Peace agreements: The case of Angola

by C. Knudsen, A. Mundt & I.W. Zartman

The pre-colonial Angolan conflict
Past attempts and failures to mediate
The Bicesse accords
Redirection and redefinition at Lusaka
Looking back, moving forward
Conclusion & Notes


Peace has eluded Angola for four decades, and has confounded politicians and policy-makers around the world for years. The multi-generational conflict has seamlessly transformed itself from an independence struggle against Portuguese colonisers, to a well-funded proxy war drawing in both superpowers, and finally, into an even deadlier and more devastating contest for personal power and resources. For Angolans, the tragedy has been overwhelming: more than five hundred thousand have been killed and more than half of the country’s population of 10 million has been displaced by war.

The search for peace and reconciliation in Angola has now stretched for more than 10 years. The thawing of the Cold War in the late 1980s, combined with the military stalemate between Uniao Nacional de Independencia Total de Angola (UNITA) and the Movimento Popular de Liberaçao de Angola (MPLA), as well as the war-weariness among Angolan people, created seemingly favourable conditions for a political settlement. The Bicesse Peace Accords, mediated by Portugal with the assistance of the US and Russia, were signed on May 31, 1991 by representatives of Jonas Savimbi and President Eduardo dos Santos. This memorable occasion officially ended 16 years of civil war. The accords were hailed from Washington to Moscow as a model for post-Cold War peacemaking. US Secretary of State at the time, James Baker III, remarked that the peace offered an ‘opportunity to show the world that a multi-party democracy can be built where before there was destruction and mistrust.’ However, within a year the carefully constructed 63-page peace accord had become a lesson on what to avoid in a new era of peacemaking - the country had entered a new and bloodier phase of conflict.

The Lusaka Accords were signed in November 1994. The signing came after 12 months of negotiations and two years of battlefield carnage, and were an attempt to correct what were assumed to be the fatal flaws of the Bicesse Accords. The Lusaka agreement mandated a flexible demobilisation time frame, provided provisions for power-sharing, and gave the UN sufficient muscle and money to implement the accords. In December 1998, after a tenuous four-year ceasefire, the accords collapsed and the country plunged back into full-scale war.

Against a background of constant conflict and failed peace attempts, this case study will examine the different contexts in which the two accords were constructed, the assumptions and motives of the players, and the lessons that have emerged from each phase of negotiation. Sadly, the Angolan conflict boasts few successes and provides myriad examples of pitfalls in the negotiation process. Conventional wisdom maintains that for a conflict to be susceptible to effective mediation efforts and de-escalation, several elements must be present: the ‘ripe’ moment requires valid spokespersons; a perception of futility in continued violent conflict (a mutually hurting stalemate); requitement; and a way out for each party. These elements, particularly a sense of requitement and a way out for each party after the elections, were absent in the Bicesse Accord. However, the redress of these perceived deficiencies in the subsequent Lusaka agreement, also proved insufficient in deterring the hostility that now engulfs Angola.

Historical perspective is essential in an analysis of this type, since the conflict has spanned several generations: definitions of the underlying causes of conflict have grown less focused and more antagonistic after 37 years of insurgency and civil war. The first section will identify the Angolan civil war as a post-colonial conflict among liberation groups, and a brief history of the move from anti-colonial to civil war upon independence in 1975 will also be provided. The second section will address past attempts to resolve the conflict following the 1988 New York Accords. Attempts to integrate the intra-strate political/ethnic conflict within the broader context of interstate proxy war will also be discussed. The final section will analyse how these attempts, particularly Bicesse and Lusaka, failed in terms of both construction and implementation. It will also examine the theoretical and practical implications of these failures.

The pre-colonial Angolan conflict

The Angolan conflict began as a struggle for independence and national liberation from the Portuguese colonial power during early 1961. However, by the time independence arrived on November 11, 1975, rivalries between the three main nationalist groups had grown into an intra-state war. A total of 15 years of anti-colonial fighting against the Portuguese served to legitimise anti-state politics and guerrilla movements.(1) The new legal state was thus faced with challenges to its legitimacy from those same groups that had fought for its legal existence. The situation was rendered even more difficult than that faced by other newly independent countries, because the Portuguese colonisers abandoned their promise to facilitate elections prior to independence and fled the country, leaving three large, factionalised liberation movements isolated, with little common ground between them. Angolan society was riddled with division long before it launched an anti-colonial movement, with each of the three groups representing different ethnic bases and ideological tenets. The pattern of colonisation emphasised division and differences, and effectively moulded and hardened the latent conflicts between these three independence movements.(2)

MPLA, founded in December 1956 under the direction of Agostinho Neto, called for the formation of a ‘single front of all the anti-imperialist forces in Angola.’(3) Its appeal, however, was primarily limited to the second largest ethnic group - the Mbundu - living in the region around Luanda, and to the Mestiços, who formed the multiracial colonial bourgeoisie. The Mbundu were integrated into Portuguese society more than any other group, and they created the drive and leadership for a nationalist movement.

The Frenta Nacional de Libertaçao de Angola (FNLA) was founded in March 1962 under the leadership of Holden Roberto. Unlike the MPLA’s nationalist goal of ‘Angolity,’ the FNLA’s original objective was the restoration of the ancient Kingdom of Congo in Northern Angola. The party’s main constituency remained the Bakongo people of the north, who were almost exclusively rural and remained largely outside colonial society. However, despite the degree of separation from the state, it was this group that suffered most from the policy of land dispossession in the 1950s.(4) The FLNA tried to expand its constituency by establishing a government in exile (GRAE) in 1962, with Jonas Savimbi as minister of foreign affairs. However, the GRAE was short-lived and by 1963, Savimbi and Roberto were no longer speaking. Savimbi resigned from the GRAE in 1964, amid accusations of ‘tribalism.’ He was accused of creating disunity when he formed his own movement.

UNITA was founded in March 1966 and drew its primary support from the Ovimbundu ethnic group. With the largest population in Angola, the Ovimbundu were well integrated into colonial society, but were also dispersed due to migrant work. This fragmentation largely explains their late entry into the nationalist movement, and UNITA became an internal vehicle for the Ovimbundu group to counterbalance the role of the other two major ethnic groups in the national liberation war.(5)

Beyond the ethnic diversity and division within Angola, the society was also characterised by the absence of intersections and social bridges among these groups. The principle areas where such intersections might have occurred - in schools or cities - were all but muted in a country with high illiteracy and rural isolation. This fractionalisation of society, as well as internal power struggles within the movements, became a determining factor in the liberation movement, and later lead to external support of the competing movements.(6)

Despite years of fighting, none of the groups were able to inflict a military victory against the Portuguese. It was not until the Movement of the Armed Forces coup in Portugal in April 1974, that independence became a real possibility. On January 15, 1975, Roberto, Neto, and Savimbi met with the Portuguese to sign the Alvor Accords, which provided for a transitional government and Angolan independence on November 11, 1975. The agreement was undermined almost immediately by renewed fighting among the three groups. Activity around the capital became so intense that, with the help of Cuban troops, the MPLA was able to claim a decisive victory over the FNLA on November 10 – the day before independence and the day the Portuguese withdrew - and declared an independent People’s Republic of Angola the next day. Thereafter, the struggle ceased to be an anti-colonial one, and was transformed into a civil war with international geo-political implications.

Internal conflict in the immediate post-independence period was perhaps inevitable, given the virulent discord among the three main nationalist factions. However, Angola’s strategic position in southern Africa, and its wealth of natural resources, made it a sought after prize in the Cold War. International interest in, and support of, different factions prolonged and transformed the internal conflict. With its strategy for ideological hegemony in southern Africa, the US entered into an unlikely alliance with apartheid South Africa, supporting UNITA against the Soviet-backed MPLA government. The ramifications of Angola’s civil war gradually spread to neighbouring states. A solution to the Angolan issue could not be found in isolation from the inter-connected problems of the region, and in the early 1980s, no progress could be made within the context of the Cold War.(7) As US covert assistance to UNITA steadily increased, Angola evolved into a pawn of the superpower struggle, and Savimbi emerged as a principled, freedom-fighting hero of American conservatives.

By 1988, the war had reached its peak. The primary combatants were militarily overextended, international sanctions against South Africa’s apartheid state were beginning to take a toll on its capacity to wage war, and superpower interest in funding a deadlocked ideological struggle was waning. The Battle of Cuito Cuanavale in 1988 marked the beginning of the end of the superpower showdown in Angola, when the MPLA and Cuban forces prevailed against UNITA and South African forces. The battle illustrated ‘the miserable fact that intervention had needlessly turned this martyred African country into a senseless, burning relic of the expiring Cold War.’(8) Against this background, the war-weary outsiders arranged a negotiated end to their war in Angola. The quadripartite talks in London in 1988 resulted in an agreement: the New York Accords, which provided for the complete withdrawal of South African and Cuban troops from Angola, expulsion of the African National Congress (ANC) and South West Africa Peoples’ Organisation (SWAPO) from Angola, and independence for Namibia.

While the de-internationalisation of the conflict failed to address the fundamental differences between UNITA and the MPLA, the withdrawal of foreign troops and the looming independence of Namibia created euphoria throughout the region, and momentum to carry forward the peace process. However, the departing international actors left a legacy of exacerbated distrust between the Angolan parties, ideological disharmony, a thoroughly destroyed infrastructure, bitter regional animosity and counterproductive interstate alliances.

Past attempts and failures to mediate

In the aftermath of the New York Accords, several attempts were made at resolving Angola’s internal civil war. Despite four concerted attempts to do so, the desire to have an agreement signed and signed quickly - rather than a desire to address and resolve the causes of conflict - was a primary element of each accord’s failure. The Gbadolite mediation attempt is an excellent case in point: African leaders sought to capitalise on the positive momentum created by the 1988 breakthrough, but failed to reach a viable and enforceable accord.

On June 22, 1989, Zairian President, Mobutu Sese Seko hosted 18 African heads of state at his palace in Gbadolite, Zaire. The meeting produced an official document calling for a ceasefire to begin at midnight, June 24. A committee would be elected to oversee its implementation. This was the first public meeting between Dos Santos and Savimbi, and their public handshake after the signing ceremony was viewed as highly encouraging.(9)

To insure the presence of Savimbi and Dos Santos, Mobutu assured them both that their respective platforms would be the basis for discussions at Gbadolite. However, this was nothing but a sleight of hand, since the two platforms did not and could not coincide. Dos Santos insisted the agenda include an end to US support for UNITA, maintenance of the Angolan constitution and one-party system, and the exile of Savimbi. UNITA demanded the talks focus on defining the terms of a ceasefire, the formation of a government of national unity, and democratic elections.(10)

Agreement on an agenda is a neutral method of building trust between wary parties and, in this instance, could have provided an opportunity to make each party aware of the other’s motivations and underlying interests. Mobutu, on the eve of a visit to the US, was clearly more concerned with improving his international prestige and made no attempt to act as a communicator of this first move toward mediation. In the end, he helped produce a document that was open to widely divergent and ambiguous interpretations, producing no agreement on the underlying issues.

Mobutu wrongly informed the MPLA that UNITA had agreed to recognise the Angolan constitution, integrate their forces into the MPLA and temporarily exile Savimbi. Meanwhile, UNITA was told that the MPLA had accepted its conditions. As soon as each party discovered Mobutu had falsely represented their own and their adversary’s platforms, the agreement broke down. Savimbi denied he had ever acceded to exile: ‘Why are we going to surrender? When we are strong, why should we say I am going into exile?’(11) This emphasis on military strength by the party that had enjoyed the most recent success, remains one of the major obstacles to resolution today, and is both a cause and effect of the conflict’s protracted nature. Although the talks did not produce any immediate results, the meetings did open the door for a series of peace talks between UNITA and the Angolan government. The talks continued into 1991, when military gains created changing positions of strength.

Between 1989 and 1991, several factors changed in Angola’s external and internal environments, which affected the domestic power configurations of both the government and UNITA. A stalemate, perceived as harmful to both parties, was created. It seemed that the ‘ripe’ moment had finally arrived after 15 years of civil war.(12) Political upheaval in Eastern Europe, and the disintegration of the Soviet Union, had a profound impact on the MPLA, which was reflected in the abandonment of Marxist-Leninist principles in July 1990. At its third party congress in December 1990, the MPLA altered the national constitution to permit multiple parties, thereby creating the legal possibility of recognising UNITA. This major concession, brought about primarily by external changes in the international environment rather than consideration for UNITA’s demands, created a more propitious setting for bilateral talks.

In addition, both the MPLA and UNITA were faced with constant food shortages by mid-1990, as crops failed for the fourth season running. Impending national famine created a natural deadline, pushing both parties out of deadlock. When UNITA proposed a ‘corridor of peace’ to transport food to civilians, the MPLA initially refused. They found any UNITA proposal unacceptable, but also suspected the US might use the corridors to transport weapons to UNITA. Eventually, the need for food prompted both parties to cooperate on a ceasefire along designated food transportation routes, but the cooperation only thinly veiled a heightened mistrust between them.

Most importantly, both sides reassessed the possibility of imposing a military solution on the political conflict, following the battle of Mavinga. The government had launched a massive offensive against the UNITA-controlled town in southeastern Angola on December 23, 1989. It quickly gained large areas of territory, yet the MPLA was unable to drive UNITA’s guerrillas far enough from the town, or its airstrip, to use it as a base. The offensive led to a stalemate, with casualties reported in the thousands on both sides. UNITA successfully increased the cost of the war to the MPLA by intensifying its attacks in the north including its Luanda base - while holding a defensive posture in the south. By March 1990, the MPLA realised it could not impose a military solution and UNITA realised that it could not risk another battle of the same magnitude. By recognising a military solution was impossible, both were forced to consider a political solution.

The Bicesse Accords

Building on these changes in perception and environment, Portugal, the US, the Soviet Union and the UN brokered a bilateral accord to end the civil war. Savimbi and dos Santos signed the final agreement on May 31, 1991 in Bicesse, Portugal. The accords called for the formation of a national army, the establishment of government administration over all of Angola, and the institution of a multi-party system. Elections were to be held between September and November 1992. In addition, a Joint Political-Military Commission (JPMC) was to be established and charged with overall responsibility for the peace process, including the ceasefire. Both UNITA and the MPLA agreed to recognise each other’s legitimacy. An interim ceasefire agreement went into effect on May 15, 1991. The MPLA insisted on control of the interim government prior to elections, with the UN Angola Verification Mission II (UNAVEM II) in the supporting role of monitoring electoral registration and voting, disarmament of troops and demobilisation of so

ldiers. UNITA agreed to MPLA authority during the transition, apparently believing it advantageous to avoid the taint of governing a broken country during the period leading up to the elections. At the time, this short-sighted bargain of convenience was hailed by the UN as a bold experiment for future peace-building operations, because the two parties had agreed to police themselves during the interim period.(13)

In theory, the peace accord ended the 16-year civil war, in which more than 300,000 people died. The accords committed UNITA and the MPLA to demobilisation. UNITA was to convert into a political party, presenting its candidates in national presidential and legislative elections. However, the accords did not foresee that UNITA would lose the elections, accuse the government of fraud and resume the war. Neither side complied with the demobilisation. More than 120,000 people were killed in the aftermath of the failed 1992 election, which ushered in a bloodier phase of the war rather than the hoped-for peace.(14)

‘The historic elections were at once peaceful and violent, free and threatened, successful and irrelevant.’(15) The two days of elections, September 29 and 30, were relatively peaceful, and hailed by UNAVEM and other international observers as free and fair. In the presidential race, dos Santos won with 49% over Savimbi’s 33% of votes cast. In the parliamentary vote, the MPLA won an unexpectedly large 55% to UNITA’s 35%.(16) Before the election results were even announced, Savimbi angrily accused the MPLA of fraud and threatened to renew the civil war. A week before the elections, Savimbi ominously told a British television crew that, ‘if I lose, then the elections were rigged and I will send my men back to the bush to fight again. We will not accept defeat.’(17) His refusal to demobilise gave teeth to this threat. Following the election, violence broke out throughout the country as Angola descended back into war.

In retrospect, the accords were flawed in many ways. The weakness of the accords can be explained, in part, by examining why the parties agreed to negotiate. The underlying assumptions and goals of those parties should also be analysed. In addition to the battlefield stalemate, UNITA and the MPLA negotiated at Bicesse purely out of pragmatic interests in gaining or maintaining power, and manipulating the changed international circumstances to their advantage. Despite the lofty language and ambitious intentions of the agreement, neither side were committed to democratic governance or motivated by concerns for the human tragedy that continued to unfold in Angola. Both Savimbi and dos Santos were convinced that elections would validate their respective claims to power, and both were convinced of their ability to prevail in any electoral contest. In this context, elections were a means to consolidate power, not the beginning of an ongoing, inclusive process to facilitate a power-sharing arrangement. The men did not see the elections as a future transfer of power through periodic elections either. All of the participants seemingly endorsed this principle: no provisions or pledges were included in the final agreement for follow-up international observation and verification of future elections. Consequently, the elections took on a zero-sum character, with both sides intent on winning total power.

UNITA and the MPLA understood the importance of international support and legitimacy for their respective causes. To rebuild the shattered economy and devastated infrastructure, dos Santos and Savimbi recognised the need for international assistance and the necessity of courting the commercial powers of the world market, as well as the World Bank and the IMF. Russia and the US, both anxious to clean up one of the last Cold War blunders, mediated an agreement in order to extricate themselves from an unpleasant situation. The agreement was short on specifics and substance, but both countries hoped it would favour their own commercial interests.

Despite the fragility of the accords, no one expected them to unravel so quickly and so completely. According to Deputy Assistant Secretary of State for African Affairs, George Moose,

‘One of the things that was clearly missing from that process was any meaningful discussion among the parties about what happened after the elections: what kind of a government they were going to participate in, what their respective roles were going to be post-elections, and what assurances each would have that, win or lose, they would still have a meaningful voice in governing the country after those elections.’(18)

By relying too heavily on a ‘winner-take-all’ election process, losers were created. Compounding that procedural flaw, there was little doubt on behalf of diplomats and many Angolans, that the populist Savimbi would be victorious in democratic elections, skewing diplomatic calculations and electoral preparations. According to one diplomat, ‘watching the demobilisation, it was clear that Savimbi was cheating more than the government, but if you’re figuring on a Savimbi victory, you keep your fingers crossed and hope it will all work out all right.’(19) However, dos Santos proved better at electoral politics than the charismatic, populist Savimbi. He hired a Brazilian polling and media company that advised him to paint his billboards and brochures in pastels: after 30 years of conflict Angolans, above all else, wanted the peace such colours conveyed. Savimbi assumed that his nation wanted a traditional strongman, tough enough to undo 16 years of communist corruption and economic mismanagement. His messages became increasingly bellicose. The fact that neither side were demobilising according to schedule was ignored. Spokesman for the US Liason Office, Joe Schreiber, said, ‘there was so much euphoria that Angola was going to have its first democratic election, that a lot of us were guilty of not making sure that all the necessary preconditions to an election had been met.’(20)

Although the burden for the post-Bicesse catastrophe lies primarily with the belligerents who refused to honour their agreement, the international community also played a large role. The UN Special Representative, Margaret Anstee, complained that the international community had ‘tried to buy peace on the cheap.’(21) Demobilisation cannot be done inexpensively: the UN budget was less that US$1,100 per soldier, compared to US$7,400 in Namibia, and only one observer for every 333 soldiers, as compared to one for every six in Namibia. Angolan camps were shoddier and weapons were poorly guarded. Both sides kept their forces largely intact, so when UNITA lost the election, they had the means at hand to return to war.(22) Ironically, it was the MPLA that insisted on a weak international monitoring presence out of concern for its own sovereignty.

While the Bicesse Accords were generally hailed as a model for post-Cold War peacemaking, the model has in fact become an example of what to avoid. Fortunately, the lessons have been favourably implemented elsewhere in southern Africa. In Mozambique the UN envoy, Aldo Ajello insisted the elections set by the October 1993 peace accords be postponed because the demobilisation was behind schedule. Angola’s precipitous decline into renewed hostilities was made possible by the refusal of both parties to demobilise their armed forces, and elections were not postponed because of Savimbi’s charges that doing so would favour the government’s position. The UN, in an attempt to avoid the same omissions and errors in implementation, deployed 7,500 UN troops to monitor demobilisation in Mozambique, compared with only 450 unarmed UN military observers in Angola. Also learning from Angola’s mistakes, the ANC and the South African government agreed, in principle, to form a transitional government of national unity for up to five years following their country’s first non-racial election. As one ANC official put it, ‘if you want to know the best argument for not having a winner-take-all election, check out Angola.’(23)

Redirection and redefinition at Lusaka

In 1989, Kenneth W. Grundy wrote: ‘We have a multi-faceted issue [in Angola], involving many actors, many of whom have the capacity to resolve the Angolan puzzle. It is as if we have a complex tumbler lock. Each party has the key. Somehow these parties...must insert and turn their keys, all at the same time. And finally, each expects that the door they unlock will provide benefits commensurate with the risk.’(24) The difficulty of unlocking the solution, however, has been further compounded by the fact that past attempts at resolution left both sides feeling betrayed by the negotiation process.(25)

Following a fresh military stalemate in 1993, the new UN Special Representative, Maitre Alioune Blondin Beye, persuaded the belligerents to talk to each other for the first time in months at secret meetings in Zambia. The government insisted that UNITA withdraw from 60% of the country that it then controlled, and it also insisted that UNITA disarm its civilian supporters. UNITA demanded a ‘serious’ role in central government (rather than the token jobs that it was offered after the September 1992 elections), autonomy for the central highlands, and the disbanding of the government’s riot police. The UN was prepared to offer peacekeeping troops to monitor the ceasefire, disarmament, and the integration of the two armed forces into a national army.(26) However, the groundwork was in place for a potential repeat of past failures: demands for unilateral disarmament, regional secession, and possibly inadequate UN supervision. Perhaps more ominously, battlefield strategy was again dictating the pace and substance of negotiations. Repeating the pattern of the Bicesse negotiations, each party’s willingness to negotiate, and the subsequent presentation of demands, depended on their ever-changing military position in the ongoing war.

Negotiation can begin only when one party decides or perceives that a solution to the conflict may be obtainable through negotiations. (27) At that point, the party attempts to convince the other that a similar possibility of mutual gain also exists through negotiations, building the essential concept of requitement between the two parties. Trust must be strengthened (or built) during the negotiation process, so that each party understands the other’s motives and underlying interests. Efforts should be made to convince each of the other’s good faith, credibility, and reliability. Otherwise, the conflict will again return to crisis the moment the conciliator turns his back, or a new incident arises.(28) The Angolan government was resistant to new peace talks with UNITA, and was reluctantly coaxed back to the table after heavy pressure from the US and Beye. For UNITA, the main reason for cooperating was to avoid the UN-backed international sanctions, which had been threatened if the rebels refused to negotiate. For both, it was perhaps the chance for a momentary break in fighting - a chance to rest and rearm - that was the strongest enticement.

Beye, who officially replaced Margaret Anstee in June 1993, exhibited a firm grasp of past mistakes and a tremendous talent for mediation. He stated that, ‘it is far better to be patient and spend time working on a peace settlement that will last, than to rush through something that will collapse immediately.’(29) By publicly announcing his disinterest in a quick-fix formula, he demonstrated a commitment to the process that was intended to evoke a similar commitment from the hostile parties. Further, he identified his main objective as addressing the need for a ‘climate of mutual confidence’ and imposed a press blackout on all activities in Lusaka, in order to isolate representatives from outside concerns and force interaction among representatives.(30)

A conciliator’s task encompasses four mutually reinforcing and cumulative phases: making contacts; creating or enhancing deadlock; forming a proposal; and implementing the agreement.(31) The first phase of conflict resolution - establishing contacts with parties - is best carried out in concert: if a number of conciliators are available to the parties, and if a number of friends of the conflicting parties can coordinate their good offices and pressure, the chances of success are improved. In Angola, the UN was able to involve the US and Russia, former patron states of UNITA and the MPLA respectively, as well as Portugal, the former colonial power. Both the US and Russia were able to exert a certain amount of leverage on their former clients: they isolated UNITA and imposed limited sanctions against them through the UN Security Council. The international community increased the cost of continued hostilities to Savimbi, thereby making the option of negotiations more attractive.

The second phase entails convincing the parties that deadlock lies at the end of their individual strategies, and that only catastrophe would break the impasse. Successful conflict resolution is possible when the actor’s fear of continued conflict exceeds their fear of settlement.(32) After creating this perception, a deadline can force both parties to alter their preferences from a non-cooperative to a cooperative stance. Boxed in, the parties begin looking for assistance out of their situation, including the new formula/proposal that the mediator can provide.(33) By shifting weight from one party to the other in the conflict, a conciliator can reinforce deadlock and enforce a deadline, particularly if he can reduce support for the party that is momentarily stronger, but not strong enough to impose a unilateral victory. This was the most difficult and longest phase in the Lusaka talks, as positions of strength - measured in military terms as hostilities continued - changed daily. However, by extracting concessions from the stronger party - the government in the case of Angola - talks continued to progress.

The proposal phase is activated when parties have been convinced they need a way out of their stalemate. The formula frames the nature of the solution by defining the terms of trade between the parties, or by establishing a principle applicable to both parties. The formula should appear relatively just and satisfactory to both parties, and should therefore cover the major issues of the conflict. The formula should also include important demands from both sides, since the parties will not give up claims without compensation.(34)

The peace talks between the government and UNITA resumed in Lusaka on November 15, 1993 under the auspices of the UN. Beye indicated he would seek a new formula through which to pursue the new series of negotiations. However, the 1992 elections, accepted as free and fair by the international community, bestowed legitimacy upon the MPLA government, and fundamentally altered the parameters of future negotiations.(35) UNITA was, in effect, in a state of rebellion against a recognised government and starting from scratch was not an option. However, rather than adopting a completely new formula - which can be time consuming and yet crucial to resolution - the Lusaka talks adopted the formula of the Bicesse Accords: some form of government power-sharing in exchange for disarmament. Thus, Beye accepted the areas of the Bicesse agreement that had worked, and targeted the acute areas of failure. In particular, Beye strove for consensus on meaningful power-sharing arrangements, rather than the token participation of the losing party mandated by the Bicesse Accords.

In December 1993, agreement was reached on the general and specific principles, as well as on the modalities relating to all military issues on the agenda: the re-establishment of a ceasefire; the withdrawal and demilitarisation of all UNITA military forces; the disarming of civilians; and completion of the formation of the Angolan armed forces, including demobilisation. Disarmament and military integration issues proved contentious at Lusaka and were rife with loopholes. The MPLA had shifted some 10,000 to 20,000 of its elite troops into the police force, rather than demobilising them during the pre-election period. On the other hand, UNITA perceived this as an area in which non-compliance was probable. In February 1994, agreement was reached on the general principles concerning national reconciliation.(36)

The military situation remained volatile throughout the negotiations in Lusaka. The government and UNITA had continued to purchase large quantities of arms, and had also continued to engage in sporadic, but intense fighting. While the talks often seemed a mere sideshow to the more com-pelling drama of the Angolan battlefields, where thousands were dying each day from war-related causes, hopes nevertheless remained high throughout the region that this time, peace would prevail. On November 20, 1994, the formal signing of the Lusaka Protocol was marred by the fact that Savimbi refused to be present for the ceremony, sending instead UNITA’s Secretary General, Eugenio Manuvakola, to represent him. Pessimism only increased when, immediately following the ceremony, the government launched a large-scale attack against UNITA.

Despite a never-ending stream of crises and disappointments, the Lusaka Protocol enjoyed limited success in its first year. The ceasefire went into effect and was basically upheld, some infrastructure was restored, the looming humanitarian catastrophe was stemmed, and in April 1997 a Government of Unity and National Reconciliation (GURN) was established. A total of 11 UNITA officials were sworn in.(37) However, the country remained physically and psychologically divided. Savimbi maintained an active army in the field, and despite the UN-imposed embargo, arms continued to flow into the UNITA military. Savimbi maintained control of the diamond rich highlands, and industry sources estimated that he was receiving up to US$400 million per year through the black-market sale of diamonds. While Savimbi dragged his feet throughout the implementation phase of the Protocal, the UN resorted to repeated threats of further sanctions against UNITA to move the process forward.

By spring 1997, Savimbi’s failure to demobilise his troops, his failure to allow the extension of state authority to UNITA controlled territories, and his failure to adhere to the agreed-upon timetables and spirit of the Lusaka Accords, led to the eventual imposition of sanctions by the UN. The sanctions neither restored the peace accords in full measure, nor did they push Savimbi entirely out of the process. Instead, both parties limped along, ambiguously committed to the peace process and increasingly distrustful of the each other’s motives. The international community watched helplessly as diamonds and oil poured out of Angola, and arms flooded in at an equal rate. In June 1998, Maitre Beye was killed in a plane crash. By the following December, the Lusaka Protocol was officially dead. UNITA, with a dramatically heightened conventional warfare capacity that made a mockery of the four-year peace effort, attacked the Angolan army with surprising force and quickly gained control of more than half of the
country’s territory.

Looking back, moving forward

Despite a seemingly serious, if sometimes grudging, commitment to the peace process by the combatants, thoughtful and constructive international mediation, and an apparent willingness to learn from and correct past mistakes in the negotiation process by all parties, both the Bicesse and Lusaka Accords collapsed in their implementation. As one observer noted ‘the bottom line of the Angolan conflict is that the MPLA and UNITA have never trusted each other. Hence, all accords, no matter how lofty, have been cosmetic.’(38) However, the extensive and unexpected cooperation of UNITA and MPLA supporters prior to the 1992 elections proved that Angolans, if not their leaders, could in fact work together to rebuild their country.

The war in Angola has ceased to be driven by ethnic hatred or ideological differences. It has descended into a profitable playground fight between a handful of very rich bullies, who show no concern for their exhausted, disenfranchised countrymen. Angola’s minister of defense has publicly admitted that, ‘senior army officers and government officials are profiting from large commissions on weapons purchases. Lower down the line of command, government soldiers are selling fuel, weapons and even uniforms to UNITA.’(39) On the other side, Zambia has emerged as UNITA’s primary point of supply, despite UN sanctions. A Zambian politician, found dealing with UNITA, was asked how he could reconcile his behaviour with his public duties. He responded saying, ‘My public life is as a government official. ‘My private life is as a businessman who seeks to do business.’(40) For many powerful actors in the region, the Angolan war has simply become too lucrative to surrender. This simple fact has posed seemingly insurmountable obstacles to those seeking a resolution to the ongoing conflict.

The failed implementation of both accords - the last and most important phase of the negotiation process - is indicative of the shallowness of the peace agreements. Theoretically, successful negotiations build trust and further momentum through agreement on less contentious issues. Building on this developing trust, the parties then gradually move to the more difficult issues that require serious concessions on both sides. Although many of the shortcomings of the Bicesse Accords were corrected at Lusaka, the most sensitive issues were either stated in very vague terms, or left completely unaddressed. For instance, while the Lusaka Protocol contained an article guaranteeing a ‘special status’ for the president of UNITA, the negotiators felt the issue could only be resolved directly, and privately, between Savimbi and dos Santos.(41) Likewise, the issue of control of the diamond mines - next to oil, the most valuable resource in Angola - was also left unresolved. The richest areas were and and still are under UNITA control, and provide them with the necessary means to maintain their war machine. According to Ambassador Paul Hare, ‘in a strictly legal sense, UNITA was obliged to withdraw its military presence from the diamond areas as part of the quartering and disarming process. However, it was understood that this issue would have to be resolved in direct talks between the two parties.’(42) Therefore, as in the Bicesse Accords, the door was again left wide open for Savimbi not only to assume the role of spoiler, but with the financial capability to do so. The lack of agreement on these two fundamental issues highlights the enduring lack of trust between the two principals, as well as the critical failure of both accords to transform the process from a zero-sum to a win-win situation.

Like the Bicesse Accords, the Lusaka Protocol also mandated unrealistic time frames for implementation. Despite their critical role in bringing the parties to the negotiating table, and in forging an agreement, the UN and troika powers (US, Russia, and Portugal) were unable to enforce the peace they helped create. While the international commitment following the Bicesse Accord has been rightfully characterised as woefully inadequate, the failure of the UN in the wake of Lusaka - despite more than US$1 billion invested in the process - is more complex, and calls into question the ability and will of the member states to build peace.

The UN Security Council, hoping to avoid an expensive, open-ended peacekeeping operation, pushed for a rapid implementation of the Protocol. However, the US Special Envoy reported to Washington that, ‘the peace process will go at a slower rate than what we and the international community want. The two timetables (Angolan and international) are not synchronised.’(43) The failure to implement the agreement according to schedule was due to a profound lack of institutional and organisational capacity within Angola to support and deepen the agreed-upon processes of peace-building. Moreover, the devastated infrastructure often posed overwhelming logistical challenges and prevented the implementation of many physical aspects of the agreement, such as demobilisation and the extension of state administration. As Jakkie Potgieter of the Institute for Strategic Studies points out, ‘the international community just wants them to stop fighting so they can get the diamonds and oil. What Angola really needs is a Marshall Plan for development, infrastructural restoration and to address socio-economic problems.’(44) The arrival of a few hundred UN blue helmets was insufficient to overcome these staggering deficiencies.

The failure of the parties (primarily UNITA) to adhere to the established time frame, forced the UN to respond to the repeated violations of the Protocol, and revealed a greatly diminished capacity to influence the behaviour of the combatants. The limited leverage of the UN, the US and other international actors in compelling the MPLA and UNITA to participate in the negotiation process, all but dissipated in the aftermath of the signing. When Savimbi again assumed the role of spoiler, the UN’s sole tool of forcing compliance was to threaten and eventually impose a series of broad, but largely unenforceable sanctions against UNITA. The effect of such traditional measures was muted by the blossoming black-market for arms and diamonds in southern Africa, and their symbolism was lost entirely on Jonas Savimbi, who remains unaffected and undiminished as a declared international pariah.

Perhaps its not surprising that traditional theories of negotiation fail to fully explain the decidedly non-traditional conflict in Angola. Nor is it surprising that international mediation in a war no longer characterised by ideological differences or ethnic divisions, but rather by personal animosities, corruption and greed, has also failed. ‘This war started before I was born,’ said a 21 year-old army draftee in hiding. ‘Then there was a reason to fight. Now the poor die and the politicians and generals get rich buying weapons, and the war just goes on forever.’(45) The latest transformation of the war, and the collapse of the peace process in Angola, pose two fundamental questions for those seeking to resolve the conflict: firstly, what could the UN and troika powers have done differently? Secondly, what steps, if any, should the international community now take to address the ‘Savimbi problem,’ as well as the looming humanitarian crisis in Angola. Paradoxically, the Angolan conflict illustrates both the need for, as well as the limitations of international involvement.

In retrospect, the post-election conflict in Angola was not ‘ripe’ for negotiation at Lusaka. Despite the appearance of a military stalemate, neither party accepted the legitimacy of the other. Savimbi’s behaviour after the 1992 election revealed his intentions to win power outright, and his actions during the implementation phase of the Lusaka Protocol proved that his commitment to peace was tactical. The only test of sincerity in the negotiating process that the international community has developed for peacefully ending civil wars, is a supervised election.(46) When Savimbi refused to accept his loss in 1992, he failed that test. Savimbi’s perception of the conflict as all-or-nothing. and the subsequent inability of the mediator to marginalise his role in the peace process, imperiled the Lusaka negotiations. However, the international community attempted to gloss over the importance of the players in the process. While UNITA was certainly blamed for a share of the Bicesse Accords failure, the sudden imposition of democracy mandated by the agreement was also deemed responsible for its collapse. The more pragmatic power-sharing provisions of the Lusaka Accord addressed the democracy issue, rather than the obvious insincerity of the actors. The international community remained convinced that Savimbi needed to be brought into the process, rather than marginalised. In 1994, when the MPLA appeared on the brink of crushing UNITA, the UN insisted on the then in-progress negotiations at Lusaka, and convinced dos Santos to withdraw his attack. Military victory by one party over the other is a viable, and sometimes preferable, outcome to a weak negotiated settlement. Given Savimbi’s destructive past behaviour, the MPLA should have been allowed to press their advantage and attempt a total military victory over UNITA at this juncture. Today, officials in Angola concede that they should not have agreed to participate in the negotiations.(47)

With the peace process in shambles, and the failure of international sanctions against UNITA made abundantly and embarrassingly clear by Savimbi’s impressive military strength, what role, if any, should the international community now play in Angola? In the aftermath of Lusaka, trust between the combatants has sunk to irreparable levels and the MPLA has ignored UNITA’s call for new negotiations as disingenuous. Moreover, the Angolan government now says that the UN no longer has any role in the country, ‘in view of its signal failure to demilitarise UNITA, as well as the failure of UN sanctions to prevent UNITA from rebuilding its military capacity.’(48)

The only remaining avenues of international engagement appear to be through commercial interests (legitimate and black-market), or a large-scale, costly military intervention. A military intervention in Angola by the international community is not only highly unlikely, but it would also re-internationalise the conflict in potentially destructive ways, which could result in a Vietnam-like quagmire. Commercially, the international community, through legitimate arrangements using oil reserves as collateral, has allowed the MPLA to mortgage the future of its country to fight the war. Angola’s cash reserves have been entirely depleted and oil production is mortgaged for the next 10 years. On the other side, business interests throughout the world and numerous governments in Eastern Europe have defied UN restrictions and purchased UNITA diamonds, or transferred arms directly to Savimbi’s army. The war will end only when the capacity of both combatants to wage war against the other is reduced, which will require the will and determination of international actors to halt the maddening spiral of arms, oil and diamond trading.

The UN has begun to take a more activist stance on enforcing its sanctions by tackling the black-market and Savimbi’s supply sources head-on. Robert Fowler, the Canadian Ambassador to the UN and head of the sanctions committee, has proposed 14 measures to impede UNITA’s capacity to sell diamonds to finance the war. He has called on all UN member countries with significant intelligence-gathering capabilities to help provide information regarding sanctions violations.(49) A total of two panels will investigate the role of the open markets - and the regulations or lack thereof - that allow illicit diamonds from war zones to enter trading centres unchecked. Fowler wants to place expert monitors to identify UNITA gems in the key diamond trading centres. He also wants to place UN civilian customs monitors at key points in Africa, where UNITA may be moving diamonds or weapons. The routes and amounts of UNITA’s diamonds were well known among dealers until the diamond embargo. Now the diamonds are being picked up before they hit the market, either being bought by trading companies or sold directly to cutters. Fowler also plans to target the arms suppliers. He has visited Kiev to question the government about allegations of Ukrainian arms and mercenaries reaching UNITA in exchange for diamonds.(50) Fowler’s attempts to combat the black-market with technical expertise, whether successful in the long-term or not, does signal a changed perception of the conflict within the UN bureaucracy, as well as the necessary means to manage it.

As the UN blue helmets leave Angola, southern African leaders have seized the initiative to resolve the war, backing their commitment to peace with the threat of military intervention. At the last SADC summit in Mauritius, regional leaders declared Savimbi a war criminal, ‘to be hunted down.’ Another political source remarked, ‘this is where the DRC (Democratic Republic of Congo) pact is very interesting, as it was undersigned by neighbouring countries and represents a new vision for peacekeeping in Africa. When we talk about underwriting an Angolan settlement, it is not just in moral terms, but in physical terms.’(51) Thus, the regional states and the wider international community have arrived at the same conclusion: negotiation alone has been ineffective and, at this point, is impossible. Angola’s war will rage on until Savimbi’s financial and military capacity is diminished.


‘If we don’t make peace this time, Angola will look like the stage at the end of a Shakespeare play - everybody dead.’(52)

The streets of Luanda, Huambo and Uíge are packed with those displaced by the war and thousands are dying everyday of starvation. Weary and hungry, they pose no threat to their leaders: it’s not about them. As oil is pumped and shipped west at record levels, Angola faces a humanitarian crisis of nightmare proportions.

The sad lesson of the Angolan experience is that not all conflicts can be resolved through negotiation. For the untrusting combatants, peace involved costs and risks that neither side seemed willing to bear. For the troika powers, a lingering sense of responsibility for a decade of destructive Cold War intervention, and a desire to tap the natural riches of Angola, did not translate into a workable political settlement. For students of conflict management, guiding elements of negotiation theory were difficult to find and, when apparent, always transitory. And for too many in southern Africa and elsewhere, the war in Angola is extremely profitable.

Peace will arrive in Angola when neither side is capable of waging war against the other. Effectively tackling the flourishing black-market in Africa, controlling the flow of goods in and out of Angola, and encouraging all UN member states to adhere to the embargo against UNITA, will likely prove far more difficult for the international community than all previous attempts at negotiation. However, the 10-year history of negotiation in Angola has proven that there is no way around these obstacles. The tragic alternative would give Shakespeare pause.


1 Zartman, I.W., 1989, Ripe for Resolution (updated edition), Oxford University Press, New York, pp.12-13.

2 For a concise and insightful view of the divisions within Angolan society on the eve of decolonisation, see Zaki Laidi, 1990, The Superpowers and Africa, University of Chicago Press, Chicago, pp.59-65.

3 Marcum, J., 1978, The Angolan Revolution, Volume two, MIT Press, Cambridge, p.30.

4 Malaquias, A.V., Angola in the 1990’s: Prospects for Peace in Swatuk, L.A. and Shaw, T.M., (eds), 1991, Prospects for Peace and Development in Southern Africa in the 1990’s, University Press of America, Lanham, Maryland, p.49.

5 Laidi, Superpowers, pp.60-1.

6 Laidi, Superpowers, p.61.

7 Anstee, M.J., 1996,. Orphan of the Cold War: The Inside Story of the Collapse of the Angolan Peace Process, 1992-93, St. Martin’s Press, New York, p.9.

8 Spikes, D., 1993, Angola and the Politics of Intervention: From Local Bush War to Chronic Crisis in Southern Africa, McFarland & Co. Publishers, Jefferson, N.C. and London, p.322.

9 McCormick, S., Angola: The Road to Peace, CSIS Notes, No.125, June 6, 1991, p.3.

10 Ibid.

11 Ibid.

12 For a concise discussion of the changes in the international and national environment at this time, see McCormick, Angola: Road to Peace, pp.8-11.

13 Ohlson, T. and Stedman, S.J., 1994, The New is Not Yet Born: Conflict Resolution in Southern Africa, The Brookings Institution, Washington DC, p.109.

14 Taylor, Seeking Peace.

15 Meldrum, A., 1992, Hungry to Vote, Africa Report, (November-December) pp.26-30.

16 Ibid.

17 Ibid.

18 Interview with Moose, G., reported in Africa Report, January/February 1994.

19 Quoted in McCormick, Angola: Road to Peace, p.4.

20 Taylor, Seeking Peace.

21 Anstee’s comments are reported in Taylor, Seeking Peace.

22 The Economist, March 19, 1994, p.48.

23 Taylor, Seeking Peace. p.A-16.

24 Grundy, K.W., The Angolan Puzzle in Onwuka, R. and Shaw, T., (eds), 1989, Africa in World Politics: Into the 1990’s, St. Martin’s Press, New York, pp.202-203.

25 Moose, G., gives an insightful interview into the problems of negotiating in Angola, Africa Report, January/February 1994, pp.24-25.

26 Why Angola’s Rebels will fight on. And on, The Economist, November 27, 1993, pp.42-43.

27 Saunders, H., 1985, The Other Walls: The Politics of the Arab-Israeli Peace Process, American Enterprise Institute. Washington DC.

28 Zartman, Ripe for Resolution, p.274.

29 Quoted in Africa Report, March/April, 1994, p.56.

30 Ibid.

31 Zartman, Ripe for Resolution, p.275

32 Ohlson, T., and Stedman, S.J., The New is Not Yet Born, p.117.

33 Zartman, Ripe for Resolution, p.276.

34 Ibid. p.278.

35 Hare, P., 1998, Angola’s Last best Chance for Peace: An Insider’s Account of the Peace Process, US Institute for Peace, Washington DC, p.131.

36 UN Document S/1994/282.

37 Tvedten, I., 1992, US Policy Towards Angola Since 1975, The Journal of Modern African Studies, Vol.30, p.97.

38 On the Angolan War (Editorial), Times of Zambia, July 17, 1998.

39 Angola: Profiteers’ war that goes on forever, Mail and Guardian, (Johannesburg) July 2, 1999.

40 Edwards, S., March 10, 1999, Mission to stop Angola’s Diamond War, Toronto National Post, p.A-12.

41 Hare, P., 1998, Angola’s last best chance for Peace: An Insider’s Account of the Peace Process, United States Institute for Peace. Washinton DC, p.107.

42 Ibid.

43 Ibid. p.103.

44 Amupadhi, T., UNITA rift could ignite Angolan war, Mail and Guardian, (Johannesburg), April 10, 1998.

45 Angola: Profiteers’ war that goes on forever, Mail and Guardian, (Johannesburg), July 2, 1999.

46 Ohlson, T. and Stedman, S.J., 1994, The New is not yet Born: Conflict Resolution in Southern Africa, p.124.

47 Gordon, C., Fierce clashes as Angola unravels, Mail and Guardian, (Johannesburg), December 4, 1998.

48 Angola experts astonished at UNITA strength, FBIS (sub-Saharan Africa), December 12, 1998.

49 Machona, G., UN links neighbours to UNITA’s arms supply, Post of Zambia, June 9, 1999.

50 Gordon, C., UNITA still open for Business in Diamonds and War, Mail and Guardian, (Johannesburg), July 16, 1999.

51 SADC to underwrite a peace deal, UN Integrated Regional Information Network (IRIN), July 14, 1999.

52 Vaz da Conceicao, A., spokesman for President dos Santos, quoted in the New York Times, ‘Will Angolan Rebel Give Peace a Chance, February 14, 1995.