Peace agreements and the termination of civil wars

Lessons from Liberia

Dr George Klay Kieh jr is Professor of Political Science and Former Dean of the College of Arts and Sciences at the University of West Georgia, Carrollton, GA.


On Christmas Eve in 1989, the National Patriotic Front of Liberia (NPFL), a rebel group led by Charles Taylor, a former official of the Doe regime, launched a military attack against the north-central region of Liberia from a base in neighbouring Côte d’Ivoire. Understandably, the Doe regime launched a counter-offensive. The resultant ‘tugs and pulls’ degenerated into Liberia’s first civil war (1989–1996).

Concerned about the adverse impact of the war on sub-regional security, the Economic Community of West African States (ECOWAS), the sub-regional organisation, intervened in the conflict using, among others, a peacemaking approach that revolved around various peace accords. However, the initial sixteen peace accords failed to end the war. Finally, the Abuja II Peace Accord ended the blood-bath. What factors accounted for these outcomes? The initial sixteen peace agreements were deficient in terms of one or more of the elements that that have to be dealt with for the successful implementation of a peace accord – ranging from the ‘spoiler phenomenon’ to the lack of enforcement. In contradistinction, the success of the Abuja II Peace Agreement was anchored in the fact that the requisite steps were taken to ensure the effective operationalisation of the battery of elements required for success. For example, the Taylor-led NPFL’s perennial role as the ‘spoiler’ was addressed by the unwillingness of Burkina Faso and Côte d’Ivoire, the warlordist militia’s two major supporters, to continue to acquiesce in the warring faction’s obstruction of the peacemaking process. Also, ECOWAS developed the political will to enforce the agreement, including the threat of establishing a war crime tribunal.


There is growing scholarly interest in the study of peace agreements as frameworks for ending civil wars (Hampson 1996; Walter 1997; Bell 2000; Spears 2008; Mutwol 2009). The reason for this development is because, as Mattes and Savun (2009:737) observe, ‘Civil wars are more frequent, more deadly, and longer in duration then interstate wars, and they prove to be more difficult to settle peacefully’. Functionally, peace agreements have impacted civil wars in three major ways. First, in cases such as Zimbabwe (1980), Guatemala (1992–1998), Mozambique (1992–1995) and Angola III (2002), peace agreements succeeded in terminating civil wars (Stedman 2001; Mutwol 2009). Second, in other cases, such as Sudan (1972), Sri Lanka (1989–1990), Somalia (1990), Angola I and II (1991 and 1994) and Rwanda (1993), peace accords failed to end violent conflicts (Stedman 2001; Murshed and Verwimp 2008). The resultant effect was the reversion to warfare and its associated cataclysmic consequences. Third, in other cases such as Lebanon (1990), Cambodia (1991–1994), and Bosnia (1995), peace agreements were partially successful in helping to arrest the tide of armed violence (Stedman 2001).

In the specific case of the first Liberian civil war (1989–1996), the focus of this article, there was seemingly an endless cycle of peacemaking and resumption of war. This was evidenced by the fact that the various peace agreements were repeatedly violated by the Charles Taylor-led National Patriotic Front of Liberia (NPFL) (Kieh 2009; Mutwol 2009). Thus, when the civil war finally ended in 1996, it had taken sixteen peace agreements to reach the termination phase (Kieh 2009).

Against this background, the purpose of this article is twofold. First, it will examine the reasons the previous sixteen peace accords failed to end the first Liberian civil war. Second, the article will probe why the Abuja II Peace Accord, the seventeenth peace agreement, was able to terminate the war. In order to address the research problem, the article is divided into six parts. In the first part, the theoretical issues are explored, including a review of the scholarly literature and of the theoretical model that guides the study. Next, the study examines the factors and forces that caused the first Liberian civil war. Third, the article explores the substantive contents of the various peace agreements. Fourth, the study tackles the reasons for the failure of the previous sixteen peace accords to end the war. Fifth, the article probes the reasons why the Abuja II Peace Accord succeeded in ending the civil war. Finally, the study draws the major conclusions.

1. Theoretical issues

A literature review

Hampson (1996) examines the raisons d’être for the success and failure of peace agreements. He proffers several major conditions. First, peace agreements can unravel because parties to conflicts simply come to the conclusion that it is no longer in their interest to abide by the agreements they have negotiated (1996:13). Second, confidence-building measures are major determinants of the outcomes of peace agreements. Third, there is the centrality of third parties that can offer carrots or wield sticks to ensure that the process does not become derailed (Hampson 1996:11–12).

Treading along the broad path regarding the outcomes of peace agreements mapped out by Hampson, Stedman (1997), in his seminal article, examines one of the major causes for the failure of peace agreements to help end civil wars. He posits that spoilers – leaders and parties who believe that peace emerging from negotiations threatens their power, worldview, and interests, and use violence to undermine attempts to achieve such peace (Stedman 1997:5) – pose the greatest threat to the successful implementation of a peace agreement. He identifies three major genres of spoilers: limited, greedy and total. Limited spoilers have narrow objectives. In the case of greedy spoilers, they are driven by particularistic opportunistic agendas that tend to wax and wane with the dynamics of the conflict. As for the total spoilers, they desire nothing short of the acquisition of power.

Gent (2007) explores the conditions for peace agreements to have successful outcomes. He argues that peace agreements that make sufficient concessions to the rebels are more credible because they decrease the ability of the government to renege on them. For example, if the rebels are represented in the transitional government, this would enable them to protect themselves against future defections by the current government (Gent 2007:5). Against this backdrop, it does not seem plausible that governments can agree to offer ‘everything’ (i.e. territorial autonomy or complete control over resources) and then be able to take it all away once it regains its strength after the civil war (Gent 2007:5).

Mattes and Savun (2009) transcend the government-based commitment problem postulated by Gent. They do so by arguing that the commitment problem is much broader in that it applies to rebel groups as well. Hence, the success of a peace agreement is contingent upon its capacity to mitigate the commitment problem between and among the warring factions (Mattes and Savun 2009:738). In this vein, they identify two types of provisions in peace agreements that can help address commitment problems and thus reduce the chance of further warfare: fear-reducing and cost-increasing provisions (Mattes and Savun 2009:738). Fear-reducing provisions are designed to help ensure mutual security. The cost-increasing provisions raise the stakes and cost of the return to warfare by either of the parties.

Derouen and Wallensteen (2009) examine the preconditions for the duration of a peace agreement. They posit that the duration is dependent upon the nature of the power-sharing provisions. For example, if the power-sharing arrangements are too costly for the government to implement, then it will abandon them. On the other hand, the rebel faction or factions would be adamant that the provisions be implemented. The resultant tugs and pulls could then unravel the peace agreement.

The Peace Agreement Model

The study uses the peace agreement model as its analytical framework for evaluating the various accords that were signed during the first Liberian civil war. The model was formulated by drawing from the lessons learned from the ‘best practices’ in peacemaking as reflected in the scholarly literature and the experiences of various practitioners, who have been involved in peace processes in civil wars across the globe. The model was ostensibly developed for two major reasons. The first is to help contribute to the development of the indicators for evaluating the successes and failures of peace agreements in civil wars. hA related reason is to provide a roadmap for scholars interested in studying peace agreements, and practitioners involved in mediating accords and working with the warring parties to implement them. The model is based on several interlocking tenets.

The characteristics of the warring factions are critical to the outcomes of peace agreements. This is because the attributes of the conflicting parties and the resulting insights provide understanding of the parties’ behaviours during the peacemaking process. Some of the major characteristics include the agendas of the various parties and their strategies for pursuing them. One of the major strategies is to be the ‘spoiler’ (Stedman 1997; Mutwol 2009). This role is performed by either the recalcitrant party’s refusal to participate in a peacemaking process, or its efforts to thwart the implementation of a peace agreement.

Another matter concerns the conflict dynamics. In other words, it is important to examine the particular state or phase of the conflict during which a peace agreement is signed. This is important because the dynamics affect the behaviour of the parties to the conflict, especially their motivations for signing the accord. Linked to this is the spoiler phenomenon.

The substantive contents of the agreement are obviously crucial. These would cover issues such as power-sharing arrangements, fear-reducing methods and cost-increasing provisions (Hampson 1996; Stedman 1997; Gent 2007; Mattes and Savun 2009). Overall, there would be a trajectory for making the transition from war to peace consisting of such activities as disarmament, demobilisation and elections.

Also important are the roles of third parties. As Mutwol (2009:8) aptly observes, ‘Third parties play a number of important roles in peace agreements.’ Broadly, there are two major types of roles: positive and negative. The former refers to activities that third parties undertake that contribute to the peace process and the success of the peace agreement. These may include activities such as peacemaking, peacekeeping and/or peace enforcement. On the other hand, third parties can play various negative roles that can undermine the success of the peace process and the peace accord. For instance, the provision of arms, sanctuary, and diplomatic or political support to the various warring functions. Such support could embolden the warring parties by making them intransigent and uncooperative in the implementation of a peace agreement.

The use of appeasement, which can serve as an effective ‘carrot’ for inducing a recalcitrant faction’s compliance with a peace agreement. In this instance, the act of appeasement should be case specific and appropriately calibrated. On the other hand, the use of appeasement can be counterproductive when it is employed recurrently, and fails to take cognizance of the specificities of a case. One major negative outcome could be the contribution to the ‘commitment deficit’ of the parties to the conflict. Such an outcome is made possible by the various parties’ belief that the ‘peacemaker’ would continuously pander to their repeated demands (Stedman 1997).

Another major contour is the centrality of an effective enforcement mechanism (Stedman 2001). In order for the parties to the conflict to comply with the terms of a peace agreement, a third party or parties must be willing to play the role of the ‘enforcer'(Stedman 2001; Derouen and Wallensteen 2009; Mattes and Savun 2009). That is, the third parties or parties must be willing to deploy the requisite resources, and bring them to bear on the implementation of the peace agreement.

Ultimately, it is important to assess the outcomes of peace agreements – as successful, unsuccessful or mixed. The evaluation is anchored on two major indicators. The extent to which the substantive contents of a peace agreement have been implemented. This is critical because it is comparatively easier to get parties in a conflict to negotiate a settlement than to actually implement it (Walter 2002). Also, consideration is given to the issue of the termination of the war and its associated violence, and the prevention of a relapse into war (Walter 2002).

2. The first Liberian civil war: A background

The first Liberian civil war was the result of the contradictions and the resultant multidimensional crises of underdevelopment – cultural, economic, political, security and social – engendered by the Liberian state (Kieh 2008; Kieh 2011). These crises were created during the two phases of the development of the Liberian state – the settler period (1820–1926) and the neo-colonial era (1926–present) (Kieh 2008; Kieh 2011). The settler phase of state-building was characterised by the primacy of ethno-cultural characteristics. Interestingly, these differences intersected with class distinctions (Burrowes 1982). That is, the ancestral origins and skin pigmentation of the various groups intersected with their respective class positions in the local political economy. For example, the light-skinned African-American repatriates occupied the middle stratum of the caste cum class structure during the colonial era (Burrowes 1982; Kieh 2008), and the upper tier during the commonwealth period and the initial four decades of the independence era (Kieh 2008). During the neo-colonial stage of state-building, the pivot was the ascendancy of class as the defining feature of the state and its local political economy (Kieh 2008; Kieh 2011).

Cumulatively, the state-building project generated multifaceted crises of underdevelopment that led to civil conflict and then warfare. Culturally, the crises revolved around two major problems: the settler-indigene divide, and the Krahn versus Gio and Mano antagonisms. In the case of the former, the foundation was laid in 1820 upon the arrival of the African-American repatriates or the settlers in Liberia (then known as the Grain Coast) (Kieh 2008:124). The repatriates were sent from the United States where they had been enslaved (Smith 1972; Beyan 1991; Sawyer 1992; Kieh 2008). When the repatriates or the Americo-Liberians or the settlers arrived in Liberia, they met sixteen indigenous ethnic groups occupying the area. These indigenous ethnic groups had their own polities replete with political, economic, social and cultural systems. Interestingly, because they had lived in the United States, albeit as slaves, the settlers came with a superior attitude. As Brown (1941:125–126) notes, ‘The American-Liberians [the settlers] considered themselves a ‘superior people’; thus, there was no sense of feeling of oneness with the Africans [the indigenes]’.

The resultant conflict between the two ethno-cultural stocks revolved around the issues of land, religion, and cultural values and norms. Although the conflict has undergone various permutations over the years, it remains unresolved.

The second major cultural crisis pitted the Krahn ethnic group against the Gio and the Mano. The conflict, which commenced in 1982, was caused by personal and political differences between Head of State Doe (Krahn ethnic group) and General Thomas Quiwonkpa (Gio/Mano ethnic group). The conflict reached its crescendo in November 1985, when General Quiwonkpa, one of the leaders of the coup that brought Doe to power on 12 April 1980, and a former confidante of Doe, staged an abortive coup (Mutwol 2009). In the aftermath of the coup, Doe instrumentalised ethnicity by portraying the Gio and Mano ethnic groups as subversive and anti-Krahn (Kieh 2008). Thereafter, the Doe regime undertook a ‘scorch the earth campaign’ in Nimba County, the base of the Gio and Mano ethnic groups, that resulted in the injury and deaths of hundreds of people from the two ethnic groups (Human Rights Watch 1990).

In the economic realm, the crises of underdevelopment were reflected in several indices. However, two will be discussed. One of the major pathologies of the country’s peripheral capitalist economy centred around class inequities: the ruling class – local state managers and entrepreneurs and the foreign-based owners of multinational corporations like Firestone and other businesses – versus the subaltern classes consisting of the working class, the peasantry and the lumpens. For example, in 1985, the upper or ruling class, which accounted for about 5% of the national population, received about 68% of the national income (Ministry of Planning and Economic Affairs 1986). By the end of the 1980s, the country’s Gini Coefficient was a high 0.53 (Peters and Shapouri 1997:45). Similarly, in 1985, the ruling class, comprising approximately 6% of the population, owned and controlled about 70% of the wealth (Kieh 1997:27). The other indicator of the crisis was the high rate of unemployment. In 1988, for example, the rate of unemployment stood at 36.2% (United Nations Development Programme 1990).

Socially, the crises found expression in many areas. Two of the critical ones were education and health care. In terms of education, fundamentally, the Liberian state has never been interested in the education of its citizens (Kieh 2008:18). This manifestation of state neglect was reflected in the lack of access to educational opportunities, including an inadequate number of schools, the inadequacy of logistics, equipment, and supplies, and the challenge of the prevalence of untrained teachers at the primary and secondary levels (Kieh 2008). Similarly, in the area of health care, the crises were manifested in the inadequacy of the numbers of medical personnel, medical facilities, supplies, and equipment, among others. For example, in 1989, there was an estimated number of 3 526 health workers – doctors, nurses, etc. – in the country (World Health Organisation 2003).

Significantly, with the Liberian state severely weakened by the afore-mentioned multidimensional crises of underdevelopment, the Doe regime, despite its repressive proclivities, became vulnerable to armed insurrection (Kieh 2009:49). Against this background, on 24 December 1989, the Charles Taylor-led National Patriotic Front of Liberia (NPFL) took advantage of the perennial legitimate grievances of the Liberian subaltern classes and launched an armed rebellion (Kieh 2009:48). Characteristically, the Doe regime responded with the battery of its military assets. The resultant ‘tugs and pulls’ occasioned the first Liberian civil war.

3. The Peace Agreements


The failure of the domestic peace initiative undertaken by the Religious Leaders of Liberia, coupled with the emergence of a state of anarchy, drew the concern of the Economic Community of West African States (ECOWAS) (Kieh 2009). Initially, the various member states engaged in a series of consultations regarding the appropriate action to take vis-í -vis the Liberian civil war (Adebajo 2002; Kieh 2009). They had to consider several major impediments. First, the organisation was primarily an economic entity; and thus, its charter did not make any provision for dealing with regional conflicts – both within and between member states (Kieh 2009). Second, the organisation’s own Non-Aggression Treaty forbids, inter alia, interference in the internal affairs of member states. Third, the Charter of the OAU, the multi-purpose continental-wide group, emphatically prohibited member states from intervening in each other’s domestic affairs.

Interestingly, despite these prohibitions, ECOWAS decided to intervene in the Liberian civil war. Salim Ahmed Salim, the then Secretary-General of the OAU, formulated the raison d’être for ECOWAS’ intervention thus: ‘Africans are one people. It is hence unacceptable that a part of that people should stand in silence and in seeming helplessness when another part is suffering’ (West Africa 1990:2691).

However, ECOWAS’ decision to intervene transcended primordial and humanitarian concerns, because the Liberian civil war directly affected the member countries in two major ways. First, several of their citizens resident in Liberia were being killed, while others were taken as hostages. Second, the member states were concerned that the Liberian civil war could have a domino effect on the rest of the sub-region. This fear was based on the fact that since the preponderant majority of the member states of ECOWAS were governed by repressive regimes, the fact that the Taylor-led National Patriotic Front of Liberia had dissidents from various West African States within its ranks became a major concern of the member states of ECOWAS. The concern was rooted in the calculation that Liberia could have become, if Taylor’s NPFL had taken control of the government, the beachhead from which these dissidents could have waged campaigns of subversion against their respective countries.

The ECOWAS Peace Plan: The Peacemaking Framework

ECOWAS devised a plan that was designed to end the civil war. Interestingly, the process of formulating the plan did not include consultations with the warring factions, civil society organisations or political parties in Liberia. In other words, ECOWAS designed a peace plan based on its assessment of the nature and dynamics of the civil war. The plan had the following elements:

  1. An immediate ceasefire.
  2. The formation of an ECOWAS Peacekeeping Force.
  3. The establishment of an interim government that would exclude both Sergeant Doe, the incumbent Head of State, and Charles Taylor, the leader of the insurgency.
  4. The holding of free and fair elections within a year, under international supervision and observation.

A Standing Mediation Committee, consisting of nine member states, was established to implement the peace plan. The plan was endorsed by the Organisation of African Unity (now the African Union) and the United Nations. However, Taylor’s NPFL (and its two principal supporters, Côte d’Ivoire and Burkina Faso) rejected the plan. So, per the tenets of the peace agreement model, this accord failed for several reasons. The substantive contents of the peace agreement were unacceptable to Taylor and his NPFL, particularly because they prevented him from serving as the head of the interim government. The related reason was that Burkina Faso and Côte d’Ivoire as major third parties emboldened Taylor by refusing to accept the terms of the peace accord as well. Yet, another major reason was the lack of enforcement. ECOWAS did not have the political will to induce compliance from Taylor and his NPFL.

The rejection of the ECOWAS Peace Plan by Taylor’s NPFL forced the organisation to convene a meeting of the Standing Mediation Committee (6–7 August 1990). The crux of the meeting was to explore ways of appeasing Taylor’s NPFL. At the end of the meeting, a final communiqué was issued. Its basic planks were:

  1. A re-affirmation of the ECOWAS Peace Plan.
  2. The call for the convening of an All-Liberia National Conference to form an interim government. The said government would include representatives from the various political parties, interest groups and the warring factions.
  3. The interim government would create the modalities for the holding of general and presidential elections within 12 months. None of the leaders of the warring parties would be eligible to head the transitional government. Similarly, the interim president would be ineligible to stand for the ensuing presidential election.

The communiqué was rejected by the NPFL. Again, the peace agreement failed because its contents, especially the prohibition against any leader of a warring faction serving as president of the interim government was unacceptable to Taylor and the NPFL. This was because Taylor wanted to head the interim regime. Another factor, as per the peace agreement model, was the lack of enforcement. Again, ECOWAS did not have the will to enforce the terms of the agreement.

The Banjul I Peace Accord

From 27 August to 30 September 1990, an ECOWAS-sponsored All-Liberia Conference was convened in Banjul, the Gambia. The meeting brought together the leaders of Liberia’s various political parties, interest groups and warring factions, with the exception of the NPFL, which refused to participate. The NPFL’s refusal was based on the fact that it made the determination that the conference would not have met its ultimate desire of making Charles Taylor, its leader, the head of the interim government. At the close of the conference, the following measures were taken:

  1. The President and the Vice-President of the Interim Government of the National Unity were elected. Correspondingly, the structure of the Interim Legislative Assembly was devised. The law-making body consisted of thirty-five members – two from each political party, one from each of the thirteen regions of the country, four from the Prince Johnson-led Independent National Patriotic Front of Liberia INPFL), and six from the Charles Taylor-led National Patriotic Front of Liberia. Additionally, the speakership of the interim legislative body was reserved for Taylor’s NPFL, and the deputy speakership was allotted to Prince Johnson’s INPFL.
  2. The 1984 constitution remained in force; but the appropriate modifications were made that were apropos to the prevailing circumstances.
  3. The President and the Vice-President of the Interim Government were not eligible to run for public office in the ensuing elections. However, the Speaker of the Interim Legislative Assembly and other government officials were allowed to become candidates in the elections.

Despite the attempts to accommodate Taylor, his warring faction denounced the conference. As the NPFL’s Justice Minister J. Laveli Supuwood asserted, ‘The interim government led by Dr. Amos Sawyer is a group without legitimacy …’ (West Africa 1990:2714).

Barely a week after the conference, an important development occurred: Head of State Doe was captured and killed by the Johnson-led INPFL. The emergent perception was that with the death of Doe, and the formal collapse of his regime, Taylor’s NPFL would be willing to accept the ECOWAS Peace Plan, and participate in the interim government. However, the accord failed to end the war for several specific reasons. The central element of the contents of the peace agreement dealing with the interim government did not satisfy Taylor, the principal warlord. This was evidenced by Taylor’s reaffirmation of his demand that he be given the presidency as the quid pro quo for ending the war (Kieh 2009). Another factor was the character of the NPFL: The militia consistently played the role of the ‘spoiler’. That is, it made the decision to undermine the peace process, as long as Taylor was not handed the leadership of Liberia. Also, no actor in the international community was willing to enforce the peace accord.

The Banjul II Peace Accord

On 24 October 1990, a meeting was held in Banjul, Gambia, under the auspices of ECOWAS. The meeting was attended by representatives of the interim government of Liberia, and the INPFL. The Taylor-headed NPFL refused to attend. The major outcome of the meeting was the establishment of a ceasefire agreement. However, in the absence of the NPFL, the other warring faction, it was difficult to achieve this goal.

This agreement failed for two major reasons. The contents of the agreement still did not meet the NPFL’s central demand that Taylor be made the head of the transitional government. Like the previous failed agreements, neither ECOWAS nor any other external actor was willing to serve as the enforcer of the agreement. Hence, the trend of appeasing the NPFL continued with the mediation of another agreement.

The Bamako Accord

After intensive diplomatic efforts, particularly with the support of Côte d’Ivoire and Burkina Faso, ECOWAS persuaded Taylor to sign a formal ceasefire agreement with Johnson’s INPFL and the remnants of Doe’s army on 30 November 1990. However, after signing the accord, Taylor claimed that although he accepted the cessation of hostilities, he did not accept the other provisions that buttressed the ECOWAS Peace Plan. This emerging trend of vacillation on the part of Taylor’s NPFL continued the prolongation of the impasse and the war.

Again, the failure of the peace agreement was due to the fact that the key elements of the substantive contents did not meet the NPFL’s central demand that Taylor be made the head of the transitional government. The related problem was that the NPFL was determined to continue performing its negative role as a ‘spoiler’ as long as its key demand was not met. In addition, the agreement lacked enforcement, which is the bedrock for the successful implementation of any peace agreement.

The Banjul III Peace Accord

On 21 December 1990, the leaders of the three warring factions – Taylor (NPFL), Johnson (INPFL), and Bowen (remnants of Doe’s Army) – met in Banjul. In a clear concession to Taylor, the group agreed to convene a second All-Liberian Conference that would hold a new election for an interim administration. However, Taylor’s NPFL made a concerted effort to stymie the development of the modalities for the proposed conference by refusing to cooperate.

Although this agreement subsequently led to the holding of the Second All-Liberian Conference, the difficulties in formulating agreed upon modalities for the conference made its failure imminent. The key reason was that the NPFL knew that the conference would not have led to Taylor becoming the leader of the interim government. Thus, the NPFL chose to continue playing its ‘spoiler’ role. Significantly, in the absence of an enforcer, the implementation of the agreement was doomed to fail, against the backdrop of the emergent trend of the NPFL not abiding by the terms of even the peace agreements that it signed.

The Lomé Peace Accord

From 12 to 13 February 1991, a meeting was convened in Lomé, Togo, to develop the modalities for the implementation of the ceasefire agreement. It was attended by members of the ECOWAS Standing Mediation Committee, the President of the Interim Government of Liberia and the leaders of the three warring factions. Again, after the meeting, Taylor denounced the agreement that he had signed.

Why was this the case? The agreement’s key content regarding the leadership of the interim regime continued to be unacceptable to the NPFL. Thus, the NPFL continued to undermine the peace accord. Moreover, the agreement lacked an enforcer, which was prepared to induce compliance from both the NPFL and the other warring parties.

The Monrovia Peace Accord

In order to further appease Taylor’s NPFL, a second All-Liberian Conference was held in Monrovia, Liberia, from 14 March to 2 April 1991, under the sponsorship of ECOWAS. The participants included Liberia’s various political parties, interest groups, the warring factions and representatives from ECOWAS, the then OAU and the United Nations. During the initial stage of the conference, Taylor’s NPFL suggested a ‘new proposal’ for ending the war. The proposal called for the interim administration to be based on a troika consisting of a chairman and two co-chairmen. Taylor insisted that he become the chair, that the other warring parties and political parties and interest groups elect one co-chair, and the second co-chair be a ‘widely respected’, ‘politically neutral’ Liberian to be agreed upon by all Liberians. Correspondingly, the NPFL called for the replacement of Archbishop Michael Francis as co-chair of the Conference. However, the conference rejected the NPFL’s proposals on four counts: 1) it contained the old demand that Taylor be the President of the Interim Government and this was antithetical to the conference’s purpose of establishing the foundation for the building of a new and democratic society; 2) the proposal was vague; 3) overall, it represented a deliberate ploy by Taylor’s NPFL to undermine the conference, by making a proposal that it knew would be unacceptable; and 4) Archbishop Francis would not be replaced because he had demonstrated impartiality. Characteristically, the NPFL delegation decided to boycott the conference. Nevertheless, the conference proceeded.

At the conclusion, the Conference elected a President and a Vice-President of the Interim Government of National Unity, and an expanded Interim Legislative Assembly. Despite the continual intransigence of Taylor’s NPFL, the conference made new concessions to Taylor: 1) it allotted 40% of the seats in the legislative body to the NPFL; 2) reserved the position of speaker for the NPFL; 3) allocated seats on the interim supreme court to Taylor’s NPFL; and 4) agreed that Taylor would be the only top official of the interim government who would be eligible to run as a candidate in the immediate presidential election. Again, Taylor rejected the offers, and continued his insistence that he must be both the President of the Interim Government, and a candidate in the ensuing election. This was because the agreement did not meet the NPFL’s key demand. Based on the pattern of appeasement, which ECOWAS had adopted as the kernel of its peacemaking architecture, Taylor knew that the NPFL would not be compelled to abide by the terms of the agreement. Also, with the continuing split between Burkina Faso and Côte d’Ivoire, which supported Taylor’s NPFL, and the other members of the organisation which did not want to acquiesce to Taylor’s principal demand that he be made the head of the interim regime and subsequently be allowed to contest the Liberian presidency in the future election, the NPFL knew that there would be no enforcement of the terms of the agreement.

The Yamoussoukro Peace Accords

ECOWAS sponsored four meetings in Yamoussoukro, Côte d’Ivoire between June and October 1991. The central purpose of these meetings was to devise the modalities for implementing the ECOWAS Peace Plan. In a classic diplomatic stroke, ECOWAS designated then Ivorian President Félix Houphouët-Boigny, the NPFL’s strongest supporter, as Chairman of the Group of Five, the committee selected to ensure the full implementation of the ECOWAS Peace Plan.

As usual, having agreed to abide by the terms of the Yamoussoukro Rounds, Taylor resorted to his old strategy of denials and the obstruction of the disarmament and encampment processes. That is, after agreeing to participate in the disarmament and encampment processes, the NPFL reneged on its promise. Two factors accounted for this. Taylor’s main demand continued to be unmet by the various peace agreements. In spite of the progress made to resolve the rift between Burkina Faso and Côte d’Ivoire, the NPFL’s principal supporters in ECOWAS, on the one hand, and the other members of the organisation, on the other, the NPFL made the determination that it could continue playing its ‘spoiler’ role without the fear of facing enforcement action.

The Geneva Peace Accords

Two rounds of discussion were held in Geneva, Switzerland. The central purpose of the first meeting, which was held under the auspices of ECOWAS in April 1992, was to clarify the mandate of ECOMOG, the ECOWAS Peacekeeping force, regarding the enforcement of the various peace accords. The second one was held in July 1993, under the auspices of ECOWAS, the then OAU, and the UN. The major purpose was to provide the Taylor-led NPFL with a final opportunity to abide by the terms of the myriad peace agreements, particularly with regard to disarmament, encampment and demobilisation.

Again, both of these peace conferences failed to end the war. Like the previous peace accords, the key substantive contents of the Geneva peace agreements did not meet Taylor’s overarching demand that he be allowed to head the transitional government, and to run as a presidential candidate in the ensuing special election to settle the country’s leadership question. Another recurring factor was the lack of enforcement. ECOWAS, the then OAU and the UN were unwilling, either singularly or collectively, to enforce the terms of the peace accords.

The Cotonou Peace Accord

The Cotonou Peace Accord was the by-product of a meeting held in Benin in 1993 under the auspices of ECOWAS. The meeting was attended by the various warring factions and the representatives of the Interim Government of National Unity. At the end of the meeting, a peace accord was formulated and signed by the warring factions. The key elements of the new peace agreement were power-sharing arrangements and security. In the case of the former, for the first time, the leaders of the major warring factions – Taylor (NPFL), Kromah (United Liberation Movement of Liberia for Democracy, ULIMO), and Boley (Liberia Peace Council, LPC) – were given direct roles in the executive leadership of the country through a new council of state. The other three members of the council were civilians with one of them serving as chair and Head of State. Dupuy and Detzel (2008:4) derisively referred to this arrangement as ‘power for gun policy’.

However, despite this major concession to Taylor, he still refused to end the war. This was due to various reasons. Although this peace agreement came closer to meeting the NPFL’s key demand regarding the arrangement of political power during the transitional period, it failed to satisfy Taylor’s ultimate desire to head the transitional government. So, the NPFL agreed to participate in the transitional government, on the one hand, while continuing to undermine the peace process, on the other hand. This strategy was also informed by the fact that Burkina Faso and Côte d’Ivoire, while remaining supportive of the NPFL, were becoming increasingly concerned about Taylor’s perennial commitment deficit vis-í -vis the peace process. Thus, the NPFL decided to appease its two major supporters by appearing reasonable – by agreeing to serve in the transitional government. But, concurrently, the NPFL continued to play its ‘spoiler’ role. Importantly, the NPFL’s half-hearted approach made it difficult for ECOWAS, even if the organisation had the political will, to compel the Taylor-led militia to end its practice of continually undermining the peace process.

The Akosombo Peace Accord

The Akosombo Peace Accord was the result of a meeting held in Ghana on 12 September 1994. The meeting, which was chaired by President Rawlings of Ghana, who was also the then Chair of ECOWAS, was attended by Taylor (NPFL), Kromah (ULIMO), and Bowen (Chief of Staff of the Armed Forces of Liberia). The resultant peace agreement had two major elements. The decision-making formula in the Council of State, the executive arm of the transitional government, was changed from consensus to majoritarian rule. The second issue was the establishment of a new deadline for the commencement of disarmament – following the installation of the new transitional government.

Although the NPFL signed the agreement, it subsequently failed to abide by the section of the substantive contents of the accord dealing with disarmament. This provision was central to ending the war. One major reason was that the NPFL’s key demand had not been met by either the previous failed peace accords or the Akosombo agreement. So, consistent with its strategy, in light of the changing conflict dynamics, to appear ‘reasonable’, the NPFL accepted and abided by the provision of the agreement dealing with the organisational modalities of the Council of State – the emergent executive arm of the transitional government. However, the NPFL made the determination that this would not have any significant impact on the continuation of the war, as long as the Taylor-led militia refused to abide by the critical disarmament provision. Also, with the continuing lack of the requisite political will to enforce the peace accord on the part of ECOWAS and the other members of the international community, the NPFL made the determination that it could, at this juncture, continue to give the superficial appearance of being committed to the peace process by abiding by the non-security related provisions, while undermining the successful implementation of the security dimension.

The Accra Peace Accord

On 21 December 1994, the leaders of the NPFL, ULIMO and the remnants of the Armed Forces of Liberia met in Accra, Ghana. Ghanaian President Rawlings chaired the meeting in his capacity as the Chair of ECOWAS. At the conclusion of the meeting, the ‘new’ peace agreement revolved around three major issues:

  1. The establishment of a new deadline for a ceasefire.
  2. The reconfirmation of the composition of the Council of State.
  3. The holding of a national election and the subsequent installation of a new government in 1996.

Once more, the NPFL signed the agreement, but subsequently undermined its security provisions that called for the cessation of hostilities, and the creation of propitious conditions for disarmament. And again, although the NPFL was now a part of the transitional government, it continued to undermine the peace process by failing to agree to a ceasefire because its principal demand that Taylor be made the head of the transitional government and be allowed to contest the presidency in any future election had not been met. Characteristically, ECOWAS continued to lack the political will to induce full compliance from the NPFL.

The Abuja I Peace Accord

The Abuja I Peace Accord was the result of a meeting held in Abuja, Nigeria, on 19 August 1995. The meeting was attended by the six warring factions – NPFL, ULIMO-J, ULIMO-K, Liberian Peace Council, Lofa Defence Force (LDF), the NPFL-CRC, a so-called break-away faction from the Taylor-led NPFL – and the representative from the Liberian National Conference (LNC), the amalgam of civil society organisations and political parties.

The resulting peace agreement covered four major areas. The power-sharing arrangements anchored by the six-member Council of State were re-affirmed. Also, amid the continuing fighting, a new date for a ceasefire was established. Further, ECOMOG was given the authority to enforce the provisions of all of the previous peace agreements. Finally, ECOWAS, the then OAU and the UN were authorised to supervise the ad hoc election commission that was to conduct the national election.

However, the accord failed to end the civil war for several reasons. First, the NPFL continued its ‘spoiler role’ by failing to abide by the proviso that all hostilities be ended. This was because its principal demand still had not been fulfilled. Second and related, although ECOMOG, the peacekeeping force, undertook some enforcement actions, they were not enough to induce full compliance from the NPFL. In other words, ECOWAS still lacked the political will to undertake complete enforcement of the peace accord. This was because neither the organisation nor the broader international community was willing to use military means to force the NPFL to abide by the terms, especially the security provisions, of the peace accord.

The Abuja II Peace Accord

A little over year after the Abuja I Peace Accord was signed, another meeting was convened in Abuja, Nigeria, on 17 August 1996. The meeting was attended by the six Liberian warring factions and the representative of the LNC. The meeting was designed to review the progress of the Abuja I Peace Accord. At the end of the meeting, the determination was made to formulate another peace accord. The new peace agreement dealt with three major issues:

  1. It selected a new civilian chair of the Council of State and Head of State of Liberia. Also, it stipulated that all other members of the council should be vice-chairmen with equal status.
  2. It developed a new schedule for implementing the modalities for the transition from war to peace.
  3. It extended the tenure of office of the transitional government to 2 August 1997.

Significantly, after the failure of the previous peace accords, the Abuja II Peace Agreement succeeded in setting into motion the process of ending the civil war. The rift between the member states that supported the NPFL and those that were opposed to its desire to capture state power through military means had been reconciled. Consequently, for the first time during the peace process, there was unanimity within ECOWAS that the war needed to end. As the peace agreement model stipulates, the collective commitment of third parties to ending a civil war through the successful implementation of a peace accord is critical. Another reason was that ECOWAS decided to change its perennial peacemaking strategy of appeasing the NPFL by insisting that the warlordist militia fully complied with the terms of the accord or face several punitive measures, including the establishment of a war crimes tribunal to try Taylor and his principal lieutenants for the commission of war crimes and crimes against humanity. Also, the decision to muster the political will to fully enforce the modalities of the accord, especially the critical provision dealing with the ‘transition from war to peace’ sent an unambiguous message to Taylor and the other warlords. (See also the section below on the Abuja II Peace Accord and the termination of the war.)

4. The failure of the Peace Accords: Banjul I to Abuja I


Several factors militated against the successful implementation of the initial sixteen agreements that were negotiated to terminate the first Liberian civil war (Banjul I to Abuja I). In this part of the article, the tenets of the peace agreement model will be used as the analytical and evaluative criteria for assessing the impact of the various peace agreements on the termination of the first Liberian civil war.

The characteristics of the warring parties

All of the warring factions had certain shared characteristics that compromised their ability to negotiate in ‘good faith’ and to abide by the terms of the various peace agreements to which they were signatories. For example, their participation in the war was not motivated by the desire to seize state power and use it as a transformative instrument for democratically reconstituting the neo-colonial Liberian state – so that it could address the needs of the country’s subaltern classes (Kieh 2008). Instead, the various warring factions served as the instruments of violence through which their respective leaders engaged in the predatory accumulation of wealth principally through the sale of the country’s resources – gold, diamonds, timber, rubber and iron ore (Reno 1996; Atkinson 1997; Adebajo 2002).

Another shared characteristic was that the various warring factions (with the exception of the Taylor-led NPFL) were designed to serve as ‘bargaining vehicles’ through which their leaders could acquire political power and consequently economic power through ECOWAS’ peacemaking formula that was based on what I term ‘rewarding warlordism’ – the allocation of state agencies and positions to various warring factions as the mainstay of the transitional power-sharing arrangements.

In contradistinction to all of the other warring factions, the Taylor-led NPFL’s primary objective was to make Charles Taylor the new President of Liberia (Kieh 2009). For example, during one of his meetings with the Inter-Faith Mediation Committee, the conflict resolution mechanism of the Religious Leaders of Liberia, an ad hoc amalgam of Christian and Islamic clerics established during the first Liberian civil war, Taylor insisted on ‘the immediate resignation of Mr Doe and his regime, and the consequent handing of power over to the National Patriotic Front of Liberia’ (Francis 1990:2). According to Taylor, ‘It was only the matter of hours for him to take Monrovia and overthrow Mr. Doe … even while the peace talks were still taking place’ (Francis 1990:2).

Thus, the NPFL recurrently assumed an antagonistic posture toward all of the peace agreements because they did not include a provision to make Taylor the President of Liberia. For example, during the Second All-Liberian Conference in 1991, the delegation of the NPFL angrily withdrew from the conference after its proposal that Taylor be made the President of Liberia was rejected by the overwhelming majority of the delegates. Subsequently, the NPFL resorted to warfare.

The conflict dynamics

The dynamics of the war adversely affected the successful implementation of the various pre-Abuja II peace agreements in several major ways. The increasing seizure of territory, especially resource-rich territory, by the Taylor-headed NPFL contributed to the warring faction’s intransigence. This was because the NPFL saw the various peace agreements as major obstacles to the predatory accumulation of wealth by Taylor and its other leaders (Reno 1996; Adebajo 2002). That is, the termination of the war by the various peace accords would have made it difficult for the NPFL to continue the unbridled sale of the country’s various natural resources – iron ore, rubber, gold, diamonds and timber (Reno 1996; Adebajo 2002).

Another factor was that the seemingly unending cycle of warfare and peacemaking enabled the NPFL to use the peace agreements as opportunities to ‘buy time’ during periods when the warlordist militias was under military pressure either from the peace enforcement actions of ECOMOG, the peacekeeping force, and/or the other warring factions. Under these circumstances, Taylor used the signing of various peace agreements as an instrument for ending the military pressure, getting new supplies of arms, and rethinking the NPFL’s ‘military strategy’. Then, once these objectives were achieved and the NPFL was strengthened militarily, Taylor would renege on the peace agreements and revert to warfare. Unfortunately, ECOWAS and the rest of the international community were vulnerable to these manipulative tactics by the NPFL, as evidenced by their willingness to continuously mediate new peace accords each time the NPFL violated the previous one.

What also happened was that the prolongation of the war, and especially the availability of natural resources as well as the opportunity to secure positions in the various transitional governments encouraged the mushrooming of new warlordist militias. For example, beginning in 1992, new warring factions – Liberian Peace Council, Lofa Defence Force, and the NPFL-Central Revolutionary Council – began to emerge, and ULIMO splintered into two factions (ULIMO-J led by Roosevelt Johnson, and ULIMO-K headed by Alhaji Kromah). In short, each emergent warring faction followed an interesting ‘logic’: the seizure of territory with natural resources led to the opportunity to privately accumulate wealth, and to instant recognition by ECOWAS as a player in the conflict. In turn, this occasioned the warring faction’s inclusion in the power-sharing arrangements and their resultant distribution of positions in the transitional governments. This cyclical process contributed to the commitment deficit of the various warring factions with regard to their adherence to the implementation of the various peace accords.

The substantive contents of the Peace Agreements

The overarching weakness of the various pre-Abuja II Peace Agreements was the failure to include effective ‘cost-increasing’ provisions (Mattes and Savun 2009). Although all of the peace accords implored the warring factions to adhere to their provisions, there were no clearly articulated stipulations for the punishment of violators. The major resultant effect was the creation of a ‘culture of impunity’ in which the various warring factions, especially the NPFL, recurrently violated provisions of the various peace accords, including ceasefire, without facing punitive measures from ECOWAS. Hence, since there was no cost for the violation of the provisions of the various peace accords, the warring factions were therefore emboldened to flaunt them.

Two specific cases were instructive. In 1992, the Taylor-led NPFL, after agreeing to abide by the disarmament and demobilisation provisions of the Geneva Peace Accords, refused to allow peacekeepers into the territories under its control (Secretary-General of the United Nations 1994; Adebajo 2002). Consequently, the exercise was stalled. Importantly, since the peace accords did not have effective ‘cost-increasing’ provisions, the NPFL did not face any punishment for this blatant violation of the peace accords. Similarly, on 28 May 1992, six Senegalese peacekeepers were killed by NPFL commandos in Vahun, a town in Lofa County, located in the northern portion of the country, and 500 ECOMOG peacekeepers were thereafter held hostage by the NPFL for one week (Adebajo 2002:609). Characteristically, no punitive measures were taken against the NPFL.


The capacity of ECOWAS to enforce the various pre-Abuja II Peace Agreements was undermined by several factors. As has been argued, the enforcement provisions in these accords were articulated in general terms devoid of the delineation of specific punitive measures for the breach of a provision or provisions. This demonstrated the lack of the required political will on the part of ECOWAS, the then OAU and the UN in holding recalcitrant warlords accountable for prolonging the war and its associated human and material costs.

Another factor related to the limitations of the capabilities of the peacekeeping force (ECOMOG). For example, ECOMOG did not have the troop size that could have enabled it to enforce the provisions of the various peace agreements. Even at its full strength of troops in 1992, including the forces from Senegal and Tanzania, ECOMOG could not effectively induce compliance from the various warring factions whose forces totalled approximately 60 000 (Berdal 1996). In addition, ECOMOG lacked adequate transportation, equipment and other logistics that could have enhanced its capacity to effectively enforce the provisions of the various peace agreements. As well, ECOMOG did not have the requisite intelligence assets that were exigent for effectively monitoring the activities of the various warring factions (Adebajo 2002).

Also, in some of the instances in which ECOMOG undertook peace enforcement measures in order to compel the recalcitrant warring factions, particularly the NPFL, to abide by the provisions of the peace accords, the efforts were undermined by the peacekeeping force’s collaboration with some of the rival warring factions. For example, on 15 October 1992, when the NPFL launched ‘Octopus’, a full-scale military assault on Monrovia, the capital city, ECOMOG enlisted and used the help of ULIMO forces to repel the attack (Adebajo 2002).


ECOWAS, the OAU and the UN’s recurrent practice of pandering to the wishes of the various warlords, even amid their violation of the various peace accords, adversely affected the successful implementation of the various peace accords. This was because the various warring factions, especially the NPFL, knew that they could get virtually what they wanted over and over again. Interestingly, as Dupuy and Detzel (2008:3) poignantly observed, ‘appeasing the warlords did not stop the civil war. But rather appeasement provided the incentives for the warring factions to continue the war thereby creating heavy transaction costs for the continuing peace process as more groups emerged and existing ones continued to fight’.

Specifically, the practice of recurring appeasement was demonstrated in two major ways. The power-sharing arrangements in the various peace accords in effect rewarded the establishment of warring factions. This was done by the allocation of various government ministries, agencies and public corporations to the various warring factions. Moreover, the powerful warring factions like the NPFL were allotted larger percentages of the state entities. In turn, like ‘Tammy Bosses’ (Lowenkopf 1976), the various warlords then appointed the members of their respective factions to these positions. This enabled the warlords to exercise enormous power in the various transitional governments.

Another major set of causes related to ECOWAS and its partners’ unending willingness to virtually scrap peace agreements, once their provisions had been violated by a warring faction or factions. This was then followed by the planning of a new peace accord that was designed to placate the recalcitrant warring faction. Charles Taylor took full advantage of ECOWAS’ willingness to pander to his whims and caprices, as he violated one peace agreement after another. The resultant effect was the continuous process of formulating new peace accords.

The role of third parties

Some of the third parties undermined the successful implementation of the various peace agreements in various ways. From 1990 to mid 1996, ECOWAS, the principal ‘peacemaker’ in the first Liberian civil war, was hamstrung by the rift between the Nigerian-led ‘Anglophone bloc’ and the Ivorian-led ‘Francophone one’. At the core of the conflict was the Francophone bloc’s argument that ECOWAS’ intervention in the Liberian civil war was designed to establish the domination of the Anglophone bloc in the sub-region. Accordingly, the Francophone member states argued that the organisation’s intervention in Liberia was an ‘Anglophone road show’ (de Costa 1990:2699).

The related problem was that some of the member states of ECOWAS were serving as patrons for the various warring factions. For example, Côte d’Ivoire and Burkina Faso supported the NPFL (de Costa 1990; Adebajo 2002). Specifically, Côte d’Ivoire served as the launching pad for NPFL’s insurgency, the conduit for transporting weapons and other materials to the NPFL bases in Liberia, a secured place of temporary residence for officials of the NPFL, and provided the conduit through which the NPFL conducted its external relations. As for Burkina Faso, it served as a conduit for funnelling arms to the NPFL.

Nigeria, the leader of the ECOWAS intervention in Liberia, was viewed as partisan by the NPFL. In part, this was attributable to the cordial relationship between President Ibrahim Babangida of Nigeria and President Samuel Doe of Liberia. Accordingly, from the NPFL’s vantage point, the Nigerian-led intervention was propelled by desire to defend the Doe regime, thereby thwarting the NPFL’s quest to overthrow the Doe regime and seize state power. This mindset informed the NPFL’s participation in the peacemaking processes, and contributed to the warlordist militia’s recalcitrant posture toward the various peace agreements.

The rivalry between the UN Observation Mission (UNOMIL) and ECOMOG also contributed to the commitment deficit that the warring factions, especially the NPFL, demonstrated toward the various post-Cotonou peace agreements. For example, although both UNOMIL and ECOMOG were responsible for the implementation of the security provisions under the Cotonou Peace Accord, UNOMIL failed to consult with ECOMOG, the lead peacekeeping force, in setting up monitoring sites and deploying observers around the country (Kieh 2009:51). This action, among others, conveyed to the various warring factions that UNOMIL neither trusted nor respected ECOMOG (Kieh 2009:51). Thus, by implication, the warring factions could follow suit (Kieh 2009:51). The resultant trust deficit emboldened the warring factions, especially the NPFL, to flaunt the various peace agreements.

5. The Abuja II Peace Accord and the termination of the war

Why was the Abuja II Peace Accord successful in ending the first Liberian civil war? Several factors accounted for this. The change of government in Nigeria, especially the emergence of General Sani Abacha as the new President of Nigeria, after the military coup of 1993, affected the NPFL’s relationship with Nigeria. Specifically, General Abacha was more willing to accommodate Taylor’s presidential ambition than his predecessor General Babangida. In turn, this made Taylor and the NPFL more cooperative and willing to abide by the peace agreement.

Another factor was that the NPFL was under tremendous military pressure, especially from some of its rival warring factions, such as ULIMO-J and ULIMO-K. In some cases, the NPFL lost territory to these rival warlordist militias. So, by adhering to the peace accord, the NPFL was able to ‘buy time’. This enabled it to get new supplies of weapons and to make adjustments in its war-making strategies.

Further, the rapprochement between the Nigerian-led Anglophone and Ivorian-headed Francophone blocs contributed to the erosion of the NPFL’s support from the latter group. Consequently, the Francophone bloc joined the Anglophone one in exerting pressure on the NPFL and the other warring factions to end the war. Being cognizant of the fact that it could no longer take advantage of the split between the two blocs in its continual efforts to undermine the ECOWAS-led peacemaking efforts, the NPFL made the pragmatic calculation of acquiescing to the peace accord. Moreover, the NPFL made the determination that since it could win state power through an election, it therefore needed to cooperate in the termination of the war.

The related point is that the modus vivendi between the two blocs within ECOWAS contributed to increased international support for the conflict management efforts. For example, the United States, which had played a minimal role in helping to end the war, provided leadership in the establishment of the International Contact Group on Liberia. Additionally, the US contributed $40 million for logistics for ECOMOG to use in helping to perform its security functions under the peace accord (Adebajo 2002). Similarly, the European Union contributed 119 trucks to ECOMOG to be used in the performance of its peacekeeping role (Adebajo 2002).

Significantly, unlike the various failed peace accords, the Abuja II Peace Accord was buttressed by the expression of political will by both ECOWAS and the OAU in increasing the cost for the continual violation of the peace agreements by intransigent warring factions. In the case of ECOWAS, it threatened to impose sanctions, including travel restrictions and the freezing of the business activities and accounts of the leaders of recalcitrant warring factions, and their exclusion from participation in the ensuing election (Adebajo 2002). As for the OAU, during its 1996 Summit held in Yaoundé, Cameroon, it threatened to draft and sponsor a resolution that would request the UN to impose sanctions against intransigent warlords, and to establish a war crime tribunal (Adebajo 2002).

6. Conclusion

The previous sixteen peace accords failed to end the first Liberian civil war both for specific and general reasons. In the case of the former, each of the peace accords failed to meet one or more of the criteria outlined in the peace agreement model for successfully ending a civil war. Three of the most recurrent factors were: a) the unacceptability of the substantive contents of the various failed peace agreements dealing with the transitional power arrangements (the NPFL’s central demand having been that Taylor, its leader, should head the interim government, and also be allowed to contest the presidency in the future election); b) the character of the NPFL (its consistent lack of commitment to the peace process as a result of its central demands not being met, and the resulting playing of the role of the ‘spoiler’); and c) the consistent lack of the political will on the part of ECOWAS and other actors in the international community to fully enforce the terms of the various accords.

Generally, the sixteen peace accords failed because of several factors. Among them were the partisan role played by external actors in supporting some of the factions, the rift within ECOWAS, the lack of effective enforcement provisions within the various peace agreements, ECOWAS’ recurrent practice of appeasing the warlords and their factions, and the resultant emergence of a ‘culture of impunity’ among the warring factions vis-í -vis compliance with the various peace accords.

In contradistinction, the Abuja II Peace Accord succeeded in ending the war for several major reasons, including the fact that the Taylor-led NPFL made the determination that against the backdrop of the prevailing circumstances the electoral route provided the best option for the realisation of its optimal goal of making Charles Taylor the President of Liberia. This shift in the NPFL’s position was shaped by the willingness of General Sani Abacha, the President of Nigeria, the primary source of financial, logistical and military support for the ECOWAS-led intervention, to accommodate Taylor’s penchant for the Liberian presidency (Interviews 1998). Another major reason was that the NPFL was under enormous military pressure from rival militias, especially ULIMO-J and ULIMO-K. Furthermore, Burkina Faso and Côte d’Ivoire, the NPFL’s principal supporters in ECOWAS, resolved their conflict with the Nigerian-led bloc that was unwilling to allow the NPFL to come to power through the use of the force of arms. Hence, there was greater unity within ECOWAS, the principal third party in the civil war. Also, ECOWAS developed the political will to fully enforce the peace accord, including the use of punitive means such as the establishment of a war crimes tribunal.

Finally, based on the Liberian case, as per the peace agreement model, there are several valuable lessons that could be learned and applied to peacemaking activities in other countries in Africa and the world both currently and in the future.

First, the warring parties have to be fully committed to the peace process both in words and deeds. And the peacemaker should accord serious attention to this as a central element of assessing the overall characters of the disputants in a civil conflict.

Second, the peacemaker should gauge from time to time the ways in which the dynamics of a conflict as reflected in the various phases of the conflict cycle impact the behaviour of the warring parties, including the ‘spoiler’ phenomenon.

Third, the substantive contents of a peace accord should seek to strike a balance among the critical variables of power sharing, fear reducing and cost increasing. In other words, peace agreements should seek to include all of the warring parties in the transitional power arrangements, allay their fears, especially about security, but explicitly delineate the cost for non-compliance with the provisions.

Fourth, by pursuing contradictory policies, including taking sides with the various warring parties, third parties can undermine the successful implementation of a peace agreement.

Fifth, while appeasement has some utility in terms of helping to ensure the development of commitment to the successful implementation of a peace agreement on the part of all of the parties involved, it must be calibrated and used sparingly so as not to develop the expectation of pander on the part of one or more of the warring parties, thereby contributing to the commitment deficit in the long-term.

Sixth, it is not enough to simply stipulate enforcement mechanisms in a peace accord. Instead, there must be the willingness on the part of a third party or parties to develop and demonstrate the political will to take concrete enforcement actions, if a violation or violations occur.

Finally, the ultimate test of the success of a peace agreement is whether it leads to the end of a civil war, and prevents a relapse into violence thereafter.


  1. Adebajo, Adekeye 2002. Liberia: A warlord’s peace. In: Stedman, Stephen John, Donald Rothchild and Elizabeth M. Cousens eds. Ending civil wars: The implementation of peace agreements. Boulder, CO, Lynne Rienner. pp. 599–630.
  2. Atkinson, Philippa 1997. The war economy of Liberia: A political analysis. RRN Paper 22. London, Overseas Development Institute.
  3. Bell, Christine 2000. Peace agreements and human rights. Oxford, Oxford University Press.
  4. Berdal, Mats 1996. Disarmament and demobilization after the Cold War. Adelphi Paper No. London, ITSS and Oxford University Press.
  5. Beyan, Amos 1991. The American Colonization Society and the creation of the Liberian state. Lanham, MD, University Press of America.
  6. Brown, George 1941. The economic history of Liberia. Washington, D.C., Associated Publishers.
  7. Burrowes, Carl 1982. The settler ruling class thesis in Liberia: Reconsideration. Occasional Paper. Chicago, Self-published.
  8. De Costa, Peter 1990. Confused State. West Africa, 1–7 October, pp. 2699–2701.
  9. Derouen, Karl and Peter Wallensteen 2009. The duration of civil war peace agreements. Conflict Management and Peace Science, 26 (4), pp. 367–387.
  10. Dupuy, Kendra and Julian Detzel 2008. Appeasing the warlords: Power-sharing agreements in Liberia. CSCW Policy Brief. Oslo, Centre for the Study of Civil War.
  11. Francis, Archbishop Michael 1990. Statement to Liberians in Sierra Leone. 7 October.
  12. Gent, Stephen 2007. Rebel strength, commitment problems, civil wars and peace agreements. Paper presented at the Annual Conference of the Midwest Political Science Association. Chicago, IL, 12–15 April.
  13. Hampson, Fen Osler 1996. Nurturing peace: Why peace settlements succeed or fail. Washington, D.C., United States Institute of Peace Press.
  14. Human Rights Watch 1990. Liberia: Human rights development. New York, Human Rights Watch.
  15. Interviews 1998. Conducted with some of the leaders of the various warring factions in the first Liberian civil war and Liberian political parties.
  16. Kieh, George Klay 1997. The crisis of democracy in Liberia. Liberian Studies Journal, 22 (1), pp. 23–29.
  17. Kieh, George Klay 2008. The first Liberian civil war: The crises of underdevelopment. New York, Peter Lang Publishing.
  18. Kieh, George Klay 2009. The United Nations Observation Mission and the first Liberian civil war. Peace Studies Journal, 2 (1), pp. 45–62.
  19. Kieh, George Klay 2011. Liberia: State failure, collapse and reconstitution. Cherry Hill, NJ, Africana Homestead Legacy Publishers.
  20. Lowenkopf, Martin 1976. Liberia: The conservative road to development. Stanford, CA, Hoover Institution of War.
  21. Mattes, Michaela and Bruce Savun 2009. Fostering peace after civil war: Commitment problems and agreement design. International Studies Quarterly, 53 (3), pp. 737–759.
  22. Ministry of Planning and Economic Affairs, Liberia 1986. Economic survey of Liberia. Monrovia, Government Printing Office.
  23. Murshed, S. Mansoob and Philip Verwimp 2008. Enforcing peace agreements through commitment technologies. UNU-WIDER Research Paper No. 2008/45. Helsinki, UNU World Institute for Development and Economic Research.
  24. Mutwol, Julius 2009. Peace agreements and civil wars in Africa. Amherst, NY, Cambria Press.
  25. Peters, May Mercado and Shahla Shapouri 1997. Income inequality and food security. Economic Research Service, November 9, pp. 44–47.
  26. Reno, William 1996. The business of war in Liberia. Current History, 95 (601), pp. 211–215.
  27. Sawyer, Amos 1992. The emergence of autocracy in Liberia: Tragedy and challenge. San Francisco, CA, Institute of Cultural Studies Press.
  28. Secretary-General of the United Nations. 1994. Third Report of the Secretary-General on the United Nations Observer Mission in Liberia. New York, UN Secretariat.
  29. Smith, Robert 1972. The American foreign policy in Liberia, 1822–1971. Monrovia, Providence Publication.
  30. Spears, Ian 2008. Understanding inclusive peace agreements in Africa: The problems of sharing power. Third World Quarterly, 21 (1), pp. 105–118.
  31. Stedman, Stephen John 1997. Spoiler problems in peace processes. International Security, 22 (2), pp. 5–53.
  32. Stedman, Stephen John 2001. Implementing peace agreements in civil wars: Lessons and recommendations for policy-makers. IPA Policy Paper Series on Peace Implementation. New York, International Peace Academy.
  33. United Nations Development Program 1990. Human development report. New York, Oxford University Press.
  34. Walter, Barbara 1997. The crucial barrier to civil war settlement. International Organizations, 51 (3), pp. 335–364.
  35. Walter, Barbara 2002. Committing to peace: The successful settlement of civil wars. Princeton, Princeton University Press.
  36. West Africa 1990. Liberia. West Africa, 8–15 October, pp. 2691 and 2714.
  37. World Health Organization 2003. Liberia: Health/Nutrition Sector report. Brazzaville, Congo, WHO.