Even before full-scale war erupted in 1996, the Zaire (country later renamed Democratic Republic of the Congo) was in crisis. President Joseph Mobutu’s kleptocratic rule and a failed transition to multiparty democracy had devastated the country’s economy. Violent clashes between migrant and indigenous communities flared up in North Kivu province in 1993.[i] In 1994-5, the mass influx of Hutu refugees as well as former Rwandan Armed Forces (ex-FAR) and militia from Rwanda in the immediate aftermath of the Rwandan genocide further exacerbated existing local tensions.[ii]
Large-scale war broke out in 1996, when Rwanda’s post-genocide government decided to militarily intervene against the ex-FAR and former genocidaires that had congregated in refugee camps across the border and launched a series of low-level attacks back into Rwanda between 1994 and 1996. In September 1996, the Alliance of Democratic Forces for the Liberation of Congo-Zaire (AFDL), a loose coalition of rebels led by Laurent-Desiré Kabila and backed by Rwandan and Ugandan forces, launched an armed campaign from the country’s east. The AFDL initially targeted refugee camps along the Zaire-Rwanda border that housed thousands of Hutu civilians. However, the rebels did not stop with the clearing the camps. Within six months they had captured most of the country’s territory and ousted long-time president Mobutu Sese Seko. In May 1997, Laurent-Desire Kabila was installed as the new president, and the country was renamed the Democratic Republic of Congo.[iii]
A brief pause in large-scale violence followed. However, relations between Kabila and his Rwandan and Ugandan backers quickly deteriorated. In August 1998, Kabila’s former allies began fuelling a new rebellion against the DRC’s leadership through the Rassemblement Congolais Democratique (RCD), a rebel group composed of many former AFDL members. The RCD was met with significant opposition and failed to oust Kabila with the ease it had expected. Instead, the conflict escalated into what has been described as Africa’s “first world war,” with the regular armed forces of the DRC, Zimbabwe, Angola and Namibia fighting on one side with some involvement of Chad and Sudan, and Rwanda, Uganda and Burundi supporting the rebel forces on the other. Rebel movements quickly proliferated. In 1999, Rwanda and Uganda turned against each other, and the RCD split between the Rwandan-backed RCD-Goma and the Ugandan-backed RCD-K/ML.[iv]
A ceasefire agreement signed in Lusaka in 1999 and the deployment of a UN peacekeeping mission (MONUC) failed to end the conflict. Up to fourteen armies continued to fight on Congolese territory until 2003, each supporting one or several of the Congolese rebel movements.[v] The conflict de facto partitioned the country into five main areas of control, with the majority of the fighting occurring between different armed groups in the country’s east and behind the Lusaka ceasefire line. The Inter-Congolese Dialogue eventually led to the withdrawal of foreign forces and a series of peace agreements between 2002 and 2003. However, thousands of civilians continued to die in the subsequent “political transition” period due to unresolved local conflicts and rivalries.
Throughout the conflict’s many phases, and even after the withdrawal of international forces, the different warring factions targeted civilians – often solely on the basis of their ethnicity, but also for a host of personal, criminal and parochial political reasons. Inter-group tensions and violence long predated the refugee crisis of 1994 – 1996 and the war that followed. Human Rights Watch reports that already in 1993, “some 6,000 people were killed and some 250,000 displaced in clashes that pitted the Nande and Hunde (considered to be indigenous to the region) against the Banyarwanda,” a term used to describe Rwandan and Burundian immigrants in Zaire, many of whom had been in North Kivu for several generations.[vi] Autesserre (2010) cites the UN Mission in the Democratic Republic of Congo as reporting “at least one thousands deaths on both sides and the displacement of more than 130,000 people.”[vii] This initially spike in violence was primarily rooted in micro-rivalries over land, resources and local power, exacerbated by Mobutu’s promotion of the principle of “indigeneity.”[viii]
Patterns of violence changed as the conflict evolved. The first phase (or “First Congolese War”), which lasted from July 1996 to July 1998, was the most intensively lethal, with sources noting a significantly higher incidence of killing during this period. The conflict pitted the AFDL (backed by Rwanda, Uganda, Burundi, Angola and other neighboring states) and the Rwandan Patriotic Army (RPA) against a coalition of former Rwandan Armed Forces (ex-FAR), Rwandan Interahamwe militia, then President Mobutu Sese Seko’s Zairian Armed Forces (FAZ) and mercenaries.[ix] The majority of deaths during this period resulted from the strategies of the AFDL and RPA, who deliberately targeted Hutu refugees from the eastern refugee camps. Not only did they target civilians while dismantling the camps, but chased refugees as they fled, killing them at multiple locations across the country. The AFDL also intentionally blocked humanitarian assistance to civilian refugees.[x] At the same time, Zairian armed forces targeted and deported Tutsi civilians.[xi]
In the later stages of the conflict, deserting soldiers of the former Zairian army also looted supplies, destroyed villages and raped civilians in their flight. The ex-FAR and its militia further targeted civilians to prevent other refugees from repatriating to Rwanda and used unarmed refugees as human shields.[xii]
With the start of the “Second Congolese War” in August 1998, Kabila’s government and its various aligned forces issued a call to target Tutsi, Rwandan, and Banyamulenge groups, whose ancestry is Rwandan but who for the most part had lived in Zaire/DRC for generations. Arbitrary arrests, disappearances, mass executions, rape and theft followed. Violence against communities of Rwandan descent was particularly high in areas where they lacked the protection of the RCD-G or Rwandan soldiers.[xiii] In areas where Rwanda- and Uganda-backed forces dominated, people affiliated with the Congolese government were targeted. Attacks sparked counter-attacks, initiating waves of violence that often washed over the same communities multiple times. Local tensions reinforced national and regional cleavages, and vice-versa. Large-scale displacement and the devastation of agriculture, health systems, public service delivery and other civilian structures led to enormous increases in mortality. Sexual violence, often perpetrated with weapons and with the intent to mutilate and permanently scar the victims, was rampant throughout the conflict.[xiv]
In Ituri region, a period of escalated violence between Hema and Lendu groups occurred in 1999 – 2000, and again in 2002 – 2003. The area, like much of the east, was impacted by the split of the RCD militia into Rwandan and Ugandan factions, which exacerbated long-standing local land disputes. Human Rights Watch estimated that at least 5,000 people were killed in Inturi between July 2002 and March 2003; this is in addition to the UN’s estimate that 50,000 people killed in the area between 1999 and 2003.[xv]Armed groups used extreme and public violence, including cannibalism, torture and rape, to terrorize civilian populations.[xvi]
The size and difficult terrain of the DRC, coupled with the highly complex nature of the conflict, have confounded efforts to capture and summarize its catastrophic impact on civilians. The Mapping Team of the United Nations Office of the High Commissioner of Human Rights and published in 2010 provides a strong, illustrative (not comprehensive) cross-country examination of the worst incidents of killing, but was limited by resources, time and a standard of proof that required multiple sources before including information.[xvii] The report documents 39,248 violent deaths and forced disappearances without release or return between 1993 and 2003—and this is surely an undercount. The vast majority of these documented violent deaths were civilian deaths, as well as a smaller number of out-of-commission soldiers killed while seeking treatment at hospitals or after having laid down their weapons. The confirmed deaths likely comprise only a small number of the actual people killed during the conflict. For example, the number of deaths for the most part does not include the thousands of child soldiers forcibly recruited during the conflict. The Mapping Team also did not include any reported killings that could not be corroborated by more than one source.
In many cases, the team was unable to confirm specific cases due to the remoteness of the atrocity sites, time restrictions, or a lack of surviving witnesses. As a consequence, the report states that the numbers of recorded violent deaths capture some of the many incidents of killing that took place in certain areas and during the time periods they prioritized. The report also cites larger numbers, while stating that the Mapping Team was unable to confirm the totals. For example, in Orientale province in 1997, a militia member reported seeing over 1,000 executions each week. The Mapping Team records only 1,835 violent deaths and disappearances without return for the whole year.[xviii]
Notes on sources: PITF data from 1995 to 2012; UCDP data from 1989 to 2014; ACLED data from 1997 to 2014; WPF data from 1993 to 2003. This chart is illustrative of patterns of killing as recorded by several datasets, it is not a record of the total number of people killed which is likely considerably higher.
 The original source of the WPF dataset is the UN Mapping Report, which does not provide annual totals, but rather groups fatalities in relation to phases of the conflict. We divided their totals evenly into annual spans, but this introduces some distortion. Their data is grouped into March 1993-June 1996 [1231 killed]; July 1996 – July 1998 [30,178 killed]; August 1998 – January 2001 [8239 killed]; and January 2001 – June 2003 [6861 killed]. Coding reflects acceptance of the higher figure where a range was included, but used zero as the number wherever the number of civilians killed was reported as unknown.
The highest numbers of civilian deaths—including all excess mortality, not just killing— is reported as up to 5.4 million between August 1998 and April 2007, was put forward by the International Rescue Committee. The IRC estimated the crude death rate (CDR) using a survey which asked respondents about the size and composition of their household, and whether there had been any deaths in the household during a specific time period. The resulting estimated CDR was then compared to the average CDR for Sub-Saharan Africa, and the difference assumed to represent conflict-induced deaths.[xix] As the IRC notes, this number includes excess mortality caused by the war and thus captures the humanitarian impact of the conflict rather than direct violence. Based on the organization’s methodology, only 0.4 percent of all deaths could be attributed to the latter. The IRC numbers have been widely criticized for being too high, largely because of sampling and baseline issues.[xx] However, with respect to direct violence and even in comparison with the Mapping Report (which, as noted, is incomplete), their numbers in fact appear relatively low: 0.4% of 5.4 million is 21,600 (1998 – 2001), and 0.4% of 2.1 million (the number of deaths reported for 2002 – 2007) is 8,400.
The Armed Conflict Location and Event Data Project (ACLED) tracks violent events and fatalities in DRC (1997 – 2014). The database indicates a total of ACLED records a total of 32,872 civilian deaths for 1997-2003. Their method of tracking fatalities, which is based on newspaper reports, provides a sense of the overall intensity of violence rather than a direct measure, as it is likely to undercount fatalities.
1993 – 1996
One estimate for the total number of people killed in ethnic violence in North Kivu between 1993 and June 1996 is between 70,000 and 100,000 individuals, but the number has proven impossible to verify.[xxi] Similarly, a local NGO cites a total of 50,000 killed during the entire campaign of persecution against Kasaian civilians carried out by JUFERI and Governor Kyungu wa Kumwanza in collusion with President Mobutu in 1994.[xxii] The United States Committee for Refugees and Immigrants suggests 6,000 to 40,000 lives were lost in the densely populated Masisi zone of North Kivu province in eastern Zaire during 1993-96, but does not offer a specific source for these estimates.[xxiii] During this same time period, the UN Mapping Team documents only 560 violent deaths.
However, there seems to be some consensus that a spike in violence occurred in North Kivu in 1993. Human Rights Watch reports that 6,000 people were killed in clashes between indigenous and Banyarwanda groups in 1993.[xxiv] Similarly, Medecins sans Frontieres estimated that between 6,000 and 15,000 people had died at the provincial level between March and May of 1993.[xxv]Autesserre (2010) cites the UN Mission in the Democratic Republic of Congo as reporting “at least one thousands deaths on both sides” for the same time period.[xxvi] For the same period, the OHCHR Mapping Team recorded between 110 and 586 confirmed deaths.
1996 – 1998
The UN Mapping report recorded the greatest number of violent incidents for this period in the whole decade under examination (238 confirmed incidents).[xxvii] The UCDP One-Sided Violence Dataset records a high estimate of 1,746 fatalities and a “best estimate” of 1,253 fatalities in 1996; a high estimate of 5,083 and best estimate of 4,160 fatalities in 1997; and a high estimate of 1,666 and best estimate of 1,291 in 1998.[xxviii]
Other sources provide more imprecise estimates. For example, Amnesty International in a 1998 report stated that during a seven-month period in 1996-97, “tens of thousands of unarmed refugees and Congolese citizens were massacred by members of the AFDL and other combatants, particularly members of the Rwandese Patriotic Army (RPA), the Rwandese government army.”[xxix] ACLED records 25,573 deaths in 1997 and 1,696 in 1998, with a total of 27,269.
1998 – 2003
This period was marked by the intervention of outside forces in the conflict. Violence against civilians reached particularly high levels of intensity in North Kivu and South Kivu, Orientale Province (in particular in Ituri), Katanga, Équateur and also Bas-Congo.[xxx] In Ituri region, a period of escalated violence occurred between 2002 – 2003. Human Rights Watch estimated that at least 5,000 people were killed in Ituri between July 2002 and March 2003; this is in addition to a UN-estimated 50,000 people killed in the area between 1999 and 2003.[xxxi]
The UCDP One-Sided Violence Dataset records a decline in killing after 1998 (high estimates of 632 in 1999 and 82 in 2000), a small uptick in fatalities in 2002-2006 (though estimates for 2003 and 2005 are missing) and a significant spike in 2009 (high estimate of 1095 and low estimate of 887).[xxxii] ACLED records 1,733 civilian deaths in 1999; 413 in 2000; 1,381 in 2001; 1,418 in 2002; and 659 in 2003. The total for the period based on ACLED data is thus 5,603.[xxxiii]
2004 – 2009
According to data found in the Mapping Report (illustrative not comprehensive reporting of violence), by 2004, existing data indicates a clear drop in fatalities that sustains through 2008, when the number jumps to over 2,000, and then 2009 when it nears 4,000. These numbers, while imperfect, suggest that the decline 2004 – 2007 was substantial and that the increase in killings in 2008 was likewise significant. ACLED reports a significant decline: 105 civilian fatalities in 2004, 335 in 2005, 72 in 2006 and 66 in 2007, confirming an overall decrease in violence until 2008. In August 2008, tensions between Mai Mai groups, army units and people of Rwandan descent again escalated into large-scale fighting in North Kivu.[xxxiv]
The pattern of ‘ending’ in DRC that fits our definition—drop below an annual number of 5,000 civilians killed, adding up to at least 50,000 deaths within a five year period–can only be described as the confluence of multiple factors that resulted in an apparent decline, though certainly not absolute ending, in the number of people being killed. The decline was possible as the various regional actors decreased their direct engagement; the UN increased their peacekeeping presence; and the national political process moved forward. The lack of state capacity, particularly vis-à-vis insurgent forces (many of whom continued to receive substantial support from neighboring countries) rendered even this limited ‘ending’ vulnerable to evolving re-alignments of power.
After the assassination of Laurent Desire Kabila in January 2001, his son Joseph Kabila took control of the country. Under Joseph Kabila, the primary belligerents in the conflict agreed to engage in an Inter-Congolese Dialogue (ICD). A series of local ceasefire agreements, the gradual demobilization of militias, and the presence of MONUC (UN Mission in the DRC) forces in some areas appear to have resulted in decreased numbers of violations and killings. In 2001, there was a ceasefire between the major belligerents in Katanga province. As part of the ICD, President Kabila and Movement for the Liberation of the Congo (MLC) head Jean-Pierre Bemba signed a power-sharing and peace agreement.
On the international level, President Kabila on 30 July 2002 signed an agreement with Rwandan President Kagame that outlined plans for the withdrawal of the RDF from the DRC and the dismantling of ex-FAR/Interahamwe militias. A further agreement with Uganda followed on 6 September 2002. Under intense international pressure, the Congolese government’s foreign supporters, Zimbabwe, Angola, and Namibia began to withdraw as well. A unified government was reinstalled in the capital and preparations for democratic elections initiated.
However, the conflict persisted at the local level, particularly in North Kivu, South Kivu and North Katanga.[xxxv] Former warring parties resisted the army integration process due to mistrust in the government and attempted to maintain territorial control and deter villagers from supporting some other faction.[xxxvi] Extremists within each of the three warring parties – the RCD-G, the MLC and the Kabila government – continued fighting in order to stall the transition process.[xxxvii] Rwandan troops remained physically present in the Kivus and continued to carry out “hit-and-run” operations in the border regions, allegedly to combat Hutu militias. Moreover, the different components of the transitional government maintained parallel military and administrative structures striving to retain control over their respective assets.[xxxviii]
In Kasai Occidental, a ceasefire between armed groups and the deployment of MONUC forces to the area led to a gradual restoration of peace between 2001 and 2003, although civilians continued to be targeted.[xxxix] During this same period in Kinshasa, there was a decrease in human rights violations, but government repression of political opponents and civil society continued in the form of arbitrary detention, extrajudicial executions, rape, and acts of torture.[xl] An interim emergency multinational force under European command and authorized by Resolution 1484 was deployed to Bunia in Orientale province in June of 2003 and restored order in the town. However, violence continued in many of the surroundings areas.[xli]
In some regions, ceasefires and military withdrawals were met with quick advances by other armed groups, who continued to target and exploit the population. For example, as the RDF withdrew from South Kivu, Mai-Mai and FDLR units expanded their control over a number of areas in the region. The September 2002 peace agreement between the DRC and Uganda, the creation of a Peacekeeping Commission in Ituri and international pressure led to the withdrawal of UPDF troops from Ituri in 2003, but local militias quickly positioned themselves to fight for control of the region.[xlii] After a spike in violence in 2003, Ituri benefited from greater international attention and resources than the other eastern provinces, despite the fact that violence in Katanga and in the Kivus caused more casualties.[xliii] Violence in North Kivu persisted due to tensions among the region’s eight ethnic communities over land and citizenship issues and conflicts between the five armed groups present in the province. New actors also played a role in continuing violence against civilians: in 2009 – 2010, for instance, the Ugandan rebel group the Lord’s Resistance Army killed over 300 civilians in Congo.[xliv]
International peacekeepers seem to have played an important if limited role in reducing the rate of killing. The deployment of MONUC forces did stem the rising tide of violence in Bunia-area in 2003, even if their presence was initially limited to the city. The decline in direct killings from 2004 – 2007 correlates with a more robust pursuit of militias by MONUC, which gained in capacity in 2003 and again in 2004 and conducted a series of offensives in 2005 – 2007. The presence of MONUC troops and observers in unstable places acted as a deterrent, and UN troops helped mediate local ceasefires between different militias.[xlv]
However, MONUC lacked a comprehensive strategy for local peacebuilding, and its overall involvement in local conflict resolution remained ad-hoc and uncoordinated.[xlvi] UN forces were unable to prevent a renewed spike in killings after 2008, despite retaining the same number of troops and the same mandate (including following the deployment of an intervention brigade in 2013). ACLED indicates that 473 civilians were killed between 2005-2007, in contrast to 1,639 civilians killed in 2011-2013. This is not to argue that the peacekeeping force, its mandate, interpretation of the mandate, and force size are irrelevant, rather than that they were clearly not sufficient. Other factors seem to have contributed to the decline in killing between 2003 and 2007. One potential explanation relates to the salience of the elections in 2006, which may have impacted the calculations of actors on the ground. However, the 2006 elections established that the RCD could not survive the turn to electoral politics, possibly increasing their return to use of violence as a means to consolidate areas of influence.
Throughout this period there was significant international engagement to pressure neighboring countries to decrease their involvement in DRC (ex: several UN reports, on illegal mineral extraction and the Mapping report; in the form of a peacekeeping force, which also assisted with elections; and through humanitarian assistance to vulnerable civilian groups particularly in the East, among other activities.
We code this period of atrocities as ending through the rise of moderating actors, both domestic and international who helped to push for a mediated solution, which also governed the withdrawal of various national forces from across the region, and the deployment of a UN peacekeeping force. The moderating shift occurred as the conflict appeared to stalemate. Throughout, multiple civilian groups were targeted for violence by myriad actors, both state and non-state.
Armed Conflict Location & Event Data Project, accessible at: http://www.acleddata.com/.
Amnesty International. 1996. “Zaire: Violent Persecution by State and Armed Groups,” 1 January. Available at: http://web.amnesty.org/library/pdf/AFR620261996ENGLISH/$File/AFR6202696.pdf
Amnesty International. 1998. Democratic Republic of Congo: A Long-standing Crisis Spinning Out of Control, 3 September. AFR 62/033/1998. Available at: http://www.refworld.org/docid/45b9a2a72.html.
Autesserre, Severine. 2006. Local Violence National Peace? Postwar ‘settlement’ in the Eastern D.R. Congo (2003 – 2006).” African Studies Review 49 (3): 1 – 29.
Autesserre, Severine. 2010. The Trouble With the Congo: Local Violence and the Failure of International Peacebuilding New York: Cambridge University Press.
Human Rights Watch. 1996a. “Forced to Flee: Violence against Tutsis in Zaire,” 1 July. A802, Available at: http://www.refworld.org/docid/3ae6a8200.html.
Human Rights Watch. 1996b. Human Rights Watch World Report 1996 – Zaire, 1 January, Available at: http://www.refworld.org/docid/3ae6a8ac14.html.
Human Rights Watch. 1997. “What Kabila is Hiding: Civilian Killings and Impunity in Congo,” Human Rights Watch Report 9, no. 5 October.
Human Rights Watch 1998. Human Rights Watch World Report 1998 – The Democratic Republic of the Congo (Formerly Zaire), 1 January. Available at: http://www.refworld.org/docid/3ae6a8a328.html.
Human Rights Watch. 2003. “Covered in Blood: Ethnically Targeted Violence in Northern DRC,” New York: Human Rights Watch. July.
Human Rights Watch. 2010. “DR Congo: Lord’s Resistance Army Rampage Kills 321,” March 28. Available at: https://www.hrw.org/news/2010/03/28/dr-congo-lords-resistance-army-rampage-kills-321.
Human Security Report Project. 2011. Human Security Report 2009/2010: The Causes of Peace and the Shrinking Costs of War. New York: Oxford University Press.
Médecins Sans Frontières. 1995. Populations in danger 1995. London: Médecins Sans Frontières.
Meger, Sara . 2010. “Rape of the Congo: Understanding Sexual Violence in the Conflict in the Democratic Republic of Congo,” Journal of Contemporary African Studies 28/2.
Pedersen, Jon . 2009. “Mortality from the Conflict in the Democratic Republic of Congo 1998-2006: A Reexamination,” Health and Nutrition Tracking Service.
Office of the High Commissioner for Human Rights (OHCHR). 2010. Democratic Republic of the Congo, 1993 – 2003: Report of the Mapping Exercise documenting the most serious violations of human rights and international humanitarian law committed within the territory of the Democratic Republic of the Congo between March 1993 and June 2003, August.
United States Committee for Refugees and Immigrants. 1997. U.S. Committee for Refugees World Refugee Survey 1997 – Democratic Republic of Congo (formerly Zaire), 1 January. Available at: http://www.refworld.org/docid/3ae6a8a70.html.
[i] Human Rights Watch 1996.
[iii] Autesserre 2010, 47-48.
[iv] Autesserre 2010, 48.
[v] Autesserre 2010, 49.
[vi] Human Rights Watch 1996b.
[vii] Autesserre 2010, 141.
[viii] Autesserre 2010, 133.
[ix] Human Rights Watch, 1997, 1.
[x] Human Rights Watch 1997, 19.
[xi] Amnesty International 1996.
[xii] Human Rights Watch 1998.
[xiii] Autesserre 2010, 145.
[xiv] Meger 2010.
[xv] Human Rights Watch 2003.
[xvi] Ibid., 8.
[xvii] Office of the UN High Commissioner for Human Rights 2010.
[xviii] Ibid., 110.
[xix] Pedersen 2009.
[xx] Human Security Report Project 2011.
[xxi] Office of the UN High Commissioner for Human Rights 2010, 64.
[xxii] Ibid., 55-56.
[xxiii] United States Committee for Refugees and Immigrants January 1997.
[xxiv] Human Rights Watch 1996a and Human rights Watch 1996 b.
[xxv] Médecins Sans Frontières 1995.
[xxvi] Autesserre 2010, 141.
[xxvii] Office of the UN High Commissioner for Human Rights 2010, 8.
[xxviii] UCDP One-sided Violence Dataset 2015.
[xxix] Amnesty International 1998.
[xxx] Office of the UN High Commissioner for Human Rights 2010, 10.
[xxxi] Human Rights Watch 2003.
[xxxii] UCDP 2015.
[xxxiv] Autesserre 2010, 213-214.
[xxxv] Autesserre 2006, 3-4.
[xxxvi] Ibid., 9.
[xxxvii] Ibid., 11.
[xxxviii] Ibid., 6.
[xxxix] Office of the UN High Commissioner for Human Rights 2010, 240-243.
[xl] Ibid., 254.
[xli] Ibid., 232-234.
[xlii] Ibid., 227-232.
[xliii] Autesserre 2010, 207-208.
[xliv] Human Rights Watch 2010.
[xlv] Autesserre 2010, 201.
[xlvi] Autesserre 2010, 199-201.