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

by Gerard Prunier

  • A core conflict in which the new Rwandese RPF regime was trying, with Ugandan help, to overthrow their rebellious puppet Kabila.

  • A second layer of powerful players (Angola and Zimbabwe, with Namibia riding shotgun) who could not care less about the Tutsi-Hutu conflict or the Uganda-Sudan confrontation but who, for their own diverse reasons, wanted Kabila to stay in power, albeit as a puppet.

  • A third layer of actors (Libya, Chad, the Sudan) who felt they had to get involved for reasons that had nothing to do with the Congo itself but had to do largely with each other and with their indirect relationship to the core conflict players.

  • A fourth layer of countries that were peripherally involved because of geography and because of other entanglements with countries not themselves neutral in the DRC war. This ranged from Burundi, which committed troops to the war (without ever admitting it),115 to the Central African Republic, which desperately tried to resist being dragged in.116

  This leaves out one major player, which, although it could not be described as the hidden puppeteer, was nonetheless the major outside presence in a conflict it was almost fated to win without ever having to get involved in fighting it. This essential actor was the Republic of South Africa.

  South Africa had no direct or indirect security concerns in the Congo; its economic stake there was peripheral to its economic core; it had no political entanglements with anybody involved in the conflict that would have steered it, against its better judgment, in one direction or the other. It never sent troops to any of the areas at war.117 Its diplomatic role was modest.118 And its real understanding of the situation remained limited.119 But South Africa is a very heavy player in a very lightweight environment. Since colonial times South African mining and transport industries have put down very deep roots in the Katanga region of the Congo Free State and later of the Belgian Congo.120 Apartheid isolationism and Mobutu’s economic mismanagement partly (but never completely) prised them apart. By 1997 both had disappeared, and economic logic had reasserted itself. Most of South Africa’s actions regarding the Congolese situation were economically motivated.

  • Pretoria was engaged in an implicit policy of economic expansion toward the whole of southern, eastern, and central Africa: plans by Trans Africa Railway Corporation to build a rail link to Kenya by way of Morogoro in Tanzania;121 plans to buy the faltering Uganda Airways and Air Tanzania Corporation as well as shares of Kenya Airways; Eskom attempts at indirectly controlling the sick Congolese electricity parastatal Société Nationale d’Electricité.122 and everywhere massive exports of South African goods and services. From that point of view Zimbabwe was a rival whose manipulation of the SADC was a constant irritant to Pretoria.

  • Gold was at $300 an ounce in 1998, a rather listless price level: prolonged marginalization of the DRC could only be a good thing from South Africa’s point of view since Kinshasa was a potential spoiler in the world gold market.

  • The “rebels” were short of funds and desperate to turn Congolese assets into a rapid cash flow. Their victory was likely to further integrate Katanga’s mining economy into the structures of its dominant southern neighbor.

  All available indicators pointed to the logic of South African support for the rebels. But up to a point: South Africa’s army was not the preferred means of intervention. Pretoria had, and rightly so, much more confidence in its economic clout and in its diplomatic capacity than in its rather inefficient armed forces. South African diplomacy was clumsy and ill-informed, but it had two very powerful trump cards: the guilt accumulated in Western countries by their long toleration of apartheid during the cold war and the eagerness of these same economically developed countries to enter a promising South African market. Both could be mined efficiently to ensure that the not-so-neutral “international community” would never stray too far from whatever positions Pretoria would take.

  Fighting down to a stalemate

  All the actors in the “Congolese” conflicts met at Victoria Falls in Zimbabwe on September 7, 1998, once the heat of Kabarebe’s Blitzkrieg had died down. But the invaders’ failure to take the capital had left an indecisive situation in which it was premature to expect any kind of settlement. As we will see, this did not discourage the belligerents from practicing with abandon Mao Zedong’s old strategic dictum “Negotiate while you fight.” According to this pattern negotiations were usually preceded, accompanied, and followed by the most extraordinary lies and exaggerations. This started right away when FAA Chief-of-Staff João de Matos declared before the Victoria Falls Conference, “Operations have been successfully carried out, the war will end pretty soon.”123 To which Modeste Rutabahirwa, chargé d’affaire at the Rwandese Embassy in Paris, retorted more realistically, “The Congo war is with us and it could last for two, four or even ten years.”124

  The Victoria Falls meeting was in itself a perfect example of deceitful pseudo-diplomacy: General Kagame denied that he had any troops in the DRC, while President Museveni said he had “fifty-one intelligence agents with the rebels and two battalions near our border for self-protection.”125 As for the cease-fire signed on September 7 (there were many more to come), it was broken the next day, when an Angolan Antonov An-12 flying from Kindu bombed Kalemie, killing twenty-five civilians.126 Everybody went home then, and the war could really start.

  Given the discomfiture of the Blitzkrieg force,127 the next best thing the “rebels” could do was consolidate their control of the east. From their bases in North and South Kivu they could move in two directions: north-northwest to occupy Province Orientale and south-southwest to enter northern Katanga.128 On the first front the still very thin RCD was to be supported by the UPDF, while the second front was largely an RPA operation, with some UPDF support in terms of artillery and tanks. Isiro was attacked in late August, and Kabila, who did not have the means of defending the northeast, flew to Ndjamena and then on to Tripoli, in defiance of the UN embargo, to request help from his allies. On September 18 the first Libyan planes started to ferry a one-thousand-strong Chadian contingent to the embattled north. But they were flown first to Gbadolite, and rallying the east from there was to prove a slow and difficult business. Kabila had also asked Central African Republic President Patassé and President Sassou Nguesso of the Brazzaville Congo for the Hutu refugees who were living on their territories. The U.S. State Department immediately sent a telegram to its Geneva mission to ask the UNHCR to stop the move at all cost.129 On September 25 a combined RCD-UPDF-SPLA force occupied Dungu. The Sudanese rebels immediately started looting the town, going as far as dismantling the local power station to take it to Yambio in the Sudan.130 They also corralled forty-six thousand Sudanese refugees from the local camps and pushed them back into western Equatoria in a rather rough way. From Dungu they moved into the Garamba National Park to scatter the Ugandan guerrillas hiding there and then went on to take Isiro on October 4. The Chadian contingent never made it to the battlefield, but it soon became involved in the fighting anyway because, even if the Chadians did not manage to move eastward, the enemy was now coming westward. Buta, on the edge of Equateur Province, had been taken on the same day as Isiro, and it soon became evident that the force moving into the province had a new component. On November 7 Jean-Pierre Bemba, a son of Mobutu’s former crony and president of the Zairian Chamber of Commerce Bemba Saolona, announced the existence of a new rebel movement, the Mouvement de Libération du Congo (MLC). His approach, as he explained to a French journalist, could be characterized as “empirical”: “I had identified the possibility of launching an armed movement. So I went looking for serious partners. There were two countries in the region which were interested but I chose to present my dossier to the Ugandans.131 They liked it and so I went in.”132 When the interviewer asked him whether he had sought the support of rich former Mobutists, he replied with a laugh, “If they want to invest, now is the time. When I get to Kinshasa they’ll have to queue up to reach my office.”

  In fact this was sligh
tly disingenuous. Bemba the younger is a Ngbaka by tribe, that is, a member of the larger Bangala group. As such he could not but be a champion of the north, and if the “rich Mobutists” had not yet flocked to his standard it was only due to diplomatic precautions and also to some doubts about the feasibility of a young Belgian-educated playboy turning overnight into an effective African guerrilla leader.133 But for the time being he was not doing badly. His Ugandan patrons had flown him into Kisangani on September 29, and from there he had loosely supervised battles fought by the Ugandan army on his behalf since his own movement was still too small to make a serious impact at the military level. The Chadian contingent, which had finally reached the front line, got whipped at Aketi and then at Bondo, losing over 70 men and 120 POWs.134 A Ugandan helicopter flew Bemba into Aketi as soon as it was taken so that he could show himself and rally potential supporters. He got a rousing welcome and, a few days later, in the company of the Ugandans, his fledgling forces ambushed the Chadians again near Buta, killing 122 and capturing 148.135

  As Bemba was driving deeper into Equateur the SPLA was finishing its housecleaning in the northeast, bringing ever greater numbers of former Sudanese refugees back into Yambio. The Chadians were frustrated at their poor showing and denied losing more than 250 casualties and over 400 POWs in a few weeks.136 On November 17 the joint MLC-UPDF forces captured Bumba, 320 kilometers northwest of Kisangani, driving deep into the north. In the next few days the MLC was to recruit over a thousand young boys, all eager to fight for the new Bangala hero.137 Bemba had earned his spurs, and he flew into Kampala two days later to meet Wamba dia Wamba, Museveni’s man at the helm of the RCD. The reason for that meeting was that problems were already beginning to develop inside the rebel camp.

  The RCD had never been popular, largely because it was seen as a tool of Rwanda. The preexisting ethnic tensions were brought to a boiling point by the “rebellion,” which all the autochthonous tribes perceived as a Rwandese invasion. In late August in Uvira the local populace killed approximately 250 Banyamulenge as the RCD approached the city, and in a reprisal action the RCD massacred over 600 people (Bavira and Bafulero) at Kasika. The RCD found it difficult to recruit local collaborators, and those they could induce to join up were usually disliked by the population. The first RCD governor of South Kivu, Jean-Charles Magabe, was sacked in mid-October and replaced by his more pliable deputy, Benjamin Serukiza. Magabe fled to Brussels, where he passed very severe judgment on the “rebels”: “We do not think these people will ever bring any form of democracy to the Congo. I can’t see such a minority group fighting for democracy because if there were elections they have almost no chance of getting their people elected… . They just use the word democracy… . I could not support massacres. I could not accept that while pretending to fight for democracy we would install the rule of a tribe or rather of a sub-ethnic group.”138

  There were rumors within the RCD that its president, Ernest Wamba dia Wamba, was not far from sharing such politically incorrect feelings. The fact that Wamba was a Ugandan protégé created tensions between Kampala and Kigali in their management of the war. The creation of the MLC, with obvious Ugandan support, only made matters worse. Jean-Pierre Ondekane, the RCD military leader, tried to keep up the pretence that the MLC was under his command and even claimed the taking of Bumba as his own victory. The Kampala meeting was supposed to bring all the “rebels” and their sponsors together; it resulted in the proclamation of a common RPA-UPDF Military Command in the Congo. This was to remain a purely paper decision, with no effect on the ground.

  Meanwhile, Rwandese forces, with some UPDF support, had pushed for a deepening of the southern front. The first step was the taking of Kindu, the capital of Maniema, in mid-October after a no-holds-barred bloody battle139 in which many of the combatants were non-Congolese. Just as it was mostly Ugandans fighting Chadians in the north, in Kindu many of the five-thousand-strong “Congolese” garrison was made up of Rwandese ex-FAR and Interahamwe and Ugandan guerrillas flown in from Juba by the Sudanese army, while the assailants were largely Rwandese and Ugandans. “After the fall of Kindu, roads are now open for the conquest of the entire country,” declared RCD Commander Jean-Pierre Ondekane. The “rebel” forces started immediately to push south, potentially a very dangerous move for the Kinshasa regime. Between November 18 and 22 Kabila used his allies’ air transport capacity140 to fly eight thousand ex-FAR and Interahamwe to Lodja (418 kilometers west of Kindu) in the hope of stemming the rebel advance, while the Zimbabwean National Army (ZNA) quickly deployed two thousand more troops in Lubao and Kabinda. What they were trying to protect was the obvious next target for the “rebels”: Mbuji-Mayi and its diamond mines, the only cash-producing territory the government had left under its control. Kabalo had fallen to the RCD on October 20 and Kongolo had been taken on November 10. From there the Rwando-rebel forces quickly extended their control over northern Katanga, taking Moba, Pepa, Manono, and Kasange. Soon the fighting was down to Pweto on the Zambian border.141 Further west as well the situation kept deteriorating for the government camp, with Lubao falling on January 27. By March the rebels had pushed to Kabinda, which was under siege.142 The Rwando-rebel forces pushed desperately because the big diamond prize was nearly theirs. By late March they had gained partial control of Kabinda, where the Zimbabweans were putting up a spirited defense. Fifteen thousand refugees had fled to the Kaputa refugee camp in Zambia, many of them with battle wounds.143 Since Kabinda was proving such a hard nut to crack, the Rwando-RCD forces tried to go around it, occupying Lusambo on the Sankuru River on June 15 to threaten Mbuji-Mayi from the north. Zimbabwe deployed another three thousand troops in June and Rwanda seven thousand more. Peace talks had started in Zambia, and both sides were hoping for military successes before international pressure proved irresistible. In the north Bemba was doing the same thing: he had captured Gemena on December 30, causing panic in Kinshasa since in his case it was not the diamond mines of Kasai that were at stake, but the capital itself.

  In the first days of January 1999 Namibian Boeings and Angolan Tupolev transport planes brought one thousand FAC soldiers to Bangui in the Central African Republic to mount a counteroffensive in Equateur. The Sudanese joined in with daily bombings by Antonov An-12 from Juba. The MLC-UPDF forces could not withstand the pressure and withdrew to Lisala after losing Gemena, Businga, and Libenge. Bemba remarked in disgust, “We are fighting a mixture of Sudanese, Chadians, Interahamwe and Central Africans; there are only very few Congolese among the FAC.”144 He conveniently forgot to mention that many of his own men had been recruited in the Central African Republic145 and that the Ugandan army made up most of his battle corps. In February the “rebels” organized the only combined offensive they were going to be able to manage during the whole conflict. Over sixty thousand men were involved on the “rebel” side if one added up the RCD, the RPA, the MLC, and the UPDF.146 In the north the targets were Gbadolite, Mbandaka, and down to Kinshasa. In the southeast they were Mbuji-Mayi, Kamina, and Lubumbashi. In case these military plans succeeded only partially, the partition of the country would follow. Bemba went on the offensive and took Ango. The FAC fled across the river into the Central African Republic town of Zemio, where Libyan aircraft came to fly them back to Kinshasa. On their own front the Rwando-RCD forces had widened their control of northern Katanga by taking Kaputo and Kasiki in mid-March, all the while desperately pushing around Kabinda. The ZNA counterattacked too soon and lost over two hundred men at the battle of Eshimba.147 But the offensive did not have the means to carry on and it progressively lost momentum.

  Meanwhile the rivalry between the Rwanda-sponsored RCD and the Uganda-sponsored MLC was growing ever sharper. When Bemba’s forces took Kateke and Bondo in mid-April he was careful to point out that “these successes belong to MLC troops, they owe nothing to the RCD.”148 But just like his rivals in Kasai and Katanga he was hoping to achieve decisive victories before the peace talks now under way in Zambia could freeze the military
situation. On July 3 he took Gbadolite, Mobutu’s former “jungle capital,” and then in quick succession kicked out the FAC from Gemena, Bokungu, and Zongo. He now controlled practically the whole of Equateur Province and, contrary to his RCD rivals, without any tension with the civilians. Both he and his movement were popular among the Bangala, who were hoping for a return of the “good old days” of Mobutist northern domination while the relatively disciplined behavior of the Ugandan soldiers also helped when compared to the constant violence of the Rwandese RPA in its own zone of operation. But Bemba could not make it to Kinshasa, at least not with his present strength and not unless the UPDF mounted an all-out offensive to support his move downriver. As for the Rwando-RCD forces, they were similarly stuck in northern Kasai and northern Katanga, where fighting was still violent around Kabinda but where the ZNA blocked all progress toward Mbuji-Mayi. The diamond capital had received large reinforcements of FAC and ZNA and its entire periphery was by now heavily mined. The “rebels” had long hoped that UNITA would cross the border and intervene in the war to break the deadlock. Already in October, at the time of the fighting for Kindu, UNITA forces had started from Tchikapa toward Mbuji-Mayi to coordinate with the Rwando-Ugandan push. But they had to turn back due to Savimbi’s needs on other battlefronts. On March 25 Savimbi had met General Kagame, Bizima Karaha, and former Republic of Congo president Pascal Lissouba in Ouagadougou at the invitation of the great west African prestidigitator Blaise Compaore. The hope was to get UNIT A to throw its weight into the “Congolese” war as a step to further operations (and victory) in Angola proper. Savimbi thought the plan too complicated and refused to go along with it, in spite of two hours of intense discussions.149 Without UNITA support (and possibly even with that) the Rwando-Ugandan troops could not swallow the huge chunk of the continent they had bitten. The situation was not that of early 1997, when the AFDL and its continentwide supporters could walk clear across Zaire, practically without opposition. The FAC, weak as they were, were trying their best, contrary to the FAZ two years before. And contrary to the FAZ, they had reasonably strong allies. By mid-April Ondekane could still declare defiantly, “The objective remains the liberation of the whole country and we are pushing on towards Kinshasa,” but it sounded increasingly like whistling in the dark. The time had come for “peace talks.”