Dr Andy Storey lectures in the Centre for Development Studies, School of Politics and International Relations, University College Dublin. He expresses his thanks to Jennifer Todd for helpful comments on an earlier version of this paper.
The 1994 Rwandan genocide occurred despite the existence of a peace and power sharing agreement (the Arusha Accords) to which all parties to the conflict had ostensibly subscribed. This paper addresses the failings of the Arusha peace and power sharing process and makes three core arguments. The first argument is that the Arusha process was more a part of the problem than it was part of any putative solution because it heightened tensions within élite circles (whose monopoly of state power was seriously challenged) and provided a channel through which aspirant élites could pursue their dangerous goals. Even more fundamentally, the Arusha process failed to tackle the most pressing problems of Rwandan society, including chronic and worsening poverty and the oppressive presence of the state in all aspects of social life. This disastrous cocktail – creating what Uvin (1998) calls a situation of ‘structural violence’ – laid the basis for mass participation in the genocide of 1994. Far from helping solve these problems, certain international interventions – especially economic ‘structural adjustment’ that ran parallel to the Arusha negotiations – worsened the situation. The Arusha Accords also therefore failed, and this is the second core argument, because they neglected (or worsened) the structural conditions of life for the vast bulk of ordinary Rwandans. The concluding section of the paper examines post-genocide Rwanda and how the legacy of the Arusha Accords has, amongst other devices, been used to legitimise new forms of repression at the same time as the abuse and violence inflicted upon ordinary Rwandans (and their neighbours) have continued. Again, and this is the third core argument of the paper, a seemingly reasonable political agreement to share power is being co-opted for a very different purpose – to legitimate the power of a new ruling élite.
Between April and July of 1994, 800–850 thousand people were slaughtered in Rwanda (Prunier 1995:265). The vast majority of the dead were members of the minority Tutsi ethnic grouping. However, members of the majority ethnic grouping – the Hutu – were also killed if they were seen as opponents of the genocide’s (Hutu) organisers. The genocide occurred despite the existence of a peace and power sharing agreement (the Arusha Accords) to which all parties to the conflict had ostensibly subscribed, and which this paper seeks to critique.
There are three main dimensions to this critique. The first dimension is an argument that the Arusha process heightened tensions within élite circles and provided a channel through which aspirant élites could pursue highly dangerous goals. In Rwanda, such goals tend to be pursued in a zero-sum or ‘winner takes all’ manner – state power is not easily shared in such circumstances. Nor were matters helped by the duplicitous and sinister role played by the French government. Thus, the Arusha Accords, while formally reasonable and conventional, generated intense reactions and dynamics by virtue of the structural characteristics of political élite competition.
However, even more fundamentally, the Arusha process, rooted as it was in power sharing modalities between various élite and aspirant élite actors, failed to tackle the most pressing problems of Rwandan society: chronic and worsening poverty; entrenched and intensifying inequality; the treatment of the poor with contempt; a pervasive sense of impunity in the context of egregious human rights abuses; and the oppressive presence of the state in all aspects of social life (Uvin 1998:45). This disastrous cocktail – creating what Uvin (1998) calls a situation of ‘structural violence’ – laid the basis for mass participation in the genocide of 1994. I argue that, far from helping solve these problems, certain international interventions – especially economic ‘structural adjustment’ that ran parallel to the Arusha negotiations – worsened the situation. The Arusha Accords therefore failed, and this is the second dimension of the critique, because they neglected not just the actual dynamics of élite political competition, but also the (worsening) structural conditions of life for the vast bulk of ordinary Rwandans.
The third dimension of the critique concerns post-genocide Rwanda and how the legacy of the Arusha Accords has, amongst other devices, been used to legitimise new forms of repression at the same time as the abuse and violence inflicted upon ordinary Rwandans (and their neighbours) have continued. A seemingly reasonable political agreement to share power is being co-opted to legitimate the inequitable and oppressive power of a new ruling élite.
In terms of structure, the paper begins with an overview of Rwandan history until 1994, before doubling back to examine: the crucial role of a state-based governing élite (the akazu); the rise of poverty, inequality and ‘structural violence’; and the rise also of protest against the akazu-dominated regime. This is the context in which the Arusha Accords are then assessed and critiqued. Finally, the paper focuses on the post-1994 experience.
Rwanda is a small country in central Africa but with a large population size – approximately 7.2 million before the genocide and growing by more than 3 percent per annum – making it the most densely populated country in Africa (World Bank 1994:1; Uvin 1998:180). Prior to 1994, Rwanda’s population consisted of two main, indigenous ethnic groups – the Hutu, who accounted for approximately eighty-five per cent of the population, and the Tutsi, who accounted for most of the remaining fifteen per cent (with the Twa group accounting for probably less than one percent). The two main ethnicities lived side by side, spoke the same language (Kinyarwanda), and shared membership of ethnically cross-cutting clan, religious and neighbourhood groups (Van Hoyweghen 2000:2).
The Tutsi (associated with a pastoral lifestyle) are frequently portrayed as invaders who came from the Horn of Africa and imposed a harsh monarchical regime on the earlier arriving Hutu (usually associated with cultivation of the soil). However, other assessments indicate that all of the different groups may have arrived in migratory waves over many centuries, and that theories of conquest must be abandoned (Takeuchi 2000:185). Some commentators suggest that the terms Hutu, Tutsi and Twa referred more to social status than to ethnicity in pre-colonial times (Hintjens 2001:27–28). In parts of the country, especially the north-west, Hutu rulers enjoyed large measures of autonomy from the royal court (Takeuchi 2000:189).
The German colonial administration was established at the end of the nineteenth century and was succeeded by that of Belgium after the end of World War I. Both the German and the Belgian administrations exploited the hierarchical structure of Rwandan society as a mechanism of indirect rule, with a certain stratum of Tutsi deployed as a colonial ruling class (Mamdani 2001:27). The Hutu kingdoms in the north-west of the country that had previously enjoyed a measure of autonomy were brought under the control of the central Tutsi court with the military assistance of the colonisers, and Hutu chiefs throughout the country were replaced by Tutsi at the instigation of the colonial powers (Van Hoyweghen 2000:4). Whatever fluidity and ambiguity had previously existed in the system was greatly restricted as a system of ethnic identity cards was introduced (in 1933) and ethnicity thus became a strict (patrilinear) inherited characteristic (Hintjens 2001:30). Tutsi were systematically favoured in employment and education and accorded the status of a superior ‘race’.
The run-up to independence (in 1962) saw a reversal of the colonially imposed order, with some Hutu seizing control and beginning a series of pogroms against the Tutsi population, with tens of thousands killed and many others forced into exile (Human Rights Watch 1999:39). The post-colonial regime was initially dominated by Hutu from the south of the country, but from 1973 onwards power became concentrated in the hands of a northern Hutu élite under the leadership of President Habyarimana, who took power in a military coup. He instituted a single-party state, with every citizen an automatic member of that party – the Mouvement Révolutionnaire National pour le Développement (MRND).1 Parallel state-party structures tightly monitored and controlled the population (Longman 1999:342). Those Tutsi who remained in Rwanda after 1962 were subject to discrimination in education, employment and other areas.
Meanwhile, those Tutsi driven into exile – many of whom grew up in refugee camps in Uganda – became the main source of a rebel movement, the Rwandan Patriotic Front (RPF) with a military wing called the Rwandan Patriotic Army (RPA), which was to attack the regime in 1990, demanding the right to return to the land they and/or their parents were expelled from.2 After being initially repulsed by the Rwandan army, the RPF regrouped and undertook a prolonged guerrilla campaign involving sporadic offensives from their northern bases and occasional (short-lived) captures of large towns. However, the RPF found little in the way of popular support inside Rwanda (from Tutsi or Hutu), and the war contributed to the ‘structural violence’ of Rwandan society (discussed below).
Following the mysterious killing of Habyarimana himself in April 1994,3 the army and government-run militias initiated and led the genocide. The RPF succeeded in militarily defeating the government forces in July 1994 and the present government is dominated by the RPF – its record will be discussed in the last section of this paper.
The akazu: at the heart of the state
Akazu is a Kinyarwanda word meaning ‘little house’. In the 1980s, it came to be applied to the country’s ruling clique – the northern-based politico-commercial network centred on President Habyarimana’s family (Reyntjens 1994:189). The akazu was heavily involved in criminality and corruption, using its control of the state to enrich itself and its allies (Braeckman 1994:109–111; Reyntjens, 1995:284). After October 1990 the akazu focus tended to be on diverting resources– including funds from the state employees’ pension fund – towards military ends (Hintjens 1999:257; Melvern 2000:64–68). The director of a local magazine published, in 1992, an article identifying 25 members – including the President and members of his family – of a group operating as ‘death squad’ organisers targeting those seen as threats to the regime (cited in Reyntjens 1996:247). A report from Amnesty International in May 1992 documented the involvement of state agents in mass murder and torture (Amnesty International 1992:1).
Amnesty’s findings were confirmed by an international commission of inquiry on human rights abuses in Rwanda – its report, in March 1993, found that the Rwandan government had, since October 1990, been responsible for the deaths of some two thousand people and that these deaths were sanctioned by the very highest forces in the land (cited in African Rights 1995:33).
What above all else facilitated the practice of illegal and repressive activities on the part of the akazu was control of the state. Without state control – direct and indirect – over the economy and society, the akazu could not function, economically or politically (Cart 1995:476). Habyarimana, for example, could demand that private enterprises contribute to his cause because they needed state approval and concessions to ensure business profitability (Human Rights Watch 1999:43), and foreign aid ‘could only be appropriated through direct control of government power at high levels’ (Prunier 1995:84). In Rwanda, battles over the distribution of economic resources were (and are) battles over control of the state. The question of who would retain or assume ‘ownership’ of this apparatus of control was the key stake of political struggle in Rwanda in the early 1990s.
Mass impoverishment, growing inequality, rising protest
Impoverishment was on the rise in the Rwanda of the 1980s and early 1990s. A principal problem lay in the evolution of global commodity markets: between 1985 and 1992, the real world price of coffee (Rwanda’s main export) fell by seventy-two per cent; between 1986 and 1992, the real purchasing power of Rwanda’s export earnings fell by 59 per cent (Woodward 1996:19, 21). This very severe foreign exchange problem arose in the context of an agricultural sector already structurally crisis-ridden by a chronic shortage of land and rapid population growth (Moodley, Gahima and Munien 2010). By the early 1990s, more than half of all Rwandan farmers occupied farms of less than one hectare, often on ecologically fragile soils, while up to 25 per cent of the population was landless (Mullen 1995:23). Forty-three per cent of all farm households lacked enough land to subsist upon (Uvin 1998:113). One desperate response to the tightening population-land pincer movement was to switch from cereal and bean cultivation towards that of root crops, so that many people’s diet became increasingly protein-deficient (Mamdani 2001:146).
In addition, the arrival of the AIDS virus in the early 1980s, drought (in 1984), excessive rain (in 1987) and plant disease (in 1988) all weighed in to contribute to declining production and food security levels (Uvin 1998:57). By 1989, an estimated one in six Rwandans was affected by famine (Pottier 1993:5), one quarter of all children was severely malnourished (World Bank 1991:1), and some 50 per cent of all children suffered from stunting (Uvin 1998:112). From October 1990 civil war was costing an estimated $100 million per annum and was causing massive displacement and disruption, especially affecting the most fertile northern regions (Marysse, De Herdt and Ndayambaje 1994:10). This resulted in the displacement of 15 per cent of the population – 1 million people (Marysse, De Herdt and Ndayambaje 1994:83).
Falling levels of income were increasingly unequally distributed. Akazu members and associates were taking over land previously under the control of smaller, often indebted farmers (Braeckman 1996:106). In the rural areas, the percentage of income held by the richest 10 per cent – themselves often traders or civil servants not actually resident in those areas – is estimated to have risen from 20 per cent in 1982 to 41 per cent in 1992 (Maton 1994:29; see also André and Platteau 1998). The inequality-enhancing effects of land concentration were compounded by the scarcity of non-farm employment opportunities, and by the fact that those opportunities were themselves unequally distributed (Clay and McAllister 1991:37).
Peter Uvin (1998) argues that the Rwanda of the early 1990s had become what he terms a ‘structurally violent’ society. This condition is characterised by extreme poverty – Rwanda, proportionately, may have had more absolutely poor people (perhaps 90 per cent of the population) than anywhere else in the world (Uvin 1998:117). But it was also characterised by (rising) inequality, injustice, discrimination, corruption, and treatment of the poor with contempt. The poor – the vast majority of the population – were subjected to humiliation and a state of permanent exclusion from the benefits of ‘development’, benefits that neither they nor their children could ever hope to achieve but which were flaunted in their faces by wealthy locals and foreigners. ‘Peasant life was perceived as a prison without escape in which poverty, infantilisation, social inferiority, and powerlessness combined to create a sense of personal failure’ (Uvin 1998:117).
Thus, the activities of a ruling élite were contributing to societal inequality, mass impoverishment and a situation of ‘structural violence’. This was generating resentment on the part of the mass of the population (the vast majority of whom were Hutu) and undercutting the legitimacy of the regime. A demonstration of this resentment was the action of southern (mostly Hutu) farmers in tearing up anti-erosion devices and destroying communal wood lots which they had been forced to dig or construct under the government’s compulsory communal labour programme (umuganda) (Mamdani 2001:147), and the uprooting of up to 300 000 coffee trees, cultivation of which was also meant to be compulsory (Kimonyo 2000:50).
The advent of multi-party democracy provided a channel through which popular discontent could be further expressed, and raised the very real prospect of the akazu losing its grip on the organs of state power. Longman (1999:344) describes a situation characterised by ‘declining legitimacy of the regime, decreasing compliance with state directives, increasing criticism of state officials and practices, and growing formal and informal protest’.
Threats to akazu control of the state
Habyarimana was forced by international pressure to legalise opposition political parties in 1991, and a large number of such parties quickly became active (Prunier 1995:126). The most important of these were: the Movement Démocratique Républicain (MDR), the largest party and broadly representing the southern Hutu who had been marginalised after Habyarimana’s ‘northern’ coup of 1973; the Parti Libéral (PL), a party associated with the business sector and including a number of Tutsi business people; the Parti Social Démocrate (PSD), a largely anti-sectarian and left-leaning party; and the Parti Démocrate Chrétien (PDC), associated with the Catholic Church. By early 1992, ‘Prominent opponents of the regime and democracy activists were confident that power was on the brink of changing hands’ (Longman 1999:339). In April 1992 Habyarimana formed a new, coalition government, consisting of ten ministers from his own party and nine from the erstwhile opposition (still mainly referred to as opposition parties in this paper, to minimise confusion). The new ministers moved to ensure that their own supporters gained key posts in central and local government, and also sought to end the systematic discrimination in education policy which had assured children from north-west Rwanda disproportionate access to school places (Human Rights Watch 1999:54–55; Prunier 1995:145–146).
Such moves prompted, by way of reaction, a powerful coalition of interests determined to defend the old order. This extended beyond the akazu itself to include state employees who feared that the new political forces would use state patronage to employ their ‘own’ people at all levels of the hierarchy. With economic crisis and ‘structural adjustment’ (discussed further below) simultaneously placing a cap on the total number of jobs available, ‘[l]ow-ranking officials in the villages – including administrators, teachers, agricultural extension workers, health workers and policemen – saw their prospects of promotion vanish, and even faced the possibility of losing their jobs altogether’ (African Rights 1994:19). If the Arusha Accords (see next section) had been implemented, all administrative positions were to be reviewed within three months of the formation of a new government (Human Rights Watch 1999:126). And to this list should be added the members of the newly expanded army, who lived ‘relatively well – from exactions if not from salary’, and who ‘dreaded demobilisation’, an especially acute fear amongst the senior officers who were targeted for first-stage demobilisation (Human Rights Watch 1999:60, 125; see also Prunier 1995:150).
Despite the ability to draw on this constituency of support, the akazu still faced a powerful threat because most Hutu were excluded from the benefits of state patronage, a situation which particularly rankled some southern Hutu who bitterly resented the north-western monopoly over power (Voyame et al. 1996:139). How was the akazu to deal with this challenge? One response was simple violence, including targeting and repression of opposition activists (Longman 1999:348). Ruling party militias disrupted opposition party rallies and became more generally involved in what Longman (1999:348–349) terms the ‘organisation of chaos’ – the carrying out of seemingly random bomb attacks, robberies, rapes and other crimes, with the apparent intention of simply heightening public insecurity and therefore generating ‘nostalgia for single-party authoritarian rule’ (Longman 1999:350).
Another tactic was to seek to manipulate the democratisation process. Crucially, in early 1992 a party called the Coalition pour la Défense de la République (CDR) was formed, pushing a Hutu extremist agenda and criticising the MRND(D) for conceding too much to the RPF and the other opposition parties. This party is widely reckoned to have been a creation of the akazu itself, and its role was to state positions that Habyarimana and the MRND(D) themselves believed but preferred not to be seen saying (Human Rights Watch 1999:52–53). Co-option of leading figures in other opposition parties was also a favoured regime tactic, with Habyarimana establishing ‘Hutu Power’ (usually referred to simply as ‘power’) factions within the main opposition parties (Prunier 1995:181; Uvin 1998:65). It was in this context of extreme violence and instability that the Arusha Accords were negotiated.
Arusha and its discontents Process and outcome
As mentioned above, in April 1992, Habyarimana had installed a multi-party government consisting of ten ministers from his own party and nine from the opposition. Between May and June 1992, representatives of three of those ‘opposition’ governing parties – the MDR, the PSD and the PL – met with the RPF and it was agreed that peace negotiations between all parties should be initiated (Mamdani 2001:210). The Arusha peace negotiations opened in July 1992 under the auspices of the Organisation for African Union (OAU) and facilitated by the government of Tanzania. As well as the various Rwandan parties, there were also delegations from other African and Western countries. The only major party to be excluded from the negotiations was the CDR. The RPF refused to negotiate with the CDR on the grounds that it was simply a front for the MRND(D) and that it was overtly racist – no one with even a Tutsi grandparent could join the CDR (Melvern 2000:54). All other parties to the talks wanted the CDR included and British and US diplomats pressured the RPF to agree to this, but to no avail: ‘Western governments saw the exclusion of the CDR as a departure from constructive negotiations, insisting that a more substantive role should be given to those who stood to lose power’ (Melvern 2000:54).
The Rwandan government delegation was first led by the Minister for Foreign Affairs – Boniface Ngulinzare of the MDR – and later by the Minister for Defence – James Gasana of Habyarimana’s MRND(D) (Mamdani 2001:210). Neither man spoke for the hardline Hutu faction – Gasana fled into exile later in 1993 and Ngulinzare was killed in the genocide in April 2004 (Mamdani 2001:210). The hardliners were represented by Colonel Théoneste Bagosora, who frequently attended at Arusha to monitor developments (but not to negotiate) and who would go on to be the main coordinator of the genocide (Prunier 1995:163). It was also far from clear that the government delegation spoke for Habyarimana himself, whose exact relationship to the hardliners was itself unclear. In November 1992, Ngulinzare stated: ‘the MRND keeps talking in contradictory ways. On the one hand, it pretends to support the peace negotiations and on the other hand it keeps sabotaging them’ (Prunier 1995:171). This tension was evident when a provisional power sharing agreement was agreed in January 1993 that envisaged the creation of a Broadly Based Transitional Government (BBTG) with 5 cabinet posts allocated to each of the MRND(D) and the RPF, 4 to the MDR, 3 to each of the PSD and the PL, and one to the PDC. MRND(D) and CDR supporters demonstrated in Rwanda against the deal and the MRND(D) national secretary claimed that his party had rejected the agreement (Prunier 1995:173).
In claimed response to government-organised massacres of Tutsis, the RPF broke a ceasefire in February 1993 and restarted the war (Prunier 1995:174). France sent military aid and troops to support the government. Between February and March 1993, the ‘opposition’ parties met with the RPF (in Burundi) and issued a call to, amongst other things, renew the peace negotiations (Prunier 1995:179). However, by now Habyarimana had created the ‘power’ factions of each opposition party (see above) and representatives of these factions simultaneously grouped with the MRND(D) and the CDR in Rwanda to condemn the RPF (Prunier 1995:179).
As late as July 1993, Habyarimana was resisting signing up to a deal. However, in late July aid donors (including the World Bank) insisted that aid to the government would be halted unless a deal was reached; even Habyarimana’s most ardent supporter, France, joined in this effort (Human Rights Watch 1999:124; Kuperman 2005:75). The composition of the BBTG (with cabinet seats to be allocated as agreed in January) was confirmed in the August 1993 agreement – this was to hold power for a maximum of 22 months until elections could take place (Melvern 2000:53). The agreement also contained provisions for a merged national army made up of the existing Rwandan army, the FAR (60%), and the RPA (40%), with the officer corps to be split 50:50, and the right of return for all refugees accepted – a crucial demand of the Tutsi exiles (Mamdani 2001:210–211). There was no provision for any amnesty for human rights abuses (Melvern 2000:53). The Accords also covered a range of other areas, including the establishment of the rule of law and the creation of institutions to oversee the political transition.
Habyarimana almost immediately sought to derail the agreement by insisting that the MDR, PL, PSD and PDC government ministers come from the ‘power’ factions of those parties and that the BBTG be broadened yet further to include the CDR (Kuperman 2004:76). The appetite for implementing the agreement was also weakened by the assassination in October 1993 of neighbouring Burundi’s first Hutu president by Tutsi soldiers and subsequent massacres of Hutus in that country, events that were portrayed as confirming the dangers of allowing any Tutsi role in government and, crucially, the army (Kuperman 2004:76).
What derailed Arusha? Or, whose interests did Arusha serve?
A fragmented Rwandan government delegation
The negotiators for the Rwandan government were disproportionately drawn from MRND(D) ‘liberals’ and from the opposition parties who had entered into coalition with the MRND(D), in part because Habyarimana wanted to distance himself from the process (Clapham 1998:203). In part, they were using the Arusha negotiations to enhance their own power against that of Habyarimana and the MRND(D) (Prunier 1995:163; Stettenheim 2002:225). (And they were also in opposition to the ‘power’ factions within their own parties). It was therefore not surprising that the final settlement – gave an extraordinary weighting in the proposed transitional government to parties with no military strength, no control of territory, and an as yet undetermined level of popular support. Confident in their ability to capitalize both on their Hutu ethnic identity (which would enable them to sideline the RPF), and on the unpopularity of the Habyarimana regime, the minor parties then hoped to establish themselves more firmly in power through early elections (Clapham 1998:205).
But these parties were vulnerable to the charge that by granting so many concessions to the RPF, they were ‘betraying’ the Hutu people – a charge vociferously levelled against them by the CDR and their own ‘power’ factions (Mamdani 2001:211). They were not helped by the lack of external support available to them relative to that available to the RPF (from Uganda and the wider Tutsi diaspora)4 and to the MRND(D)/CDR (from France especially). Nor were they helped by the belligerence of the RPF.
The strategy of the RPF
Kuperman (2004) attributes much of the blame for the genocide to the RPF. His list of charges includes: invading in the first place; launching military offensives in 1991 and 1992; being opposed to compromise during peace negotiations in 1992 and 1993; breaking the ceasefire in early 1993; refusing any renegotiation of Arusha in late 1993; refusing ceasefire offers at the start of the genocide in April 1994; and pursuing a military strategy during the genocide that prioritised military victory over the protection of ordinary Tutsi (Kuperman 2004:62). The RPF is itself estimated to have massacred tens of thousands of civilians between April and September 1994 (Reyntjens 2004:194). In addition to those civilians they themselves killed, from the beginning of their campaign, Kuperman (2004:61) argues, ‘the rebels expected their invasion to trigger a violent backlash against Tutsi civilians in Rwanda’. To this charge list may be added the (contested) claim that it was the RPF that shot down Habyarimana’s plane and directly triggered the genocide in April 1994 (Lemarchand 2006:6; Robinson and Ghahraman 2008).
Specifically in the context of the Arusha Accords, the RPF – the most capable and determined party to the negotiations – pursued a maximalist agenda, especially with regard to the division of the military. According to the then US Assistant Secretary of State for Africa, ‘RPF demands concerning the future of the military were guaranteed to push the regime into a state of total paranoia’ (Kuperman 2004:75). The insistence on excluding the CDR may be seen as another example of this approach (Mamdani 2001:211; Melvern 2000:54).
Excluding the CDR
Spears (2000:115) argues that ‘the lesson of Rwanda is that one cannot afford to leave anyone out of the political process’. A diplomat who was involved in the Arusha negotiations is quoted as claiming that ‘the 1993 Arusha Accords were the perfect example of the failure of power-sharing because of a basic decision to exclude a group of people’ – the CDR, who were left to choose between losing power or violently subverting the Accords (Lemarchand 2006:5).
But whether the CDR could have been included in the government has been forcefully challenged by Clapham (1998:205–206):
These groups [CDR/akazu] were fundamentally irreconcilable to any resolution of the conflict through a negotiated settlement … The incorporation of such groups into the Arusha process could only have aborted the process itself. It could certainly be argued that this would have revealed the futility of the negotiations, and compelled a resort to war … but there is no plausible basis for the belief that it could have led to a viable settlement.
A ‘settlement’ involving the CDR could only have been premised on their being in charge or their being defeated (Clapham 1998:209). But the CDR could not easily be defeated so long as they received external backing from France.
France consistently and substantively supported the Habyarimana government and the akazu (Prunier 1995:162–163). French military support was crucial in repelling the RPF offensive of early 1993 (300 new French troops were rushed to the country), and French instructors deployed at this time trained the militias who would go on to perpetrate the genocide the following year (Prunier 1995:164–165, 176). The French Secret Service spread disinformation about the RPF offensive (such as massacre allegations) to help justify further French intervention (Prunier 1995:176).
In February 1993, the French Minister for Cooperation, during a visit to Kigali, asked non–MRND(D) parties to ‘make a common front’ with Habyarimana in opposition to the RPF (Prunier 1995:178), a direct undermining of the ostensible French commitment to inclusive negotiations at Arusha. Though the French government did press Habyarimana to agree to the deal in July 1993, Habyarimana expected the French to back him in subverting the Accords after their signing (Stettenheim 2002:226). France supplied arms to Rwanda in January 1994 in contravention of the Arusha Accords (Stedman 1997:23), and French military aid continued even after the genocide had begun in 1994 and a UN Security Council arms embargo had been imposed (Andersen 2000:441).
Arusha: preparing the apocalypse?
Those who had most to lose from a power sharing agreement in Rwanda did not meaningfully participate in the Arusha negotiations, and even if they had done so they would not have been willing to genuinely commit to any significant diminution of their power. And their stubborn refusal to cede power was backed up by the military and political support of France. Those who did participate on the ‘opposition’/government side were mostly seeking to enhance their own political prospects rather than, necessarily, instituting a stable and sustainable settlement. The RPF pursued a hardline approach that heightened the insecurities of the akazu and their allies, making it more rather than less likely that extreme violence would be precipitated, which may well have been what they really wanted all along (Kuperman 2004). This was a recipe for disaster.
The role of structural adjustment
Economic ‘structural adjustment’, simultaneously pushed at this time by Western actors and institutions, also wreaked considerable harm. The most notable short-term impact of adjustment was a massive increase in development aid to help (or so it was intended) the Rwandan government implement the economic reform measures (Uvin 1998:87–88, 91). This funding of the Rwandan state enhanced its legitimacy and may well have encouraged it to believe it could get away with still further abuses (Storey 2001).
Structural adjustment also impacted directly on ordinary Rwandans through devaluation-induced food price rises (Woodward 1996:20), while increased user fees for health and education services ‘contributed significantly to social tensions and fears’ (Newbury 1995:14). This last point is also made by Sellström and Wohlgemuth (1996:20) who cite ‘ample evidence that the introduction of higher fees for health and education, among other things, added to the already heavy burden on Rwanda’s poor’. Even for those not at the cutting edge of poverty – state employees – adjustment’s cap on public sector recruitment would have contributed to fear and insecurity, especially when allied with the Arusha-related threat of new political masters making new appointments (see above).
The structural violence under which Rwandans lived has been persuasively argued to be a key motive force for mass participation in genocide (Uvin 1998). Hundreds of thousands of people perpetrated atrocities (Mamdani 2001:5–6). Uvin links this to the frustration, hopelessness and anger engendered by structural violence. These, in turn, Uvin argues, provoked a desire for scapegoating because the identification and persecution of a scapegoat, at a socio-psychological level, helped to combat low self-esteem and provided some sense of hope and direction. The existence of deeply rooted racism meant that a scapegoat (the Tutsi) was readily to hand and élite (akazu) manipulation ensured that was the direction towards which anger was channelled. By enhancing élite capacities and adding to structural violence, structural adjustment made its own contribution to this explosive mix.
Back to the future: the legacy of the Arusha Accords
Rwanda at the end of the genocide was a devastated country: at least 800 000 people were dead; 2 million refugees had fled abroad; 1 million people were living in ‘internally displaced’ camps inside the country; some 500 000 ‘old caseload’ Tutsi refugees had returned after many years in exile; and most civil servants were dead or were refugees (Reyntjens 2004:178). Furthermore, the country’s infrastructure lay in ruins, crops and livestock were mostly destroyed, and banks and businesses had been ransacked (Reyntjens 2004:178). In the midst of this chaos, the victorious RPF affirmed it remained committed to the Arusha Accords, though insisting that the MRND(D) be excluded as having been a party to genocide. A minister in the new government argued that ‘Arusha was well negotiated. It offered the promise of political stability. It was our bible’ (Bruce 2007:11).
The government that was inaugurated on July 19, 1994, was a genuine government of national unity. It was fully in the spirit of the Arusha Peace Agreements of August 1993 … The new president, Pasteur Bizimungu, was an RPF Hutu who had been a government civil servant in the 1980s. Of the twenty-one ministries, the lion’s share (eight) had gone to the RPF; the rest were evenly distributed, with four ministries going to the … [MDR], three to the … [PSD], three to the Liberals, two to independent personalities, and one to the small Christian Democratic Party. In ethnic terms fifteen of the new ministers were Hutu and only six were Tutsi. After such a catastrophe the new cabinet looked like a small miracle of reason in a sea of madness’ (Prunier 2009:7).
And yet, within barely a year, that ‘miracle of reason’ would be revealed as a shallow façade. On a range of issues – justice (the often arbitrary arrest and detention of alleged genocide suspects), the possession/repossession of property, monopolisation of economic resources (Dorsey 2000:324–326) and others – a clique within the RPF leadership began to monopolise power (Prunier 2009:43). In April 1995, government troops massacred thousands of Hutu ‘internally displaced people’ at a camp in southern Rwanda (Prunier 2009: 37–42). The Minister for the Interior, Seth Sendashonga (unusual in being a Hutu member of the RPF), opposed this and other human rights abuses. In August 1995 Sendashonga, along with other ministers, was fired and he and the Prime Minister (Faustin Twagiramungu, an MDR Hutu) were placed under house arrest (Prunier 2009:46). Both fled the country in late 1995, and Sendashonga was murdered by RPF agents in Kenya in 1998 (Prunier 2009:365–368).
Tens of thousands of civilians were killed by government forces in 1997 and 1998 in counter-insurgency operations (Reyntjens 2004:195). Another wave of high-profile political resignations (forced or otherwise) followed in 2000, including that of the Hutu President, Pasteur Bizimumgu (to be replaced as president by the head of the RPF, and de facto national, ruler, General Paul Kagame), and the MDR was banned as a political party in 2003 on the grounds of promoting ‘divisionism’ (Reyntjens 2004:184). Kagame won a presidential election in 2003 in a context of widespread intimidation and vote rigging (Prunier 2009:295). Parliamentary elections later in 2003 saw all candidates not members of, or allied to, the RPF debarred or intimidated out of the electoral running (Reyntjens 2004:186), following a pattern established at deeply flawed local elections in 2001 (Reyntjens 2004:182). Kagame won another presidential election in 2010, again in a context of systematic exclusion and harassment of the opposition (Beaumont 2010).
After 2001, abuses such as extra-judicial killings had escalated even outside periods of electoral contestation (Front Line 2005). Media freedom became increasingly circumscribed through legal and extra-legal measures (Reyntjens 2004:181; Human Rights Watch 2010:151). The space within which civil society can genuinely contest government policy is severely circumscribed (Beswick 2010; Front Line 2005). In 2004, a number of national and international non-governmental organisations (NGOs) were accused by a parliamentary organisation of contributing to ‘divisionism’ and were either suppressed or expelled (Buckley-Zistel 2009:47).
Meanwhile, ‘[t]he benefits of the country’s economic progress have been channelled almost exclusively to the new elites living in their large villas in Kigali, while 90 per cent of the people continue to scrape together an existence below the poverty line in rural areas’ (Oomen 2005:900). Food and asset vulnerability, along with continuous food shortages, remain pervasive in rural areas (Hintjens 2008:20). ‘Distress sales’ of land are once again common, dependence on food aid for survival is common, and many of the rural poor lack access to even basic healthcare and education services (Hintjens 2008:20, 21).
The population is, as in the past, closely surveilled and tightly controlled (Ansoms 2009:304–305; see also Ingelaere 2010a). This has a pronounced ethnic dimension: control is exercised through appointed (not elected) officials who draw state salaries and whose career prospects are dependent on following government diktats – these people are often Tutsi (though a small number of all Tutsi) with no prior connection to the area and who are accountable upwards to central government but not downwards to the people (Ingelaere 2010b:288–290).
In addition to its regressive record at home, the RPF government has attracted significant international condemnation for its direct interventions in neighbouring Democratic Republic of the Congo (DRC). Initially, in 1996, this involvement had the main objective of closing the refugee camps in which the former genocidal forces and hundreds of thousands of civilians had become ensconced after their flight from Rwanda in 1994 – these camps were indeed closed, and mass killings of civilians by the RPF and its allies were carried out in the process (Reyntjens 2004:205). A second major phase of conflict began in August 1998 and has led to considerable criticism of all parties involved (including several other governments in the region) over extensive civilian deaths, other human rights abuses, and the looting of DRC resources (Human Security Report Project 2010:36–48). A UN Panel report on the issue in 2001 constituted a damning indictment of the Rwandan government’s piracy of DRC resources – including coltan and gold (UN Panel 2001:3). Another UN-established expert group on the DRC reported in 2008 that Rwanda maintained support for murderous militia groups within the DRC (Group of Experts 2008), and a further UN report released in October 2010 documented massive human rights violations in the DRC, in which the armed forces of Rwanda (and other countries) are implicated (Onyiengo 2010).
Despite overwhelming evidence of its murderous brutality (at home and abroad), the absence of any real democracy in the country, and the (at best) uneven nature of economic progress, the Rwandan government retained substantial external support, particularly from Britain and the USA. Hayman concludes that ‘donors have remained largely supportive even in the face of signs of increasing authoritarianism and poor political governance’ (2009:177). Why should this be the case? One answer is that some donors are impressed at the government’s reconstruction and development efforts (Beswick 2010:246; Uvin 2001). Another is that the RPF government has been highly effective at creating and cashing a ‘genocide credit’. By portraying itself as the saviour and defender of the Tutsi ethnic minority, and highlighting the past failure of external actors to protect that minority, it has enhanced its negotiating hand with Western powers in particular (Pottier 2002). The Rwandan government has also proved very skilled in speaking the language of international development, as manifested in strict adherence to neo-liberal economic policies and stated commitments to ‘poverty reduction’, ‘participation’, and other beloved buzzwords of the international development sector (Hayman 2009:175; Oomen 2005:901). For the staff of aid agencies who rarely venture into rural areas this may be a persuasive discourse (Ingelaere 2010a).5
The RPF élite has, in summary, decisively established its domination over Rwandan society and has been able to align itself sufficiently well (rhetorically or in reality) with the goals of international actors so as to attract substantial international support and ward off any particularly significant international censure. Crucially, one such international goal is the idea of power sharing and ‘national reconciliation’, as conceived under the Arusha Accords. According to the official Rwandan government website,6 the ‘[Arusha] principal provisions now constitute the Fundamental Law of the Republic of Rwanda’. Non-RPF and Hutu ministers continue to hold cabinet positions (however non-existent their real power is). The letter of Arusha is (partly) observed even as its spirit is violated.
At the start of this paper, I argued that the Arusha peace process was flawed in three main ways. The first was that it provided a channel through which élites and aspirant élites could pursue state power – power that they had little or no interest in genuinely sharing. The second was that it did not address the structural violence of Rwandan society – the chronic impoverishment, and deep senses of deprivation and humiliation, that laid the bases for mass participation in the 1994 genocide. Indeed, another external intervention at that time – structural adjustment – actually made this structural violence worse. The third was that Arusha has, since 1994, been co-opted by the RPF élite to legitimise its grip on power.
Indeed, the current Rwandan situation parallels the early 1990s in important ways. The new ruling élite uses its claimed commitment to the Arusha Accords as a device to legitimise its monopolisation of state power – both because that power is essential to its (political and economic) reproduction as an élite, and also because any diminution of its grip on state power would expose it to severe, quite possibly fatal, threat. The fig leaf of power sharing conceals a naked refusal to share any real power with anyone outside the élite. And while the Arusha Accords are being strategically appropriated to legitimise the élite claim to monopoly power, the underlying condition of ‘structural violence’ for the vast mass of ordinary people is, if anything, intensified.
Allowing a state to claim adherence to a power sharing settlement (as the current Rwandan government does vis-í -vis Arusha) while it is violating the rights of its citizens discredits the whole conflict resolution project.
African Rights 1994. Rwanda: Death, despair and defiance. London, African Rights.
African Rights 1995. Rwanda: Death, despair and defiance (revised edition). London, African Rights.
Amnesty International 1992. Rwanda: Persécution de la Minorité Tutsi et Répression des Détracteurs du Gouvernment, 1990–1992. London/Paris, Amnesty International.
André, Catherine and Jean-Philippe Platteau 1998. Land relations under unbearable stress: Rwanda caught in the Malthusian trap. Journal of Economic Behaviour and Organisation, 34, pp. 1–47.
Andersen, Regine 2000. How multilateral development assistance triggered the conflict in Rwanda. Third World Quarterly, 21 (3), pp. 441–456.
Ansoms, An 2009. Re-engineering rural society: The visions and ambitions of the Rwandan elite. African Affairs, 108 (431), pp. 289–309.
Beaumont, Peter 2010. Paul Kagame: A tarnished African hero. Observer, 18 July.
Beswick, Danielle 2010. Managing dissent in a post-genocide environment: The challenge of political space in Rwanda. Development and Change, 41 (2), pp. 225–251.
Braeckman, Colette 1994. Rwanda: Histoire d’un Génocide. Brussels, Fayard.
Braeckman, Colette 1996. Terreur africaine: Burundi, Rwanda, Zaire: les racines de la violence. Brussels, Fayard.
Bruce, John 2007. Drawing a line under the crisis: Reconciling returnee land access and security in post-conflict Rwanda. Humanitarian Policy Group Working Paper, Overseas Development Institute.
Buckley-Zistel, Suzanne 2009. Nation, narration, unification? The politics of history teaching after the Rwandan genocide. Journal of Genocide Research, 11 (1), pp. 31–53.
Cart, Henri-Philippe 1995. La coopération suisse au Rwanda ou les limites de l’aide extérieure. In: Guichaoua, Andre ed. Les crises politiques au Burundi et au Rwanda (1993–1994), pp. 465–485. Paris and Lille, Karthala and Université des Sciences et Technologies de Lille.
Clapham, Christopher 1998. The perils of peacemaking. Journal of Peace Research, 35 (2), pp. 193–210.
Clay, Daniel and Jim McAllister 1991. Family development cycle, social class, and inequality in Rwanda. Rural Sociology, 56 (1), pp. 22–40.
Dorsey, Michael 2000. Violence and power-building in post-genocide Rwanda. In: Doom, Ruddy and Jan Gorus eds. Politics of identity and economics of conflict in the Great Lakes region, pp. 311–348. Brussels, VUB University Press.
Front Line 2005. Front line Rwanda: Disappearances, arrests, threats, intimidation and co-option of human rights defenders 2001–2004. Dublin, Front Line.
Group of Experts 2008. Final Report of the Group of Experts on the Democratic Republic of the Congo. New York, United Nations.
Hayman, Rachel 2009. Rwanda: Milking the Cow. Creating policy space in spite of aid dependence. In: Whitfield, Lindsay ed. The politics of aid: African strategies for dealing with donors, pp. 156–184. Oxford, Oxford University Press.
Herman, Edward and David Peterson 2010. Rwanda and the Democratic Republic of Congo in the propaganda system. Monthly Review, May, pp. 20–6.
Hintjens, Helen 1999. Explaining the 1994 genocide in Rwanda. The Journal of Modern African Studies, 37 (2), pp. 241–286.
Hintjens, Helen 2001. When identity becomes a knife: Reflecting on the genocide in Rwanda. Ethnicities, 1 (1), pp. 25–55.
Hintjens, Helen 2008. Post-genocide identity politics in Rwanda. Ethnicities, 8 (5), pp. 5–41.
Human Rights Watch 1999. Leave none to tell the story: Genocide in Rwanda. New York, Washington, London and Brussels, Human Rights Watch.
Human Rights Watch 2010. World Report 2010. New York, Washington, London and Brussels, Human Rights Watch.
Human Security Report Project 2010. Human Security Report 2009. New York, Oxford University Press.
Ingelaere, Bert 2010a. Do we understand life after genocide? Center and periphery in the construction of knowledge in postgenocide Rwanda. African Studies Review, 53 (1), pp. 41–58.
Ingelaere, Bert 2010b. Peasants, power and ethnicity: A bottom-up perspective on Rwanda’s political transition. African Affairs, 109 (435), pp. 273–292.
Jacques, Bahati Ntama and Beth Tuckey 2008. Rwanda and the War on Terrorism. Foreign Policy in Focus, 21 February.
Kimonyo, Jean Paul 2000. Revue critique des interpretations du conflit rwandais. Cahiers de Centre de Gestion des Conflits, (1), Université Nationale du Rwanda.
Kuperman, Alan 2004. Provoking genocide: A revised history of the Rwandan Patriotic Front. Journal of Genocide Research, 6 (1), pp. 61–84.
Lemarchand, Rene 2006. Consociationalism and power sharing in Africa: Rwanda, Burundi, and the Democratic Republic of the Congo. African Affairs, 106 (422), pp. 1–20.
Longman, Timothy 1999. State, civil society and genocide in Rwanda. In: Joseph, Richard ed. State, conflict, and democracy in Africa, pp. 339–358. Boulder and London, Lynne Rienner.
Mamdani, Mahmood 2001. When victims become killers: Colonialism, nativism and the genocide in Rwanda. Oxford, James Currey.
Marysse, Stefaan, Tom de Herdt and Elie Ndayambaje 1994. Rwanda: Appauvrissement et ajustement structurel. Cahiers Africains, (12).
Maton, Jef 1994. Développement économique et social au Rwanda entre 1980 et 1993: Le dixieme decile en face de l’apocalypse. Universite de Gand, Faculte des Sciences Economiques.
Melvern, Linda 2000. A people betrayed: The role of the West in Rwanda’s genocide. London and New York, Zed Books.
Moodley, Vadi, Alphonse Gahima and Suveshnee Munien 2010. Environmental causes and impacts of the genocide in Rwanda: Case studies of the towns of Butare and Cyangugu. African Journal on Conflict Resolution, 10 (2).
Mullen, Joseph 1995. From colony to nation: The implosion of ethnic tolerance in Rwanda. In: Igwara, Obi ed. Ethnic hatred: Genocide in Rwanda, pp. 21–33. London, the Association for the Study of Ethnicity and Nationalism.
Newbury, Catharine 1995. Background to genocide in Rwanda. Issue: a Journal of Opinion, xxiii (2), pp. 12–17.
Onyiegio, Michael 2010. UN Congo Report released amid protest from Uganda, Rwanda. voanews.com, 1 October.
Oomen, Barbara 2005. Donor-driven justice and its discontents: The case of Rwanda. Development and Change, 36 (5), pp. 887–910.
Pottier, Johan 1993. Taking stock: Food marketing reform in Rwanda, 1982–89. African Affairs, 92, pp. 5–30.
Pottier, Johan 2002. Re-imagining Rwanda: Conflict, survival and disinformation in the late twentieth century. Cambridge, Cambridge University Press.
Prunier, Gerard 1995. The Rwanda crisis 1959–1994: History of a genocide. London, Hurst and Company.
Prunier, Gerard 2009. From genocide to continental war: The ‘Congolese’ conflict and the crisis of contemporary Africa. London, Hurst and Company.
Reyntjens, Filip 1994. L’Afrique des Grands Lacs en crise: Rwanda, Burundi: 1988–1994. Paris, Karthala.
Reyntjens, Filip 1995. Rwanda: Background to a genocide. Bulletin des Séances de 1’Académie Royale des Sciences d’Outre-Mer, 41 (3), pp. 281–292.
Reyntjens, Filip 1996. Rwanda: Genocide and beyond. Journal of Refugee Studies, 9 (3), pp. 240–251.
Reyntjens, Filip 2004. Rwanda, ten years on: From genocide to dictatorship. African Affairs, (103), pp. 177–210.
Robinson, Peter and Golriz Ghahraman 2008. Can Rwandan President Kagame be held responsible for the killing of President Habyarimana? Journal of International Criminal Justice, 6 (5), pp. 1–10.
Sellström, Tor and Lennart Wohlgemuth 1996. The international response to conflict and genocide: Lessons from the Rwanda experience, Study 1: Historical perspective: Some explanatory factors. Odense, Steering Committee of the Joint Evaluation of Emergency Assistance to Rwanda.
Spears, Ian 2000. Understanding inclusive peace agreements in Africa: The problems of sharing power. Third World Quarterly, 21 (1), pp. 105–118.
Stedman, John 1997. Spoiler problems in peace processes. International Security, 22 (2), pp. 5–53.
Stettenheim, Joel 2002. The Arusha Accords and the failure of international intervention in Rwanda. In: Greenberg, Melanie, John Barton and Margaret McGuinness eds. Words over war: Mediation and arbitration to prevent deadly conflict, pp. 213–236. New York, Carnegie Commission on Preventing Deadly Conflict.
Storey, Andy 2001. Structural adjustment, state power and genocide: The World Bank and Rwanda. Review of African Political Economy, 28 (89), pp. 365–385.
Takeuchi, Shinichi 2000. Hutu and Tutsi: A note on group formation in pre-colonial Rwanda. In: Goyvaerts, Didier ed. Conflict and ethnicity in Central Africa, pp. 177–208. Tokyo, Institute for the Study of Languages and Cultures of Asia.
UN Panel 2001. Report of the Panel of Experts on the illegal exploitation of natural resources and other forms of wealth of the Democratic Republic of the Congo. New York, United Nations.
Uvin, Peter 1998. Aiding violence: The development enterprise in Rwanda. West Harford, CT, Kumarain.
Uvin, Peter 2001. Difficult choices in the new post-conflict agenda: The international community in Rwanda after the genocide. Third World Quarterly, 22 (2), pp. 177–189.
Van Hoyweghen, Saskia 2000. From human(itarian) disaster to development success? The case of Rwanda. Consortium for Political Emergencies (COPE) Working Paper No. 18, Centre for Development Studies, University of Leeds.
Voyame, Joseph, Richard Friedli, Jean-Pierre Gem and Anton Keller 1996. La cooperation suisse au Rwanda. Rapport du Groupe d’Etude par le DFAE.
Woodward, David 1996. The IMF, the World Bank and economic policy in Rwanda: Economic, social and political implications. Oxford, report for Oxfam.
World Bank 1991. Rwanda: Agricultural strategy review. Washington, D.C., Report No. 8704-RW.
World Bank 1994. Rwanda: Poverty reduction and sustainable growth. Washington, D.C., Report No. 12465-RW.
In July 1991, the words ‘et la Démocratie’ were added to the party’s name, thus turning the acronym into MRNDD. I mainly use the term MRND(D) to denote the party under both monikers.
The RPF also contained some prominent Hutu members who had broken with Habyarimana’s regime (Mamdani 2001:159–184).
It has long been thought that the most likely explanation of his death is that he was killed by members of the ruling élite itself, concerned at his alleged betrayal of the Hutu extremist cause (Prunier 1995:213–229). However, more recent evidence from (not necessarily fully reliable) RPF defectors suggests that it was the RPF which shot down Habyarimana’s plane (Lemarchand 2006; Robinson and Ghahraman 2008).
There is circumstantial evidence that the RPF was backed, to some extent, by the US government (Herman and Peterson 2010).
Jacques and Tuckey (2008) also claim that US support (the US is the largest single aid donor to Rwanda) is based on the interests of US companies in accessing DRC resources and Rwandan cooperation in the US-led ‘war on terror’.