The devil is in the details: The challenges of transitional justice in recent African peace agreements

 Andrea Armstrong
 MPA (Public Affairs) (Princeton)
 Student, Yale Law School, USA
This email address is being protected from spambots. You need JavaScript enabled to view it.

The International Center for Transitional Justice (ICTJ) supported the research and writing of this paper while the author was a Research Associate at the ICTJ. Andrea Armstrong thanks Darren Geist for his able research assistance. An earlier draft of this paper also benefited from comments by former colleagues at the ICTJ, Federico Borello, Priscilla Hayner, Ian Martin, Alexander Mayer-Rieckh, Paul Seils and Marieke Wierda.

 Gloria Ntegeye
 MA (International Affairs) (Columbia)
 Political Affairs Officer, United Nations Department of Peacekeeping Operations Headquarters, New York, USA
This email address is being protected from spambots. You need JavaScript enabled to view it.

Gloria Ntegeye thanks Olanrewaju Shasore for her expert analysis of the political situation in Liberia, as well as colleagues in the UN Departments of Peacekeeping Operations and Political Affairs for their perspective and guidance. The views expressed here reflect the personal opinions of the authors and not those of their respective institutions.

 Edition: AHRLJ Volume 6 No 1 2006
  Pages: 1 - 25
 Citation: (2006) 1 AHRLJ 1-25
 Download article in PDF

Summary

Over the last seven years, warring parties in Burundi, the Democratic Republic of Congo, Liberia and Sierra Leone have signed peace agreements that include detailed provisions aimed at securing transitional justice. The novelty is not the growing use of transitional justice mechanisms in the aftermath of violent conflict, but rather that these mechanisms are being increasingly designed within the peace negotiation process. An examination of these four agreements illustrates a curious phenomenon: Alleged human rights violators are involved in the articulation of transitional justice mechanisms at the initial stages, without victim representation, transparency and dialogue. This article examines three underlying justifications for including transitional justice in peace agreements and finds that all three fail to adequately justify the inclusion of transitional justice blueprints in the initial stage of the peace process. First, including blueprints for transitional justice within peace agreements may actually weaken, rather than strengthen, the likelihood of a holistic and integrated transitional justice strategy by allowing alleged perpetrators to dictate the terms of justice. Second, including these details within peace agreements is not necessary to further conflict resolution efforts. Third, including detailed designs through undermining justice may also undermine state building. The inclusion of detailed transitional justice processes in peace agreements is not necessarily a clear victory for victims and human rights activists. At best, peace agreements may provide a foundation on which future transitional justice strategies can build.

1.  Introduction

According to traditional wisdom, warring parties will not agree to peace if such peace also includes measures to hold them accountable for their actions during the conflict. The peace agreements for El Salvador (1992) and Guatemala (1996), which both included the establishment of a truth commission, stand in stark contrast to at least eight peace agreements signed between 1989 and 1999 that did not make reference to transitional justice mechanisms, much less include operational transitional justice design elements. [1] Yet, since 1999, peace agreements that plan transitional justice mechanisms have been signed by parties to conflicts in Burundi, the Democratic Republic of Congo, Liberia and Sierra Leone. [2] It is becoming increasingly common for peace agreements to include processes that help societies account for past abuse. [3] It cannot, however, be assumed that the inclusion of transitional justice operational blueprints in peace agreements is a positive development for victims of abuse, often at the hands of signatories to the agreements. This raises an important question: What reasons underlie the inclusion of such detailed transitional justice processes in peace agreements? [4]

This article begins with an overview of transitional justice, followed by an assessment of the key transitional justice mechanisms included in each of the four peace agreements. These cases are then used to inform a broader analysis of three primary justifications for designing transitional justice within peace agreements: justice and reconciliation, conflict resolution and state building. Note that the question presented here is not whether transitional justice writ large should be included. Indeed, there are compelling moral, legal, and strategic arguments supporting the inclusion of transitional justice. Rather the question explored in this essay is whether the inclusion of detailed operational transitional justice mechanisms can be justified.

2.  An overview of transitional justice

Transitional justice strategies are usually pursued where national institutions lack the capacity and/or legitimacy to provide justice and redress to victims, particularly in the context of mass abuse. It consists 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 (ie backward-looking mechanisms that support forward-looking processes). While certain elements of transitional justice, such as prosecutions, have a long history, it is only in the last decade that individual transitional justice strategies have coalesced into a field of study and action. At present, most practitioners agree that transitional justice includes four basic areas: prosecutions, truth-telling, institutional reform and reparations. Each aspect of transitional justice addresses a particular need on the part of victims, and indeed for the larger society. For example, prosecutions, among other things, respond to society's need for establishing a judicial record of events and providing the means for punishment when found guilty. Truth-telling can help victims express their experiences and provide a forum for both more general public listening and understanding.

These four aspects of transitional justice are also complementary. The effective provision of justice in one area, such as prosecutions, can buttress efforts in the other three, contributing to greater victim satisfaction. [5] For rural survivors of abuse, judicial trials of military leaders in the country's capital may seem remote from their immediate experience at the hands of their neighbours. In this context, truth-telling measures can assist victims in integrating and understanding their experience within the larger conflict. [6] When implemented in isolation, on the other hand, transitional justice measures may not seem credible to victims. Institutional reform, without a public discussion on the role of certain institutions in the conflict, can be seen as an empty gesture unlikely to address victims' needs. Moreover, without an adequate sense of how institutions may have facilitated abuse in the past, success in reform may be limited. Hence, experts in transitional justice have argued that holistic integrated transitional justice strategies are more likely to succeed in reconciling divided communities to a common future through addressing the abuses of the past. [7] This, in turn, has implications for how transitional justice is pursued.

Certainly, transitional justice can be implemented in a top-down fashion. But it is unlikely to result in the societal internalisation that is regarded as central to eventual reconciliation. [8] While the catalogue of abuses may be the same from conflict to conflict, societies differ not only in the scope and depth of abuse, but also in how those experiences are addressed. Ideally, transitional justice acknowledges the different experiences of victims, perpetrators and observers, as well as the relationships among them, in an effort to promote justice and reconciliation. The goal of transitional justice is not to create one version of the truth, but rather to create the space for victims to receive justice and redress as part of a larger public commitment to reconciliation. For these reasons, current practice eschews a cookie-cutter approach to transitional justice, preferring an inclusive participatory process to the design of transitional justice mechanisms. [9]

3.  Case studies in peace agreements

Peace agreements generally include commitments by the parties to respect the Universal Declaration of Human Rights (Universal Declaration) (or reference key human rights concepts, such as equality and individual rights). Forward-looking mechanisms, such as a human rights ombudsman or department within the Ministry of Justice responsible for human rights, are often established to serve an important function in delineating standards that may contribute towards the effective protection of human rights in the future. They do not, however, directly address the needs of society as a result of abuse perpetrated during the conflict, or establish accountability for past atrocities.

The primary focus of this discussion is to detail the obligations undertaken by signatories at the negotiating table to shed light on the degree to which designing transitional justice mechanisms within peace agreements is justified. These peace agreements are not reached in a vacuum, however. The political interests and considerations of national, regional and international actors are relevant at every step, including at the negotiation table.

The agreements below include both forward and backward-looking transitional justice mechanisms. Examined, in chronological order, they are:

•      Sierra Leone: Lomé Peace Agreement, signed 7 July 1999;

•      Burundi: Arusha Peace and Reconciliation Agreement, signed 28 August 2000;

•      Democratic Republic of Congo (DRC): Global and Inclusive Agreement on Transition, signed 17 December 2002; and

•      Liberia: Comprehensive Peace Agreement, signed 18 August 2003.

3.1  Reaching the negotiating table

Peace agreements become necessary when, as in these cases, the governments fail to meet the primary responsibility of states: the establishment and preservation of public security. Negotiations between belligerents are critical aspects of restoring public security, ideally through the successful implementation of a carefully negotiated peace agreement. The success of these agreements to effectively restore peace varies significantly from case to case. Each of the peace agreements examined here was preceded by numerous failed agreements. However, at the time of writing, the peace processes in Burundi, Liberia and Sierra Leone showed significant progress.

To use the example of Sierra Leone, the peace negotiations that began in Lomé in May 1999 were held in the absence of a viable alternative for the government. President Ahmed Tejan Kabbah was experiencing serious setbacks in containing the rebels, particularly Foday Sankoh's Revolutionary United Front (RUF). The new civil government in Nigeria, concerned about the costs in terms of life and money of peacekeeping in a country where the peace would not hold, was inclined to shift resources towards internal needs. [10] This prospect would remove Sankoh's military options for control of the territory of the state, and would have resulted in the de facto partitioning of Sierra Leone, with the rebels holding a large percentage of the resource-rich areas. [11] On his side, Sankoh needed to resolve the issue of his conviction for treason, and subsequent death sentence.

A number of actors, both internal and external, were critical in bringing the government and rebels to the negotiating table. The active participation of Sierra Leonean civil society, particularly the Inter-Religious Council of Sierra Leone, which gained the confidence of government and rebels alike in earlier peace negotiations, was particularly important. The Council was instrumental in bringing President Kabbah and Sankoh to meet in advance of the negotiations. Recognising the importance of having the potential sub-regional spoilers on board, the Council persuaded Liberian President Charles Taylor to attend the negotiations. [12]

The negotiations, which began in May with the signing of a ceasefire, concluded in July with the signing of a peace agreement between the government of Sierra Leone and the RUF. During the two months that negotiations were underway in Lomé, negotiators focused on a primary objective: bringing to an end the violent conflict ravaging the country. Lomé was the result of political expediency, rather than justice. [13] Numerous realities had to be factored in, particularly the position of strength the RUF was negotiating from, vis-à-vis President Kabbah's government. While much of Sierra Leonean society was against power-sharing with the RUF, the pressure to end the war at all costs, and the very real threat of the RUF re-igniting the war, compelled a peace agreement with numerous concessions to the demands of Sankoh and the forces he controlled. [14]

3.2  Prosecutions

Amnesties are very often one of the most controversial elements in a peace agreement. All of the peace agreements examined refer to amnesties for combatants and leaders. Continuing the example of Sierra Leone, the Lomé Peace Accord extended the widest possible immunity from prosecution by 'grant[ing] absolute and free pardon and reprieve to all combatants and collaborators in respect of anything done by them in pursuit of their objectives, up to the time of the signing of the present Agreement'. [15] Indeed, the pardon was enormously controversial, resulting in outrage on the part of Sierra Leonean citizens and human rights activists. [16] The Special Representative of the Secretary-General entered formal reservations to the agreement, on the grounds that the United Nations (UN) does not recognise amnesty for crimes against humanity, genocide or war crimes. [17] The inclusion of amnesty for those perpetrating the war in Sierra Leone, and specifically for Sankoh, was a direct concession to the rebels in Lomé who, as during previous peace agreements, demanded a blanket amnesty. At the time of the negotiations in Lomé, they were in a strategic position to receive what they wanted — not only would Sankoh's conviction for treason not end in execution; he would be awarded the Vice-Presidency and control over important state resources.

In Burundi, the peace agreement provided for a partial amnesty, prohibiting amnesty for acts of genocide, war crimes, crimes against humanity and coups d'état. [18] These prohibitions reflect the issues and sensitivities relevant to Burundi. They can also be understood in the context of the sub-region Burundi inhabits, as well as the involvement of strong regional mediators determined to avoid any possibility of catastrophe on the scale of Rwanda. The 1993 assassination of the first democratically elected, and first Hutu, president, Melchior Ndayaye, and the subsequent coup attempt staged by the Tutsi-dominated army, were watershed moments in the history of Burundi. When a second coup staged in 1996 to re-establish a political order led by an army-installed President Buyoya resulted in the region imposing an embargo on Burundi, the government agreed to move toward negotiations in 1998. [19] However, the role of regional actors in compelling a political solution to the conflict in Burundi cannot be underestimated. Nelson Mandela, who served as mediator for Burundi following the death of Julius Nyerere in 1999, insisted that the rebel movements be included in the negotiations, though the political elite wanted to exclude them from the talks.

Part of the agreement reached in Arusha was that the transitional government of Burundi should request the UN Security Council to establish an international judicial commission of inquiry to investigate acts of genocide, war crimes and other crimes against humanity from independence in 1962 to the signing of the agreement in 2000. [20] At the time of signing, however, some predominately Tutsi parties expressed reservations on the provisions contained in the Accord, and stated that they did not 'subscribe' to it. [21] Nevertheless, the agreement signed by the 19 parties in Arusha required the government of Burundi to formally request the establishment of an international tribunal should the Commission's report 'point to the existence of acts of genocide, war crimes and other crimes against humanity'. [22] While addressing the call for the exploration of the nature of the conflict in Burundi, as well as acknowledging the history of exclusion that is central to the post-colonial history of Burundi, the agreement displaced accountability for initiating and implementing a process of investigation to the international community — perhaps necessary in a historically divided society, where reconciliation between the ethnic groups would have to be a very 'delicate process' requiring impartiality. [23]

The most vaguely defined amnesty clause of these peace agreements is the Liberian Comprehensive Peace Agreement, which states that the transitional administration 'shall give consideration to a recommendation for general amnesty to all persons and parties engaged or involved in military activities during the Liberian civil conflict'. [24] There is no specification on whose recommendation or which criminal acts would be excluded from receiving amnesty. While presenting a risk that crimes constituting serious violations of international law may be granted amnesty, the vague language provides space for Liberians to define and then advocate for prohibitions against amnesties based on the particular needs and expectations of Liberian society. The objections of the UN and local civil society have reportedly postponed the adoption of an amnesty in Liberia.

Although prosecution is not mentioned in the text of the Global and All-Inclusive Agreement for the DRC, Resolution DIC/CPR/05, adopted at the Inter-Congolese Dialogue in Sun City, tasks the transitional government with requesting the UN Security Council to establish an 'International Criminal Court' for the DRC. [25] The proposed court would investigate serious crimes committed since 30 June 1960, the date of Congolese independence from Belgium. There are a variety of reasons why negotiators may have deemed it fit to include both an amnesty and a prosecutorial process. Negotiators representing armed groups may support prosecutions to eliminate rival leaders within their own group. Other parties, aware of the cost of such tribunals to the international community, might have risked endorsing calls for a tribunal that they knew was unlikely to be established. By asking for an international tribunal, parties to the agreement can improve their domestic and international reputations and consolidate their shaky bona fides as representatives of the people. For a variety of reasons, then, leaders of civil society, armed groups and President Joseph Kabila regularly and publicly called for the establishment of such a tribunal. [26]

All the agreements that refer to prosecutions designate an explicit and activating role for the international community upon which the entire process hinges. If the international community fails to respond or ignores the request, then the prosecutorial process envisioned in these agreements never starts.

At the same time, amnesties have not been dependent on the international community. The Sierra Leone government granted absolute and free pardon to Foday Sankoh on day one of the agreement. DRC President Joseph Kabila decreed amnesty for 'acts of war, political crimes and crimes of opinion' during the conflict. [27] In Burundi, the parliament passed a law granting 'temporary immunity to political leaders returning from exile who committed crimes from 1962 to 27 August 2003' - the date the law was passed. [28] In addition, as one of the conditions of the CNDD/FDD joining the Burundi peace process after the Arusha Agreement was signed, all belligerents (government and rebels) are to receive temporary immunity for both leaders and soldiers. [29]

3.3  Truth-telling

Truth-telling [30] mechanisms can include the establishment of investigatory commissions as part of the prosecutorial process, commissions mandated to collect the truth to promote national reconciliation, or projects that 'map' patterns of violations through the collection of written evidence. Peace agreements for Liberia and Sierra Leone include relatively few provisions regarding the actual powers of the proposed truth and reconciliation commissions, giving space for these to be elaborated by civil society and the citizenry at large. The Comprehensive Peace Agreement tasks the Liberian Commission with three specific activities: to provide a forum for addressing impunity and allowing victims and perpetrators to share experiences; to address the root causes of the Liberian crises; and to make recommendations on the rehabilitation of victims. [31] It requires that commissioners be drawn from 'a cross-section of Liberian society' — a relatively common phrase in these peace agreements. The truth commission is also not referenced in the otherwise extensive timetable annexed to the agreement. Liberia brings to the fore the question of timing — while some may interpret the Commission's failure to begin its work during the mandate of the transitional government, it could also be argued that the delay provided the space for an inclusive process. When negotiators worked to reach an agreement among warring factions in Accra, one consideration would have been the fact that many of those victimised by war in Liberia were displaced, and therefore would return home before being able to participate in the Commission. 

Similarly, the Lomé Peace Agreement for Sierra Leone provides a broad outline of the Commission's jurisdiction, namely crimes committed since 1991 and duties, such as recommending measures to rehabilitate victims, but refrains from specifying operational guidelines. The agreement specifies that commissioners should 'be drawn from a cross-section of Sierra Leonean society with the participation and some technical support of the international community'. [32] Its most restrictive provision requires that the Commission be established within 90 days of signing the agreement in July 1999. [33] However, the situation in the country, as well as operational constraints, precluded the Commission from meeting these deadlines.

In contrast to Sierra Leone and Burundi, the jurisdiction, activities and operations of the Truth and Reconciliation Commission (TRC) in Burundi are fully specified in the peace agreement. [34] The Truth Commission negotiated by parties to the Burundi Arusha Accord appears to be closely modelled on the South African Truth and Reconciliation Commission of 1994. The proposed truth commission will have the power to grant amnesty, which, excepting the South African truth commission, has not been granted to the approximately 30 such truth commissions to date. The Commission is mandated to 'establish the truth' about serious acts of violence since independence in 1962, including the identification of victims and perpetrators. In the name of reconciliation, it will have the ability to make recommendations on reparation measures and on more general social and political measures to foster healing. The Arusha Agreement provides for a two-year mandate (with the possibility of a one-year extension) and outlines the selection criteria and process for commissioners. The agreement also includes a rigid implementation timetable that includes the establishment of the truth commission as well as other transitional measures. [35] Although these timetables are routinely violated, if this timetable had been adhered to, the general public's involvement in, and engagement with, the truth commission process would have been seriously curtailed for lack of time.

Similarly, the Inter-Congolese Dialogue Resolution on the institution of a 'truth and reconciliation commission' provides little space for discretionary judgment or creative design. In marked contrast to the Sierra Leonean and Burundian agreements (which had both been signed prior to the DRC negotiations), commissioners are drawn from the ranks of the negotiating parties (ie each group is represented), which clearly presents problems of holding objective and impartial discussions about past impunity. [36] Second, the DRC Commission is mandated to 'identify the nature, causes and extent of the political crimes and large-scale violations of human rights . . . since independence', including those crimes committed outside of DRC territory, but nevertheless related to the conflicts in the DRC. [37] It has the power to award a limited amnesty [38] for full confession of crimes of a political nature, which presumably is applicable to both combatants and leaders. The Commission is also tasked with several responsibilities incommensurate with its abilities as an agency independent of government. 'Deciding the fate' of victims and 'taking all necessary measures to compensate them and completely restore their dignity' can only be accomplished by the government; the Commission has neither the reach nor resources necessary to achieve these objectives. [39] The DRC Truth and Reconciliation Commission is tasked with a series of largely unrealisable objectives, including the reconciliation of the main political actors (both among themselves and with the Congolese people); the emergence and consolidation of the rule of law; and the rebirth of a new national and patriotic consciousness, among other things. Indeed, many truth commissions are saddled with unrealistic expectations, In this respect, the DRC is no different.

The cases of Burundi and the DRC are prime examples of over-determining the shape and functioning of transitional justice mechanisms in peace agreements. Transitional justice mechanisms in general, and truth telling and reparations in particular, require the involvement of the general public to be successful. Through broad consultation, institutions and mechanisms can be shaped to directly respond to the most pressing needs as articulated by the people themselves. In addition, the process of involving the public in designing truth commissions, in and of itself, can contribute towards healing and solidarity. [40] At the same time, wide discretion in the design of transitional justice measures is not a guarantee of public involvement. Initial discussions on the Truth and Reconciliation Commission in Liberia, including the appointment of commissioners, started as an exclusive process that only opened following the protest of Liberian civil society as well as the regional and international actors that have invested time and resources into the Liberian peace process. The selection process was finally conducted under the oversight of the Economic Community of West African States (ECOWAS) and the UN, in consultation with civil society and political party representatives. The names of all nominees, including those appointed by the Chairperson, were published in the Liberian newspapers, with an invitation for the public to evaluate all names and present any reason why any nominee was not fit for the role of commissioner. In February 2006, newly elected President Ellen Johnson-Sirleaf inaugurated the Liberian Truth and Reconciliation Commission, following two years of discussion and planning. [41]

It may be considered that actors opposing discussion on past impunity can use the discretion provided in the agreement to delay implementation, or to manipulate the process. However, an expedited process is not necessarily the best case scenario — and, within reason, mechanisms with the objective of reconciliation should not be dependent on the timetable of external actors. For example, representatives of civil society in Sierra Leone used the delay between the Lomé agreement and the inauguration of the Truth and Reconciliation Commission in 2002 to discuss how the Commission could best assist victims and suggest priorities for the Commission's attention. So discretion and delay are not necessarily counter-productive for ensuring responsive transitional justice measures, but they do need to be accompanied by formal guarantees of broad public consultation and discussion.

3.4  Institutional reform

Institutional reform, within a transitional justice context, involves the altering of the government's structures, mandates and composition through the creation of new entities, the abolishment of specific agencies, and/or the retooling of existing departments and personnel with the aim of improving the government's protection of human rights. It can include the creation of new oversight mechanisms (such as commissions) or mandate existing institutions, such as the legislature, with greater oversight powers. All four peace agreements include provisions for the establishment of a governmental Human Rights Commission or Ombudsman. [42] This is not particularly new, since most peace agreements have little trouble adopting forward-looking mechanisms. The ways in which institutional reform is implemented may achieve different degrees of accountability for the past. The establishment of a new institution with a mandate to investigate current and future abuses entails little accountability for past crimes. However, some practitioners argue that reforming the composition of an institution, in terms of its ethnic, gender, religious or geographic representation, can play an important role in preventing future violence, while correcting for discrimination in the past. [43]

It is worth noting that the Arusha Accord, including the Annexes, details a comprehensive forward-looking vision of the structure of the new government and its responsibilities towards its citizens. For example, it calls for the reform of the judiciary to promote judicial independence and correct existing gender and ethnic imbalances. [44] Of the cases examined, Burundi goes the farthest in addressing structural inequalities, including gender inequality. It calls for adopting legislation on women's inheritance rights and makes numerous references to gender balance in government ministries and commissions. The agreement also mandates the creation of an institution to (1) identify the problems faced by women as a result of Burundi's crises and (2) propose solutions to the transitional government that would be necessary 'to promote and support the advancement of women', given the difficulties that women have had and still continue to face. [45]

As the conflict in Burundi centres on the allocation of political, economic and military power, historically linked to ethnic affiliation, the peace agreement reached in Burundi explicitly requires ethnic balance in every governing institution. For example, when the ethnic diversity of the community is not reflected in the make-up of the Commune Council, the Senate is mandated to 'order the co-option of persons . . . from an underrepresented ethnic group . . . provided that no more than one-fifth of the Council may consist of such co-opted persons'. [46] The Arusha Accord also requires that the first transitional President and Vice-President be of different ethnicities and represent different political parties. [47] This provision is in recognition of the difficult balance transitional justice will have to strike in Burundi — all sectors of public life require reform, however, this reform must not be seen as punitive.

When it comes to backward-looking institutional reforms, the agreements are far from comprehensive. An important element of the Liberian Comprehensive Peace Agreement is the complete dismantling of the armed forces of Liberia. Decommissioned former army personnel interested in joining the new army must be re-recruited following a vetting exercise whereby their personnel records are examined for indications of human rights abuse. [48] It was clear to all present in Accra during the peace negotiations that the Liberian army had been discredited completely within most segments of Liberian society.

The peace accord in Burundi does not provide for the systematic vetting of its military personnel, nor could it — suggestions of an overhauling reform of the historically Tutsi-dominated army, even in spite of the role it has played in perpetrating coups in that country, could turn the Tutsi elite against any new dispensation, and would likely be strongly resisted by Rwanda. Nevertheless, the Arusha Agreement does exclude from government service soldiers found guilty of genocide, war crimes, and crimes against humanity. [49]

3.5  Reparations

Reparations [50] are efforts to specifically address the variety of harms suffered by victims of human rights abuse. [51] To that end, reparation programmes may seek to rehabilitate victims through the provision of 'medical and psychological care, as well as legal and social services'; [52] compensate victims for physical or mental harm, lost opportunities, and material damages; [53] and provide restitution to victims through restoring their legal rights, employment and property, for example. [54] Reparations can also include symbolic measures, such as apologies, reburials, monuments, or days of remembrance in honour of victims, which can provide victims with redress that material reparations are incapable of providing. None of the four agreements contain an explicit reference to reparations writ large; instead they focus on repairing harm through rehabilitation.

The Liberian agreement briefly refers to rehabilitation in article XXXI, where it tasks the National Transitional Government of Liberia with 'design(ing) and implement(ing) a programme for the rehabilitation of war victims'. The TRC has subsequently given itself the task of providing recommendations to the Heads of State on (1) the reparations on rehabilitations of victims and perpetrators; (2) the need for continued investigations; (3) legal, institutional and other reforms; and (4) prosecution of certain cases. It is instructive that the focus of rehabilitation is not solely on those defined as victims, but also includes perpetrators of the war. This may indicate that the focus in Liberia is centred more on reconciliation than on persons who fought in the war, primarily youth combatants, who are now regular Liberians looking for an opportunity to rebuild a life in a war-ravaged country. Other actors who were once active or complicit in the civil war are now political actors, which can be, depending on their degree of liability, a constructive transformation.

In Sierra Leone, where conflict resulted in thousands of amputees, several mechanisms have been established to support the rehabilitation of war victims. The National Commission for Resettlement, Rehabilitation and Reconstruction (later re-named the National Committee for Social Action (NaCSA)) is tasked with the needs of women and their potential contribution to reconstruction. [55] Generally, while the Lomé Peace Agreement provides for a number of options to access rehabilitation services, reparation is often linked to general developmental goals unrelated to the abuse or victimisation experienced by individual citizens.

The Arusha Accord for Burundi takes an expansive view of reparations, including a series of recommendations designed to support healing and foster reconciliation. The agreement calls for the establishment of a National Commission for the Reintegration of War Victims to address the needs of the large internally displaced and refugee populations and creates a National Fund for War Victims, to support their reintegration into their communities. [56] Compensation is also addressed, though only as it applies to property that was plundered and cannot be restored to its original owner (ie restitution). [57] Arusha was also the only of the four agreements to include symbolic measures of reparations (although not referred to as such in the agreement). These measures include the construction of a monument, inscribed with the words 'never again'; the identification of mass graves and if desired, the dignified reburial of a loved one; and a day of remembrance for victims of the conflict. [58]

A common option is for proposed truth commissions to be tasked with making recommendations on reparation measures to rehabilitate war victims or promote reconciliation and forgiveness. [59] Often, however, the Commission's mandate outpaces its authority. However, these expressed authorities are limited, as the Commissions do not have budgetary authority to implement their findings or to compel governments to implement their recommendations. Making reparations contingent on the government's implementation of a TRC recommendation may mean no reparations at all. [60] It is questionable if a desired role for truth commissions would be the implementation of a reparations programme; arguably, their function is to provide a space for healing, rather than compensation.

4.  Justifications

In all of the cases examined, thousands (or in DRC, millions) of people have been killed, displaced or otherwise victimised in the execution of the violent conflict. Justice, in this context, ideally contributes to restoring individual dignity and promoting reconciliation. [61] The ends of transitional justice can be supported by conflict resolution and state building. Establishing a durable peace arguably ends situations of massive abuse, promoting individual dignity and security. Building a just and effective state provides essential guarantees of non-repetition that can promote reconciliation. But simply because the ends may be mutually supportive does not mean that the means to achieve those ends are equally supportive. Indeed, the means to achieve one end may undermine other equally legitimate ends. For example, limiting the peace negotiations to warring parties may promote conflict resolution, but not transitional justice. (Alternatively, broad participation may further the objectives of transitional justice, but not conflict resolution.)

4.1  Justice

One possible justification for including such detailed provisions in each of these agreements is simply to ensure justice for victims within the new state framework. Transitional justice becomes a price attached to the re-allocation of political authority among armed groups. In each of the cases discussed, the transitional government was composed of members of competing armed groups. The detailed inclusion of transitional justice may reflect a lack of faith on the part of mediators that armed groups will actually adhere to less detailed commitments. But it is not yet clear that including these blueprints results in their implementation. For every Sierra Leone, where the most prominent mechanism, the Truth and Reconciliation Commission, has been implemented, there is a Burundi, where progress in implementation has been both intermittent and slow.

Moreover, these blueprints are likely to be incomplete. Even if parties to the conflict undertook the design of transitional justice mechanisms with the genuine intention of promoting full accountability and redress, they lack the necessary information from victims to do so. All peace agreements in this study allowed for some degree of input from unarmed political parties and civil society. In some instances, as in Sierra Leone and Liberia, there was very active and important participation from local leaders and women's groups. In Burundi, 19 political parties (some linked to armed groups) participated in the Arusha negotiations. It could be argued that these groups had little leverage vis-à-vis armed groups in the negotiating process. But the broader point is whether these groups can realistically be expected to effectively represent specific victim demands for justice in exclusive peace negotiation forums when violence is still the norm. Decades of cyclical violence in the region have resulted in one of the largest regionally dispersed populations in recent history. Different armed groups may control separate sectors of a country, further complicating travel and open discussion. Second, the closed-door nature of the negotiations certainly prevented non-armed representatives from consulting with their membership as the mechanisms were developed. Any operational details resulting from peace agreement negotiations would still require the input of society at large.

Designing transitional justice mechanisms within peace agreements could also reflect the idea that some justice is better than none. By having armed groups explicitly agree to detailed though imperfect provisions, future conflict over the scope and depth of these mechanisms would be avoided. But this assumes that the interests and positions of the armed groups are static and fixed. Alliances both among and within armed groups can change. Particularly where the battles involve multiple armed groups, the new political reality can shift existing alliances among the various groups. Even within armed groups, leadership battles may lead to less support for amnesties, for example, if the ascendant sub-group believes it is less culpable. The deals made during peace negotiations then become rigid mechanisms, which neither armed groups nor the general public are able to amend or supplement once the agreement has been signed.

Under either argument, lack of faith or some justice is better than none, the actual content of these transitional justice mechanisms is still relevant. The impact of armed groups is most clearly seen in the amnesty provisions (accompanied by relatively weak provisions on prosecution) in each of the agreements. But other areas of transitional justice are similarly rendered less effective.

The composition of the Truth and Reconciliation Commission for the DRC, as stated in the Global Agreement and the Transitional Constitution, requires that each of the negotiating parties be represented in its offices and that a representative of civil society serves as President. [62] In concrete terms, this means that the commissioners are representatives of the warring parties, that is, the Kabila-led government, the RCD-Goma, the MLC, the RCD Mouvement de liberation, the RCD-N, the Mai Mai militias, in addition to a representative each from civil society and the unarmed opposition. [63] The government and many of the armed groups have each been accused of serious violations of human rights, including systematic rape and targeted ethnic cleansing [64] — so it would be fair to question the ability of such representatives to impartially and objectively 'identify the nature, causes and extent of the political crimes and large-scale violations of human rights committed in the DRC, since the country's accession to independence'. [65] Through the staffing of mechanisms, the design of the mandate, the powers and independence granted, armed groups can wield considerable influence on the shape and weight of transitional justice strategies.

In addition to overt obstruction, the peace agreements also omit measures that might be desirable. Of the four agreements, only Burundi provides for symbolic reparations, such as monuments. Broader participation may have resulted in stronger institutional reform measures, such as the vetting of military and police forces.

Allowing armed parties to negotiate the form and content of transitional justice strategies devoid of transparency or public comment creates serious difficulties in achieving justice, which contributes to reconciliation and healing in post-war societies. There is a difference between soliciting 'buy-in' (or at least guarantees of non-interference from militarised groups) and giving these same groups almost sole control over the design and implementation of a country's transitional justice strategy.

At best, the detailed design of transitional justice mechanisms within the context of peace agreements is imperfect in achieving justice for victims. At worst, armed groups party to the agreement can pay lip service to the demands of victims, while creating transitional justice mechanisms devoid of substance or applicability. It is far from clear that including blueprints for justice within peace agreements enhances, rather than detracts from, the desired outcome of dignity and reconciliation. 

4.2  Conflict resolution

Given the less than hopeful impact on achieving justice, perhaps including operational details of transitional justice, enhances the aims of a complementary goal: conflict resolution. Indeed, one could argue that when negotiated as part of a peace agreement, transitional justice positively impacts peace negotiations by expanding the issues in controversy. In Burundi, the negotiating groups did agree on the need for an international commission of inquiry into genocide and crimes against humanity, but disagreed on whether the inquiry should cover events since 1965 or only events beginning in 1993. [66] This is not unique to Burundi; the same issue has plagued discussions in the DRC. The Rassemblement Congolais pour la Democratie (RCD) proposed that the jurisdiction of the international tribunal included in the peace accord include events since independence, 30 June 1960, but the Mouvement de Liberation du Congo (MLC) and the RCD-National (RCD-N), also signatories to the DRC accord, preferred 1996 (the start of the first Congolese war that ended in the removal of Mobutu and the installation of Laurent Désiré Kabila as President). [67] In such contexts, transitional justice is manipulated to justify the abuse that occurred — by legitimising the reasons one party went to war or by marginalising the abuse within a larger political context.

The expansion of issues through including transitional justice may or may not support the goal of ending the conflict. On the one hand, expanding the number of issues enlarges the 'negotiating pie' and provides negotiators with more opportunities to co-operate through trading or exchanging concessions. In addition, the fact that these issues are being negotiated represents a first step to settling conflicts through discussion and not violence. While this does not necessarily bode well for achieving justice (since most armed groups could agree that impunity is better than justice) it may enhance the end of violent conflict in the short term. [68]

In such scenarios, however, transitional justice is reduced to a means of conflict resolution, rather than an independent end. It is also unclear to what extent detailed transitional justice mechanisms are necessary to enlarge the options available to negotiators. It is certainly worth asking whether other incentives could perform the same function without sacrificing the goals of transitional justice.

Another argument for including details for future transitional justice mechanisms could be its potential as a deterrent to violating the peace agreements. Armed groups, it could be argued, would be on notice that future violations could not only result in prosecution, but could undermine their political legitimacy during the transition. This is a compelling argument in theory, but violence in the Democratic Republic of Congo and Burundi after the signing of the agreements casts empirical doubt. [69] Moreover, it does not address why the details are necessary for deterrence. Does a commitment to certain transitional justice mechanisms (without the operational details) provide less deterrence than one that incorporates the negotiating positions of armed groups? Indeed, the more concessions made during the negotiations, the less of an impact for any future mechanism reducing its deterrent value.

4.3  State building

Could state building, then, justify the inclusion of detailed transitional justice benefits despite the imperfect justice result? Transitional justice also includes an explicit state-building component: institutional reform. Liberia is engaged in a comprehensive effort to vet its new security forces, as mandated by its peace agreement. Burundi explicitly provides for measures to enhance the role of women within the society and the government. Both of these measures contribute to the construction of a just state. Both of these measures address the legacy of past abuse.

Transitional justice and state building are similar in other ways. Neither process is particularly fast in producing results. [70] Neither process can be imposed externally and still be effective. [71] In addition, the two processes are mutually supportive. Restoring individual dignity and facilitating reconciliation can play an important role in ensuring the sustainability of newly created institutions and creating a shared vision for the state. Conversely, creating effective, responsive, and effective institutions is an important signal to victims that they are respected and included within the new state. Moreover, transitional justice can help lay the foundation for a 'social contract', which is an essential component of state building. [72]

Yet, all of these similarities do not address whether including the operational details of such mechanisms in peace agreements enhances the state-building project. It is clear that transitional justice can contribute to state building (whether it should is another matter beyond the scope of this paper). But if the impact of transitional justice is lessened through including operational details in peace agreements, does that not detract from its potential contribution to state building?

5.  Conclusion [73]

Negotiations focused on the primary purpose of bringing violent conflict to an end are not conducive to designing an inclusive process that takes into consideration the full range of needs society has for reconciliation and justice. While it makes sense to limit discussions at the negotiating table to mediators and warring parties, peace negotiations, exclusive by design and nature, are not necessarily the appropriate forum to design transitional justice mechanisms. [74] While accountability and reconciliation should clearly be a part of the peace process, viability and inclusiveness dictate that planning these mechanisms be outlined through a separate process. But this analysis does not imply that there should be a strict separation between transitional justice writ large and peace agreements.

Including a broad foundation for transitional justice within peace agreements can lead to increased support for the eventual mechanisms. First, establishing the principle of justice for past crimes in peace agreements also creates additional opportunities to provide justice and redress for victims outside of the obligations contained in the agreement. The negotiation of peace agreements is but one element of a larger peace process. The Lomé Accord, through establishing a Truth and Reconciliation Commission mandated to 'address impunity, break the cycle of violence, provide a forum for both the victims and perpetrators of human rights violations to tell their story, get a clear picture of the past in order to facilitate genuine healing and reconciliation', [75] committed the negotiating parties to accounting for the past. These public commitments do not preclude the establishment of additional transitional justice measures not included in the original agreement, [76] such as the Special Court for Sierra Leone, which found that this amnesty does not apply to crimes under international law, including acts of genocide, crimes against humanity and war crimes. [77] Although progress has been slow, the Liberian peace accord — of the four surveyed here — best achieves the balance between broad support for transitional justice mechanisms and detailed design.

Second, including transitional justice can formally involve international actors in the transitional justice process. [78] In some cases, but not all, international involvement can provide a degree of objectivity in deeply divided societies, particularly in the area of prosecutions. Formally involving international actors in the design and implementation of transitional justice mechanisms may provide a perception of impartiality. It is also more likely to result in assistance and funding. [79]

Justice for victims of massive abuse is increasingly a non-negotiable issue. [80] Peace agreements can make an important contribution in providing a foundation for a deliberative and inclusive design process, but mediators should be cautious in simultaneously designing transitional justice mechanisms and negotiating peace. First, peace agreements should include processes that facilitate broader discussion of the past and ways in which to address it. The focus should be on guaranteeing as much space as possible for public participation and input into transitional justice strategies and creating institutions flexible enough to respond to the public's demands. For example, the peace agreement could create an independent human rights commission, ideally one that bars adherents of armed groups from membership, mandated to initiate such discussions. Second, peace agreements can enumerate the types of abuses perpetrated (allowing for the addition of more) and commit negotiators to seeking justice for victims. By providing a platform for a more inclusive design process, less determinative peace agreements may contribute more towards justice in the long term.