Anbookstore com Copyright 2013

government identified a common threat—the inyenzi (or cockroach) raids that began in late 1961. The

so-called inyenzi was an armed faction of the UNAR (Reyntjens 1992, 172). Small bands made

frequent forays with the aim of “creating a sense of insecurity within Rwanda, and the prime targets

were Hutu officials and European administrators” (Weinstein 1977, 61). The worst attack was in

December 1963 when between 250 and 1,000 inyenzi rebels crossed into Rwanda from Burundi and

came within twelve miles of Kigali, after a foot trek of some 140 miles (Lemarchand 1970, 220;

Weinstein 1977, 64). The reaction of the Kayibanda government was swift. Authorities rounded up

twenty Tutsi leaders affiliated with UNAR and RADER (a moderate and ethnically inclusive party)

and publicly executed them (Lemarchand 1970, 225). Local militias were created in each province to

guard against future Tutsi attacks. In some areas, Hutu authorities attacked Tutsi homesteads, burning,

looting, and pillaging (Weinstein 1977, 65). By the time the attacks were contained, more than ten

thousand Tutsi had lost their lives (Lemarchand 1970, 225). The national police rounded up a few

hundred influential Tutsi and some Hutu; some were executed, others charged with treason or held

without charge (Reyntjens 1985, 463). By the end of 1963, between 130,000 and 300,000 Rwandans

(mostly Tutsi, along with Hutu and Twa who followed their patrons) had fled to neighboring countries

(Lemarchand 1970, 172; Reyntjens 1985, 455).

In identifying the inyenzi rebels as the common enemy, Kayibanda was able to unite the Hutu

leadership. His tactics, however, eventually backfired as the population—both ordinary peasant people

and the educated class—recognized that government-led efforts to unify against the inyenzi threat

were designed to consolidate Kayibanda’s position, rather than actually addressing the sources of fear,

insecurity, and dissatisfaction that had motivated Hutu to participate in the 1959 revolution in the first

place. Ordinary people continued to suffer the same abuses of power that their local officials (now of

Hutu ethnicity) had exacted upon them in the past. Educated Hutu—students, teachers, and junior civil

servants—felt shut out of the rewards of the revolution, from which much had been promised. Despite

their initial commitment to Kayibanda’s cause of Hutu unity, many educated Hutu quickly saw

corruption among state elites, recognized that the promises of increased pay would not be honored,

and judged that opportunities for career advancement were virtually nonexistent (Lemarchand 1970,

238–41). The response of the Kayibanda regime to these grievances was to blame all Tutsi, regardless

of social status or economic class.

Kayibanda expanded PARMEHUTU’s reach into local communities in an effort to “mould the

loyalties of citizens” (Lemarchand 1970, 247). Kayibanda retained the identity cards that the Belgians

introduced in 1931 to label Rwandans as “Hutu,” “Tutsi,” or “Twa,” and he instituted a quota system

to ensure proportionality in education and employment. The Kayibanda regime also limited Tutsi to 9

percent of the total number of seats in schools, the civil service, and even the private sector (Reyntjens

1985, 501). Contrary to the version of history found in the policy of national unity and reconciliation,

elite Tutsi continued to populate the ranks of the educated elite because of their historical advantage.

In addition, Kayibanda’s quota system was regularly bypassed as close-knit intra-ethnic kin and local

networks procured prominent positions in government and private sector employment for elite Tutsi.

This was possible because colonial policy had limited access to formal education to Tutsi, which

meant that few Hutu had the necessary skills to compete for these jobs (C. Newbury 1992, 197). It also

meant that close to 50 percent of the teachers and students at secondary school and in universities

were Tutsi (Reyntjens 1985, 501). At the National University, the primary training ground for the civil

service, Tutsi accounted for 90 percent of the student body (Lemarchand 1970, 260). The residual

effect of these colonial practices prompted members of the Hutu elite from the northern region of





Rwanda to call for radical reforms to remedy this ethnic imbalance.

Individuals from central Rwanda, where Kayibanda grew up, dominated his government (C.

Newbury 1992, 197). Corruption was the order of the day; Kayibanda’s long-term associates and

family held key administrative positions, which caused a rift among Hutu elites. Northern factions

“began to openly criticize the regime” (Jefremovas 2000, 303). In efforts to protect his power,

Kayibanda reframed regional dissent as ethnic violence. Rumors of Tutsi attacks against Hutu like

those of the inyenzi in 1959–63 spread quickly. The population was quick to believe them, relying on

the stories of refugees from Burundi who had fled that country’s 1972 Hutu genocide, in which an

estimated one hundred thousand to two hundred thousand individuals died (Lemarchand 1998, 6). As

Burundian Hutu streamed into Rwanda, Kayibanda instituted public safety committees, which were

essentially vigilante groups created to monitor the civil service, universities, schools, and businesses

to ensure that the ethnic quota regulations were being followed (Des Forges 1999, 40).12 In late 1972

and early 1973, almost all Tutsi students were run out of schools and the national university

(Reyntjens 1985, 503). Anti-Tutsi sentiment affected elites and ordinary people alike. The public

safety committees ensured that educated, salaried workers were fired, and blacklists were posted in

offices to intimidate local Tutsi. Ordinary folk were asked to leave their homes, which were looted

and burned by members of public safety committees (Reyntjens 1985, 503). In the hills, the violence

was not directed solely at Tutsi as it was in economic centers like Kigali, Butare, and Gitarama.

Instead, the wealthy, rural elite Hutu and Tutsi alike—not ordinary peasant Tutsi—were the victims.

Those from the north targeted southerners, while those with outstanding grievances against the local

authorities used the opportunity to exact payback (Reyntjens 1985, 504).

By July 1973, the violence had abated, and Major-General Juvénal Habyarimana, the most senior

officer in the northerner-controlled army, took power from Kayibanda in a “bloodless” coup (Des

Forges 1999, 41; C. Newbury 1992, 197–98; Reyntjens 1985, 506).13 Under Habyarimana, the ethnic

question took a backseat, and his coup was “welcomed” because it reduced both ethnic violence and

rampant government corruption (Jefremovas 2000, 303). Habyarimana was popular with Tutsi, and

some Hutu groups accused him of favoring Tutsi. Relations between elite Hutu and Tutsi were

amicable early in Habyarimana’s rule, although top positions in government were reserved for

northern Hutu. There was also considerable intermarriage, “not only between southern Tutsi and Hutu

but also between northern Hutu families and economically powerful Tutsi families” (Jefremovas

2000, 303).

Peace and the semblance of stability nonetheless came at a cost. Habyarimana consolidated a highly

centralized state apparatus to monitor and control the activities of the population that today is part and

parcel of the RPF regime (Purdeková 2011; Reyntjens 2011). Habyarimana banned political parties in

the name of ethnic unity and national security in much the same way that Kagame has (Amnesty

International 2011; Cooke 2011; Reyntjens 2011). In 1975 Habyarimana created the Mouvement

révolutionnaire national pour le développement (MRND), of which all Rwandans were members (Des

Forges 1999, 41). The party structure was then extended down to the most local levels of government

as the central party apparatus appointed all officials. Each local official was granted a leadership

position in the MRND, so that the state and the party became one entity in people’s daily lives

(Prunier 1997, 76). The central government required that residents of each commune register with the

local authorities, which in turn reported all births, deaths, and movement in and out of their bailiwick

on a monthly basis, a practice that continues today (Des Forges 1999, 42; field notes 2006). Identity

cards continued to categorize Rwandans according to their ethnicity. This measure was necessary to

implement Habyarimana’s policy of “ethnic and regional equilibrium,” which reserved seats in





educational institutions and in the state apparatus in order to rectify the favoritism shown by the

Belgians to Tutsi. On paper, the policy aimed to ensure ethnic equality; in practice, it excluded Tutsi

and non-northern Hutu from lucrative government posts. Only 9 percent of positions were reserved for

Tutsi, despite the fact than an estimated 15 percent of Rwandans held Tutsi identity cards.

The “hierarchical ethos” and top-down decision-making structures that shaped Habyarimana’s

regime affected ordinary people (C. Newbury 1992, 199). Mobilizing the population for public work

projects such as road repairs, ditch digging, and brush clearing was common practice in most rural

communities, harkening back to the colonial practice of akazi (or labor conscription by the Europeans)

and continuing today with the same practice of umuganda (community work). The stated difference in

the practice under Habyarimana was that it was voluntary (not mandatory, as it had been under

Belgian rule, although naming and shaming were employed to ensure full “voluntary” participation).

Such “voluntary” labor was limited to only two days per month (not per week, as it had been). In

practice, the government required ordinary people to work on umuganda projects on average four days

per month, and local officials enforced everyone’s full participation. Individuals who did not report

for work duty were fined (Des Forges 1999, 42). Some men in their late forties and early fifties who

participated in my research reported that the poorest of the poor “had to report for umuganda at least

one or two days a week or suffer imprisonment or worse” (interviews 2006). They also told me that as

the poorest of the poor had to report more frequently in those days because the authorities told them,

“Since you don’t work for money, you will come work for the state and we will feed your families.”

Local officials often used this tactic to entice men to participate in umuganda projects, as the full

participation of the population showed “those at the top” (senior government officials) that the local

official was able to control the activities of peasant people in his jurisdiction, just as it does today

through imihigo local accountability performance contracts (Ingelaere 2011; interviews 2006). When

local officials actually delivered food to families as promised, it was “usually rotten or infested”

(interviews 2006).

The ordinary peasants I consulted also made it clear that, despite the years of social revolution

(1959–62) and Habyarimana’s policy of ethnic equilibrium, ethnicity was a tool that elites used to

gain and keep power. Burnet notes that many of the women she worked with during her field research

“learned who they were when they went to school, obtained their national identity cards (around the

age of 16), or applied for jobs. The majority of women I interviewed . . . became aware of their ethnic

identity while at school” (Burnet 2005, 68). In 1972 and 1973, before the Habyarimana regime

introduced reforms to the education system, teachers forced Tutsi students to stand up in their

classrooms and identify their ethnicity, as only Hutu would be allowed to sit for national exams

(Burnet 2005, 69).

The highest level of formal education attained among my participants was the equivalent of the

third grade in the American system. Most participants had had only a year or two of formal schooling,

meaning that none of them had sat for national exams and none had graduated from primary to

secondary school. They also all lived in southern Rwanda, where relations between Hutu and Tutsi

were more amicable and cooperative than in other regions of the country, particularly the north. For

these individuals, ethnic identity had little impact on their daily lives. The ordinary peasants who

participated in my research understood that their lives were shaped by their social and class positions

and by their daily interactions with local authorities, which in turn were formed by the prevailing

regional and class politics at the time. All of the individuals who participated understood themselves

to be peasants. As Martin, a Tutsi man who survived the genocide told me,

the genocide. Why do you think they [the government] kept the cards? Even they

can’t tell who is Tutsi or who is Hutu. They needed cards to tell the killers, just

like they needed cards before [under Habyarimana] to determine who would

benefit from politics.

But me, and my family, and the others around us, Hutu or Tutsi or what, we

are peasants. And the authorities don’t care about us. Any of the benefits of

being Hutu didn’t matter; we [peasants] didn’t get anything out of being this

ethnicity or that one. We were told what crops to grow and always support the

party [MRND]. Coffee was important in those days.14 But we didn’t have land

like that. We can hardly feed our families. Some grew coffee instead of food

because they were forced. Others grew coffee and stole food from other plots.

There were a lot of problems among us [peasants], but they [the authorities]

didn’t care. We are peasants, and we don’t matter for much.

Being Hutu or Tutsi, that was the business of the government and other

important people [elites]. We just hoped, like we do now with this government,

that we could get some peace. (Interview 2006)


The Habyarimana regime created a strong, centralized, and effective state that served as an

instrument of domination and control. In the process, his government sought to contain the regional

and ethnic tensions from which Habyarimana’s power was born. The system of ethnic equilibrium was

unable to contain rivals for state power. Two senior officers in Habyarimana’s army made an

unsuccessful coup attempt in April 1980 (Gasana 2002, 30). The coup sharpened the divisions within

the northern Hutu elite, meaning that those from Habyarimana’s home province of Gisenyi were

favored over individuals from bordering Ruhengeri province. Habyarimana’s most favored individuals

were those from the abahinza lineage, of which his wife was a member. Habyarimana himself was

from an unimportant Hutu lineage; he consequently relied on “his wife’s clan . . . to be his ears and

eyes” (Prunier 1997, 86). Agathe Habyarimana was nicknamed Kanjogera (the name of King