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

by Gerard Prunier

  Sendashonga and his Forces de Résistance de la Démocratie had been the voice of the moderate opposition to RPF policies. After over a year in the political wilderness they were just beginning to come to the fore as a possibly credible alternative. Sendashonga was a Hutu who was completely acceptable to many Tutsi, especially the “survivors.” But he was also a seasoned politician, and the RPF hard core knew that any political combination that would accommodate him and his friends would demand a heavy price and could lead to a fundamental redivision of the power structure. His murder in Nairobi on May 16 cleared the way for extremist policies.

  James Kabarebe was in the paradoxical position of heading an army (the FAC), half of which was ready to follow him into a rebellion while the rest wanted him dead. But up to the month of July, when he was eliminated from his position as chief of staff, there had been diverging voices in Kigali about what to do with the growing ALIR threat.82 Outright invasion of the Kivus looked like a case of military overkill, even if many young ambitious officers in the RPA preferred it for the economic opportunities it could offer them. With Sendashonga dead and Kabarebe kicked out of Kinshasa, the RPA hard-liners had a nice international casus belli and a domestically clear field of fire. The invasion-in-the-guise-of-a-rebellion meant carrying out a counterstrike against ALIR and at the same time opening up economic opportunities in the Congo for the young, undereducated, hungry, foreign-born Tutsi who formed the hard core of the RPA officer corps.

  Another factor pushing Rwanda into a Congo war was the infighting within the RPF itself. Kagame and several of his close friends were seen, rightly or wrongly, as the center of what the ordinary Rwandese called “the new akazu,” in parallel with Habyarimana’s corrupt kitchen cabinet. The new akazu had a majority shareholding in the TriStar Company, which had been awarded all the road contracts financed by UNDP and the European Union and seemed to get its cut of all the foreign aid money. Kayumba Nyamwasa, recently promoted to the rank of brigadier general, and Major Nsesibera, the assistant director of the RPA medical services, were spearheading a (possibly self-interested) anticorruption drive which had taken them close to trying a military coup against the akazu.83 A short and successful war would be a nice way out of that tension and an occasion for all to get their share of the spoils. All the more so because Kigali felt that it had the blessing of the U.S. government, a feeling that had been reinforced by President Clinton’s visit in April84 and by the fact that Kigali’s mediation had been solicited by the U.S. State Department in May 1998, when Ethiopia and Eritrea went to war with each other.85 Thus by August 1998 the Rwandese leadership believed, for a variety of reasons, that it had wide diplomatic leeway, even after Kabarebe’s Blitzkrieg attack on the Congolese capital failed. As President Kabila had warned his people, the war was going to be a long one.

  The situation was quite different when seen from Kampala. President Museveni’s spokeswoman Hope Kivengere immediately denied any Ugandan involvement in the Congo,86 and when Beni fell to “the rebels” on August 10 Kampala reiterated its denial. This was of course false. The “rebel” occupation of Bunia three days later involved a UPDF contingent.87 In his careful examination of Kampala’s motives for getting involved in the Congo Professor J. F. Clark considers as “the most plausible explanation” the necessity for Museveni to support his Rwandese ally, saying that if the Rwandese regime had fallen this would have damaged Museveni’s regional prestige and caused him great difficulties, as large numbers of Tutsi refugees would have poured into Uganda.88 This seems like a far-fetched explanation because in August 1998 Paul Kagame’s regime in Rwanda was in no danger of collapsing at all.89 Even if the danger had existed there would have been no need to send thousands of troops to take Kinshasa in order to shore up Rwanda’s security; a simple broad sweep into the Kivus would have been enough. Professor Clark is right in saying that economic and ideological reasons, even if they were present, were not the main reasons to move into the Congo. The main reason, as in the previous war, was Sudan.

  Early in 1997 Museveni had already said, “We have run out of solutions with the Sudan. We are now seeking a solution on the battlefield.”90 Obviously by mid-1998 the solution still had not been found: the UPDF had had to enter Sudan itself on several occasions to strike back at the LRA;91 the ADF guerrillas who got regular arms air drops from Juba92 were wreaking havoc in Bundibugyo (there were 50,000 IDPs by late January and 70,000 by July); and when an ADF agent was caught in Mbarara in March and gave the names of his friends, all were Muslims. Amama Mbabazi, security adviser to President Museveni, declared, “Khartoum’s plan is to destabilize the region to prepare the ground for the spread of Islamic fundamentalism and Arabism.”93 In addition to the various guerrilla groups (ADF, UNRF II, FUNA, LRA), all supported by Sudan, there were another three thousand hostile troops in Garamba National Park, two hundred kilometers west of Arua, who were also occasionally supplied from Juba.94 None of these forces had the capacity to overthrow the regime, but together they kept grinding it down and costing it an inordinate amount of money, which Museveni felt was not only spent largely in vain, but tended to create problems with the donors, who reproached him for an oversize military budget. Although fighting Khartoum directly remained an option, the Ugandan government felt that it was better to leave that task to the SPLA95 and to strike at the wild Congolese northeast, where Sudanese military intelligence operated freely.

  The first strategic move of the UPDF as soon as it went seriously over the border was to head for Garamba and clean it up.96 But does this overriding security concern mean that there was no economic motive in the Ugandan intervention? Hardly. In fact, one strange thing was the fluctuations in Ugandan gold exports around the time of the war.

  Uganda Gold Exports (in U.S. $ millions)

  1995

  1996

  1997

  1998

  23

  60

  105

  19

  Source: Economist Intelligence Unit.

  One notices a rapid increase in gold exports since the first Congo war and then a sharp drop as Uganda went into the second. Three years later, when the UN commissioned a research panel to investigate the looting of natural resources in the Congo, President Museveni said that the increase in gold exports was due to “liberalization of trade,” a not altogether wrong explanation if by this it is understood that wildcat gold-mining products from the Congo were allowed to transit freely though Uganda. The real problem was that from 1995 on, President Museveni started gradually to lose his hold over UPDF finances. In 1991, when the UPDF was 100,000 strong, the cost of keeping these troops was $42 million per year. In 1996, after the demobilization exercise, the cost of only 50,000 troops had climbed to $88 million. Over $400,000 a month was being stolen from the anti-LRA operational budget. In the Congo, where the UPDF had remained on anti-ADF duty with the agreement of Kinshasa, Col. Peter Kerim and his ADC Lt. Col. Napoleon Rutambika were accused of stealing over five million dollars’ worth of goods from Congolese traders, and General Kazini admitted that he had only 6,000 men under his command and not 10,000, as shown in his books.97 When the auditor general James Kahooza produced a report on UPDF finances it read like a catalogue of horrors: a $1.5 million swindle in military equipment customs clearance in Dar-es-Salaam harbor; $1 million vanished from a “special account” opened by Brigadier Kazini; two Mi-24 combat helicopters bought in Belarus for $1.5 million apiece that could never fly; over 30 percent of the T-54 tanks bought in Bulgaria that could not run.98 In many ways the Congo gold flurries were part and parcel of the same thing. Uganda produced very small quantities of gold at its Kaabong mine in Karamoja,99 mined by Branch Energy Uganda Ltd.,100 which denied any responsibility in the massive gold export increases of 1996–1997. It was obvious that since the late 1996 Ugandan involvement in the overthrow of Mobutu, UPDF officers had found ways of quickly enriching themselves in the Congo that were not unique but rather were part of the multifaceted swindle operations the Ugandan a
rmy staff had deftly developed. The 1998 drop did not reflect a decrease in illegal gold mining; it simply reflected a rechanneling of product flows through the new possibilities that opened up with the occupation of Kisangani, a city with a strong attraction for seasoned “behind-the-counter” gold and diamond dealers. This, like a similar appetite among their RPA colleagues, explains why a deep thrust into the DRC was preferred to a limited border security operation. It was not an either/or proposition, security or illegal mining. It was a combination of these, wherein the officer corps in both armies hoped to kill two birds with one stone. But there was a major difference between the RPA and the UPDF concerning the attitude toward what could be called “national interest”: the Rwandese officers were under a strong obligation to surrender a share of their gains to the Ministry of Defense, which had a special Congo Desk to deal with such matters;101 not so in Uganda, where the loot remained in the private hands of perhaps up to two hundred well-connected officers, their civilian friends, and their families. As the darling of the IMF, Uganda lived up to its entrepreneurial reputation.

  The last of Kinshasa’s foes was also the least. Burundi’s FAB collaborated with the UPDF and the RPA, but only on a limited scale, both geographically and in terms of numbers. Burundian forces in the Congo were to fluctuate between one thousand and two thousand, and never went very far away from the Uvira-Baraka-Fizi area directly to the west of Lake Tanganyika’s northern end. This was the shore facing Burundi and one of the entry points for the FDD guerrillas led by Jean-Bosco Ndayikengurukiye, who had supplanted the tamer CNDD of Léonard Nyangoma in early 1998. Bujumbura’s military was not out to overthrow the Kinshasa government, nor was it trying to control gold mines: it was mostly minding its back door. In a paradoxical way, although Burundi never admitted to having troops in the Congo, it was the only one of Kinshasa’s foes that could legitimately claim to be there purely for reasons of local security.

  Fence-sitters and well-wishers

  This was a very large category indeed, regrouping all the governments and ethnic groups whose sympathies or interests linked with one side or the other or, even more incongruously, with both sides simultaneously.

  The epitome of ambivalence was probably Zambia, where, as we will see, President Chiluba was actively promoting his country’s role as a diplomatic go-between while serving at the same time as the main conduit for military supplies going to UNITA.

  As for Jonas Savimbi, he had a long-standing relationship with Zambia. Because he had not been particularly welcome there at the time of Kenneth Kaunda in the 1960s (his attacks on the Benguela railway were a security problem for Zambia), he worked with Kaunda’s enemy, Simon Kapwepwe. Being in opposition to UNIP, Kapwepwe was bound to be hostile to UNIP’s ally, the MPLA. So when Frederick Chiluba swept into power in October 1991 several of his close associates who had won their opponents’ spurs with Kapwepwe happened to also be friends of Jonas Savimbi. The situation got so embarrassing after the signing of the Lusaka Peace Protocol in 1994 that Chiluba used the pretext of a failed mini-putsch in October 1997 to sideline some of the most compromised UNITA supporters. But the changes were largely cosmetic because Chiluba needed those men’s support. Thus Benjamin Mwila, who had been removed from his position as defense minister, was kept in the cabinet as minister of energy. The minister of commerce and industry, Enoch Kavindele, was removed but kept all his connections within the MMD. Vice President Christian Tembo stayed on. As the war went on things grew dangerously tense between Angola and Zambia, forcing Paris and Washington to use their influence on Luanda to stop the MPLA from attacking Ndola Airport in April 1999.102 Chiluba kept denying everything and frantically protesting his country’s innocence. In a way he was right: it was not the Zambian government helping UNITA, or even the MMD; it was men with private business interests and long-standing friendships. But this in itself spoke eloquently of the state’s weakness, when it could not even halt the momentum that could lead to war.103

  Another uneasily positioned fence-sitter was Tanzania. Dar-es-Salaam had sent six hundred military instructors to train Kabila’s fledgling army in May 1997. They were still at the FAC Kamina base when the war broke out and they had to be quickly repatriated.104 But the presence of the Tanzanian People’s Defense Force in Kamina was ambiguous, in that one of their jobs was to train to combat readiness a number of survivors of the Hutu refugee trek across the continent a few months before.105 In a similar vein, when Seth Sendashonga was seriously thinking of opening up an eastern front against the RPF regime in Rwanda, it was in Tanzania that he found a ready welcome.106 It was also a well-known fact that the CNDD and later the FDD recruited freely among the huge Burundian refugee population in Kigoma.107 In spite of all this, Paul Kagame was very careful never to say anything hostile against Tanzania, banking on Mkapa’s pusillanimity, Nyerere’s desire to achieve peace in Burundi, and the general confusion of the Chama cha Mapinduzi (Party of the Revolution), which made it an unlikely candidate for military adventures. He was eventually proved right, and in spite of many near-misses when Tanzania and Burundi looked like they were going to war with each other, which would have led to an eastward extension of the conflict, things always seemed to mend at the last minute.

  A short mention should also be made of Kenya, which remained peripheral to but not uninvolved in the crisis. There were two connections between Nairobi and the exploding Great Lakes crisis: one was the long-standing hostility between Presidents Moi and Museveni:108 the other was the massive Rwandese akazu presence in Nairobi after 1994. Many of the leaders, such as Mrs. Habyarimana herself, her brother Séraphin Rwabukumba, and top Garde Présidentielle officers such as Colonels Nkundiye and Mpiranya, were all in Nairobi, where they bought large houses. They recreated a little akazu culture away from home, and for a while their cash bought a lot of influence in Kenyan political circles. This created a climate in which the KANU government was, from the start, hostile to the rebellion. But later, when Rwanda and Uganda fell out after the August 1999 clashes in Kisangani, Nairobi got on much better with Kigali and discreetly squeezed out the old génocidaire crowd, whose presence was becoming an embarrassment and whose money and therefore influence had dwindled as time went on. This way Kenya eventually edged a bit closer to true neutrality.

  Another quasi-player who never declared its hand until it was unavoidable was Congo-Brazzaville. The Republic of Congo had recently known two periods of armed conflict (in 1993–1994 and again in 1997),109 which had been characterized by the use of “tribal” militias;110 a growing fragmentation of the conflict in which southerners of Niari, Lekoumou, and Bouenza first fought among themselves and with central Pool Province dwellers while the northerners from the Sangha and Likouala later took advantage the southerners’ divisions to eventually “win” the civil war in October 1997; and a central role for the oil money, which had been used by former president Lissouba till the last moment to keep buying weapons and which the new president Sassou-Nguesso was desperately trying to lay his hands on. Sassou-Nguesso started by trying to persuade Elf into paying $600 million in April 1998, arguing that the company had helped Lissouba fight him.111 When this failed he scaled down his demands to $180 million, and then finally settled for an increase in government royalties from 17.5 to 32 percent.

  In fact, the fighting that had restarted sporadically during April 1998, quickened in August, and finally exploded in December was but a continuation of the two previous bouts of civil war. Same actors, same causes, same methods. There was only one difference: this time the fighting moved away from the capital and focused on the “rebel” areas of the Pool, where the Ninja militiamen of Bernard Kolelas had taken refuge, and in the Niari-Bouenza-Lekoumou (Nibolek) region, where Lissouba’s Cocoyes had gone into hiding.112 But Brazzaville was caught in a web of contradictory loyalties and enmities that made it very difficult this time around to contain the conflict within the borders of the Republic of Congo (or simply to understand what was going on).

  • In Apri
l 1997 many fleeing Zairian DSP soldiers crossed into the Congo. They later fought for Sassou-Nguesso during the June–October 1997 war.113 Their presence remained a permanent threat to Kabila.

  • There were 11,000 Rwandese Hutu refugees in three UNHCR camps in the Republic of Congo. They had also fought on the side of Sassou’s Cobra militia, but when fighting broke out in the DRC the men started leaving the camps and crossing the border to join the FAC and to tangle with Kabarebe’s Tutsi.114

  • Sassou was embroiled in a complicated quarrel with Central African Republic President Ange-Félix Patassé because he had given asylum to a French adviser of former president Kolingba (Patassé’s enemy), which had caused Patassé to lend a favorable ear to Lissouba’s demands for support. Patassé himself was worried that he could not keep aloof for long from the fighting in the DRC as the war spread northward into the Equateur Province and Congolese rebels were recruiting ethnically friendly tribal fighters in the Central African Republic.

  • Sassou, who had received the backing of a four-hundred-strong Chadian expeditionary corps, was forced to mediate between Idriss Deby, president of Chad, and Patassé because Deby was angry at Patassé, a Sara by tribe, alleging his support for Sara rebels in Chad who were endangering oil exploration in the Doba region.

  Does the reader at this point want to throw in the towel and give up on the ethnopolitical complexities of the region? I would not blame him, although I can assure him that I am honestly trying to simplify the picture. If we stand back for a moment and try to assess the situation, what do we see?