Rwanda is about to complete one of the most ambitious transitional justice experiments in history, blending local conflict-resolution traditions with a modern punitive legal system to deliver justice for the country’s 1994 genocide. Rwandan President Paul Kagame described the initiative as an “African solution to African problems.” Since 2005, just over 12,000 community-based gacaca courts—deriving their name from the Kinyarwanda word meaning “grass” (the place where communities gather to resolve disputes)—have tried approximately 1.2 million cases. They will leave behind a mixed legacy.
Some Rwandans have welcomed the courts’ swift work and the extensive involvement of local communities, stressing that gacaca has helped them better understand what happened in the darkest period of the country’s history and has eased tensions between the country’s two main ethnic groups (the majority Hutu and minority Tutsi). Others are more skeptical: some genocide survivors complain that not all perpetrators were arrested or punished adequately for their crimes. Some of those convicted and sentenced to decades in prison maintain that trials were seriously flawed, that private individuals and government authorities manipulated the course of justice, that gacaca became politicized over the years, and that ethnic tensions remain high. On both sides, there are doubts, as well as tentative hopes, about gacaca’s contribution to long-term reconciliation.
This report acknowledges the enormous challenges the Rwandan government faced in choosing a system that could rapidly process tens of thousands of cases in a way that would be broadly accepted by the population. It explains the government’s decision to use gacaca to deal with the extraordinary circumstances it faced after the genocide and describes the government’s attempt to strike a balance between conventional due process and the overwhelming need for swift justice.
The report notes some of gacaca’s main achievements. Using dozens of cases, it also illustrates the price paid by ordinary Rwandans for the compromises made in the decision to use gacaca to try genocide-related cases, including apparent miscarriages of justice, the use of gacaca to settle personal and political scores, corruption, and procedural irregularities.
This report is not the first evaluation of the gacaca process. Avocats Sans Frontières (ASF) and Penal Reform International (PRI) have monitored the process closely since it began and have issued dozens of detailed reports on a range of topics related to gacaca. Rwandan human rights organizations, in particular the Human Rights League of the Great Lakes (LDGL) and the Rwandan League for the Promotion and Defense of Human Rights (LIPRODHOR), have also followed the process and have reported their findings. Books and scholarly articles have been written on gacaca as well. This report draws inspiration from these writings and raises some problems which have already been documented by others, but strives to analyze the gacaca process specifically from a human rights perspective, noting its accomplishments and its limitations in this context.
When the Rwandan Patriotic Front (RPF), currently the country’s ruling party, first took power in July 1994 after ending the genocide, it was confronted by the need to deliver justice for the killings of more than three-quarters of the country’s Tutsi population, as well as numerous Hutu who opposed the killings or tried to protect Tutsi. In total, more than half a million people perished in the span of only thirteen weeks. The challenge would have overwhelmed even the world’s most advanced justice system. In Rwanda, the task was made even more difficult because the genocide had killed a large number of judges and other judicial staff and had destroyed much of the judicial infrastructure.
A few months after the end of the genocide, Rwandan prisons were bursting at the seams with genocide suspects. By 1998, around 130,000 prisoners were crammed into space meant for 12,000, resulting in conditions that were universally acknowledged to be inhumane and that claimed thousands of lives. Conventional courts began trying genocide cases in December 1996, but had only managed to try 1,292 genocide suspects by 1998. At that rate, genocide trials would have continued for more than a century, leaving many suspects behind bars awaiting trial for years and even decades. The process might have been accelerated had foreign lawyers and judges been brought in to help, but the Rwandan government rejected such proposals.
Instead, the government proposed to set up community-based courts to try genocide-related crimes using the customary gacaca model. Aimed at speeding up genocide trials, reducing the prison population, and rapidly rebuilding the nation’s social fabric, the new form of gacaca, like its customary predecessor, would be run by local judges and would encourage participation of local community members. One of the government's aims in encouraging community participation was to make ordinary Rwandans the main actors in the process of dispensing justice and fostering reconciliation. A series of gacaca laws would regulate the genocide trials, mixing certain basic fair trial standards with more informal procedures.
Some government officials feared that gacaca might not be the right mechanism for genocide trials, given the gravity and complexity of the crimes. The customary form of gacaca had only been used for minor civil disputes—involving property, inheritance, personal injury, and marital relations—with more serious cases, such as murder, reserved for resolution by village chiefs or the king’s representative. These government officials worried that judges would struggle to correctly apply the law, given that many had no formal education or training. They warned of the risk of bias, stressing that the local setting meant judges would inevitably know the parties in a case which would reduce their objectivity and increase the risk of corruption. Most significantly, these government officials warned that gacaca procedures would fail to comply with Rwanda’s international fair trial obligations. Nearly 10 years after gacaca began, many of these concerns have turned out to be well-founded.
The concerns were overruled and, in June 2002, the Rwandan government launched a contemporary form of gacaca to try genocide cases, run by a new institution which later became known as the National Service of Gacaca Jurisdictions (SNJG). For more than two years, gacaca courts in 12 pilot areas used information provided by local community members to compile files on what had happened in each of these areas between 1990 and 1994. The courts drew up lists of victims and suspects, and classified the latter into four categories according to the severity of the alleged crimes. The most serious cases (category 1), involving mass murderers, rapists, and leaders who had incited killings, were transferred to the conventional courts; the rest were to be tried in gacaca.
The first gacaca trials started in 2005. They were set to end in late 2007, but the deadline was repeatedly extended over the following three years. In mid-July 2010, the government announced that the last gacaca trials in the country had been completed. However, two months later, it unexpectedly declared that gacaca would continue. This latest extension will allow the SNJG—tasked with oversight of the gacaca process—to review a number of cases of suspected miscarriages of justice and to allow for revision where appropriate. However, gacaca courts are not expected to handle new cases.
Rwanda’s experiment in mass community-based justice has been a mixed success. Many Rwandans agree that it has shed light on what happened in their local communities during the 100 days of genocide in 1994, even if not all of the truth was revealed. They say it helped some families find murdered relatives’ bodies which they could finally bury with some dignity. It has also ensured that tens of thousands of perpetrators were brought to justice. Some Rwandans say that it has helped set in motion reconciliation within their communities.
Yet there are multiple shortcomings and failures with gacaca: basic violations of the right to a fair trial and limitations on accused persons’ ability to effectively defend themselves; flawed decision-making (often caused by judges’ ties to the parties in a case or pre-conceived views of what happened during the genocide) leading to allegations of miscarriages of justice; cases based on what appeared to be trumped-up charges, linked, in some cases, to the government’s wish to silence critics (journalists, human rights activists, and public officials) or to disputes between neighbors and even relatives; judges’ or officials’ intimidation of defense witnesses; corruption of judges to obtain the desired verdict; and other serious procedural irregularities.
Many of these shortcomings can be traced back to the single most significant compromise made in choosing to use gacaca to try genocide cases: the curtailment of the fair trial rights of the accused. Although these rights are guaranteed by both Rwandan and international law, the gacaca laws failed to put in place adequate safeguards to ensure that all accused persons appearing before the gacaca courts would receive a fair trial. The gacaca laws tried to strike a balance by protecting some rights, including the right to be presumed innocent until proven guilty; modifiying others, such as the right to have adequate time to prepare a defense; and sacrificing others altogether, including the right to a lawyer. Dozens of cases mentioned in this report show how these due process shortcomings have directly contributed to flawed gacaca trials.
The government argued that traditional fair trial rights were unnecessary because local community members—who witnessed the events of 1994 and knew what really happened—would participate in the trials and would step in to denounce false testimony by other community members or partiality by the judges. Contrary to these expectations, however, Rwandans who witnessed unfair or biased proceedings decided not to speak out because they were afraid of the potential repercussions (ranging from criminal prosecution to social ostracism) and instead passively participated in the gacaca process. Without active popular participation, trials were more easily manipulated and did not always reveal the truth about events in local communities.
Another significant factor restricting the success of gacaca was the limited training given to gacaca judges, most of whom had little or no formal education and, in the vast majority of cases, no formal legal experience or training. Judges were not bound by evidentiary rules (explaining what types of evidence are admissible and the level of proof needed to convict a person) and were expected instead to rely on common sense and general principles of fairness. Courts had to provide reasons for their decisions, but were free to weigh the evidence as they saw fit. This led to contradictory results in different cases based on similar facts; to flawed decisions based, for example, on over-reliance on hearsay (words a person attributes to another who is not present at trial), and to convictions based on weak evidence. The fact that gacaca judges received no state remuneration also made the judges vulnerable to corruption.
Originally tried in conventional courts, genocide-related rape cases were transferred to gacaca courts in May 2008. Many rape victims based their initial decision to seek prosecution of the alleged rapist on the fact that conventional courts could enact measures to respect their privacy and could keep a woman’s identity confidential where necessary. The government’s decision to transfer their cases to gacaca courts, by definition involving the local community, took them by surprise and left some feeling betrayed. The SNJG justified the decision by claiming that many rape victims were dying of AIDS and that the conventional courts were unable to deal with these cases sufficiently quickly. It emphasized that the decision was based on requests by thousands of women who were raped in 1994. However, it would also enable the Rwandan government to complete all genocide trials as quickly as possible and to end this chapter of its history. Although the law provided for gacaca courts to hear rape cases behind closed doors, victims still feared that the community-based nature of the courts would mean that the local population would know what the closed-door trials were about. On the other hand, some rape victims whose cases were heard by closed-door gacaca courts said that the experience was less traumatic than they expected.
One of the serious shortcomings of the gacaca process has been its failure to provide equal justice to all victims of serious crimes committed in 1994. Between April and August 1994, soldiers of the Rwandan Patriotic Front (RPF), which ended the genocide in July 1994 and went on to form the current government, killed tens of thousands of people. They also carried out other killings later in the year, after the RPF had gained full control of the country. Gacaca courts have not prosecuted RPF crimes. Initially, in 2001, gacaca courts had jurisdiction over crimes against humanity and war crimes, in addition to genocide. But the following year, as gacaca courts began their work, President Kagame cautioned against confusing crimes committed by RPF soldiers with genocide and explained that RPF crimes were merely isolated incidents of revenge, despite evidence to the contrary. Amendments to the gacaca laws in 2004 removed war crimes from the jurisdiction of the courts and a national government campaign followed to make sure that these crimes were not discussed in gacaca. Nearly 17 years after the genocide, Rwandans who suffered or lost relatives at the hands of the RPF are still waiting for justice.
As gacaca draws to a close, the Rwandan government faces another challenge: correcting the grave injustices that have occurred through this process. There have been numerous gacaca cases involving miscarriages of justice or serious procedural irregularities, many of which have not been resolved by existing gacaca appeals procedures. The government’s recognition in late 2010 of the need to correct miscarriages of justice is a positive step. However, the proposal to have such cases reheard in gacaca risks replicating the same problems and may not remedy the situation. A more appropriate mechanism might involve a specialized unit within the conventional court system, staffed with professional judges or other trained legal professionals, to review the cases. Fair and impartial handling of these cases is of paramount importance to the legacy of gacaca and to strengthening the Rwandan justice system in the longer term.
To the Rwandan Government
- Announce a definitive deadline for the closure of gacaca and confirm that all outstanding and new genocide-related cases will be decided by the conventional courts.
- Direct the Ministry of Justice, in consultation with the SNJG, to create a mechanism within the conventional courts to review serious cases of injustice alleged to have occurred in gacaca.
- Order government officials and state agents not to interfere in gacaca and conventional court proceedings and not to attempt to influence decision-making.
- Order all police officers and state agents to refrain from conducting unlawful arrests and detention; prosecute agents suspected of such conduct and compensate persons unlawfully arrested and detained.
- Revisit the policy of using camps for community service (“travaux d’intérêt général” or “TIG”) and ensure, where possible, that community service is performed in the local community, rather than in camps, to facilitate reintegration of prisoners into their communities.
- Ensure that convicted prisoners and persons participating in community service are released as soon as their sentence has been served and compensate persons not released on time;
- Broaden the official definition of “genocide survivor” to include persons who lived through the genocide and were targeted or lost family members (i) because either they or their relatives were Tutsi or (ii) because they opposed the killings or tried to protect Tutsi; ensure that all such survivors are eligible for the government-run program of assistance to genocide survivors (provided they meet the other requisite criteria).
- Provide victims of sexual violence with trauma counseling and other assistance programs.
- Order credible investigations and allow prosecution of members of the RPF responsible for war crimes and crimes against humanity.
To Rwandan Justice Officials
- Create a specialized unit within the conventional justice system, for example within the Supreme Court, to review appeals from accused persons who claim to have suffered miscarriages of justice or serious due process violations in gacaca; develop a two-part review process which would provide: (i) an initial screening of appeals based on certain pre-determined criteria and (ii) a review of those cases appearing to have merit by specialized panels, headed by professional judges (not gacaca judges) or other trained legal professionals. These professional judges may consider a range of sources of information (including the written record from relevant gacaca proceedings and written submissions from the parties) and may hold short hearings where necessary before issuing a final decision affirming the judgment handed down in gacaca or revising the judgment (and sentence) where miscarriages of justice are found to have occurred. They may prioritize review of appeals for individuals still serving (or facing future) custodial sentences in prison or community service programs.
- Ensure that any new allegations of genocide are properly reviewed by trained prosecutors and judges before a person is prosecuted in the conventional courts.
- Verify that no person has been prosecuted twice for the same crime; review all convictions where a person was tried both before a gacaca jurisdiction and a conventional court or in at least two different gacaca jurisdictions in connection with the same events to identify and rectify violations of double jeopardy.
- Review all cases in which gacaca courts convicted persons solely for their presence at roadblocks during the genocide; confirm that each case contains adequate evidence of intent and criminal conduct to support the conviction.
- Prosecute persons who falsely accuse others.
- Investigate, prosecute, and punish appropriately members of the RPF responsible for war crimes and crimes against humanity.
- Monitor the execution of prison rules and regulations relating to the punishment of “life imprisonment with special provisions” to ensure that the punishment meets national and international standards, including prisoners’ right to have regular contact with other prisoners, in addition to outside visits from relatives or friends.
- Pursue the current proposal to convert any remaining prison time for convicts who have satisfactorily completed the community service program to a suspended sentence and allow convicts to return home.
- Ensure that the gacaca archives, including the database currently being compiled by the SNJG and the National Commission for the Fight against Genocide, are available to the general public in a comprehensive and easily accessible format.
To the Rwandan Legislature
- Amend the laws on divisionism and genocide ideology to bring them in line with international standards, narrowing the scope of prohibited conduct and requiring a specific intent of the actor, in order to ensure free speech and to encourage individuals to testify freely in judicial proceedings.
- Include in the draft penal code a provision establishing sanctions for state agents who intimidate or tamper with witnesses or judges, fail to execute judicial orders, or obtain statements or confessions under duress or coercion.
- Encourage the Rwandan government to create a mechanism to review gacaca cases involving serious injustices.
- Support the new review mechanism through funding and technical expertise.
- Continue to raise cases involving miscarriages of justice or due process violations with the Rwandan authorities and press them for corrective action.
- Bring to the attention of the SNJG any past cases of concern which have not been adequately remedied so that they may be reviewed and corrected.
- Provide the Rwandan judicial system with additional funds and technical assistance to strengthen the Victim and Witness Support Unit and to ensure equal access to this unit's services by prosecution and defense witnesses.
- Call on the government to end interference in gacaca and other judicial proceedings and to punish state agents who abuse their power or who try to influence judicial cases.
- Ensure that the external review of the gacaca process, currently financed by the European Union (EU) and the Netherlands, provides a meaningful and independent assessment, with recommendations to address shortcomings and to correct miscarriages of justice.
To Countries Considering the Use of Dispute Resolution Mechanisms Similar to Gacaca to Prosecute Serious Crimes
- Ensure fair trial rights are guaranteed.
- Offer all accused persons and victims access to pre-trial legal advice.
- Ensure equal access to justice for all victims of crimes committed by any party during the relevant time period or conflict.
- Provide adequate protection for witnesses, survivors, and judges, and ensure that police and prosecutors promptly investigate allegations of intimidation or corruption.
- Guarantee an environment in which witnesses may testify openly and freely, without fear of repercussions, and in which freedom of expression is respected.
- Create a mechanism to examine accusations before prosecutions are undertaken in order to protect against misuse of the judicial process by private citizens or government officials.
- Establish clear requirements and written guidelines on burden of proof (which should be on the prosecutor or accusing party), standard of proof (with all substantive elements necessary to convict a person), and admissibility of evidence.
- Vet potential judges vigorously and impose educational requirements; provide adequate judicial training in advance of all trials and throughout the process.
- Provide some form of payment or benefits to judges to help reduce the risk of corruption.
- Introduce mechanisms for independent trial monitoring, with emphasis on identifying corruption and trial manipulation by private individuals or government officials.
- Confer jurisdiction for sexual violence offenses to conventional or specialized courts located outside victims’ local communities.
- Create an ombudsman or other independent oversight body to receive complaints and oversee appropriate investigation of allegations of errors of law, violations of due process, and other abuses.
Human Rights Watch has closely monitored the work of gacaca courts since their creation in June 2002. It has observed all phases of the gacaca process: information gathering, categorization of suspects, trials, appeals, and final revision of judgments. Human Rights Watch followed more than 350 gacaca cases in total, some from initial trial hearings to the last review or revision stage and others only at the later appellate or revision stages. Many of these cases continued over the course of several years. Cases were chosen from different parts of the country, with trial observers dispersed among Rwanda’s four provinces and the capital Kigali.
Human Rights Watch researchers and consultants observing and translating gacaca proceedings spent more than 2,000 days observing trials, conducting interviews, and investigating cases over the course of eight years. Researchers conducted hundreds of interviews with participants from all sides of the gacaca process, including accused persons, genocide survivors, witnesses, other community members, judges, district coordinators, and local and national government officials.
Human Rights Watch met privately with officials at the National Service of Gacaca Jurisdictions (SNJG) and other parts of the Justice Ministry and with international donors supporting the gacaca program and participated collectively in larger consultation meetings. Human Rights Watch also regularly conferred with other nongovernmental organizations (NGOs) monitoring gacaca and, at times, undertook joint initiatives with them to raise particular areas of concern with SNJG authorities. Human Rights Watch also met with members of the National Human Rights Commission to discuss particular cases.
On March 30, 2011, Human Rights Watch wrote to the Minister of Justice and the Executive Secretary of the SNJG with a summary of the findings of the present report and its main recommendations. Human Rights Watch invited comments from the government with a view to reflecting its perspective in the final version of the report. On May 5, 2011, the Minister of Justice responded to Human Rights Watch’s letter with written comments. A copy of the Minister of Justice’s comments and Human Rights Watch’s letter is annexed to this report.
This report seeks to provide a representative sample of Human Rights Watch’s findings on gacaca on issues of key importance. Given the large number of trials observed and interviews conducted, Human Rights Watch is unable to include references to all of the cases in which it found areas of concern or abuses. A small number of cases with multiple irregularities are used throughout the report to illustrate broader patterns observed in the course of researching the report and to underscore how trials often suffered from numerous procedural errors.
The report provides as much detail as possible on the trials Human Rights Watch observed, including case names where possible. Many individuals interviewed said they feared reprisals by the authorities for speaking openly with Human Rights Watch and only agreed to comment on their personal experiences and on the gacaca system as a whole if their confidentiality was guaranteed. The identity and other details about some interviewees are therefore omitted from references.
Between April and July 1994, Rwanda experienced the darkest and most brutal period of violence in its history. On April 6, 1994, President Juvénal Habyarimana was returning from peace negotiations in Tanzania with the Rwandan Patriotic Front (RPF)—a rebel group consisting mainly of Rwandan Tutsi refugees in Uganda—when the plane in which he was travelling was shot down over Kigali by unknown assailants. All on board were killed. Within hours, Hutu extremists seized control of the government and military and began executing the political elite who might oppose their plans.
Assisted by tens of thousands of soldiers, local militia, and ordinary citizens, the extremists launched a three-month nation-wide genocidal campaign to wipe out the country's minority Tutsi population. More than half a million Tutsi and Hutu who tried to stop the massacres or protect Tutsi were killed while the world looked on. Meanwhile, the RPF entered the country from Uganda and began taking over parts of the country. By mid-July, the RPF had ended the genocide, seizing control of Kigali and the rest of the country.
Seventeen years later, the RPF remains in power and the genocide continues to weigh heavily on the country. Many Rwandans continue to see each other through an ethnic lens and distrust persons of different ethnicity. People frequently speak of the relatives they lost or the harm they suffered in 1994 and struggle with trauma and vivid memories of the genocide.
Delivering justice for mass atrocities is a daunting challenge, and the scale and complexity of the genocide would have overwhelmed even the best-equipped judicial system. In Rwanda—where the justice system was under-resourced before the genocide—the task was made even more difficult because of the vast number of judges and other judicial staff killed during the genocide and the destruction of much of the country’s infrastructure.
Tens of thousands of suspects were arrested after the genocide, often on the basis of a single unsubstantiated accusation of participation in the genocide. The number of detainees grew rapidly and quickly overwhelmed the prison system. By October 1994, an estimated 58,000 persons were detained in prison space intended for 12,000, and by 1998, the number of prisoners had reached around 130,000. Extreme overcrowding and lack of sanitation, food, and medical care created conditions that were universally acknowledged to be inhumane and which claimed thousands of lives. Many persons were held for years without charge and without their cases being investigated.
In December 1996 the government began to prosecute genocide suspects in conventional courts. By early 1998, only 1,292 persons had been judged and relatively few people had confessed to their crimes. The authorities realized that, at this rate, it would take decades to prosecute the large number of detainees. Yet they turned down proposals for foreign judges and other legal personnel to work alongside Rwandan judicial officials to help speed up the process.
In January 1998, Vice-President Paul Kagame announced that Rwanda could no longer afford the US$20 million a year necessary to support the huge prison population. The government proposed that the most notorious perpetrators be executed (the death penalty being the maximum penalty for genocide at that time) and that others be tried through a customary judicial mechanism, with some sentenced to prison terms and others serving terms of forced labor on public work projects.In April 1998, 22 persons convicted of genocide were executed, the first and only formal executions carried out in connection with the genocide. Most had been convicted in unfair and summary trials.
The government then set up a commission to assess problems facing the country and to propose means of addressing them. Between May 1998 and March 1999, the commission met weekly to discuss issues of unity, democracy, justice, security, and the economy and consulted with representatives of Rwandan society on these issues. One of the concerns discussed was how to provide justice for the genocide. The commission contemplated whether it would be possible to modernize the customary dispute resolution mechanism of gacaca to enable it to handle genocide-related cases. In August 1999, the commission set out the blueprint for the new gacaca system.Its report provided insight into the government’s broader political objectives and included a range of views, including some skeptical of the proposal to use gacaca.
Discussions within the commission focused heavily on the concepts of accountability and national unity. Three main arguments were advanced for using gacaca for genocide trials. First, it would accelerate the process of delivering justice for the genocide and would ease prison overcrowding. The commission estimated that without gacaca, it would take conventional courts approximately 200 years to try these cases. Second, like the conventional courts, gacaca would break the cycle of impunity by holding individuals responsible for crimes, rather than entire families or larger communities. Third, the participatory nature of the gacaca process could help reunite local communities. With trials taking place in the very location where the crimes had occurred and with neighbors, families, and friends looking on, local communities would play an important role in the proceedings and would see justice being done; this in turn would give them greater ownership of the process.
Some of the commission’s members, however, expressed concerns about whether gacaca was the best means of resolving genocide-related cases. They feared that using gacaca courts—traditionally reserved for small civil disputes—would minimize the seriousness of the crimes. Some also questioned whether ordinary citizens, often uneducated and with no formal legal training, had the skills to manage the trials and to apply national laws correctly. Others worried that relatives and friends with close connections to the community might be unduly influenced and show partiality in their decisions, creating new conflicts and tensions.
Some members expressed concern that witnesses with personal scores to settle or perpetrators with crimes to hide might give false information. They stressed that in certain parts of the country, there were no survivors left to testify or to challenge false testimony. Finally, some members worried that gacaca trials might not meet international fair trial standards. Those with reservations suggested that gacaca might be better used as an investigative tool to gather evidence at the local level which could assist conventional courts. Many of these concerns turned out to be well-founded.
Proponents of gacaca argued against these reservations and ultimately won. They emphasized that using gacaca for genocide crimes would not trivialize the crimes, but rather would force communities to deal with the crimes at the level where they happened and would help end impunity locally. They also argued that ordinary citizens could be trained to apply the law correctly and could receive assistance from lawyers where necessary.
Some members said that carrying out the trials in public would reduce the risk of judges taking sides and would discourage community members from giving false testimony. In their thinking, the advantages of using gacaca to individualize guilt, to dispel the notion that all Hutu committed genocide, and to give ordinary Rwandans an active role in delivering justice for the genocide far outweighed any potential limitations.
In June 2002, Vice-President Kagame officially launched gacaca courts to try genocide-related cases and announced five core objectives:
- Reveal the truth about what happened;
- Accelerate genocide trials;
- Eradicate the culture of impunity;
- Reconcile Rwandans and reinforce their unity; and
- Prove that Rwanda has the capacity to resolve its own problems.
Gacaca draws its inspiration from past efforts by local communities to resolve disputes. However, in designing gacaca for genocide-related cases, the government made significant changes to the customary model, transforming it into a more formal, state-run judicial apparatus. The courts’ work began in stages, allowing for amendments and other fine-tuning of the system before gacaca was rolled out nationwide in 2005. Originally set to end in 2007, the date for its completion has been postponed several times. At the time of writing, it is not clear when the process will finally end.
Differences between the Customary and Contemporary Gacaca Systems
The Rwandan government portrayed its decision to use gacaca for genocide-related cases as “revert[ing] to our traditional methods of conflict resolution.” However, other than in name and certain general characteristics, the version of gacaca used to try genocide-related cases bears little resemblance to the customary form.
Little documentation exists about gacaca before 1994. The practice is believed to have come into existence in the pre-colonial period but continued to be used during colonialism and after independence in 1962. In customary gacaca, respected community elders known as inyangamugayo (literally “those who detest disgrace”) came together as required to mediate family and inter-family disputes related to property, inheritance, personal injury, and marital relations. More serious matters such as cattle theft, murder, or other crimes were left to community chiefs or the king’s representative for resolution.
The customary gacaca gatherings usually involved only community elders, the disputing parties, their relatives, and immediate neighbors. The gatherings were dominated by older men since women were not permitted to speak. Customary gacaca emphasized restoring social harmony, with punishment of the perpetrator and compensation to the victim being of lesser importance. Punishment was not individualized, meaning that family and clan members of the accused were also held responsible. Often, the losing party had to provide beer to the community as a means of reconciliation.
The modern version of gacaca continued the ritual of hearing cases at the local level but differed in five main ways. First, it handled serious crimes—with genocide arguably being the gravest of crimes—rather than minor, civil disputes. Second, it was fundamentally retributive or punitive in nature, with the exception of cases involving property crimes. Gacaca courts could impose prison sentences ranging from short terms to “life imprisonment with special provisions.” Reconciliation and restoration of social order remained objectives of contemporary gacaca, but they were secondary to the punitive process.
Third, gacaca was governed by an official state institution under the Ministry of Justice (the SNJG) and was therefore intimately linked to the state apparatus of prosecutions and incarceration. Fourth, gacaca applied codified, rather than customary, law. Finally, gacaca judges were not community elders but were instead elected community members (inyangamugayo) and were often relatively young. Women also made up a significant percentage of the judges.
The Legal Framework Governing Genocide Cases and Gacaca Courts
Rwanda’s first genocide law
In 1996, the Rwandan Parliament provided the legal framework for conventional courts to try crimes of genocide committed between 1990 and 1994. The Genocide Law contained two novel features.
First, it classified suspects into four groups: category 1 included planners, leaders, organizers, and instigators of the genocide, well-known killers, and rapists; category 2 included persons who committed homicide; category 3 included those who killed or inflicted bodily harm without the intention to kill; and category 4 included those who stole or damaged property. The law prescribed capital punishment for those in category 1, and varying prison sentences and damages for the remaining categories.
Second, the law introduced the common law practice of plea-bargaining, allowing courts to reduce sentences for those who confessed to their crimes and named their accomplices. These provisions aimed to speed up the trials of genocide suspects, but initially, very few people confessed to their crimes. Consequently, the government found itself looking for alternative solutions to deal with the huge backlog of genocide cases.
The gacaca laws
In 2001, Parliament adopted legislation creating gacaca courts, giving them jurisdiction over serious crimes committed between October 1, 1990 and December 31, 1994, and the ability to judge all suspects except those in category 1 (whose cases remained before the conventional courts). The definition of “genocide” in this law largely followed the definition of the Genocide Law but also required that violations be committed with genocidal intent in order to be qualified as genocide.
Since 2001, Parliament has amended the gacaca law four times, usually to simplify and accelerate the way in which the courts process cases. The 2004 law, for example, decreased the number of levels of gacaca jurisdictions, reduced the number of categories from four to three, and reduced the number of judges required to hear cases from 19 to 7.
Under the 2004 law, gacaca courts were to operate at two local levels (known as the cell and sector levels) in each jurisdiction. Cell level courts handled the information gathering phase and classified suspects. They also tried category 3 cases relating to property offenses. All other genocide-related trials (involving category 2 and later category 1 offenses) occurred at the sector level. Separate gacaca courts at the sector level handled all appeals.
The 2007 law increased the number of courts in each area and gave them jurisdiction over well-known killers, previously classified under category 1, who until then had remained under the jurisdiction of the conventional courts. The law also declared that judges could decide cases so long as five of the seven judges were present.
Then, in 2008, the government decided to transfer most of the remaining genocide cases in conventional courts to gacaca jurisdictions to alleviate the backlog in the conventional courts.Since then, the jurisdiction of the conventional courts has only covered cases of those accused of being ringleaders and individuals who held official positions at the préfecture (province) level or higher.
Unlike conventional courts, gacaca courts had no prosecutors. Cases depended largely on accusations initiated by a “civil party,” usually the victim of the crime or his or her relatives. Panels of five to seven judges heard each case, with one of the judges presiding over proceedings. Judgment was reached by majority rule.
Gacaca’s Pilot Phase
Contemporary gacaca courts were launched on June 18, 2002 in 12 pilot sectors around the country.The pilot phase took place in three stages:
The information gathering phase (“collecte d’information” as it is more commonly known) involved community members aged 18 or above meeting every week in a “general assembly” to gather information on who lived in the area in 1994, who was killed (and if possible whether it happened in that area or in another area), whose property had been damaged, and who had participated in the killings and destruction.
Once all relevant information had been collected, the cell level courts categorized suspects according to the severity of the alleged crimes. Category 3 suspects, accused of property damage and looting, were to be released from prison and their files sent to the cell level courts for resolution. Category 1 and 2 suspects were to remain in detention, with category 2 suspects to be heard by the sector level gacaca courts and category 1 suspects to be transferred to the national prosecutor’s office for trial in the conventional courts.
It was not until the end of 2004, two-and-a-half years after the pilot phase began, that the first two steps were completed and that gacaca courts were ready to proceed to the trial phase. Rather than wait for the information gathering phase and the categorization of suspects to be completed nationwide, the government authorized pilot jurisdictions to begin trials on March 10, 2005. Trials proceeded expeditiously with more than 650 persons tried in the first six weeks.
Trials typically took place on the grass outside of the community’s local administrative office, although occasionally sessions were held in school classrooms or other public buildings (particularly during the rainy season). Judges wearing official sashes sat on benches at the front and often had a table in front of them so that at least one of the judges could take notes of the proceedings. Accused persons sat to one side of the judges or in front of the community, with the civil party sitting to the other side. Interested parties rose and stood in front of the judges and community when testifying. Community members gathered on the grass and under nearby trees facing the judges and could speak freely after witnesses had completed their testimony. Gacaca sessions lasted anywhere from an hour to an entire day, with some trials concluded in a single session and others requiring several weekly sessions.
National Implementation of Gacaca
The government launched the nationwide information gathering phase on January 15, 2005. There was one important procedural change compared to the pilot phase. Instead of gathering information through weekly community meetings, the SNJG tasked local authorities known as nyumbakumi (in charge of “10 households”) to collect information. These local officials gathered information by assembling small groups or by going door-to-door and later presented the written accusations to the whole community for verification. There was no meaningful community debate on the nature or veracity of accusations during the verification process. The changes were designed to accelerate the collection of relevant information and to make the process more efficient. However, the new measures limited the transparency of the process and made it easier for officials and others working with them to construct false or poorly documented accusations against individuals for personal or political ends.Consequently, the use of nyumbakumi compromised the integrity of the nationwide information gathering phase.
By the end of the information gathering phase nearly 18 months later, the SNJG reported accusations against 818,564 persons broken down in the following categories:
Category 1 *
Of this total, more than 100,000 suspects were believed to have died or to be living outside Rwanda. Once persons classified in category 1 (to be tried in the conventional courts) had been removed from the list, gacaca jurisdictions faced a daunting caseload of 610,028 persons.
Trials began nationwide on July 15, 2006 in more than 12,000 jurisdictions. Three months later, the SNJG reported that 16,801 accused had been judged, 2,546 of whom had been acquitted. By December 2006, the number of judgments had jumped to almost 40,000. The Minister of Justice Tharcisse Karugarama announced that all trials would be completed by the end of 2007.
Repeated Extensions of Gacaca’s Closing Date
The initial target may have been too ambitious. By the end of February 2007, gacaca courts had tried only 50,000 cases—a seemingly large number, but a small proportion of the overall total. The relatively slow pace of trials, coupled with new cases appearing around the country, meant that gacaca courts were unlikely to meet the December 2007 deadline. Parliament attempted to remedy the situation by adopting a new law in March 2007 that allowed gacaca jurisdictions to have multiple courts. The move—which permitted multiple trials to take place in a given community at the same time—accelerated trials but resulted in potential witnesses and interested parties sometimes having to choose between attending different trials.
Pressure mounted on gacaca judges as the deadline grew near, and trials began to take place at an alarming speed with some individuals sentenced to life imprisonment in trials lasting less than an hour. Human rights groups monitoring the processexpressed concern that the quality of decision-making was being sacrificed for the sake of speed.
The SNJG eventually realized that it could not meet the deadline and extended it to 2008. At the end of September 2008, it announced that it had 1,127,706 cases on record (involving category 1 and 2 suspects), of which only 4,679 remained pending in gacaca courts. However, genocide accusations continued to emerge and gave rise to new cases.
Meanwhile, genocide cases in the conventional courts progressed slowly with only 222 cases concluded between January 2005 and March 2008. Realizing that at that pace it would take decades to complete category 1 trials, Parliament transferred most remaining category 1 cases to gacaca jurisdictions in June 2008. The SNJG indicated that 90 percent of the cases transferred (at least 8,000 cases) involved rape or sexual violence and would be held behind closed doors. The SNJG announced that all gacaca trials would be completed by June 2009, but later extended the deadline to December 2009 and then again to April 2010. The SNJG announced the completion of all gacaca trials in July 2010, but three months later, surprisingly announced that the gacaca process would continue. At the time of writing, the SNJG reported completion of the following number of cases:
The Final Phase of Gacaca
The SNJG’s announcement, in October 2010, that the gacaca process would continue surprised many who believed the process had already ended. The decision to extend gacaca was taken after a September 2010 meeting among representatives of the Ministry of Justice, the SNJG, the Office of the Ombudsman, the National Human Rights Commission (NHRC), and the Ministry of Local Government, all of whom had received complaints from citizens concerning gacaca. A confidential report sent by the NHRC to President Kagame, citing more than 25 cases of serious injustice, and another internal government report compiled by the Office of the Ombudsman, citing more than 230 complaints it had received in connection with gacaca, may have been catalysts for the meeting. The meeting concluded with an agreement that all national institutions would transmit their complaints and gacaca files to the SNJG by early November 2010 and that the SNJG would implement a procedure to review these complaints and any others it had previously received.
According to the Ministry of Justice, by May 2011 government institutions had received approximately 1,000 applications for review. In March 2011, the SNJG spokesperson told Human Rights Watch that the SNJG had identified 40 cases as meriting additional review.The SNJG has not publicly divulged the criteria used in determining which cases merit review or explained the manner in which cases are being reviewed. Once the SNJG has completed its examination of the case files, it expects to send those cases meriting review back to gacaca jurisdictions for a new hearing and possible corrective measures.
The question of how to expeditiously resolve the backlog of genocide-related cases without compromising fair trial rights has been at the forefront of discussions on gacaca for years. The Rwandan government chose gacaca because it would be quick and informal. Yet it faced the daunting task of balancing these benefits with more formal fair trial standards enshrined in Rwandan law and international treaties to which Rwanda is a party. The government made a number of substantial compromises, particularly in relation to the rights of the accused, judges’ qualifications, and applicable legal standards. It believed that the transparency of the gacaca process and the participation of the entire population would legitimize the process and protect the rights of all participants, rendering formal fair trial guarantees unnecessary. Human Rights Watch believes that these compromises did not adequately protect the rights of the parties and led in many instances to unfair trials.
Limited International Fair Trial Rights in Gacaca
The Rwandan constitution, domestic laws, and international treaties to which Rwanda is a party guarantee certain minimum fair trial rights. These include: the right to a lawyer, the right to be presumed innocent, the right to be informed of the charges against oneself and to have adequate time to prepare a defense, the right to be present at one’s trial and to confront witnesses, the right against self-incrimination, the right not to be tried twice for the same crime, and the right to be free from arbitrary arrest and detention.
The Rwandan government has expressly or implicitly attempted to guarantee some of these rights but has modified others, such as the right to have adequate time to prepare a defense. Other rights, such as the right to a lawyer, have been entirely sacrificed in order to achieve a rapid resolution of cases. In 2009, the UN Human Rights Committee assessed Rwanda’s compliance with its obligations under the International Covenant on Civil and Political Rights (ICCPR) and concluded that the gacaca system did not operate in accordance with basic fair trial rules. The Committee raised particular concerns about the protection of the rights of the accused and the impartiality of judges. UN High Commissioner for Human Rights Louise Arbour raised due process concerns during her visit to Rwanda in May 2007, citing the “worrisome haste” of trials, the lack of legal training for gacaca judges, and the heavy penalties imposed on convicted persons. The Rwandan government broadly disregarded these criticisms and made clear, in discussions with donors and human rights organizations, that compliance with its international obligations in this context was not its top priority.
The right to counsel
Both Rwandan and international law guarantee the right to legal counsel.Gacaca jurisdictions remain an exception to this rule, with the accused having no access to counsel at any stage of proceedings. The right to counsel is not expressly curtailed in any of the gacaca laws, but the SNJG has repeatedly made clear that such representation is not permitted.
The government justified its decision to exclude defense lawyers from gacaca courts on four grounds. First, the high number of accused persons would make it impossible for all of them to have lawyers without significantly delaying the trials. Second, lawyers might unduly influence the non-professional gacaca judges who have a limited understanding of the law. Third, the local community’s participation at trials would be sufficient to guarantee a fair trial because community members could speak out if a witness lied and could question witnesses. Finally, emphasizing community participation instead of the use of lawyers would maximize the community’s sense of ownership.
There have been a handful of cases where lawyers have nonetheless tried to appear on behalf of an accused person. In one instance, a lawyer defending an accused man in a conventional court was allowed to continue advising his client after the case was transferred to a gacaca court (although he was not allowed to wear his robe).
In the 2009 case of human rights activist François-Xavier Byuma, described below, the SNJG also reluctantly permitted a lawyer to be present but then allowed the court to take measures rendering his assistance ineffective.
The government’s need to complete all genocide trials in years rather than decades made the usual type of legal representation for each and every accused impossible. However, the Rwandan government could have put in place alternative measures to guarantee the accused the right to legal assistance. For example, in May 2002, the Danish Institute of Human Rights proposed that the Rwandan Ministry of Justice allow national and international judicial defenders (legal professionals, but non-lawyers), funded by foreign donors, to provide the accused, as well as genocide survivors, with pre-trial legal advice. Aimed at informing the accused and victims of their rights and enhancing the transparency of the gacaca process, the proposal also sought to speed up the judicial process by encouraging the guilty to confess in exchange for reduced sentences under the system’s plea-bargaining scheme.The Ministry of Justice never responded to the proposal.
The story of François-Xavier Byuma
Human rights activist François-Xavier Byuma was convicted of genocide-related charges following a gacaca trial that violated both Rwandan law and the fundamental principle that accused persons must be tried before a fair and impartial court.
At the time allegations of genocide first surfaced against him in early 2007, Byuma headed an organization for the defense of children’s rights (Turengere Abana) and had recently started investigations into an allegation of rape of a 17-year-old girl by a local gacaca judge. Knowing that this judge would preside over his case, Byuma immediately wrote to the SNJG expressing concern that he may not receive a fair trial. His letter was found to be “baseless and unfounded.”
Byuma’s trial began in Kigali on May 13, 2007. He was accused of having been present at a roadblock erected to prevent Tutsi fleeing the genocide, having a firearm, and participating in weapons training. At the outset of the trial, Byuma asked to have the judges dismissed on the grounds of conflict of interest, but the court declined his request.In protest, Byuma refused to testify. The judge threatened to charge him for his refusal to testify. Byuma decided to subject himself to the jurisdiction, despite overt hostility shown by the presiding judge throughout the remainder of the trial.
At a second hearing a week later, Byuma defended himself against the charges, but the presiding judge cut off many of his answers and those of witnesses who tried to speak in his defense. In one instance, the presiding judge accused a defense witness of lying. The court found Byuma guilty of participating in weapons training and several other counts (including participation in an attack and abduction and assault of a Tutsi woman) which were not mentioned when the charges were first read to Byuma before trial. It sentenced him to 19 years in prison. The court acquitted two co-accused of the same crimes, despite one of them having admitted to being guilty of one of the charges.
On August 18, 2007, an appeals court upheld the 19-year prison sentence despite numerous irregularities. Byuma had presented court records revealing that one prosecution witness who accused him of assault had previously testified that a different person committed the crime (and whose name the witness never mentioned in the gacaca case). Byuma pointed out that the trial court declined to hear some of the witnesses whom he sought to call in his defense and failed to reconcile contradictions in the evidence. The appeals court gave no justification for its decision affirming the conviction and offered no explanation for its failure to deal with the fact that the presiding judge of the lower court had a clear conflict of interest with Byuma.
After a strong public outcry from local and international organizations, the SNJG accepted Byuma’s request for revision and brought a bench of judges from the eastern part of the country to decide the case. The SNJG reluctantly agreed to allow a lawyer to assist Byuma in his defense (as long as he did not wear his robe). During the hearing, however, the lawyer provided by Avocats Sans Frontières (ASF) was not permitted to sit next to his client and was repeatedly denied the opportunity to question witnesses. These restrictions, coupled with the open hostility shown by the presiding judge to the lawyer’s presence, rendered his assistance ineffective.
The court deciding Byuma’s request for revision gave little consideration to additional defense witnesses who testified but concluded that new evidence had been offered by accusing witnesses, even though some of this information was inconsistent with earlier testimony given at trial and on appeal. The court also found Byuma guilty of possessing a firearm, in violation of a 2006 SNJG directive which stated that having a firearm or being at a roadblock did not in itself constitute a crime. The court upheld Byuma’s conviction but reduced his sentence to 17 years’ imprisonment.
Byuma’s case is also discussed in connection with the right to be presumed innocent, the right to present defense witnesses, and the right against self-incrimination.
The presumption of innocence
The Rwandan Constitution, the Rwandan Code of Criminal Procedure, the ICCPR, and the African Charter on Human and Peoples’ Rights (ACHPR) to which Rwanda is also a party, all guarantee that an accused person be presumed innocent until he or she is proven guilty. However, this fundamental right has not always been respected in gacaca trials.
The UN Human Rights Committee, in its General Comment on Article 14, in reference to the presumption of innocence has advised that there is a “duty for all public authorities to refrain from prejudging the outcome of a trial.” Yet senior Rwandan government officials and pro-government media have at times repeatedly and publicly labeled persons as guilty of genocide-related crimes before their gacaca trials were concluded, and in some cases even before the individuals were formally charged in gacaca. Most often this occurred in high-profile cases against political opponents or critics of the government. Such statements created an atmosphere in which it was difficult to ensure that a person would be presumed innocent and would be judged solely on the basis of evidence presented at his or her trial.
The same is true for persons accused of “sectarianism” (more commonly known as “divisionism”) and “genocide ideology”—both vaguely defined by law to prohibit ideas, statements, or conduct that might lead to ethnic animosity or violence. The government’s campaign to denounce persons suspected of these crimes, including pre-trial statements by public officials, often receiving significant media coverage, has further called into question the extent to which a person facing trial can effectively enjoy a presumption of innocence. Between 2003 and 2008, four parliamentary commissions investigated and denounced purported cases of “divisionism” and “genocide ideology,” with little or no verification of the facts and no judicial process. In a May 2007 statement responding to accusations that police officers had killed 20 detainees, the Commissioner General of the Rwandan National Police Andrew Rwigamba (formerly chief military prosecutor) said that the detainees had been “of extreme criminal character ready to die for their genocide ideology.” In fact, the detainees, all recently arrested, had not been tried for any crimes and none had been convicted of holding “genocide ideology.” The Commissioner’s statement was made in the context of the larger government campaign against genocide ideology and at a time when gacaca trials were operating at full speed.
Government officials have used these accusations—along with accusations of “revisionism,” “negationism,” and “gross minimization of genocide,” all of which are proscribed by the Rwandan Constitution and a 2003 law punishing genocide—as tools to quash debate on sensitive issues, silence independent opinion and criticism, and pursue political opponents. In April 2009 the government suspended the BBC Kinyarwanda service for two months for trying to give a public platform to individuals the government described as “genocide deniers,” including Faustin Twagiramungu, the first prime minister in the government formed by the RPF after the genocide and a 2003 presidential candidate against President Kagame. The BBC program was never aired. Individuals calling for justice for victims of crimes committed by RPF soldiers in 1994 or attempting to challenge the ruling party in presidential elections in 2010 faced public denunciation and, in some cases, formal criminal accusations. Government officials publicly accused political opponents Victoire Ingabire (president of the FDU-Inkingi opposition party) and Bernard Ntaganda (president of the PS-Imberakuri opposition party) as well as outspoken critic Déogratias Mushayidi of “divisionism” and “genocide ideology.” All were later formally charged with these crimes.
These examples involved prominent opposition figures and critics, but ordinary citizens have also faced “genocide ideology” accusations. More than 2,000 cases have been brought before Rwandan courts, some even before the 2008 genocide ideology law was adopted. A significant proportion of these cases have resulted in acquittals, but often after accused persons spent long periods in detention. The laws on “divisionism” and “genocide ideology” have had a chilling effect on respect for the presumption of innocence, as well as freedom of expression (discussed later in the report).
The right to be presumed innocent also means that a court will not prejudge an accused or treat him or her as if the person is guilty, regardless of the likelihood of conviction. Yet Human Rights Watch documented dozens of cases in which judges demonstrated preconceived notions of guilt or treated the accused as if he or she were guilty from the outset of trial. Often these types of cases involved judges making disparaging remarks or using a hostile tone toward the accused. For example, in a January 2008 case in the west of the country, a presiding judge opened the trial by asking whether the accused wanted to plead guilty. When the accused said no, the judge said “you are not innocent because you are being prosecuted for crimes of genocide committed in this prefecture.”
In another case in 2009, discussed below, the presiding judge encouraged two of the accused to plead guilty at the beginning of their trial. When they declined, the judge stated in a sarcastic manner that the accused clearly did not understand the benefits of pleading guilty—intimating that the accused should have pleaded guilty. The judge proceeded to treat both accused with hostility throughout the rest of the trial.
Human Rights Watch also observed cases where judges demonstrated bias towards the accused or defense witnesses. The trial of human rights activist François-Xavier Byuma, discussed above, is a prime example. Judges at both the trial and revision stages showed overt hostility to the accused, frequently interrupting his testimony and that of his witnesses. At the trial level, the reason for the presiding judge’s hostility was obvious: Byuma had previously investigated rape allegations made against the judge. There was no known conflict of interest with the judges at the revision stage, yet there too, the judges—and in particular the presiding judge—showed overt bias against the accused from the outset of proceedings, for example through the presiding judge’s hostile tone toward the accused and his frequent interruptions of Byuma’s testimony.
Gacaca courts also violated the presumption of innocence by shifting the burden of proof to the accused and relying on the accused to prove that he or she did not commit the alleged crime.This issue will be discussed in further detail later in this report.
The right to be informed of the case and to have time to prepare a defense
Under the Rwandan Constitution, the Rwandan Code of Criminal Procedure and the ICCPR, the fair trial rights of an accused include the right to be informed of the accusations against him or her and the right to have sufficient time to prepare a defense.In gacaca cases, these rights have not always been respected: many accused did not receive the legally prescribed notice of cases pending against them, were not provided with sufficient pre-trial information about the charges against them, and were not given enough time to prepare their defense. Many accused only learned of the real nature of the allegations against them on the day of their trial. The inability of the accused to involve a lawyer only aggravated these problems.
A person facing the prospect of gacaca charges may have learned through information provided by community members during the public information gathering phase that he or she would be summoned before gacaca. However, in 2005, the procedure changed: local officials would go door-to-door or meet with small groups of community members to gather information on crimes committed (as opposed to weekly meetings where community members raised and debated potential allegations against an accused), with the result that accused persons were less likely to learn of their impending trial.
By law, the gacaca jurisdiction must deliver a formal summons to any person asked to appear at a trial. The summons should indicate the following information: whether the person is requested to appear as an accused or as a witness; whether the person is incarcerated and, if not, the person’s address; the charges against the person and the category of the alleged crimes; and the time, date, and location of the hearing. The summons should be signed by the gacaca jurisdiction’s secretary and countersigned by the person to whom it is addressed at the time the person receives it.
The summons should be delivered to an accused at his or her current home or last known place of residence at least seven days before the person is scheduled to appear before gacaca. The district coordinator, who assists gacaca judges and oversees the management of their caseload, normally delivers the summons to the person called to appear. Where the accused has no known residence in Rwanda and is believed to live abroad, the summons should be posted at least one month in advance on the wall of the gacaca office in the jurisdiction and at the district and provincial government offices. If the person receives the notice less than seven days before the hearing, the court should automatically postpone the hearing and issue a new summons that complies with the prescribed time period.
Failure to comply with summons procedures
In some cases, simple error explained why an accused did not receive a summons according to the required procedure. In other cases, failure to send a summons may have been deliberate in order to trigger a conviction, since courts will decide a case in the absence of the accused if he or she fails to appear three consecutive times. Occasionally, local or gacaca officials did not deliver summonses to the accused or to his or her residence, and instead simply gave them to relatives or friends to pass on to the accused.
In one case in 2007, an individual learned by accident that he had been accused of genocide-related crimes. Former neighbors attending the weekly gacaca trials heard the presiding judge announce that a default judgment would soon be issued against the man. The accused had not received a summons, even though the authorities regularly sent notice of other administrative decisions to his address in the area where the crimes were alleged to have taken place. Two years later, when the same case reappeared before the same gacaca court, the presiding judge tasked with delivering the summons never gave it to him.
Another accused, Innocent Nizeyimana, happened to learn of an ongoing gacaca case against him through an acquaintance in 2007. Neither he nor his family had received a summons, despite the fact that he continued to own property in the area where he had lived in 1994 to which a summons could have been sent. He also resided in a nearby neighborhood of Kigali. When Nizeyimana spoke to the presiding judge, he was told that he would be convicted at the next session if he did not appear. The presiding judge was unable to say whether a summons had ever been issued. In later proceedings, gacaca authorities sent Nizeyimana a text message with the date and location of the proceedings but never delivered a summons. Other accused persons also received notification of upcoming gacaca hearings by text message.
Human Rights Watch documented many cases in which summonses were delivered less than seven days before the hearing, which prejudiced the ability of the accused to prepare a defense. In a striking case in December 2009, Symphorien Kamuzinzi received his summons at 6 p.m. on the eve of his trial.Most often, however, summonses were delivered at least the day before the person was to appear in gacaca.
Detainees were the most likely to receive their summons late and, as a result of the late notice, typically had the greatest difficulty in ensuring that their witnesses came to the trial. In some cases, detainees only learned of the trial the morning of their scheduled appearance. In one instance, a detainee first realized that he was to appear in gacaca when a prison guard came and told him to get into the prison vehicle. The man requested that the hearing be delayed, but the presiding judge refused and the trial went ahead. He was convicted and sentenced to “life imprisonment with special provisions.”
In some instances, the gacaca court immediately addressed the failure to give sufficient notice by postponing the hearing. In other instances, the authorities’ failure to provide notice enabled persons convicted at the first instance to successfully appeal and obtain a new hearing. But in other cases, the accused were forced to proceed with their trial despite not having received adequate notice.
Failure to provide adequate information on the charges against the accused
In many cases, summonses did not contain enough information about the charges against the accused, as required by law. In most of the cases where Human Rights Watch detected irregularities, the “accusations” line was simply left blank, leaving the accused with no idea of the accusations against him or her. Where charges were specified, they usually consisted of general accusations such as “genocide” or “murder” with no details of the specific incident or crime. Such vague information did not enable the accused to prepare a defense in advance of his or her trial.
In several particularly troubling cases, individuals were informed that they should appear as a “witness” in someone else's trial and only discovered upon arrival at the hearing that they themselves were accused.Some of these cases demonstrate the risk that defense witnesses face, an issue discussed in further detail in section VII of this report.
In a 2009 case in the southern province, a gacaca court summoned Domina Nyirakabano as a witness but, upon her arrival, arrested her and imposed a sentence of 30 years' imprisonment.She appealed the conviction, and it was overturned more than five months later, at which time she was released.
Where individuals did not have enough information about the allegations against them or were unsure whether accusations were pending against them, they sometimes approached the district coordinator, gacaca judges, or local authorities in their area for additional information. Many also conducted their own investigation into the accusations, with the help of friends and acquaintances in the community. The gacaca laws are silent on whether an accused has the right to receive supplemental information from gacaca or local administrative officials in advance of their trial.
In some cases, the authorities willingly provided the individuals with the information they requested. In others, individuals were compelled to pay to obtain information on charges pending against them even though payment was not legally required. In these instances, the payment amounted to a bribe in exchange for the requested information. For example, in an area near Gitarama, a farmer sold the only cow she owned to pay several gacaca judges to tell her whether any genocide accusations had been made against her.Similarly, gacaca officials required a university student in Kigali to pay 50,000 Rwandan francs (approximately US$82) to find out whether there was a case pending against him.
Failure to postpone hearings to give the accused adequate time to prepare a defense
Human Rights Watch documented numerous cases in which an accused requested an extension of time to obtain documents or to secure the appearance of defense witnesses. Some gacaca jurisdictions granted more time, but others refused and proceeded with the trial.
In a number of appeal cases, the convicted person had not been given a copy of the trial judgment or had not been given enough time to review it before the appeal hearing. Former Cyangugu sous-préfet (local official) Théodore Munyangabe and his co-accused, Abbé Aimé Mategeko, asked the appellate court to give them a copy of the trial judgment and to postpone the hearing until they had time to review it. The presiding judge responded that the judges had a copy and read a portion of it aloud. The accused then told the bench that the judgment did not conform to what had actually happened at trial, noting that the judgment listed new charges against Munyangabe and stated that Abbé Mategeko had confessed to certain crimes when he had not. In response, the presiding judge said, “Let’s forget those details and move on to the issue raised by this appeal.” Yet the issue raised was central to Munyangabe’s appeal because he claimed he was being retried for the same crime of which a conventional court had acquitted him. The appeals court proceeded to judge the case and upheld his conviction. Munyangabe’s case will be discussed in more detail later in this report.
In other cases, late notice of a hearing prejudiced the ability of an accused to gather witnesses in time for the hearing. Detained persons often had limited access to relatives who might otherwise have helped them find witnesses.Several examples are discussed in the next section.
The story of Pascal Habarugira
Dr. Pascal Habarugira was a doctor in the gynecology department at the University Hospital of Butare in 1994 and cared for a number of women and newborn children during the early period of the genocide. He returned briefly to his native town of Cyangugu in May 1994 before reaching Kigali in August where he began work at the Centre Hospitalier de Kigali (CHK). In 1995, he returned to the University Hospital of Butare and took up his prior position. The following year, Habarugira accepted a two-month internship in Paris, leading to rumors that he had fled the country, but he returned to Rwanda later that same year. The rumors persisted, and in 1999, he followed his wife to Côte d'Ivoire for her studies. The couple returned to Rwanda in 2003, and Habarugira resumed his functions at the hospital in Butare. Accusations against Habarugira were made during the gacaca information gathering phase in 2005, and police arrested him in March of that year as he left a medical conference in Kigali.
In August 2007, Habarugira faced trial before gacaca on five counts: participation in a genocide planning meeting, turning a Tutsi patient over to soldiers to be killed, attendance at a roadblock where killings occurred, membership in a crisis committee, and strangling a Tutsi newborn child.On September 5, 2007, the trial court convicted Habarugira of all but the first charge. Five other doctors were also convicted of having played a role in the death of Tutsi at Butare university hospital during the genocide. All were sentenced to 30 years’ imprisonment.
Habarugira appealed his conviction, arguing that the trial court had not considered the testimony of certain defense witnesses (including eight genocide survivors who worked with him at the hospital in 1994) and had ignored important inconsistencies in the testimony of several women who accused him. He tried to prove that one of the main witnesses against him was not in the area during the genocide as she had claimed, but the court refused to call a detained witness who could confirm this fact or to summon court records from another trial in which the woman had admitted to being elsewhere. Testimony on appeal revealed that Habarugira did not turn the Tutsi woman over to soldiers as alleged and that the newborn child whom he had allegedly killed was still alive (and present at the hearing with his mother). Despite the powerful testimony presented, the appeals court upheld the conviction on February 6, 2008, but reduced his sentence to 19 years’ imprisonment without any explanation for the reduced sentence.
The National Human Rights Commission (NHRC), which had sent monitors to observe the case, found so many due process violations at the appeals hearing that it wrote two separate letters to the SNJG executive secretary, calling for revision of the conviction. Of particular concern, the NHRC noted:
Habarugira’s request for revision was denied, first in April 2008 by the local gacaca jurisdiction, and again in June 2008 by the SNJG.At the time of writing, Habarugira remains in prison.
Habarugira’s case is also discussed in connection with the right to be presumed innocent and the right to present defense witnesses.
The right to present a defense
I cannot understand how you ask me to present my defense witnesses when I do not even know the charges against me in this case?
— An accused man at his trial.
The fact that many accused only learn of the precise allegations against them on the day of trial impedes their ability to prepare their defense and to find defense witnesses. This is particularly worrying given that most genocide prosecutions in Rwanda depend almost entirely on witness testimony.
Rwandan law does not guarantee an accused the right to summon witnesses in his or her defense, but the Rwandan Code of Criminal Procedure suggests that such a right exists because it sets out the procedure for witnesses to provide testimony. The ICCPR states that an accused is entitled “[t]o examine, or have examined, the witnesses against him and to obtain the attendance and examination of witnesses on his behalf under the same conditions as witnesses against him.” The ACHPR also guarantees the “right to [a] defence.”
The government’s campaign against “divisionism” and “genocide ideology” has proved to be a significant obstacle to securing defense testimony in gacaca courts. A number of persons interviewed by Human Rights Watch expressed fear that they might be accused of “genocide ideology” and imprisoned if they spoke in defense of accused persons or denounced survivors’ false testimony. With genocide ideology punishable by up to 25 years’ imprisonment, or life imprisonment for repeat offenders and those convicted of genocide, the perceived risks were high and unlikely to prompt lone voices for the defense to come forward.
The genocide ideology law’s impact on securing defense testimony was so significant that it contributed to the decision by the International Criminal Tribunal for Rwanda (ICTR) —created by the UN Security Council and tasked with prosecuting crimes that took place in Rwanda in 1994— and several foreign jurisdictions to deny the transfer of genocide-related cases to Rwanda for domestic prosecution. In response, the government adopted legislation in 2009 that precludes prosecution of witnesses for any in-court statements they make (other than perjury). Still, the new law does not seem to have assuaged the fears of Rwandans with whom Human Rights Watch spoke.
In April 2010, the Minister of Justice announced that the government was reviewing the “genocide ideology” law and had commissioned a study to examine weaknesses in the law.In January 2011, the Minister told the UN Human Rights Council that a proposal would soon be tabled before the Cabinet and later sent to Parliament for review.In May 2011, he informed Human Rights Watch that “a significant revision [of the genocide ideology law] has been drafted to address concerns that the law was overly vague and subject to abuse”. This may represent a significant move toward respecting free speech, but it came too late to positively impact gacaca. At the time of writing, neither the contents nor the exact timeframe for the adoption of any amendments are known.
Public officials and prominent community members occasionally intimidated or tried to influence witnesses and their testimony, further hindering efforts to secure defense witnesses. These issues will be discussed in greater detail in section VII of this report.
Human Rights Watch documented a number of cases in which courts obstructed the right of an accused to call witnesses in their defense, including refusing to hear defense witnesses who were physically present or declining the request of the accused to summon potential defense witnesses, such as in the case of Pascal Habarugira, discussed above.In its 2009 annual human rights report, the United States State Department also expressed concerns over gacaca courts’ refusal to allow the accused to present witnesses in their defense.
A lack of sufficient notice to accused persons in detention seriously compromised their ability to ensure their witnesses appeared at trial. In one case in 2008, police arrested a man in Kigali and held him at the police station for five days before transferring him back to his native region for trial the following day. As a result of his arrest and detention away from the place of trial and without any contact with his family, the man was unable to notify persons who could have appeared in his defense.
In another case in 2008, courts denied a detained woman the opportunity to present defense witnesses at her trial and at her appeal. The trial court convicted her and sentenced her to 30 years in prison. The appeals court refused to postpone the hearing or to summon witnesses who the woman believed had exculpatory evidence. It justified its decision on the grounds that the woman had not provided the court with the names of these witnesses in her written appeal. The 64-year-old woman explained that she was unable to read or write and had asked another person to draft the letter challenging the trial decision. She stated that she had been unable to confirm what was written in the appeal, but the judges were not persuaded and emphasized that it was her responsibility to ensure the witnesses’ presence.
Gacaca courts also occasionally denied the accused the right to confront witnesses against him or her. Human Rights Watch observed cases in which the accused was physically present at trial but was not allowed to follow his or her own trial in any detail. For example, in the south of the country, two different gacaca courts made the accused move away from the proceedings so that they were unable to hear or see what was happening in their own trials. A similar case occurred in Kigali in October 2008, when a court told five co-accused to sit apart and well away from trial proceedings, until it was their turn to testify. It was not immediately clear why courts ordered segregation of the accused in these types of cases. It is possible that the judges, who did not have adequate legal training, confused the practice of keeping witnesses outside of earshot of trial proceedings (in order to prevent their testimony from being influenced by other witnesses) and applied it to accused persons. Denying an accused the right to follow witness testimony implicating him or her in an offense and the right to cross-examine those witnesses clearly violated the right of these individuals to defend themselves.
The right to testify in one’s defense and the right against self-incrimination
The ICCPR guarantees an accused the right “[n]ot to be compelled to testify against himself or to confess guilt.”
The 2004 Gacaca Law fails to guarantee this right, as its preamble states that all Rwandans have a legal duty to testify. Article 29 goes on to say that “[a]ny person who omits or refuses to testify on what he/she has seen or on what he/she knows, as well as the one who makes a slanderous denunciation, shall be prosecuted by the Gacaca Court which makes the statement of it.” Prison sentences range from three to six months, with longer sentences for repeat offenders. While Article 29 does not specifically refer to the obligations of the accused, they have normally been expected to testify in their defense and have not been offered the right to remain silent at trial. Requiring the accused to testify effectively inverted the presumption of innocence by making the accused prove that he or she did not commit the alleged crimes. Without lawyers present in gacaca, the burden on the accused to defend themselves has been even greater.
Human Rights Watch is aware of only one case where an accused person refused to testify: that of human rights activist François-Xavier Byuma, discussed above. After requesting the disqualification of the presiding judge at trial on the grounds that he had a conflict of interest, Byuma refused to testify. The presiding judge responded by threatening that “the bench can judge people who refuse to testify,” effectively forcing Byuma to concede and defend himself.
Avocats Sans Frontières (ASF), which operated a gacaca-monitoring program nationwide from 2005 until 2010, repeatedly expressed concern that accused persons were required to take an oath to tell the truth before speaking in gacaca , in violation of their right against self-incrimination. The organization documented cases in which the accused were convicted not only of genocide-related crimes but also of perjury or of having failed to confess his or her crimes. In late 2006, the SNJG issued an instruction to judges telling them that an accused cannot be prosecuted for false testimony given during his or her own trial.However, this instruction did not direct courts to warn accused persons that, by testifying about what happened, they may incriminate themselves and that any statements made could form the basis of a conviction.
The government ran a number of programs that provided incentives for accused persons, particularly detainees, to confess to the charges they faced. The gacaca law offered significantly reduced punishments to individuals who confessed, including shorter prison sentences, the possibility of serving portions of a sentence through community service (known as “travaux d’intérêt général” or “TIG”), and suspension of portions of a sentence. While a significant number of detainees confessed to crimes initially, only a third had done so by 2002.However, the proportion increased in the following years, with over half the prison population having confessed by the end of 2004.
In many parts of the country, prison authorities, with the encouragement of government and judicial officials, organized committees to hear detainees’ confessions, even before gacaca began. They also invited evangelical Christians to proselytize in prisons and to try to persuade prisoners to confess. In addition to the prospect of reduced sentences, those who confessed could benefit from better prison conditions and the promise of an early release.In order to be accepted, a confession had to include the names of victims, accomplices, and a detailed description of the crimes committed. Failure to implicate other individuals by name could be a basis for rejecting the confession. The various advantages offered to prisoners who confessed led to a rash of partial and even false confessions.Some prisoners were prepared to confess to crimes they had not committed, to minor offenses where other crimes were committed, and to denounce others wrongfully.
Encouraging confessions was an obvious way to reduce the backlog of genocide-related cases, but the circumstances in which many prisoners confessed meant that the information they provided was often unreliable. There were also numerous contradictions between confessions. The questionable quality of confessions undermined confidence in some gacaca trials. In addition, officials who exerted pressure on people to confess failed to provide enough information to ensure they understood the rights they would forfeit through confession. If a gacaca court later found that a person had made a partial confession or had confessed to crimes which he or she had not committed, it could hand down harsh prison sentences and could exclude the person from participating in the community service program. Indeed, a number of prisoners were returned to prison on the grounds that their confessions were incomplete.
Protection from double jeopardy
Most legal systems accept the general principle that an accused cannot be tried twice for the same offense (known as "double jeopardy"), unless new evidence comes to light or there is evidence that the first trial involved a miscarriage of justice. Protection from double jeopardy provides accused persons with a guarantee that once judged, the case is over, and helps nurture public confidence in the judicial system. Both the Rwandan penal code and the ICCPR prohibit double jeopardy.
Genocide-related charges can be multi-faceted, potentially involving a number of distinct criminal acts that may have been committed at different times and in different locations. This can make it difficult to clearly distinguish between cases involving a violation of the principle of double jeopardy and cases in which an individual is charged in separate cases with unrelated offenses. However, both the minister of justice and the SNJG executive secretary have acknowledged that dozens of accused persons repeatedly brought to trial in gacaca proceedings have suffered a violation of their right to be protected from double jeopardy.
The double jeopardy legal loophole
In theory, appeals against conventional courts’ judgments should be heard by conventional appeals courts. Similarly, decisions rendered in gacaca jurisdictions should be decided by gacaca appeals courts.
However, the 2004 Gacaca Law provided an exception to this rule and gave gacaca courts the power to prosecute persons for crimes for which they had already been tried in first and second instance conventional courts, regardless of whether they had been convicted or acquitted. Without explanation, the law simply states that any discrepancy in judgments between two courts in the same case should be resolved by the gacaca appeals court. This provision led one Supreme Court judge to conclude that gacaca courts had become the new Supreme Court.
Judges and others became aware of the risks of double jeopardy as early as 2005 when gacaca courts began to investigate and prosecute persons already judged by conventional courts. Supreme Court judges asked the minister of justice to remedy the problem in 2006, either through legislative reform or by some other means. Human Rights Watch and other international organizations following the gacaca process raised similar concerns with the SNJG, providing it with detailed examples of cases where violations had occurred.
In May 2008—several years after judges and nongovernmental organizations first raised the issue with the SNJG—Parliament amended the law to close the legal loophole. Under the new law, cases tried by gacaca courts may only be appealed to gacaca appeals courts in the same jurisdiction and cases tried by conventional or military courts may only be reviewed by their respective appellate courts.
However, the law was poorly drafted and a loophole remains. Cases judged by first instance conventional courts which have not been appealed to the highest level can be brought again in gacaca, even after the deadline for appeal has expired in the conventional courts. In addition, the law does not provide a remedy for cases where double jeopardy violations have already occurred. It also leaves open the thorny question of whether a case can be revived if new evidence comes to light, or new allegations have been made relating to events which have been the subject of a previous case.
Cases involving a violation of double jeopardy
In some instances where a case previously judged in the conventional courts reappeared in gacaca, the gacaca court did not know how to handle the issue and asked the SNJG for legal advice, usually by briefly adjourning the hearing. According to the SNJG’s executive secretary, one gacaca judge excused himself in the middle of a hearing and pretended to use the bathroom in order to call her for advice on whether to proceed with the case. The SNJG executive secretary advised the judge to dismiss the case, which he did. In other cases, the court declared it was not competent to hear the case. The SNJG occasionally reminded gacaca judges that they should not decide cases which had already been ruled on by conventional courts, and that they should examine whether the allegations in the two cases were identical to determine whether to adjudicate the case.
In a number of other cases, however, gacaca courts rejected the argument that a case should be dismissed because the accused had already been prosecuted for the same offense. In one case, a soldier arrested a man in 1997 on the basis of a single witness statement. The man spent seven years in prison before a conventional court acquitted him on the grounds that he had been mistaken for another person of the same name. Two years later, in August 2006, a gacaca court summoned him to appear as a witness in another case. Upon arrival, the court accused him of the same offense of which he had previously been acquitted. The court convicted him and imposed a 30-year prison sentence. Four months later, an appeals court overturned the decision. He was released two weeks later.
Gacaca judges in Huye district told Human Rights Watch that two accused men faced charges in gacaca that were identical to cases previously heard in the conventional courts. In the first case, the gacaca court convicted and sentenced a man to 19 years' imprisonment. In the second case, the gacaca court convicted a former parliamentarian in December 2009 and sentenced him to “life imprisonment with special provisions,” even though a conventional court had already acquitted him of the same charges.
Human Rights Watch also documented cases where the accused were tried twice by the same or a neighboring gacaca jurisdiction on identical charges. Usually the second case was brought after the person making the accusations was dissatisfied with the original verdict.
In one case in the south of the country, a man was accused of involvement in the death of an elderly Tutsi woman in three different jurisdictions. In the first jurisdiction, the court acquitted the man. No appeal was filed. The victim’s son (the civil party) had already accused another person of killing his mother and that person had been convicted in gacaca.The following summer, in July 2008, the case re-emerged in a neighboring jurisdiction at the appellate level. The court declared that it lacked the competence to hear the case since the matter had already been decided by another court. A month later, the case reappeared before a third jurisdiction. This time, the court convicted the man and sentenced him to 19 years in prison.
Odette Uwimana had a similar experience in December 2009. A gacaca court acquitted her of involvement in the death of a Tutsi woman. The decision was affirmed on appeal. Dissatisfied with the acquittal, the victim’s relative then brought the same charges against Odette Uwimana's husband, Vincent Uzarama, in the same jurisdiction. Uzarama was acquitted but his wife—who was neither a witness nor an accused in the case—was again convicted and sentenced to 15 years in prison, a decision affirmed on appeal. The SNJG later appointed a new bench of judges to hear the case. The new court acquitted Uwimana and ordered her release in March 2010, by which time she had spent nearly six months in detention.
Double jeopardy violations have also occurred in more subtle ways in an attempt to circumvent the bar against repeat prosecutions. In some cases, accused persons found themselves faced with slightly modified charges or new witnesses who had not testified in the original case. In one case, a man acquitted in gacaca in October 2007 was re-arrested 10 months later on the same allegations. The man challenged the fact that the civil party in the new case had not come forward in the original case, but the gacaca court proceeded to convict and sentence him to 19 years in prison. The case was one of 18 cases in the jurisdiction where accused persons were retried at the direction of the district coordinator.
In a 2008 case, a man found himself accused twice, first in a conventional court which acquitted him, then before gacaca, for allegedly being involved in killings at Gahini hospital in eastern Rwanda. When the man challenged the case on the basis of double jeopardy, the gacaca court claimed he was being tried not for killing Tutsi at the hospital (the allegations in the first trial) but rather for mutilating the corpses. The gacaca court sentenced him to 30 years’ imprisonment, a decision affirmed on appeal.Given that these acts had not been raised in the first trial, the case appeared to be an attempt to circumvent the bar against double jeopardy. The man’s request for revision of his case was denied, and he remains in prison.
In another case, a man found himself tried twice in the same gacaca jurisdiction. The initial case involved category 2 charges relating to his alleged presence at a roadblock where people were killed, although not in his presence. The appeals court reversed the decision and ordered his release. However, in September 2008, a new court prosecuted him on category 1 charges based on allegations that he had told the local mayor to erect roadblocks in an area where killings later took place. The man challenged the validity of the trial on the grounds that no new information had been introduced that might justify a second trial. He also noted that no one had accused him of wrongdoing and that the only evidence in the case consisted of a statement he had made during the information gathering phase about a conversation he had with the local mayor in 1994 discussing the need to protect the community. Despite more than a dozen witnesses testifying in his defense, and none against him, the court convicted and sentenced him to “life imprisonment with special provisions.” A local official admitted to Human Rights Watch that the case was ill-founded and that the man’s earlier statement was being used to bring a new case with slightly amended charges. The official said that the second decision led to an outcry among the local population. At the revision stage, a court affirmed the conviction but reduced the sentence to seven years’ imprisonment. In addition to being an example of double jeopardy, this case illustrates how local community participation does not always protect accused persons against unfair trials.
Similar double jeopardy scenarios occurred when gacaca jurisdictions began a new round of information gathering across the country in 2009. For example, in October 2007 a gacaca court in the Western province acquitted a man who had been detained for 13 years without trial. Almost two years later, a detainee raised nearly identical accusations against the man. The court failed to consider whether the new allegations were identical to those raised in the first case and convicted the man, imposing an eight-year prison sentence.
The story of Théodore Munyangabe
Théodore Munyangabe served as a high-ranking local official (sous-préfet) of Cyangugu préfecture before, during, and after the genocide. He was one of the few government officials who remained in service after the genocide, having received much praise for his actions to protect and assist Tutsi.
In March 1995, he was arrested by the police on accusations of involvement in the genocide. A conventional court tried and convicted him, sentencing him to death which was the maximum penalty at the time. An appeals court reversed the conviction in July 1999 and ordered his release. Police placed Munyangabe under house arrest within days of his release, however, and formally rearrested him a month later on new, unspecified accusations of genocide.
Munyangabe remained in prison for nine years with no further hearing or trial until he was finally brought before gacaca in November 2008. The trial court concluded that the case was identical to the one decided in the conventional courts and dismissed the charges. Instead of being released, however, Munyangabe was brought before a neighboring gacaca jurisdiction and charged with the same crimes.
Munyangabe asserted that the case should be dismissed on double jeopardy grounds, but the court disagreed and proceeded to try him. One of the most flagrant irregularities at trial was the presiding judge’s coercion of a man into making a written statement against the accused. After a local RPF representative testified that Munyangabe’s former driver told him that the accused had attended a secret genocide planning meeting, the court summoned the driver to testify. The driver denied having ever made the statement, adding that the RPF representative and a second man had unsuccessfully tried to pressure him into accusing Munyangabe. The presiding judge immediately scolded the witness and threatened to arrest him for perjury. When the witness continued to insist that he had never made the statement, the judge adjourned the proceedings and told the witness to go home and think about the consequences of giving false testimony. The following day, the witness reappeared and reluctantly gave the court a written statement implicating the accused in the meeting in question. The court convicted Munyangabe largely on the basis of this statement and sentenced him to “life imprisonment with special provisions.”
On appeal, Munyangabe argued that he had been unlawfully convicted of the same crimes for which a conventional court had acquitted him. The court rejected this argument and found that the alleged genocide planning meeting constituted a new accusation. Munyangabe pointed out that the issue of such meetings had been raised in the conventional court case but had been dismissed. He also noted that the new allegation had not been raised during the national information gathering phase of gacaca and that the new witness had contradicted himself several times. He asked the court to summon the driver, but the court refused and upheld Munyangabe’s conviction.
During the course of the appeal, two genocide survivors were intimidated and arrested for trying to defend Munyangabe. The two men showed up to testify on the first day, but the hearing was postponed. Shortly after the local community dispersed that day, police arrested the two men without explanation and kept them in detention overnight. The local prosecutor secured their release the following day after other survivors in the community complained about the arrests.Despite the harassment, both men still decided to testify. The first man faced hostility from the judges but had no other problems during his testimony.
The second man testified that the civil party had bribed other survivors to accuse Munyangabe and that Munyangabe had not committed any wrongdoing during the genocide. After the man left the hearing, the civil party (a woman) accused him of having tried to intimidate her. The court ordered the man to be arrested and brought back to gacaca for questioning. Meanwhile, the judge scolded other survivors present at the hearing and reminded them of the “need to speak with one voice and not fight with each other.” He threatened to send anyone else criticizing the civil party to jail for perjury. Soldiers went to the second witness’s house and brought him back to gacaca. By the time they returned, however, the day’s proceedings had ended and the man was detained. After strong protests from other local survivors at the scene, the presiding judge ordered his release.
The gacaca jurisdiction denied Munyangabe’s request for revision in November 2009.Munyangabe then wrote to the SNJG to complain but had not received a response at the time of writing.
Munyangabe’s case is also discussed in connection with the right to have adequate time to prepare a defense and the risks faced by defense witnesses.
The right to be present at one’s own trial
Rwanda allows trials in absentia, that is trials without the accused present. The justification for such trials is that individuals should not be able to evade justice by not showing up for their trial. While this practice is not permitted in common law systems, it is standard procedure in civil law countries and has generally been accepted as lawful under international law provided certain procedures are followed. The UN Human Rights Committee has emphasized two procedural requirements: first, the accused should be given proper notice of the trial; and second, the court should strictly protect all of the due process rights of an accused.
Over the past few years, gacaca courts have prosecuted hundreds, and perhaps thousands, of individuals in their absence. This was not necessarily in breach of the law, but given that gacaca courts often fail to protect other basic rights set forth above, trials in absentia are particularly problematic.
Politically motivated in absentia trials
In a number of apparent politically-motivated cases, individuals suddenly learned that gacaca courts had convicted them in absentia. Some cases involved allegations that arose quite recently, and which were not raised during the national information gathering phase (2002-2004). Others appear to have resulted from Rwandan judicial officials seeking to gain custody over a suspect living abroad. For example, in 2006 the ICTR acquitted Emmanuel Bagambiki, the former Cyangugu préfet, of genocide. Rwandan judicial officials called the acquittal “unforgiveable” and “ridiculous.”Soon after, Rwandan prosecutors brought rape charges against Bagambiki, which ICTR prosecutors had considered but rejected due to lack of evidence (Rwandan judicial authorities did not protest against this decision at the time). A gacaca court convicted Bagambiki of rape in absentia in October 2007 and has since sought hisextradition from Belgium, where he is living in exile.
Rwandan lawyer and former ICTR defense investigator Léonidas Nshogoza found himself accused in gacaca just as the ICTR contemplated bringing a case against him for bribing a prosecution witness to change his testimony before the tribunal. When the allegations first emerged in June 2007, Rwandan police detained Nshogoza and charged him with corruption and genocide denial. He remained in prison without trial for more than five months before being released in November 2007. Meanwhile, in December 2007, the district coordinator in his native area of Mahembe sent local gacaca judges instructions to open a file against him. They obliged, rapidly assembling a file, charging him with involvement in the deaths of four of his sister’s children, and setting a trial date for January 22, 2008 which was postponed by a week after Nshogoza did not appear.
Three gacaca judges told Human Rights Watch separately that they were surprised to hear of Nshogoza’s case because all gacaca trials in the sector had officially been completed. Two of the judges reported that the district coordinator sent a letter on January 29 directing them to immediately decide on the case and to issue a default judgment on February 7 if the accused did not appear. One of these judges and a local government official told Human Rights Watch that they believed the case to be unfounded, a position echoed by several community members who lived in the area in 1994. The trial court acquitted Nshogoza in absentia, a decision affirmed on appeal. On February 8, 2008 Nshogoza surrendered himself to the ICTR after the court issued a warrant for his arrest. The ICTR convicted Nshogoza for contempt of court for meeting with a prosecution witness in violation of the tribunal’s protection orders, and for disclosing protected information about the witness to a third party, but acquitted him of the corruption charge. The ICTR sentenced him to 10 months’ imprisonment but released him as he had already spent more than a year in detention.
The case of Léopold Munyakazi is a striking example of an apparently politically-motivated genocide case. In September 2008, Munyakazi, then living in the United States, faced allegations in gacaca proceedings at exactly the same time the Rwandan government sought an international warrant for his arrest for the second time. Soon after the genocide, Munyakazi spent time in prison on genocide-related charges, but in 1999 the national prosecutor’s office ordered his release for lack of evidence. He went on to work for a national university in Rwanda, a position for which he required and obtained certificates of good standing from several government authorities. In July 2004, while he was attending a teaching conference in the United States, friends contacted him to warn him about worrying rumors circulating about him in Rwanda. Munyakazi decided not to return to Rwanda. Soon after, a parliamentary commission accused him—and hundreds of other persons and organizations—of “divisionism.”He sought asylum in the United States and began teaching at a university in Maryland. In October 2006, Munyakazi gave a faculty speech calling into question official Rwandan discourse on the genocide, which attracted the Rwandan authorities’ attention. One month later, the Rwandan national prosecutor’s office issued an international arrest warrant for Munyakazi on charges of genocide and genocide denial.
In the fall of 2008, the Rwandan prosecutor’s office renewed efforts to secure Munyakazi’s extradition and teamed up with the American television channel NBC (and its “To Catch a Predator” series) to confront Munyakazi in the United States. Around the same time, in September 2008, the Rwandan prosecutor’s office issued a second warrant for his arrest, and relatives in Rwanda received a summons for him to appear before gacaca. The gacaca trial began a year later in late 2009 in Munyakazi’s absence, but the case appears to have been dismissed later at the direction of the SNJG. Human Rights Watch could not confirm the SNJG’s involvement and was unable to find out the justification for the dismissal.
Other in absentia trials
The story of Jean-Népomuscène Munyangabe
In November 2007, Jean-Népomuscène Munyangabe learned that he had been convicted of genocide and sentenced to 18 years’ imprisonment in a gacaca court in southern Rwanda. At the time, he was working for the UN World Food Program in Chad (where he had been since 2005). Neither he nor his family, who still resided in Rwanda, had been notified of the case against him, even though his employment at the time was well known in the area.
Confident of his innocence, Munyangabe took leave from his post in Chad and returned to Rwanda to challenge the conviction.He filed his written appeal in January 2008. He was arrested and held in police custody for nearly three months until his case was heard.A final verdict was handed down in July 2008 after a series of flawed gacaca proceedings, discussed below.Munyangabe was convicted of involvement in the death of two Tutsi who had sought refuge at his family’s house and was sentenced to 19 years’ imprisonment.
The trial revealed that rather than being responsible for killings during the genocide, Munyangabe and his family had actively sought to protect Tutsi from the killers in 1994 by hiding them at their house. The two victims came to their house seeking refuge but changed their mind when they found other Tutsi already hiding there and feared being discovered by local militia. The victims convinced Munyangabe’s father to help them flee towards Burundi, but they were intercepted along the way and both were killed. Munyangabe’s father was killed shortly after the genocide upon returning to Rwanda.
Seven witnesses, including four neighbors, and three genocide survivors who had taken refuge at his family’s house testified in Munyangabe’s defense. One relative of the victim claimed she saw Munyangabe traveling with his father and the victims that evening before the killings took place, but her testimony was contradicted by several other witnesses. Others accusing Munyangabe said merely that he was friends with a neighboring family that committed crimes during the genocide and that he must have committed genocide too. One person speculated that Munyangabe had left his studies in Butare and returned to his family’s house in 1994 to commit crimes. All were family members of the victim, and none were eyewitnesses to the events in question.
The gacaca court acquitted Munyangabe and ordered his immediate release on April 4, 2008. The district coordinator delivered the release order to the prison where Munyangabe was detained, but later returned to retrieve it and prevented his release. The civil party appealed the acquittal, and Munyangabe was kept in prison, allegedly so that he would not flee the country.
The appeals trial, which began the following month, opened with the presiding judge refusing to disqualify himself despite allegations that he was a close friend of the civil party in the case. Three additional hearings took place, during which the presiding judge acted in a biased manner, took decisions without consulting other judges, reacted angrily to statements made by the accused, interrupted and detained at least three defense witnesses on allegations of perjury, and tried to manipulate the written record of proceedings. On June 17, 2008, the trial was suspended indefinitely.
The SNJG stepped in to change the jurisdiction hearing the case, but the newly appointed bench sat for just a single session and convicted Munyangabe despite the absence of new evidence. It sentenced him to 19 years’ imprisonment. In May 2010 the SNJG denied Munyangabe’s request for revision, leaving him with no other recourse.At the time of writing, he remains in prison.
Munyangabe’s case is also discussed in connection with the right not to be arbitrarily detained and the right to impartial justice.
In some cases, private grievances helped explain the decision to hold trials in the absence of the accused. The above case of Jean-Népomuscène Munyangabe is one example, with the civil party—an influential family in the south with several prominent family members living and working in Kigali—using the accused’s presence abroad to secure a judgment without him or his family even knowing about the case.
Not all gacaca trials in absentia were ill-founded or brought for political or personal reasons. In some cases, the accused fled the country or went into hiding, apparently to evade justice. In other cases, individuals chose not to appear because they thought they would not receive a fair trial or feared they might be tried twice or charged with additional accusations, particularly during the new information gathering phase in 2009. However, the authorities have also publicly accused large numbers of Rwandans, who may have left the country for legitimate reasons, of evading justice.For example in the late summer and early fall of 2009, hundreds of Rwandans crossed the border to Burundi. Human Rights Watch’s interviews with a number of those who fled to the Kirundo and Ngozi provinces of Burundi in October 2009 suggested that many may have had a credible fear of persecution—not prosecution—in Rwanda. Similarly, Rwandan refugees interviewed by other NGOs in early 2010 in Uganda said that they had left in part because they feared that the Rwandan authorities and private individuals were manipulating gacaca courts for their own purposes. Human Rights Watch is not in a position to ascertain whether some of these individuals may have participated in the genocide or may have gacaca cases pending; however, it cannot be assumed, as the government has done, that most or all were seeking to evade justice simply because they left Rwanda.
The right not to be arbitrarily detained
Rwandan and international law guarantees the right not to be arbitrarily arrested or detained. Rwanda has made important strides on this front over the past 17 years, but substantial concerns remain. In late November 2010, the NHRC Executive Secretary presented the Commission’s annual human rights report to Parliament and noted continuing problems of arbitrary arrest and detention and prolonged pre-trial detention.
Under the ICCPR, victims of unlawful arrest or detention have a right to compensation. However, neither the gacaca laws nor the Criminal Procedure Code provide for such a right. In December 2003, the National Unity and Reconciliation Commission recommended the creation of a compensation fund for individuals who were wrongfully imprisoned in the immediate aftermath of the genocide and for the heirs of innocent persons who had died in prison. The NHRC has made similar recommendations, including in its 2010 presentation to parliament. The government has never taken steps toward awarding compensation for cases of wrongful detention and does not appear to be contemplating such measures following the NHRC’s recent recommendations.
In the years immediately following the genocide, tens of thousands of individuals were arrested on the basis of a single, unverified accusation of participation in the genocide and detained for prolonged periods (in many cases years) without any form of due process. By 1998, the prison population reached around 130,000, with detainees held in life-threatening conditions. The enormous cost and logistics needed to support such a huge prison population were among the factors which led to the government’s decision to launch gacaca for genocide cases. By 2008, following several thousand releases, prison overcrowding had eased.
Today, the prison population has stabilized at just over 60,000, which is still well above the full capacity of the country’s prisons.Nearly two-thirds of the prison population has been convicted of genocide-related charges. Prison conditions remain harsh. In February 2011, approximately 130 persons remained in pre-trial detention on genocide-related charges, some having already spent many years in prison.
Over recent years, the authorities have introduced a number of positive steps to reduce the risk of arbitrary detention and to ensure that prisoners who have served their full sentence are released. In 2004, changes to the Criminal Procedure Code gave judges habeas corpus powers to compel police and prosecutors to bring to court detainees who may be illegally held. The changes also authorized judges to punish state agents responsible for arbitrary detention, but the detailed sanctions available to judges have yet to be set out in a revised penal code. The 2004 changes also specified that detainees must initially be held at police brigades, making it easier for relatives and others to locate them.
Arbitrary arrest and detention remain a problem in Rwanda. Human Rights Watch came across a number of cases in which police arrested persons without a legal basis and detained them for several days. In some cases, police detained accused persons before or after gacaca trials without a court order. For example, in a 2006 case in the north of the country, police arrested a presidential guard officer as he attended the funeral of his grandfather. They held him in the cell coordinator’s office for two days until his trial, guarded by members of the local defense forces (LDF), the government-sponsored paramilitary forces that patrol local communities.Police arrested another man in June 2008 and detained him for three days until his trial, at which time he was convicted. In the western part of the country, police arrested another man in August 2008 and held him for five days pending his trial.
Human Rights Watch also documented cases where gacaca courts ordered the detention of accused persons or witnesses without establishing that the person intended to flee or might cause harm to others or him- or herself if the person remained at liberty. In April 2010, a gacaca court in the southern part of the country ordered a man attending an appeal hearing against his acquittal to be detained pending completion of proceedings two days later (when he was convicted).
Human Rights Watch also documented a handful of cases in which individuals were kept in detention despite being acquitted. The case of Jean-Népomuscène Munyangabe, discussed above, illustrates the problem.
Similarly, in February 2008, a gacaca court acquitted Justin Nsengimana of charges of distributing arms, carrying an illegal weapon, and rape during the genocide. He remained in Butare prison for two years on the grounds that he had a category 1 case pending against him in the conventional courts. However, Nsengimana was never brought before a conventional court. Instead, in February 2010, he faced new accusations in the same gacaca court which had tried and acquitted him on the earlier charges. The court convicted him of the same two arms charges and of having participated in killings in Butare and sentenced him to “life imprisonment with special provisions.”
In another case, Viateur Munyandekwe was acquitted in gacaca three times: first on August 26, 2007, on accusations of having failed to assist a neighbor in danger and of participating in killing a man; then on January 18, 2009 of accusations of rape; and finally on March 17, 2010, on the same accusations of having killed a man.However, he remained in detention throughout all three trials, and was finally released in 2010 after more than 31 months in detention.
Differences in Judicial Standards between Conventional Courts and Gacaca
Other differences between Rwanda’s conventional justice system and gacaca courts relate to judges’ qualifications and the legal standards applied in cases. In opting for the gacaca system, the Rwandan authorities made compromises which they believed were necessary to accelerate the resolution of cases and to avoid cases getting bogged down by legal formalities. However, the absence of adequate safeguards led to serious irregularities.
Judges: qualifications, training, remuneration and removal
It would have been impossible to staff the more than 12,000 gacaca courts with legally trained professionals. The war and genocide had devastated the judicial system, with only 237 judges able to resume work in August 1994 out of more than 600 judges in service before the genocide. While the number of judges had more than tripled by 1996, it was still insufficient to deal with the huge caseload of genocide-related cases. Consequently, as part of the decision to move these cases to gacaca, gacaca judges (inyangamugayo) were to be elected by local communities and would be trained to handle complex cases and to uniformly apply legal standards.
Qualifications of gacaca judges
In October 2001, the population elected approximately 259,000 laymen and women to serve as gacaca judges in genocide cases. The first gacaca law, adopted in 2001, established the criteria for the candidates: judges must be at least 21 years old, persons of “integrity” within their community, and ordinary citizens. “Persons of integrity” were defined as individuals with high moral character who had not participated in the genocide, who did not hold sectarian or “divisionist” beliefs, and who had not been sentenced to more than six months’ imprisonment.
At the time of their election in October 2001, a significant number of gacaca judges had not finished primary school, although those at district and province levels tended to have a higher level of education. Similarly, the majority of judges at the cell and sector levels were farmers, whereas a large number of judges at the higher levels were teachers or civil servants. Women were well-represented amongst gacaca judges but remained in the minority.
In April and May 2002, gacaca judges attended six full-day compulsory training sessions around the country (spread over three weeks), which were led primarily by magistrates and law students. Each group of 70 to 90 judges received instruction on the basic principles of the gacaca law, management skills, ethics, and trauma. Given the low education and literacy levels of many judges, and the complexities and ambiguities of the gacaca law, it is difficult to see how such training could have been sufficient to prepare the judges to decide genocide-related cases.
The Supreme Court issued a gacaca manual for judges, which focused primarily on procedural matters, with little explanation of the material elements that need to be proven in order to convict a person, of what weight should be given to different forms of evidence, and of the applicable standards of proof. The U.S. Justice Department’s resident legal advisor in Rwanda expressed concerns at the time that “judges and prosecutors who were providing legal training to individuals responsible for the actual training of the Gacaca judges were teaching vastly different instructions on categorization.” The effects of divergent instructions soon became clear when various gacaca courts adopted strikingly different approaches to the categorization of offenses and decided cases involving similar facts very differently.
In 2005, the SNJG also circulated simplified instruction booklets to assist judges with procedural matters but gave no additional guidance on who carries the burden of proof, how to evaluate evidence, or what level of proof is needed to convict a person.In 2006 and 2007, other short training sessions, usually consisting of several days, were also provided to judges.
In 2008, after the decision to transfer most category 1 cases (the majority of which involved rape) to gacaca, the SNJG launched a new training program to sensitize judges to the issues involved in sexual violence cases. The program targeted only judges selected to hear category 1 cases and consisted of two parts. First, SNJG officials traveled around the country and instructed judges on how to handle procedural aspects. Second, the Institute for Legal Practice and Development, with Dutch funding, sent a team of Rwandan lawyers and counselors around the country to conduct role-playing exercises with judges to sensitize them to trauma and other relevant issues in rape cases.
Gacaca judges do not receive any monetary remuneration for their services, although in recent years the SNJG gave them some in-kind compensation as well as a small sum of money (discussed below). From the very beginning, commentators expressed concern that requiring judges to take one or two days per week away from their own work for several consecutive years without adequate monetary compensation could be an incentive for corruption.Consequently, the Rwandan government tried to develop alternative means of compensating gacaca judges and of ensuring their commitment to the judicial process, including providing them with national health insurance and holding official ceremonies at the local community level to recognize their service.
In 2005, the SNJG and the Belgian Technical Cooperation proposed other ways to improve recognition of judges’ role, including distribution of bicycles or goats, financial contributions to a local bank for microcredit loans, and a national day to celebrate the judges. The government rejected many of the proposals, but the SNJG distributed radios to all judges in 2007 and bicycles to each jurisdiction in 2008. It also gave a one-time payment of 4,300 Rwandan francs (approximately US$7) each to all judges.
Removal of judges
By law, gacaca judges may be replaced if they repeatedly fail to appear at hearings without good reason, are convicted and receive a sentence of six months' imprisonment or more, incite sectarianism, occupy political or government positions, or do anything that is incompatible with their role as persons of integrity. In 2005, the SNJG issued a special directive on the dismissal and replacement of judges.
Initially, many judges were removed for alleged participation in the genocide. The Belgian Technical Cooperation reported that by December 2003, more than 650 gacaca judges and 15 district coordinators had been removed due to such allegations. The SNJG later reported that 45,396 judges had to be removed (and replaced) because of their alleged involvement in the genocide. In later years, a number of judges were also removed for corruption, namely soliciting and accepting bribes from accused persons or other interested parties. The SNJG executive secretary reported in January 2008 that 56,000 ineffective or corrupt judges had been removed from service. In total, more than 92,000 judges (or 35 percent of the total number) have been removed since gacaca’s inception. The SNJG's resolve to remove allegedly corrupt or criminal judges was a positive move. However, the fact that such a move was necessary reinforced Rwandans’ lack of confidence in the courts and increased their concerns over judges’ impartiality.
Burden and standards of proof
Unlike the conventional justice system, gacaca courts have no procedures governing what evidence is admissible or inadmissible, who has the burden of proving that a person committed a crime, and what standard should be used to determine guilt. Certain general legal principles appeared to apply from Human Rights Watch’s trial observations, such as an understanding that eyewitness testimony is preferable to hearsay, that relatives’ testimony may be biased, and that corroboration by several witnesses makes an allegation more credible and reliable. Still, gacaca practices lacked uniformity or consistency. Hearsay testimony was routinely relied upon and given significant weight without taking steps to summon the person who made the original statement. Convictions were also often based on uncorroborated or inconsistent statements by witnesses, some of whom had no direct knowledge of the events in question.
Burden of proof
Because gacaca trials do not involve a prosecutor, at the beginning of a trial presiding judges announce the charges against the accused and provide a general overview of the allegations. The accused is then given the floor to provide information and set out his or her defense. Judges often ask follow-up questions. Then, witnesses to the events are called, with those testifying against the accused appearing first, followed by any defense witnesses. The civil party, normally the victim or relatives of the victim, usually makes a statement. Once the witnesses have been heard, the proceedings are opened to the general population for statements or questions to anyone who has already spoken.
Although the law requires that an accused be presumed innocent, in practice the burden has generally fallen on the accused to prove that he or she did not commit the alleged crime. The absence of a public prosecutor placed the burden of proof even more squarely on the accused. Many judges openly demonstrated hostility to the accused, made disparaging remarks or interrupted the testimony of the accused. The accused also had to bring his or her own witnesses to help defend him or herself against the allegations. If he or she was unable to find defense witnesses, the accused was usually convicted. Human Rights Watch documented a number of cases where courts convicted a person despite the fact that no witness testified against the person and only defense witnesses exculpating the accused appeared at trial.
Standard of proof
Often, gacaca jurisdictions applied divergent standards of proof. The gacaca laws gave no objective guidance on how much weight to give to witness testimony, the necessary level of corroboration to establish facts, and the amount of evidence needed to convict a person. As a result, judges were left to subjectively decide on these matters. The only requirement under the 2004 Gacaca Law was that “[j]udgments must be motivated” and must be signed or marked by all members of the gacaca court.Gacaca judgments differ from regular court judgments in that they are not formal written opinions. Rather, they are short handwritten summaries (known as “fiches de jugement”) which are included in the register of minutes for each jurisdiction and are signed by the judges and the accused.Many judgments were not justified by reasoning to explain what evidence was relied upon or discredited in arriving at a decision. In some cases, even the charges that were retained or dismissed against the accused were missing from the judgment. These deficiencies made the appeals process more difficult for accused persons, as well as for judges hearing the appeals.
In 2004 the SNJG took some steps to assist judges in deciding cases and to ensure some degree of consistency between jurisdictions. It launched an initiative through which gacaca judges confronting particularly complex issues could ask legal experts for help. The SNJG had a Kigali-based team of experts who fielded telephone calls from jurisdictions throughout the country and who occasionally visited judges to discuss issues.
Two areas that illustrate the extent of divergence in courts’ decisions are legal intent and witness credibility. The requirement of “intent,” under which the court must establish the state of mind of the accused and conclude whether or not he or she intended to commit the alleged crime, understandably proved to be one of the most difficult concepts for judges to grasp. In order to convict a person for genocide under Rwandan and international law, a court must find that the person intended “to destroy, in whole or in part, a national, ethnical, racial or religious group.” If the intent requirement is not proven, a court should acquit the accused of the specific crime of genocide. In practice, however, judges rarely considered the issue of intent and almost never included it in the reasoning of judgments. The result is that many people were convicted of genocide without any proof that they intended to destroy, in whole or in part, the Tutsi ethnic group. Given that most gacaca cases involved genocide, the SNJG should have instructed judges on the need to consider the intent of the accused and should have provided detailed guidance on the issue.
This issue proved particularly problematic when judges confronted the question of accomplice liability. Under the gacaca laws, an accomplice is someone who “by any means, assisted to commit offenses” and is punished to the same degree as the main perpetrator of the crime. However, the gacaca laws were silent on whether a person must intend to assist someone else in committing an offense before he or she can be called an accomplice, leaving the decision to the discretion of individual judges. Some courts held that individuals present at roadblocks where killings later occurred, or who were forced to participate in night patrols to ensure security in their area, were accomplices. Other courts required that persons be physically present when killings occurred or that they intended their actions to cause later deaths.
Given that many individuals were obliged to participate in neighborhood patrols in 1994, the SNJG could easily have foreseen that the question of whether or not accomplices needed to have intended certain consequences would arise in a significant number of cases. Yet the SNJG failed to provide guidance, even after it was made aware of judges’ divergent approaches by independent monitors such as ASF. More specifically, the SNJG should have told judges that a person’s mere presence at roadblocks or any other crime scene was not enough to convict him or her of being an accomplice and that they should require proof that the person’s actions clearly constituted assistance in, or encouragement of, the commission of a crime. It was not until March 2007 that the SNJG’s executive secretary finally stated publicly that a person’s presence at a roadblock was not in and of itself enough to convict that person of a crime.No further guidance was provided on accomplice liability.
Courts also regularly accepted hearsay instead of summoning the person who made the original statement or asking whether that person could appear as a witness. While hearsay evidence is allowed in many jurisdictions, including the conventional courts in Rwanda and many civil law jurisdictions in Europe, courts generally recognize that it is a secondary form of evidence which must be probed for its reliability. Gacaca courts did not appear to regularly make this distinction and instead often afforded significant weight to hearsay statements.
In the case of Jean-Népomuscène Munyangabe, discussed above, a community member told the court that another man had told him, during an earlier gacaca hearing, that the accused had participated in an attack in April 1994. The presiding judge asked the man why the person who made the statement had not testified to this fact at the earlier hearing but did not ask whether the man could be brought before the court to testify and did not consider postponing the hearing to summon the man to appear. In other cases, courts accepted written testimony—usually in the form of handwritten notes—as reliable evidence without any meaningful discussion of whether the person who wrote the note could have appeared before the court to testify and be questioned by the judges and the accused. Courts also failed to verify the authenticity of such handwritten notes.
In some cases, judges also struggled to assess the quality of testimony. At times, they failed to identify evident bias on the part of witnesses against one of the parties or failed to probe further when obvious inconsistencies arose within a witness statement or between different witnesses.
In the case of Pascal Habarugira, discussed above, the accused pointed out several inconsistencies in the testimony of two key witnesses at the appeals stage. He said he could provide the court with proof of the inconsistencies by presenting written judgments from other trials in which these witnesses had testified. The presiding judge scolded the accused, telling him he was trying to turn the case into a formal affair as if it were being heard in a conventional court. He said that the discrepancy was irrelevant to the value of the witnesses’ testimony and that the accused should “stop wasting time.”
In the case of Théodore Munyangabe, also discussed above, a key witness significantly contradicted himself between the trial and appeals stages. When the accused identified the inconsistencies on appeal, the presiding judge rejected his argument and read out the witness’s statement from the trial. The accused and a number of community members objected, saying that the testimony read by the judge (as reflected in the trial court’s record of the proceedings) was not the same testimony presented at the trial and that the statement must have been altered after the trial. The judges rejected the argument, and, relying on the witness’s earlier statement, upheld the conviction.
Sentencing and Reparations
The question of what constitutes an appropriate punishment for genocide and related offenses has been hotly debated both inside and outside Rwanda. Genocide is among the most heinous of crimes, and as such the punishment should reflect the gravity of the crime. The Rwandan Government occasionally expressed disappointment at what it viewed as “lenient” sentences handed down by the ICTR. Since Rwanda abolished the death penalty in 2007, the maximum penalty for genocide in Rwandan courts (whether conventional courts or gacaca) is “life imprisonment with special provisions.”
Gacaca courts follow sentencing guidelines which may be roughly summarized as follows:
- Category 1 offenders receive a mandatory sentence of “life imprisonment with special provisions;”
- Category 2 offenders receive sentences ranging from five years to life imprisonment, depending on the nature of the crime and whether the person intended to kill; and
- Category 3 offenders are ordered to pay civil reparations in the amount of damage caused.
Persons who confess receive a reduced sentence, with those confessing before they are accused receiving the lightest sentences.
By law, those convicted can also be stripped of certain civil rights, including the right to vote, the right to engage in public or military service, and the right to be a teacher or work in the medical profession. Children under 14 at the time of the crime cannot be prosecuted, while children between the ages of 14 and 18 receive reduced sentences.
Over the years, the Rwandan government has attempted to reduce the prison population by releasing certain categories of detainees, primarily the elderly, the chronically ill, minors, and those without files. On January 1, 2003, President Kagame announced the provisional release of prisoners who had confessed to their crimes (except those in category 1) and who had already served their sentences. Those released included genocide suspects and suspects of criminal offenses unrelated to genocide.
By March 2003, the government had released more than 24,000 detainees and sent them to “solidarity camps” (known as “ingando”) for two months of reeducation, before reintegrating them into their communities. In mid-2005, the government provisionally released another 20,000 detainees and, in early 2007, it released yet another 9,000 prisoners. The releases helped reduce the prison population but caused many genocide survivors to fear for their safety. A number of prisoners who benefited from these provisional releases were later re-arrested once gacaca trials began.
“Life imprisonment with special provisions”
The sentence of “life imprisonment with special provisions” replaced the death penalty in 2007 and has been the mandatory sentence for all category 1 offenders who do not confess or plead guilty to their crimes.
Rwandan law originally defined “special provisions” as imprisonment in “isolation” and provided that supplemental legislation would establish more specific modalities for its application. The United Nations and the ICTR expressed concern over whether the punishment amounted to prolonged solitary confinement and would therefore constitute inhumane treatment. International and Rwandan human rights groups similarly criticized the penalty and called for its abolition.
The mandatory application of “life imprisonment with special provisions” has been problematic in gacaca as the penalty has sometimes been imposed following a flawed process before non-professional judges, in circumstances where all the rights of the accused to due process were not respected. In some cases, this meant that the most stringent penalty was imposed following summary trials which may have lasted no more than an hour.
The Rwandan government maintained that “life imprisonment with special provisions” did not amount to solitary confinement, but accepted that a more precise definition was needed. It asserted that no person would be placed in isolation until new legislation outlining the penalty entered into effect. In any case, Rwanda did not have the facilities to put the measure into effect given the large number of persons sentenced to “life imprisonment with special provisions” and the very limited prison space available. In November 2008, in response to concerns expressed by the ICTR and countries contemplating extradition of genocide suspects to Rwanda, the government adopted legislation barring application of the sentence to cases transferred to Rwandan courts by the ICTR or foreign jurisdictions.
In September 2010, more than three years after the penalty’s introduction into Rwandan law, Parliament enacted legislation further defining the punishment.
The new law provides that:
1° a sentenced person is not entitled to any kind of mercy, conditional release or rehabilitation, unless he/she has served at least twenty (20) years of imprisonment;
2° a sentenced person is kept in an individual cell reserved to the guilty people of the inhuman crimes...
The cell must have sufficient dimensions and requirement equipment [material].
The law guarantees those sentenced to the punishment certain basic rights, such as the right to be visited by relatives, to physical exercise, to medical care, to leisure activities, and to worship. However, it does not specify the frequency with which these rights may be exercised, all of which are to be determined by internal prison rules and regulations. It also does not guarantee regular interaction with other prisoners, which is the only dependable form of contact for prisoners (especially those detained for long periods whose relatives, friends, and lawyers may stop visiting them over the years).
Human Rights Watch welcomes the government’s measures to bring the penalty of “life imprisonment with special provisions” in line with its domestic and international legal obligations and recognizes that, to date, prisoners have not been held in isolation. However, Human Rights Watch remains cautious about the potential application of this sentence until it is demonstrated in practice that prisoners are granted all these basic rights, including regular contact with other prisoners.
Alongside the establishment of gacaca, the government introduced an alternative to imprisonment in genocide and genocide-related cases: community service. While most countries reserve community service for low-level offenders, Rwanda introduced it only for genocide-related cases. The alternative sentence offered three main benefits. First, community service would alleviate overcrowding in prisons. Second, it could help reintegrate convicted persons into their local communities. Third, it would provide a means for indigent convicts to make reparations to society and to contribute to national development.
The community service program, known by its French acronym “TIG” (“travaux d’intérêt général”), became operational in 2005 and allowed category 2 offenders who confessed to their crimes (and whose confessions were accepted as complete and truthful) to serve the first portion of their sentence in prison and the second portion doing community service. The program was originally consensual: prisoners could decide whether to serve their full sentence in prison or commute half of it into community service. The government later removed the requirement of prisoner’s consent, and the program became mandatory for anyone that qualified. In 2008, Parliament amended the gacaca laws to require individuals sentenced to prison and community service to serve the community service portion of the sentence first, with the possibility of having the remainder of the sentence suspended if the person satisfactorily completed the TIG program.
Rwandan law provides for two types of community service: it can be performed in either a convict’s local community or a special TIG camp. In recent years, the government has prioritized the use of camps. Individuals who perform community service in their home communities live with their families and do community service three days a week. The work usually consists of construction and repair of roads, schools, and housing settlements for genocide survivors. Individuals often spend the remainder of the week tilling their own land or doing other remunerated work. In contrast, those who live in TIG camps work six days a week but complete their sentences in half the time: for example, a person sentenced to eight years’ community service can complete his or her sentence in only four years in a TIG camp. In both instances, projects involve intense manual labor for many hours each day and can be extremely physically demanding.
The Rwandan government has described the community service program as a huge success. Ordinary Rwandans’ perspectives have been mixed. Genocide survivors expressed two main concerns in interviews with Human Rights Watch. First, some categorized community service as a lighter sentence than imprisonment and as inconsistent with the gravity of the crime of genocide. They therefore considered the “tigistes” (or persons doing community service) to be getting off lightly.
Second, some survivors expressed fear at having to live alongside the convicted persons in their local communities and worried that the convicts might take revenge on them. Human Rights Watch is not aware of acts of retaliation committed while a convicted person participated in TIG. A few survivors reported that they were later relieved that there had been no significant tensions between them and the tigistes. Still, for many survivors the bitterness and fear of revenge remained ever present.
Several genocide convicts, on the other hand, told Human Rights Watch that they regarded the community service program as a form of forced labor and that they felt exploited by the government. Others complained about the conditions in TIG camps and, in particular, that they did not receive enough food to sustain them in carrying out the long hours of manual work.
In the course of one interview in a TIG camp, an interviewee revealed that he had completed his community service sentence but had not yet been released. As Human Rights Watch tried to follow up with the interviewee, a high-ranking government official who had overheard the remarks promptly escorted the Human Rights Watch researchers away from the camp.
By mid-2009, more than 90,000 persons had been sentenced to community service.Approximately 26,000 persons had completed TIG by the end of 2010, while more than 19,000 continued to serve their sentence. More than 27,000 had yet to start the program due to limited capacity. Whether the community service program will achieve its objectives remains to be seen. It has certainly succeeded in reducing the prison population and has contributed to the physical rebuilding of the country. The program’s success in reintegrating tigistes into their local communities is more debatable, particularly for those in TIG camps who live far from their home communities and have little opportunity to interact with the outside world.
Compensation to victims has been a contentious issue from the very start. Persons accused of category 3 offenses, defined as property crimes (resulting from looting and pillaging), have been made to pay reparations to their victims for the amount of the damage caused. However, the gacaca laws have never provided for direct indemnification of victims by category 1 and 2 convicts. The 1996 Genocide Law and 2001 Gacaca Law provide for gacaca courts to draw up lists of damages suffered by victims (including bodily injuries and related costs) and to pass them on to a government compensation fund which had yet to be established. The 2004 Gacaca Law states that “[o]ther forms of compensation the victims receive shall be determined by a particular law,” which gave many genocide survivors hope that they would receive monetary compensation. However, to date, no one has received any monetary or other compensation.
Most genocide survivors interviewed by Human Rights Watch cited the lack of compensation as one of the main shortcomings of the gacaca process. In a country where most of the population draws its livelihood from farming and has limited financial resources, many victims looked to reparations as a tangible punishment that would recognize their suffering and would help them in their daily lives. Realistically, it would have been very difficult to set a price on the damage suffered by victims of the genocide and other crimes committed in 1994. Moreover, most accused persons are poor and would have been unable to pay any compensation. Yet the decision not to make perpetrators indemnify victims and their families and not to provide government indemnification has disappointed many survivors.
In December 2008, the government announced an overhaul of the Fund for the Support and Assistance of Genocide Survivors (known by its French acronym “FARG”). The fund was established in 1998 with government financing but encountered difficulties over the years, including allegations of corruption, financial mismanagement, and poor construction of housing for genocide survivors. FARG’s premise was simple: to provide financial assistance to genocide survivors in the form of children’s school fees, medical assistance, building of houses, and support for income-generating activities. Under the new law, the government in January 2009 began the process of confirming that FARG beneficiaries were indeed genocide survivors who qualified for assistance. Within days, the authorities found more than a dozen instances of fraud, including ghost beneficiaries and mismanaged funds. A scandal ensued resulting in hundreds of individuals being removed from the beneficiary list, the dismissal of several senior FARG officials, and the arrest of more than 100 people. Even after the shake-up, FARG has continued to have some of the same difficulties.
The genocide survivor fund has been a mixed success. Its benefits—especially medical fees, school tuition, and housing—have provided valuable assistance to many genocide survivors. However, it has a narrow definition of who qualifies as a “survivor.” It excludes Tutsi women who were married to Hutu before the genocide and children of such marriages, as well as Hutu widows who lost their Tutsi husbands during the genocide. Hutu men and their wives or children who were injured or killed do not qualify as survivors, even if they were killed trying to protect Tutsi.
One Tutsi widow married to a Hutu man who died during the genocide lamented the fact that she could not receive medical care even though she was handicapped as a result of injuries suffered during the genocide. Another Tutsi widow with children remained homeless after being denied FARG assistance for the same reason. FARG officials refuted Human Rights Watch’s claim that it fails to assist those who qualify. However, Human Rights Watch and local NGOs working with these vulnerable groups documented a number of such cases.Allegations of corruption and mismanagement of funds in FARG have left many genocide survivors disillusioned by the government’s promises to assist them.
One Tutsi woman said:
Ibuka will not help me because my children are Hutu. They refuse to give me the certificate of a survivor because I was married to a Hutu. Now I’m sick with HIV as a result of being raped during the genocide, and I don’t have money to continue to get medicine. My children find the situation unfair. Their father was killed because of their mother and yet they aren’t seen as victims of the genocide.
Gacaca’s success has been predicated on local community participation. The aims of involving the community have been to uncover the truth about what happened during the genocide, to safeguard the fair trial rights of perpetrators and victims, and to contribute to the healing process of the community as a whole. At times, local participation has helped to maintain the integrity of the process and to achieve these goals. However, the waning interest of a significant part of the population and the silence of others (who have attended trials but did not speak publicly) have limited gacaca’s success, as the public could not always be relied upon to denounce false testimony or miscarriages of justice. Individuals had well-founded reason to fear that if they spoke out, they risked being prosecuted themselves or incurring problems with neighbors or the government. Manipulation of some trials, with private citizens using gacaca to try to settle scores or the government using it to to silence critics, and inappropriate influence exerted by other actors such as district coordinators, further contributed to a certain level of disillusionment.
When gacaca began, local communities around the country attended trials in huge numbers. Judges turned up early and appeared motivated to perform their duties. Rwandans were curious to see how the process would unfold.Perhaps expectedly, this level of enthusiastic involvement has declined sharply over the years.
When interviewed in 2002, members of different communities told Human Rights Watch that they had found the preliminary stages of the process—creating lists of families, victims, and perpetrators, and the gathering of initial information— bureaucratic, slow, and tedious. Some people complained that individual testimonies were too long or that debates among community members should have been postponed until the actual trial. In some urban areas, especially Kigali, residents who had not lived in the area during the genocide saw little need to participate in gacaca. In areas where few genocide survivors remained, there was even less interest.
Several factors deterred genocide survivors from taking a more meaningful and sustained interest in the gacaca process. First, as one former Ibuka president put it in 2003, “There are no incentives for survivors [in gacaca]: there has not been compensation or reparation…” Second, survivors risked being re-traumatized in gacaca, particularly if they showed emotion – a sign of weakness in Rwandan culture. Third, many genocide survivors feared retaliation in their local communities as a result of describing what had happened to them or challenging other persons’ testimonies.
Hutu often stayed away from gacaca, afraid of being publicly denounced or concerned they might not be given an opportunity to defend themselves. According to one judge, people were also reluctant to speak out in response to false testimonies for fear of being accused themselves. Hutu whose relatives were killed and property destroyed by RPF soldiers were unable to raise these cases, which left them frustrated and disappointed with the process.
Both genocide survivors and genocide perpetrators worried that speaking about what they knew in gacaca would lead to social ostracism or repercussions from relatives and neighbors or would create problems with local government officials. As a result, the practice of “ceceka” (meaning “to keep silent”) emerged, with local residents attending gacaca but deliberately choosing not to speak.Repercussions for speaking out included prosecution for perjury, genocide ideology, minimization of genocide, or even complicity in genocide. There may also have been an implicit pact among some Hutu not to denounce other Hutu. Regardless of the reasons, the fact that residents in many communities did not participate actively in gacaca undermined the reliability of proceedings and weakened the government’s argument that popular involvement was ensuring fair trials.
Many Rwandans could also not afford to sacrifice a day or more away from cultivating their fields or from other forms of paid employment. With the population already devoting one day of the week or month (depending on the area) to mandatory community work (known as “umuganda”), many people were reluctant to devote an additional day or two every week to gacaca.
As community participation declined, local officials and gacaca judges tried to persuade individuals to attend. When persuasion failed, they closed shops on the day of gacaca hearings and threatened to fine residents who failed to attend the sessions. One individual told Human Rights Watch that fines ranged from 1,000 to 2,000 Rwandan francs (up to US$3.30). In certain areas, the local defense forces also went house to house, rounding up community members and bringing them to gacaca.Gacaca judges occasionally used local defense forces to prevent people from leaving gacaca sessions early.
By 2004, the government was so concerned about attendance that it introduced a provision making participation in gacaca hearings compulsory when Parliament revised the gacaca laws that same year. Nevertheless, absenteeism increased over the years, particularly as trials dragged on and the deadline for gacaca’s closure was extended several times. In late 2007 or early 2008, judges and local officials lost control over attendance at the weekly sessions and stopped fining individuals for their failure to attend.
Risks for Witnesses
Why is it that any person who tells the truth and defends a man is seen as a traitor?
— A genocide survivor testifying as a defense witness in gacaca
The gacaca law makes it a legal duty for all Rwandans to state what they know. But individuals speaking out in gacaca proceedings, either as formal witnesses or as community members, have sometimes done so at great personal risk. One local official told Human Rights Watch that “witnesses are scared to be arrested under Article 29 [which prescribes penalties for those who perjury themselves, making slanderous statements, or refusing to testify]. Testifying for the defense risked having your statements qualified as lies.” A genocide survivor who had been raped during the genocide said that “even people who know things don’t speak because they don’t want to cause problems with their neighbors.”
Rwanda’s ill-defined laws on “divisionism” and “genocide ideology” also had a chilling effect on individuals’ willingness to speak out in gacaca. Many individuals interviewed by Human Rights Watch between 2005 and 2010 expressed fear of being accused of these crimes, or of “minimizing the genocide,” if they testified in gacaca proceedings.
The risk of reprisal was a particular barrier for individuals who lost relatives at the hands of the RPF. These individuals were unable to use gacaca to seek redress for these deaths because gacaca’s jurisdiction only covered genocide-related crimes committed against Tutsi (discussed below). People who spoke publicly about RPF crimes or challenged the official tenet of the genocide—that only Hutu were killers and only Tutsi were victims—sometimes found themselves swiftly facing charges of “genocide ideology” themselves as a result of their testimony.
At a gacaca trial in southern Rwanda in October 2006, defense witness Célestin Sindikuwabo stated that the accused person had fled to Burundi in 1994 because he and others had seen RPF soldiers killing people. The court acquitted the accused, but police arrested Sindikuwabo several days later in connection with his statement. In March 2007, a conventional court convicted and sentenced Sindikubwabo to 20 years in prison for “gross minimization of the genocide.”
Another man found himself accused of minimizing the genocide after he stated, at a 2006 weekly gacaca session during the information gathering phase, that a group of Tutsi seeking refuge at a church in 1994 pillaged sweet potatoes from neighboring Hutu farms and should also be forced to apologize in gacaca. The man, who happened to be a genocide survivor, spent nearly 10 months in detention before a conventional court acquitted him.
Threats and intimidation of witnesses, discussed below, also deterred potential witnesses from coming forward. In some cases, witnesses were even killed. According to the government, 120 individuals were killed between 2004 and the end of 2008 because of either their ethnicity or their participation in gacaca – a sharp rise from 42 people killed between 1995 and 2003. The rate of killings more than quintupled during the time of gacaca, with the highest number of deaths in 2006 when trials began nationwide. The government reported that most of the individuals killed were genocide survivors but that a number of Hutu who appeared as witnesses in gacaca were also killed.The highest concentration of killings occurred in the southern part of the country, particularly in the Karongi district around the town of Kibuye. Human Rights Watch was not able to ascertain the number of deaths in 2009, but the Rwandan government appears to have reported six deaths to the US embassy, while Ibuka reported 24. As of November 2010, the Victim and Witness Support Unit (“VWSU”) had recorded only one death of a genocide survivor in 2010.
Risk of arbitrary arrest and detention or being charged with committing perjury or complicity in genocide
Some witnesses have been arbitrarily arrested, detained and, in some cases, prosecuted for giving false testimony. Initially, gacaca courts were permitted to immediately charge a witness with perjury and convict him or her during the same hearing. However, in 2006, the SNJG officially instructed the courts to try perjury cases only after the case in which the alleged perjury occurred had been concluded and to do so in a separate hearing. The same instructions also stated that individuals accused of perjury should not be taken into custody pending trial. However, Human Rights Watch documented a number of cases in which individuals, usually defense witnesses, were immediately tried for perjury, even after the 2006 instructions.
In one case in 2007, a defense witness working in a hospital during the genocide told the court he did not know how victims found dead in the hospital had been killed because he was not at the hospital at the time. He suggested the court ask his former supervisor who was also present at the hearing. The supervisor, a genocide survivor, accused the witness of genocide denial, leading another community member present at the trial to do the same. The court immediately accused, tried, and convicted the witness of perjury, sentencing him to five months’ imprisonment.
In a 2009 case, a court threatened to charge all 12 defense witnesses with being accomplices of the accused, without explanation, even though none of them had said anything to implicate themselves in the alleged crimes. The court then ordered all of them to be arrested. The 12 remained in detention until the end of the three-week trial, at which time the court convicted them of perjury and sentenced them to prison terms ranging from three to six months.
In a third case in 2010, two men who had already confessed to their own crimes and completed their prison sentences appeared as defense witnesses in a separate case. The presiding judge interrupted their testimony and instructed them to sit on the ground next to the judges for the remainder of the hearing. Police detained the two men overnight and the court convicted them of perjury the next morning, for having testified that they did not see the accused in the community during the genocide.
In a fourth case, that of former sous-préfet Théodore Munyangabe discussed above, the judge coerced a witness into making a statement implicating the accused in planning killings during the genocide. At the appeal stage of Munyangabe’s case, two genocide survivors were arrested and detained overnight after they came to testify in his defense.
In a 2008 case marred by irregularities, a gacaca court charged 13 defense witnesses with perjury and found that they had all lied to defend the accused. At the very next hearing, the court convicted them, together with another five defense witnesses. All but three of them were given prison terms ranging from six to 12 months.
In other cases, individuals summoned to appear as witnesses found themselves charged as co-accused. In some of these cases, the court did so intentionally in what appeared to be an attempt to trick persons into appearing at a hearing. In others, the sudden charges resulted from the person’s testimony as a witness, usually on behalf of the accused. In one case, a woman who testified in defense of a man accused of involvement in her relative’s death because she knew him to be innocent was convicted as a co-conspirator and sentenced to 19 years in prison. The decision was affirmed on appeal but was overturned at the revision stage after the SNJG intervened. In another case, a court punished a man for appearing as a defense witness by charging and convicting him of a crime of which he had already been convicted. Similarly, in another case, a court accused a defense witness, Célestin Rusanganwa, of genocide and placed him in detention. The court acquitted the accused in the original case but convicted Rusanganwa and sentenced him to 19 years' imprisonment.
Fear of being ostracized by the community
People with relevant information sometimes chose not to come forward, fearing repercussions in their local communities or with the government. In several cases documented by Human Rights Watch, persons with information that could have helped accused persons defend themselves against genocide-related charges but who chose to remain silent later apologized to the accused or his or her family.
One genocide survivor broke down in tears in September 2007 as he told a Human Rights Watch researcher how ashamed he was at having refused to testify as a defense witness at the gacaca hearing of a man accused of genocide who had saved his life and those of more than a dozen members of his family.The fact that some of the accused or their relatives said they understood why potential defense witnesses had not come forward and excused them for not testifying is indicative of how real the fear for potential defense witnesses was.
In September 2008, a gacaca judge in the southern part of the country tried to cause problems for a man who spoke out in defense of an accused person. The judge, who was not deciding the case and who was merely attending the trial as a member of the public, asked for the man’s identity card after he testified. When the man refused to give it and asked why the judge was interfering in the proceedings, the judge ordered the man to present his identity card to the judges deciding the case. An SNJG lawyer observing the trial then accused the witness of being an intelligence agent and intimidating the population. The man reacted by vehemently accusing the judge and the SNJG lawyer of trying to intimidate him into not testifying for the accused and of trying to cause problems for him within the local community. Human Rights Watch is not aware of whether the man suffered any reprisals after the hearing.
Individuals appearing as prosecution and defense witnesses in gacaca faced intimidation, most often by police and other state agents, but also by genocide survivors and civil parties. In some cases, individuals accused of genocide were believed to be behind the intimidation, targeting community members or witnesses who accused them of crimes.
The 2004 Gacaca Law provides that anyone found guilty of exerting pressure on, or threatening, a witness or judge is liable to between three and six months’ imprisonment. The punishment is doubled for repeat offenders. Individuals should be prosecuted in the conventional courts under the penal code. Where appropriate, a special protection unit in the national prosecutor’s office investigates and prosecutes these cases. In 2009 alone, there were 473 such investigations resulting in 181 cases heard in the conventional courts.
According to the VWSU, most intimidation occurred during the national trial phase of gacaca. The majority of cases documented by the VWSU involved accused persons or their relatives verbally threatening genocide survivors and their close relatives. Such cases came to the attention of VWSU when genocide survivors sought assistance or protection from the unit. The VWSU also recorded incidents where accused persons threatened other detainees or released convicts and their close relatives in response to these individuals’ statements accusing them in gacaca.
The VWSU also documented a significantly smaller number of cases where genocide survivors intimidated other survivors who defended individuals before gacaca and where judges or community members intimidated defense witnesses.
In an interview with Human Rights Watch, the VWSU coordinator Théoneste Karenzi stated that “we were contacted by lots of survivors but many less 'non-survivors'.”In Rwanda, only Tutsi may be considered “survivors.” Based on interviews conducted with defense witnesses, Human Rights Watch has concerns that ethnicity may have been a factor in witnesses’ confidence in contacting the VWSU and may have resulted in underreporting of incidents involving defense witnesses.
The VWSU took steps in 2009 to raise awareness of its services through radio announcements and a television documentary describing its services, as well as meetings with local authorities and police in every district. However, it may be too soon for this awareness campaign to show results in terms of encouraging "non-survivors" (meaning Hutu) or defense witnesses to seek assistance from the VWSU.
Human Rights Watch documented a number of cases in which judges were intimidated. Two gacaca judges from the north of the country contacted Human Rights Watch in November 2006 to report that they had been instructed by the district coordinator to formulate accusations against a particular man. A number of judges resisted because no accusations had been made against the man during the information gathering phase. According to the judges, however, two other judges in the jurisdiction quickly produced written testimony against the accused, which they believed was fabricated, and a hearing date was set. Before the trial, the district coordinator told the judges to convict the man, warning, “If you don’t, you will be punished.” Later, during deliberations among the judges, a soldier burst into the room and tried to speak with the judges. The judges made him leave, but the two judges who spoke with Human Rights Watch said that they were frightened as the case was being monitored “from above” (they claimed not to know by whom or for what reason).
In another case, a local police commander arrested two gacaca judges on August 14, 2005, after they declined to bring genocide-related charges against an individual whom the police commander wanted to be convicted. The police commander accused the judges of planning to derail the gacaca process. A new presiding judge was appointed to take over the case, but soon found himself in a similar situation. When he presented the case file to the police commander, the commander threw the document on the floor and said that the judge needed to change the report, as previously instructed, or he too would go to prison.The two other judges were charged with gross minimization of the genocide. They spent more than 14 months in detention and were eventually acquitted in October 2006.
Police and military presence at gacaca trials often appeared to create anxiety among local residents. The 2004 Gacaca Law allows security agents to be present during trials, both to ensure order and to participate as members of the community, but is silent on whether they may bear arms. It is customary to see security agents in conventional courts but their presence in the more informal gacaca courts—particularly when armed and in uniform—seemed to influence participants’ willingness to speak out in gacaca proceedings. In general, Human Rights Watch observed less community participation when police officers or soldiers attended proceedings. Some participants told Human Rights Watch that they were fearful of speaking out when police or soldiers came to gacaca trials and worried that they might be arrested.
In some cases armed police or soldiers deliberately misused their position of authority at hearings to influence witnesses and community members. For example in one hearing in 2007, a Human Rights Watch monitor overheard a member of the civil party, an influential man who worked at the local hospital, say on the phone that he would send a car to bring people to the hearing. Approximately 30 minutes later, a car belonging to the hospital, which the man often drove, arrived at the site where gacaca proceedings were under way. The car carried three police officers: the local police commander and two armed officers. All three joined the hearing and remained there throughout its duration.
At the end of the hearing, the trial was not completed and the accused was free to leave. However, the Human Rights Watch monitor saw the police commander lead the accused into the hospital’s car with the other two police officers and overheard the commander say, “You won’t escape me.” A large number of local residents, including genocide survivors who had testified in defense of the accused, stood in front of the vehicle and blocked the road to prevent the police from detaining the accused. After a few minutes, the police managed to disperse the crowd and took the accused to the police station.He remained in detention until the next hearing the following week. Several people told Human Rights Watch that the accusations against him were false. During the hearing, Human Rights Watch heard some of the same individuals, including a local gacaca judge who had not participated in the case, testify that the civil party had tried to pressure them into accusing the man. The trial and appellate courts both acquitted the man, but a gacaca court convicted him of genocide-related charges and sentenced him to 19 years’ imprisonment at the revision level.
In the trial of former presidential candidate Théoneste Niyitegeka in 2008, four soldiers arrived more than an hour after the hearing had begun and went to the front of the crowd, momentarily disrupting the proceedings and making their presence well known. Later the same day, a military police vehicle arrived and parked next to the place where proceedings were taking place while the judges were deliberating, causing fear among the population and leading some to believe that the outcome of the trial was predetermined.
In another case, several uniformed police officers sat with the civil party and were seen talking together throughout the trial. Their presence led some residents to conclude that the police’s support of the civil party meant that the accused would be convicted and deterred at least two individuals from speaking out in the accused’s defense.
Gacaca as a Means of Resolving Personal Grievances
In gacaca there were a lot of personal disputes that had nothing to do with the genocide.
— Genocide survivor, Butare, August 14, 2009
Between 2005 to 2010, Human Rights Watch documented dozens of cases in which individuals used gacaca to try to settle personal scores, falsely accusing someone of genocide or genocide-related crimes. In most instances, both the accuser and the accused had resided in Rwanda for more than a decade and the accuser offered no reason for having failed to make the allegations sooner (e.g. during the information gathering phase). Ethnic hostility between Hutu and Tutsi sometimes appeared to explain the behavior, but usually simple personal grievances and financial motives were the cause. Human Rights Watch’s trial monitoring suggests that such cases increased from around 2007, perhaps in part because people saw how the process worked and felt increasingly confident that they could use gacaca to resolve disputes over land, inheritance, and local economic inequalities.
A dramatic increase in such cases in late 2007 and early 2008 led international organizations following the gacaca process to call on the SNJG to announce an end date for gacaca. There is no statute of limitations for genocide, which means that any cases arising after a cut-off date would be handled by the conventional courts. These courts would have trained legal professionals to properly review new cases and to help identify—and hopefully discourage—false accusations. The SNJG dismissed these concerns in a meeting with the international organizations and appear not to have considered the proposal.Cases of false accusations based on personal vendettas and other outside interests continued throughout 2009 and 2010.
An illustrative case of how gacaca risks being misused to settle personal scores is one in which a family used gacaca to try to settle a land dispute under a 1959 agreement involving a neighbor and his son.One family accused their neighbor and his son of genocide. While the gacaca court acquitted the father,it convicted and sentenced the son to 30 years in prison. During the trial, the civil party bringing the case acknowledged that the families had a land dispute, but denied that this was why his family had brought genocide allegations to court. The trial court believed him. However, the appeals court accepted evidence that the civil party had made false accusations and had tried to use gacaca to settle the old score over land and convicted the civil party of perjury; a second gacaca appeals court overturned that decision.
In another case, a man who in 1994 had lived in the northwestern town of Gisenyi was accused of committing genocide-related crimes more than 125 kilometers away in Kibuye, his native town where he returned after the genocide. According to the accused, the case arose out of a private dispute he had with the family of a local genocide survivor. The genocide survivor’s child had stolen items from his house in March 2000 and was ordered to pay back 20,000 Rwandan francs (approximately US$33). According to the accused, the survivor—who had served as the Ibuka representative in the community—then brought genocide-related accusations against him in retaliation for having raised the theft case with local authorities. The accused managed to prove that he was 125 kilometres away in Gisenyi at the time of the massacres and that he arrived in Kibuye later in 1994. However, he admitted to having been in possession of a grenade, allegedly for protection (he later turned it over to the RPF once they had taken control of the country). The gacaca court convicted him of illegal possession of a grenade during the genocide and imposed a 19-year prison sentence; his request for revision of the judgment was denied.
Cases can also involve parties who have competing business interests. The case of Aphrodis Mugambira, a businessman in Kibuye who owns a hotel and other valuable property in the area, is a striking example of personal interests driving gacaca cases, as well as of the violation of the right not to be tried twice for the same crime. Mugambira spent nearly 10 years in prison without trial after the genocide before finally being prosecuted. In November 2002, a conventional court convicted him as a category 1 offender, but the decision was reversed on appeal in 2003 and he was acquitted. During his detention, a former high-ranking government official had appropriated Mugambira’s hotel. After Mugambira’s release, the official, perhaps worried that Mugambira would want to reclaim his land, joined forces with an influential policeman and other businessmen in the area to bring renewed charges against him in gacaca. In August 2008, police rearrested Mugambira on the very same charges for which he had already been tried in the conventional courts. The gacaca court convicted him and sentenced him to “life imprisonment with special provisions,” a decision affirmed on appeal. After the SNJG expressed concern about a possible violation of the double jeopardy rule, the case was dismissed at the revision stage. This case is an example of the disappointingly infrequent occasions in which the SNJG stepped in to correct procedural errors or miscarriages of justice. The SNJG’s intervention in this case resulted in a positive outcome, but the fact that the SNJG had to intervene at all to correct such problems points to the inherent weaknesses within the appeals mechanisms and the gacaca system as a whole.
Another case involved a genocide survivor who had hidden in a man’s house to escape the killings, but who later fell out with the man. Once appointed as a gacaca judge, the genocide survivor brought charges against her former rescuer, apparently because he had not married her. At trial and again on appeal, the accused claimed that the allegations were false and had been brought as retaliation for their falling out as friends. He pointed to the fact that all of the witnesses against him belonged to the woman’s family. Several community members, including genocide survivors, confirmed that the two had been close friends and that the relationship had soured when the accused declined to marry the woman. The trial court acquitted him. At the appeal hearing, one of the judges had to be disqualified because he was related to the woman. The appeals court then affirmed the acquittal, but the police detained him on allegations that he had tried to bribe the woman to drop the case. A conventional court acquitted him on those charges too, but again the police kept him in custody pending a revision of the original case. Despite no new evidence or proof of manifest error in earlier proceedings, the gacaca court convicted him on the charges of involvement in the death of a woman and a genocidal attack in the area and sentenced him to 15 years’ imprisonment.
In a 2008 case, two nurses at Gahini Hospital in eastern Rwanda had fallen out shortly after the genocide, leading one to accuse the other of having refused to suture the wounds of a young Tutsi boy injured in the genocide (who was later killed at the hospital). During the trial, other genocide survivors accused the woman of “hating Tutsi,” although they provided no evidence. The court convicted the woman and sentenced her to 15 years’ imprisonment. One of the trial level judges was known to have a personal conflict with the accused’s family in connection with their local parish. The judgment was overturned on appeal, but at the revision stage, the court again convicted the woman and sentenced her to six years’ imprisonment with conversion of the prison sentence into community service.
In some cases, gacaca was allegedly even used within families to settle arguments. In a 2009 case, a female gacaca judge was said to have resented her sister over their inheritance when their parents died. She used her status in the community to persuade others to accuse her sister of genocide-related offenses. After a gacaca court had convicted her sister, several residents in the community, including genocide survivors, jointly wrote to the SNJG explaining what had happened. The SNJG intervened and the woman was released.
Silencing Opponents and Critical Voices
Private citizens were not the only ones to manipulate the gacaca process for personal ends. Government officials and influential politicians also lodged allegations with gacaca courts in contexts which strongly suggest that the goal was to silence outspoken critics and potential political opponents.
The case of Dr. Théoneste Niyitegeka
Dr. Théoneste Niyitegeka, a surgeon at a central Rwandan hospital, who cared for many wounded persons during the genocide, ran into problems shortly after he decided to run for President in the 2003 elections. His candidacy was rejected and he filed a complaint, after which the police detained him for three days of questioning on statements he had recently made.After his release, Dr. Niyitegeka continued practicing medicine, and occasionally commented on Rwandan politics in the local and foreign press. In 2005 he criticized gacaca in a radio interview with the Voice of America. After the police interrogated him about his comments on the program, Niyitegeka left the country for a short period, returning to Rwanda after he thought the situation had calmed down. Shortly after his return, unidentified persons blew up his car outside his house. Soldiers later came to his house and unsuccessfully tried to pressure him into publicly retracting his criticism of gacaca.
Accusations were then brought against him in gacaca. In October 2007, a gacaca court acquitted him on charges of having turned patients over to soldiers to be killed in 1994. An appeals court overturned the decision, sentencing Niyitegeka to 15 years’ imprisonment, without providing any explanation for the reversal. The decision surprised many because the only two witnesses against Niyitegeka contradicted themselves and more than a dozen witnesses—including doctors, nurses, and patients—gave exculpatory testimony. One genocide survivor testified that the doctor had provided her with excellent care for a serious wound and that he had treated Hutu and Tutsi patients without distinction. Niyitegeka requested revision of his conviction, but his request was denied. At the time of writing, he remains in prison.
The case of Father Guy Theunis
Another troubling case which appears to have been politically motivated is that of Father Theunis, a Belgian priest, human rights activist, and journalist who lived in Rwanda between 1970 and 1994. Theunis was the editor of the periodical Dialogue, originally published in Rwanda and now released in Belgium, which often featured articles critical of the Rwandan government. In 1990 Theunis helped launch one of the Rwanda’s first human rights organizations, the Rwandan Association for the Rights of the Individual and Public Freedoms (ADL). Before and after the genocide, he documented human rights violations affecting Tutsi and Hutu alike.
Theunis returned to Belgium shortly after the genocide began. He returned to Rwanda briefly in 2004 without facing any problems. However, in September 2005, he was arrested as he transited through Rwanda en route from the Democratic Republic of Congo to Europe. The prosecutor’s office hastily cobbled together a case against Theunis, and he was brought before a gacaca court just five days after his arrest (in stark contrast to the tens of thousands of Rwandans awaiting trial for long periods in prison). Theunis was charged with incitement to commit genocide through his writings.
The real impetus for the case appeared to have come from persons hostile to the Catholic Church for its role in the genocide, including some high-ranking RPF members who were seeking to gain control of Dialogue and its assets.
A score of witnesses, several of them prominent RPF members, denounced Theunis for having supported the genocide. They relied on a distorted reading of some his writings, ignoring, for example, the distinction between his own words and those he was quoting (indicated by quotation marks). His efforts to alert others to the genocide were misrepresented as efforts to discourage international involvement. Some of the witnesses read from prepared statements, highly unusual in gacaca sessions where participants generally speak spontaneously. The case was highly politicized; one high-ranking military officer in the audience remarked to a Human Rights Watch researcher during the proceedings that he was “gratified” to see the Catholic Church humiliated.
At Theunis’ hearing, an estimated 1,700 persons, some alerted by repeated announcements on the radio, attended.The usual restrictions on the attendance of foreign nationals and on audio and visual recordings were relaxed, apparently to attract greater attention to the proceedings.
The gacaca judges concluded the hearing by classifying Theunis as a category 1 suspect and ordering his trial in the conventional courts. Theunis remained in detention. Following a request by the Belgian government, the case was transferred to Belgium in November 2005. Belgian police released Theunis (who had returned to Belgium) while they investigated the case. Concluding that the file was “empty of any real proof,” Belgian judicial authorities have since closed the case.
In recent years, several parliamentarians have faced genocide accusations in cases that appear to have little connection to the genocide. Alfred Mukezamfura, a journalist who later became a prominent member of parliament after the genocide and was speaker of the Chamber of Deputies from 2003 to 2008, was accused of incitement to genocide for the first time in the spring of 2008. Mukezamfura led the Centrist Democratic Party, which supported Paul Kagame’s candidacy in the 2003 presidential election, but had been known to speak out against the official government line. He traveled to Belgium in March 2008 for medical care, and claimed asylum there after rumors began to circulate inside Rwanda about his involvement in the genocide. Gacaca courts tried him in absentia and sentenced him to “life imprisonment with special provisions,” having concluded that several of his articles published in the government-run weekly newspaper Imvaho in 1994 had called on the population to take up arms and begin killing Tutsi.
Another politician, Stanley Safari, who served under the government of Juvénal Habyarimana (the president of Rwanda from 1973 until his assassination in 1994) and who later became a member of parliament for the Prosperity and Solidarity Party (PSP), first faced genocide accusations in the spring of 2009. Safari had become increasingly critical of the government. Among other things, he had told a high-level delegation visiting Rwanda to consider whether the country should be admitted to the Commonwealth, that political parties were restricted from freely expressing themselves and that there was no real democracy in Rwanda. Safari fled the country just days before a gacaca court convicted him of genocide, sentencing him to “life imprisonment with special provisions.” Several days later, the Senate expelled him, citing his failure to appear for work. Shortly after he fled, a parliamentary commission heard accusations that Safari’s divisionist ideas had caused infighting within the PSP.
Several months before, Béatrice Nirere, an RPF member of parliament, faced similar problems due to what appeared to have been another RPF member’s political ambitions. Nirere had been a sous-préfet of Byumba before 1994. During the nationwide information gathering phase for gacaca, noone accused her of any wrongdoing. It was not until several months after she had been elected to the lower house of Parliament in September 2008 that genocide-related accusations first surfaced. A gacaca court sentenced her to “life imprisonment with special provisions” in March 2009, a decision affirmed on appeal. At the time of writing, she remains in prison. An RPF member who had been lower down on the RPF nominee list than Nirere (and who had therefore not been selected), instigated the case and took over Nirere's parliamentary seat after her conviction.
Other outspoken critics of the Rwandan government have also faced apparently politically-motivated accusations in gacaca courts. Léopold Munyakazi, a Rwandan academic in exile, became a target after challenging the government’s official discourse on the genocide while teaching in the United States. The Rwandan government reacted by issuing a series of international arrest warrants against him and later initiating gacaca proceedings against him in his absence. The gacaca case was later dropped at the direction of the SNJG.
Jean-Léonard Rugambage, an independent journalist, was arrested on genocide accusations in September 2005, just 10 days after he published an article in Umuco newspaper accusing gacaca officials in the Gitarama region of mismanagement and interference with witnesses. He was tried in gacaca in November 2005. When Rugambage accused one of the judges of bias and called for him to be disqualified, the court sentenced him to one year’s imprisonment for disrespecting a judge (contempt of court). He was later placed in category 1 and set to be tried before the conventional courts. However, in July 2006, the SNJG stepped in and ordered the journalist’s release due to a lack of evidence and procedural irregularities. The Committee to Protect Journalists argued that Rugambage was “a victim of abusive procedures designed to punish him for critical reporting.”
On June 24, 2010, an unknown assailant shot and killed Rugambage outside his home, the same day that the newspaper he worked for, Umuvugizi, published an article alleging the involvement of senior Rwandan government officials in the attempted assassination of former Rwandan general (turned outspoken government critic), Faustin Kayumba Nyamwasa, the previous week in South Africa. The government prosecuted two men for Rugambage’s murder. One of the men had immediately confessed upon arrest, and claimed that he and his co-accused were avenging the death of a brother whom they alleged was killed by Rugambage during the genocide.These allegations had formed the basis of the case against Rugambage in gacaca but were never proven, and Rugambage had not been convicted of murder. In October 2010, both men were convicted and sentenced to life imprisonment. One of them was ordered to commence his sentence immediately; however, the second—a police officer— was allowed to remain free on bail pending an appeal scheduled for July 2011.
Human Rights Watch conducted its own investigation into Rugambage’s murder and identified several leads suggesting Rugambage may have been murdered in retaliation for his critical reporting. Rugambage had also complained of increased surveillance in the days before his murder. However, there is no evidence that the police made any effort to explore these leads, and advised Human Rights Watch that their investigation was closed after one of the suspects confessed. The police exclusively presented the theory that this was a revenge killing linked to events in 1994, in an apparent attempt to exclude the possibility of official collusion in Rugambage’s murder.
The creation of gacaca was a good thing because it allowed the population to play a large role in the gacaca process. But I deplore you [the judges] for taking sides…
—Man testifying as a witness in gacaca, Save, September 18, 2008
Gacaca judges try cases relating to events that happened in their own area. Having lived through the genocide, many have their own strong views about what happened and know some or all the parties in any given case, whether they are relatives, friends, neighbors or business partners. Rwandan and international observers believe these factors have given rise to potential conflicts of interest or inherent partiality, and that with even the best will in the world, most gacaca judges inevitably struggle to evaluate evidence impartially.
Gacaca has also seen widespread corruption and a pattern of political interference with the judiciary. Both phenomena occur in the conventional justice system too but appear to have been more pronounced in gacaca. Judges were not the only ones who profited: accused persons and genocide survivors also sought personal gain by engaging in corruption. At times local officials, particularly district coordinators, interfered with the decision-making process. Both the lack of independence of the courts and corruption weakened public confidence in the system and led to decisions that did not reflect what really happened during the genocide.
Potential Conflicts of Interest for Judges
As discussed above, to date the SNJG has removed more than 45,000 gacaca judges from their positions because of accusations of their involvement in the genocide. Many of these judges were tried in gacaca courts after they were dismissed from their position. Judges who are themselves genocide survivors of the genocide or who lost close relatives may also have found it hard to remain impartial. But beyond these clear-cut cases, it has not always been easy to identify less obvious conflicts of interest, such as little known family or business ties to key parties in a case.
In one case that demonstrates the emotional vulnerability of judges, a judge accused a woman of involvement in the death of her own child. The woman replied: “You people said we should tell the truth and yet you are a judge and you don’t tell the truth.” In response the judge shouted, “Keep quiet! I know that my child will never rise again from death.” Human Rights Watch has not documented many cases involving such emotional exchanges with judges, but the case illustrates the difficulties some judges have in putting aside their personal experience when deciding gacaca cases. One gacaca judge, who is a genocide survivor, openly told Human Rights Watch that she found it difficult to remain impartial in many cases because the victims and the accused were all neighbors. However, there were also many judges who are genocide survivors, and who showed no bias, demonstrating an apparent capacity to put aside their feelings and focus on the evidence at hand.
Under the law, a judge must disqualify him- or herself if (i) one of the parties is a spouse or relative (defined as parents and siblings all the way to the level of cousin), (ii) a serious conflict or a close friendship exists between the judge and one of the parties, or (iii) the judge is the guardian of one of the parties. Usually, at the beginning of each trial, the presiding judge asks the parties if anyone has an objection to any of the judges. If someone raises an objection, the judges withdraw to decide on the matter. Many cases are resolved properly, but Human Rights Watch documented a number of cases where judges refused to disqualify themselves in these situations.
Corruption and Personal Gain through Gacaca
Many Rwandans—genocide survivors, accused, witnesses, and judges alike—told Human Rights Watch that, over the years, gacaca became a lucrative “business.” Almost everyone interviewed agreed that corruption affected decision-making in gacaca courts. Some spoke about their own stories or about cases of which they had direct knowledge.
Cases included judges accepting bribes from wealthy accused persons in exchange for acquittals or asking the accused to pay money in exchange for an acquittal; genocide survivors accusing wealthy people in the community of crimes in order to receive monetary compensation to drop the case; witnesses taking bribes from the accused; and civil parties bringing cases in exchange for making false allegations, changing their testimony, or defending an accused person. The SNJG has been aware of such corruption and, in some cases, the police attempted to arrest those responsible. The SNJG told Human Rights Watch that it does not keep statistics on such cases. Human Rights Watch also asked Rwanda’s Ombudsman for his views on the issue of corruption in gacaca, but he declined to provide any information. In January 2008, the SNJG executive secretary reported that 56,000 ineffective or corrupt judges had been removed from service. It was not clear how many of these cases involved corruption instigated by judges as compared to judges accepting payment from one of the parties to a case. On the basis of its own research and observations, Human Rights Watch believes that there were many more undetected cases.
Corruption also occurs in the conventional courts, although the phenomenon of money physically changing hands between judges and parties does not appear to be as pervasive as in gacaca, according to Human Rights Watch’s research. In a 2008 report, the Office of the Ombudsman ranked the judiciary as the second most corrupt state institution, falling after the traffic police.In February 2011, the chief justice of the Supreme Court denounced the continued problem of corruption in the conventional courts and reiterated that the government intends to prosecute offenders.
Judges requesting bribes
In gacaca, the largest number of corruption-related cases documented by Human Rights Watch involved judges taking bribes from accused persons. As one accused said: “You have to give money. Gacaca judges were not paid so they sometimes made arrangements to receive money from those who were accused.” Several genocide survivors who saw accused persons walk free despite strong evidence against them agreed, as did the genocide survivor organization Ibuka.Penal Reform International (PRI) reported an increase in the number of corruption cases after gacaca trials accelerated in 2007 and after gacaca activities were listed as a component in local government “performance contracts” (benchmarks set by the national government).
In a number of cases, judges used intermediaries—persons known to both the judge and the accused—to contact the accused or his or her family to request money in exchange for an acquittal. The accused or the family paid in cash, wrote checks, or deposited money into the intermediary’s bank account. The amounts paid depended in large part on the socioeconomic status of the accused, with documented cases ranging from 100,000 Rwandan francs (approximately US$165) to 5 million Rwandan francs (approximately US$8,200). In a 2009 case, the accused wrote a check but then reported the incident to the police who forced the intermediary to return the check and arrested the intermediary. In another case, a priest told the police that after his acquittal, one of the judges approached him and asked for money so that the judge could “discourage” the civil party from appealing the verdict. Police arrested the judge and he was prosecuted in a conventional court. More often, however, the transaction went ahead as planned and the accused received a favorable resolution to his or her case, as documented below.
In some cases, the initial payment turned out not to be enough to secure an acquittal. In a 2009 case in Kigali, a man paid an intermediary 100,000 Rwandan francs (approximately US$165). Soon after, the intermediary returned and said that the judges would require a further 300,000 Rwandan francs (approximately US$495) to guarantee an acquittal, to be paid when the case was over. The man was acquitted and he paid the remaining funds. Human Rights Watch documented a similar case in another part of Kigali, where an accused paid a total of 1.3 million Rwandan francs (US$2,140), some before and some after, to help secure an acquittal.
In a few cases, the accused or his or her relatives refused to pay a bribe. In a 2009 case, a man (correctly, as it turned out) felt confident that his wife would be acquitted of any wrongdoing in gacaca and declined to pay. She was later acquitted.For others, failure to pay a bribe resulted in a conviction.
In most cases, only one or a few of the judges were involved in the arrangement. One genocide survivor recounted that a large number of cases had to be reviewed in her sector because the district coordinator had cooperated with judges in taking bribes from accused persons. In a few isolated cases, judges approached accused persons known to have paid bribes in the past to seek similar remuneration to make new cases “go away" or to acquit them in such cases.
Accused persons seeking exoneration
Human Rights Watch also documented cases in which the accused approached judges or genocide survivors, either directly or through an intermediary, and offered them money in exchange for an acquittal or to encourage the victim to drop the case. In some of these cases, the person admitted to Human Rights Watch researchers that they were guilty, but said that they did not want to suffer the humiliation of a conviction or to be sent to prison. In other cases, individuals maintained their innocence, but said they had offered to pay a bribe because they were afraid they would be convicted on the basis of outside considerations, or because they did not have sufficient defense witnesses to help prove their innocence.
In a 2009 case in Kigali, an accused man’s brother approached one of the gacaca judges whom he knew personally and asked her whether she would be willing to accept money in exchange for an acquittal. The judge consented, but the case was later transferred to another jurisdiction for a new trial. The brother approached one of the new judges and paid him 100,000 Rwandan francs (approximately US$165). The court nonetheless convicted the man and sentenced him to 19 years in prison. On appeal, the brother paid the presiding judge, a friend of the family, 250,000 Rwandan francs (approximately US$412) to be split between the five judges. The court still convicted the man but imposed a reduced sentence, less than the period of time the man had already spent in pre-trial detention. He was therefore released.
Another man told Human Rights Watch that his family had offered judges 120,000 Rwandan francs (approximately US$198) in exchange for an acquittal. They accepted and he was acquitted in 2009.
In another case, a man confessed to bribing a gacaca judge in order to secure an acquittal and was prosecuted for corruption in a conventional court. The court sentenced the man to eight years’ imprisonment and imposed a fine amounting to double what he had paid to the gacaca judge. The prosecutor also argued that the act of bribing a gacaca judge indicated that the accused was minimizing the gravity of the genocide. The court disagreed and acquitted the man of that charge in 2008.
Human Rights Watch documented only one case in which an accused offered to pay a victim in exchange for the victim dropping the case. A rape victim told Human Rights Watch she had accepted money from a man who raped her during the genocide in exchange for dropping the case.
Genocide survivors seeking compensation
Many genocide survivors’ destitution and frustration at the lack of compensation for their losses and injuries explained, at least in part, why some approached accused persons and offered to drop the case against them in exchange for money. Human Rights Watch documented only a handful of such cases, all of which involved intermediaries. In one particularly troubling case in 2009, a civil party and several other genocide survivors in the community offered to drop a case against an accused woman in exchange for payment, but later denounced the woman for corruption when she was unable to pay an additional sum requested of her.
Another accused, Jean-Népomuscène Munyangabe, whose case has already been discussed, requested the disqualification of the presiding judge because he suspected that he had taken a bribe from the civil party bringing the case. After the presiding judge refused to disqualify himself, Munyangabe’s younger brother told the community members attending the hearing that he had witnessed private meetings between the civil party and the presiding judge on two separate occasions, and suggested this was evidence that the judge was corrupt. The presiding judge reacted angrily, forcing the young man to sit beside the police for the remainder of proceedings and later opening a file against him for perjury.
A woman who accused a man of involvement in a Tutsi woman’s death later recanted her testimony after being convicted of perjury. She explained to the court that the victim’s son, who was the civil party in the case, had given her money, clothes and a metal roof for her house in exchange for her testimony implicating the man.
In another case, a housekeeper accused her employer at gacaca of raping her. The trial court convicted him of genocide-related charges, but the rape charge was not considered because it was a category 1 offense. At his appeal, the housekeeper retracted her accusation, saying that her uncle had encouraged her to falsely accuse her employer by promising her a cow.
External Interference in Decision-Making
In some cases, third parties interfered with gacaca proceedings. Most cases involved the district coordinator, who sometimes wielded considerable influence over gacaca judges and the gacaca process more generally. One gacaca judge told Human Rights Watch that the district coordinator regularly influenced decisions in his jurisdiction. As discussed earlier, coordinators sometimes failed to deliver summonses to the accused and detainees in a timely manner or failed to deliver release orders (known as “billets d’élargissements”) to prisons, with the result that acquitted persons remained detained. In some of these cases, the omissions appeared deliberate.
One troubling case documented by Human Rights Watch is that of Prudence Nsabimana. After being acquitted at trial and on appeal by gacaca courts in the southern part of the country and released from prison, Nsabimana reported to the SNJG executive secretary that an SNJG legal adviser had conspired with the district coordinator to delay his release from Muhanga prison. When the legal adviser learned that Nsabimana had gone to the SNJG to report on him, he worked with the district coordinator again to bring about a new summons for Nsabimana’s arrest on corruption charges, which the police executed the following morning. The corruption case never saw the light of day, but Nsabimana’s original case reappeared at the revision stage. The court convicted Nsabimana of injuring a Tutsi woman and pillaging her vehicle. The court sentenced Nsabimana to 15 years’ imprisonment, which he is currently serving.
Equally problematic were cases in which the district coordinator appeared to have directed the course of gacaca proceedings. Usually, the involvement occurred behind the scenes and took the form of district coordinators telling judges to initiate a case or how to decide a case. Sometimes, district coordinators joined judges during deliberations in a particular case and were said to have improperly influenced their decisions. The motives for district coordinators varied from their own private interests to assisting relatives or friends who sought to pursue genocide accusations against a person.
In one location, Human Rights Watch traced three separate cases against a man to the district coordinator who initiated the accusations. In the first case, the court acquitted the man. In the second case, the court said it had no jurisdiction because the allegations were identical to those heard in the first case. In the third case, the court convicted the man and sentenced him to 19 years in prison for the same crimes as in the first case. After the conviction, the genocide survivor organization Ibuka wrote to the SNJG denouncing what had happened and calling for the SNJG to reverse the conviction.At the time of writing, the man remains in prison.
Until 2008, genocide-related rape cases were heard in conventional courts. Because only a limited number of women came forward in the years immediately following the genocide, the government repeatedly encouraged women to report rape cases by reassuring them that their cases would be heard confidentially in the conventional courts.
In May 2008, the government changed course and passed a new law which transferred all such cases to the gacaca courts. The new law provided that the cases be heard behind closed doors (known as “huis clos”) in order to protect the victims’ privacy.
There were two main problems with this decision. First, despite the closed-door nature of the proceedings, placing these cases in gacaca courts meant that entire communities became aware of rape cases involving women who had initially decided to report the crime because their privacy would be better respected in conventional courts and their stories told behind closed doors where necessary. As a result, the goal of protecting rape victims’ privacy was seriously compromised and these women’s trust betrayed. Second, the decision to hold the trials behind closed doors in gacaca, which was meant to rely on community participation to challenge the veracity of testimony, led to considerable risks for both victims and the accused. Given the other fair trial concerns set out above, closed gacaca trials raised grave risks of miscarriages of justice. Gacaca courts derived their legitimacy from popular participation, so hearing these cases behind closed doors undercut the very rationale for using the local courts. While the decision to hold these trials behind closed doors was no doubt well-intentioned, it was simply not compatible with the nature of gacaca.
As outside observers were not allowed to observe these trials, little first-hand data exists on how rape trials were handled. Human Rights Watch conducted more than 20 interviews with rape victims, as well as judges and trauma counselors around the country, who were involved in gacaca hearings. Human Rights Watch also spoke with women’s and genocide survivors’ groups that provided trauma counseling to rape victims whose cases were tried by gacaca. Due to limited access to prisons, Human Rights Watch was unable to conduct interviews with persons accused of rape.
The Decision to Transfer Rape Cases to Gacaca
In May 2008, Parliament adopted a law transferring all category 1 cases to gacaca courts except for cases in which the accused occupied government positions at the préfecture level or higher.Just over 8,000, or 90 percent of these cases, involved rape or sexual violence.
The decision to transfer these cases to gacaca came as a shock to many of the rape victims interviewed by Human Rights Watch, some of whom had been reluctant to come forward in the first place and did so only after receiving assurances that their cases would be heard in the conventional courts and not in their local communities. The main women’s groups, including Avega (the association of widows of the genocide), Haguruka, Profemme, and the Rwandan Association of Trauma Counselors (ARCT), as well as Ibuka, opposed the transfer of rape cases to gacaca. However, the groups said they only met once with senior SNJG officials to raise their concerns and, in the absence of a serious public campaign to make their concerns and opposition well known, the proposal was adopted. Avega’s legal representative candidly said, “We knew the law would pass, so we didn’t publicly oppose it.” Another admitted that “we didn’t perhaps fight as hard as we could have for the rape cases not to be transferred.”
Defending the government’s decision, the SNJG’s executive secretary told Human Rights Watch that she traveled the country meeting with rape victims who told her that they wanted their cases to be heard by gacaca courts because many of them were dying of HIV/AIDS and wanted to see justice before they died. She said that, after the decision had been taken, she had received only a few letters from women who did not want their cases heard in gacaca.Of the over 20 rape victims Human Rights Watch interviewed in different parts of the country (more than a quarter of whom were infected with HIV/AIDS), only one said she preferred her case to be heard before a gacaca jurisdiction, because the procedures were less formal and she could “speak more freely.”
Most victims interviewed by Human Rights Watch said they were scared at the thought of speaking in gacaca about their rape and had only reluctantly gone ahead with their cases. They gave various reasons for their reluctance. First, most feared that their statements would not remain confidential, given that the judges were all members of their local communities and were sometimes even related to the accused. Four of the women’s groups and Ibuka also cited the lack of confidentiality as one of the main reasons they had opposed the transfer of rape cases to gacaca, while another organization said that many of the women they had assisted in gacaca felt that the confidentiality of their statements had not been protected. Second, some said that even if the precise nature of their case was protected, they felt that everyone in the community would still know that the case involved rape because, on the day of gacaca sessions, whether behind closed doors or in public hearings, community members would see a woman and a man enter a room (with others) and therefore guess the nature of the case.
Two women said they did not believe their cases would be judged fairly and impartially, given the judges’ ties within the community. One woman said that she did not have confidence in the process, because the brother of the man who had raped her had served as a judge in separate rape cases and that she believed he had also committed crimes during the genocide. Another rape victim also said that the people judging rape cases in her area were often closely related to persons accused of involvement in the genocide.
Several rape victims and a representative of a women’s group spoke about corruption in rape cases. One woman told Human Rights Watch that she would have preferred her case to go before the conventional courts where it would be more difficult for the family of an accused person to corrupt the judges. The representative of a women’s organization said that in gacaca, accused persons sometimes asked women to accept money in exchange for dropping their cases. Two trauma counselors who accompanied women through the gacaca process reported that some women accepted bribes from community members to claim falsely they had been raped. The above case of Dr. Justin Nsengimana—in which the woman initially accusing him of rape later changed her story and revealed that her uncle had offered her a cow to accuse her former employer—serves as an example. According to a gacaca judge who handled rape cases, individuals were sometimes falsely accused of rape when it became difficult to convict them of other offenses. Some Rwandans believed rape allegations were easier to prove because they depended largely, if not entirely, on the testimony of a single witness—the victim.
Two women felt that having their cases tried in gacaca minimized the seriousness of rape. Several women also said they believed the sentences to be too lenient, particularly when the accused confessed. Women’s groups and Ibuka agreed. A third of the rape victims interviewed by Human Rights Watch expressed frustration with the fact that they had received no monetary compensation after the accused was convicted. Under statutory law, a rape victim whose case is heard in the conventional courts is entitled to civil damages.
Rape Cases that Were Not Brought before Gacaca
A few women asked the SNJG to discontinue their case once it had been transferred to the gacaca courts. In one such case, the SNJG offered to appoint a gacaca jurisdiction from a different area to hear the case. In other cases, women decided not to pursue their case because they had not told their spouses or other relatives about the rape and did not want them to know or create problems for their family in their community. Some women chose not to proceed with their cases because they feared renewed trauma if they were to speak about what happened to them again.
Half of the rape victims who spoke to Human Rights Watch, including those who had suffered multiple rapes and who had seen other women pursue their cases in gacaca, said they had been unable to bring their rape cases at all because either the accused was not known to them or the accused had died. As one trauma counselor said, “many women who were raped by military or Interahamwe [militia who participated in the genocide] could not bring their cases because they didn’t know the perpetrator’s identity.” One victim who was unable to bring her case said, “I knew the face of the person but not his name.” Another rape victim reported that she had been unable to bring a claim against any of the nine men who raped her as they had all died. Yet another woman explained: “It would have been a relief to have him confess but there is nothing I can do. You are left with the trauma of him not coming.” These stories suggest that gacaca trials, and prosecutions more generally, were not sufficient to provide closure for some rape victims.
Rape Victims’ Perspectives on Gacaca
Women who appeared in gacaca in connection with rape cases had mixed experiences, with some feeling quite negative about the experience and others finding it less difficult than they expected. Under the gacaca rules, women have the right to bring one trauma counselor and a relative or friend to accompany them to the hearing, even behind closed doors. A number of organizations provided rape victims with trauma counselors, including Avega, ARCT, and Ibuka. The Victim and Witness Support Unit also took women who expressed fear of testifying in gacaca to the communal rooms where their trials would take place to familiarize them with the surroundings in advance of the trial.While many women received trauma counseling ahead of their trial and again at the hearing if needed, others were less fortunate due to the limited number of trauma counselors around the country and appeared on their own or with a relative or friend.
Gacaca courts often disposed of rape cases in a single hearing lasting anywhere from several hours to a full day, but some needed three to four sessions to decide a case. One gacaca judge told Human Rights Watch that the adjudication of rape cases had been “problematic” because the victims and the accused typically appeared, but that summoned witnesses often failed to show up. In some cases this led the court to adjourn the case two or three times, after which the court simply decided the case with or without the witnesses. In the majority of cases described to Human Rights Watch, the only persons who testified in the gacaca hearing were the victim and the accused.
Two procedural tools were introduced to make the experience easier for rape victims. First, like in other cases, victims had the right to request that a judge be disqualified from hearing their case. Unlike in other category 1 and 2 cases, however, the ability to disqualify a judge appeared almost automatic and did not require the victim to demonstrate a judge’s actual bias or conflict of interest. One woman disqualified a judge because she thought he would not respect her right to confidentiality. Second, rather than appear in person, women could write a letter containing their allegations, which was given to the district coordinator who then presented it to the gacaca court.This procedure would not have been possible in the conventional courts. While the procedure compromised the right of an accused to confront his accuser (the rape victim) directly and to challenge her credibility, it provided some degree of relief to women who were too frightened to appear in gacaca or to confront the men who had allegedly raped them. Human Rights Watch documented two cases where this procedure was used.
For most women, the experience of appearing in gacaca was emotionally difficult, and more difficult than they believed a conventional court trial would have been, but their cases proceeded relatively smoothly. One woman said that she could not reveal everything that happened to her in the hearing because she knew all of the judges from her community and did not feel comfortable telling them about the incident. Another woman believed that the gacaca judges asked “bad” or insensitive questions during the hearing. Most of the interviewees, however, believed the judges acted appropriately and in a manner that was sensitive to the situation, with one woman describing how judges “seemed to be listening” to her and another recalling how judges gave her a moment to calm down when she broke down in tears.
A few women said that they experienced problems as a result of their rape cases. One woman said that people had thrown stones at her house four times following the trial and had left handwritten notes with threatening words.She said she felt she had to go home early every evening and felt deprived of her ability to move freely in her community. Two women reported that community members had accused them of making false accusations simply because they had decided to pursue their rape cases. Another rape victim, who happened to be a judge of category 2 cases (unrelated to her own rape case), reported that community members’ threats and intimidation (including stones thrown at her house) had been so intense that she was forced to move to a different location. Another rape victim also had to relocate due to threatening notes she received at home. Avega documented several cases where rape victims had received threats from the individuals they had accused.
One troubling case involved a woman whose alleged rapist tried to intimidate, or perhaps even harm, her on the eve of his trial in November 2008. When she saw him and another man arrive in front of her house on a motorcycle, she immediately called the police and hid, leaving another woman in the house to deal with the men and to try to buy her time. The police arrived on the scene quickly and managed to arrest the alleged rapist. The man told the police that he had raped the woman and was coming to request forgiveness and offer her 200,000 Rwandan francs (approximately US$330). However, when they frisked the man, they found that he had little money on him—nowhere near the amount had he claimed to be offering her. The trial went ahead the following day, and the man–who did not confess–was convicted and sentenced to life with “special provisions.”
The biggest problem with gacaca is the crimes we can’t discuss. We’re told that certain crimes, those killings by the RPF, cannot be discussed in gacaca even though the families need to talk. We’re told to be quiet on these matters. It’s a big problem. It’s not justice.
—Relative of a victim of RPF crimes, May 30, 2004
One of the gacaca law’s most serious shortcomings is that it does not cover war crimes and crimes against humanity committed by the RPF as it sought to end the genocide between April and July 1994 and consolidated its control on the country in the months that followed. According to at least four UN bodies and a number of NGOs who collected testimonies, RPF soldiers committed war crimes and crimes against humanity during this period. A study by the UN High Commissioner for Refugees estimated that RPF soldiers killed between 25,000 and 45,000 persons between April and August 1994. These crimes are not equivalent to genocide, but the rights of the victims are equivalent: under Rwandan and international law, all citizens have the right to justice regardless of their ethnicity and political affiliation or that of the alleged perpetrator, and whether the crime is genocide, a war crime or a crime against humanity.
Under the 2001 Gacaca Law, gacaca courts had jurisdiction over war crimes. However, political considerations soon eliminated any hopes that victims of RPF crimes and their relatives might be able to seek justice through gacaca. In his June 2002 speech launching the gacaca process, President Kagame said it would be a grave error to confuse genocide with “acts of vengeance taken by individuals.” The 2004 Gacaca Law removed war crimes from the jurisdiction of the courts, limiting their remit to genocide and crimes against humanity, and a government-sponsored national public campaign insisted that RPF crimes were not to be talked about in gacaca.
Government officials have frequently said that anyone who suffered at the hands of a soldier should report him or her to the police for prosecution. But given that discussing RPF war crimes has been and continues to be equated with holding “genocide ideology” or arguing that a “double genocide” occurred, few Rwandans were likely to file such complaints.
The failure to deal with these crimes in gacaca and to provide people who lost relatives at the hands of RPF soldiers with some form of redress has caused bitterness and frustration for some Rwandans. The exclusion of these crimes from the jurisdiction of gacaca courts might not have been so serious had there been other avenues for victims of these crimes to seek justice. But very few RPF soldiers, and even fewer officers, have been charged or tried in connection with these crimes, and it is almost taboo to talk about these events publicly in Rwanda. The result is that most victims and relatives of victims of RPF crimes have all but given up on seeking justice. In 2009, the UN Human Rights Committee called on Rwanda to investigate and prosecute RPF soldiers responsible for the “large number of persons, including women and children, reported to have been killed from 1994 onwards.” No further actions have been taken since that time.
The Rwandan government maintains that RPF crimes have been prosecuted. However, to date, the military justice system has prosecuted only 36 former or current officers for killing or otherwise violating the rights of civilians during 1994. Most of those convicted were ordinary soldiers or of lower ranks and received punishments of less than four years that were not proportionate to the gravity of the crimes. The ICTR, for its part, has failed to prosecute any RPF crimes, even though these fall squarely within its mandate.
Many Rwandans are reluctant to speak openly on the subject of accountability for RPF crimes, but those willing to discuss the subject expressed frustration and dissatisfaction with the gacaca process. Some believed that the government has tried to impose an inaccurate single historical narrative—that the RPF stopped the genocide and saved the people of Rwanda from the atrocities without committing any crimes themselves—while others thought the RPF did not want to admit to the crimes because it might weaken its moral authority. The inability of victims of RPF crimes to raise their claims in gacaca courts, and the very limited options for doing so in any other forum, have hindered reconciliation efforts.
Has gacaca achieved its stated objectives? Has it revealed the truth about what happened during the genocide, accelerated trials, eradicated the culture of impunity, reconciled Rwandans, and proved that Rwanda has the capacity to settle its own problems?
Over the course of five years, Human Rights Watch interviewed a wide range of people involved in gacaca, including victims, genocide survivors, perpetrators, witnesses, other community members, judges, local and national government officials, and nongovernmental organizations. These Rwandans told Human Rights Watch how they viewed gacaca and its role in the aftermath of the genocide. While their views related specifically to gacaca trials, some of their concerns might have been equally relevant to the conventional courts.
Genocide Survivors’ Perspectives
A number of genocide survivors told Human Rights Watch that the gacaca process played a positive role in their lives. They said that most importantly, they and the broader community learned about what happened to their loved ones and that the process helped them to give their relatives “a proper burial.” Other genocide survivors challenged this position, saying that not all of the truth had been revealed during gacaca due to partial confessions, false accusations by all parties involved in the process, and judgments that did not always reflect the evidence presented at trial. Most agreed that they learned some valuable information about the events of 1994. Former Ibuka executive secretary Benoît Kaboyi summarized the success of gacaca as “having more or less informed us [the population] about what happened” and as “informing us of where the dead are.”
Genocide survivors had more mixed views on whether gacaca was the appropriate forum for genocide-related cases and on how gacaca trials played out in their local communities. Nearly all those interviewed agreed that gacaca reduced the prison population and processed cases faster than the conventional courts.
Many genocide survivors had concerns about corruption and judges’ partiality. A number of people referred to the community service program and the early release of certain categories of prisoners as examples of soft sentencing practices. Only a few people said they thought the sentences matched the crimes committed against them or their families. A number of genocide survivors also complained that gacaca courts provided no financial compensation to victims who had lost relatives or who were themselves injured or raped: only those whose property was looted or destroyed received reparations.
Opinions also differed on whether the gacaca process had eradicated the culture of impunity and would deter future violence. A significant number of genocide survivors, particularly widows, expressed a fear of renewed violence. Some believed the individuals whom they accused might take revenge on them once released from prison. Others who had received threats or been intimidated worried that individuals who took part in the genocide might come back to finish what they had started.
Human Rights Watch also encountered a wide range of views on gacaca’s role in promoting reconciliation. A number of genocide survivors said they were now able to greet their neighbors who had committed wrongs against them or could finally attend community events at which those neighbors were present. One judge declared that “gacaca has helped the situation because people are slowly approaching each other when they didn’t before.” However, many of the same genocide survivors indicated that these encounters were superficial and that tensions remained high between victims, perpetrators, and their families. As one genocide survivor said, “We say hello to each other but we don’t visit each other even though we were friends and shared beer together before the genocide.”
One woman said that the process had reduced her hatred towards the man who had raped her. But most genocide survivors said they remained distrustful of those who had wronged them. A number of survivors also raised the lack of remorse on the part of the perpetrators, saying that only those who destroyed or stole personal property expressed genuine remorse and asked for forgiveness. As one genocide survivor put it:
The young man who raped me whispered to me at the trial that if I forgave him, he would honor me in the future. He has never come to see me since he was released. I never see him even though he lives in the same neighborhood.
Other genocide survivors gave similar accounts. Some felt that the confessions were incomplete or lacking in detail, often because the confessions were aimed primarily at securing release from prison. Some genocide survivors explained that they felt forced to publicly forgive those who had wronged them even though in their hearts they had not forgiven them. As one woman said, “This is government enforced reconciliation. The government forced people to ask for and give forgiveness. No one does it willingly…The government pardoned the killers, not us.” Others spoke of the government’s “insistence” on reconciliation but reiterated how dire genocide survivors’ economic circumstances remained. A number of genocide survivors expressed bitterness over the government’s failure to give them financial assistance and to ensure their security.
According to many genocide survivors, reconciliation remained precarious. Many survivors spoke of the need to live peacefully and to co-exist with their Hutu neighbors, but most admitted that they still saw people through the lens of “Hutu” and “Tutsi.”
The Perspectives of Those Accused of Genocide and their Families
Many individuals accused of participating in the genocide echoed the views of genocide survivors by saying that gacaca’s main success was to help people understand what had happened during the genocide and to allow people to find and bury their loved ones. In general, however, they remained more critical about the role that gacaca has played in rebuilding the country. Their families often expressed similar views.
Most of the accused (some of whom were later convicted) believed that gacaca trials helped reduce the prison population and ensured that some of the innocent were released. Most understood that it would not have been possible to resolve the large caseload of genocide-related cases as quickly through the conventional courts. However, many believed that political considerations heavily influenced the gacaca process and that the resulting judgments were not always fair or based on facts. Most individuals referred to irregularities in their cases and felt that their rights had been sacrificed for the expediency of the process. Similar to many genocide survivors, individuals raised concerns over corruption, false accusations, and certain judges’ partiality. In addition, some believed their cases had little to do with the genocide and more to do with private disputes with neighbors, friends, or even relatives.
Individuals who confessed to their crimes tended to be more optimistic about the process. One man said that he had admitted to all of his crimes and that the community and his family welcomed his confession. While he acknowledged that he had only confessed because he feared accusations by genocide survivors, he expressed relief at having done so. Most persons who confessed agreed that it helped to reveal what had happened and that it accelerated their release from prison to rejoin their families.
A number of those accused lamented the fact that Hutu could not seek justice for crimes committed by the RPF. Many hesitated to openly discuss this question, in part because they were afraid of what might happen to them if they spoke to outside observers. Those who were willing to do so said they thought it was unfair that only certain crimes could be raised in gacaca and that the loss of their relatives at the hands of the RPF remained “unrecognized.” Others claimed the gacaca process had insincere aims and was designed to impose a sense of collective guilt on all Hutu. One individual described gacaca as a “means of targeting Hutu.” For these individuals, the gacaca process was not likely to break the cycle of impunity and had instead only caused more problems.
Commenting on gacaca’s contribution to reconciliation, the wife of one convicted man said: “Gacaca has left Hutu and Tutsi even more divided than before.” A number of interviewees agreed and spoke of increased tensions between the two ethnic groups. A few—far fewer than the number of genocide survivors—said that gacaca had helped relieve ethnic tensions and gave examples of individuals who were now able to greet or speak with each other.
The interviews conducted by Human Rights Watch suggest that many genocide survivors and persons accused of involvement in the genocide view gacaca as having had some success, notably in bringing to light new information about the genocide and in accelerating efforts to achieve justice. Interviewees disagreed on whether gacaca was the appropriate forum to resolve these cases, whether gacaca courts had operated fairly, and whether the sentences handed down were commensurate with the crimes. The largest variation of opinions came with respect to the issue of reconciliation.
Human Rights Watch did not carry out an in-depth study on the reconciliation aspect of gacaca. The above perspectives were gathered from conversations with Rwandans in the course of gathering information on gacaca trials and other human rights research in Rwanda. Nonetheless, three main conclusions can be drawn from these interviews. First, justice alone may not bring reconciliation and may only be one step in a much longer and more complex process. Gacaca may have placed Rwandans on the path to reconciliation, at least superficially, by allowing them to live together in relative peace and to greet each other or exchange a few words, but–unsurprisingly just 17 years after the genocide–there is still distrust within communities between the two main ethnic groups.
Second, gacaca has reopened certain wounds and reinforced ethnic divisions. The government has effectively banned public mention of the words “Hutu” or “Tutsi,” in an attempt to allay ethnic tensions and reinforce the notion of “one Rwanda,” but gacaca has reinforced alternative labels along ethnic lines: that of “victim” and “perpetrator.” Only Tutsi can be victims in gacaca and generally only Hutu can be perpetrators.Gacaca courts’ failure to give redress to all victims, Tutsi and Hutu alike, has caused bitterness for some Rwandans and has led to increased tensions in certain communities. Third, reconciliation in Rwanda is more about “cohabitation,” or peaceful co-existence as a matter of daily necessity, than genuine forgiveness that comes from the hearts of genocide survivors.
Gacaca would not have been possible without the significant support of international donors. Belgium, the Netherlands, and the European Union (EU) have been the largest funders over the past 10 years. Austria and Switzerland also contributed to the process.
Human Rights Watch could not ascertain the motivations behind foreign donors’ decision to finance gacaca, in part because many of the representatives of donors involved in the initial policy decisions no longer work on Rwanda and could not be reached for comment. Investing in the gacaca process was a risky decision, as one academic wrote in a 2000 report prepared for the Belgian government as it contemplated funding gacaca:
How to decide on a policy towards the gacaca proposal?....[I]t is clear that the proposal is simultaneously extremely promising and very dangerous; long thought-out and full of uncertainty; locally owned and weakly socially implanted; containing the seeds of reconciliation and potentially leading to increased conflict; preparing a decrease in the (current) prison population while possibly leading to increases in new detainees. There is no way to be sure of anything: it is a giant bet for the Rwandan authorities and population, as it would be for any donor supporting it (with th[e] difference that for donors it is not a matter of life and death, whereas for the Rwandans it is).
Despite the inherent risks, a number of European countries made the choice to support the gacaca process.
Belgium was the earliest and largest contributor to gacaca, giving approximately €8.1 million (approximately US$11.3 million) to the SNJG between 2000 and 2008. Most of this funding went toward training gacaca judges and providing logistical support, including tables, chairs, notebooks, and sashes for the judges. Belgium also supported initiatives to improve the standard of living of gacaca judges, conducting a study to determine judges’ needs and later providing radios and a one-time payment of 4,300 Rwandan francs (approximately US$7) each to all judges. Belgian funding facilitated the purchase of one bicycle for each gacaca jurisdiction in 2008. In addition to funding the SNJG, Belgium provided €1.5 million (approximately US$2.1 million) each year, for a total of €12 million (approximately US$16.8 million), to NGOs monitoring the gacaca process and the justice system more generally—the two main organizations being ASF and PRI.
The Netherlands has been another important contributor to gacaca, providing more than €5 million (US$7 million) to gacaca between 2002 and 2009. A significant portion of this funding was channeled through a basket fund, joined by Switzerland and Austria, which provided technical assistance to the SNJG. The main assistance offered through the basket-fund, or Bureau d’Appui Technique (“Office of Technical Support”), consisted of training gacaca judges. Switzerland and Austria, as discussed below, dropped out of the project after the pilot phase ended in 2005, but the Netherlands continued its support for training judges in later years. The Dutch government contributed significantly to an initiative to train judges in 2008 when rape cases were transferred to gacaca.
The EU contributed approximately €3 million (US$4.2 million) to gacaca between 2002 and 2009. The funding went directly to the SNJG and was used mainly for training judges and publication of the Inkiko-Gacaca newsletter—a government initiative to report on gacaca activities. Like other donors, the EU funded NGOs involved in the gacaca process, from those monitoring trials to those providing psychological counseling to rape victims whose cases would be decided by gacaca. In 2010, the EU stopped funding specific projects and instead moved to sector budget support—providing financial backing to the justice sector as a whole and allowing the Rwandan government to determine how the money was to be spent.
Austria provided €1.2 million (US$1.68 million) to the gacaca process between 2002 and 2010, with funds initially directed to the basket-fund supporting the Bureau d’Appui Technique and later to SNJG directly for the establishment of an audio-visual documentation center.Austria devoted an additional €570,000 (approximately US$796,000) over the same period to civil society groups monitoring the gacaca process.
Switzerland contributed one million Swiss Francs (approximately US$1.11 million) to gacaca during the 2002-2004 pilot phase. It withdrew its support before trials began nationwide in 2005, however, after concluding that the process appeared to be aggravating social tensions. Switzerland tried to redirect its funding towards reform of certain problems in gacaca, but the Rwandan government was not receptive to the proposal and no further funding was provided. Switzerland continued to fund at least one NGO monitoring gacaca trials until 2008.
Some diplomats have effectively raised individual cases where miscarriages of justice occurred and, at times, urged the SNJG to be more transparent in providing information on the number of cases pending and judged. Diplomats often relied on NGOs monitoring the gacaca process to alert them to problematic cases but also occasionally sent local embassy staff to monitor particular trials.
Donors rarely used their influence, however, to address the more fundamental and systemic problems described in this report. Given the extent of their financial and political support for the judicial system, donors should have used their position to insist on the incorporation of certain minimum standards for gacaca trials and to press the Rwandan government for corrective action to end corruption and abuse of the gacaca process for personal or political ends.
Rwanda has faced enormous challenges in the aftermath of the genocide. There was never going to be an easy solution for dealing with the hundreds of thousands of genocide-related cases within a reasonable timeframe. The Rwandan government’s decision to consider gacaca was not an unreasonable way to offer some form of justice for the genocide. It had the advantages of expeditiously bringing suspects to justice and providing redress to the victims, while also reducing the prison population.
However, as documented in this report, the compromises made in adapting the customary community-based practice to try grave criminal offenses led to significant due process violations being built into the system and a degree of disappointment on the part of many Rwandans.
A number of compromises may have been inevitable in the context of gacaca, but certain fundamental rights—such as the right not to be prosecuted twice for the same crime, and the right to be informed of the charges with enough specificity and adequate time to prepare and present a defense (including through defense witnesses)—should have been better protected. Absent trained legal professionals to assist the parties or to weigh the evidence and decide cases, the protection of these rights was even more important to ensure fair trials.
The government did not provide gacaca judges with adequate training and legal guidance, despite the complexity of the criminal concepts that they would need to address. Nor did it pay them for their work. With judges elected by their local communities, it was eminently foreseeable that it would be difficult, if not impossible, for many to prevent their own perspective of the genocide, their relationship with community members, and their own economic interests from interfering with their decision-making. A stronger and more robust legal framework was needed to ensure judges’ impartiality and to insist upon reasoned and fact-based judgments.
Similarly, more safeguards were necessary to prevent private individuals and government officials from misusing gacaca proceedings to serve their own narrow interests. Gacaca’s informal nature and dependence on local actors, many of whom had their own agendas, meant that accused persons, genocide survivors, influential community members, judges, and state agents all exerted undue influence on the gacaca process at times.
Gacaca unfolded differently across the country and evolved over the years, in part because of changes to the law and in part due to local variation on the ground. The government made certain improvements to the process, such as abolishing the death penalty and allowing convicts to suspend portions of their prison sentence and serve community service first. At times, the SNJG responded positively to reports of irregularities—for example, providing legal guidance to gacaca judges or directing remedial action in particular cases. Yet these instances were sporadic and inconsistent. The SNJG’s inability or unwillingness to effectively monitor and remedy problems in the gacaca system as a whole stemmed from inadequate resources, a lack of political will, and a failure to proactively monitor cases and listen to local communities and outside observers about worrying trends that developed countrywide. By 2008, and perhaps even earlier according to some, the SNJG had become unresponsive to many NGOs’ and donors’ expression of concern about the extent and scope of irregularities. Its failure to respond to increasing reports of misuse of gacaca for personal and political ends was particularly serious.
However, gacaca’s structural and systemic weaknesses that compromised its suitability to provide fair and impartial trials have been most seriously compounded by the prevailing political climate in the country and the restrictions on free speech. The government’s campaign against “divisionism” and “genocide ideology” has had a chilling effect on Rwandans’ ability and willingness to express themselves. This has been particularly detrimental in the context of gacaca: it has sometimes prevented members of local communities from speaking freely about what they saw in 1994 and has made them fearful of the repercussions of testifying in defense of individuals accused of genocide. Rwandans have come to realize that any statement given as part of gacaca can have negative repercussions for them, and many individuals with relevant information chose to remain silent. While only a handful of individuals who testified in gacaca were later formally charged with “genocide ideology,” “divisionism,” or minimization of the genocide, many more were accused of perjury or complicity in the genocide as a result of their testimony—most often when they defended accused persons.
The government’s decision to remove crimes committed by the RPF in 1994 from gacaca courts’ jurisdiction—which meant that some victims would never see justice through the community-based courts or even be recognized as victims—also limited gacaca’s potential to foster long-term reconciliation.
As the government seeks to bring gacaca to a close, it has recognized that certain miscarriages of justice must be corrected and has begun formulating a process to review these cases. Some justice officials have been candid, both in public and in interviews with Human Rights Watch, on the importance of this last stage in securing the legacy of gacaca.
Human Rights Watch agrees that this step is of critical importance. Yet the proposal to have cases which have been identified as potential miscarriages of justice reheard in gacaca, where they first occurred, risks repeating many of the problems outlined in this report. A more judicious option would be for the government to establish a special unit within the conventional justice system to assume this role. The unit, which could be located within the Supreme Court, would encompass a two-part review mechanism. First, it would receive appeals from accused persons who claim to have suffered miscarriages of justice or serious due process violations, and would provide an initial screening of these appeals in accordance with certain pre-determined legal criteria. In order to limit the number of cases, a review would only be accorded to individuals serving (or facing future) custodial sentences in prison or community service programs. Second, the unit would pass on cases which appear to be well-founded to specialized review panels, headed by professional judges (not gacaca judges) or other legal professionals. While the number of applications received by the unit might be in the thousands, the initial screening process would determine those cases meriting review.
The specialized review panels—consisting of several persons—would consider a range of sources of information on the case, including the written record from the gacaca court that handled the case and written submissions from the parties. Where necessary, the review panels could hold short hearings with the parties, hear additional important witness testimony, or request supplemental documentation. With the information gathered, the review panels could issue a final decision, either affirming judgments previously handed down in gacaca or revising judgments (and sentences) where miscarriages of justice are found to have occurred.
This will not be an easy task, and it will require financial support and technical assistance from foreign donors. Human Rights Watch believes this would be a worthwhile investment in maximizing the full potential of gacaca, securing fair and impartial justice for the genocide, and strengthening the Rwandan justice system in the longer term.
Minister of Justice
Ministry of Justice
P.O. Box 160
March 30, 2011
Dear Minister Karugarama,
Re: Forthcoming Human Rights Watch Report on Gacaca
I am writing to inform you that Human Rights Watch will be publishing a report on gacaca in the coming months. Based on several years of research and first-hand observation of gacaca trials, the report will analyze the gacaca process from a human rights perspective, outline achievements and areas of concern, and make recommendations aimed at strengthening the justice system in Rwanda.
We would like to ensure that the Rwandan government’s perspective is reflected in our report. The report already includes information provided to Human Rights Watch by officials in the Ministry of Justice and the National Service of Gacaca Jurisdictions (SNJG) over the last several years. However, to ensure that the report is comprehensive and accurate, we would appreciate your response to our main findings and recommendations (summarized below) by April 29, 2011. We are also writing to the SNJG to seek their point of view.
Aside from these specific issues, we would welcome any additional updates ongacaca and on plans for further reform of the Rwandan justice system relating to future prosecutions for genocide and genocide-related crimes.
Summary of findings
The forthcoming Human Rights Watch report acknowledges the enormous challenges the Rwandan government faced in choosing the gacaca system to process such a large number of genocide cases. It notes some of gacaca’s main achievements, including the swift work of the courts, the extensive involvement of local communities in testifying to events which took place during the genocide, and the opportunity gacaca provided to genocide survivors to learn what had happened to their relatives.
The report also notes a number of human rights concerns, as well as irregularities and violations of due process – all of which have resulted in certain compromises in the delivery of justice for the genocide.
Our concerns relate primarily to the absence of fair trial safeguards and limitations on the ability of accused persons to defend themselves effectively. Human Rights Watch documented, among other things, cases where:
- the presumption of innocence was undermined;
- the accused were not provided with adequate information on the charges against them in advance of their trial;
- the accused did not have sufficient time to prepare a defense; and
- individuals were tried twice for the same offenses, for example first in a conventional court, then in gacaca, or twice by different gacaca jurisdictions.
The report also documents:
- the misuse of gacaca by private individuals as a way of settling scores or resolving personal grievances unrelated to the genocide;
- political interference in a number of trials, particularly those of individuals viewed as government critics;
- corruption of judges, accentuated by the lack of remuneration;
- intimidation of witnesses;
- obstacles to witnesses testifying freely in gacaca hearings;
- the broader political climate in Rwanda which has further discouraged many people from speaking out in gacaca trials for fear of repercussions. Repercussions for potential defense witnesses may include arbitrary arrest, accusations of perjury, charges of “genocide ideology” or charges of complicity in the genocide.
Other concerns relate to some of the fundamental premises of gacaca from the outset, for example the lack of professional legal training for judges and the absence of lawyers for the accused. The lack of professional legal training for judges, in particular, has resulted in divergent practices in different gacaca jurisdictions, for example in standards of proof, decision-making, and sentencing, as well as attitudes towards the accused, the civil parties, and members of local communities participating in gacaca trials.
The report will include individual case studies and examples from across the country to illustrate these and other patterns observed during gacaca trials.
We would appreciate your response to the concerns summarized above, and in particular, any information on action taken to remedy these problems.
Like all Human Rights Watch reports, our report on gacaca will make a number of recommendations to the Rwandan government and justice officials, as well as to donors to help strengthen the justice system following the closure of gacaca.
The closure of gacaca and mechanisms for additional review
Human Rights Watch believes that the fair and impartial handling of outstanding cases, following the closure of gacaca, will be of paramount importance to the legacy of gacaca and to strengthening the Rwandan justice system in the longer term.
We understand that the majority of gacaca jurisdictions have closed, but that a small number of cases are still under consideration. Could you confirm how many cases are still open, at what stage they currently stand, and by when you expect them to be resolved?
We understand that the current plan is for the SNJG, in consultation with other state institutions, to review cases in which serious irregularities or miscarriages of justice are alleged to have occurred and to determine whether these cases should benefit from additional review. Could you provide information on the specific criteria for review and the process or mechanism through which they will be reviewed?
Human Rights Watch will propose the creation of a specialized unit within the conventional justice system, for example within the Supreme Court, to review appeals from individuals who claim to have suffered miscarriages of justice or serious procedural violations in gacaca. We would recommend that the review be conducted by professional judges (not gacaca judges) or other trained legal professionals, and that precise criteria be established for prioritizing the cases to be reviewed. For example, they may prioritize review of appeals for individuals still serving (or facing future) custodial sentences.
We welcome your comments on this proposal.
Future prosecutions of genocide or genocide-related cases
We understand that a new bill is currently under consideration concerning the prosecution of genocide and genocide-related cases after the completion of the gacaca process. Justice officials have informed Human Rights Watch that any new cases would be brought before the national courts.
We would appreciate information on action taken, or planned, in the following areas:
- ensuring that any new allegations of participation in the genocide are properly reviewed by trained prosecutors and judges before a person is prosecuted in the conventional courts;
- rectifying violations of double jeopardy, to ensure that no one is prosecuted twice for the same crime, and reviewing all convictions where a person was tried both before a gacaca jurisdiction and a conventional court, or in two or more different gacaca jurisdictions for the same offenses.
To ensure that our report reflects measures which may already be underway, we would be grateful if you could inform us of progress in terms of government action in the following areas:
- measures taken to ensure that state agents do not interfere in gacaca or conventional court trials and do not attempt to influence decision-making;
- measures taken to ensure that all police officers and state agents refrain from conducting unlawful arrests and detention, and any prosecutions or disciplinary action taken against individuals responsible for such conduct;
- measures taken to compensate individuals who have been unlawfully arrested and detained;
- the prosecution of individuals who have falsely accused others;
- proposals to convert any remaining prison time for convicts who have satisfactorily completed community service (“TIG”) to a suspended sentence; and
- revision of the 2008 law on “genocide ideology” announced in 2010.
Thank you in advance for your responses to these questions and any additional information you are willing to share with us. As mentioned above, we would appreciate a response by April 29, 2011 to enable us to incorporate any new information you may provide in our report.
You can reach me at firstname.lastname@example.org
This report was researched and written by Leslie Haskell, Rwanda Researcher at Human Rights Watch, and contains information gathered by several local gacaca observers and previous Human Rights Watch researchers. The report was edited by Carina Tertsakian, Senior Researcher, and by Rona Peligal, Deputy Africa Director. Valuable feedback was provided on an earlier draft of this report by Sara Darehshori, Senior Counsel in the International Justice Program, Lars Waldorf, former Human Rights Watch researcher in Rwanda and senior lecturer at the Centre for Applied Human Rights (University of York), and Zarir Merat, former head of mission at Avocats Sans Frontières in Rwanda. It was reviewed by Aisling Reidy, Senior Legal Advisor, and Babatunde Olugboji, Deputy Program Director. Additional assistance was provided by Rachel Nicholson, Lianna Merner, Grace Choi, Anna Lopriore, and Fitzroy Hepkins. Danielle Serres, with the assistance of Simon Marrero, translated the report into French. Peter Huvos, French website editor, vetted the translation.
This report would not have been possible without the support of Alison Des Forges, Senior Advisor in the Africa Division, who died tragically in February 2009, and her husband Roger Des Forges.
Human Rights Watch would like to thank the hundreds of Rwandans who shared their personal experiences and views of gacaca with us over the past 10 years. Some individuals were fearful of the consequences of speaking with Human Rights Watch but still came forward courageously to provide their stories. We would also like to thank the Minister of Justice, the SNJG Executive Secretary, and the many other justice officials who agreed to be interviewed and who provided valuable information.