Africa's World War: Congo, the Rwandan Genocide, and the Making of a Continental Catastrophe

by Gerard Prunier

  The civil war was immediately internationalized when the South Africans decided to intervene and were progressively deployed between November 1975 and January 1976. Encouraged by the U.S. nonrecognition of the new Angolan state and by the South African intervention, the FNLA thought that the time was ripe to take Luanda by storm. But it had overestimated its military strength and underestimated the amount of support brought in by Cuba70 and the USSR. In January 1976 the FNLA was smashed by Russian heavy artillery fire and withdrew in disarray to the Zaire border. Mobutu then sent his troops into the war, but they were also promptly and decisively defeated. For all practical purposes the FNLA then disappeared from the Angolan political equation, leaving no other option for the anti-MPLA actors, the United States and South Africa, but to support Savimbi and his fledgling UNITA.71

  The stance the United States took against the MPLA was straight cold war strategy, but for Pretoria the situation was more complicated. The main motivating factor, apart from opposing the rise of a communist-backed regime in Angola, was to protect South African control over South West Africa. This former German colony had been given to Great Britain as a Mandate Territory after World War I and retroceded by London to its South African colony. But when the 1948 election brought the South African National Party to power, leading to the proclamation of apartheid and the subsequent break between Pretoria and the Commonwealth, it left South West Africa in the hands of a regime that was isolated by international reprobation. Given the mineral riches of South West Africa and its strategic importance as a buffer state against subversion by the African National Congress (ANC), Pretoria was strongly committed to its security. During the 1960s the South West Africa People’s Organization (SWAPO) had worked in close collaboration with the Angolan FNLA and, when he broke with Holden Roberto, with Savimbi’s UNITA.72 But in 1976, given the rise to power of the MPLA, SWAPO had to make a sudden about-face and fight its former UNITA allies in order to be able to stay in Angola at all. The break was particularly cruel because, since it knew UNITA’s internal organization very well, SWAPO could betray it effectively to the MPLA, and many UNITA militants were killed. Savimbi then became a resolute enemy of his former friends and betrayed all the information he had on SWAPO bases to the notorious South African Secret Service, which made good use of it.73 But by 1976 the situation did not look very promising for the South Africans: the FNLA had collapsed, there were now thirty-six thousand Cuban troops in Luanda, and the ill-prepared 1975–1976 campaign into Angola by the South African Defense Force (SADF) had been less than conclusive from a military point of view. The poor SADF showing in Angola brought about active support for SWAPO, under the condition that the Namibian movement would actively fight its erstwhile ally UNITA in the south. The UN then voted a new resolution asking for “the withdrawal of South Africa’s illegal administration of Namibia and the transfer of power to the people of Namibia with the assistance of the United Nations.”74 In Pretoria’s eyes the UN had just become an objective ally of Moscow and Havana and military means were the only answer. At a secret conference in December 1977 the SADF top brass persuaded the South African prime minister John Vorster to move across the border. Vorster’s unenthusiastic endorsement was taken as an absolute green light by the military, who in May 1978 attacked the Kassinga SWAPO refugee camp inside Angola, killing over six hundred civilians. Pretoria’s so-called Total National Strategy policies, outlined in a key 1977 Defense White Paper, recommended nothing less than “coherent military, diplomatic and economic actions aiming at the creation of a South African–dominated cluster of interdependent states in Southern Africa.”75 UNITA had become a part of that grand strategy.

  In the United States the 1980 election of President Ronald Reagan radically modified the rules of the international game when Assistant Secretary of State for Africa Chester Crocker pronounced his famous “linkage” speech.76 making the application of UN Resolution 435 dependant on the withdrawal of Cuban troops. This was an unacceptable approach for Luanda; determining to win militarily at all cost, the MPLA purchased over a billion dollars’ worth of weapons from the Eastern Bloc during 1986–1987. But there were key political changes under way in the Soviet Union and Fidel Castro knew that he might not have too much time left to win some kind of decisive victory before he would be forced to the negotiating table. But if Soviet Perestroika was putting indirect pressure on Havana’s commitment to the MPLA,77 Pretoria was feeling another kind of constraint: the degradation of its economic and financial situation. In 1988 South Africa’s external debt had reached $24 billion, $12 billion of which was due for repayment in 1990–1991.78 In the words of a South African analyst, “Foreign debt repayment may have been one of the single biggest factors putting pressure on South Africa to end its occupation of Namibia.”79 On December 22, 1988, a tripartite UN-guaranteed agreement was signed in New York between Angola, Cuba, and South Africa. SWAPO, whose fate was being decided, was not invited to attend. The Cubans had until July 1, 1991, to go home; during the same period the South African government would comply with UN Resolution 435 and organize free and fair elections in Namibia, with SWAPO participation.80 SWAPO had been handed a victory by the MPLA as a ready-made package, and this after years of being used by Luanda as an auxiliary military force in the war with UNITA. This godfather role toward Namibia was decisive, and it was going to remain a permanent feature of the relationship between the two countries.

  The New York Agreement was a fine piece of diplomatic work. The only problem was that it had nothing to say about the situation in Angola itself, where the war went on as before. After years of harping on Savimbi-the puppet-of-the-South-African-racist-regime the MPLA had begun to believe its own propaganda. The propaganda was partly true; Savimbi had indeed played South Africa’s game. The man had an absolute single-mindedness which enabled him to adapt to any circumstances or to any ally as long as this served his long-term strategy. But he was nobody’s puppet, and the problem of having a Lusophone Creole elite numbering 500,000 at best ruling a country of over 12 million Africans remained. Worse, the war had now turned the MPLA into a subculture of its own, with its own rules, tricks, and arrangements.

  Oil was a big part of the problem. From its humble beginnings in the Cabinda Enclave during the Salazar years, oil had grown to represent Angola’s major economic resource, and it was going to grow even more in future years until its value peaked at 89 percent of government revenue in 1997.

  Angola’s Oil Production (in barrels per day)

  1973

  1988

  1994

  1995

  1999

  172,000

  470,000

  580,000

  680,000

  770,000

  Source: Sonangol, Angola’s national oil corporation.

  The oil economy had grown out of Cabinda and then pervaded the whole Angolan economic landscape. But it employed only ten thousand people and tended to concentrate rather than distribute wealth.81 Between the war and the neglect of productive investments, traditional and even plantation agriculture had fallen by the wayside. A whole candonga (black market) economy had grown around the empty bunkers of “socialist production,” social and health facilities had collapsed, and it was only the Cuban doctors who managed to keep minimum service going; prostitution and delinquency were rife, and the whole public sector ran almost exclusively on kickbacks and payoffs. This did not prevent the favored members of the nomenklatura from living well and from practicing in their private lives the capitalist opposite of the socialist asceticism they were preaching every day in public. The old left-wing ideals of the past had been buried, together with the pretence of nonracialism. The MPLA had become a single party in the best communist tradition: intellectual and press freedoms were unknown, the radio was controlled, there was a single (compliant) trade union, and the Direcião de Inforrnacião de Segurança de Angola (the MPLA’s secret police) was a looming threat in people’s daily life.

  As for Jonas Savimbi,
he had been a tough fighter and an astute politician from his early days with the FNLA, but he had also been a systematic tyrant, a megalomaniac, and a killer. If the MPLA had developed an oil-based nomenklatura, UNITA had built its own candonga economy, in which its garimpeiros (illegal miners) exploited secret diamond mines, mostly in Lunda Norte Province, making up to $600 million or $700 million per year on the world market.82 Often the two candonga families would blend, as when officers of the Forças Armadas Populares de Libertação de Angola (FAPLA) either organized their own garimpeiros networks to sell the diamonds on the black market83 or even dealt directly with UNITA to commercialize the enemies’ gems. Savimbi ruled his outfit with an iron hand, and diamond thieves and dissenters were regularly shot. Violence was not reserved for high-ranking movement members: soldiers were regularly shot for misdemeanors, ordinary peasants were mercilessly taxed and their children press-ganged into the Forças Armadas de Libertação de Angola (FALA). Nevertheless, UNITA not only survived but even thrived. Why? First and foremost there was a basic unspoken reality nobody wanted to acknowledge publicly: the MPLA was not the legitimate representative government of Angola, and this for three reasons: it had refused to abide by the 1975 Alvor Agreement, which provided for a realistic tripartite government, and it was and had remained the political expression of a Creole class of mestiços and assimilado blacks. These were unified by their use of the Portuguese language, their transnational Lusophone culture, and their fear and disgust of the pretos (“blacks”), of the matumbos (boorish bush dwellers), and of the deeply African peasant masses.84 There were plenty of reasons for the international community to remain silent about these facts. It was not politically correct to say that the “progressive” MPLA was a cultural holdover of Portuguese colonization, and moreover there was a form of unconscious racism: the MPLA Creole elite was “more like us,” the whites, the civilized people. They spoke a European language and they talked in intellectual terms. As for Savimbi, he had concentrated in his persona a near-perfect cornucopia of all the contradictory evils of postmodern demonology: Maoism, Salazar’s fascism, South Africa’s apartheid, Ronald Reagan, the CIA, and antiwhite racism, and, last but not least, he had dared to threaten U.S. oil companies in 1992. Short of being a Nazi child molester, it is hard to do worse in terms of political image. But this international image was in complete contradiction to the internal perception of many of the despised matumbos. For them Savimbi was a hero. Never mind that Savimbi himself was a pure product of the assimilado group. He stood for the matumbos, and many matumbos stood for him.

  The December 1988 New York Agreement had taken care of the Namibian problem,85 but not the Angolan one. President Mobutu had organized a major meeting among African heads of state in Gbadolite in June 1989, where Savimbi and José Eduardo dos Santos86 had been able to meet, but the results had been inconclusive. It was only on May 31, 1991, that the two mortal enemies signed a peace agreement at Bicesse in Portugal. The main features of the agreement were the decision to hold free and fair elections within a year and a process of fusion between FAPLA and FALA, setting an unreasonable target of fifty thousand men for the unified Forças Armadas Angolanas.87

  The country lay in ruins. There were almost one million IDPs and another 760,000 refugees in Zaire and Zambia. The 1990 per capita income was 45 percent of its 1974 level, and agricultural production had practically caved in. The financial situation was critical, with the service of the national debt representing roughly 30 percent of the value of exports.88 Military expenditures stood at around 60 percent of the total budget, and the signing of the Bicesse Agreement did not prevent the government from immediately spending another $25 million for military hardware from Spain.89

  After many delays and much mutual bickering elections were set for September 29 and 30, 1992. The UN created a special mission to “verify” them: UNAVEM, the United Nations Angola Verification Mission.90

  The opening up of the Angolan political landscape to multipartism had some interesting consequences. Nobody could be sure of a monopoly any more: two UNIT A generals, N’Zau Puna and Tony Da Costa Fernandes, had defected to start their own party, and several other small parties had appeared, with limited constituencies and a hope of making their voices heard. Savimbi, who was not used to democratic politics, was conducting an unnecessarily aggressive campaign in which he managed in a record few weeks to threaten the foreign oil companies (“We will re-negotiate your contracts”), to scare not only the mestiços but practically all the city dwellers by sounding more matumbo than it seemed possible, and to antagonize UNAVEM members through clumsy militaristic and propagandistic displays. All the nice people wanted him to lose but were scared of what the monster would do if he did. The elections carried out on schedule were of the tropical variety: phantom polling stations, unreliable voters’ lists, lost ballot boxes, and variable geometry results. UNAVEM claimed to have verified that 53.7 percent voted for the MPLA and 34.1 percent for UNITA, with the balance going to the small parties.91 As for the presidential election, it was even more fantastic, with first results giving 51.54 percent to dos Santos and 38.83 percent to Savimbi. This meant that dos Santos was elected president on the first round. On second thought, the results were changed to 49.57 percent against 40.09 percent, and it was decided to organize a second round. Savimbi called the whole process “a masquerade”92 and called for verification. Within days the atmosphere in Luanda became particularly electric. Fearing for his life Savimbi fled the capital on October 8, and everything finally exploded on October 31. The bloody events of that day and night were called by the MPLA “the battle for the cities” and by UNITA “the All Saints Day massacre.” Fifty percent of UNITA’s three hundred soldiers and two thousand cadres in Luanda were killed.93 Symbolically Jeremias Chitunda and Salupeto Pena were among the dead; they were respectively Number 2 and Number 3 of the rebel movement and had negotiated the Bicesse Agreement, which was now coming apart. Within days the war had started again.

  It restarted badly for the MPLA, which immediately began losing ground and responded by panicky massacres of civilians accused of being UNIT A supporters.94 The hysteria reached its peak after UNIT A took the strategic oil town of Soyo on the coast. Because Soyo is close to the Zaire border, the fall of the city was blamed on the Bakongo, and during Luanda’s infamous “bloody Friday” (January 22–23, 1993) over one thousand were slaughtered in the streets of the capital.95

  In April talks restarted between the two warring factions in Abidjan. The tone of the international community was distinctly anti-UNITA: President Clinton threatened to recognize the MPLA government as legal,96 and in June the UN voted Resolution 843. Savimbi had become a kind of tragic Falstaff, the former boon companion of the West during the period of its worst anticommunist excesses whom everybody now wanted to get rid of in order to forget those shameful years. But the brute refused to die and even claimed he had a cause. The country was now cut in two, with one administration (MPLA) in Luanda and another (UNITA) in Bailundo. Deprived of any foreign support and making less money from its diamonds than the MPLA was making from its oil, UNITA started to lose ground. In September 1993 the UN voted a global embargo on UNITA, which drove it to finally sign the Lusaka Peace Protocol on November 20, 1994. But both sides were acting in bad faith, waiting for the other one to make a mistake.

  With over half a million barrels per day the Angolan oil economy had entered the big league, and foreign oil companies were now careful to harmonize their political positioning with their exploration interests. As for American oil companies, they never had any problem dealing with the Angolan Marxist state their government was actively fighting. In Cabinda Cuban troops even stood guard, protecting the Chevron oil wells against possible attacks by U.S.-supported rebels of the Frente de Libertação do Enclave de Cabinda (FLEC). But now Luanda wanted to diversify its weapons supply channels and so started to get closer to the French oil company Elf Aquitaine.97 The balance of trade showed a $3.1 billion surplus for 1995, but de
bt service was $1.6 billion annually and there was $5.66 billion in arrears. So every penny had to be squeezed out of the oil companies, especially since the MPLA was secretly rearming. UNITA was doing the same thing, using its diamond money.

  Estimated UNITA Diamonds Revenue (in U.S. $ millions)

  1992

  1993

  1994

  1995

  1996

  600

  600

  600

  320

  700

  Source: Global Witness, A Rough Trade: The Role of Companies and Governments in the Angolan Conflict, London, 1999, 4, quoting various years of the Economist Intelligence Unit’s Quarterly Reports on Angola. The French diamond expert Olivier Vallée considers these figures to be much too high and puts UNITA’s net yearly diamond income at around $200 million. Professional sources in Antwerp lean toward a $450 million to $500 million yearly figure for 1992–1997.