International law determines that when one attacks legitimate military targets one can inflict collateral damage as long as this damage is proportional in terms of military necessity. Following the war in Kosovo, complaints were directed against NATO regarding the indiscriminate use of force. For example, one accusation was that NATO forces struck a television broadcasting station. All sides admitted that the station was deliberately bombed. Some 10 to 17 civilians were killed in the attack. The question that was asked at a special tribunal that examined the issue was whether this was a legitimate target, because the result was the suspension of broadcasts for only a few hours. The verdict was that the attack was proportional to the objective of silencing the station's activity for a few brief hours. The question of proportionality is something that is difficult to quantify; nonetheless, the aforementioned precedent illustrates what is acceptable under international law. In Gaza, no attacks took place that even approached these ratios.This seemed interesting, so I found the actual report. JCPA is oversimplifying NATO's stated objectives and justifications a bit, but the incident is still very relevant:
Amnesty International disagreed.
71. On 23 April 1999, at 0220, NATO intentionally bombed the central studio of the RTS (state-owned) broadcasting corporation at 1 Aberdareva Street in the centre of Belgrade. The missiles hit the entrance area, which caved in at the place where the Aberdareva Street building was connected to the Takovska Street building. While there is some doubt over exact casualty figures, between 10 and 17 people are estimated to have been killed.
72. The bombing of the TV studio was part of a planned attack aimed at disrupting and degrading the C3 (Command, Control and Communications) network. In co-ordinated attacks, on the same night, radio relay buildings and towers were hit along with electrical power transformer stations. At a press conference on 27 April 1999, NATO officials justified this attack in terms of the dual military and civilian use to which the FRY communication system was routinely put, describing this as a
"very hardened and redundant command and control communications system [which …] uses commercial telephone, […] military cable, […] fibre optic cable, […] high frequency radio communication, […] microwave communication and everything can be interconnected. There are literally dozens, more than 100 radio relay sites around the country, and […] everything is wired in through dual use. Most of the commercial system serves the military and the military system can be put to use for the commercial system […]."
Accordingly, NATO stressed the dual-use to which such communications systems were put, describing civilian television as "heavily dependent on the military command and control system and military traffic is also routed through the civilian system" (press conference of 27 April, ibid).
73. At an earlier press conference on 23 April 1999, NATO officials reported that the TV building also housed a large multi-purpose communications satellite antenna dish, and that "radio relay control buildings and towers were targeted in the ongoing campaign to degrade the FRY’s command, control and communications network". In a communication of 17 April 1999 to Amnesty International, NATO claimed that the RTS facilities were being used "as radio relay stations and transmitters to support the activities of the FRY military and special police forces, and therefore they represent legitimate military targets" (Amnesty International Report, NATO/Federal Republic of Yugoslavia: Violations of the Laws of War by NATO during Operation Allied Force, June 2000, p. 42).
75. NATO intentionally bombed the Radio and TV station and the persons killed or injured were civilians. The questions are: was the station a legitimate military objective and; if it was, were the civilian casualties disproportionate to the military advantage gained by the attack? .... Insofar as the attack actually was aimed at disrupting the communications network, it was legally acceptable.
77. Assuming the station was a legitimate objective, the civilian casualties were unfortunately high but do not appear to be clearly disproportionate....
78. Assuming the RTS building to be a legitimate military target, it appeared that NATO realised that attacking the RTS building would only interrupt broadcasting for a brief period. Indeed, broadcasting allegedly recommenced within hours of the strike, thus raising the issue of the importance of the military advantage gained by the attack vis-à-vis the civilian casualties incurred. The FRY command and control network was alleged by NATO to comprise a complex web and that could thus not be disabled in one strike. As noted by General Wesley Clark, NATO "knew when we struck that there would be alternate means of getting the Serb Television. There’s no single switch to turn off everything but we thought it was a good move to strike it and the political leadership agreed with us" (ibid, citing "Moral combat, NATO at War," broadcast on BBC2 on 12 March 2000). At a press conference on 27 April 1999, another NATO spokesperson similarly described the dual-use Yugoslav command and control network as "incapable of being dealt with in "a single knock-out blow (ibid)." The proportionality or otherwise of an attack should not necessarily focus exclusively on a specific incident. (See in this regard para. 52, above, referring to the need for an overall assessment of the totality of civilian victims as against the goals of the military campaign). With regard to these goals, the strategic target of these attacks was the Yugoslav command and control network. The attack on the RTS building must therefore be seen as forming part of an integrated attack against numerous objects, including transmission towers and control buildings of the Yugoslav radio relay network which were "essential to Milosevic’s ability to direct and control the repressive activities of his army and special police forces in Kosovo" (NATO press release, 1 May 1999) and which comprised "a key element in theYugoslav air-defence network" (ibid, 1 May 1999). Attacks were also aimed at electricity grids that fed the command and control structures of the Yugoslav Army (ibid, 3 May 1999). ... Not only were these targets central to the Federal Republic of Yugoslavia’s governing apparatus, but formed, from a military point of view, an integral part of the strategic communications network which enabled both the military and national command authorities to direct the repression and atrocities taking place in Kosovo (ibid, 21 April 1999).
79. On the basis of the above analysis and on the information currently available to it, the committee recommends that the OTP not commence an investigation related to the bombing of the Serbian TV and Radio Station.
An interesting analysis of the entire question of disproportionality in international law can be seen in this paper. Although this was written under the auspices of the IDF, the paper appears to be as objective as possible; it is quite critical of the vagueness of the current "disproportionality" rules by saying that they give free rein to military commanders. The point of the paper was to begin to find an effective way to objectively calculate proportionality. Here is her synopsis of the RTS incident:
NATO aircraft attacked the RTS television and radio studios in Central Belgrade, killing sixteen civilians.184 The discussions of this incident revolved mainly around questions of targeting and distinction, as the military nature of the studios was controversial.185 However, the question of proportionality was raised as well, as the result of the attack was only (it appears) a brief interruption in the studios’ broadcasting, whereas the collateral damage amounted to sixteen civilians killed, and a further sixteen wounded.186 According to Laursen, one ought not to make too much of this, as it is difficult to tell what effect this disruption had on the military communications of RTS.187 This, of course, is the perennial problem: we never know exactly what the military advantage was, so we are always in effect missing half of the equation.By any measure, however, the IDF seems to have gone way beyond NATO in 1999 (as well as the US in the first Gulf war, which was also touched on in the paper) in attempting to avoid civilian casualties. And almost certainly no purposeful IDF actions in Gaza approached the ratio of casualties to military gain that NATO's bombing of RTS did. (There is information in the three reports about accidental casualties, which would be an interesting topic on its own.)
Amnesty argues that the attack was disproportionate.188 The prosecutor’s report, in contrast, concludes that the civilian casualties were high, but not disproportionate.189 In its discussion of the law regarding target selection, the report states that proportionality must be assessed on a “case by case basis.” In discussing the attack on RTS, however, the committee reaches its decision on the basis of a cumulative assessment of the collateral damage in relation to the military objective, of which the RTS studios were an integrated part: the prosecutor defines this objective as the entire “Yugoslav command and control network.”190
It is unfortunate that the prosecutor’s report uses a cumulative assessment, as this is arguably inappropriate,191 and moreover, precludes any real debate about the proportionality of this specific attack. Still, this is one case where the question is addressed directly: Amnesty argues that sixteen civilians are too many, and the prosecutor argues that they are not. This attack falls into the simpler, type I category of proportionality decisions, as explained above (section II): Whether or not the destruction of a TV studio is worth the lives of sixteen civilians. With a real debate on proportionality, a consensual, customary law answer to this question might in time become possible. Under the present circumstances, however, too little has been written on the subject: NATO states that the death of sixteen civilians is not disproportionate; Amnesty disagrees. At this point, there is no external and independent discourse to provide a framework within which the question can be decided.