What is similar to Rwanda is that there is in Burundi a lot of debate about the nature of key historical trends and concepts as well, starting with the very basic ones: are Hutu, Tutsi and Twa different races? Different ethnic groups, with different historical origins? Or simply different castes, socio-professional organizations that are rather closed off and hereditary but leave some measure of flexibility? Could people change from Hutu to Tutsi and vice versa? Was the king originally Hutu? And the Ganwa?
There were no wars or conflicts between Hutu and Tutsi during these years: this does not mean that equality prevailed or that stereotypes were absent – traditional proverbs clearly show that not to be the case – but that the system displayed a fair degree of legitimacy and was capable of addressing social conflicts.
The colonial period
Germany, briefly, and then Belgium, for four decades, controlled Burundi through indirect rule. The king and his court and administrators continued to run the country, with the colonial authority simply an extra layer on top. Serious reforms of the system were, however, undertaken by Belgium from 1926 to 1933. The Ganwas and Tutsis were seen by the colonial power as the ruling group while the Hutus were naturally destined to obey, and all Hutu authorities were dismissed. As a result, while formally the old political structure of the country remained intact, colonization profoundly altered its nature. Political, social, and economic relations became more rigid, unequal, and biased against Hutu. The power and wealth of the Ganwa particularly, as well as some Tutsi, increased (Prunier 1994). Higher education was rarely extended to the population, and the few who had access to it were, again, powerful Ganwa and Tutsi.
At the same time, the state intervened more heavily in people’s lives. New taxes were introduced, as was mandatory cropping and occasional forced labor for the maintenance of streets and buildings. Some of these measures were ostensibly for the people’s benefit – mandatory manioc cropping, for example, to reduce the risk of famine, or ditch digging to combat water-induced erosion. Others were needed by the colonial state to pay for its upkeep – mandatory coffee production to pay for taxes, for example. A deeply interventionist but low-capacity state that would turn independent on 1 July 1962 was created.
The first few years of independence
In 1958, a nationalist party, UPRONA (Union pour le Progrès National), had been founded by Prince Louis Rwagasore – a popular, modern, pro-independence son of a deposed king, with good links to the Hutu community. In a countermove, the Belgian administration helped create the PDC (Parti Démocrate Chrétien), which was led by chiefs close to Belgium. The Ganwa split between the two parties. UPRONA dominated the 1961 legislative elections, gaining 58 of the 64 seats. The party was truly multi-ethnic: of those elected, 25 were Tutsi, 22 Hutu, 7 Ganwa, and 4 of mixed parentage. Prince Rwagasore was assassinated by agents of the PDC on 13 October 1961. The historic significance of Rwagasore’s murder is enormous: it is truly a day on which doors were closed for Burundi. Note that all this took place against the backdrop of Rwanda’s ‘social revolution’ (1959–62), in which the monarchy was overthrown, thousands of Tutsi lost their lives, and tens of thousands fled the country – including to Burundi. From now on, increasingly, the Rwandan term for demokarasi, referring to ethnic majority politics, would sound appealing to some Burundian Hutu and scary to most Tutsi (Chrétien 2000).
During the next four years, Burundian politics was extremely unstable and gridlocked. The main parties became divided internally, the Hutu–Tutsi division became much more important, government after government fell, extremist positions increased. A failed 1965 Hutu gendarmerie coup d’état led to exemplary retribution, with thousands dead – a pattern that would repeat itself many more times over the next decades. The Tutsi military officer in charge of repressing that operation, Major Michel Micombero, was soon offered a ministerial position in the government. A few months later, in a bloodless coup d’état, he took over the government and declared the First Republic, with himself as president.
This was the start of almost three decades of military rule by a small group of Tutsi-Hima from Bururi province: Michel Micombero (1966–76), Jean-Baptiste Bagaza (1976–87), and Pierre Buyoya (1987–93). Their rule constituted the creation of a low-caste Tutsi dictatorship.
Burundi was dominated by one party – UPRONA. With the party’s women’s and youth movement, all Burundians were theoretically members. There was little separation of power between executive, legislature, judiciary, the single party, and the army. The central clique derived its power from control over the higher echelons of the army, the key levers of the state (and, consequently, aid flows) and party, as well as the small business sector. Dissent was crushed increasingly violently.
The events of 1972 had a profound impact on Burundi’s politics. After an uprising by Hutu and Congolese rebels, during which Tutsi were killed, the army went on a two-month pogrom, systematically killing all educated Hutu throughout the country. At the very least 80,000 were killed (but some estimates are much higher); many more fled the country. Hutu thereafter lived in fear of a repetition of what Lemarchand (1996) has called a ‘selective genocide.’
In the following decades, Burundi developed a system of almost total exclusion of Hutu. By 1985, there were only 4 Hutu cabinet ministers (out of 20), 17 Hutu MPs (out of 65), and 2 Hutu members of UPRONA’s Central Committee (out of 52). Only 1 out of 22 ambassadors was Hutu, and only 2 provincial governors out of 15. Hutu amounted only to 10 percent of the teachers and 20 percent of the students at the National University; 89 percent of public corporation managers were Tutsi. All 37 highest command positions within the army were Tutsi (of which 27 were from Bururi province) (Nkurunziza and Ngaruko 2002). Jackson (2000) notes that just one commune of Bururi province, Mugamba, accounted for 15 percent of the 6,000 students of the University of Burundi, and that in the late 1980s the government allocated about 60 percent of donor aid to education for Bururi. Given that the formal labor market is dominated by public employment and that access depends on education, these government policies meant that the large majority of the population was structurally excluded from advancing. Yet Burundi’s form of apartheid went undiscussed in aid circles or international liberation politics.
The state became further centralized, imposing its controls in all domains of the country’s social, political, and economic life. A plethora of state enterprises was created, allowing for clientelism in job distribution and graft of the proceeds. Corruption became widespread. Resources were drained; land was confiscated through various extralegal means; enormous profits were made by the use of monopolies with import and sales licenses. The state became primarily ‘a milking cow’ for the elites that controlled it (Gasana 2002; Nkurunziza and Ngaruko 2002).
This inefficient and unjust system came increasingly under attack. As the economy stalled and debts mounted, structural adjustment was imposed. Even though only partly implemented, it upset the system and increased political and economic competition among elites and aspiring elites. Internally, in late 1988, Hutu mobs, organized by the Parti pour la Libération du Peuple Hutu (PALIPEHUTU), a clandestine radical movement born in the Tanzanian refugee camps in 1972 and with bases in Rwanda, attacked local Tutsi in the northern communes of Ntega-Marangara (close to where I did my research). Hundreds of Tutsi were killed. The army answered with the usual indiscriminate massive reprisals. At about the same time, the international community – including, importantly, the French, who had provided the main international support for the regime – began talking seriously about democratization.
1993 elections and the beginning of the crisis
In the wake of this crisis, and recognizing the growing international and internal pressures, President Buyoya began a series of important reforms. He assigned twelve Hutu and twelve Tutsi to the National Commission to Study the Question of National Unity. A Charter of National Unity was subsequently ratified. He also created parity in government by assigning many Hutu to senior positions,
including that of prime minister (to make sure, however, Buyoya retained the functions of president of the country and of the party as well as minister of defense; the key ministries of the interior, justice, and the police also remained under Tutsi control, and the entire army top brass remained mono-ethnically Tutsi, with devastating consequences). The atmosphere of inclusion led to the creation in 1992 of the country’s first NGOs, two human rights organizations (SONERA, closely connected to the Tutsi cause, and League ITEKA, which came to be seen as associated with the Hutu cause). A new, multiparty constitution was drafted, with provisions that all parties should be multi-ethnic. The new system of cooptation seemed to be working.
Yet the 1993 elections took place in a climate of growing ethnic antagonism and radicalization. Prunier (1994) suggests that the events in Ntega-Marangara were deliberately provoked by hardliners in the government and the army who wanted to undermine Buyoya’s ‘liberalizing intentions.’ PALIPEHUTU, in turn, was only too glad to oblige, for it too feared it would lose its clout if successful change took place. PALIPEHUTU infiltrations and tracts spread; scary rumors of failed coups, killing plans, and militia were a daily affair. At the same time, the FPR1 was recruiting people throughout the region, including in Burundi; it invaded Rwanda and started a brutal civil war there. The then Rwandan president Habyarimana, in turn, supported PALIPEHUTU. According to some, PALIPEHUTU cadres infiltrated the Front pour la Démocratie au Burundi (FRODEBU, a semi-clandestine organization that originated in 1990) at the local level, and they are responsible for the organized murder of thousands of Tutsi throughout the country after the failed coup d’état of October 1993.
The elections eventually mainly pitted Uprona against FRODEBU, with both parties clearly identified along ethnic lines, even though they were theoretically bi-ethnic (Reyntjens 1995). The 1 June 1993 presidential elections were won by Melchior Ndadaye, the FRODEBU candidate, who received 64.75 percent of the vote, while Buyoya received 32.39 percent. The FRODEBU victory in parliament was even greater. In ethnic terms parliament comprised 85 percent Hutu and 15 percent Tutsi representatives, closely paralleling the supposed ethnic make-up of the country.
Ndadaye took the same cooperative line as Buyoya, appointing several Tutsi to cabinet positions, including as prime minister. Forty percent of the ministers in the new government came from other parties. At lower levels, however, FRODEBU held more posts: all the governors were replaced, with fourteen out of the new sixteen being from FRODEBU. This tendency also existed at communal levels.
After 100 days in power, President Ndadaye, as well as the president and vice-president of the National Assembly (i.e. the full constitutionally described succession), were killed in a coup d’état. It is generally believed that this coup was the counter-reaction to the rapid ‘FRODEBU-ization’ of the middle and lower levels of the state (many Tutsi and UPRONA loyalists lost their jobs in these weeks), and the fear that the army, the prime tool of protection of the Tutsi, would soon follow. The coup itself formally failed a few days later, after an international outcry, bolstered by freezes of aid. Yet the dynamics it had set in motion remained: a constitutional crisis that was to last for years, mass violence throughout the country, and further confirmation for both sides that the other was not to be trusted.
Thousands of Tutsi were killed in the hills in the days after the coup. Pruniers writes that
the first violent acts appear to have been spontaneous and to have been triggered by the news of President Ndadaye’s arrest and death. But quickly FRODEBU local cadres ‘organized the resistance,’ an ambiguous term since in the first days nobody attacked them. In fact, they organized the indiscriminate massacre of ordinary Tutsi peasants who were collectively scapegoated for the murder of the President. Pro-UPRONA Hutu were also massacred along with Tutsi as they were considered ‘accomplices’ of the ‘UPRONA coup.’ Two days later, when the Army moved to stop these killings, it immediately started its own indiscriminate killings of Hutu.
Another specialist, Reyntjens (1995: 16), disagrees, seeing the killings of Tutsi as partly spontaneous popular anger and partly the act of some local FRODEBU politicians. He concludes that ‘there is no evidence that a genocidal plan ever existed, and the allegations that it did were part of a strategy to exonerate the army and to implicate Frodebu.’
For thirty years, political competition in Burundi had become increasingly violent and ethnic in nature: now, the floodgates were open, and civil war had begun. As no side managed to acquire the upper hand, a decade of violence began. The civil war and ensuing genocide in neighboring Rwanda deepened the ethnic dimension. The year 1993 was the defining moment for many Tutsi, who feel that they were victims of a genocide that was only stopped by the (belated) intervention of the army.
A political stalemate followed, which UPRONA, as well as a slew of one-person radical Tutsi parties, used to work their way (back) into government, eventually coming to control the government far beyond what the election results warranted; in Reyntjens’ (ibid.: 16) words, this ‘creeping coup’ consisted of ‘the imposition of a de facto constitutional order which in effect consolidated the achievements of the coup.’ After long negotiations, a new president was chosen in January 1994 – Cyprien Ntyamira (FRODEBU, Hutu). He was killed a few months later in the same plane crash that killed Rwandan president Habyarimana and marked the start of the genocide there.
More arduous negotiations followed, leading to a new convention in October 1994 that gave as many ministerial posts to UPRONA as to FRODEBU. The new government was riddled by infighting and conflict, and incapable of ruling: each side totally distrusted the other and saw its main function as sabotaging any plan of the other side. The country descended into terror. The city of Bujumbura continued to be rocked by extreme violence by mainly Tutsi but also Hutu militias. The years up to 1996 were years of absolute terror for people living in Bujumbura: no urban person has forgotten those terrible days.
In the summer of 1994, Léonard Nyangoma, until recently a FRODEBU interior minister and leader of a militia that controlled Kamenge (a neighborhood in Bujumbura we did part of our research in), took up arms, claiming his party had ceded too much power to UPRONA. This was the birth of the second major armed rebel movement, the Conseil National pour la Défense de la Démocratie (CNDD), after the Front National de Libération (FNL), the armed wing of the PALIPEHUTU. The CNDD eventually repeatedly split. At the end, one wing was led by Nyangoma (which participated in the 2005 elections as CNDD-Nyangoma), the other by Peter Nkurunziza, who would become Burundi’s president, his party known under the original acronym of CNDD/FDD (Forces de Défense de la Démocratie). It is estimated that the FDD had about 18,000 soldiers and the FNL 5,000.
The war and the negotiations
Hutu rebel groups emerged, split, and launched attacks from Tanzania and the Democratic Republic of Congo. The political process was deadlocked. Presidents came and went. Chaos reigned. Violence prevailed. The city of Bujumbura became ethnically cleansed: Tutsis and Hutus lived in separate worlds, cut off from each other. Crossing into a zone of another ethnicity meant risking your life. Thousands fled their homes, either to safer havens around communal headquarters and military garrisons (mainly Tutsis, as they felt safer near the army and the police); or to the hills far from the army; or to neighboring countries, notably Tanzania. Tutsis were summarily executed by the FDD. Hutus were forcibly rounded up by the FAB (Forces Armées Burundaises) in ‘camps de regroupement’ without food and safety. The FNL fired shells indiscriminately from the hills around Bujumbura. Around 300,000 people were killed in Burundi, over 500,000 refugees fled, and another 800,000 were displaced internally, often for many years.
In July 1996, Buyoya launched a second coup, presumably aimed at stabilizing the situation and avoiding a possible international military intervention. A regional embargo was immediately imposed, and years of further fighting, negotiation, and economic impoverishment followed.
The international community from early on sough
t a negotiated solution to the crisis. The USA, the EU, South Africa, Tanzania, the OAU, and the neighboring heads of state all played major roles in these negotiations. Ten summits were held between June 1996 and August 2000 alone. Under enormous international pressure, including some last-minute arm-twisting by South Africa, the Arusha Peace and Reconciliation Agreement was eventually signed on 28 August 2000 (Chrétien 2000; ICG 2004). This agreement and its protocols marked the beginning of the transition out of war toward the development of new institutions designed to support and maintain peace, integrate the army, adopt a new constitution, organize elections, and kick-start development – a tall agenda.
While the Arusha agreement started by saying that the conflict in Burundi was ‘fundamentally political, with extremely important ethnic dimensions; it stems from a struggle by the political class to accede to and/or remain in power,’2 it went on to deal mainly with the ethnic issue. The most relevant sections for our purposes are in Protocol II, entitled ‘Democracy and Good Governance,’ which stipulates that:
• There will be two vice-presidents from different political parties and ethnic groups.
• The government will contain 60% Hutu and 40% Tutsi.
• The same proportion holds in Parliament. In addition, there will be a minimum of 30% women, and three Twa deputies will be coopted.
• No ethnic group may have more than 67% of the positions of communal administrator.
Life After Violence Page 2