I. The Technological Evolution and Early Proliferation and Use of Cluster Munitions
A half-century ago, cluster munitions were a little known instrument of warfare. They have since become common—if controversial—weapons for most modern militaries. Cluster munitions gained preferential status through a combination of technological innovations, changing combat needs, industrial interests, permissive laws, and lack of public awareness or debate. These factors produced an area effect munition that exacts a lethal and predictable, even if unintentional, toll on civilians.
From their first major use, the civilian harm inflicted by cluster munitions has outweighed their military benefits. During the Vietnam War, the United States blanketed Southeast Asia with the weapons, causing civilian casualties at the time of attack and leaving millions of unexploded submunitions that continue to kill and maim decades later. Since then, cluster munitions have proliferated widely and been used in almost every region of the world. While the design of cluster munitions has evolved in ways that theoretically could reduce humanitarian harm, technological fixes have failed to eliminate the weapons’ negative effects. The history of development, use, and proliferation illuminates the major problems of cluster munitions and foreshadows the impact they still have today.
Early Development and Use
The technology that produced the earliest cluster munitions also gave rise to their devastating effects. Through experiments conducted in the early twentieth century, scientists determined that small, high-velocity projectiles were the most effective means of maximizing injury. Equipped with this insight, weapons designers worked to develop the controlled fragmentation of explosive devices by using certain metals and pre-fragmented materials. While increased and more predictable fragmentation ensured the lethality of individual submunitions, new fuze technology made their wide dispersal possible. In particular, mechanical time fuzes installed on the large container that carried submunitions released submunitions after the passage of a certain period of time, allowing them to spread over a wide footprint and hit a large number of targets. While the design of cluster munitions would continue to evolve, developments in fragmentation and fuze technology created a deadly weapon with a broad area effect.
Modern cluster munitions date back to the First World War, when Britain had the idea of dropping a group of munitions for incendiary bombing. Cluster-type weapons were also used during World War II, but at the time, military officials did not consider cluster munitions very effective because they were unable to control dispersal patterns. Although cluster munitions were not used during the Korean War, it sparked technological innovations that would make submunitions less expensive and more effective, and therefore more widespread. The United States sought weapons with maximum antipersonnel impact to offset the disadvantage of being outnumbered by enemy troops. Controlled fragmentation munitions, with their ability to incapacitate through debilitating wounds, offered a solution.
The Vietnam War made cluster munitions a staple of military operations, and a weapons expert described it as “a proving ground” for the weapons. Faced with Cold War insurgencies in the late 1950s and early 1960s, the US government pursued the development of conventional weapons that could combat guerilla enemies, who were often difficult for US soldiers to locate. The United States also wanted weapons that could attack anti-aircraft artillery and surface-to-air missions when their locations could not be pinpointed. Officials, therefore, pushed forward arms research on fragmentation, dispersal, and detonation. For example, new dispensers were designed to hold more submunitions and disperse these submunitions more widely. A mechanical device that would trigger submunition detonation based on the submunition’s spin rate was invented to prevent submunitions from exploding until they had penetrated the forest canopy (thus slowing the spin rate). By the time US troops began their major buildup in Southeast Asia, these branches of research had converged to enable the production of area effect cluster munitions that were affordable and ready for use in battle.
The United States made widespread use of a variety of cluster munitions during the Vietnam War. According to an analysis of bombing data by Handicap International, over the course of the conflict, US forces dropped approximately 80,000 cluster munitions (containing 26 million submunitions) on Cambodia, more than 296,000 cluster munitions (containing nearly 97 million submunitions) on Vietnam, and more than 414,000 cluster munitions (containing at least 260 million submunitions) on Laos. While the most common models were antipersonnel, some could also attack vehicles, a multi-effects trend which continues today. The area over which Vietnam War-era cluster munitions could spread submunitions ranged from 10,000 to 200,000 square meters.
US cluster munition attacks wrought tremendous harm against civilians in Indochina, and continue to do so. Michael Krepon, later founder of the Henry L. Stimson Center in Washington, DC, called cluster munitions “the most indiscriminate antipersonnel weapon used in the Vietnam War.” Much of the harm resulted from munitions that failed to explode when originally released but were later triggered by civilian passersby. Assuming a conservative dud rate of 5 percent and relying on Handicap International’s estimated number of total submunitions used, cluster munitions would have left more than 19 million unexploded submunitions. According to the International Committee of the Red Cross (ICRC), unexploded submunitions have killed or injured some 11,000 people in Laos, more than 30 percent of whom have been children. Civilians continue to this day to be killed and injured by unexploded submunitions in all three countries. There is no credible estimate of the amount of land still contaminated, but it will likely take decades to clear.
The use of cluster munitions provoked increasing public opposition to the Vietnam War in the United States and elsewhere. Though napalm was the weapon antiwar protesters targeted most, cluster munitions had a mobilizing effect as well. Opponents criticized the manufacturers of cluster munition components, and peace activists and antiwar journalists visiting North Vietnam discovered and reported on the humanitarian impact of the weapons. Serious doubts also existed as to whether the weapon on balance helped US military efforts. The uncoordinated and unsustained opposition to cluster munitions, along with a deliberate government policy of secrecy and lack of debate, precluded successful international initiatives to regulate or ban cluster munitions at this stage.
Early Proliferation of Cluster Munitions: 1970s and 1980s
The conflict in Southeast Asia significantly raised the profile of cluster munitions and transformed them from a military experiment into a mainstream weapon. By 1973, for example, cluster munitions comprised 29 percent of the US Air Force’s entire ordnance procurement budget. The weapons also quickly proliferated.
Use in the 1970s and 1980s extended to Africa, the Americas, the Middle East, and South Asia. In Africa, unknown forces left cluster munition remnants in Zambia (1970s), Morocco used cluster munitions against a non-state armed group in Western Sahara and Mauritania (1975-1991), the United States attacked Libyan ships (1986), and France and Libya launched attacks in Chad (1986-1987). In the Americas, the United Kingdom dropped 107 cluster munitions on the Falkland Islands/Malvinas (1982), and the United States dropped 21 in Grenada (1983). In the Middle East, Israel used cluster munitions in Syria in 1973 and in Lebanon in 1978 and 1982. The United States used cluster munitions against Syrian units in Lebanon in 1983 and Iranian ships in 1988, and Iraq used cluster munitions in its war with Iran beginning in 1984. Finally, the Soviet Union, which would become another major user, used cluster munitions during its invasion of Afghanistan from 1979 to 1989, primarily to attack mujahiddeen strongholds and exposed fighters.
Timeline of Cluster Munition Use
A Spike in Use: 1990s
The Gulf War of 1991
The United States and its allied coalition opened the 1990s with the most extensive use of cluster munitions since the Vietnam War. Cluster munitions accounted for about one-quarter of the bombs dropped on Iraq and Kuwait during the Gulf War of 1991. Between January 17 and February 28, 1991, the United States, France, the United Kingdom, and to a limited extent Saudi Arabia dropped about 61,000 cluster bombs, releasing twenty million submunitions, only about 15 percent of which were new models. They used an additional estimated 30 million surface-launched submunitions. Coalition armed forces notably targeted mobile SCUD missiles and Iraqi tank and vehicle columns retreating from Kuwait. As a result, unexploded submunitions littered roads, culverts, and bridges. Coalition forces also used cluster munitions in urban areas, leading to attacks on infrastructure and dual use targets frequented by civilians during and after the war. Exacerbating the humanitarian threat, use of low-altitude cluster munitions at medium to high altitudes decreased the accuracy of strikes and increased the dispersal pattern of the submunitions.
While the lack of precision exacerbated the risk to civilians during strikes, duds caused most of the civilian cluster munition casualties in the Gulf War. As of February 1993, unexploded submunitions had killed 1,600 civilians and injured 2,500 more. Post-war research revealed an “excessively high dud rate” due to the high altitude from which they were dropped and the sand and water on which they landed. The large quantity of cluster munitions added to the problem; even a 5 percent dud rate would have left more than 2 million unexploded submunitions. The plethora of duds on major roads put both refugees and foreign relief groups at risk. The duds particularly endangered children; 60 percent of the victims were under the age of fifteen.
Unexploded submunitions caused other significant side effects. First, they slowed economic recovery because duds needed to be cleared before people could restore industrial plants, communication facilities, and neighborhoods and extinguish the oil fires in Kuwait. Second, during and after the war, unexploded ordnance (UXO), including submunitions, represented the “greatest threat” to US troops. Submunitions killed or injured more than 100 US soldiers and killed an additional 100 clearance workers.
Other Conflicts in the 1990s
Over the course of the rest of the decade, militaries used cluster munitions in armed conflicts in Africa, Central Asia, and Europe. In Africa, both Eritrea and Ethiopia used the weapons in their 1998 territorial dispute over the Badme border area, causing hundreds of civilian casualties. Other cluster munition attacks in Africa affected Angola, Sudan, Sierra Leone, and the Democratic Republic of the Congo during this time period.
The use of cluster munitions also accompanied the breakup of the Soviet Union. Cluster munition remnants were found from conflicts in Nagorno-Karabakh and Azerbaijan (1992-1994) and Tajikistan (1992-1997). The Russian government used cluster munitions extensively in Chechnya between 1994 and 1996 and again in 1999. The attacks culminated in at least 636 casualties, including 301 deaths. Russia directed many, if not most, of its cluster munition attacks, including the 1999 attack on the Grozny market, at civilian areas. According to one estimate, the Grozny attack killed 137 people.
Conflict in the Balkans led to use of cluster munitions in Europe, which became the decade’s most affected region. Internal and NATO forces used cluster munitions in Bosnia and Herzegovina between 1992 and 1995, reportedly causing at least 92 casualties, of which 13 individuals were killed and 79 injured. Armed forces of the self-proclaimed Republic of Serbian Krajina (RSK) used cluster munitions in Croatia in 1995, causing 221 known casualties (7 individuals killed and 214 injured) at the time of attack alone. The International Criminal Tribunal for the Former Yugoslavia ultimately found RSK President Milan Martic guilty of crimes against humanity for cluster munition attacks that landed on Zagreb’s commercial center in May of that year. During the 1999 conflict in Yugoslavia, which will be discussed in detail in the next chapter, NATO forces launched an extensive air campaign that used cluster munitions, and both NATO and Yugoslav armed forces used cluster munitions in Albania, which was not party to the fighting. Overall, cluster munitions were likely used more extensively during the 1990s than in the previous two decades combined.
Modern Technological Developments
After the Vietnam War, as use of cluster munitions spread, the technology of cluster munitions continued to evolve. A trio of developments has attempted to improve cluster munitions’ accuracy and reliability, primarily for military reasons but also to decrease humanitarian harm. None, however, has succeeded in eliminating the inherent problems of the weapons.
Perhaps the most important technological change was the addition of devices designed to reduce dud rates, including self-destruct, self-neutralization, and self-deactivating mechanisms. In theory such devices would minimize the number of civilian casualties, but as exemplified by the use of M85 submunitions with self-destruct devices, they failed to do so. British ground forces used M85s for the first time in combat during major hostilities in Iraq in 2003. One British officer told Human Rights Watch that his troops were more careless about using the M85 in populated areas because they assumed the self-destruct mechanism had eliminated the humanitarian impact and they neglected to consider the danger of cluster munitions during strikes. The true test of M85s came when Israel, which produces the submunitions, used them extensively in south Lebanon in 2006. Many military experts at that time consider it to be the most reliable submunition model produced because it had a 1.3 to 2.3 percent dud rate in tests. Clearance groups in Lebanon found numerous unexploded M85 submunitions with self-destruct mechanisms, however, indicating that the M85’s self-destruct component did not always work as designed. In an in-depth study of strike locations where these submunitions landed, C. King Associates, the Norwegian Defence Research Establishment, and Norwegian People’s Aid concluded that the failure rate of such submunitions in Lebanon was in fact around 10 percent. The UN Mine Action Coordination Center–South Lebanon (MACC SL) echoed this finding. In both Iraq and Lebanon, a touted “technological fix” had failed to eliminate the humanitarian problems of cluster munitions.
Modern cluster munition technology has also sought to increase the accuracy of the container itself. More precision would improve the chances of hitting the intended target, which would have military and humanitarian benefits. During its bombing of Afghanistan in 2001, the United States used for the first time in combat the CBU-103, regarded as a technical improvement over the CBU-87 used in the Gulf War. It was outfitted with a Wind Corrected Munitions Dispenser (WCMD) designed to improve accuracy by compensating for wind encountered during its fall and narrowing the pattern of submunition dispersal. Despite pronouncements by the US Air Force that the CBU-103 was “highly successful,” Human Rights Watch did not find evidence during field missions in Afghanistan and Iraq to conclude that the modification provided a technological fix to the humanitarian problems caused by cluster munitions. In particular this model still released 202 submunitions, many of which did not explode on impact as designed, and was vulnerable to poor targeting. Other more precise munitions, such as the Joint Standoff Weapon-A (JSOW-A) and BGM-109D Tomahawk Land Attack Missile (TLAM-D), have been equipped with submunitions but used much more rarely.
Cluster munitions that have guided submunitions with fail-safe mechanisms, often called sensor fuzed weapons after a US model, are the most cutting-edge types in existence. The United States introduced its air-delivered Sensor Fuzed Weapon, or CBU-105, when it dropped 88 of them in Iraq in 2003. Equipped with a WCMD to direct the canister, it contains 10 BLU-108 submunitions that each include four hockey puck-sized “skeet” warheads. Infrared and laser sensor guidance systems on the skeets are designed to direct them to targets with high heat sources, such as armored tanks, parked airplanes, and vehicles. If they fail to find such a target, one of a trio of fail-safe mechanisms is supposed to activate. The manufacturer, Textron Defense Systems, claims that these redundant mechanisms “are key elements that distinguish [Sensor Fuzed Weapons] from traditional munitions, preventing hazardous unexploded ordnance and ensuring a clean battlefield for follow-on troop movement and civilian habitation of the area.” Not enough evidence is available, however, to determine what kind of humanitarian impact sensor fuzed weapons would have in the field, or whether they would function as designed under battle conditions.
Despite these multiple technological developments, states have also continued to use Vietnam War-era cluster munitions. The United States used updated versions of the Rockeye, containing 247 dart-shaped dual-purpose Mk-118 submunitions that are known to leave behind a high number of unexploded duds, in Yugoslavia in 1999, Afghanistan in 2001 and 2002, and Iraq in 2003. In Yugoslavia and Iraq, the United Kingdom used variants of the air-dropped BL-755, modeled after another Vietnam War-era cluster munition containing 147 submunitions. In Lebanon in 2006, Israel used US-manufactured and -supplied air-dropped CBU-58B cluster munitions containing 650 BLU-63 antipersonnel submunitions each. Deminers after that conflict discovered CBU-58B canisters marked with a September 1973 load-date that suffered catastrophic failures, meaning that they failed even to dispense their submunitions.
Over the past five decades, militaries increasingly have come to choose cluster munitions as an important element of their arsenals. At least 86 countries acquired stockpiles of the weapons and their use spread to Africa, Asia, Europe, Latin America, and the Middle East. Newer, more technologically advanced submunitions have been developed but have failed to solve humanitarian problems. At the same time, models from the 1970s continue to be used. While the Vietnam War may have been the most egregious case of civilian harm from cluster munitions, it was only the beginning.
For a discussion of the science of wound ballistics, see Eric Prokosch, The Technology of Killing (London: Zed Books, 1995), pp. 10-27.
For a discussion of early developments in controlled fragmentation, see ibid., pp. 37-39. The US Army defined controlled fragmentation as “the technique of design and fabrication of a projectile, mine, grenade or bomb to cause a predetermined size, shape, density, velocity and pattern of the fragments upon detonation.” Ibid., p. 49 n. 22.
Ibid., p. 103.
Ibid., p. 82. The first known use of a weapon resembling the cluster munition appears to have occurred during the 1840s, when the Swedish military fired a cluster of grenades from a mortar. Rae McGrath, Cluster Bombs: The Military Effectiveness and Impact on Civilians of Cluster Munitions (London: UK Working Group on Landmines, 2000), http://www.landmineaction.org/resources/Cluster_Bombs.pdf (accessed September 17, 2010), p. 11.
 In 1943, Soviet forces employed cluster munitions against German armor while the German military used SD-2 butterfly bombs, antipersonnel devices resembling cluster munitions, against Britain. The United States dropped cluster-like munitions in its incendiary bombing of Tokyo, carpeting the city with groups of six-pound napalm explosives in an operation that killed tens of thousands. International Campaign to Ban Landmines (ICBL), Cluster Munition Monitor 2010 (Ottawa: Mines Action Canada, 2010), p. 240, www.the-monitor.org (accessed November 1, 2010) [hereinafter Cluster Munition Monitor 2010]; Eric Prokosch, Cluster Weapons (Colchester: Human Rights Centre, University of Essex, 1995), p. 1; Prokosch, The Technology of Killing, p. 82; Major Thomas J. Herthel, “On the Chopping Block: Cluster Munitions and the Law of War,” Air Force Law Review, vol. 51 (2001), p. 236.
 Prokosch, The Technology of Killing, pp. 32-33; Stockholm International Peace Research Institute, Anti-personnel Weapons (London: Taylor & Francis Ltd, 1978), p. 161. Early submuntion models were primarily antipersonnel weapons because they had been designed to counter waves of enemy troops of the kind encountered in the Korean War. Later models, particularly those designed to attack Soviet armor, usually had dual purposes.
 Prokosch, The Technology of Killing, p. 7.
 Ibid., p. 81.
 Herthel, “On the Chopping Block,” Air Force Law Review, p. 237.
 Prokosch, The Technology of Killing, p. 82.
 Ibid., p. 104.
Cluster Munition Monitor 2010, p. 13.
 The most common cluster munition used was the CBU-24, or Sadeye, a cigar-shaped dispenser containing nearly 700 one-pound spherical BLU-26 submunitions, each of which was filled with 300 steel ball bearings. Although Air Force officials claimed the Sadeye was designed for suppressing enemy anti-aircraft installations, a secretary of the Navy characterized it as a “large area antipersonnel weapon.” Vietnamese civilians called the BLU-26 submunitions “guavas.” In addition to the CBU-24, US forces made extensive use of CBU-2/A and CBU-14 cluster munitions, containing the antipersonnel BLU-3 submunition, which dispersed 255 steel balls and which the Vietnamese civilians dubbed the “pineapple.” The submunitions in a Rockeye targeted armored vehicles with a shaped charge but also had antipersonnel effects when the submunitions’ metal case fragmented. The largest US cluster munition employed during the Southeast Asia conflict was the Hayes dispenser, a boxlike aluminum device capable of holding 12,744 BLU-26 or 38,520 M40 submunitions. Michael Krepon, “Weapons Potentially Inhumane: The Case of Cluster Bombs,” Foreign Affairs, vol. 52 (1974), p. 597; Prokosch, The Technology of Killing, pp. 81, 85, 98-99, 101-102, 105; Prokosch, Cluster Weapons, p. 4.
 Stockholm International Peace Research Institute, Anti-personnel Weapons, p. 29.
 Krepon, “Weapons Potentially Inhumane,” Foreign Affairs, p. 595.
 ICRC, “Explosive Remnants of War: The Lethal Legacy of Modern Armed Conflict,” July 2003, www.icrc.org/Web/Eng/siteeng0.nsf/htmlall/p0828/$File/ICRC_002_0828.PDF!Open (accessed September 18, 2010), p. 6.
 Unexploded submunitions killed or injured 33 people in Laos in 2009. Landmine and Cluster Munition Monitor, “Lao PDR: Casualties,” October 13, 2010, http://www.the-monitor.org/index.php/cp/display/region_profiles/theme/330#_ftnref11 (accessed November 1, 2010).
 Prokosch, The Technology of Killing, pp. 134-135.
 In 1967, British philosopher Bertrand Russell and a group of writers, scientists, and civic leaders organized a panel to investigate the US military conduct in Vietnam. The panel, which called itself an International War Crimes Tribunal despite having no formal legal power or state sponsorship, sent field investigators to Indochina to report on cluster munition attacks and visit cluster munition victims. Two sets of hearings were held, one in Stockholm and one in Copenhagen. The US government, however, declined to attend the Tribunal’s hearings, and ultimately the Russell-led initiative received little coverage in the world media. Prokosch, The Technology of Killing, p. 93.
 Certainly, the weapons succeeded in destroying many anti-aircraft installations and their crews. “Guavas,” however, became important propaganda tools for the North Vietnamese, who would present the submunitions to visitors as symbols of US brutality. As information about the weapons spread, cluster munitions thus fueled opposition to the war both domestically and abroad. Krepon, “Weapons Potentially Inhumane,” Foreign Affairs, pp. 603-604; Prokosch, The Technology of Killing, pp. 112-114 (discussing Chief Warrant Officer Swearington, Staff Study on Pernicious Characteristics of US Explosive Ordnance (Washington: US Marine Corps, 1969)).
 With regard to US secrecy, see Prokosch, Technology of Killing, p. 90; Krepon, “Weapons Potentially Inhumane,” Foreign Affairs, p. 600.
 Krepon, “Weapons Potentially Inhumane,” Foreign Affairs, p. 604. Krepon’s analysis includes the following models: CBU-25, CBU-55, CBU-58, MK-20 (Rockeye), and BLU-32.
 For a complete list of cluster munition use, see timeline reprinted in this chapter from Cluster Munition Monitor 2010, pp. 13-14.
 Israel’s use may have violated an agreement not to use US-supplied cluster munitions in populated areas. The United States imposed a moratorium on exports of cluster munitions to Israel in July 1982 but quietly lifted it in November 1988. For more information, see Human Rights Watch, Flooding South Lebanon: Israel’s Use of Cluster Munitions in Lebanon in July and August 2006, vol. 20, no. 2(E), February 2008, http://www.hrw.org/reports/2008/lebanon0208/, pp. 102-103.
Cluster Munition Monitor 2010, p. 13.
Cluster Munition Monitor 2010, pp. 13-14.
 Human Rights Watch, Fatally Flawed: Cluster Bombs and Their Use by the United States in Afghanistan, vol. 14, no. 7(G), December 2002, http://www.hrw.org/reports/2002/us-afghanistan/, p. 40. For more information on cluster bomb use in the Gulf War, see Human Rights Watch, Ticking Time Bombs: NATO’s Use of Cluster Munitions in Yugoslavia, vol. 11, no. 6(D), June 1999, http://www.hrw.org/reports/1999/nato2/index.htm#TopOfPage, pp. 8-12; Human Rights Watch, Cluster Bombs in Afghanistan, October 2001, http://www.hrw.org/en/reports/2001/10/31/cluster-bombs-afghanistan, pp. 5-6.
Cluster Munition Monitor 2010, p. 13; Human Rights Watch, Fatally Flawed, p. 40. The new model was the CBU-87 cluster munition; other, less reliable Vietnam War-era types included the Rockeye and CBU-52, CBU-58, and CBU-71.
 Colin King, commissioned by the ICRC, Explosive Remnants of War: A Study on Submunitions and Other Unexploded Ordnance, August 2000, p. 16 (citing Donald Kennedy and William Kincheloe, “Steel Rain: Submunitions,” U.S. Army Journal, (January 1993)).
 U.S. Air Force, Gulf War Air Power Survey, vol. IV, pt. I (1993), pp. 290, 231 [hereinafter GWAPS] (cited in Human Rights Watch, Fatally Flawed, p. 40).
 Ibid., pp. 43, 48 (cited in Human Rights Watch, Fatally Flawed, p. 40).
 Pilots also had more experience with delivery of cluster munitions closer to the ground. Ibid., vol. II, pt. II, pp. 162-63 (cited in Human Rights Watch, Fatally Flawed, p. 40). The new CBU-87 fared better because of its radar proximity fuze and a spin mechanism that controlled dispersal, but it still had a lower accuracy rate than designed. Murray Hamrick, “Aerial Views: USAF Air-to-Air Combat,” International Defense Review (July 1991), p. 743 (cited in Human Rights Watch, Fatally Flawed, p. 40).
 Human Rights Watch, Fatally Flawed, p. 40.
 GWAPS, vol. II, pt. I, p. 261 (cited in Human Rights Watch, Fatally Flawed, p. 40).
 Human Rights Watch, Fatally Flawed, pp. 40-41.
In addition to being less cautious in battlefield areas, children were attracted by the colorful submunitions, which one reporter described as resembling “white lawn darts, green baseballs, [and] orange-striped soda cans.” James Vincent Brady, “Kuwaitis Dying from Old Menace: Unexploded Bombs,” Forth Worth Star-Telegram, January 12, 1992, p. 1 (quoted in Human Rights Watch, Fatally Flawed, p. 41).
 Iraqi authorities said that they removed tens of thousands of submunitions from such areas. Human Rights Watch, Fatally Flawed, p. 41.
 The General Accounting Office reported, that in some cases, “ground movement came to a halt because units were afraid of encountering unexploded ordnance.” US General Accounting Office, “Operation Desert Storm: Casualties Caused by Improper Handling of Unexploded U.S. Submunitions,” GAO/NSIAD-93-212, August 1993, p. 9 (cited in Human Rights Watch, Fatally Flawed, p. 41).
 Patrick J. Sloyan, “U.S. Bomblets Killed 14 Americans in Gulf War,” Newsday, September 19, 1991, p. 4; Brady, “Kuwaitis Dying from Old Menace,” Fort Worth Star-Telegram, p. 1 (both cited in Human Rights Watch, Fatally Flawed, p. 41).
The Eritrea Ethiopia Claims Commission found Eritrea liable for two strikes in Mekele, including one on a school, on the grounds that it failed to take all feasible precautions to minimize civilian casualties. Virgil Wiebe, “For Whom the Little Bells Toll: Recent Judgments by International Tribunals on the Legality of Cluster Munitions,” University of St. Thomas School of Law, Legal Studies Research Paper No. 07-23 (2007), pp. 11-12. For more information on the use of cluster munitions by Ethiopia and Eritrea, see also McGrath, Cluster Bombs, p. 41; Handicap International, Circle of Impact: The Fatal Footprint of Cluster Munitions on People and Communities (Brussels: Handicap International, 2007), http://en.handicapinternational.be/Circle-of-Impact--report-on-the-human-impact-of-cluster-bombs_a467.html (accessed October 30, 2010), pp. 50-51.
Cluster Munition Monitor 2010, pp. 13-14.
 All but 24 of the documented casualties came during strikes, not afterwards. Many post-conflict casualties, however, may not have been reported. Handicap International, Circle of Impact, p. 85.
 For information on other incidents in Chechnya, see Mennonite Central Committee, “Clusters of Death: Global Report on Cluster Bomb Production and Use,” 2000, http://clusterbombs.mcc.org/clusterbombs/resources/research/death/ (accessed October 30, 2010), chap. 3.
 HALO Trust, a UK-based demining organization, estimated that 137 people were killed while Chechen President Aslan Maskhadov alleged that 282 died in the attack. Ibid.
 Yugoslav forces and a non-state armed group used cluster munitions. NATO aircraft dropped two CBU-87 cluster bombs due to an apparent “operational miscommunication” in violation of its own regulations. Handicap International, Circle of Impact, pp. 60-61.
Prosecutor v. Milan Martic, International Tribunal for the Former Yugoslavia, Case No. IT-95-11-T, Judgment (Trial Chamber I), June 12, 2007, http://www.icty.org/x/cases/martic/tjug/en/070612.pdf (accessed September 9, 2010), paras. 456-463, 470-471, 518.
Cluster Munition Monitor 2010, p. 14.
 These M85 submunitions were delivered in L20A1 artillery projectiles, carrying 49 M85s each. Human Rights Watch, Off Target: The Conduct of the War and Civilian Casualties in Iraq (New York: Human Rights Watch, 2003), http://www.hrw.org/reports/2003/usa1203/, p. 85.
 Human Rights Watch telephone interview with Col. Gil Baldwin, commanding officer, First Queen’s
Dragoon Guards, Cardiff, Wales, July 2, 2003.
 Military experts from numerous countries that stockpile the M85 or variants of it have made this claim in discussions with Human Rights Watch during meetings related to the Convention on Conventional Weapons (CCW) in recent years. For a discussion of the tests that produced these failure rates, see Human Rights Watch, Flooding South Lebanon, pp. 30-31.
 Email communication from Dalya Farran, media and post clearance officer, Mine Action Coordination Center–South Lebanon (MACC SL) to Human Rights Watch, January 15, 2008.
 C King Associates, Norwegian Defence Research Establishment, and Norwegian People’s Aid, M85: An Analysis of Reliability (Norway: Norwegian People’s Aid, 2007), http://www.npaid.org/?module=Articles;action=Article.publicShow;ID=5662 (accessed September 18, 2010), pp. 6, 15.
 Chris Clark, program manager, MACC SL, “Unexploded Cluster Bombs and Submunitions in South Lebanon: Reliability from a Field Perspective,” paper presented at ICRC Expert Meeting, Montreux, Switzerland, April 18-20, 2007.
 Human Rights Watch, Fatally Flawed, p. 6.
 The accuracy of air-launched munitions is measured by “circular error probable” (CEP), which is “the radius of a circle within which half of a missile's projectiles are expected to fall.” US Department of Defense, “Dictionary of Military Terms,” http://www.dtic.mil/doctrine/dod_dictionary/ (accessed September 18, 2010). The WCMD is designed to have a CEP of 85 feet (26 meters), according to manufacturer specifications. Robert Hewson, ed., Jane’s Air-Launched Weapons, Issue 44 (Surrey, UK: Jane’s Information Group Limited, 2004), pp. 544-545; Directory of US Military Rockets and Missiles, “WCMD,” http://www.designation-systems.net/dusrm/app5/wcmd.html (accessed September 18, 2010).
 Frank Wolfe, "Air Force Employing WCMD, Flex Targeting on B-52s, Linking Predator and AC-130s," Defense Daily, November 28, 2001 (cited in Human Rights Watch, Fatally Flawed, p. 6).
 In Afghanistan, for example, the US Navy used a small number of Joint Standoff Weapon-A (JSOW-A) cluster munitions, also known as AGM-154 cluster munitions, which carried 145 BLU-97 submunitions. Human Rights Watch, Fatally Flawed, p. 1. US forces used a small but uncounted number of JSOW cluster munitions and precision-guided BGM-109D Tomahawk Land Attack Missiles-D (TLAM-Ds) containing 145 and 166 BLU-97 submunitions, respectively, in Iraq in 2003. Lt. Gen. T. Michael Moseley, US Air Force, “Operation Iraqi Freedom: By the Numbers,” April 30, 2003, http://www.globalsecurity.org/military/library/report/2003/uscentaf_oif_report_30apr2003.pdf (accessed September 18, 2010), p. 11.
 Human Rights Watch, Off Target, pp. 60-61. CBU-105 cluster munitions without the WCMD are referred to as CBU-97 cluster munitions. In Iraq, the United States also used 121 artillery shells with Sense and Destroy Armor Munitions (SADARMs), which are similar to CBU-105s but have only two submunitions. For details on these weapons types, see Human Rights Watch, Off Target, pp. 82, 84. SADARMs are not considered cluster munitions under the definition established by the 2008 Convention on Cluster Munitions. See Convention on Cluster Munitions, adopted May 30, 2008, Diplomatic Conference for the Adoption of a Convention on Cluster Munitions, CCM/77, entered into force August 1, 2010, art. 2(2), http://www.clustermunitionsdublin.ie/pdf/ENGLISHfinaltext.pdf (accessed October 6, 2010); Human Rights Watch, Twelve Facts and Fallacies about the Convention on Cluster Munitions, April 2009, http://www.hrw.org/en/news/2009/04/14/twelve-facts-and-fallacies-about-convention-cluster-munitions, p. 3.
Textron Defense Systems, “BLU-108 Submunition: Sensor Fuzed Submunition with Clean Battlefield Protection,” 2008, p. 2, http://www.textrondefense.com/pdfs/datasheets/blu108_datasheet.pdf (accessed October 28, 2010); “CBU-97/B Sensor Fuzed Weapon System (SFW) (with BLU-108),” in Duncan Lennox, ed., Jane’s Air Launched Weapons (Surrey, UK: Jane’s Information Group Limited, 1999).
 According to the manufacturer, Textron Defense Systems, “The first two modes enable the Skeet to self destruct after eight seconds from launch or within a 50-foot altitude above the ground. The Skeet’s third feature is a time-out device that will yield the warhead inert within minutes of hitting the ground.” Textron Defense Systems, “Delivering Confidence: Sensor Fuzed Weapon,” 2010, http://www.textrondefense.com/pdfs/datasheets/sfw_datasheet.pdf (accessed October 28, 2010), p. 1. See also Textron Defense Systems, “BLU-108 Submunition,” p. 1.
 Textron Defense Systems, “Delivering Confidence,” p. 1.
 Landmine Action, Cluster Munitions in Kosovo: Analysis of Use, Contamination and Casualties (London: Landmine Action, 2007), http://www.landmineaction.org/resources/Cluster%20Munitions%20in%20Kosovo.pdf (accessed September 18, 2010), p. 9 (citing data provided by NATO to the UN Mission in Kosovo (UNMIK)); Human Rights Watch, Fatally Flawed, p. 1; Moseley, “Operation Iraqi Freedom—By the Numbers,” p. 11.
 Human Rights Watch, Memorandum to Delegates to the April Prepcom for the 2001 Review Conference for the Convention on Conventional Weapons, April 2, 2001, http://www.hrw.org/press/2001/04/clusterbombs0402.htm; Landmine Action, Cluster Munitions in Kosovo, p. 9 (citing data provided by NATO to UNMIK); Human Rights Watch, Off Target, p. 61; UK Ministry of Defence, “Operations in Iraq: First Reflections,” July 2003, http://www.globalsecurity.org/military/library/report/2003/iraq2003operations_ukmod_july03.pdf (accessed September 18, 2010), p. 24.
 Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 18, 2008.
 Human Rights Watch interview with Allan Poston, chief technical advisor, National Demining Office, UN Development Program (UNDP), Beirut, November 29, 2006; Presentation by Chris Clark, program manager, MACC SL, to CCW Delegates, Geneva, August 30, 2006, notes by Human Rights Watch.