.5

In order to ensure a smooth coordination between these two new public administrations, the parties established an eight-member National Commission (four from each side) that remained in place until the end of the ‘transition’. Inasmuch as the agreement could be commended for providing ample details regarding the country’s new political system, the electoral process and the reform of the security sector – among other things, it fell short of providing a clear and meaningful leadership on the issue of transitional justice as shall be shown later.

In Sierra Leone, the Lomé Agreement represented the main framework that helped end the civil war. It provided for power-sharing under a government of national unity comprising representatives of President Tejan Kabbah’s government, the RUF and other national stakeholders such as political parties and civil society. To this effect, Foday Sankoh was appointed Vice-President and Head of the Commission for the Management of Strategic Resources, National Reconstruction and Development; while Major Johnny Paul Koroma (who led the military coup against President Kabbah in 1997) was appointed Chairperson of the Peace Consolidation Commission (Tejan-Cole 2009:244). Furthermore, the RUF was awarded four ministries (trade and industry; energy and power; land, housing, city planning and environment; as well as tourism and culture) and four deputy ministries in the national cabinet (Hayner 2007). However, it ought to be mentioned that the Lomé Agreement lacked popular support among Sierra Leoneans mainly as a consequence of the blanket amnesty it granted for atrocities committed by the RUF during the war.

Transitional justice mechanisms in Burundi, Mozambique and Sierra Leone

Burundi’s Arusha Agreement contained provisions relating to transitional justice. It called for amnesty for all crimes committed during the conflict as long as they did not amount to genocide, war crimes, crimes against humanity and coups d’état. To this effect, the agreement called on the UN Security Council to establish an international judicial commission of inquiry tasked with investigating acts of genocide, war crimes and crimes against humanity committed in the country from independence in July 1962 to the date of the signing of the agreement (28 August 2000). Furthermore, the agreement called upon the Burundian government to work with the UN Security Council towards the establishment of an international tribunal, should the commission referred to above identify cases of genocide, war crimes and crimes against humanity committed in the country during the earmarked period. Lastly, the agreement called for the establishment of a Truth and Reconciliation Commission (TRC) tasked with establishing the truth about serious crimes and acts of violence committed in the country between 1962 and 2000, identifying victims and perpetrators and, after thorough assessment of individual cases, granting amnesty. The TRC was also required to make recommendations on reparation measures for victims and other social and political initiatives designed to foster national reconciliation and healing.

Following up on the provisions contained in the Arusha Agreement, in August 2000, parliament adopted a law granting temporary immunity to political leaders returning from exile, including those who may have committed crimes between July 1962 and August 2000. This provision was later extended to rebel leaders and fighters following their decision to join the transitional dispensation.

The UN Security Council never established the international judicial commission alluded to in the Arusha Agreement. Instead, the Council encouraged the Burundian political leadership to set up the TRC as provided for in the agreement6 and remained opposed to the government of Burundi granting any amnesty on the gross crimes as mentioned above.

On 15 May 2014, President Pierre Nkurunziza enacted Law No 1/18 relating to the establishment, mandate, composition, organisation and working of the TRC.7 This was followed by Decree No 100/286 signed by President Nkurunziza on 8 December 2014 and through which he appointed the TRC commissioners after they had been approved by parliament as provided for in Law No 1/18. It ought to be highlighted that, although commendable, this development came very late. If anything, the work of the TRC was undermined by the tense socio-political situation in which it was established as related to the controversial 2015 general elections and their aftermath. The TRC found itself outpaced by political events while the relevance of its work was questioned by different role players throughout the country’s socio-political spectrum.

In contrast to the Arusha Agreement, the General Peace Agreement for Mozambique did not provide for a transitional judicial mechanism. Instead, the Mozambican government unilaterally granted a legislative amnesty to all combatants well before the start of actual peace negotiations with RENAMO (Almeida, Sanches and Raimundo 2010:11). Such a move was designed to assure RENAMO’s leadership of the government’s commitment to a peaceful settlement.

Although both parties openly acknowledged the human and material costs of the civil war, none of them was eager to apologise publicly for their role in the violence. The shared sentiment among them was that the people of Mozambique needed not to focus on the past but rather invest their energy in a future peaceful and developing country. As Igreja (2009:278) observes, the parties to the Mozambican peace agreement

…deliberately precluded any possibility for the enactment of a mechanism for justice that could reckon with the grave abuses and war crimes. Justice was considered inimical to the peace-building process and was therefore replaced with a discourse of reconciliation that was expressed through oblivion and silence.

To this effect, just 10 days following the signing of the peace agreement, the FRELIMO-controlled Mozambican People’s Assembly adopted Law No 15/92 that effectively granted unconditional amnesty for all crimes committed in the country between 1976 and 1992. However, although the agreement emphasised the need for reconciliation, the Mozambican government never set up a truth and reconciliation commission. Furthermore, no specific reparation programmes targeting war victims were put in place. With their ‘forward-looking’ approach, the FRELIMO-led government and RENAMO rejected any idea of justice, be it retributive or restorative, in post-civil war Mozambique.

As far as Sierra Leone is concerned, it ought to be recalled that amnesty from prosecution was central to the Lomé Agreement. Article 9 of the agreement enjoined the government to ‘take appropriate legal steps to grant Corporal Foday Sankoh absolute and free pardon.’ The same measure applied to all combatants and collaborators on both sides of the conflict spectrum in respect of anything done by them in pursuit of their objectives as members of their respective organisations for the period between 1991 and 1999. In conforming to the official discourse, these measures were expected to contribute toward consolidating the peace and promoting national reconciliation.

Furthermore, the agreement called for the adoption of legislative and other measures necessary to guarantee immunity to former combatants, exiles and other persons, within and outside the country for reasons related to the armed conflict in order to ensure the full exercise of their civil and political rights and to fast-track their reintegration. It ought to be noted that all these provisions were consistent with demands from the RUF that requested a blanket amnesty for all its actions during the civil war.

The Lomé Agreement also provided for the establishment of a TRC tasked with addressing impunity, breaking the cycle of violence, providing a forum for victims and perpetrators of human rights violations to tell their stories and better understand the past, making recommendations on the rehabilitation of war victims as well as facilitating genuine reconciliation and healing (Hayner 2007:224).

The Sierra Leonean TRC was set up in 2000 through an act of parliament and lasted until 2005 when its final report was released to the public. However, the actual work of the commission only started in late 2002 as a consequence of disagreements over the suitability of commissioners put forward by the parties to the conflict, the resurgence of violence in the country and the reluctance of international donors to support a process that they regarded as controlled by former belligerents.

In its early days, the TRC focused its work on raising public awareness on its mandate, role, implementation plan and difference with the country’s other transitional justice mechanisms so as to foster public participation in its work. This initial phase was followed by the actual work of the commission that involved statements taking, hearings and the writing of the report (Ekiyor 2009).

Ultimately, the TRC received close to 8 000 statements from a variety of stakeholders (Ekiyor 2009). The recommendations put forward by the commission included a call for national, community and inter-personal reconciliation; the implementation of a reparations programme for victims (access to pensions, micro-credits, free health care, education, skills and training) as well as the establishment of a permanent Human Rights Commission (HRC). While the HRC was set up as early as 2004 and some reparation measures were carried out, the country’s post-civil war economic predicament prevented the government from undertaking any large-scale reparation programme for the victims.

Lastly, transitional justice mechanisms in post-civil war Sierra Leone also included the establishment of the hybrid Special Court for Sierra Leone (SCSL). The SCSL was not provided for in the 1999 Lomé Agreement; yet, it was popular among ordinary Sierra Leoneans who had borne the brunt of the violence perpetrated by warring parties, especially the RUF. In fact, the idea of establishing the court emerged as a result of the RUF’s determination to pursue military activities in spite of signing the Lomé Agreement. The move was regarded by President Kabbah’s government and the international community as an opportunity to put into context the open-ended amnesty provisions contained in the agreement, as they applied to the RUF.

The agreement establishing the SCSL was signed between the UN and the government of Sierra Leone on 16 January 2002. Located in Freetown, the court was tasked with prosecuting individuals who bore the greatest responsibility for serious violations of international humanitarian and Sierra Leonean law committed in the country from 30 November 1996 onward (Tejan-Cole 2009:226, 228). The court was thus designed to prosecute only those who played leadership roles in the conflict.
Hence, the low number of those who were prosecuted and eventually sentenced, including Moinina Fofana and Allieu Kondewa from the government-allied Civil Defence Forces (CDF); Issan Sesay, Morris Kallon and Augustine Gbao from the RUF as well as Alex Brima, Brima Kamara and Santigie Kanu from the Armed Forces Revolutionary Council (AFRC), all sentenced for jail terms of between six and 50 years. But, the most prominent case handled by the SCSL relates to the former Liberian President Charles Taylor who was sentenced in May 2012 to a 50-year imprisonment for his role in the Sierra Leonean civil war.

Politicised power-sharing and impeded transitional justice in post-civil war Burundi, Mozambique and Sierra Leone

The section above has highlighted the manner in which Burundi, Mozambique and Sierra Leone have implemented transitional justice processes following their respective civil wars. Despite some similarities, a number of differences among the three countries have been noted – attributable to the specific conditions within each state and the peculiar context of their respective civil wars and post-war situations. However, in taking into account the role played by power-sharing in the design and the implementation of transitional processes in the three countries, the paragraphs below highlight some salient trends derived from Burundi, Mozambique and Sierra Leone that may be applicable to countries having to pursue transitional justice in the aftermath of civil war ended through power-sharing.

Manipulated power-sharing and its negative impact on retributive and restorative justice

Whereas transitional justice in post-conflict countries where internal war has ended in military victory by one of the parties tends to turn into victors’ justice, the cases of Burundi, Mozambique and Sierra Leone have revealed that where war ended through negotiations and power-sharing, it tends to turn into warriors’ justice. In these latter settings, justice is only pursued in its dimensions that do not threaten the political survival (and sometimes the physical freedom) of the main protagonists, in spite of their possible involvement in the commission of crimes and atrocities during the war period. This is mainly due to the fact that, in most cases, the suspected criminals are directly involved in state management during the transition and beyond. This state of affairs helps understand the total absence of lustration measures in transitional justice processes implemented in post-war Burundi, Mozambique and Sierra Leone as analysed in this article.

Although in Sierra Leone, some senior armed groups’ leaders (including Foday Sankoh) were successfully prosecuted, this only became possible after international peacekeeping intervention had significantly boosted the position of President Kabbah’s government, bringing the country into a de facto situation of a civil war won by the government.

Furthermore, while Mozambique chose to totally ignore all aspects of retributive and restorative post-war justice and Burundi failed to effectively implement several aspects of transitional justice mechanisms contained in the Arusha Agreement, Sierra Leone ought to be commended for actually establishing a fully functional TRC and the SCSL that successfully prosecuted a number of individuals implicated in the commission of the gravest crimes during the country’s civil war.

Elite-driven Truth and Reconciliation Commissions with a negative impact on retributive justice

The fascination with TRC processes, as far as Africa is concerned, can be traced back to South Africa. In order to deal with the past legacy of violence that included centuries of white minority domination, 45 years of legalised racial discrimination and decades of armed resistance, South Africa adopted the TRC in 1995 on the premise that the country could only move forward by uncovering and documenting the truth as well as reconciling with itself. Furthermore, the context of the emergence of the South African TRC model was conducive to its adoption in many African countries that, like South Africa, had chosen negotiation and power-sharing as preferred mechanisms to end their respective civil wars.

As highlighted in the previous section, no truth and reconciliation commission or mechanism was ever provided for in the General Peace Agreement for Mozambique. In Burundi, the process of establishing the TRC provided for in the Arusha Agreement experienced several setbacks as a result of the eagerness of the two largest political parties to the Agreement and the transition (UPRONA and FRODEBU) – joined since by the ruling CNDD-FDD – to ensure their direct control over this transitional justice mechanism (International Centre for Transitional Justice 2011:10).

In Sierra Leone, the TRC achieved much under very difficult circumstances. Its work enabled an open national debate on the root causes of the country’s civil war and the implementation of some reparation programmes for the victims. Yet, it needed the political elite’s support for the effective implementation of its recommendations. In both Burundi and Sierra Leone, the main perception is that the TRC was designed and managed in a way that does not threaten the physical freedom and political worthiness of political elites. In spite of the opportunity it provided to the country’s people to uncover the truth surrounding the civil war, the Sierra Leonean TRC still fell short of imposing meaningful retributive measures (judicial prosecution or vetting) to senior political elites, including those involved in the commission of wartime crimes and atrocities.

Power-sharing and the imbalance between demanded amnesties and deserved reparations

Amnesty represents the only transitional justice mechanism that cuts across all the three countries analysed in this article. Of course, internal predicaments specific to each country determined the scope and content of amnesty laws and measures adopted by each of them.

It should be admitted that the prominent place afforded to amnesty in the three countries was, to a very large extent, a consequence of the adoption of power-sharing as the principal means to end the war in the three countries.

In all three countries, some forms of amnesty laws and measures were enacted prior to the conclusion of peace agreements between the government and rebel groups as a demonstration by the former of its commitment to a peaceful settlement. On the other hand, the adoption of amnesty laws or measures was a response to demands from armed groups that regarded such action as sine qua non for their commitment to a negotiated settlement. The case of Sierra Leone is eloquent in this regard.

Still, after inclusive transitional institutions were put in place, former warring parties (either affiliated with the former government or with former rebel groups) used their newly earned privileged positions to issue further amnesty measures and laws designed to benefit wartime crime suspects. This was for instance the case in Burundi where a number of amnesty measures and laws have been taken since the signing of the Arusha Agreement in 2000.

While power-sharing was central to the adoption of wide amnesty measures and laws in post-civil war Burundi, Mozambique and Sierra Leone, it did not make similar provisions for reparation measures aimed at victims. In this regard, Mozambique totally ignored the victims of what was actually a very atrocious civil war with devastating consequences for very large numbers of the civilian population. On its part, the Arusha Agreement for Burundi failed to include victims’ reparation as a specific aspect of the country’s transitional justice package. Perhaps, there was an expectation from all involved parties that the issue of victims’ reparation would be addressed by the country’s TRC as provided for in the Agreement. Only Sierra Leone implemented victims’ reparation measures following on the TRC’s recommendations.

As may be learned from the Burundian, Mozambican and Sierra Leonean cases analysed in this article, power-sharing (as a means to end civil war) tends to provide former warring parties an opportunity to request and secure amnesty laws and measures for themselves. It does not necessarily cater as much for the victims. To this end, it is not exaggerated to argue that, in spite of efforts to address its impunity aspects, power-sharing brings an imbalance between amnesty and victims’ reparations in post-civil war societies. Indeed, power-sharing significantly reduces the possibility for those suspected of committing wartime crimes and atrocities to face retributive justice. At the same time, as a result of the central role it affords to those suspected of committing wartime crimes and atrocities in the post-war dispensation, power-sharing tends to place the fate of war victims – including their quest for justice – in the hands of their very victimisers.

Power-sharing which avoids or politicises referrals to the international judicial system

In Mozambique, parties to the General Peace Agreement, i.e. the FRELIMO-controlled government and RENAMO, never made any reference to the international judicial system as a means for addressing crimes and atrocities committed during the country’s civil war. In Burundi, the Arusha Agreement left open, in theory, the possibility for a recourse to the international judicial system in the form of an international tribunal. However, in practice, no internationally-inspired court was set up in/for Burundi. In similar vein, no referral to the international judicial system was made by Burundian transitional and post-transitional authorities.

Sierra Leone represents the only country, among the three analysed in this article, to have made use of the international justice system in dealing with atrocities committed during the country’s civil war. However, as may be recalled, the 1999 Lomé Agreement did not provide for any retributive justice mechanism. Instead, as a result of the RUF’s pressure, the agreement made extensive provisions for amnesty. In this context, the establishment of a retributive justice mechanism, as symbolised by the SCSL, was only made possible after the UN Secretary General’s Special Representative (for Sierra Leone), Francis Okello, reminded the parties that the amnesty provisions contained in the 1999 Lomé Agreement could not apply to war crimes, crimes against humanity and crimes of genocide. This position was further reinforced after the RUF attacked and killed civilian populations in Freetown as they marched to protest against the group’s lack of commitment to peace. Still, as argued earlier, the eagerness of the international community and the Sierra Leonean government to explore judicial prosecution as one of the country’s transitional justice mechanisms was only made possible after significant international military intervention (mainly from Britain) had helped strengthen the Kabbah government’s position against the RUF and its allies.

It is important to observe that the SCSL spread the blame of civil war atrocities on all non-state parties by sentencing leaders of the RUF, the AFRC and the CDF. Although CDF leaders, allied to the government, were prosecuted, there is need to emphasise that government and national army leadership were absolved of any wrongdoing during the civil war. This was despite the government’s involvement in the Liberian civil war, albeit, in a retaliatory move against Charles Taylor’s meddling with the Sierra Leonean civil war.

Perhaps, the most blatant case of politicisation of the international judicial system, as far as Sierra Leone is concerned, relates to Charles Taylor. He had agreed to vacate the Liberian presidency in 2003 after he was guaranteed by ECOWAS leaders that he would not be prosecuted for his involvement in the Sierra Leonean civil war (Souaré 2008:212–213). This guarantee was later abandoned and Taylor was prosecuted and eventually sentenced to a 50-year jail term, raising the concern that threat of judicial prosecution (or withdrawal thereof) had become a tool to constrain political actors to specific actions.

The experiences of Burundi, Mozambique and Sierra Leone analysed herein reveal that post-civil war countries where conflict ended through a power-sharing-based agreement tend to avoid recourse to the international judicial system. This can be seen as forming part of former warring parties’ strategy not to expose themselves to the risk of prosecution in judicial institutions over which they have no meaningful control. Sierra Leone was able to steer away from this dominant trend after involvement from international role players contributed to weakening significantly the position of Foday Sankoh’s RUF and its allies. Yet, as shown in this article, although commendable, the overall success of the SCSL was limited.

Nonetheless, commendable, yet limited efforts towards institutional reforms

Besides providing for amnesty and the establishment of TRCs, the Arusha and Lomé Agreements for Burundi and Sierra Leone called for institutional reforms, including the establishment of national human rights commissions and, in the case of Burundi, the office of ombudsman. In Burundi, the Office of Ombudsman was established in January 2010, a year before the establishment of the National Independent Human Rights Commission (CNIDH), while Sierra Leone set up its Human Rights Commission as early as August 2004.

In Mozambique, the end of the country’s civil war brought an end to the single party system and enabled the adoption of multiparty democracy. In Burundi, the process involved in resolving the country’s conflict also incidentally led to the overhaul of the country’s political system through the establishment of an ethnic-based democratic regime that remains in place to this day. The Arusha Agreement is also credited with the establishment in May 2006 of the National Commission on Land and Other Assets (CNTB) tasked with, amongst other things, resolving land disputes as they relate to those displaced by the country’s cyclic episodes of political violence (both Internally Displaced Persons and returning refugees).8 Lastly, the processes involved in ending civil wars have also enabled the establishment of autonomous election management bodies in all countries analysed in this article.


Countries emerging from civil wars face daunting challenges, most of which they can hardly overcome even with meaningful external assistance. Transitional justice constitutes one among such challenges. As this article has shown, Burundi, Mozambique and Sierra Leone have all had to deal with demands for transitional justice in the aftermath of their respective civil wars. To a large extent, these demands were the result of the highly violent nature of these civil wars that affected large numbers of the
civilian populations.

The article has found that, by affording warring parties a prominent role in the post-settlement environment, the power-sharing mechanisms set up in Burundi, Mozambique and Sierra Leone inadvertently impeded the pursuit of both restorative and retributive justice in all three countries.

As an instance of warriors’ justice, power-sharing was used by such actors as an instrument to avoid facing retributive justice. Indeed, due to the central position they held within the power-sharing dispensations, former warriors emphasised amnesty while paying lip service to reparations. In all three countries, the decision to revert to the international judicial system or not was mainly motivated by political calculations rather than any genuine concern for justice. Lastly, notwithstanding the shortcomings above, the consensus generally brought about by power-sharing dispensations enabled all three countries to effect institutional reforms, albeit with limited and different levels of success.

This article analysed the role of power-sharing in the design and the operationalisation of transitional justice mechanisms in post-civil war Burundi, Mozambique and Sierra Leone. Future research ought to go beyond just power-sharing to look into the contribution of other variables such as the economic situation and external influences in shaping transitional justice in African countries emerging from protracted internal wars ended through negotiations.

Meanwhile, in learning from Burundi, Mozambique and Sierra Leone, the proposals below ought to be considered with regard to addressing the tensions between power-sharing and transitional justice in such post-war settings:

  • Transitional justice processes ought to include the larger public and not just be elite-focused.
  • There is need to acknowledge the virtual impossibility to prosecute all wartime crimes perpetrators in the short term. Provisions ought therefore to be made for war crimes, crimes against humanity and genocide to be prosecuted – even if a longer period than that of the initial transition is required.
  • The provision of amnesty should always be conditional on potential beneficiaries committing to open up to the TRC.
  • Care ought to be taken to ensure that the TRC’s recommendations on victims’ compensations are fully complied with by the government and all relevant stakeholders.
  • Lastly, adequate space, capacity and resources ought to be afforded to civil society entities – especially community-based organisations – so as to enable them to play a meaningful role in the transitional justice process.


  1. Almeida, Cláudia, Edalina Sanches and Filipa A. Raimundo 2010. ‘Bringing fighters together’: A comparative study of peacebuilding and transitional justice in Angola and Mozambique. Paper presented at a Conference on Legacies of Conflict, Decolonisation and the Cold War, Lisbon, May 2010.
  2. Armstrong, Andrea and Gloria Ntegeye 2006. The devil is in the details: The challenges of transitional justice in recent African peace agreements. African Human Rights Law Journal, 6 (1), pp. 1–25.
  3. Bhargava, Rajeev 2000. Restoring decency to barbaric societies. In: Rotberg, Robert I. and Dennis Thompson eds. Truth v. Justice: The morality of truth commissions. Princeton/Oxford, Princeton University Press. pp. 45–67.
  4. Binningsbø, Hlega M., Cyanne E. Loyle, Scott Gates and Jon Elster 2012. Armed conflict and post-conflict justice, 1946–2006: A dataset. Journal of Peace Research, 49 (5), pp. 731–740.
  5. Ekiyor, Thelma 2009. Reflecting on the Sierra Leone Truth and Reconciliation Commission: A peacebuilding perspective. In: Lekha-Sriram and Pillay 2009, pp. 153–170.
  6. Falch, Åshild 2008. Power-sharing to build peace? The Burundi experience with power-sharing agreements. Centre for the Study of Civil War (CSCW) Policy Brief 2. Oslo, Peace Research Institute Oslo (PRIO).
  7. Fombad, Charles M. 2008. Transitional justice in Africa: The experience with truth commissions. Available from: <> [Accessed 27 September 2013].
  8. Hayner, Priscilla 2007. Negotiating peace in Sierra Leone: Confronting the justice challenge. Centre for Humanitarian Dialogue / International Centre for Transitional Justice (CHD/ICTJ) Report.
  9. Hirsch, John L. 2009. Peace and justice: Mozambique and Sierra Leone compared. In: Lekha-Sriram and Pillay 2009, pp. 202–219.
  10. Igreja, Victor 2009. The politics of peace, justice and healing in post-war Mozambique: ‘Practices of rupture’ by Magamba spirits and healers in Gorongosa. In: Lekha-Sriram and Pillay 2009, pp. 277–300.
  11. International Centre for Transitional Justice 2011. Le processus de justice de transition au Burundi: défis et perspectives. ICTJ Briefing.
  12. Isbell, Thomas 2017. La commission nationale des terres du Burundi: résout-elle les conflits fonciers avec équité? Afrobarometer Dispatch Number 163.
  13. Jarstad, Anna K. 2008. Power sharing: Former enemies in joint government. In: Jarstad, Anna K. and Timothy D. Sisk eds. From war to democracy: Dilemmas of peacebuilding. Cambridge, Cambridge University Press. pp. 105–133.
  14. Järvinen, Taina 2004. Human rights and post-conflict transitional justice in East Timor. Finnish Institute for International Affairs (FIIA/UPI) Working paper 47.
  15. Lekha-Sriram, Chandra and Suren Pillay eds. 2009. Peace versus justice? The dilemma of transitional justice in Africa. Scottsville/Cape Town, University of KwaZulu-Natal Press/James Currey.
  16. Lemarchand, René 2006. Consociationalism and power sharing in Africa: Rwanda, Burundi, and the Democratic Republic of the Congo. African Affairs, 106 (422), pp. 1–20.
  17. Lijphart, Arend 1977. Democracy in plural societies: A comparative exploration. Yale, Yale University Press.
  18. Lijphart, Arend 1999. Patterns of democracy: Government forms and performance in thirty-six countries. Yale, Yale University Press.
  19. Mehler, Andreas 2009. Peace and power sharing in Africa: A not so obvious relationship. African Affairs, 108 (432), pp. 453–473.
  20. Olsen, Tricia D., Leigh A. Payne and Andrew G. Reiter 2010. Transitional justice in the world, 1907–2007: Insights from a new dataset. Journal of Peace Research, 47 (6), pp. 803–809.
  21. Papagianni, Katia 2008. Power sharing, transitional governments and the role of mediation. Background Paper for the OSLO Forum.
  22. Reydams, Luc 2013. Let’s be friends: The United States, post-genocide Rwanda, and victor’s justice in Arusha. Institute of Development Policy (IOB) Discussion Paper 1, University of Antwerp.
  23. Sadiki, Koko 2013. The tensions between power sharing, justice and human rights in Africa’s ‘post-violence’ societies: Rwanda, Kenya and the Democratic Republic of the Congo. African Human Rights Law Journal, 13 (2), pp. 254–280.
  24. Snyder, Peter 2014. Bangladesh Special Tribunal convicts opposition leader of war crimes. Available from: <> [Accessed 9 April 2019].
  25. Souaré, Issaka 2008. Le dilemme de la justice transitionnelle et la réconciliation dans les sociétés postguerre civile: les cas du Libéria, de la Sierra Leone et de l’Ouganda. Études Internationales, 39 (2), pp. 205–228.
  26. Tejan-Cole, Abdul 2009. Sierra Leone’s ‘not so’ Special Court. In: Lekha-Sriram and Pillay 2009, pp. 223–247.


  1. Although other African countries would meet the set criteria, space did not allow for all to be included. In fact, the initial research for this article involved six countries, namely Angola, Burundi, the Democratic Republic of the Congo, Liberia, Mozambique and Sierra Leone. However, the need to comply with length limits made it necessary to split the initial article into two, each sharing the same theoretical framework. This first one focuses on Burundi, Mozambique and Sierra Leone while the second will be dedicated to Angola, the DRC and Liberia. Also, although a potentially very relevant case for this study, South Africa is not included given the fact that it has already received adequate attention in
    existing literature.
  2. This debate has taken place in Bangladesh, for instance, where a number of high profile politicians have, as recently as 2014, been charged and convicted for crimes allegedly committed during the 1971 war with Pakistan. These include Zahid Hossain Khokon of the Bangladesh Nationalist Party, Motiur Rahman Nizami of the Jamaat-el-Islami and Ghulam Azam of the Bangladeshi Islamist Party. They were all found guilty of war crimes and sentenced to penalties ranging from death to life imprisonment (Snyder 2014).
  3. Front for the Liberation of Mozambique, the country’s ruling party since independence in 1975. It was established in 1962 and successfully led the armed struggle (1966–1975) for independence against Portugal.
  4. The Union for National Progress (UPRONA) was established in 1958 by Burundi’s first prime minister, Prince Louis Rwagasore. It was the country’s single party from 1966 to 1976 and between 1987 and 1992.
  5. See General Peace Agreement for Mozambique, 1992 – Protocol V: Guarantees, Specific guarantees for the period from the cease-fire to the holding of the elections, III 9.(d).
    It ought to be mentioned that the long-term consequences of the peculiar territorial power-sharing model applied in Mozambique has been the projection of wartime confrontational logic onto Mozambican politics: there were limited levels of disarmament, demobilisation and reintegration of former RENAMO combatants, and an entrenchment of wartime loyalty among former RENAMO combatants toward the group’s leadership while serving under the new national defence and security forces. As a result, armed confrontation resumed in 2015 between loyalist troops and RENAMO combatants in the latter’s stronghold in central Gorongosa. Dissatisfaction with the political system (including the outcomes of the 2014 general elections), the need for inclusion of remaining former rebel fighters in the regular army, and internal political dynamics within RENAMO are said to be behind this resumption of violence.
  6. Beside the provisions of the Arusha Agreement, the case for the TRC was further emphasised during the national consultations on transitional justice organised in 2009. The vast majority of the 3 887 respondents indicated their preference for the country’s transitional justice programme to be anchored on truth, justice and reparations (International Centre for Transitional Justice 2011:16).
  7. It ought to be noted that the Burundian parliament adopted the law establishing the TRC on 17 April 2014 amid a boycott of the parliament session by UPRONA’s members of parliament, despite their party’s alliance with the ruling CNDD-FDD.
  8. The Commission succeeded in providing the citizenry with an alternative framework to the conventional court system to resolve land-related conflicts. For a discussion on the working of the CNTB, see for instance Isbell (2017).