Thursday, August 13, 2009

The Hutu's Willing Executioners?

Note: The following analysis is a guest post by Alex Gourevitch of Columbia University. It is based on a previously unpublished 2005 paper on the RPF and international intervention in the Rwandan civil war, with a freshly written introduction. This is a provocative argument, and while I don't agree with every aspect of the analysis (eg, I don't think the phrase 'well-intentioned' belongs in the lexicon of international relations), it is well worth engaging with. Though there is a bibliography supplied, there are also hundreds of footnotes that I haven't been able to include.

This paper was written in order to dispel the myth that the great mistake in Rwanda was the ‘failure to intervene.’ The reality is that the Rwanda disaster was largely a product of overlapping, uncoordinated, and overwhelming forms of international intervention at the diplomatic, political, and economic level. Moreover, some of these interventions are not even appropriately seen as ‘interventions’ so much as the kinds of international forces that can destroy a country as small and peripheral as Rwanda. One example of this kind of force is the dramatic decline in coffee prices in the late 1980s, which occurred after the collapse of the International Coffee Agreement (OPEC style price setting for coffee). However, it is important to point out that, despite this global forces that buffeted Rwanda, the impact of conscious interventions by everyone from the IMF and World Bank (imposing structural adjustment policies during a civil war) to the United States and France was more important. Of course these were not direct military interventions, but as the paper hopefully shows, the obsession with these forms of intervention obscures the various ways outside powers intervene in and determine the fate of a country. There was, of course, an internal political process, and the paper hopefully shows how internal dynamics became assimilated into broader international ones. To be clear, the argument is not that any of these intervenors tried to create a disastrous civil war, that culminated in the 3-month jacquerie of 1994. However, through a mixture of well-intentioned, ignorant, and downright duplicitous and insidious forms of intervention, they not only created the civil war, but sustained it, and radicalized it, to the point where Hutu Power was able to unleash the horrifying sequence of events.

A few further points should be clarified. First, none of this is meant to excuse those who participated in the killing. Instead, it is to point out that international actors do bear responsibility for the killings - but largely for having created those conditions through their interventions, not for having failed to intervene. Rwanda should be seen as a case not for endorsing more humanitarian intervention elsewhere, but for keeping it in radical check. It is time to abandon that myth of non-intervention in Rwanda, not just because, as many have noted, the killing was so swift, it is unlikely a military intervention could have done much to begin with, but more fundamentally because there is a deeper issue here. That deeper issue is the destructive consequences of the endless, ignorant (though sometimes well-intentioned), multi-pronged interventions already going on in weak, dependent countries.

Second, this paper does not cover all aspects of the Rwanda case. For one, it basically ends the narrative in 1994. Yet once the Tutsi-led RPF took power, it invaded eastern Congo, where millions of Hutus fled, and killed hundreds of thousands of civilians, all in the name of eliminating ‘genocidaires.’ It is important to note how little attention this event received (not to mention the killing in the early 1990s in neighboring Burundi of hundreds of thousands of Hutus by the Tutsi-led government), relative to the Rwanda massacre of 1994. No doubt that is because taking in all these facts disrupts the easy, morally simplistic, but ultimately dangerous, ‘genocide’ narrative. Another relevant factor this paper does not address is that the RPF, and its civilian form after the war, have governed Rwanda in authoritarian fashion ever since 1994, on the grounds of its supposedly unimpeachable authority as the terminator of genocide. Here again we can see how problematic seeing the whole situation in the black and white terms of genocide is actually dangerous and simplistic. There is one final fact to note, which is often forgotten in the genocide narrative that dominates the events of 1991-1994. Though many Hutus participated in the killing, according to the best estimates I had available when I wrote this paper, 90% of the killing was performed by 10% of the participants, while 10% performed by 90%, many of whom appear to have been intimidated, coerced and otherwise shamed into killing. The ‘Hutu’s willing executioners’ were not a uniform body of eternally hateful or racist killers, just waiting for the chance to unleash violence, and it would be wrong to see them, as a body, as responsible for the deaths, horribly and tragic as they were.

This paper is based on sources that have since been improved and expanded since it was written, and no doubt it misses out some key details and points of argument. However, hopefully it succeeds in the central task of cutting through the combination of moral simplism and interventionist mythology that still dominates our understanding of Rwanda.

The RPF and humanitarian intervention
On August 15, 2004 a contingent of 150 Rwandan soldiers arrived in Sudan as part of an African Union sponsored cease-fire monitoring mission. The mission was aimed at preventing what some observers were calling ‘genocide’ in Sudan’s Darfur region. The symbolism of having the Tutsi-dominated Rwandan military spearhead the international mission in what has been called ‘The Next Rwanda’ was obvious. Ten years earlier, a terrible slaughter of mainly Tutsi civilians had occurred in Rwanda, a tragedy widely considered to have been preventable by foreign intervention. The thought that the international community had failed to prevent a genocide has shadowed all subsequent violent conflicts in Africa and elsewhere. Indeed, Rwanda continues to be perhaps the strongest reference point grounding current arguments for humanitarian intervention in the Third World. This makes it all the more important carefully to scrutinize existing understandings of why and what happened in Rwanda ten years ago.

The predominant interpretation of Rwanda’s recent history has two dimensions. First: that the violence Rwanda was a genocide. In this view, the violence was the product of ‘ethnic conflict.’ Second: that the genocide was preventable by foreign intervention. The most significant ‘failure’ of the international community was its failure to act and its indifference to the state of Rwanda’s political affairs. I intend to challenge both aspects of this argument. It is inaccurate to label Rwanda’s conflict ‘ethnic violence’ and the most significant ‘failure’ of the international community was its intense and sustained interference in Rwandan affairs prior to 1994.


The view against which this paper will contend is more nuanced than that outlined above, so it is first necessary to lay out in greater detail the theoretical and explanatory terrain. It is no longer acceptable, in academia or elsewhere, to attribute the Rwandan violence to ancient ethnic hatreds. It is not even a plausible or interesting straw man against which to defend one’s own theory about what happened in the early 1990s. This is no doubt a positive step, and this paper should not be seen simply as a critique of the ancient hatreds view. However, most of the more sophisticated theories that have replaced the ‘ancient hatreds’ argument do not escape a view of the conflict in Rwanda as ‘ethnic violence.’ In fact, the ethnic interpretation of the Rwandan civil war has been reproduced in a more sophisticated form more in line with contemporary theorizing about culture. It is argued that a culture of obedience led the average Hutu automatically to obey orders to kill Tutsis. For instance, Prunier’s still influential account argues that “belief and obedience” were the main motivation. A “strong state authoritarian tradition going back to the roots of Rwandese culture” combined with an “equally strong acceptance of group identification” to produce a pliant and willing killing machine. Others argue that it was not a long-standing culture of obedience but the contingent social construction of a discourse of racial hatred the contributed to the massacres. By this argument, the discursive dehumanization of the Tutsis laid the groundwork for a ‘campaign of extermination’.

These constructivist and/or culturalist arguments often go hand in hand with ‘elite manipulation’ arguments that are superficially more political. The Economist, for instance, recently argued that “militias had to be organized, machetes bought and distributed, and Hutu peasants persuaded, through skillful propaganda, that all Tutsis were their enemies.” That is to say, the argument is ostensibly not based purely on ‘ideational’ factors but also on interests. Yet, while it is no doubt true that certain Hutu elites made every effort to stir up ethnic conflict, the ‘elite manipulation’ argument merely ends up arguing that the elites had ‘rational’ motivations while the peasants were driven by ethnic ones. The Economist, for instance, recently argued that “the small gang of Hutus who organized the genocide were rational men in pursuit of a rational – albeit evil – objective” who “harnessed ethnic hatred as a means to” maintaining their power. There is an empirical problem with this view. After all, the most virulently ideological racists were the journalists, academics, and political activists that constituted the core of the right-wing elite that organized the violence, while those who followed were driven by various mixtures of fear, need, and insecurity. However, more important than this specific empirical question are the conceptual issues at stake. If primordial or tribalistic arguments no longer dominate, many of the substitutes both in academia and the press are still based on the idea that what happened was a genocide and that the violence was ‘ethnic’ rather than ‘political.’ To be sure, one can still find primordialist accounts of the Rwandan civil war, and other conflicts for that matter. One need only think of the current understanding of the Sudanese conflict as one between ‘Africans’ and ‘Arabs.’ However, what is most significant about current thinking is that it has maintained an ethnic interpretation of the Rwandan case while distancing itself from the ancient hatreds view. Indeed, articles from The Economist past and present serve as a useful barometer of this shift. In 1994 it argued that “ethnic hostility in both [Rwanda and Burundi] stems from ancient rivalries between the tall Tutsi …and the Hutu majority,” whereas now it fully rejects the idea that “primeval ethnic antagonism” was the driving force but still insists that it was an ethnic conflict ending in genocide. It may not be primeval, but it is still a genocidal evil.

There is a theoretical question as to whether the ethnic and the political are inherently opposites. However, they tend to be seen as such. Ethnic violence seems to be the embodiment of irrational racially motivated violence, while political violence has a rational character, including at heart the pursuit of some interest. Put another way, ‘ethnic violence’ is conceived as violence between identity groups engaged in some kind of existential struggle to the death, where the highest purpose is the destruction of the other group. The genocide argument is the application of this ethnic violence notion to Rwanda: each individual decision and action is interpreted in terms of a final end which is the elimination of the Tutsis. This view of the Rwanda conflict has informed even those accounts that give more political explanations. So, for instance, even when it is observed that the old regime killed Hutu opposition opponents first, before the violence against Tutsis, this is seen as simply a necessary means to the end of eliminating Tutsis. The ‘elite manipulation’ argument similarly takes ethnic hostility as a given, dominant feature of the political situation, cynically ‘harnessed’ towards factional ends, and meticulously planned from the very beginning. The politicization of ethnicity itself is not itself explained or historicized. In whatever form, then, the political is only understood insofar as it is circumscribed by the ethnic.
This emphasis on the ethnic character of the violence has tended to shift the dynamic political relations between the different actors to the background. It also inverts the means and ends of violence. That is why, in this paper, I will attempt to right what is upside down. The ethnic needs to be situated within the historically specific political contexts that shape and define it. I will show that the ruling elite around Juvenal Habyarimana concocted the anti-Tutsi and civilian violence as part of a desperate, last resort strategy to shore up their political power against various challengers. This strategy was chosen, but they were also pushed towards it by various forces and actors beyond their control. Likewise, their ability to pull off this strategy relied on a number of contingent factors beyond their control, and that cannot be explained away merely as ‘elite manipulation.’ Rather, it was the internationalization of political reform and the peace process amidst a burgeoning social crisis that led to a radicalization of the civil war and its bloody culmination.

The Theoretical Terrain: What Exists
There are already a number of political explanations of the Rwanda affair, which I will draw on to form my own account. For instance Barrie Collins (2002), David Newbury (1998) and Jack Snyder (2000) have, in different ways, focused on the importance of the democratization process in triggering the violence. Catharine Newbury (1988) has done the most extensive research on the impact of colonialism on the production of modern ethnic identities in Rwanda. Mahmood Mamdani (2001) has in many ways extended Newbury’s historically situated approach to show how ethnicity in Rwanda continued to change after independence, and he has drawn attention to the way a political explanation of the violence can be built on the fact of widespread fear and insecurity as opposed to culturally determined ethnic motivations. John Pottier (2002), too, has noted the way the politicization of ethnicity has changed with shifting political alignments over time. Alan Kuperman (2004) has drawn attention to the way the militaristic political motivations of the Rwandan Patriotic Front (RPF) contributed to the radicalization of the civil war. Scott Straus (2004a, 2004b) has shed important light on both the demography of the violence as well as pointed to the critical importance of intra-Hutu political competition before and during 1994. Collins, Newbury, Mamdani, Bruce Jones (1995) and others have pointed to the way external political pressures radicalized the conflict, while Peter Uvin (1998), Michel Chossudovsky (1998), and David Smith (2002) have paid close attention to the influence of foreign interference in the Rwandan economy on political dynamics.

However, most have focused on only one or two aspects of the complex political dynamics, rather than relating them each to the other as whole. For instance, Kuperman does not situate the development of the RPF’s political strategies in relation to the shifting priorities of foreign donors and the crisis of the Habyariman regime. Straus makes little mention of the RPF’s militarism and willingness to break cease-fires. Snyder’s emphasis on democratization describes the political dynamic as one between anti-democratic elites and democratic masses, without including either the anti-democratic RPF or the donor countries as politically significant actors. Also, for the most part, the various analyses have not clearly placed the ethnic dimension in the broader context of the crisis of legitimacy and struggle for power amongst the various groups that emerged as the Habyarimana regime weakened at the end of the Cold War. My contribution, then, is to try to draw on and synthesize these various politically oriented explanations.

Theoretical Terrain: Political Explanations of Ethnicity
There are some ostensibly political explanations that I shall reject. For instance, the ‘ethnic security dilemma,’ suggested as a general theory by Barry Posen does not explain what happened in Rwanda. First, the state did not disintegrate, at least on in the way traditionally understood by the ethnic security dilemma. As many have noted, it was the continued salience of coercive state apparatuses that allowed the Hutu Power to organize the violence. Second, as I shall show, the violence had more to do with political struggles between different elements of Rwandan society and with foreign intervention than with the collapse of state structures per se. Third, the Tutsi-dominated RPF invaded long before any real threat developed to Tutsis, nor did the RPF represent in any meaningful way Tutsi interests within Rwanda. Thus it cannot be said that the civil war began in response to ethnic insecurities in the wake of crumbling state structures. In fact, the RPF’s behavior during the civil war was not oriented towards reducing the threat of violence to Tutsi civilians or protecting them, as the ethnic security dilemma would predict, but rather oriented towards winning state power through force. The other ‘political’ explanation of the violence that will be rejected is the ‘elite manipulation’ argument because it exaggerates the strength of the right-wing elite, and fails to explain the conditions in which a Hutu Power strategy of anti-civilian violence either emerges or could have been effective. As discussed above, the ‘elite manipulation’ argument, though somewhat conscious of the contingent nature of ethnic identities, leaves the ethnic violence argument intact, and does not adequately explain the political forces that determine the character of the violence.

The political explanation for Rwanda given here will be more historical in nature. A number of commentators note that the character of ethnic relations in Rwanda has changed dramatically over time. In particular, the nature of ethnicity and the dynamics of ‘ethnic politics’ has been shaped by the underlying political economy of Rwandan society. The shifting social relations both between peasants and the ruling class, as well as conflicts within the ruling class itself have come to a head at key moments within Rwandan history. At each moment, what were long-term historical processes have suddenly and rapidly transformed the nature of ethnic politics in Rwanda. For example, as we shall see, the very idea of differentiating the ‘ethnic’ from ‘class,’ often assumed as a meaningful distinction throughout time, only emerged in the post-independence period with the creation of Hutu elite centered around control of state offices. More significantly, the politicization of ethnicity has depended on underlying power struggles between the different class elements as well as on the actions of foreign powers on Rwandan soil.
That the distinction between ‘Hutu’ and ‘Tutsi’ has been given shape by historical, social factors is obvious by comparison with the alternative. The physiological differences between ‘Hutus’ and ‘Tutsis’ are slight, not the least because of a great deal of intermarriage. Other markers of difference are also hard to find. As Uvin notes, at least by the 1980s, other markers of difference were are also hard to identify:

The integration had gone far: they spoke the same language, believed in the same god, shared the same culture, belonged to joint clans, and lived side by side throughout the country. There are few cases anywhere in the world of different ethnic groups sharing so many of the same characteristics.

If ethnicity has therefore been a social product, its social meaning has changed over time. The Rwabugiri kingdom of the late 19th century was a ‘Tutsi’ kingdom, but the distinctions between ‘Hutu’ and ‘Tutsi’ mainly had to do with cattle ownership, and one could become a ‘Tutsi’ by owning enough cattle (or become a ‘Hutu’ by losing them.) In addition, as Catherine Newbury notes in her superb history of Rwanda, other kin-based relations cross-cut ‘Hutu’ and ‘Tutsi’ relations up through the early colonial period. As she and many others note, it was under colonialism that ‘Hutu’ and ‘Tutsi’ developed into strict, racialized classifications. The Germans, and then later the Belgians, imported rigid, biologically based notions of racial difference, appropriate to the highly stratified and immobile political and economic hierarchies they sought to establish under colonialism. Especially during late colonialism, when the Belgians used Rwanda (then still ‘Ruanda-Urundi’) to produce for the mines in the Congo and to produce coffee for export, the Rwandan economy was dramatically changed. The intense introduction of market relations transformed many of the kin relations, and precipitated the emergence of incipient modern class formations. At the same time, the Belgians turned the ‘Tutsis’ into a privileged oligarchy in control of land, labor and the administrative apparatus, all justified by their racial ideology. Many have suggested this is when the modern notion of ‘Hutu’ and ‘Tutsi’ took hold and cemented into Rwandan consciousness. However this implies a somewhat ahistorical conception of ethnicity as form of consciousness that locks in a particular moment. It is more appropriate to see this period as the point at which a specifically ‘Hutu’ identity first emerged as a consciousness of general subjection under a class dominated state distinct from more regionally based peasant identities.

This consciousness would change and develop further after independence. For while this period is the first in which ethnic identities become something more than class categories, they did not simply ossify into timeless culturally specific modes of perception and interaction. We shall see later how the politics of ethnicity continued to change after independence, especially with the creation of a state-based Hutu elite and with the emergence of regional political competition between northern landowners and southern and central state bourgeoisies. The purpose of the above discussion has simply been to show how the nature of ethnic politics has changed over time according to changes in socio-political relationships. In Rwanda, the redefinition of ethnicity has occurred in periods of political instability, when the nature and basis of Rwandan society was called into question, and when there was a decisive shift in Rwandan social relations. The move from Rwabugiri monarchy to Belgian colonialism, with the concomitant penetration of market forces and a modern state apparatus, was one such crucial moment. Later I shall identify three other key political moments that lead to similar transformations: The 1959 Hutu Revolution and the early period of consolidation of the southern and central elite’s control of the state; the 1973 coup that brings Habyarimana and the northern landholders to power; and the 1990-94 internationally managed ‘democratization’ process that eventually results in the RPF’s military and political victory.

Theoretical Terrain: Democratization and Internationalization in the early 1990s
While this paper will briefly discuss the shifts in Rwandan politics from independence to the early 1990s, the purpose of the historical sections is mainly to set the stage for explaining the events of 1994. Such historical detail is necessary because the political struggles that emerge in the early 1990s are a product of short-term changes mediated by historical processes. What, then, are these short-term shifts in political relations that transform Rwandan society? There are two theories of the immediate causes of the Rwanda debacle, which can be termed ‘democratization’ and ‘internationalization.’ The democratization thesis comes from a set of arguments one can find in the international relations literature as well as in historical studies. Some have argued that democratization pressures produces nationalist counter-reactions by the ruling elite who seek ways of diverting mass energies away from challenges to the status quo. That mass pressures can spark nationalist violence is a recurring theme in sociological and historical accounts of war. Eric Hobsbawm has argued that the rise of right-wing nationalism in Europe between about 1870 to 1914, was a response to the emergence of an organized working class demanding a greater share of political and social power throughout Europe. In a similar vein, Gabriel Kolko argues that the nationalist World Wars were driven by European elites attempting to divert and forestall the pressures within their society to give up existing privileges. In particular, Kolko notes that Hitler pursued (and German elites followed) irrational war planning and strategies because he “preferred courting defeat abroad to revolution at home.” Jack Snyder argues that in many countries, past and present, authoritarian elites have responded to democratic movements with nationalist gambits, but these response have taken various forms and achieved different degrees of success based on institutional factors. In societies with weak civic institutions, the elites are successful, and right-wing nationalist violence emerges in some ethnic or counter-revolutionary form. Snyder explicitly includes Rwanda as an example of where democratization has led to reactionary nationalist violence because of weak institutions.

While I shall argue that ‘democratization’ pressures did lead to the civil war and the re-emergence of ethnic politics in Rwanda, as the theory stands, it is inadequate. First, there are some crucial differences between the European experience with democracy and nationalism, and Rwanda’s (not to mention other late 20th century Third World countries’). The most important is that, whereas the major pressure for democracy came from within Europe in the late 19th and early 20th century, in Rwanda, the pressure came from the outside. The major theme of this paper is that the major reason why the Rwandan ‘democratization’ process ended in tragedy was because the demands were shaped by external priorities, and the process removed from most Rwandans. In fact the international program for Rwanda focused mainly on certain formal attributes like a relatively free press, multipartyism, elite power-sharing and elections. Indeed, ‘democratization’ went hand in hand with increasing foreign control over the Rwandan economy, in the form of structural adjustment policies over which the average Rwandan had absolutely no control. The difference in strategy and focus between the popular classes and the elite opposition, as well as the incentives it created to pander to international favor, rather than establish a political base in Rwanda, were two aspects of the democratization process that led to the initiation and radicalization of the civil war. This does not mean popular pressures were unimportant. As we shall see, they were. But they mainly emerged in response to political changes in Rwanda, including international interference in these changes. In short, it was not merely democratization but the internationalization of the democratization process that led to the radicalization of Rwandan society and the politicization of ethnicity.

The full impact of the internationalization of political change in Rwanda cannot be understood without noting another key difference between Rwanda and most other ‘democratizing’ countries. Democratization threatened not just the old regime but also the political future of a second ‘elite’: the Tutsi dominated Rwandan Patriotic Front. As a relatively unknown expatriate group, with at best a moderate expectation of support from the small local Tutsi population, the RPF stood little chance of winning any elections; and yet by the early 1990s, they had become committed to regaining power in Rwanda. This made the international decision to force democracy on Rwanda even more disastrous as it pitted two groups – the ancien regime elites, and the RPF rebels -- against each other. In pursuit of their interests both of these groups attempted to strike a balance between pleasing foreign powers and avoiding any political advance towards democracy. The consequence was a civil war conducted amidst extensive foreign interference, of which the killings of 1994 were the bloody culmination.

It is therefore inadequate to argue that the problem was democratization in Rwanda in the context of weak civic institutions. We cannot understand why civil war ensued, nor why the various groups made the decisions they did, without noting that political change was forced on Rwanda and managed from the outside. In addition, we cannot explain the start and later radicalization of the civil war without noting that there was not one but two elite groups who found the possibility of electoral democracy threatening and who responded in accordance with their interests to the rapidly changing Rwandan political context. The domestic political struggles between regionally differentiated land-owning and state-based Hutu elites, and between the peasants and the ruling class, were a product of historical changes in Rwandan society. Their assimilation into an internationalized democratization process precipitated and radicalized the civil war, bringing an otherwise marginal Hutu racist political tendency to the fore.

Periodization of the Conflict and Plan for the Paper
Given the theoretical premises of this paper, it will proceed historically. The paper cannot show how the ‘ethnic violence’ was in fact political, arising from the internationalization of the democratization process, without first providing the historical context for the underlying political struggles. Therefore, the paper will be broken into a brief section on the First and Second Republics, and then a more extended section on the 1989-1994 period. Also, it should be noted that ending this paper in 1994 in one sense contradicts its basic premise. Accounts that end with the violence of 1994 tacitly (and often explicitly) accept the premise that the violence against Tutsis marked a culmination or end point in a period of Rwandan history defined by a bloody ethnic conflict or genocide. However, the civil war did not end after that period, although it changed course. The Rwandan invasion of the Democratic Republic of the Congo, and attacks on Hutu refugee camps that killed tens perhaps hundreds of thousands of Hutus, in 1996-97 was clearly a continuation of the conflict that emerged in the early 1990s. However, due to the time and space limits of this paper, it is impossible to include a discussion of these events. In addition, it is still possible to give a political explanation of the Rwandan civil war up through 1994 without accepting the premises of an ethnic interpretation of the events.

One final word is in order before proceeding to the concrete reconstruction and explanation of Rwanda’s past. Although many of these issues sound theoretical they have clear political implications. The ethnic interpretation of the Rwandan tragedy, and the view that the main failure was international indifference, supports a certain view of intervention not only in Rwanda but elsewhere. It tends to cast the affairs within Rwanda, and other countries, as beyond their control. When seen as ethnic conflicts they are thought to be irrational, without a civilized or identifiable purpose, and irresolvable without a ‘moderate’ and ostensibly disinterested third party. The international role is seen as positive, rationalizing, constructive, and even ethical. This view of international involvement is only sustainable if one ignores the truly destructive effects of international interference in Rwanda prior to and during 1994. On the flip side, the destructiveness of the civil war had less to do with intractable ethnic differences and more to do with the way the internationalization of political change altered the political calculations of the different groups involved, and pushed some towards extreme measures. If there is a political implication of this analysis, then, it is that the lesson of Rwanda is not that intervention was necessary, but that it was to a great degree responsible for what happened. In a broader sense, this analysis is also a warning against facilely labeling all anti-civilian violence or intense violence ‘genocide’ or ‘ethnic violence.’ Such labels often have the effect of simplifying complex situations. They forestall more thoughtful and critical discussion of the issues at stake in favor of blind action.


Rwanda achieved its political independence on the back of the ‘Hutu Revolution’ of 1959. The Revolution was organized by a Hutu nationalist party called PARMEHUTU against the Belgian established Tutsi ruling class, and set the contours of post-independence politics. The PARMEHUTU was led by a small Hutu elite, whom the Belgians supported against the more populist APROSOMA mainly because the latter was directly anti-colonial and more left-wing. In fact, the Belgian involvement in Rwandan independence followed the general pattern of an attempt to manage as extensively as possible the decolonization process. McNulty, for instance, notes that “there was no real decolonization of sub-Saharan francophone Africa” but rather a “limited concession of autonomy to French fostered elites” - or in this case Belgian fostered elites. The Belgians believed they would be better able to influence a post-independence government more focused on eradicating Tutsi privilege than Belgian presence, and with more conservative, at times monarchical, political aspirations than APROSOMA. The Belgians even were involved in some of the violence in 1959, and between 1959 and 1962 engineered the political emergence of a set of Hutu chiefs close to Gregoire Kayibanda, leader of PARMEHUTU, and secured their political dominance over the newly independent government.

The post-independence period was a time of great insecurity as the PARMEHUTU attempted to consolidate its rule against internal and external challenges. In order to build support amongst Hutu peasants, Kayibanda abolished the colonial institution of ubureetwa or forced labor, and appropriated Tutsi lands for redistribution. It also replaced the authoritarian colonial era Tutsi-dominated politics with a series of national and local elections. However, in what was the most important change in post-independence society, Kayibanda used PARMEHUTU political power to foster the development of a state-centered Hutu ruling class alongside the now reduced Tutsi elite. This new elite was drawn from the central and southern regions of Rwanda, to the exclusion of the northern landlords. Conscious of the peasant question without seeking to resolve it, and aware that other parties made appeals to the poor, Kayibanda attempted to divert any class appeals into anti-Tutsi politics. The Tutsi were legally constituted as an alien presence in Rwanda and excluded from participation in the political process. A series of small-scale Tutsi restorationist raids in the early 1960s provided the context for Kayibanda to kill thousands of Tutsi in a series of reprisals during the mid-1960s. The continued politicization of ethnicity throughout the 1960s was in fact a sign of the Kayibanda regime’s political insecurity, as it sought to parlay the popularity of the Hutu Revolution into continued support and to ward of political challenges. It also introduced a new element into the dynamics of ethnic and class conflict in Rwanda. With the growth of a Hutu ruling class, centered around the control of state power, new conflicts within the Hutu ruling class emerged. Simultaneously, ethnicity came to have decreasing relation to socio-economic cleavages or political power, and the politicization of ethnicity revolved less and less around the contemporary order and more around the arousal of fears about the restoration of a past order. Contemporary inequalities, then, were hidden behind the fear of a return to the inequalities of the past.

Kayibanda’s political strategy was from the start short-termist, as his conservatism meant he did not directly address popular concerns, and his regional favoritism created enemies in the north. By the late 1960s an increase in unemployed educated Hutus created social unrest, as did the persistence of Tutsis in positions of social power like churches, colleges, banks and industry amidst continued Hutu poverty. At the same time, given the privileges enjoyed by the new state-centered Hutu elite, it was becoming increasingly evident that ‘ethnicity’ did not correspond to social and political privilege. Kayibanda had no real answer to the emerging conflicts within the middle class, nor increasing dissatisfaction amongst the poor, and reacted defensively to the instability. As Mamdani notes, the re-emergence of racial tension and the First Republic’s political paralysis

brought to the surface the tension both within power and within society: the former between Hutu of the north and those of the south, and the latter between the poor and the rich.

The massacre of hundreds of thousands of Hutu by a mostly Tutsi army in early 1973 in neighboring Burundi raised tensions even more. The northern elite capitalized on the instability and mounted a coup that brought Juvenal Habyarimana and the Second Republic to power.

The Habyarimana regime, drawn mainly from landholders and other conservative, elite social elements was at heart authoritarian yet initially popular and capable of defusing ethnic tensions. These were in fact two sides of the same coin. It was only by building on its ability to pacify racial tensions, and to secure some national unity behind a period of initial growth, that it managed to consolidate and preserve its otherwise right-wing politics. As Mamdani notes, the coup was initially met with relief, as there was an increasing sense that ethnicity was not the real problem so much as disparities of economic and political power. The Habyarimana regime made a show of deracializing the Tutsi, politically and legally constituting them as a Rwandan ethnicity, rather than a foreign race. They were allowed to participate in politics, although to a limited degree, and a Tutsi served in Habyarimana’s cabinet. The effect on Hutu-Tutsi relations was dramatic. As Mamdani notes “no major anti-Tutsi political violence was reported from Rwanda between the time Habyarimana came to power in 1973 and the onset of the war with the RPF [Tutsis] in 1990.” Simultaneously, the Habyarimana regime pursued an affirmative action program that redistributed land, and created state quotas for education and employment, away from Tutsis and to Hutus. In practice, this attempt at justice was simultaneously a way of consolidating the north’s hold on power, as the policies were aimed at redressing not merely ‘ethnic’ injustices but also ‘regional’ ones.

Deracialization of Tutsis and the affirmative action program were the means by which Habyarimana attempted to execute his ‘Moral Revolution’, but it was also cover for a dramatic shift in the balance of power within the Rwandan elite and the rise of an authoritarian class. Catharine Boone has noted that in many post-colonial African countries, colonial economic and political institutions persisted because they were the most secure way for the post-independence elite to secure their rule. This elite was often cultivated by colonial powers and shared their interest in maintaining the status quo. Therefore, they tended not to dramatically reform existing economic arrangements or introduce new technologies, because this would create new classes that would challenge the status quo. A more secure strategy was to maintain and reproduce the exploitative and authoritarian institutions inherited from the colonial period. In Rwanda this is oddly more true of the Second Republic under Juvenal Habyarimana than the First Republic under George Kayibanda. While the latter maintained the colonial coffee marketing boards as a lever of accumulation, he banned the hated colonial practice of forced labor, eliminated customary controls over land, and brought into being a new state-centered Hutu middle class. Habyarimana, however, in many ways was closer to the model Boone describes. The Second Republic reinstated forced labor, both because the Habyarimana regime was dominated by landlords that sought to preserve state domination of the peasantry, and because it was a way of earning foreign exchange for the state budget without having to introduce more efficient forms of exploitation like peasant wage-labor and advanced technology. Such social and technological innovations would have brought into being social classes that could have undermined the landlords’ political and social dominance. The Habyarimana regime was only able to reinstate forced labor in the context of the temporary legitimacy it earned by calming the political tensions of the First Republic and because, initially, the use of forced labor went hand in hand with economic growth. In the 1970s, food output per capita increased in spite of a population explosion, and forced labor was used to build roads, clear land, and for other infrastructure projects.

In fact, these initially positive seeming developments were grounded in an untenable political economy. First of all, the state was increasingly used simply to consolidate the north’s dominance of Rwanda, not merely against the peasantry but the southern and central elites. At one point, Gitarama and Kibuye, two central prefectures especially favored under Kayibanda, received just 1 percent of rural investments, whereas three northern and a northwestern prefecture received almost 90 percent. As Jones notes, “The Habyarimana regime was in fact a clan-based northern Hutu regime that was as discriminatory against Hutus from southern Rwanda as against Tutsis.” The only way for the Rwandan economy to grow without dramatically changing the underlying social relations was through increased exploitation of the peasantry – either by changing the price the marketing boards paid for coffee, by increasing the number of days they had to work for the state, or by appropriating and increasing the amount of land under cultivation. In the long run this strategy could not work. There are limits to the extent a state can exploit its peasantry, the surpluses were not reinvested in new kinds of production or better technology but increasingly siphoned off, and the economy remained dependent on the vicissitudes of the world prices for the few commodities it produced, mainly coffee.

In fact, as the world economy slowed down in the late 1970s, the Rwandan economy felt the impact in decreased demand for exports, and the initial grace period gradually transformed into a political problem for the Habyarimana regime. The Second Republic had not been able to prevent the market from changing Rwandan society, and the average (Hutu) peasant had seen a real erosion of his social position. Land-ownership had become increasingly unequal and concentrated. Whereas in the 1960s the average peasant owned 2 hectares, by the mid 1980s it had declined to 1.2. By then, more than half worked a single hectare or less, while only just over a quarter owned more than 1.5 hectares. Tenantship and landlessness had increased so that “nearly half of all farms were rented by otherwise landless tenants…and a wealthy minority (16 percent) owned nearly half the land.” According to Smith, when asked in 1982 “whether they wanted their children to become farmers, nearly four out of five Rwandan peasants said no.” For a regime whose rule was built around the notion of a ‘peasant republic’, and which defended forced labor in the name of building said republic, this was clearly a budding legitimacy crisis.
The Habyarimana regime had reached its political limits by the early 1980s but it managed to forestall change through a parallel set of defensive, reactionary measures. First, it restructured the state in a defensive move to concentrate power almost exclusively in the north. The developmental overtones of the 1970s gave way to patronage politics in administrative positions, educational opportunities, military offices, land distribution, state banks, and government contracts. Where some of these benefits had been spread more widely in the 1970s, the Rwandan ruling class was now so tightly composed of personal favorites almost exclusively from Habyarimana’s own clan that it acquired the name ‘akazu’ or little house. As Jones notes, “the akazu was a classic oligarchy.” Second, it sought to it sought to further consolidate itself by securing more support from abroad.

The Rwandan state, like many post-colonial states, was in fact never fully independent in the way rich, industrial states are. Recall McNulty’s comment that “there was no real decolonization of sub-Saharan francophone Africa” but rather a

limited concession of autonomy to French-fostered elites, and the institutionalization of patron client relationships, whereby the patron’s influence hinged on the client’s survival, while the client’s survival often depended on the patron’s protection. Independence was conceded, but only to…small, artificial and weak [states]. Control was retained of these new states’ economies, currency, and of their foreign and defense policy.

As with Boone’s observation about the preservation of colonial institutions, McNulty’s general thesis is more appropriate to the Second than the relatively autonomous First republic. The formerly Belgian territories, including Rwanda, were absorbed into the French sphere of influence in the mid 1970s, and the Habyarimana regime in particular became a close ally of the French. Increasing political illegitimacy and declining economic fortunes also drove the Habyarimana regime to look for help outward, as the only other way of improving the economic and political situation within Rwanda would have been revolutionary political and social change.
In the 1980s, then, the Habyarimana became extremely dependent on foreign political and economic support. As Uvin notes, “since the early 1980s, the state faced a persistence balance-of-payments deficit, as well as an increasing dependence on foreign aid.” This support came not just from the French, but also from the IMF, World Bank, and various development agencies eager to find a poster child for their newly minted structural adjustment programs. So much money flowed in that “from 1982 to 1987 foreign aid financed more than two-thirds of all public investment.” The World Bank and IMF considered Rwanda a model developing country, mainly for having implemented structural adjustment programs, beginning in 1982, in exchange for much needed cash. In fact, all the cash did was keep a corrupt and illegitimate regime afloat over a stagnant economy and simmering social tensions. The donor institutions simply refused to acknowledge that, far from being a developing country, Rwandan social relations were tremendously backward – still reliant on forced labor, small holder farming, single commodity export (coffee), and authoritarian, single party politics. The fetish of a rising GDP concealed this underlying socio-political stagnation. Likewise, donor countries remained oblivious to the way in which they simply perpetuated and deepened popular antipathy to a regime that nonetheless managed to use loans to forestall political challenges. The French continued to maintain military alliances and financial support; the US declared Habyarimana a valuable ally; the IMF and World Bank, excited to have a pliant and willing partner, continued to pour in money.

Yet while the Habyarimana became increasingly dependent on external support, foreign involvement also began to lead to its undoing. Structural adjustment meant the increasing penetration of the market that promised to disrupt the stagnant but moderately stable status quo. Indeed, one of the strange facts about structural adjustment was that it was bad for nearly everyone involved. It meant a transformation of Rwandan society during the 1980s that would inescapably result in the emergence of political pressure on the unpopular Habyarimana regime. The elimination of marketing boards and price controls forced millions of peasants to suddenly adjust to the demands of the market without any transitional measures to assist in the adjustment, meaning many started to find themselves unable to support their families through traditional small-scale production. Landlessness and poverty increased, even as wealth became increasingly concentrated. Meanwhile, the Habyarimana regime began to demand more forced labor on coffee plantations in order to earn more foreign exchange. The Second Republic was rapidly squandering what little domestic legitimacy it had left.

Already we can see how the progression of events leading to the crisis of the early 1990s do not fit the ethnic interpretation of Rwandan politics. While the ‘ethnic violence’ view tells only the story of episodic fighting between ‘Hutus’ and ‘Tutsis’ over the course of Rwandan history, it is unable to describe or explain the changing meanings of these political categories. It also overlooks the way in which the most decisive and constant political conflicts were not between ‘Hutus’ and ‘Tutsis’ but between different and shifting social classes within Rwanda. By the 1970s, these cleavages existed along two lines. First, it operated within the regionally divided Hutu ruling class, both sides of which were highly dependent on control of the state for the maintenance of their social dominance. Second, the ruling class played a risky game of trying to mobilize lower class, peasant support while maintaining socio-political inequality, often through the politicization or depoliticization of ethnicity. Alongside this emergence of post-independence class politics, goes a second trend: the increasing international influence over domestic politics. This became particularly acute as the Habyarimana regime attempted to hold on to power amidst an intensifying legitimacy crisis and the world-historical social changes brought on by the end of the Cold War.

In a process that quickly took on a life of its own, the more the Habyarimana regime sought to protect its privileges, the more dependent on foreign support it became. Rwanda’s foreign dependence was guaranteed in 1989, when world prices for coffee and tin, dropped precipitously, with coffee prices falling 50% in the summer of 1989. Since the two exports accounted for nearly all of Rwanda’s foreign exchange, the price collapse had widespread consequences for the livelihoods of most Rwandans and left the government unable to service foreign debts -- “Rwanda was now firmly in the grip of forces it did not control.” GDP fell 5.7% in 1989 and another 2% in 1990, with the economy getting even worse after the civil war started. By 1990 foreign assistance provided more than 70 percent of public investment, and “supplied the fuel on which the machinery of the state ran.” The international financial institutions, aid agencies, and national governments, however, were so wrapped up in pursuing their own objectives in Rwanda that they remained ignorant or indifferent to the way their aid was propping up a corrupt, authoritarian, and unpopular government. Smith notes that

as late as 1989-1991 – when the coffee market and the economy were in crisis, war and famine had broken out, and ethnic tension and repression had worsened – the World Bank praised the Habyarimana regime for its human spirit and prudence.

In fact, the IMF and World Bank proposed another structural adjustment program to deal with the economic crisis. In 1990 the cash-strapped Habyarimana regime agreed to implement one. This entailed not only liberalization and privatization, but also a devaluation of the Rwandan franc, which dramatically increased the cost of living for the Rwandans.

The most important effect of the new round of structural adjustment was not its economic but its political impact. It called the legitimacy of the regime into question, as the reforms so obviously ignored the needs of most Rwandans, and reflected the will of foreign creditors:

the question ‘who rules Rwanda?’ became pertinent when Habyarimana, under pressure from the European Economic Community (EEC), agreed upon a Structural Adjustment Programme…in the wake of the crash in coffee incomes.

The Habyarimana regime was increasingly at odds with most Rwandans. While the state insisted, in line with the IMF’s plan for rehabilitating Rwandan finances, that peasants plant more coffee trees in order to increase meager export earnings, the peasantry was moving in the opposite direction. Peasants sought to replace unprofitable coffee with crops like maize, sorghum, wheat and rice that could be eaten if they fetched too low a price on the market or none at all. The Habyarimana regime also scaled back public sector jobs, raised fees on health, water, and school services, which antagonized many Rwandans and “exacerbated the already severe pressure on most farm households.”

With the Habyarimana regime suddenly weakening, the political dynamics of the past forty years reasserted themselves. The southern and central businessmen, professionals, and politicians began to form political parties to press for political reform and a greater share of power, while the peasants and urban poor began to demand, if in disorganized ways, remedies for increasing hardships. Subtler, end of the Cold War changes, also made it clear that the authoritarian, client regimes of the Cold War were now in an increasingly insecure position domestically and internationally. We shall never know what would have happened if these forces and new political movements had been allowed to run their course on their own. For these emergent domestic pressures were quickly subsumed under a demand for political reform by former international backers of the Habyarimana regime. As Uvin notes “Following the end of the cold war, the international community suddenly rediscovered a strong attachment to democracy and put pressure on the regime to democratize.” Instead of retreating from managing Rwanda’s economy, various international agencies and foreign governments decided to transform its political structure as well. These different external actors wielded an influence over Rwandan society that arguably had not been seen since the colonial period. In the end, they were quite possibly more destructive.


1990 - From Elections to Invasion

Initially, it was the French whom proposed the ‘democratization’ process, and it took the form of urging Habyarimana to deal with the Rwandan Patriotic Front (RPF) and to plan for elections. The RPF was a military-political organization of expatriate Tutsis based in Uganda that had transformed in late 1987 from the peacefully oriented politico-cultural Rwanda Alliance for National Unity into the militaristic RPF. Around this time the RPF, which was virtually unknown within Rwanda, had embraced a “‘zed option’—the use of military force…to return to Rwanda.” By 1988 it was an “open secret” that the RPF was considering an invasion of Rwanda, but “the RPF leadership had yet to make the final decision for invasion.” The motives behind the embrace of the ‘zed option’ were complex, initially having to do with Ugandan President Yoweri Museveni’s increasing marginalization of his former allies, but also with the decision by various donor countries that Rwanda should democratize and extend the right of return to Tutsi refugees. For although the RPF claimed it simply wanted repatriation, they had little reason to support democracy. With almost no following in Rwanda, they had little chance of winning a significant share of political power through elections. The only other option was power by force.
Meanwhile, it looked increasingly likely that elections would come to Rwanda. Mamdani notes that regardless of Habyarimana’s resistance to real change, he was constrained by the demands of his foreign backers:

as late as January 1989, Habyarimana considered any political change feasible only within the one-party system...only a year and a half later, on 5 July 1990, he agreed to the necessity of a separation between the party and the state, possibly within the context of multipartyism.

The conception of ‘democratization’ as multiparty elections, free press, and right of return for Tutsis, betrayed how deeply political priorities were shaped by external rather than domestic interests. That the most urgent needs of most Rwandans, such as land reform and food scarcity, were hardly discussed, and that substantive issues of economic planning remained in the hands of international experts from the World Bank and IMF, makes it difficult sincerely to label the process as ‘democratic’ in any meaningful sense, unless one subscribes to the meaningless formalism that elections alone make a democracy.

Although Habyarimana was reluctant to follow these international demands for fear of losing power, he initially appeared to conform. He agreed to a right of return for Tutsi refugees, to liberalize the press, to hold elections, and to recognize newly forming political parties. However, these moves towards ‘democracy’ were cut short by the RPF invasion on October 1, 1990. The RPF claimed that Habyarimana was not acting in good faith and never fully intended to implement the promised measures. However Kuperman notes that RPF officials have privately admitted “they still would have launched the invasion” even if Habyarimana had conceded to the refugee demands. Their goals by this point included an overthrow of Habyarimana and claiming state power, something they could not achieve through the electoral process. The invasion, then, seems to have been meant to pre-empt any further moves towards electoral democracy, and with the hope of using military force to win power.

It is at this moment that the internationalization of political reform began to show its disastrous effects. Not only had the democratization process inadvertently triggered the invasion, foreign governments also sent mixed messages that de facto encouraged RPF militarism and further weakened the Habyarimana regime. With the exception of the French, foreign donors remained surprisingly silent about the RPF invasion. This set the tone for the RPF, who were to receive repeated signals that their use of military force would not be held against them in spite of their anti-democratic means, and could even end up being beneficial to them in future political negotiations. As Collins notes

Having successfully complied with the terms of Western [political] conditionality, the Rwandan government was shocked that the invasion was not condemned by the international community outside of France. The suspicion that the RPF enjoyed discreet American approval in addition to significant Ugandan support placed a question over the sincerity of the stated motivations behind political conditionality.

At the same time, while the Habyarimana regime responded to the invasion with a series of mass arrests, it was also forced to continue with the script the international community had laid out for political liberalization. This included negotiating over including opposition parties in government, setting a new timetable for elections and expanding press freedoms. Indeed, it is a sign of just how utterly dependent the Habyarimana regime was on foreign donors that he took the step of expanding certain civil liberties in the midst of war, something nearly unheard of in wartime. In the midst of this, the existing unease amongst Hutu peasants regarding the economic crisis was intensified with the alarming reports of an invasion by the unknown RPF rebels.

The upshot of this initial stage of democratization, then, was to make the Habyarimana regime appear that it was selling out to the economic and political demands of foreign actors. And to a degree this was true. Newbury notes that the “process had produced only weak links to the population as a whole” and that it

was forced on Kigali by outside powers who largely dictated the form of this political transformation. These external actors achieved their leverage because they controlled the resources on which the government depended to address a series of recent economic setbacks.

The alienation of most Rwandans from political reform was mirrored by the fact that most of the elite parties spent as much time at the internationally managed negotiating table as organizing constituents. This socio-political division appeared from the very beginning. While Hutu opposition leaders initially welcomed the invasion as a form of pressure on Habyarimana and even started to form a tactical alliance with the RPF, most Rwandans “resented the RPF for perpetrating violence and provoking government retaliation against them.” The relatively weak links between opposition leaders and the population, and their divergent political interests, were to haunt them later when an organized resistance was necessary to stop right-wing Hutus.

1991-1992 – Internationalization; Peace Process Begets War and Social Crisis
Notably, at this time ethnic politics were still very marginal, in spite of the RPF invasion. There was no Hutu Power movement. As Mamdani notes, “neither the power nor the opposition was organized along ‘ethnic’ lines.” The most significant issue was the familiar struggle within the ‘Hutu’ elite, which took on the regionalized character “that power be anchored in a base broader than simply the northwest of the country.” With power up for grabs the most important socio-political cleavages were regionalized inter and intra class divisions, whose dynamics were to redefine once more in dramatic fashion the politicization of ethnicity. As foreign donors became increasingly involved in this multi-level political contest, occurring at one level within the Rwandan ruling class, at another level between the RPF and Habyarimana, and always in the background between the old regime and the peasantry, they radicalized the situation.

Part of the problem was that, in spite of the fact that the RPF suffered a serious defeat at the beginning of the invasion, various actions by external powers ensured that the war persisted. Collins notes that

The war was sustained largely due to the extent to which both sides received external backing: the FAR [Tutsis] most significantly from France; the RPA from Uganda, which was in turn reliant upon military and economic support primarily from the United States and Britain.

The split external backing reflected a certain degree of great power rivalry. The old regime had been a staunch French ally in Francophone Africa. France had traditionally seen Francophone Africa as its natural reserve, had staked its position as a great power partially on its control of this region, and the US had granted France the mission of monitoring Francophone Africa, including former Belgian colonies like Zaire, Rwanda and Burundi, against Soviet influence during the cold war. At the end of the Cold War, there was a renegotiation of spheres of influence in Africa, and the French became pre-occupied with the ‘anglo-saxon conspiracy.’ They feared they would lose ground to the US and the UK via an expansion of Anglophone control. To a degree, events of the mid-1990s bore this out as they saw anti-French, pro-‘Anglo’ candidates, parties or movements emerge in nearly every Francophone country.

In Rwanda, French fears took concrete form in the RPF, which was not only backed by the US and UK ally Museveni, but which itself was anglophone, and some of whose leaders, including then general Paul Kagame, had trained in the United States. The initial French interest in ‘democratization,’ in Rwanda and elsewhere, reflected a desire to manage inescapable post Cold War political change, with the hopes of ensuring that pro-French, rather than anti-French, candidates would take power. Likewise, the later US and UK interest in and control over the peace process was in part driven by their desire to take leadership in Rwanda away from France, which also explains why they seemed to favor the RPF during negotiations. The French saw Habyarimana as the legitimate arbiter of change, and the RPF as anti-democratic, illegitimate and foreign. They therefore poured money, over $160 million in economic aid between 1990 and 1994, as well as huge amounts of military assistance, with the hope of defending their ally, Habyarimana. The US and its Anglophone allies saw the RPF and friendly opposition parties as agents of change and democracy, and Habyarimana as outdated, authoritarian, and intransigent. It may also have been easier for the US and UK to sustain the legitimacy of their interference by focusing on discrediting the main object of reform – the old regime – while overlooking the flaws of those like the RPF who would be part of the ‘democratic opening.’ It was also easier for Anglophone powers to claim credit for ‘democratizing’ Rwanda – and gain political points back home – if their project in some way differed from France’s. Indeed, the great power struggle was not merely over whose ally would control a new Rwandan government, but also over which foreign governments would get to claim the moral highground for having reformed Rwanda. The problem for Rwanda, then, was not merely the internationalization of political reform, but that this internationalization included great power competition. This had the destabilizing effect of creating an international struggle for control over the pace and direction of reform, a struggle the US and UK eventually won out once ‘democratization’ became a ‘peace process.’

In the short run, this created the paradoxical situation in which even though there was a general demand for peace, each party to the war was supplied directly and indirectly by competing powers. This had the disastrous consequence of extending the war. Even as different nations supplied and supported the two sides, external insistence on a ceasefire gave the RPF and the Rwandan government time to reconnoiter, recruit, and retrain. This became a recurring pattern. A final military confrontation that would have decisively settled the war was constantly forestalled by a new ceasefire, which simply gave combatants breathing room for a fresh round of fighting. Initially this was most important for the RPF, which lost its top commander and many fighters in the first invasion, but later ceasefires became equally important for the Habyarimana regime as the regular Rwandan army faltered against a reorganized RPF. By 1992, France had helped arm and train the newly formed self-defense militias, or Interahamwe, which had emerged both as a response to the crumbling army and to the burgeoning ranks of unemployed. The Interahamwe did much of the killing during the massacres in 1994. In short, the international insistence on ceasefires and the ‘peaceful’ re-negotiation of power was remarkably counterproductive. The ‘peace process’ helped ensure that the civil war continued long after it should have ended, with disastrous social consequences.

The most serious consequence of the civil war was to turn incipient economic collapse into a full out crisis. Indeed, Rwandan society was buffeted by extreme economic destruction, and the suddenness of the collapse made it especially disruptive to the Rwandan social order. Already in 1991, the UN had concluded “that 43 percent of all Rwandan farm households had fallen below the absolute survival threshold.” Between 1989 and 1993 Rwanda saw a 40% decline in its GDP and its first famine since 1943. It also saw a 34% increase in its balance of payments problem and massive inflation in the price of consumer goods. Peasants were uprooted, both by RPF gains in the northwest and the failure to make a living, unemployment increased, and fertility declined. Even before civil war reached their immediate lives, many Rwandans already found themselves struggling to survive.

The Habyarimana regime had few answers to the climbing social crisis. By 1992 the ranks of the regular army had swelled from between 3,000 and 5,000 to 30,000 to 40,000. Given that it was still trying to hold on to power and privilege, the massive allocation of spending to military purposes seemed to serve all purposes at once. It provided employment, a new form of social bonds as others deteriorated, and increased resistance to the RPF forces. But the government needed tremendous amounts of external fiscal assistance to sustain such policies. In exchange for more loans, the lending institutions, in a remarkable act of ignorance and stupidity, had Rwanda implement another structural adjustment in mid-1992. The program further devalued the franc, raised the cost of living for average Rwandans, eliminated price supports and, worst of all, forced the state try to increase its foreign exchange earnings by planting more coffee.

These ‘recommendations’, as Uvin notes, were handed down “as if politics did not exist.” The only way to increase coffee exports so suddenly was to use the coercive apparatus of the state to force peasants to plant more trees and endure more forced labor. In response, in the summer of 1992 peasants uprooted at least 300,000 coffee trees, refused to perform forced labor, destroyed property, refused to participate in state rallies, and “occupied Western-sponsored demonstration and reforestation projects.” As Smith notes, “these moves were all acts of defiance and desperation” and “terrified the embattled rulers, who saw their political prospects and their coffee profits plunge still further.” This sudden emergence of the peasants as a social force in Rwandan society meant that any future political arrangement in some way would have to address their needs. The question was who and how.

In the short-term the peasants saw their needs addressed neither by lending institutions nor by the internationally managed political process. In fact, by mid-1992 the US and the Belgians had taken took over from the French, gradually redefining the ‘democratization’ process as a ‘peace process’ in the context of which political reform would take place. Habyarimana made “additional concessions on democratization and refugees to satisfy international demands.” By April 9 Habyarimana had installed opposition members as ministers, as well as granted two prefectures to opposition parties, as part of a new government. The administrative structures the prefect commanded penetrated deep into Rwandan society; therefore, prefectureships were very important positions of power in the Rwandan state. Even if Habyarimana remained in effective control of the national government and the military, these were nonetheless serious concessions. They also showed the way in which the political victories of the regional opposition were contingent upon international pressure, not a strong domestic following.

In spite of Habyarimana’s moves, the RPF broke the cease-fire in March 1992, attacking the northeast, and then launching another attack in the northern Byumba province in June. The attack was at once a renewal of the attempt to seize power militarily, and a way of testing the waters of international sentiment. The RPF was very sensitive to the signals it received from the international community. Given its lack of political support inside Rwanda, it would have to rely on external allies if it gained control of the state. It could therefore only hope to use force to gain power if it thought such actions were at least tacitly sanctioned, which they appeared to be. Once again, there was little international condemnation of the RPF’s militarism. (Only France responded by sending troops to (ineffectively) bolster the Rwandan army. ) With the exception of the international response to the massive February 1993 offensive (see below), when the RPF broke another cease-fire, the RPF hardly ever faced serious criticism for its constant violation of the peace and democratization process. In fact, soon after the March 1992 attack and the formation of the new government, the international community forced Habyarimana to enter official peace negotiations with the RPF in Arusha, Tanzania, called the ‘Arusha Peace Process’. Arusha in effect recognized the RPF as a legitimate political party alongside those that had been formed within Rwanda. This official recognition effectively rewarded them for breaking the cease-fire, and sent the signal that they could gain at the bargaining table of the democratization/peace process through the use of military force. Subsequent decisions by the RPF, including breaking more cease-fires, reflected calculations based on the possibility of repeating this early victory.

The most generous interpretation of the actions of the donor countries is that they simply were unaware of the impact they were having on Rwandan politics, perhaps blinded by their own ideologically motivated theories of ‘democratization.’ Jones has found in confidential interviews that “analysis by Western diplomats suggested that democratization must run parallel to peace negotiations if a viable peace was to be found.” In other words, the donor countries insisted on continuing one process – democratization - that was producing social conflict, even as they insisted on peace negotiations between warring parties. Somehow ‘process’ was supposed to resolve fundamental conflicts of interest. In reality, Arusha inspired the RPF to continue its militarism, began to drive the Habyarimana oligarchy further into its increasingly crowded corner, while compromising the political legitimacy of the opposition parties and doing nothing to address social crisis.

1992 - The Rise of Hutu Power and the Racialization of Social Crisis
The accumulated set-backs of 1992 produced a shift in political strategies amongst those close to the regime. The Habyarimana government faced intense internal pressure as the collapse of the regular army, and economic crisis, discredited its current course of cooperation with the peace process and its political decisions so far. In the short-term, it organized the Interahamwe militia as regionally based ‘self-defense’ units mobilized against the RPF. Although the Interahamwe did most of the eventual killing of Tutsis, as Mamdani notes, its original purpose was to form a cheaper, back-up to the weakening regular army. In the long-term, however, it was clear that the akazu was losing ground to the southern and central opposition, as well as to the RPF not just militarily but in negotiations, and that these two opponents enjoyed more international favor than Habyarimana and the akazu.

Facing defeat both on the battlefield and at the negotiating table the old regime now began to think it could only preserve its power by extreme measures. The racist faction grew in power. Anti-civilian, ethnic violence appeared to be the only political strategy by which the cornered, reactionary Hutu elite could cohere Rwandan society against an enemy, and preserve its power not just against the RPF but the encroachments of opposition parties. The Coalition pour la Defense de la Republique (CDR), an extremist party composed of some of the right-wing akazu elites, emerged and at times took positions even in opposition to Habyarimana’s own party, especially when the more moderate Habyarimana insisted on ‘ethnic reconciliation.’ Habyarimana’s Mouvement Revolutionnaire National pour le Developpement (et la Democratie) MRND(D) became increasingly shrill in its denunciations of the RPF and in its ethnic appeals. This was the beginnings of what came to be known as the ‘Hutu Power’ movement that eventually dominated not only the CDR and MRND(D) but also split the opposition parties. As opposed to Habyarimana’s original achievement of de-racializing the Tutsis, and considering them a Rwandan ethnic group, the “objective of the [Hutu Power] was to reracialize the Tutsi” as an alien race. At this point we can provide a political account of what the constructivist, discourse analysis tends only to describe: the emergence or ‘construction’ of a racist, ethnic discourse. The intensely racist politics amongst the ruling Hutu elite and parts of the opposition leadership did of course draw on long-standing and familiar elements of Rwandan politics. However, it had been politically marginal in 1990, and only emerged with any strength as the old regime’s contingent, last-ditch response to encroaching military and political defeat. (Indeed, the infamous and propagandistic Radio et Television Libres des Mille Collines was only established in late 1993.) The RPF was beginning to make gains on the battlefield, even as foreign pressure to make concessions at the negotiating table intensified, donor insistence on further economic reforms continued apace. Here we can see how the Hutu Power was a product of the crisis politics brought about by the civil war and internationalization of the democratization/peace process.

The Opposition Fragments as Rwandans seek representation
It is not enough to provide a political explanation of the emergence of anti-Tutsi politics, but also why it ultimately managed to acquire some degree of ‘appeal’ and support beyond its old regime base. It is in fact remarkable that the akazu managed to find participants amongst the peasants whom it had so viciously exploited for the past twenty years and continued to exploit in pursuit of a political project so inimical to their interests. In fact, in 1992, one might have assumed that the Habyarimana regime, facing mass protests by peasants uprooting coffee trees and refusing to attend state sponsored rallies, would be precisely the group least likely to find political support amongst the lower classes. It was more likely that one of the opposition groups would be a vehicle for representing peasant and lower class interests. In the late 1980s and early 1990s, when the Habyarimana regime’s crisis of legitimacy created a opening in which the national project of Rwandan politics could be redefined, it seemed that the opposition would indeed play such a role.

However, the ‘democratization’ process as a political alternative was becoming increasingly discredited in the eyes of the peasantry, because it was so removed from their concerns. Yet the opposition remained committed to the process because it appeared to be an arena in which the southern and central middle classes could achieve their interests. As David Newbury notes, many opposition leaders seem to have used the negotiation process as a way of gaining access to “politically-defined arena of privilege.” They were unperturbed by the fact that the process had “little to do with bringing any meaningful participation of broader classes into the policy-making arena.” As a consequence, the opposition pursued power through maintaining political links with the international mediators, and making early alliances with the RPF against Habyarimana in negotiations. The opposition was not altogether wrong to believe that cooperating with international mediators was the path to power – they were rewarded at the negotiating table, and it was clear that any who did not follow the script and fell into disfavor with the international community would have remain politically marginal. The net effect was that by 1992, the opposition parties had weak links with the peasantry and lower classes, and had staked out positions, especially alliance with the RPF, that were unpopular. They were therefore in no better a position to claim to represent the interests of the popular classes than the old regime.
When the RPF broke the cease fire of 1992, they exposed the political vulnerability of the opposition. Originally supportive of the RPF, some parts of the opposition now felt betrayed by its erstwhile allies, not to mention hung out to dry by the Belgian, American and Ugandan sponsors of the peace process. Each opposition party – Parti Liberal (PL), Parti Socialiste Democratique (PSD), Mouvement Democratique Rwandaise (MDR) – began to fragment internally between a faction that thought it should support Habyarimana against the RPF, and those that sought to continue peace negotiations and alliance with the RPF. International pressure to continue the peace process meant that endorsing armed opposition to the RPF was at the same time to a degree a rejection of the donor countries and their peace/reform agenda, especially since resistance to the RPF now took the form of a reborn Hutu Power movement. The political confusion of the opposition parties developed into sharp internal divisions. One faction was more receptive to the increasingly vehement, right-wing ‘Hutu Power’ propaganda coming out of the old regime’s elites while another moderate faction pinned its hopes on a foreign donor managed political compromise. Struggles emerged at the local level over which faction – Hutu Power or not – of each party would control political offices and party privileges. At the same time, the main new parties still remained ‘opposition’ parties, and jockeyed for control of ministries and prefectures. The fragmentation and internal division of these parties left them unable to present themselves as an independent project, representing the interests of Rwandans, or later to stand in the way of the right-wing reactionaries who took the civil war to its bloody culmination. This political mayhem, in turn, signaled the way in which the donor community had so severely compromised the opposition by forcing the renegotiation of power through the democratization and peace process. The opposition began to realize only too late the consequences of having pursued their class commitments by pandering to international favor – a choice that had never been fully theirs to make in the first place. Unlike the internationally mobile donor community, the opposition now had to make the best of an inescapably bad situation, and live (and die) with the consequences.

Racism fills the gap
We can now better understand the almost inexplicable fact of how members of a regime widely considered illegitimate and exploitative by most Rwandans in 1990 managed to acquire some political support amongst those it had most abused. Although the brutal and discredited old regime elites were widely loathed as late as 1992, when peasants were protesting state policies and uprooting trees, they were given a new lease on life by the even greater political blundering of the international community and the resulting social crisis. The Hutu elite did not so much manipulate the masses as capitalize on the opposition’s political disarray, the international community’s obsession with peace negotiations, and the conditions of chaos and uncertainty existing in Rwanda at large. It gradually developed a message tying all popular anxieties to the ‘foreign Tutsi threat.’ The major priorities of the peasantry and lower classes were ending the civil war, opposing the RPF (who clearly did not represent peasant interests), and securing some relief from the economic crisis. Though the peasants were politicized by war and economic collapse, they were not independently organized through their own party according to their own class interests as peasants. Thus there was little chance these concerns would be met outright, and in fact nobody offered any consistent and substantive solutions to the economic problems besetting the majority of Rwandans. The various donor countries and lending institutions mainly ignored the social crisis; the RPF actively pursued a policy of uprooting and displacing peasants in the territories it conquered as part of a strategy of strengthening its hand at the bargaining table; the fragmenting opposition parties pursued power and privilege at the negotiating table. The government continued to obey lending institution demands to increase coffee production, which meant exploiting the peasantry.

Tragically, the Hutu Power managed to seize on this political vacuum. Without official sanction from the government but through control of local state apparatuses, the Hutu Power initiated a strategy in March of 1992 of linking massacres of Tutsis to the acquisition of land. The first of these killings began in Bugesera

where landless Hutu from the north-west had resettled. Competing for land with Bugesera’s Tutsi, themselves resettlers from the 1950s, and ‘encouraged’ by the exceedingly explicit, ‘Hutu Power’ threats…the northern Hutu migrants took out their anger on Tutsi and members of opposition parties, killing at least 300 Tutsi…

There are two key elements to these massacres. First, they signal the way in which the continuation of the war discredited the regular army and the political class, thereby creating the opening in which more extreme elements were able to wield political power. Indeed, Mamdani notes that the eventual slaughters were “not the project of the entire army, but of a fragment…[B]orn of defeat in the civil war, this fragment re-created a sense of national unity.” Even as Habyarimana continued to be forced to negotiate internationally, his gradual political and military defeat weakened his control over his own allies, and extremist forces expanded their power. Second, Mamdani notes that, in contrast to a few earlier massacres that were a retaliatory response to turning points in negotiations or RPF victories, these new killings were offensive. All of the massacres prior to 1994 have been called ‘dry runs’ but in fact this was the first one to have an offensive rather than reactive character, and its political content was distinct. It is easy to mistake this linking of killing to land as a tactic meant to prey on peasant land-hunger and poverty. Insofar as this is true, it had more to do with tapping into desperation than ‘greed.’ However, the real point is that linking killing to land was part of the Hutu Power’s strategy of tapping into peasant apprehensions about the RPF’s advance and reconstituting it as the threat that the Tutsis sought to reinstate the old monarchy. It was an element in the overall strategy of politicizing ethnicity by exaggerating the return of old inequalities, and its success was owed not just to peasant disorganization but also to the concrete gains the RPF made through war. As Mamdani notes

The growing appeal of Hutu Power propaganda among the Hutu masses was in direct proportion to the spreading conviction that the real aim of the RPF was not rights for all Rwandans, but power for the Tutsi. This is why one needs to recognize that it was not greed-not even hatred- but fear which was why the multitude responded to the call of Hutu Power the closer the war came to home.

Ending the civil war had become a central preoccupation of the average peasant because of the way it destroyed social life. By 1993, the effects of civil war had displaced one in seven Rwandans. The desire for peace doubled as resentment of the RPF for starting the civil war, and for increasingly appearing as if it merely sought power. In fact, this fear was not altogether unfounded. The RPF was constantly breaking cease-fires, purposefully uprooting peasants, and subverting the peace process, all the while making political gains. Some have argued that this was always their intention, although it is more likely that it developed and hardened in response to the doors opened by international control over political change. Either way, the longer the war continued the larger were the territories that experienced fighting or came into contact with refugees fleeing the fighting. This expanded the number of peasants seeking a solution to the fear and insecurity of the situation, and intensified this demand. The struggle to survive economic crisis was also associated with war. The war, with the RPF as the antagonist, thereby circumscribed and defined all facets of the peasants’ needs. Resisting the RPF and ending the war increasingly became the most obvious and available solution to their problems.

It was this the Hutu Power was able to exploit. They linked it to their racist politics and to a discourse about the return of the ‘Tutsi feudalism.’ Of course, the latter was apocryphal, not the least because it was many of the elite elements backing the Hutu Power and the old regime that during the 1970s and 1980s had profited on the backs of the backwards and exploitative economic relations based on forced labor, land inequality, and authoritarianism. Nonetheless, as Mamdani notes

The irony was that the more successful the RPF was on the battlefield, the more this view came to define the political center stage, bringing Hutu Power back from a fringe preoccupation to the mainstream of respectable politics.

Various interviews with participants in the 1994 massacres suggest that they killed Tutsis (and moderate Hutus) as supporters of one side of the civil war, not merely as members of an ethnic group. According to one interviewee, “the Tutsis were not killed as Tutsis, only as sympathizers of the RPF.” Although the RPF did increasingly try to recruit from Tutsis within Rwanda, for the most part there was little direct connection between civilian Tutsis and the RPF rebels. Nonetheless, as the only political discourse that made much of an effort to make sense of the peasants’ social reality, Hutu Power racism managed therefore to garner some followers, particularly amongst the most desperate – landless, displaced and unemployed.

An example of the way in which the Hutu Power racism tapped into and yet distort the concerns of the disorganized peasants and poorer classes is in the alarmist and vitriolic language about the inyenzi. The word means ‘cockroach’ in Kinyarwanda and was used by the Hutu Power to refer to the RPF. It has been mentioned as an example of the way the Hutu Power created a ‘dehumanizing’ discourse that prepared the ground for ‘genocide.’ However, the word inyenzi was first used by the Tutsi monarchist insurgents in the early 1960s as a term of self-reference because it was supposed to signify their invincibility and ineradicability. The Hutu Power was well aware of the historical significance of the term, and the historical fears it would evoke amongst the peasantry. It was therefore not meant simply to be ‘dehumanizing’ but rather to create a (spurious) association between the RPF and a political project - Tutsi monarchism. That it had any effect on the peasants has more to do with circumstances mainly beyond the control of the Hutu Power – the civil war and economic crisis. Thus ‘elite manipulation’, just like the construction of a ‘dehumanizing’ discourse, does not accurately describe nor explain the social and political context in which the Hutu Power made the most significant efforts at popular mobilization. What degree of actual popular support for the Hutu Power movement there was seems to have been a response to the unusual crisis conditions in which there were few voices clearly and consistently directed at public concerns. As such, the essence of the popular response lay in the fear and uncertainty of war and social collapse, rather than deeply ingrained and widespread racism.

From Arusha to April 6, 1994
If early and mid-1992 were a turning point for the old regime, and Rwandan politics in general, the Hutu Power was not yet strong enough to pursue the racialized political strategy they were developing. The Radio et Television Libre Mille Collins (RTLMC) did not yet exist, the Hutu Power still did not possess adequate control of local and regional state apparatuses, and they needed to secure and expand their control over the militia. The few massacres they did execute occurred in their regional strongholds, and were tests of a new political strategy. Nor had they managed, yet, to parlay social crisis into support for their project. In the short-term, Habyarimana, who was in fact more moderate than the Hutu Power leadership, had no other options but to continue to bow to international demands. Under pressure from the US and Uganda, the president decided to negotiate a new cease fire on July 10, 1992, agreeing in principle to a number of new concessions, including power-sharing with the RPF in the military and administration.

For the rest of the year there was foot-dragging on implementation of the agreement. Repeating the course of prior cease-fires, all parties reorganized and regrouped. According to Kuperman, “two senior rebels” in the RPF say the military organization “used the next six months to politically indoctrinate recruits, conduct military training, and enhance ties to the Hutu opposition.” Even now, in spite of the shifts in political strategies amongst a number of key actors, the eventual slaughter was not inevitable. The Hutu Power did not have a way of re-acquiring prefectures not under its administrative control, nor did it possess much popular support, at least outside the regime’s strongholds in the north and northwest.

The straw that broke the camel’s back was the RPF’s February 1993 offensive. It did more than the Hutu Power’s own efforts at recruitment and ideological indoctrination to radicalize the peasantry. Ostensibly a response to a massacre of Tutsis a few weeks earlier, the attack was rather part of its strategy of forcing Habyarimana’s hand at the negotiating table while continuing to seek power militarily. The tactic worked but at tremendous cost. Over the course of the month-long assault, the RPF doubled the size of the territory under its control including parts of the right-wing Hutu stronghold in Ruhengeri, killing many Hutu civilians along the way. In March the RPF stopped its offensive and pulled back from some territory in response to international criticism. Kuperman quotes one senior RPF official saying “we could have won, but the international community wouldn’t let us. France would aid the army and the international community would criticize us.” In the immediate aftermath of the February offensive, this was no doubt true. However the ‘international community’ – mainly the US, which directed the Arusha negotiations – quickly returned to sending signals that the RPF would be rewarded for breaking cease-fires and resorting to military action rather than keeping to democratic process. In August 1993, the Arusha Accords were rewritten to increase the RPF’s share of officers in the new army from 40% to 50% and gave it control of the important Interior ministry. It also gave the RPF and its allied opposition parties “the majority of seats in the interim cabinet and legislature preceding elections,” which as Kuperman notes made it seem to the old regime like Arusha was more or less a “negotiated coup.” These changes seemed to signal that if the RPF managed to change the facts on the ground through military force, foreign diplomats would reward them politically at the negotiating table. The take away message for the RPF appeared to be that it could continue to pursue a strategy of Tutsi power without too much concern for cease fires or democratic process. Meanwhile, for the old regime, the prospect of being the minority in the new government meant Arusha spelled dramatic political defeat, matching their declining fortunes on the battlefield.

The offensive had other effects. It displaced a million Hutus, radicalized many Hutus, broadened the already widespread fear and resentment of the RPF, and intensified the social and economic crisis stemming from the war. The power-sharing agreement also ensured that 40% of the regular army would be composed of Tutsis, which, in a country where the “army had ballooned into six times its original size…[and] plagued with massive unemployment …was literally like serving an unemployment notice to young recruits in the army.” Some of the most violent participants in the 1994 massacres came from these young recruits. The February offensive also decisively compromised the Hutu opposition, destroying what bonds of trust that still existed between the RPF and the opposition leadership. According to Jones, “as the shape of the Arusha deal became clear, many of the outer members of the akazu rallied to the center, including some who had flirted with opposition groups. This led to chaos among the opposition groups.” The Hutu Power factions gained ground within the major parties. In addition, since the agreement banned the CDR and dramatically reduced the power of the old guard, it ensured that they would fight to the death against any possibility of implementing the Arusha Accords. In short, the February offensive and new Arusha Accords increased popular and elite support for the Hutu Power movement, as well as stiffened hardline leadership’s resolve by closing the door to all other options.

The Arusha Accords embodied the disastrous consequences of the internationalization of political reform in Rwandan. Instead of marginalizing the Hutu Power, Arusha “confirmed the Hutu Power claim that the opposition had betrayed the nation. In doing so, Arusha sealed the political fate of the opposition.” It also created “the growing realization” amongst many Hutus “that the real objective of the RPF invasion was not rights but power—specifically Tutsi Power.” It allowed Habyarimana to co-opt “virtually all of the Hutu opposition parties into his ‘Hutu Power’ alliance against the Tutsis.” After all, the Hutu Power movement was the only political force in Rwanda that could credibly claim it was not a selling out Rwanda to foreign interests in a time of extreme social crisis and civil war.

Habyarimana knew he could not afford to sign and implement the Arusha Accords. There was no domestic political support for them. The Accords were almost entirely an international affair. Yet Habyarimana was walking a fine line, stalling on implementation while desperate to maintain some foreign support given that international aid and loans were the only source of income left in the war-ravaged economy. At this point, however, the international community’s apparent neutrality appears to have transformed into outright hostility towards Habyarimana. France had now withdrawn its military and economic support for Habyarimana. By the end of the year, the donor countries had abandoned all signs of neutrality, and now clearly supported the RPF: “at the end of 1993 the crucially important donors round table on Rwanda took place in Mulindi, north eastern Rwanda - the base of RPF operations against the government.” In December, the UN escorted 600 RPF soldiers into the capital. Every foreign creditor was withholding money unless Habyarimana signed Arusha, and the World Bank and IMF also attached political conditionalities to their funds. In addition, Habyarimana was afraid that if he rejected Arusha, the donor countries would completely abandon him to the mercies of the RPF, against whom the tattered Rwandan forces could no longer defend themselves. For Habyaramina, Arusha was inescapable no matter how ill-conceived. As Collins notes, “with no other source of funds available, Habyarimana was obliged to sign along with the other parties, on 4 August 1993.”

Signing Arusha did not achieve peace. Instead it created an “untenable situation that would lead only to more and greater violence.” For different reasons, neither the Hutu Power nor the RPF wanted to see Arusha implemented. The Hutu Power feared that implementation would mean a massive erosion of their power and possibly reprisals by the RPF; the RPF believed it could gain more by conquering Kigali and acquiring exclusive control over the state. As a result, Arusha produced “an arms race between the warring parties.” The RPF once again subverted the peace process even as it appeared to comply. It prepared for a final assault on the capital, with the hopes of conquering state power once and for all. Meanwhile the Hutu Power moved rapidly towards organizing a massive campaign of anti-Tutsi violence. The RTLMC was formed a few days after Arusha was signed, and began protesting the deal in virulent, racist terms. Indeed, it is really only after Arusha that the Hutu Power was able to co-opt so many opposition politicians, partly out of desperation and partly because the accords had discredited those who appeared to be cooperating with the RPF. The fall and winter of 1993 is when the Hutu Power, particularly those who believed massive anti-civilian violence against the Hutu opposition and Tutsis, began really flexing its muscle. Although Habyarimana attempted to delay implementation of Arusha, well aware of the reaction it would garner by extreme elements he could not control, by early spring 1994 the donor pressure was unavoidable. On April 6, 1994, Habyarimana was flying back from Arusha having made commitments to implement the accords when his plane was shot down over Kigali, killing all of its passengers. When the plane crashed the politics of ‘ethnic reconciliation’ fell with it and both sides set into motion their final gambits.
April 6 to July 19, 1994 - Ethnic or Political Violence?
The desperate and horrific actions of the Hutu Power had a number of dimensions all of which reinforce the notion that the massacres of 1994 cannot be seen simply as ‘ethnic violence.’ While civilian, Rwandan Tutsis were a main target of violence the development of an anti-Tutsi strategy must be situated in a concrete, socio-political context. Without doing so, many aspects of the 1994 massacres are unintelligible. It becomes impossible to comprehend why the Hutu Power would spend so much energy on eliminating Tutsi civilians in the south while the RPF advanced in the north, or why the Interhamwe killed Hutu civilians, or why participation in the killing was sociologically varied, rather than simply ‘mass participation’ in the way that is normally understood.

The argument so far is that the anti-civilian violence directed mainly but not exclusively at Tutsis emerged as political strategy for the old regime elite in the face of political and military defeat. The RPF had managed to play the peace process in a way that allowed it to gradually gain more territory, and the Rwandan regular forces were in disarray following defeat, demoralization, and economic crisis. Defending themselves against the RPF in direct military confrontation, then, was nearly impossible. Targeting helpless civilian Tutsis, however, was a way of striking a series of indirect blows at the RPF. First, it was possible that the RPF would be shocked into pulling up short in the face of the massacres, so as to prevent further killing of the Tutsis. In a sense this was a parallel strategy to the RPF’s, which sought to use the displacement of hundreds of thousands of Hutu peasants in territory under RPF control, and even some limited killing, to force Habyarimana’s hand at the negotiating table. Civilians had already become (dispensable) pawns in the complex bargaining of the civil war and international negotiations. Second, as de Waal notes, “forcing ordinary people to kill,” makes the killers feel “worthy only of the company of other killers.” The machinery of the state was used to coerce Hutu peasants and others to kill so as to taint as many as possible, making subsequent RPF rule effectively impossible. The thought, it seems, was that Tutsi rulers would be spontaneously rejected, or have no stomach to rule, if the subject population was generally implicated. Third, de Waal’s observation has another dimension. The Hutu Power needed as many Rwandans willing to participate in direct resistance to the RPF, but the only way to develop a force committed to such resistance was through anti-civilian violence. It was in this sense a desperate recruitment strategy. No doubt a great deal of wishful and irrational thinking went into this strategy, but desperation breeds desperate plans. This does not make them less political.

Indirect resistance to the RPF was not the only political element to the anti-Tutsi violence. It contained a very instrumental aspect of creating the cover under which the Hutu Power could eliminate the Hutu opposition leadership from the south and center. In fact, as de Waal notes “the first and most prominent victims of the killing were Hutu opposition politicians, academics, journalists, human-rights activists, lawyers, priests and businessmen.” Immediately after Habyarimana’s plane was shot down, the first to be rounded up and killed were “individuals who had criticized the Habyarimana regime…leaders of the MDR, PL, PSD, and PDC who rejected Hutu Power, members of the judiciary, human rights activists, clergy, journalists, and other leaders of civil society” most of them Hutu. Des Forges notes that, “by mid-day April 7, the Presidential Guard, with the help of soldiers of other elite battalions and some National Policemen, and eliminated those leaders who could have legitimately governed.” After this, the Hutu Power felt strong enough to move against rivals in opposition strongholds. They replaced less reliable military and administrative officials with those committed to Hutu Power, and as the violence expanded from Kigali to the south and center, the Interahamwe and other military forces under the control of the Hutu Power were sent to kill opposition leaders in positions of regional power. For instance, Prefect Jean-Baptiste Habyalimana, who had successfully opposed the killings in his prefecture of Butare, was arrested, killed and replaced with a Power operative. The violence, then, was driven by the awful political logic of a “physical liquidation of all advocates of democratization.” These advocates were labeled enemies of the regime and eliminated. The ‘southern strategy,’ as opposed to direct confrontation of the RPF from the north, becomes more intelligible and less bizarre when we realize that the Hutu Power were actually fighting a number of opponents at once. The anti-Tutsi civilian violence provided the means by which to confront these various opponents in the extreme circumstances of 1994.

Indeed, that the Hutu opposition sought to fight two groups at once accounts for the somewhat chaotic nature and regional variation of the violence. There was immediate killing in Kigali, where much of the opposition leadership was and the Hutu Power the most clearly in control, as well as in the northeast and east, where much of the civil war had taken place. Yet it took longer for the violence to arrive in the south because of greater resistance by regional political leadership. As Straus notes “intra-Hutu dynamics” were key in those regions because, after Habyarimana’s plane was shot down, “there ensued a competition for power among Hutus at the local level, the outcome of which was violence against Tutsis.”

Desperation measures are designed to solve all problems at once. The threat to Hutu Power emanated not just from the RPF or the southern and central counter-elite, but also from the peasants themselves. In fact, as much as we have discussed the ways in which the lower classes found some degree of representation in the Hutu Power project, it is misleading to see their participation as ‘mass participation,’ as if most or all Hutus were involved to the same degree. We have already seen that one of the main purposes of the killing was to taint the peasants, in part because of the way it might threaten future RPF rule. However, it also had the effect of driving peasants into the hands of the same elite it had so despised for the past twenty years. The purpose of the killing, then, was not just to kill as many Tutsis as possible, but to cohere a peasant population on the brink of revolt around a collective project of resistance against real and imagined enemies. It was a ‘solution’ to the peasant question and the regime’s legitimacy crisis without actually addressing the real issues of inequality, landlessness and unemployment, except insofar as it allowed desperate peasants to loot and claim land and cattle.

All of this was not easily achieved, and participation was not as voluntary as often implied. There was a great degree of coercion involved. The post-Habyarimana Hutu Power government used the institution of forced labor under supervision of state authorities to coerce peasants into digging graves, remove bodies, and often force them directly to kill. Hutus that resisted were killed, intimidating many into participating – a fact that is often obscured by the label ‘ethnic violence.’ It is difficult to find reliable numbers on how many Hutus were killed as ‘traitors,’ some estimates run in the tens of thousands. But the key point is that the voluntarist connotations of the phrase ‘mass participation’ or ‘ethnic violence’ need to be more carefully investigated, as does the very nature of ‘mass participation.’

It is extremely difficult to get reliable numbers on how many Hutus actually participated in the violence. The current Rwandan government claims nearly the entire adult Hutu male population participated, or about three million Hutus, while President Kagame is on record suspecting as many as a million. Others suggest it was ‘hundreds of thousands,’ while according to yet others it was tens of thousands. Prunier’s suggestion that the “main agents of the genocide were ordinary peasants” tacitly embraces high-end estimates. The one to three million estimates are clearly too high, however the ‘hundreds’ versus ‘tens’ of thousands should not be seen as an ‘either or’ debate. Rather, beyond simple data collection problems, the ambiguity in estimates reflects some of the complexities of conceptualizing the events themselves. Participation is in fact an ambiguous concept. Different estimates are based on different definitions of participation, and in fact there were different kinds of participation. As Straus notes, “explicit statements on what ‘perpetrator’ means…[is] critical.” Likewise, Des Forges notes that there were some who were the “first to kill” and who “attacked Tutsi frequently” and many more who participated “reluctantly, some only under duress or in fear of their own lives.”

A first cut notion of participation is someone who attacked with intent to kill. For Straus, a participant is “any person who participated in an attack against a civilian in order to kill or to inflict serious injury on that civilian.” Using this definition and his own research, Straus has produced the most accurate estimates we have on participation and his data paints a nuanced view of the sociology and nature of the violence. As a first cut, he estimates that between 175,000 and 210,000 participated in the violence between April 6 and July 19, 1994. However, he qualifies this estimate by noting that some were involved in numerous attacks, while others reluctantly or only once. By his calculations, “a small minority of perpetrators did the majority of killing.” In fact, according to Straus, “soldiers, paramilitaries, and extremely zealous killers” represented 10% of all perpetrators but were responsibly for 75% percent of the killing, while the other 90% of the perpetrators were responsible for the other 25% of the deaths. And in most places general killing did not begin until roving bands of militias appeared to initiate the violence. Of course, a larger number of Hutus were indirectly involved in and implicated in tasks like digging graves, identifying Tutsis, and carrying and burying bodies. But given the tremendous uncertainty, the fear of RPF takeover and the awareness of Hutu Power reprisals against non-participants, it is not unreasonable to see these forms of indirect participation as ways of getting around direct participation in a situation where outright resistance was extremely difficult. As Mamdani notes

The predicament of ordinary Hutu is clear from a single fact: it is not only the political opposition that got massacred in the days that followed the president’s assassination. As they grew in scope, the massacres targeted anyone, peasant or professional, who refused to join in the melee.

Des Forges also notes the degree to which “fear induced many others to attack or to refuse help to Tutsi”:

People were afraid of the RPF who, the radio said, were killing Hutu with great cruelty. But many Hutu were more immediately afraid of fellow Hutu, including local authorities and political leaders.

The threat to the survival of the ordinary Hutu came from a number of quarters, and there were few forms of organization that could form the social basis for resisting the violent demands of the Hutu Power and militias. Most participants were not the ‘Hutu’s willing executioners’ but bewildered and frightened peasants uncertain whom to trust, whom to resist, and how they would survive. Indeed, the participation of most Hutus needs to be understood within the context of multiple threats to their survival and no clear means of resistance besides the fragmented and compromised opposition.

These qualifications are not often made, nor carefully explained. Not only is Prunier’s despairing “terrible statement” that the “main agents of the genocide were the ordinary peasants themselves” unsupported, but the vagueness of the concepts ‘mass’ and ‘participation’ can be misleading. To be sure, between 175,000 and 210,000 is a large number (7-8% of the adult Hutu population ), but it is definitely not the ‘average’ Hutu. Most Hutus were not direct participants. And simply calling such numbers ‘mass participation’ gives the impression of spontaneity, without trying to account for the dramatic disparity in nature and intensity of participation between the aggressive 10% and the more reluctant 90%. The distinction is sociological. The small group of “soldiers, paramilitaries, and extremely zealous killers” was drawn from those lumpen elements of society most likely to respond to the Hutu Power project because their lives had been the most profoundly and directly affected by the civil war: the landless, unemployed, and refugees/internally displaced. In some places, as in the north, the militias were composed mainly of those fleeing areas where fighting was actually taking place, in others the recruits comprised urban unemployed and landless peasants, and in the east they drew on similar populations as well refugees from the late 1993 massacre of Hutus in Burundi. These various groups were the most fertile recruiting ground of the various paramilitaries and militias that did most of the killing. Having been stripped of their relationships to people in a specific, most likely peasant community, and thrust into a military environment, they were the least likely to be embedded in social relations that could form the moral foundation for resisting the racist and depraved calls of the Hutu Power. That is to say, the existence of a large pool of deracinated, atomized and desperate young men with a great deal of resentment and fear is not a social constant, but a product of extreme circumstances. The Hutu Power would not have found its willing executioners at any time, but only in the particular environment of civil war and social crisis. And even then they were forced to use coercion and fear to sustain the violence. ‘Elite manipulation’ is inadequate as a causal explanation, just as ‘mass participation’ is descriptively inaccurate, because it fails to identify the historically specific conditions that made the Hutu Power project possible.


These are not merely academic distinctions, but political ones. The uncritical notion ‘mass participation’ tends to criminalize the entire adult, Hutu male population. It also forms the backbone of the ideology supporting RPF authoritarianism. On this view, the RPF has the right to govern Rwanda undemocratically – banning parties, limiting elections, and restricting press freedoms – because most of the population cannot be trusted with political power. Alongside the ‘mass participation’ concept stands the ‘elite manipulation’ thesis. The problem with this explanation for ‘mass participation’ is that it fails to account for the events along two dimensions. First, it does not explain the extreme socio-political circumstances that made it possible for the Hutu Power to carry out its plan. Second, it sets up the crucial political dynamic as an issue of the relation between the akazu elites and an undifferentiated Hutu mass. The RPF, donor countries and institutions, and opposition parties disappear, and we lose sight of the various forces at work in Rwandan society. Most crucially, with these actors sidelined we lose sight of the political character of the violence, as well as the emergence of Hutu Power as a desperate movement to hold on to power in the face of the military and political defeat of the old regime.

The emphasis on the ethnic character of the violence has tended to shift the complex political relations amongst the different actors – great powers, lending institutions, Habyarimana elite, opposition leadership, RPF, and Rwandan peasants and lower classes – to the background. The use of the term ‘genocide’ has therefore turned all aspects of the violence, and in many cases all aspects of the events between 1990 and 1994, into bloodthirsty maneuvering by the akazu to carry out its ‘master plan’ of exterminating Tutsis. As we have seen this inverts the nature and causes of the violence. The violence was highly political. It was partially instrumental, creating the conditions in which the Hutu Power could kill opposition leaders it could not defeat politically. It was partially linked to a desperate strategy of indirect engagement with the RPF, as well as to a tactic of sublimating peasant social and economic woes into racist violence. These various dimensions of the violence arose out of the rapidly shifting alignments of the socio-political struggles within Rwanda during the early 1990s. It was not the ethnic that circumscribed and defined the political but the political that defined the ethnic.

In turn, we have seen how the conditions that led to the emergence of the Hutu Power and its last gasp were a product of the internationalization of the ‘democratization’ and peace process. The least acknowledged aspect of the Rwandan tragedy is the degree to which the Habyarimana regime, as well as other local political actors, were dependent on foreign support and constrained by the externally conceived project for political reform prior to 1994. Internationalization transformed the regionalized class struggle within the Hutu elite, and the growing political consciousness of the peasantry, which emerged with the Habyarimana regime’s end of the Cold War legitimacy crisis. The ‘democratization’ and later ‘peace process’ triggered and perpetuated the war, and radicalized Rwandan society. The result was the political and military defeat of the akazu, the rise of the Hutu Power, the acceleration of RPF militarism, the weakening of the opposition, and a massive social crisis that deracinated and politicized the Rwandan peasantry. These were the extreme and unusual circumstances that made the massacres of 1994 possible, as well as set the stage for a takeover of Rwanda by a new authoritarian regime that has enjoyed international support for over a decade. The terrible three months from April 6 to July 19 was not a sudden outbreak of ‘ethnic violence’ but a bloody culmination of a political struggle that foreign interference had decisively shaped. As Jones puts it, over those five years:

third-party interventions in Rwanda effectively transformed a civil conflict which had claimed roughly 6,500 lives into a power struggle between old and new guards: the old guard then chose to claim the lives of a million Rwandans rather than surrender its power.

Jones comments are a minority position. Most of the criticism of international action and inaction has been that there was ‘international indifference’ towards the killing. There are even accusations that the killings were covered up at the time so that foreign powers would not have to intervene, and there are the well known tortuous debates during the time over whether what was happening was a ‘genocide’ or not.

However, it seems the real ‘indifference’ was related not to the lack of foreign intervention, but the near total lack of concern various donor countries and institutions seemed to have for the effects of its deep and profound interference in Rwandan political affairs. Mamdani writes

The donor community force-fed Rwanda a reform agenda out of a textbook, without regard to the situation on the ground and secure in the knowledge that they would not have to suffer the consequences of their actions.

This indifference to the consequences of intervention was mixed with a decided ignorance about its effects. Foreign mediators were not fully aware of the signals they sent to the different groups – RPF, opposition, Habyarimana regime – nor of how this led to the perpetuation of war and crisis, mainly because they were narrowly focused on the negotiations themselves. This is always a problem with attempts by international actors to manage political reform, especially when what is sought is a smooth and quick transition to elections that the international community can chalk up as a victory for democracy. The great tragedy was not that the international community failed to intervene, but that it blindly and haphazardly forced a political experiment on a weak and poor country immensely dependent on foreign donors and in the throes of social crisis.

The sad irony of such experiments in political reform is that it is easy to blame failures on those who ‘fail to cooperate’ or refuse to act out the script international experts have written. When the history is written, somehow it turns out that the international community is not to blame for moving from ‘democratization’ to supporting the RPF which, without significant domestic support, was nothing more than an “army of occupation” and still rules in a deeply authoritarian fashion. The RPF’s first years were marked by its own massacres of civilians, estimated at about 25,000 to 45,000 immediately after the war, and then another 200,000 Hutus killed in a subsequent invasion of neighboring Zaire (Democratic Republic of the Congo). It also saw an increase of the prison population from 4,000 to 120,000, now only down to 90,000 with arbitrary detentions and even disappearances a continuing practice. More recently, President Kagame tried to ban the second largest party, jailed opposition leaders, and maintains restrictions on the press. Under the RPF, Rwanda is once again completely dependent on foreign donors, now mainly the US. Hopefully someday Rwandans will regain control of their own future.
In the meantime, it is difficult to see the Rwandan tragedy as one caused by the failure of the international community to intervene but rather the other way around. Somehow the story has been rewritten, with the focus mainly on the irrationality and violence of 1994, and the disastrous international interference quietly downplayed or left unmentioned. This has allowed Rwanda to become the reference point for subsequent interventions, in places like Kosovo and Darfur. A careful look at Rwanda’s history, however, does not confirm the common view that there was a crime of inaction for which the international community must make amends. Quite the opposite. The true crime was too much foreign intervention in Rwandan affairs prior to 1994. If anything, Rwanda is a lesson in why international interference in political change and social crisis in the Third World should be viewed with a great deal of apprehension.

Andersen, Regine. 2000. How Multilateral Development Assistance Triggered the Conflict in Rwanda. Third World Quarterly 21, no. 3: 441-56.
BBC, News. 2003. Profile: Rwanda's Strongman BBC [cited January 15 2004].
———. 2004. Rwandan Soldiers Arrive in Sudan
Boone, Catharine. 1994. States and Ruling Classes in Post-Colonial Africa: The Enduring Contradictions of Power. In State Power and Social Forces, edited by Joel Migdal. Cambridge: Cambridge University Press.
Burgess, Stephen F. 1998. African Security in the Twenty-First Century: The Challenges of Indigenization and Multilateralism. African Studies Review 41, no. 2: 37-61.
Carroll, Rory. 2004. Kagame Set Genocide in Motion, Paris Judge Says. The Guardian, Friday, March 12.
Chossudovsky, Michel. 1997. The Globalisation of Poverty. Penang: Third World Network.
Collins, Barrie. 1998. Obedience in Rwanda: A Critical Question. Sheffield: Sheffield Hallam Press.
———. 2002. New Wars and Old Wars? The Lessons of Rwanda. In Re-Thinking Human Rights, edited by David Chandler, 157-75. New York: Palgrave Macmillan.
Crawford, Barry. 1997 Debate: Why Did Rwanda Explode? Paper presented at the Rwanda The Great Genocide Debate, London, July 27.
Cumming, Gordon. 1995. French Development Assistance to Africa: Toward a New Agenda? African Affairs 94: 383-98.
Des Forges, Alison. 1999. Leave None to Tell the Story. New York: Human Rights Watch.
Economist. 1994. Burundi and Rwanda; Joined in Death. The Economist, April 9, 46.
———. 2004a. Rwanda, Remembered: Lessons of a Genocide. The Economist, March 27.
———. 2004b. Rwanda. The Economist, March 25.
Gourevitch, Philip. 1998. We Wish to Inform You That Tomorrow We Will Be Killed with Our Families : Stories from Rwanda. New York: Farrar Straus and Giroux.
Hobsbawm, Eric. 1992. Nations and Nationalism since 1780: Programme, Myth, Reality. Cambridge: Cambridge University Press.
Huliaras, Asteris C. 1998. The 'Anglosaxon Conspiracy': French Perceptions of the Great Lakes Crisis. The Journal of Modern African Studies 36, no. 4: 593-609.
Jones, Bruce D. 1995. 'Intervention without Borders': Humanitarian Intervention in Rwanda, 1990-94. Millenium: Journal of International Studies 24, no. 2: 225-49.
Kaldor, Mary. 1999. New & Old Wars. Stanford: Standford University Press.
Kasfir, Nelson. 2004. Domestic Anarchy, Security Dilemmas, and Violent Predation: Causes of Failure. In When States Collapse: Causes and Consequences, edited by Robert I. Rotberg. Princeton: Princeton University Press.
Kolko, Gabriel. 1994. Century of War: Politics, Conflicts, and Society since 1914. New York: New Press.
Kuperman, Alan. 2004. Provoking Genocide: A Revised History of the Rwandan Patriotic Front. Journal of Genocide Research 6, no. 1: 61-84.
Luttwak, Edward. 1999. Give War a Chance. Foreign Affairs 78, no. 4: 36-44.
Mamdani, Mahmood. 2001. When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda. Princeton: Princeton University Press.
McNulty, Mel. 1997. France's Role in Rwanda and External Military Intervention: A Double Discrediting. International Peacekeeping 4, no. 3.
Melvern, Linda. 2000. A People Betrayed: The Role of the West in Rwanda's Genocide. New York: Zed Books.
Mironko, Charles. 2004. Igitero: Means and Motive in the Rwandan Genocide. Journal of Genocide Research 6, no. 1: 47-60.
Monbiot, George. 2004. The Victim's Licence. The Guardian, Tuesday, April 13.
Mueller, John. 2000. The Banality of Ethnic War. International Security 25, no. 1: 42-70.
Newbury, Catherine. 1988. The Cohesion of Oppression: Clientship and Ethnicity in Rwanda 1860-1960. New York: Columbia University Press.
Newbury, David. 1998. Understanding Genocide. African Studies Review 41, no. 1: 73-97.
Nolan, Stephanie. 2004. The Next Rwanda?: Sudan's Neglected Nightmare. The Globe and Mail, Saturday, June 5, F1.
Olsen, Gorm Rye. 1997. Western Europe's Relations with Africa since the End of the Cold War. The Journal of Modern African Studies 35, no. 2: 299-319.
Pender, John. 1997a A Glimpse into Rwanda's Future. Paper presented at the Rwanda The Great Genocide Debate, London, July 27.
———. 1997b Understanding Central Africa's Crisis. London, July 27.
Posen, Barry. 1993. The Security Dilemma and Ethnic Conflict. Survival 35, no. 1: 27-47.
Pottier, John. 2002. Re-Imagining Rwanda: Conflict, Survival and Disinformation in the Late Twentieth Century. Cambridge: Cambridge University press.
Power, Samantha. 2002. A Problem from Hell: America and the Age of Genocide. New York: Basic Books.
Prunier, Gerard. 1995. The Rwanda Crisis: 1959-94: History of a Genocide. Kampala: Fountain Publishers.
Schraeder, Peter J. 1997. France and the Great Game in Africa. Current History 96, no. 610: 206-11.
Shawcross, William. 2001. Deliver Us from Evil: Peacekeepers, Warlords, and a World of Endless Conflict. New York: Touchstone Book.
Smith, David Norman. 2002. Globalization and Genocide: Inequality and Mass Death in Rwanda. In One the Edge of Scarcity, edited by Michael N. Dobkowski and Isidor Walliman. Syracuse: Syracuse University Press.
Snyder, Jack. 2000. From Voting to Violence: Democratization and Nationalist Conflict. New York: Norton.
Straus, Scott. 2004a. Origins and Aftermath: The Dynamics of Genocide in Rwanda and Their Post-Crime Implications. In Mass Crimes and Post-Conflict Peace-Building, edited by Beatrice Pouligny Simon Chesterman, Albrecht Schnabel. Tokyo: United Nations University Press.
———. 2004b. How Many Perpetrators Were There in the Rwandan Genocide? An Estimate. Journal of Genocide Research 6, no. 1: 85-98.
Uvin, Peter. 1998. Aiding Violence: The Development Enterprise in Rwanda. West Hartford: Kumarian Press.
Waal, Alex de. 1994. The Genocidal State. Times Literary Supplement, July 1, 3-4.
———. 1997. Becoming Shameless: The Failure of Human-Rights Organizations in Rwanda. Times Literary Supplement, February 21, 3-4.