As time passes and we are able to collect more data on the Rwandan genocide, it is becoming increasingly important that we recognize not only the externalities that created socio-political pressure in Rwanda leading up to the genocide, but also systematically study why the perpetrators of violence made the decision to participate. Generally we see a disconnect between the analysis of state level externalities and the personal stories told by perpetrators. In The Order of Genocide Scott Strauss is able to deftly weave a narrative that brings the structural and personal levels together, using a mixed methods approach to finding commonalities in the stories told by genocedaires and the empirical data on Rwandan political structures leading up to, and during, the genocide.
The central question Strauss aims to answer is why otherwise non-violent individuals would decide to join in the violence being perpetrated against the Tutsi minority. He dismisses theories of collective madness, but also makes a point to problematize overly structural arguments that make the genocide part of an overarching Hutu plan. In the middle of these two interpretations is a set of variables that combine institutional continuity, population distribution, and the manipulation of fear and obedience that were the key factors in why many of the participants joined in the violence.
Strauss lays out a very important limitation of the argument from the beginning. He is focusing on a single case study in order to identify core factors that led to genocide in Rwanda. He points out that doing a cross-national comparison becomes too general since the details of the socio-politics of each case can be so different. With that in mind he starts his analysis at the individual level, and then induces the information gathered from interviews with perpetrators to build a structural model of why participants would choose to kill during the genocide instead of opting out of the violence.
Since Strauss is making a generalizable argument about the nature of participation in violence in Rwanda, he starts by finding a novel way to randomize the perpetrators he interviewed. He selected jails where perpetrators were being held, and then randomized the names of the inmates. In order to control for bias and to protect the legal process for genocediares who were awaiting trial, he only interviewed those inmates who had already been convicted and sentenced. This came as close as one could to finding a random sample of perpetrators of violence whose stories could be deemed credible. The sample size, over 200, was also sufficient to draw general conclusions about the larger population of perpetrators. From this sample he draws out themes of coercion and social pressures that would be used to demonstrate the validity of the larger structural aspects of violence in Rwanda.
To supplement the interviews he not only uses statistical methods to test relationships between responses and interviewee characteristics, but also uses historical and political structures to interrogate the data he collected from the interviewees. By pairing the interview data with a historical analysis of the institutional continuity of the state, from the national down to the sub-local level, he is able to corroborate the stories told by the interviewees. Many of the stories focused on social pressure, especially the tendency to conflate Tutsi civilians with the RPF and the war going on in the northeast of the country. In doing this, he goes a step beyond just basic structural analysis of violence and ties the experiences of the individual perpetrators to a generalizable structure of coercion and obedience that only required a small number of well organized extremists to kick start.
While the methodology is strong, and the multi-method approach takes qualitative data into a generalizable space admirably well, there are some issues that can be highlighted. One area that seems difficult to square is the relationship between the individual narratives and the structural pressures that led to participation in violence. It is hard to say that the interviewees haven’t learned over time about the structures set up by the extremist wing of the Hutu government, and that this knowledge can allow an informant to formulate a narrative around the extremist structure, as opposed to a narrative that was driven by the Hutu ‘Pawa’ structure. There are also problems of recall bias that could be highlighted. He is doing these interviews years after the event, and while these kinds of events would obviously be important memories, how the respondents have shaped their recall of the events since 1994 could lead to externally invalid data in the dataset.
With these critiques in mind, Strauss does a solid job of integrating a mixed methods approach to understanding a qualitative aspect of participation in violence. He blends his statistical methodology into the narrative in a way that is accessible to non-mathematical readers while also being valid for those interested in analyzing his data. Overall, this analysis, while being a single case, provides a generalizable methodology that can be used to understand genocidal violence in other countries even though it is focused on Rwanda exclusively.