Making a Genocide: The Role of Top-Down Leadership
By Liam Hollen
Pierre-Yves Beaudouin / Wikimedia Commons / CC BY-SA 4.0
In all of the progress that has been made for humanity in recent centuries, there still remains mankind's ability to commit one of the greatest crimes against humanity. Having emerged across all types of cultures and at various historical periods, studying and comparing certain cases of genocide is essential for coming to grips with humanity's most dangerous sociological phenomenon. For these purposes, I define genocide as the systematic extermination of a large group of people based on their ethnicity, nationality, religion, or race. When asked to give examples of genocide, the first case most would offer would be Nazi Germany's Holocaust, with its industrial eradication of the Jews. But other instances of genocide are not so clear cut. In many cases, the motives are numerous and the perpetrators are dispersed. The Rwandan genocide was one instance in which a lack of reporting allowed ambiguities to remain for the outside world while the violence was in full swing. Gone unchecked, over a period of just 100 days, the Hutu majority swept across the nation killing over 800,000 of the Tutsi minority. In this case where a peaceful citizen on one day took up arms against their neighbors in the next, I am interested in the role of those in authority during a genocide, and how the existence of a genocidal mastermind at the top of a hierarchical chain of command may impact a genocide's possibility and potency. In short, I ask the question: to what degree is a centralized leadership with a top-down command hierarchy necessary for genocide to occur?
To answer this question, the Rwandan genocide will be used as the primary case study. The Rwandan genocide was chosen because of its unique circumstances of widespread participation by the general population. These circumstances provide a high degree of potential for exhibiting the bottom-up, decentralized networks that could be most capable of proving a strict top-down command structure unnecessary. In addition, using the twin nation of Burundi that was divided from Rwanda in the late 20th century as a control variable allows for the relatively isolated analysis of certain social causes and effects. To establish appropriate context, this paper will begin by laying out a general background of Rwandan history preceding the genocide with a focus on historical factors that will come into play later on. I will then detail how the genocide took place in a step-by-step chronology, which will provide the benefit of discerning from which causal direction the genocide's string of events came. In tandem with this approach, I will investigate the nature, salience, and effectiveness of the Hutu perpetrator's command structure. Next, I will establish the social and psychological conditions of Rwanda's population before genocide and discern whether or not these conditions provided enough momentum for genocide to take place, while keeping an eye on the influence of a central figure in shaping these conditions. Continuing along the lines of a leader's role in shaping social psychology, I will discuss the role of group dynamics in committing genocidal acts, especially in terms of how systematically implemented and government-imposed social paradigms may have enabled widespread participation within these groups. Finally, I will address the ways in which sub-national coalitions and networks matter for shaping the outcome of genocide. Here I will ask if alliances between national and local actors are necessary for genocide to occur, and if policies of genocidal mass violence are initiated or accelerated at the local level. Recalling the value of Burundi as a comparative source, I will finalize my assessments by analyzing whether the differences in outcomes in Burundi are directly and cohesively explained by my conclusions.
Rwanda is a small landlocked nation in central Africa, bordering Burundi, Tanzania, Uganda, and the Democratic Republic of the Congo that began as a Belgian colony. Hutu historically comprised 85 percent of the population, Tutsi about 15 percent, and the Twa less than one percent. The division between Hutu and Tutsi began as an economic one, with the commoners working in agriculture labeled as Hutu, and the elites who took part in cattle breeding labeled as Tutsi. We know that for several centuries Hutu, Tutsi, and Twa coexisted within a common society, spoke the same language, and shared the same religious ideas and social customs. The arrival of the European colonists, however, played an especially prominent role in exacerbating the Hutu-Tutsi divide. The Belgians promoted the idea of the Tutsi as a racially superior people, solidified their differences by issuing identification cards, and put the Tutsi in positions of power to act as a proxy in enforcing their colonialist agenda. Although it was believed that the Tutsi were generally taller and slimmer with more refined facial features, the difference between a Hutu and a Tutsi often cannot be determined by physical appearance. Regardless of how arbitrary the original classification was, in the years leading up to the genocide, the prevailing view amongst the Rwandan population was that the Tutsi were foreign invaders from the north, and there was a difference between the Hutu and the Tutsi that could not be changed.
On July 1, 1962, Rwanda officially declared independence. By this point in time, Belgian loyalty had shifted in favor of the Hutu, and with the Belgians' political and military support the Hutu were able to forcibly take control of important government positions and instate a Hutu president, Habyarimana. Rwanda was at first not an obvious location for genocide to take place, with a sound economy, generous foreign aid, and relatively muted ethnic tensions under Habyarimana. However, by the late 1980s, Rwanda's model image began falling apart. Coffee beans, their primary export, suffered a precipitous decline in price, and with it came famine and widespread unemployment. The following years were marred with occasional outbursts of violence against the Tutsi, which led to large-scale Tutsi refugee movements and social displacement. In response, the Tutsi formed the Rwandan Patriotic Front (RPF) with the goal of eventually taking back their homeland. In October of 1990, the RPF invaded Rwanda in an effort to retake the nation but were repelled by government forces. Under international pressure, Habyarimana had no choice but to acquiesce to peace talks that became known as the Arusha Accords, but these actions prompted strong dissent from members within his regime. Likely consumed by exclusionary ideologies centered on creating a "pure" homogenous population, the dissenting leaders began to meticulously plan a Tutsi genocide.
The plan went into action on April 6, 1994, when the plane carrying Habyarimana was shot down, killing everyone onboard. Within three hours, the RTLM radio broadcasted throughout Rwanda that the plane had been shot down by the RPF and that Rwanda was in a state of invasion. All Tutsi were branded as allies of the RPF, and Hutu were called to take up arms against their neighbors. As part of the first stage of the plan, assassins from the presidential guard were dispatched to target prominent Tutsi leaders and Hutus who were Tutsi sympathizers. Soon after, roadblocks were set up by youth militias known as Interahamwe—"those who attack together"—who were then dispersed about the nation to form killing groups of their own. Amid daily and persistent killing, the chronology of the genocide from then on is punctuated by a number of massacres in "safe havens" where fleeing Tutsi congregated. The genocide finally came to an end when the RPF seized control over the government.
By all accounts, the execution of the Rwandan genocide was highly coordinated and planned well in advance. The instigators were a small and tight-knit group of people who had dutiful followers and authority. In preparation, they organized and trained approximately 7,000 Interahamwe and other militia, ordered thousands of machetes, conjured lists of key resistance leaders, and ensured control over the national radio. This preparation was key in enabling the perpetrators to outmaneuver their opponents, striking with such dazzling speed that they had no time to respond. The hierarchical structure of the perpetrators consisted of three tiers: the highest tier was the planners and instigators who held high-ranking positions within the government or military, the second tier was the leaders who mainly consisted of the Interahamwe, and the third tier was the joiners, who were primarily made up of ordinary citizens. In the words of a witness:
It was organized on all levels. Everyone had a superior and subordinates and told them what to do. Those authorities in neighborhoods called meetings and gave orders and encouraged people. There was a member of the Interahamwe in every group. That member would lead the killing in that particular group.
When analyzing how the events unfolded in this light, it is clear that there was a direct, coordinated, and authoritative line of communication from the head instigators to the massacres on the ground. Also notable about this hierarchy is that there were few means for the joiners to send information in the opposite direction. Although this evidence points towards the effectiveness of a highly rigid and centrally based source of command, it would be a mistake to immediately assume that this form of organization was a causal factor for the genocide's success rather than a structural accompaniment to it. In order to settle this ambiguity, we must investigate various other structural components of the genocide—such as the psychological state of the population—and determine if the aforementioned top-down command hierarchy led these structural components or if the command hierarchy was in fact just a complementary force.
The ability to bring a regular citizen to a state of openness towards committing genocide is an extremely multidimensional issue. Some of the most robust empirical findings in recent literature show a strong connection between the presence of war—especially civil war—and genocide. However, there is less consensus on the causal mechanisms linking war to genocide. Some argue that the logic of war can be easily applied to the logic of genocide, in that civilian groups can be constructed as enemies that must face a military means of destruction. Another probable hypothesis is that war creates conditions of insecurity and vulnerability within society, amplifying the sense of threat to an extent that will trigger disproportionate responses. These hypotheses seem to have played out in Rwanda, where the Tutsi minority were framed as a foreign enemy force that was outside of the national community and a threat to the Hutu nation as a whole.
The classification, symbolization, and polarization of minority groups, in that order, is also a common pattern amongst societies on the path to genocide, and the actions of the Belgians in Rwandan history closely mirrored these steps. Most groups that perpetrate genocide are in control of the government and make up the majority of the population but are a class of lower social status and claim to be victimized and taken advantage of by the minority. This was reflected in Rwanda in how the Hutu who made up most of the population's lower class were quick to frame the Tutsi as the root of all of society's problems.
The final key variable that will be discussed is the influence of the state-owned RTLM radio station. Although RTLM did issue orders when the genocide began, it was arguably just as important for its role in propaganda. The lively and casual RTLM broadcasts were a key force in sticking the "cockroach" label onto the Tutsi. This tactic is a textbook example of distinguishing a certain group as the "other" and assigning to them negative value judgments, which can be highly effective in dehumanization. In addition, RTLM normalized hatred towards Tutsi, playing popular songs so inflammatory that some featured artists were later put on trial as war criminals at the International Criminal Tribunal for Rwanda, a special court established specifically to account for the most influential and egregious perpetrators of the genocide. The same central body that incited the genocide owned RTLM, and they wielded it as a powerful tool in shaping public opinion. However, RTLM's influence should still be placed alongside cultural, historical, and social circumstances to paint a broader picture of the forces that shape social sentiment. In addition, it is doubtful that social sentiment alone has the capacity to instigate genocide, since there have been many other instances throughout history in which ethnic tensions did not escalate into synchronized nationwide violence. In all, when investigating social sentiment as a causal force, we repeatedly see that the psychological priming of populations for genocide acted less as a gear itself, and more as the lubricant to keep various other gears moving. If not exclusively sociological priming, what exactly compelled the Rwandan population to turn against their neighbors?
When compiling all causal factors, group dynamics emerges as the key component that opens the floodgates to genocide. One highly indicative observation is that killings in the Rwandan genocide took place almost exclusively within groups. People are social beings, and conformity research has shown that compliance and consensus mechanisms are an extremely powerful force in shaping human behavior. Once within a group, members often feel a strong inner urge to adhere and adapt to the group norm and personally identify with group values. Within fanatical groups, such as those run by the Interahamwe, these pressures are amplified. Socio-psychological research has also shown that groups make people feel anonymous and less accountable for their behavior, lowering their regular social inhibitions. Another common group phenomenon is competition emerging between group members as to who is the best member, with the best member being the individual who best adheres to the group norms and is most successful in contributing to the main aim of the group. Needless to say, this group dynamic became particularly deadly during the Rwandan genocide, when the group aim was to kill as many Tutsi as possible.
But in comparison to ordinary delinquent groups, Rwandan killer groups were different. Rwandan killer groups were organized from the top-down rather than spontaneously from the bottom-up, and this difference proved to be instrumental to their ubiquity, and consequently, their success. Violence within Rwandan killer groups was instigated, ordered, and condoned by authorities rather than committed in deviance, as is the case in normal delinquent groups; by operating on orders of the ruling political power-holders, groups could operate in an environment in which their actions were considered legitimate and justified. This expanded the reach of killer groups beyond the regular fanatics, sadists, criminals, and profiteers, and enabled killer groups to draw in ordinary individuals who believed they were doing the right thing on behalf of authorities.
Taking a step further, I argue that Rwandan killer groups could not have formed without a coordinated, instantaneous, and drastic change in social norms. Within hours after the plane crash, the social default left that of a normal society and turned into a "kill or be killed" state. Put simply, the social duty not to kill was suddenly replaced with the social duty to kill. A major change in intersubjective social understanding is only possible if it is a view adopted by the majority of society, and the majority of society cannot adopt such a polar opposite of social understanding in such a short amount of time unless conducted in a coordinated fashion. Thus, this sort of paradigm shift must have and can only have been implemented by the dictum of an organized, capable, and unified leadership. In the Rwandan genocide's execution of an intricate plan through an established command structure, we see exactly that.
To summarize the chain of logic of the previous few pages, the inputs that shape socio-ethnic hatred are highly multivariable, with the leaders' input being one of many factors. In order to turn socio-ethnic hatred into face-to-face violence perpetrated by civilians, the spell of group psychology comes into play. But for genocidal groups, certain social paradigms are necessary, and these necessities can only be fulfilled by a unified leadership that operates under an organized and hierarchical command structure. Therefore, from the evidence gathered so far, top-down leadership is a necessity for genocide to occur.
However, my investigation up until this point has omitted an underlying force within society that may have the ability to incite genocide on its own, and must be dealt with in order for my logic to stand. This force is that of sub-national networks. I define "sub-national networks" as prominent social networks at the local and regional level that are capable of shaping societal outcomes. Influential actors within sub-national networks can take the form of town or province-level civilian administrators, local security forces, industry professionals, as well as religious and community leaders. These sub-national networks operate dependent of one another and are densely interwoven in the functioning of society. They are networks built on the sum of shared knowledge and shared activities that involve regular, face-to-face interaction. On a more granular level, sub-national networks can be broken down to the building blocks of civil society, such as circles of friends, local business partners, sports teams, clubs, members of a neighborhood, and so on. These multiplex ties give rise to multiple channels of contact, and serve as conduits through which people express and maintain social order.
Considering the prevalence and potential of these inherently decentralized networks, the question should be asked: must national actors ally with sub-national coalitions for genocide to occur? In the case of the Rwandan genocide, killer groups were enabled by the dense set of communal ties that linked leaders, collaborators and joiners. It is clear that sub-national networks determine how genocides occur; however, they do not seem to determine if genocides occur. To illustrate this point, if a leader was intent on pursuing genocide but lacked the support or faculty of sub-national networks, the genocide would still take place, just through organizations that are managed at the national level, such as the state-police or the military. Almost all genocides are carried out by militaries because of the ease of execution offered by their built-in command structure and operational capacity, and it is highly revealing that although Rwanda lacked a capable national military, it still directly mimicked the command structure of one.
Leading off of the question that led the previous paragraph, perhaps just as revealing a conclusion can be reached by asking the same question from the perspective of sub-national networks: must sub-national networks ally with national actors for genocide to occur? In other words, can sub-national coalitions initiate genocide? One may argue that if these sub-national networks were interconnected enough, a single sub-national network taking up genocide could sweep up the nation through a domino effect. However, this is logically implausible, since it requires that all groups be equally open to genocide and that no counter-coalitions would form. This is because if any sub-national coalition resisted, since sub-national networks are so amorphous in nature, one sub-national actor would have to establish dominance and organization over multiple other coalitions. Doing so would relinquish its role as a sub-national network and turn it into a microcosm of a state actor. Likewise, for a sub-national group to expand to the extent necessary to harness a genocide on a national level, it would have to no longer be sub-national and become national. Regardless of theory, in practice the Rwandan networks were isolated enough from one another to exclude the domino effect as an option.
Although initially counterintuitive, the importance of sub-national groups in fact demonstrates the importance of cohesion in leadership. A dense network of sub-national groups and opposing factions would dramatically weaken and potentially counteract any ambitions for genocide in nations that do not have strong national organization. Therefore, the existence of powerful decentralized networks further points towards the necessity of a centralized and singular leadership for genocide to occur.
Judging by its heavy reliance on the regularly non-participatory local population, we can determine that the Rwandan genocide was the most sub-national group-dependent genocide in recent history. Significantly, in this instance in which sub-national groups had the greatest capacity to form bottom-up chains of command, a top-down hierarchy prevailed and went into force as strong as ever. In other words, in this exceptional case when an opposing bottom-up argument had the most potential, not only did the top-down argument prevail, but the example became a point of further evidence to reinforce the necessity of a top-down structure.
Thus, reconciling the existence of powerful sub-national networks, a unified top-down form of leadership remains as the critical lever for genocide. What makes conclusions formed from the Rwandan genocide particularly valuable and more universally applicable than other case studies is that Rwanda has a control variable. Burundi is a nation directly south of Rwanda, and since the Belgians ruled it as one half of a joint entity, Burundi is a near-identical twin to Rwanda in terms of geography, colonial history, and ethnic makeup. Fundamentally, it shares Rwanda's circumstances in all variables but genocide. Even the plane crash that killed Rwanda's president and began the Rwandan genocide also had Burundi's president on board. So why did Rwanda fall to genocide when Burundi did not?
The answer, as expected, lies in Burundi's leadership. The withdrawal of colonial leadership had a different effect in Burundi than it had in Rwanda in that the Tutsi minority maintained control over the central government. Anti-Tutsi violence took place in groups and was relatively common but, lacking a central node of organization, was always sporadic and short-lived. Since the Tutsi dominated high-ranking positions but remained a minority in general society, they lacked the support and participation of sub-national networks, and we therefore see that all acts of violence against the Hutu were carried out through national military forces. In the largest of incidents, Hutus in a single region would form riotous groups, usually killing hundreds of Tutsi, and the Tutsi would respond with state-run military with crushing force, usually killing Hutus in the tens of thousands. The most prominent of these incidents took place in 1965, 1972, 1988, 1991, and 1993, making them not uncommon.
Any sort of Hutu violence against Tutsi that was planned and organized through a centralized body was always about insurrection and political overthrow, and never entered into the realm of genocide. Since political unification is a precondition to genocide, it only makes sense that missions to gain preeminence over the national government would precede any execution of genocide on an organized scale. In almost all cases, the manner in which events transpired within Burundi illustrate isolated social variables manifesting themselves in ways that fall directly under the previously established logic of sub-national networks and command structures that were developed from analysis on the Rwandan genocide.
In light of the preponderance of both contextual and theoretical evidence for the importance of top-down leadership, it is easy to come to the conclusion that whenever leaders have the means, opportunity, and motive to pursue a strategy of genocide, they could do so with relative ease. After polarizing and dehumanizing certain identities, the assumption goes, a unified leadership with a capable enough command hierarchy would be able to generate the necessary force to drive even ordinary people to commit mass violence.
Yet it is dangerous to place too much stress on personal figures and attribute all causality to them while ignoring structural factors. Decentralized forces can shape social sentiment, can direct social order, and can form groups who commit violence in a genocidal manner. Cohesive groups need dense networks of interaction and communication that link leaders with their followers, and this especially applies on the national scale. Unless a leadership enjoys the privilege of having an independently operating and fully capable military body under their authority, sub-national networks will dictate the degree to which genocide can occur. Even in the case of powerful and military-run regimes, there is a dialectical relationship between external conditions and a leader's intent, and there are many happy cases when authoritarian and autocratic states do not feel the need to instigate genocide even if other violence is used routinely.
That said, there remains a causal force reserved for centralized leadership that must exist in order for genocide to occur. Although decentralized forces can foster genocidal tendencies within the population and engage in isolated acts of genocide, they cannot form an aggregate of multiple instances of violence that are repeated in a consistent and systematic fashion, which is the essence of what makes a genocide a genocide. A variety of decentralized factors act as complementary forces to facilitate genocide, but a top-down command structure is precisely and exclusively the mechanism that can initiate widespread violence in the coordinated fashion defined above. In Rwanda, the civilian killer groups that were instrumental to bringing the genocide into fruition were only operational because they were implemented in a highly organized, hierarchical, and simultaneous manner by a centralized leadership.
Due to the overwhelming breadth of genocide studies, this paper's discussion has been relatively limited to factors of genocide initiation and implementation that a centralized body or a decentralized network can exclusively provide. There remain various other avenues of considering the role of leadership in genocide that have great potential for future research, such as topics involving a leader's motive, maintenance of power, and continuity in relation to genocide. But amid this discussion, it would be a disservice to forget genocide's human toll. Whether it was the day of the plane crash or after weeks of living in a state of fear for one's life, victims of the Rwandan genocide were often not simply killed, but hunted down, beaten, raped, and mutilated before being killed by gunshots or machetes. Bellies of pregnant women were slashed open, babies and small children were smashed against walls, and relatives were forced to watch how their loved ones were raped, tortured, and killed. With the conclusion that group dynamics and organized leadership is instrumental to enabling this kind of evil, I hope that we may be one step closer to being able to fight genocide in a more targeted and effective manner.
 Jason McCoy, "Making Violence Ordinary: Radio, Music and the Rwandan Genocide," African Music 8, no. 3 (2009): 86.
 William R. Pruitt, "Toward a Modified Collective Action Theory of Genocide: A Qualitative Comparative Analysis," (PhD diss., Northeastern University, 2012, 140, https://repository.library.northeastern.edu/files/neu:1020/fulltext.pdf.
 McCoy, "Making Violence Ordinary," 86.
 Ibid., 87.
 Pruitt, "Toward a Modified," 140.
 McCoy, "Making Violence Ordinary," 87.
 Sabby Sagall, Final Solutions: Human Nature, Capitalism and Genocide (London: Pluto Press, 2013), 233.
 Alette Smeulers and Lotte Hoex, "Studying the Microdynamics of the Rwandan Genocide," British Journal of Criminology 50, no. 3 (May 2010): 439.
 Eric Weitz, A Century of Genocide: Utopias of Race and Nation (Princeton: Princeton University Press, 2003), 14.
 Sagall, Final Solutions, 239.
 Smuelers and Hoex, "Studying the Microdynamics," 435.
 Ibid., 440.
 Ibid., 442.
 Scott Straus, "'Destroy Them to Save Us': Theories of Genocide and the Logics of Political Violence," Terrorism and Political Violence, 24, no. 4 (2012): 546; Jay Ulfelder and Benjamin Valentino, "Assessing Risks of State-Sponsored Mass Killing," SSRN, February 1, 2008, 14, https://ssrn.com/abstract=1703426.
 Martin Shaw, War and Genocide: Organized Killing in Modern Society (Cambridge: Polity, 2003).
 Manus Midlarsky, The Killing Trap: Genocide in the Twentieth Century (New York: Cambridge University Press, 2005), 94.
 Gregory Stanton. "Countries at Risk Report—2012," Genocide Watch: The International Alliance to End Genocide, 2012, 4-5, http://www.genocide-watch.com/images/Countries_at_Risk_Report_2012.pdf.
 McCoy, "Making Violence Ordinary," 87-88.
 Pruitt, "Toward a Modified," 97.
 McCoy, "Making Violence Ordinary," 92.
 Smuelers and Hoex, "Studying the Microdynamics," 451.
 Ibid., 446.
 Ibid., 448.
 Ibid., 449.
 Ibid., 446.
 Straus, "'Destroy Them to Save Us,'"556-557.
 M. Catherine Barnes, "Beyond Conflict: The Structure and Purposes of Genocide in the 20th Century" (PhD dissertation, George Mason University, 2000), 706.
 Ibid., 130.
 Smuelers and Hoex, "Studying the Microdynamics," 435.