Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia
Table of Contents
36. Where objects are concerned, the definition has two elements: (a) their nature, location, purpose or use must make an effective contribution to military action, and (b) their total or partial destruction, capture or neutralization must offer a definite military advantage in the circumstances ruling at the time. Although this definition does not refer to persons, in general, members of the armed forces are considered combatants, who have the right to participate directly in hostilities, and as a corollary, may also be attacked.
37. The definition is supposed to provide a means whereby informed objective observers (and decision makers in a conflict) can determine whether or not a particular object constitutes a military objective. It accomplishes this purpose in simple cases. Everyone will agree that a munitions factory is a military objective and an unoccupied church is a civilian object. When the definition is applied to dual-use objects which have some civilian uses and some actual or potential military use (communications systems, transportation systems, petrochemical complexes, manufacturing plants of some types), opinions may differ. The application of the definition to particular objects may also differ depending on the scope and objectives of the conflict. Further, the scope and objectives of the conflict may change during the conflict.
38. Using the Protocol I definition and his own review of state practice, Major General A.P.V. Rogers, a former Director of British Army Legal Services has advanced a tentative list of military objectives:
The list was not intended to be exhaustive. It remains a requirement that both elements of the definition must be met before a target can be properly considered an appropriate military objective.
39. In 1956, the International Committee of the Red Cross (ICRC) drew up the following proposed list of categories of military objectives:
I. The objectives belonging to the following categories are those considered to be of generally recognized military importance:
II. The following however, are excepted from the foregoing list:
The above list will be reviewed at intervals of not more than ten years
by a group of Experts composed of persons with a sound grasp of military
strategy and of others concerned with the protection of the civilian
40. The Protocol I definition of military objective has been criticized by W. Hays Parks, the Special Assistant for Law of War Matters to the U.S. Army Judge Advocate General as being focused too narrowly on definite military advantage and paying too little heed to war sustaining capability, including economic targets such as export industries. (W. Hays Parks, "Air War and the Law of War," 32 A.F.L. Rev. 1, 135-45 (1990)). On the other hand, some critics of Coalition conduct in the Gulf War have suggested that the Coalition air campaign, directed admittedly against legitimate military objectives within the scope of the Protocol I definition, caused excessive long-term damage to the Iraqi economic infrastructure with a consequential adverse effect on the civilian population. (Middle East Watch, Needless Deaths in the Gulf War: Civilian Casualties during the Air Campaign and Violations of the Laws of War (1991); Judith G. Gardam, "Proportionality and Force in International Law," 87 Am. J. Int'l L. 391, 404-10 (1993)).
41. This criticism has not gone unexplored. Françoise Hampson, a British scholar, has suggested a possible refinement of the definition:
42. Although the Protocol I definition of military objective is not beyond criticism, it provides the contemporary standard which must be used when attempting to determine the lawfulness of particular attacks. That being said, it must be noted once again neither the USA nor France is a party to Additional Protocol I. The definition is, however, generally accepted as part of customary law.
43. To put the NATO campaign in context, it is instructive to look briefly at the approach to the military objective concept in history of air warfare. The Protocol I standard was not applicable during World War II. The bomber offensives conducted during that war were conducted with technological means which rendered attacks on targets occupying small areas almost impossible. In general, depending upon the period in the conflict, bomber attacks could be relied upon, at best, to strike within 5 miles, 2 miles or 1 mile of the designated target. The mission for the US/UK Combined Bomber Offensive from the UK was:
"To conduct a joint United States-British air offensive to accomplish the progressive destruction and dislocation of the German military, industrial and economic system, and the undermining of the morale of the German people to a point where their capacity for armed resistance is fatally weakened. This is construed as meaning so weakened as to permit initiation of final combined operations on the Continent."
(A. Verrier, The Bomber Offensive (1968) 330).
The principal specific objectives of the offensive were designated as:
construction yards and bases.
Notwithstanding the designation of specific targets and the attempt, at least by US Army Air Force commanders on occasion, to conduct a precision bombing campaign, for the most part World War II bombing campaigns were aimed at area targets and intended, directly or indirectly, to affect the morale of the enemy civilian population. It is difficult to describe the fire bombing of Hamburg, Dresden and Tokyo as anything other than attacks intended to kill, terrorize or demoralize civilians. Whether or not these attacks could be justified legally in the total war context of the time, they would be unlawful if they were required to comply with Protocol I.
44. Technology, law, and the public consensus of what was acceptable, at least in demonstrably limited conflicts, had evolved by the time of the 1990-91 Gulf Conflict. Technological developments, such as precision guided munitions, and the rapid acquisition of control of the aerospace by coalition air forces significantly enhanced the precision with which targets could be attacked.
Target sets used during the Gulf Conflict were:
"Leadership; Command, Control, and Communications; Strategic Air Defenses; Airfields; Nuclear, Biological, and Chemical Research and Production; Naval Forces and Port Facilities; Military Storage and Production; Railroads and Bridges, Electrical Power; and Oil Refining and Distribution Facilities. Schwarzkopf added the Republican Guard as a category and Scuds soon emerged as a separate target set. After the beginning of Desert Storm, two more categories appeared: fixed surface-to-air missile sites in the KTO and breaching sites for the ground offensive."
(W. Murray, Air War in the Persian Gulf (1995) 32)
45. In the words of the Cohen, Shelton Joint Statement on Kosovo given to the US Senate:
46. The NATO Internet Report Kosovo One Year On (http://www.nato.int/kosovo/repo 2000, 21 Mar 00) described the targets as:
47. Most of the targets referred to in the quotations above are clearly military objectives. The precise scope of "military-industrial infrastructure, media and other strategic targets" as referred to in the US statement and "government ministries and refineries" as referred to in the NATO statement is unclear. Whether the media constitutes a legitimate target group is a debatable issue. If the media is used to incite crimes, as in Rwanda, then it is a legitimate target. If it is merely disseminating propaganda to generate support for the war effort, it is not a legitimate target.
48. The main problem with the principle of proportionality is not whether or not it exists but what it means and how it is to be applied. It is relatively simple to state that there must be an acceptable relation between the legitimate destructive effect and undesirable collateral effects. For example, bombing a refugee camp is obviously prohibited if its only military significance is that people in the camp are knitting socks for soldiers. Conversely, an air strike on an ammunition dump should not be prohibited merely because a farmer is plowing a field in the area. Unfortunately, most applications of the principle of proportionality are not quite so clear cut. It is much easier to formulate the principle of proportionality in general terms than it is to apply it to a particular set of circumstances because the comparison is often between unlike quantities and values. One cannot easily assess the value of innocent human lives as opposed to capturing a particular military objective.
49. The questions which remain unresolved once one decides to apply the principle of proportionality include the following:
a) What are the relative values to be assigned to the military advantage gained and the injury to non-combatants and or the damage to civilian objects?
b) What do you include or exclude in totaling your sums?
c) What is the standard of measurement in time or space? and
d) To what extent is a military commander obligated to expose his own forces to danger in order to limit civilian casualties or damage to civilian objects?
50. The answers to these questions are not simple. It may be necessary to resolve them on a case by case basis, and the answers may differ depending on the background and values of the decision maker. It is unlikely that a human rights lawyer and an experienced combat commander would assign the same relative values to military advantage and to injury to noncombatants. Further, it is unlikely that military commanders with different doctrinal backgrounds and differing degrees of combat experience or national military histories would always agree in close cases. It is suggested that the determination of relative values must be that of the "reasonable military commander". Although there will be room for argument in close cases, there will be many cases where reasonable military commanders will agree that the injury to noncombatants or the damage to civilian objects was clearly disproportionate to the military advantage gained.
51. Much of the material submitted to the OTP consisted of reports that civilians had been killed, often inviting the conclusion to be drawn that crimes had therefore been committed. Collateral casualties to civilians and collateral damage to civilian objects can occur for a variety of reasons. Despite an obligation to avoid locating military objectives within or near densely populated areas, to remove civilians from the vicinity of military objectives, and to protect their civilians from the dangers of military operations, very little prevention may be feasible in many cases. Today’s technological society has given rise to many dual use facilities and resources. City planners rarely pay heed to the possibility of future warfare. Military objectives are often located in densely populated areas and fighting occasionally occurs in such areas. Civilians present within or near military objectives must, however, be taken into account in the proportionality equation even if a party to the conflict has failed to exercise its obligation to remove them.
52. In the Kupreskic Judgment (Case No: IT-95-16-T 14 Jan 2000) the Trial Chamber addressed the issue of proportionality as follows:
This formulation in Kupreskic can be regarded as a progressive statement of the applicable law with regard to the obligation to protect civilians. Its practical import, however, is somewhat ambiguous and its application far from clear. It is the committee’s view that where individual (and legitimate) attacks on military objectives are concerned, the mere cumulation of such instances, all of which are deemed to have been lawful, cannot ipso facto be said to amount to a crime. The committee understands the above formulation, instead, to refer to an overall assessment of the totality of civilian victims as against the goals of the military campaign.
53. In its report, Civilian Deaths in the NATO Air Campaign, Human Rights Watch documented some 500 civilian deaths in 90 separate incidents. It concluded: "on the basis available on these ninety incidents that as few as 488 and as many as 527 Yugoslav civilians were killed as a result of NATO bombing. Between 62 and 66 percent of the total registered civilian deaths occurred in just twelve incidents. These twelve incidents accounted for 303 to 352 civilian deaths. These were the only incidents among the ninety documented in which ten or more civilian deaths were confirmed." Ten of these twelve incidents were included among the incidents which were reviewed with considerable care by the committee (see para. 9 above) and our estimate was that between 273 and 317 civilians were killed in these ten incidents. Human Rights Watch also found the FRY Ministry of Foreign Affairs publication NATO Crimes in Yugoslavia to be largely credible on the basis of its own filed research and correlation with other sources. A review of this publication indicates it provides an estimated total of approximately 495 civilians killed and 820 civilians wounded in specific documented instances. For the purposes of this report, the committee operates on the basis of the number of persons allegedly killed as found in both publications. It appears that a figure similar to both publications would be in the range of 500 civilians killed.
54. During the bombing campaign, NATO aircraft flew 38,400 sorties, including 10,484 strike sorties. During these sorties, 23, 614 air munitions were released (figures from NATO). As indicated in the preceding paragraph, it appears that approximately 500 civilians were killed during the campaign. These figures do not indicate that NATO may have conducted a campaign aimed at causing substantial civilian casualties either directly or incidentally.
55. The choice of targets by NATO (see paras. 38 and 39 above) includes some loosely defined categories such as military-industrial infrastructure and government ministries and some potential problem categories such as media and refineries. All targets must meet the criteria for military objectives (see para. 28-30 above). If they do not do so, they are unlawful. A general label is insufficient. The targeted components of the military-industrial infrastructure and of government ministries must make an effective contribution to military action and their total or partial destruction must offer a definite military advantage in the circumstances ruling at the time. Refineries are certainly traditional military objectives but tradition is not enough and due regard must be paid to environmental damage if they are attacked (see paras. 14-25 above). The media as such is not a traditional target category. To the extent particular media components are part of the C3 (command, control and communications) network they are military objectives. If media components are not part of the C3 network then they may become military objectives depending upon their use. As a bottom line, civilians, civilian objects and civilian morale as such are not legitimate military objectives. The media does have an effect on civilian morale. If that effect is merely to foster support for the war effort, the media is not a legitimate military objective. If the media is used to incite crimes, as in Rwanda, it can become a legitimate military objective. If the media is the nerve system that keeps a war-monger in power and thus perpetuates the war effort, it may fall within the definition of a legitimate military objective. As a general statement, in the particular incidents reviewed by the committee, it is the view of the committee that NATO was attempting to attack objects it perceived to be legitimate military objectives.
56. The committee agrees there is nothing inherently unlawful about flying above the height which can be reached by enemy air defences. However, NATO air commanders have a duty to take practicable measures to distinguish military objectives from civilians or civilian objectives. The 15,000 feet minimum altitude adopted for part of the campaign may have meant the target could not be verified with the naked eye. However, it appears that with the use of modern technology, the obligation to distinguish was effectively carried out in the vast majority of cases during the bombing campaign.
57. In the course of its review, the committee did not come across any incident which, in its opinion, required investigation by the OTP. The five specific incidents discussed below are those which, in the opinion of the committee, were the most problematic. The facts cited in the discussion of each specific incident are those indicated in the information within the possession of the OTP at the time of its review.
58. On 12 April 1999, a NATO aircraft launched two laser guided bombs at the Leskovac railway bridge over the Grdelica gorge and Juzna Morava river, in eastern Serbia. A 5-carriage passenger train, travelling from Belgrade to Ristovac on the Macedonian border, was crossing the bridge at the time, and was struck by both missiles. The various reports made of this incident concur that the incident occurred at about 11.40 a.m. At least ten people were killed in this incident and at least 15 individuals were injured. The designated target was the railway bridge, which was claimed to be part of a re-supply route being used for Serb forces in Kosovo. After launching the first bomb, the person controlling the weapon, at the last instant before impact, sighted movement on the bridge. The controller was unable to dump the bomb at that stage and it hit the train, the impact of the bomb cutting the second of the passenger coaches in half. Realising the bridge was still intact, the controller picked a second aim point on the bridge at the opposite end from where the train had come and launched the second bomb. In the meantime the train had slid forward as a result of the original impact and parts of the train were also hit by the second bomb.
59. It does not appear that the train was targeted deliberately. US Deputy Defense Secretary John Hamre stated that "one of our electro-optically guided bombs homed in on a railroad bridge just when a passenger train raced to the aim point. We never wanted to destroy that train or kill its occupants. We did want to destroy the bridge and we regret this accident." The substantive part of the explanation, both for the failure to detect the approach of the passenger train and for firing a second missile once it had been hit by the first, was given by General Wesley Clark, NATO’s Supreme Allied Commander for Europe and is here reprinted in full:
General Clark then showed the cockpit video of the plane which fired on the bridge:
60. Some doubt has since been cast on this version of events by a comprehensive technical report submitted by a German national, Mr Ekkehard Wenz, which queries the actual speed at which the events took place in relation to that suggested by the video footage of the incident released by NATO. The effect of this report is to suggest that the reaction time available to the person controlling the bombs was in fact considerably greater than that alleged by NATO. Mr. Wenz also suggests the aircraft involved was an F15E Strike Eagle with a crew of two and with the weapons being controlled by a Weapons Systems Officer (WSO) not the pilot.
61. The committee has reviewed both the material provided by NATO and the report of Mr. Wenz with considerable care. It is the opinion of the committee that it is irrelevant whether the person controlling the bomb was the pilot or the WSO. Either person would have been travelling in a high speed aircraft and likely performing several tasks simultaneously, including endeavouring to keep the aircraft in the air and safe from surrounding threats in a combat environment. If the committee accepts Mr. Wenz’s estimate of the reaction time available, the person controlling the bombs still had a very short period of time, less than 7 or 8 seconds in all probability, to react. Although Mr Wenz is of the view that the WSO intentionally targeted the train, the committee’s review of the frames used in the report indicates another interpretation is equally available. The cross hairs remain fixed on the bridge throughout, and it is clear from this footage that the train can be seen moving toward the bridge only as the bomb is in flight: it is only in the course of the bomb’s trajectory that the image of the train becomes visible. At a point where the bomb is within a few seconds of impact, a very slight change to the bomb aiming point can be observed, in that it drops a couple of feet. This sequence regarding the bomb sights indicates that it is unlikely that the WSO was targeting the train, but instead suggests that the target was a point on the span of the bridge before the train appeared.
62. It is the opinion of the committee that the bridge was a legitimate military objective. The passenger train was not deliberately targeted. The person controlling the bombs, pilot or WSO, targeted the bridge and, over a very short period of time, failed to recognize the arrival of the train while the first bomb was in flight. The train was on the bridge when the bridge was targeted a second time and the bridge length has been estimated at 50 meters (Wenz study para 6 g above at p.25). It is the opinion of the committee that the information in relation to the attack with the first bomb does not provide a sufficient basis to initiate an investigation. The committee has divided views concerning the attack with the second bomb in relation to whether there was an element of recklessness in the conduct of the pilot or WSO. Despite this, the committee is in agreement that, based on the criteria for initiating an investigation (see para. 5 above), this incident should not be investigated. In relation to whether there is information warranting consideration of command responsibility, the committee is of the view that there is no information from which to conclude that an investigation is necessary into the criminal responsibility of persons higher in the chain of command. Based on the information available to it, it is the opinion of the committee that the attack on the train at Grdelica Gorge should not be investigated by the OTP.
ESS Home Page
Kosovo Index Page
Holocaust Index Page
Genocide Index Page