For Brazil, South Africa, and the United States, whether institutions breed races, or races breed institutions is an ongoing debate. It is difficult to determine, in other words, whether the categories in a national census are engendered and informed by existing racial delineations in the larger society, or whether the census categories actually influence and help “lock in” these categories for society at large.

Rwanda, however, is a clear example of how institutions breed races. Two institutions are explored here – identity cards, which Belgians began to assign Rwandans during colonial rule, and the census. Both of these changed racial dynamics within the country in significant ways.

Identity Cards

Identity cards issued during Belgian colonial rule profoundly shaped racial identities in Rwanda. What had previously been fluid and ill-defined identities became solidified under this policy, in a pseudo-scientific attempt to distinguish the two races through phenotypical measurements, such as the width of their nose and skull size (Uvin, 1997). Based on the results, Rwandans were categorized as either Hutu (the African race) and Tutsi (the ‘Hamitic’ race, and long-lost brothers of Europeans).

The need for rigid categorization during colonial rule was clear. The Belgians employed indirect rule, placing Tutsis as administrators over Hutu in an attempt to facilitate their detached style of governance. Such a system could not survive if Rwandans were able to readily switch identities.

The institution of identity cards had two effects in Rwandan society. For one, it made the races mutually exclusive, dismissing the possibility of identifying with both races (which, given the high rates of miscegenation and the lack of distinct identities, most Rwandans could). Secondly, it politicized these groups, and created a racial hierarchy. Therefore, ‘Hutu’ and ‘Tutsi’ identities, in the context of the pre-genocide Rwanda, derived their political meaning from these colonial identity cards.

The Census

The 1994 genocide marked Rwanda’s transition from a state that emphasized racial identification to a state that does not recognize any racial or ethnic boundaries.

Here is a portion of Rwanda’s census in 1978:

1978

The question asks subjects for their nationality, and for Rwandans, it asks them to designate “RWA” for nationality and to additionally specify their ethnicity. There is no room in the answer section to specify two ethnicities, or two nationalities, even though Hutus and Tutsis often intermarried and some Rwandans probably had parents from both categories; this census implicitly establishes Hutu, Tutsi, and Twa categories as mutually exclusive identities.

That the question would be formulated this way makes sense given the historical context of the year in which this census was administered. In 1978, Rwanda’s colonial past was still fresh in its memory. The Belgians had assigned every Rwandan under one of the three categories, without much flexibility or mobility in categorization. The concept of identities as rigid and mutually exclusive categories survived in the years after independence.

1991

The Census of 1991 mimics the one in 1978. The question is asked in Kinyarwanda, the language that all Rwandans speak, perhaps as a Afrocentric, nationalistic reaction against remaining pieces of colonial legacy. Nevertheless, the rigid racial lines are still in the census – it asks Rwandans to identity their race or “ethnicity” and non-Rwandans to identify their nationality.

The post-genocide censuses diverge from their pre-genocide predecessors. The Census from 2001 shows that the corresponding question only asks for nationality, and does not ask for additional identification along racial lines for Rwandans. Paul Kagame’s regime, instituted in the aftermath of the genocide, outlawed not only the incitement of racial violence and hatred, but banned any mention of Hutu/Tutsi/Twa distinctions. ¹ Kagame’s hope is to create a homogeneous Rwandan society, united by a single national identity.

2001