For nearly three decades, armed conflict has been waged in the country by a variety of groups and continues to the time of writing. While violence has been seemingly relentless, there have been distinct, distinguishable phases, with mass atrocities likely reaching the 50,000 threshold between the end of 1988 and beginning of 1989, and falling below the 5,000 per year for two years in 1994 and 1995. Overall violence within the atrocity period of six years (1988-1993) likely resulted in 50,000 to 100,000 civilian deaths as a direct result of violence or hostilities.
While Somalia is ethnically, religiously, and linguistically homogenous, the country’s population is divided into clans who can draw their lineage back to a common ancestor. During the colonial period under Italy and Great Britain clans became a central feature of state administration and political competition. Colonial administrators established a patrimonial system of resource distribution, employed tactics of divide and rule along clan lines, and engaged in collective punishment of clans. All of these tactics would be employed during later periods of violence.[i] Somalia gained independence after a ten year period under a UN Trusteeship from 1950 to 1960. The northern and southern regions were united under multi-party democracy that lasted from 1960 to 1969.
In 1969, a bloodless coup resulted in the installment of President Siad Barre. From 1969 to 1978, the Barre Regime enjoyed relative popularity and financial support from both the Soviet Union and Western institutions.[ii] While projecting an image of Somalia as a constitutional state to international actors, Barre cultivated a patrimonial state that increasingly revolved around clan identity.[iii] Clan-based paramilitaries were funded and armed by the government, a practice that exacerbated relations between communities that had previously lived adjacently and intermarried with little conflict.[iv] Rather than completely excluding particular clans, Barre coopted key actors in certain sub-clans, causing divisions within the larger clans.[v] During this time, the regime passed legislation giving the state wide powers of detention and execution. A number of paramilitaries, militias, and security agencies were founded, including the National Security Service and the Victory Pioneers.[vi] While there were several incidents of political violence, this caused relatively low numbers of civilian deaths; no single incident from 1945 to 1975 seems to have caused more than 100 civilian deaths.[vii]
The Barre regime became increasingly oppressive and violent in the late 1970s through the 1980s, although mass atrocities did not begin until later. In 1977 Somalia entered the Ethio-Somalia or Ogaden war with Ethiopia. After a number of initial victories, the Soviet Union withdrew support from Somalia in favor of Ethiopia, and Somalia lost the war in 1978.[viii] Discontent with the Siad Barre regime began to spread after the military loss against Ethiopia. Siad Barre had eighty-two high level military officers executed in Ethiopian territory for their opposition to the way the war was handled. The military failure and execution of military officers prompted a 1978 coup attempt. Despite somewhat diverse clan participation amongst the coup leaders, Barre portrayed the coup as orchestrated by the Majeerteen clan. In a pattern similar to what would be used later against the Isaak clan in 1988, Barre responded by purging the government and military of Majeerteen, and committing reprisal killings against the Majeerteen civilian clan members that left roughly 2,000 dead.[ix]
The 1980s saw the rise of opposition armed movements, the largest of which was the Somali National Movement (SNM), drawn principally from members of the Isaak clan in northwestern Somalia, which developed in response to state marginalization and abuse including the purge of Isaak from civil service posts, confiscation of businesses, arrests, detention, and violence against Isaak civilians.[x] Throughout the 1980s, the Siad Barre regime responded to oppositional militias by employing increasingly violent and restrictive measures on various clan populations. Beginning in 1982, the state imposed curfews in certain areas that were used as a pretense for the detainment and extortion of civilians. Detainment and looting became a lucrative source of funding for state forces and paramilitaries that were referred to as the ‘meat market’.[xi] The government employed Mobile Military Courts (MMCs) to combat opposition militants and their associated civilian populations. MMCs were superficial judicial proceedings conducted by military officials, and followed almost immediately by executions. Although wholesale targeting and decimation of the Isaak population did not begin until the SNM offensive in 1988, a confidential report from General Morgan to President Barre that was leaked in February 1987 revealed a government intention to “liquidate” the “Isaak problem” through violent tactics.[xii]
In 1988, the SNM received information that they were about to be expelled from their base of operations in Ethiopia on account of a peace agreement between the Somali and Ethiopian governments. In response, on May 27, 1988, the SNM launched a sudden attack on Burao, followed by an attack on Hargeisa on May 31. The SNM dispersed amongst the civilian population, with some SNM in uniform and others in partial uniform or plain clothes. These tactics likely resulted in higher civilian casualties than necessary, however, the government clearly responded to the SNM attack with a purposeful program of reprisal against Isaak civilians.[xiii] Compagnon estimates that 15,000-20,000 civilians were killed directly from the bombing of Hargeisa and Burao.[xiv] The Africa Watch report, “A Government at War with Its Own People” estimated that roughly 50,000 to 60,000 people were killed between May of 1988 and the beginning of 1990.[xv] While it is unlikely that all of these deaths are civilian, with SNM membership estimated at only 10,000[xvi], it is clear that many civilians were killed, the large majority of which were Isaak. Deaths were inflicted through indiscriminate government bombing of the towns of Hargeisa and Burao. Civilians fleeing from fighting were strafed by government planes.
Although the largest spike in killings is in May and August of 1988, the government’s Somali National Army (SNA) continued to target the Isaak community over many months through round-ups and mass executions of Isaak civilians at the town level. Other government tactics against Isaak civilians included laying land mines around towns, destroying water points, killing or looting livestock, burning of villages, and arbitrary detention. A report by Robert Gersony, “Why Somalis Flee,” documented many of these tactics and calculated that, “5,000 unarmed civilian Issaks were purposefully murdered by the Somali Armed Forces between May 1988 and March 1989, in the absence of resistance and in contexts which presented no immediate danger to these forces.” In general, civilian casualties were slightly more likely to be male than female. In particular, men were targeted for execution and detention.[xvii]
Although violence from 1988 to 1990 was largely one-sided, the SNM forces were also responsible for civilian deaths, including about 400 deaths that resulted from attacks on refugee camps in the northwestern area of Somalia that housed ethnic Somalis from Ethiopia and from which the government had recruited SAF and militia members. All SNM attacks resulting in civilian deaths occurred in the period of May and August of 1988. With roughly 400,000-500,000 people displaced from the Northwestern area of Somalia by the government’s violent tactics and fewer Isaak left to attack, the government violence in Northwestern Somalia slowed.[xviii]
The bombing of Hargeisa and Burao came at a high cost to the Barre regime. Not only had the fighting been expensive, but the departure of many of the Somali elite, the withdrawal of vital U.S. support for Barre, and the sympathy the SNM increasingly attracted from neighboring countries threatened Barre’s regime.[xix] With a sense that the Barre regime was weakening, a number of clan based militias arose to secure control over their respective areas of the country. The United Somali Congress (USC) representing the Hawiye clan in Central Somalia emerged in late 1989. From the start, the USC was divided by an “internal” faction fighting within Somalia, and an “external” branch based in Italy. By 1990 the Somali Patriotic Movement (SPM) was established by Ogaden clansmen in the South. SNM retained some control in the Northwest and the Somali Salvation Democratic Front (SSDF) originally created in 1978 but revived in 1989 was present in the Northeast.
The Barre regime continued to launch targeted reprisals against the civilian population throughout 1990. Siad Barre’s presidential brigade, the Red Berets, was responsible for civilian executions in and around Mogadishu. In one incident, the Red Berets killed 100 civilians gathered in a stadium.[xx] In another incident in Buli Burti, the Red Berets killed fifty unarmed civilians (most were prominent locals such al elected officials, clan elders, and Islamic leaders) in retaliation for a USC attack on SAF troops.[xxi] Nontheless, the government lost ground, and eventually retained control over only roughly ten to fifteen percent of the state territory.[xxii] At the end of 1990, the USC launched an offensive on Mogadishu, and on 27 January 1991, Barre fled the capital.
Violence continued for many years thereafter, but is unlikely to have risen above 50,000 deaths caused by a single perpetrator, such as characterizes the cases in this study. Nonetheless, below, we provide an over view of the events that followed.
As Barre fled, the ‘external’ branch of the USC declared Ali Mahdi Mohamed president. The “internal” branch of the USC led by General Mohamed Farah Aydeed contested this decision, and civil war between the two factions enveloped Mogadishu shortly after Barre’s departure,[xxiii] with significant civilian casualties as a result of heavy artillery being used within the confines of a densely populated urban environment. While some of the fatalities were undoubtedly accidental, Amnesty International reports the intentional shelling of neighborhoods known to be associated with opposing factions[xxiv] and Prunier describes a situation in which prisoners were executed and ambulances routinely fired at.[xxv]
While violence raged in Mogadishu, other parts of the country were also enveloped in fighting between clan based militias, none of which is reflected in the above death tallies calculated for Mogadishu. In 1992, both the USC and SNF committed atrocities against civilian populations in the Gedo region of Somalia. Amnesty International recorded the testimony of survivors, who described tactics that included massacres of up to thirty or forty people at a time, cutting off and burning of body parts with acid, and the widespread use of rape.[xxvi] The SNF under General Morgan and SPM under Colonel Omar Jess battled over the port area around Kismayo.
Kapteijns categorizes the violence used during this time as “clan cleansing as a tactic to capture the state.”[xxvii] Civilians were intentionally and brutally targeted, and sexual violence, which had not previously been a prominent feature of the violence in Somalia, became pervasive.[xxviii] Kaptejins cites a breakdown of law and order following the collapse of the Barre regime, the erosion of cultural scripts, and increased impunity as causes for the escalation in tactics.[xxix] She further describes deliberate targeting of Daarood as the USC gained ground. She writes:
… when Mogadishu passed into its hands, the leaders of the USC, followed by USC fighters and civilian supporters, adopted a politics that defined as mortal enemy all Somalis encompassed by the genealogical construct of Daarood, which also included the president. Although the vast majority of these individuals had not been associated with, or benefited from the regime—in many cases as little or even less than those now hunting them down—they were nevertheless targeted for elimination and expulsion not only in Mogadishu but also, over a period of two years, in central, south-central, and southern Somalia. This is the violence that is central to this study. It did not just represent violence against civilians based on clan, which is in itself not new in Somalia, but a shift to a new kind of collective, clan-based violence, namely that of clan cleansing, in a new political context and with a new dominant discourse (Kapteijns, 2).
Massive displacement and disruption of the livelihoods of agro-pastoral communities resulted in a famine beginning in 1992. Although famine deaths can be viewed as a direct result of violence, they are not included in the casualty numbers of this paper, which reflect only intentional deaths violently inflicted on the civilian population. Various ranges emerge to capture the number of deaths resulting from the famine. Hansch et al. estimate 250,000-300,000[xxx] deaths while the International Committee of the Red Cross (ICRC) estimates 500,000 may have died.[xxxi] Famine in Somalia paved the way for an international military and humanitarian intervention in Somalia.
Joint U.S.-UN forces entered Somalia under the United Nations Task Force (UNITAF) banner in December of 1992. UNITAF was later transformed into UNOSOM and UNOSOM II. While inter-clan militia fighting continued during this period, new casualties arose from fighting between international forces and Somali militias. Newspaper reports from this period record anywhere from two to 357 Somali casualties as a result of clashes between international troops and Somali militias, although it is not always clear whether these deaths have been confirmed as entirely civilian.[xxxii] Civilian casualties and high numbers of civilians wounded by bullets appear to be the result of international military equipment such as the Cobra helicopter gunship. Human Rights Watch estimated that between December 1992 and October of 1993, at least 500 to 600 Somali civilians and combatants were killed, while 2,000 were wounded by U.S. or UNOSOM forces.[xxxiii] A December 8, 1993 New York Times article reported that the U.S. Special Envoy to Somalia and the Marine Corps General directing U.S. military operations there estimated 6,000 to 10,000 Somali fatalities occurred during the summer of 1993, either as a result of Somali factional fighting, or confrontations with UN Peacekeepers. General Zinni reported that he believed two-thirds were women and children used as shields by militia fighters in the fighting in Southern Somalia. The ICRC responded that the 10,000 figure seemed high, but that they believed several thousand to have been killed.[xxxiv]
Investigations into the actions of various UNITAF/ UNOSOM forces also revealed abuses against civilians. One investigation of Belgian forces revealed 58 cases of killing or injury to unarmed civilians, although the number of actual abusive killings may have been much larger than the inquiry suggests.[xxxv] Abuses committed by the Italian troops include the looting of displaced persons camps and rape. Malaysians also engaged in looting while Pakistanis and Nigerians indiscriminately fired on protesting crowds.[xxxvi] The United States adopted a practice of excessive force intended to achieve military victories with minimal loss of U.S. military lives. This policy, coupled with U.S. military technology in the urban Mogadishu environment led to breaches of the Geneva conventions and hundreds of civilian fatalities, including an attack on a hospital that led to at least two civilian casualties, and an attack on a mainly civilian political meeting of Aideed supporters, that resulted in 54 deaths, according the Red Cross (U.S. estimates where lower and Somali estimates were higher).
On September 9, 1993, a U.S. helicopter fired on an unarmed crowd killing roughly 60 civilians. Anywhere from 60 to 500 Somali deaths resulted from the October 3, 1993 “Black Hawk Down” incident.[xxxvii] After the “Black Hawk Down” incident, the United States declared a de facto truce with Gen. Aideed and the fighting diminished, and thereafter UNOSOM troops shifted their mission to a more defensive strategy, and as a result, civilian deaths caused by tensions between Somali and international troops dropped off.[xxxviii] UNOSOM forces were gradually withdrawn throughout 1994 and made a final exit in February of 1995. Before departing, some UNOSOM forces are reported to have sold military equipment to Somali militants, potentially causing an increase in the amount of heavy weapons on the ground.[xxxix]
By and large the casualty numbers for the period of mass atrocities in Somalia fail to distinguish clearly between civilian and militant deaths, and between deaths as a result of crossfire versus intentional killings. Among the better documented incidents are the Barre government offensive against the Isaak population, deaths in Mogadishu following the collapse of the Barre regime, and the subsequent inter-clan USC fighting. Very little information exists quantifying the loss of life that resulted from clan cleansing in semi-urban and rural areas after the flight of Barre.
Scholar Daniel Compagnon estimates that 15,000-20,000 civilians were killed directly from the bombing of Hargeisa and Burao that began in 1988.[xl] The Africa Watch report, “A Government at War with Its Own People” estimated that roughly 50,000 to 60,000 people were killed between May of 1988 and the beginning of 1990.[xli] The sustained and brutal targeting of civilians in the wake of these bombings was also corroborated by the Gersony Report. At the time of the bombing, Hargeisa had a population of 500,000, and the bombardment caused the destruction of 14,000 buildings and the damage of 12,000 more,[xlii] which reflects a destruction of 70% of the buildings.[xliii] Given the size the population and the level of destruction, the numbers presented by Compagnon and Africa Watch do not seem implausible.
The 1990 battle for Mogadishu resulted in high casualties. The US based NGO, Physicians for Human Rights, indicated that about 4,000 civilians may have died in the USC offensive on Mogadishu,[xliv] and up to 500 may have died in only a two day period during the height of the USC takeover.[xlv] These numbers are supported by Clodfelter who places the number of civilian casualties from 1990 to 1991 at 5,000.[xlvi]
Fighting between warlords continued to cause high numbers of civilian fatalities, even after Barre fled the country. Based on Mogadishu hospital records, Africa Watch estimated that the factional fighting between Aideed and Ali Mahdi’s USC factions caused 14,000 deaths between November 1991 and February 1992.[xlvii] Bradbury calculates that during all of 1991 and 1992, 25,000 people died in the factional fighting in Mogadishu.[xlviii]
The figures of 4,000 – 5,000 and 14,000 civilian casualties from the fighting within Mogadishu do not appear to overlap. They are collected from hospital reports and records at the time and appear reliable, particularly given the larger figure of 25,000 from Bradbury that encompasses a longer timeline from 1991 to 1992. Further, daily death tolls from this period are high with the United States Bureau of Citizenship and Immigration Services estimating deaths due to gunshot wounds in Mogadishu between August and September of 1991 at 70 to 100 per day.[xlix]
General Zini’s estimate of 6,000-10,000 deaths with two-thirds being civilian (an average of roughly 5,300 civilian deaths) in the summer of 1993 was not fully supported by the ICRC. However, given that no numbers are included to capture deaths as a result of semi-urban and rural clan cleansing, the inclusion of this higher statistic should result in unrealistic overall estimates.
This case study formally ends with the overthrow of Siad Barre in 1991.
In the mid 1990s, with killings and kidnapping of foreign nationals on the rise, the looting of World Food Program (WFP) stores in January 1995, and the withdrawal of international troops who had previously offered some protection to humanitarians, nearly all of the humanitarian assistance programs were suspended. With very few international organizations on the ground, there is little concrete data on civilian deaths in 1994 and 1995. Amnesty International, who continued reporting on Somalia in this period, documented intentional targeting and killing of hundreds of unarmed civilians as a result of the clan warfare.[l] No estimate of killings in this time period reaches into the thousands. It is difficult to know if the decrease in numbers is due to an actual drop-off in civilian deaths, or simply a lack of reporting and investigation into deaths. Despite a series of peace negotiations, some relatively successful local reconciliation programs, and some bilateral peace accords signed from 1993 to 1995, violence has not ended in Somalia.[li]
Prunier’s description of this period in Somalia is consistent with a lowering of casualties on account of violent parties reaching an acceptable stalemate. According to Prunier, the exit of international actors and armed forces deprived many factions of their economic bases. No single militant group possessed the strength or capability to overcome others and capture the state, but peace was not sought by these groups because the condition of statelessness and violence was beneficial to militias. Rather than pursue the re-establishment of a peaceful centralized Somali state, the militias pursued only enough stability as would allow them to conduct profitable business activities.[lii] The decline of mass civilian deaths is also correlated with the rise of the independent statelets of Somaliland (with Hargeisa as its capital) and Puntland.
We code this case as ending through defeat of Siad Barre. Violence continued thereafter, with many groups perpetrating atrocities, but it is not clear that any one surpassed the 50,000 fatality threshold for this study.