Abstract
More than one million people participated in the 1994 genocide against the Rwandan Tutsi. How did Rwanda, whose criminal justice infrastructure was decimated by the genocide, attempt to bring the perpetrators to justice? In this article, we provide the first analysis of the outcomes of the gacaca courts, a traditional community-based justice system that was greatly modified to address crimes of genocide. After briefly reviewing the creation of the National Service of Gacaca Jurisdictions, we explain the court process. Then, we present an overview of the outcomes of the courts with a focus on the specific sanctions given to those found guilty. This article provides the first systematic analysis of these sanctions, contributing both an empirical overview and new insights into how Rwanda attempted to bring justice to the many citizens who took part in the genocidal violence. We conclude by briefly highlighting some successes and failures of the gacaca system and its broader lessons for justice in other contexts.
On April 6, 1994, the plane carrying the President of Rwanda and the President of neighboring Burundi was shot down as it prepared to land in the capital of Rwanda. The crash killed the plane’s occupants immediately, and within hours, targeted killing of Tutsis and those associated with them began. Members of the government had deliberately engendered animosity between the two main ethnic groups—Hutu and Tutsi—which had become politicized and polarized during Belgian colonialism. Just a few months later, over 1,050,000 people were dead (Center for Conflict Management of the National University of Rwanda, 2012). To put this number in perspective, these 1,050,000 people comprised approximately 14% of Rwanda’s 7.5 million 1 inhabitants, a rate of killing 2,800 times higher than the current U.S. homicide rate of 5 per 100,000.
Government leaders were largely responsible for the planning and execution of genocide. Yet, it is also well documented that priests, doctors, nurses, judges, and even human rights activists took part in the violence by murdering their neighbors, looting houses, destroying property, and raping women (Center for Conflict Management of the National University of Rwanda, 2012; Fujii, 2009; Hatzfeld, 2008; Mamdani, 2001; Straus, 2006). Overall, more than one million perpetrators participated in the genocide; and as new government leaders debated how to rebuild the country, an important part of the conversation concerned holding these participants accountable.
How has Rwanda, whose criminal justice system was destroyed by the genocide, attempted to bring the massive number of perpetrators to justice? This article offers a first look at some of the outcomes of the gacaca courts, a traditional community-based justice system that was modified to address crimes of genocide. We briefly explain the creation of the National Service of Gacaca Jurisdictions before discussing how these courts functioned in law and in practice. Next, we draw on court records to present an overview of the outcomes of these courts. We focus particularly on the sanctions given to different groups of perpetrators, which have yet to be systematically analyzed. Finally, we briefly place the successes and failures of the gacaca courts in context, discussing the system’s lessons for transitional justice and other settings.
Transitional Justice: Responding to Genocide
Due in part to the magnitude and severity of the crimes, transitional justice responses to genocide often occur at both the local and international levels. Internationally, formal trials have only been legitimized as appropriate responses to genocide and similar atrocity crimes since World War II. While prosecution was considered after the Armenian genocide and for war crimes committed during World War I (Bass, 2000; Heberer & Matthaus, 2008), the 1945 Nuremberg trials are widely recognized as the key precedent for international trials for mass atrocity. 2 After the Nuremberg Tribunal closed, members of the newly formed United Nations began discussing the possibility of a permanent international tribunal. Due in part to Cold War rivalries, however, that court (known today as the International Criminal Court) was not yet in existence when genocide occurred in Rwanda.
Instead, in November 1994, the United Nations Security Council passed a resolution that created the International Criminal Tribunal for Rwanda (ICTR). The mandate of this ad hoc tribunal was to try the Rwandans deemed most responsible for the genocide. The ICTR held its first trial in Arusha, Tanzania, in 1997 and handed down the world’s first conviction by an international court for the crime of genocide 1 year later. As of 2014, the tribunal has completed 75 cases (ICTR, 2013b).
While the ICTR was heralded as a milestone in international criminal justice, it lacked both the mandate and the capacity to try even a fraction of those who participated in the genocide. Instead, Rwandan leaders were faced with the task of deciding how to hold people accountable while simultaneously facilitating reconciliation. The mass public participation in the violence, as well as the country’s traditional focus on restorative justice, led many observers to recommend forming a truth commission (Clark, 2010). 3
In fact, at the insistence of international donors and non-governmental organizations (NGOs), a truth commission operated in Rwanda just 1 year before the genocide to investigate alleged human rights violations by the government (Bornkamm, 2012). Rwandan leaders considered forming another truth commission after the genocide but ultimately rejected this approach on the grounds that it would not adequately punish those who had taken part in the violence (Clark, 2010). Thus, while some of the key orchestrators of the genocide would be tried at the ICTR, a mechanism was still needed to respond to the violence within Rwanda.
Rwanda’s Response: The
Gacaca
Courts
As discussions about a large-scale response to the genocide continued, the new Rwandan government—comprised mainly of members of the Rwandan Patriotic Front (RPF) rebel army that stopped the genocide—was beginning to punish suspected perpetrators. In 1994 and 1995, government officials searched for people suspected of participating in the genocide and transported them to Rwandan prisons (Bornkamm, 2012; Clark, 2010). An estimated 120,000 people were brought to facilities that were built to hold 45,000 people. The 19 prisons in Rwanda were thus operating at over 200% capacity, with the vast majority of the incarcerated awaiting genocide trials (International Center for Prison Studies, 2013).
Just as the prison infrastructure was inadequate to handle the sheer number of genocide participants, the existing judicial system was ill-suited to manage the newly incarcerated suspects or the many other suspected genocidaires who were not in prison. Moreover, many lawyers and judges had been killed during the genocide or had fled the country. In November 1994, there were only 12 prosecutors and 244 judges in Rwanda, compared to 70 prosecutors and 758 judges in Rwanda before 1994 (Gacaca Report Summary, 2012). Even at full capacity, however, the existing Rwandan justice infrastructure would have been straining at the seams in responding to claims against over one eighth of the nation’s population.
In response, multiple international NGOs donated money, time, and other resources to help the Rwandan national courts train new judges and lawyers. Many new legal personnel were indeed trained (estimates in 1996 included 210 prosecutors and 841 judges), and the national court system began hearing cases related to the genocide (Gacaca Report Summary, 2012). Yet, by 2000, only a few thousand cases had been processed (Bornkamm, 2012; Clark, 2010). 4
The government of Rwanda thus turned to a different mechanism that combined retributive and restorative justice in 2001—the gacaca courts. Gacaca courts date back to before Rwanda was colonized (Bornkamm, 2012). Gacaca means “grass” in Kinyarwanda, and, as the name implies, the hearings were traditionally held outside in empty markets, school yards, and other public places within each community. In an ideal hearing, defendants would confess their crimes, express remorse, ask for forgiveness, provide restitution, and then offer food and drink to all parties as a symbol of reconciliation (Vandeginste, 1999). This traditional method of dispute resolution, which bears a close correspondence to the principles of restorative justice (Drumbl, 2002), had been established as an officially sanctioned court system for common crimes since the 1940s.
After the genocide, however, the institution changed significantly (for detailed information, see Bornkamm, 2012; Clark, 2010; Jones, 2009), to the dismay of critics who lament the shift from a community-based informal system to a formal institution with close ties to the state (Waldorf, 2006). Government leaders decided that the modified gacaca courts, called inkiko gacaca but shortened here to gacaca, were to be organized at local levels throughout the country. Its goals would be simultaneously punitive (e.g., fighting impunity) and restorative (e.g., contributing to national reconciliation), 5 and the courts would have jurisdiction over genocide and crimes against humanity committed between October 1, 1990, and December 31, 1994. 6
As there were too few lawyers or judges for the trials, and as the courts were intended as community courts, lay members of the community would serve as judges (inyangamugayo). In October 2001, the first round of elections for gacaca judges began. Legal training was not required to serve as a judge. Instead, judges were selected on the basis of their commitment to justice, truth, and a “spirit of sharing.” They were required to be 21 years or older, have no previous criminal convictions, and could not have been serving in a government or political leadership role. They also could have never been suspected of committing crimes against people during the genocide (Organic Law, 40/2000). More than 250,000 male and female judges were elected, and in April 2002, the elected judges underwent training. Finally, on June 18, 2002, the first pilot phase of the gacaca began. 7
Court Procedure
The judgment phase of the courts began in March 2005. Suspects were divided into three categories (Organic Law, 13/2008), as defined in the Appendix. In brief, Category 1 was reserved for planners or organizers of the genocide, officials, and leaders who participated or incited others to participate, and those who committed rape and sexual torture. Category 2 included “notorious murderers,” those who tortured others or defiled their bodies, suspects who killed or intended to kill, and those who served as accomplices in such acts. Finally, Category 3 was comprised of property offenders who had not yet come to settlement with victims or authorities before the law took effect.
Courts were operational at both the cell (akagari) and sector (umurenge) levels of geographic administration. Those at the cell level (the smallest administrative region in Rwanda) were responsible for trying people who were accused of Category 3 crimes, while courts at the sector level were responsible for Categories 1 and 2. 8 Nationwide, there were 9,013 cell courts, 1,545 sector courts, and 1,545 courts of appeal (National Service of Gacaca Jurisdictions, 2012). Each level of court consisted of a general assembly, a bench of judges (originally 19, later reduced to 14, 9, and finally 5), a president, and a coordinating committee. Lawyers did not participate in the trials, ostensibly to avoid adversarial proceedings, and there were no prosecutorial teams. Instead, the judges were tasked with investigating the crimes and then conducting the trials, which lasted an average of 1 to 3 days (Mukantaganzwa, 2012).
Trials took place in classrooms, conference rooms, or outdoors when indoor locations were not available. At each trial, inyangamugayo (judges) sat together on the bench, wearing green, yellow, and blue sashes—the colors of the Rwandan flag. The victims sat between the inyangamugayo and the public, with the perpetrators on the other side. Public confession was a cornerstone of the trials and their capacity to simultaneously serve the punitive and restorative goals of justice. As such, trials were open to the adult Rwandans in every community, whose participation was seen as key in moving the nation past the atrocity. In fact, participation was a duty (Organic Law n°16/2004, Article 29), and there are records of fines for non-participation (Sommers, 2012).
The gacaca laws based punishments on the category of crime, recognition of guilt, and time to confession (if any). The most severe punishment that judges could assign was lifetime in prison. This was reserved for those committing Category 1 crimes without confession before or during the trials (Organic Law, 13/2008). Sentences were then to be decreased depending on the severity of the crime and time of confession (with lighter sentences for those who confessed before rather than during their trial). Community service, fines, and other restoration-based reparations were also possible sanctions, which we discuss in more detail below.
Special procedures were put in place for minors. Rwandans who were between the ages of 14 and 18 at the time of the offense could be tried and receive up to 20 years in prison for a Category 1 crime. Those who were less than 14 at the time of the genocide were not to be prosecuted but sent to training camps instead (Organic Law, 13/2008). With regard to appeals, if defendants, victims, or others were not satisfied with a judgment, they were to register their intention to appeal within 15 days. The court then assessed the grounds of the appeals claim based on the gacaca laws.
Assessing the Courts
The courts closed on June 18, 2012, after completing the vast majority of the cases brought to their attention. To date, research on the gacaca courts has focused on scholars’ and practitioners’ assessments of the orientations and operations of the courts rather than case outcomes. Indeed, as the gacaca were heralded as a new form of transitional justice that uniquely combined mechanisms of punitive and restorative justice, diverse scholars sought to document how the courts functioned in law and in practice (Bornkamm, 2012; Clark, 2010; Daly, 2002; Longman, 2009; Schabas, 2005).
In the course of this documentation, divergent views emerged regarding the purpose of the gacaca. For many practitioners and advocacy groups, the gacaca were primarily a retributive-based institution that existed to punish perpetrators of genocide. Others, such as Karekezi (see, for example, Karekezi, Nshimiyimana, & Mutamba, 2004) and Drumbl (2000), argue that a view of the gacaca as strictly retributive is inherently limiting. Rather, in combining punishment and reconciliation, the gacaca blend elements of retributive and restorative justice. Clark (2010) similarly views the gacaca’s goals as both punitive and restorative, though he moves beyond this binary distinction in proposing six ideal aims—truth, peace, justice, healing, forgiveness, and reconciliation—along with three more pragmatic objectives of processing the backlog of cases, improving living conditions in the prisons, and facilitating economic development.
Beyond documenting the process and objectives of the gacaca, scholars have assessed Rwandan public opinion of the courts, with many finding generally favorable public sentiment (Gasibirege & Babalola, 2001; Longman, Pham, & Weinstein, 2004; Rettig, 2008). 9 The courts have also sustained harsh criticism, however, as scholars and practitioners have critiqued the implementation of gacaca and the institution more generally (Apuuli, 2009; Fierens, 2005; Ingelaere, 2009; Rettig, 2008; Staub, 2004; Waldorf, 2006). These commentaries address a range of gacaca processes, including but not limited to the traumatic effects of testifying, the retroactive application of laws, the lack of due process rights for the accused, the reduced role of local communities in the creation of the courts, the inclusion of crimes against property, and the lack of trials against members of the RPF, the ruling party who assumed power after the genocide and who were accused of committing crimes in the process of stopping the genocide and in its immediate aftermath. Notably, the Rwandan government also echoes some of these concerns. In fact, the closing report of the National Service of Gacaca Jurisdictions listed many difficulties impeding the functioning of the courts, such as responding when gacaca judges and then-current government leaders were accused of genocide, conspiracy among both the accused and the witnesses to conceal information, destruction of court records, violence against those who testified, trauma manifested during (and likely after) court proceedings, accused perpetrators who fled Rwanda, and corruption within the court system.
These critiques identify important problems, especially when measured against an idealized liberal legal model. When measured against the post-genocide Rwandan context, however, many of the legalistic critiques of the courts seem disproportionate, given the scale of crimes and the lack of other options to respond to them (McEvoy, 2007). Rwanda had just lost more than one million people to a genocide that its own citizens had planned and perpetrated, its institutions were completely shattered, and its pre-genocide legal system already differed from Western legal systems in numerous ways. With both context and critiques in mind, we here detail a form of justice the gacaca produced—the sanctions given to over one million perpetrators.
Analytic Strategy
To detail these sanctions, we draw on official records that were kept by the National Service of Gacaca Jurisdictions. This unit was the administrative arm of the gacaca courts, and they compiled records from all trials. Millions of these records were kept in official notebooks, which are currently being archived in Kigali. This process will take years, though gacaca employees also transferred abbreviated forms of court records into Excel files that contain information about the perpetrators (including sex, date of birth, and family identifiers), whether they were found guilty, the punishment, the category of crime, the location of the crime, and the date of the trial. Through a partnership with the Rwandan National Commission for the Fight Against Genocide, we obtained the Excel files for each administrative region within Rwanda. We then compiled these abbreviated court records from 10,558 courts 10 into a single database of gacaca files with almost two million records. We caution that this is the first presentation of these data and it is likely that a small percentage of the almost two million cases in our Excel files contain clerical errors. Nevertheless, after taking care to clean these data and minimize such errors, we are confident about the data we present.
In this article, we focus specifically on the initial sanctions imposed at the cell and sector levels. To avoid confusion and duplication, we omit appeals (explained in more detail below) from the analysis. In addition, we exclude trials in absence, which involved those who had fled or, for some other reason, did not appear in court 30 days after they were summoned. These trials were conducted not for the alleged perpetrator but for the victims, their families, and the community, illustrating the restorative aspect of the trials. Yet, as the sanctions were likely influenced by the inability of the accused to confess or testify on their own behalf, we exclude them from this analysis.
We also exclude a relatively small percentage of cases that include errors. Official data in any country are never perfect, and Rwandan gacaca data are no exception. Therefore, while we strove to include as many cases as possible in this analysis, we exclude cases with errors, such as sanctions recorded incorrectly (with nonsensical numbers, words, or symbols). Thus, the total number of trials included in the analysis is 1,441,555. Note that the unit of analysis is the trial rather than the individual; a small proportion of individuals were tried for crimes in multiple categories and jurisdictions, whereas others were tried in separate cases for crimes falling into different categories.
We begin by detailing the total number of cases concluded by the gacaca system. Then, we restrict the analysis to our analytic sample as described above and detail the sanctions received in each category of crime. 11 Finally, we briefly examine how these sanctions varied by the age and sex of the perpetrators.
The Gacaca Courts: An Overview
Overall, the gacaca courts completed 1,958,634 cases (National Service of Gacaca Jurisdictions, 2012). Table 1 illustrates the case breakdown by category of genocidal crime, showing that, by far, the majority of the cases were crimes against property (Category 3). Thirty-three percent of the cases involved crimes against people, with 3% of all cases falling in Category 1, the category reserved for the cases deemed most serious. Table 1 also includes the percentage of individuals found guilty in each category. The likelihood of a guilty verdict was highest for property crimes in Category 3 (96%), followed by those in Category 1 (88%). The high percentage of those convicted in Category 3 may be explained in part by the relatively modest sanctions, as detailed below. In addition, the discrepancy between those convicted in Category 1 as opposed to Category 2 (63%) may be due in part to the percentage of confessions in each category (approximately 41% and 30%, respectively) or the notoriety of the perpetrators.
Gacaca Case Completion.
Approximately 9% of the total cases tried by the gacaca were appeals cases. As the table illustrates, the vast majority of these cases were for crimes in Category 2, which included crimes against people, such as killing. Twenty-six percent (45,839) of all appeals cases were acquitted. As noted above, we exclude these cases, as well as cases tried in absentia and cases with erroneous data, resulting in an analytic database of 1,441,555 trials.
Category 3 Sanctions: Crimes Against Property
Table 2 shows the distribution of sanctions for Category 3 crimes—crimes against property—that were tried at the cell (akagari) level. The overwhelming majority of perpetrators in this category were sanctioned via fines that were to be paid to victims and their families. The median fine amount was 7,100 Rwandan Francs (RWF), which corresponds to approximately US$11 at the current exchange rates (649 RWF per U.S. dollar) and a relatively modest percentage of Rwanda’s per capita gross national income of US$560 (World Bank, 2013, data for 2011).
Frequency Distribution of Category 3 Sanctions.
Precise fine amount information is missing for approximately 1% of the cases.
Gacaca law stipulated that prison sentences were not to be given for Category 3 crimes. It is unclear whether the .01% of prison sentences in the database were errors in sentencing or errors in recording the sentence.
Approximately half of the fines were between 1,285 and 25,000 RWF, or US$2 to US$39. As shown in Figure 1, almost one fourth of the fines were less than 1,000 RWF, and another third were below 10,000 RWF, likely reflecting the economic circumstances of the perpetrators. At the other end of the distribution, 9% of fines exceeded 100,000 RWF, or US$154 (and 1% of fines exceeded 680,953 RWF, or US$1,049). The distribution is skewed by a small number of fines over 1,000,000 RWF, which may not have been accurately recorded.
Fines for Category 3 perpetrators (in Rwandan Francs).
To assess the fine amount, the victim and/or family members would provide a list of the property that was destroyed. This list would be read aloud to the public during the hearing, and anyone present with additional information could provide it to the court. This was especially important when the entire family was killed; in such cases, the neighbors would generally assist the gacaca courts in assessing the damage to property and compiling the list. Then, if the suspects were found guilty of stealing or destroying the property in question, the gacaca court judges would decide how much was owed for the damaged or stolen property. The fines were intended to be paid to the family rather than the state. If no family member survived the genocide, the fines were held in a special account at the district level, to be paid to any relatives who came forward at a later date.
Beyond fines, over 9% of cases were settled by “agreement” (ubwumvikane). This represents a negotiated arrangement between the perpetrators and the victim or families who lost property, initiated by the perpetrator who wished to be forgiven and to pay the victim for the stolen or destroyed property. For example, an accused individual took four doors from a victim’s house during the genocide. By the time of the hearing, however, the accused had damaged the doors and could no longer find the same wood to replace them. If the accused recognized his crime and asked for forgiveness, he could ask the victim for an agreement. Together, the two parties would then choose another wood to replace the damaged doors. Once chosen, their agreement would then be signed before the gacaca court, with the court retaining the ultimate authority to accept or reject the agreement.
In addition, over 25,000 cases (2.3%) were resolved via gusonerwa or “exemption.” In such cases, both perpetrator(s) and victim(s) recognized the crime. The perpetrator requested exemption from all or part of the financial obligation, and the victim decided whether to grant this request. If the victim agreed, the perpetrator’s crimes would still be recognized as such, though no fine or other punishment would be assigned. For example, an individual was accused of looting two cows from a victim’s family. During the trial, he was found guilty and sentenced to pay 80,000 RWF to the victim’s family. He sold many of his belongings to try to come up with 80,000 RWF. After a few months, he visited the victim’s family. He brought 60,000 RWF—the most he could get at that time—and promised to bring the remaining money soon. After consultation, the family decided to forgive him and exempted him from paying the remaining 20,000 RWF.
Beyond this, about 1% of Category 3 cases were resolved through the return of goods. These included cows, pigs, sheep, metal sheets, clothing, glasses, basins, knives, coffee, beans, peas, pitchers, blankets, saucepans, and other animals, foods, and household goods. The remaining sanction categories listed were imposed less frequently but remain important for understanding the variety of approaches to restorative justice for crimes against property. For example, perpetrators could request “forgiveness” (imbabazi) for their crimes. This did not deny the crime’s occurrence or the suspect’s participation, since recognizing the crime is an important part of the gacaca process. Forgiveness was also an element of other sanction processes, such as exemption or agreement. Yet, when officially granted, the victim agreed not to pursue the perpetrator for damages. Finally, some of those found guilty were given the option to build a house or work for the victims. This served as a form of restitution and reconciliation, particularly in cases in which the guilty party lacked the means to pay fines or repay stolen goods.
Category 1 and Category 2 Sanctions: Crimes Against People
Cases involving crimes against people, such as killing, rape, and torture, were addressed through courts at the sector (umurenge) level, a higher administrative level within Rwanda. Nevertheless, they were designed to operate according to similar procedures and were governed by the same laws. Due to the greater severity of these crimes, the sentences for these categories were much more likely to involve incarceration.
In fact, almost all of the Category 1 and Category 2 perpetrators were given prison sentences, as shown in Table 3. Approximately 93% received a sentence varying from several months to 30 years. In addition, 5% received a life sentence. 12 There were also 329 agreements recorded in the data, which functioned as explained above for property offenses, and 100 sentences to training camps, where those sentenced took classes in Rwandan history and learned skills pertaining to agriculture, woodwork, and art, among others.
Frequency Distribution of Category 1 and Category 2 Sanctions.
Community service, a hallmark of the gacaca system and restorative justice more generally, remained an option for perpetrators in these categories who confessed to their crimes. Overall, 29% of cases (91,556 cases) involved a community service obligation. When perpetrators recognized their crimes and confessed, the sentence length could be halved through community service, with 50% of the sentenced time served in TIG (Travaux d’Interet General or “works of public interest”) rather than prison. Community service work generally depended on the needs of the region, often including tasks such as building (or rebuilding) homes for survivors, repairing roads and bridges, planting fauna, or creating terraces. This work was completed prior to the prison sentence.
Figure 2 shows the distribution of sentence length for these two types of genocide offenses. 13 Category 1 crimes are considered significantly more serious than Category 2 offenses, meriting longer prison sentences. The modal Category 1 sentence was 25 to 30 years, with 17% of the perpetrators receiving life sentences. Excluding the life sentences, the median sentence length was 19 years—or a year less than the presumptive minimum of 20 years. In Category 2, by contrast, the median sentence was 15 years, the modal interval was 10 to 14 years, and only 2% received life sentences. These distributions thus show a degree of proportionality in sentencing across this severity gradient. As a point of reference, 23% of U.S. homicide offenders were given life sentences in 2006. The remainder were sentenced for an average of 21 years (Rosenmerkel, Durose, & Farole, 2009, p. 6), significantly longer than sentences given to genocide perpetrators tried in the gacaca courts. While the United States remains an outlier in regard to sentence length, the sparing use of life sentences for gacaca Category 2 crimes also appears tough but fair by international standards (Bernaz, 2013).
Prison sentences for Category 1 and Category 2 perpetrators.
Figure 3 shows the distribution of sentence length in years for Category 1 and Category 2 crimes combined, as well as the number within each 5-year grouping that received community service. Notably, the most common prison sentences were 10 to 14 years in length, though more than one third of the individuals receiving these sentences were able to lessen their time in prison by doing community service. In addition, some prisoners were eligible to have up to one third of their sentence suspended. 14
Sentences and community service for Category 1 and Category 2 perpetrators.
Age and Sex Distribution of Sentences
We have only limited information about the demographic and social characteristics of perpetrators, but it is instructive to compare the sex and age distributions of those receiving punishments (for more information on the age and sex of perpetrators, see (Nyseth Brehm, Uggen, & Gasanabo, 2012). 15 In general, a greater percentage of men received fines for Category 3 crimes, with a median fine amount of 7,480 RWF for men and 5,000 RWF for women. Yet, for Categories 1 and 2, the median length of years in prison (15) was the same for men and women, and community service was given in a similar percentage of cases. With regard to age at the time of offense, the average fine was 8,155 RWF for people at or below the median age of 32, relative to 7,435 RWF for people 33 and above. In addition, the median length of prison sentences was 12 years for those less than age 33 and 13 years for those over age 33 and above.
Figure 4 shows life sentences by age and sex. As discussed, life sentences were the most severe punishments given by the courts. Overall, 17% of men convicted of Category 1 crimes were given life sentences, relative to 7% of women in this category. In Category 2, men are also more likely to receive life sentences, which again could reflect differences in severity or other characteristics of the case or the court. About 16% of perpetrators below the median age of 32 received life sentences for Category 1 crimes, relative to 11% of their older counterparts. There is little difference in the age distribution of life sentences for Category 2 crimes.
Life sentences by age and sex.
Overall, these differences are small, with more difference seen between sexes and in Category 3 crimes. As there are no similar studies of sanctions for perpetrators of other episodes of genocide, we are unable to compare these discrepancies with other cases. Moreover, we lack the sort of detailed information about crime severity that would allow us to analyze discrimination on the basis of sex and age. Nevertheless, age and sex are likely linked to both crime severity and sentencing outcomes (Steffensmeier, Ulmer, & Kramer, 1998). Discrepancies in fines may also reflect differences in the ability to pay financial obligations. Finally, there may have been gender- or age-specific differences in confessions (which, again, influenced the sanctions given) or differences in sentence assignment by the judges. As more complete gacaca data become available, a full accounting of the relative influence of such factors may be undertaken.
Conclusion: Gacaca in Perspective
In the past decade, the National Service of Gacaca Jurisdictions processed an astounding number of cases in an attempt to bring perpetrators of genocide to justice and to promote reconciliation within Rwanda. This system represents a complex amalgamation of seemingly disparate elements. It is at once formal and informal, community-based and state-driven, traditional and contemporary, and punitive and restorative. The underlying tensions within the gacaca model, combined with the enormity of the courts’ responsibilities, have led to sustained criticism with regard to both process and outcomes (Apuuli, 2009; Fierens, 2005; Ingelaere, 2009; Rettig, 2008; Staub, 2004; Waldorf, 2006).
We do not seek to refute these criticisms. Nevertheless, the courts’ ability to try and sanction a crushing caseload of genocide perpetrators represents a truly historic achievement in the administration of justice. Whereas one cannot measure the quality of justice by the number of cases processed, it is important to reiterate that the gacaca courts did their difficult work in the wake of a mass tragedy that shattered Rwanda’s legal infrastructure and left the country reeling. To put these accomplishments in perspective, consider that the International Criminal Tribunal for Rwanda has completed 75 cases since 1997, with annual budgets over US$200 million (ICTR, 2013a). In contrast, the gacaca processed almost two million cases at an estimated total cost of US$46 to US$65 million (National Service of Gacaca Jurisdictions, 2012).
While punishment was not the only purpose or outcome of the courts, it is clearly an important aspect of the institution that merits scholarly attention and analysis. This article provides the first comprehensive overview of the sanctions given by the gacaca courts and the first attempt to organize and analyze these data. Although much work remains, we can draw three basic conclusions from the data presented here.
First, the sanctions of the gacaca courts reflect a mix of fines and restorative justice alternatives for genocide offenses against property, illustrating that restorative aims were indeed embedded in the institution. Second, the severity-graded prison sentences for genocide crimes against people provide evidence of proportionality and a more punitive approach toward crimes deemed more serious. While it is difficult to characterize any genocide sanction as proportional, the sentences appear to reflect the severity gradient intended by the basic categorization scheme. It is perhaps noteworthy that, on average, sentences for murder and non-negligent manslaughter in the United States are significantly longer. Third, while the gacaca courts made extensive use of life sentences and long-term imprisonment, they also made extensive use of community service (Travaux d’Interet General or “works of public interest”). Further research on these restorative efforts might explore their potential to serve tremendously disadvantaged communities while also alleviating prison overcrowding, costs, and a severely overtaxed prison infrastructure.
In short, the gacaca courts represented a powerful response to mass crime and an important element in the struggle to address society-wide tragedy and move forward. While these courts represent a “home-grown” Rwandan solution in many ways, their blending of punitive and restorative aims and traditional and contemporary elements holds important insights for justice pursuits around the world. To the extent that a hybrid model such as gacaca can address the “crime of crimes,” it should encourage other innovative approaches to lesser offenses in wide-ranging social contexts.
Footnotes
Appendix
Acknowledgements
We owe special thanks to Marie Berry, Suzy McElrath, Naphtal Kisakye, Lindsay Blahnik, and Andrew Johnson for valuable feedback.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the University of Minnesota Graduate Research Partnership Program. Uggen is additionally supported by a Robert Wood Johnson Health Investigator Award, whereas Nyseth Brehm is supported by the National Science Foundation.