Prlić et al.: The Destruction of the Old Bridge of Mostar and Proportionality

Rogier Bartels is a legal officer in Chambers at the International Criminal Court. The views expressed in this post (and in the article referred to, which was written before he joined the Court) are his own and do not necessarily represent the views of the ICC. He blogs at the Armed Groups and International Law Blog.

 

These last months, most of the blog posts on the ICTY have focused on debatable Appeals Chamber judgements and the associated letter by Judge Harhoff. A significant Trial Judgement (in French) that would normally perhaps have received attention was therefore largely ignored – although the international courts and tribunals’ judgements and decisions in French are often overlooked. The recent judgement in the multi-accused trial Prlić et al. is significant not just because of its length (six volumes totaling almost 2600 pages, including a 576 page dissent) and the time it took to draft (it was rendered more than three years after the end of the Defence case, and well over two years after the closing arguments were held). Most striking perhaps is the Trial Chamber’s determination that the Croatian leadership, including president Tudjman, formed part of a joint criminal enterprise that pursued the establishment of a Croat-only part of Bosnia-Herzegovina (for it to be joined with Croatia); this is especially interesting after the Appeals Chamber’s ruling that the Gotovina Trial Chamber was wrong in its finding in this regard. However, when listening to the delivery of the Prlić Judgement, what really caught my ear was the following “finding” on the destruction of the Old Bridge of Mostar:

On 8 November 1993, as part of the offensive, an HVO tank fired throughout the day at the Old Bridge until it was unusable and on the verge of collapse. The Bridge then collapsed on the morning of 9 November 1993. The Chamber finds, by a majority, with the Presiding Judge dissenting, that although the Bridge was used by the ABiH and thus constituted a legitimate military target for the HVO, its destruction caused disproportionate damage to the Muslim civilian population of Mostar (Judgement Summary, p. 3).

I have recently published an article in the Israel Law Review (open access until half August) on the use of the principle of proportionality in international criminal law, which discusses extensively the ICTY practice with respect to the principle. When listening to (the English translation of) Judge Antonetti reading out the summary of the judgement, I feared that my article was outdated already, as it appeared that the Majority had applied the proportionality principle in coming to this finding. And not just any finding: one that appeared to include a dual use object and a novel determination on the weighing of the long-term expected incidental damage. I say “appeared” twice because, as will be explained below, the summary was actually quite deceiving in this respect and no such finding was in fact made. (Old Mostar Bridge pictured right, credit)

In the article referred to above, I show that the Tribunal’s practice can be divided into three categories of cases where the different chambers: i) state and clarify the principle, but do not apply it to the facts of the case; ii) make findings on disproportionate or excessive use of force that cannot actually have resulted from an application of the principle; iii) the Gotovina case, in which the Trial Chamber did apply the principle to the evidence, but was then quashed by the Appeals Chamber. In this post, I will briefly discuss the ways in which the principle has been addressed in the ICTY’s case law in order to see where the Prlić case fits in.

The principle of proportionality has hardly been dealt with by the ICTY. This is due in part to the relatively limited number of cases dealing with conduct of hostilities, and partly because the ICTY’s subject matter jurisdiction. The Tribunal’s Statute lists war crimes as “grave breaches of the Geneva Conventions of 1949” (Article 2) and “violations of the laws or customs of war” (Article 3). A violation of the principle of proportionality obviously could not fall under Article 2, as the grave breaches of the four 1949 Geneva Conventions are made up of an exhaustive list (Articles 50, 51, 130, and 147 of the Conventions, respectively)and the regulation of conduct of hostilities was not yet included in the grave breaches regime in 1949 (this was done in 1977 by means of Article 85 of Additional Protocol I).

Since the crimes under Article 3 ICTY Statute are limited to serious violations of those norms of IHL that, at the time of the conflict in the former Yugoslavia, had beyond any doubt reached the status of customary international law (Report of the Secretary-General, para. 34), the Tribunal has not considered serious violations of in bello proportionality to constitute a separate crime. Quite the contrary, in fact: in its case law, such violations have merely served as evidence of the existence of attacks directed against civilians or against civilian objects. When the ICTY did address the proportionality principle in its case law, it did so firstly in cases where a legal analysis of what constitutes the principle was made, but in which the principle was not actually applied to the facts and/or evidence. Examples of such cases are Kupreškić, Galić, Strugar, Martić, and Dragomir Milošević.

The customary status of the principle, and what it entails, was set out correctly in these cases – as also acknowledged by the ICRC Customary IHL Study (see the extensive reference to these ICTY judgements in the Study’s practice for Rule 14), which anyway relies heavily on case law of the ad hoc Tribunals. Particularly, the Galić Trial Chamber explained clearly what the principle entails and how the proportionality of an attack should be assessed (Galić TJ, paras. 58-61). Be that as it may, nearly all of these findings are no more than obiter dicta, as the concerning chambers did not actually apply the balancing test that underlies the principle (i.e. anticipated military advantage versus expected incidental damage) to the facts of the case. In Strugar, for example, the Trial Chamber explains what the proportionality principle entails, but then goes on to conclude that the principle did not arise in that case because no military objects were located in the old town of Dubrovnik, which was attacked by Serbian forces (Strugar TJ, paras. 214, 281 and 295).

The second category of cases conclude that disproportionate or excessive force was used, but do not explain how the chamber arrived at this conclusion. Or, when such a finding was made, the chambers do not appear to have balanced properly the expected military advantage against the expected collateral damage. In Blaskić, Milutinović et al. and Đorđević, the respective Trial Chambers mention the use of “disproportionate force”, “force out of proportion to the military necessity” and “excessive force” numerous times (see, e.g., Blaskić TJ, para. 651; Milutinović TJ, para. 920; and Đorđević TJ, para. 980). And yet, these findings do not seem to have resulted from a proportionality analysis (the one that was correctly set out in Galić (para. 58)). In Blaskić, the Trial Chamber concludes that no military objects were attacked, and both the Kosovo cases do not even concern unlawful attack charges, but deal rather with the murder of detained persons and the deliberate destruction of civilian houses. In these three cases, the reference to disproportionate and excessive use of force thus indicates the manner in which the HVO and Serbian forces, respectively, used force, rather than the principle’s legal meaning as framed in Articles 51 and 57 of Additional Protocol I. As mentioned above, Galić correctly discusses the law related to the principle of proportionality; nevertheless, it applied the principle only once to the facts – sort of. And when it did, it did so without reason because the issue of proportionality could not, in fact, arise for the attack in question. Since the attacking Bosnian-Serb forces could not see the targeted location (a parking lot, surrounded by tall buildings; Galić TJ, para. 387), and were unable to make a distinction between legitimate targets and civilians due to the use of artillery (with mortar shells of “at least 81 mm”, para. 377), the crowd made up of football players and spectators, both civilians and members of the Bosnian-Muslim forces, was targeted as a whole. The strike was thus indiscriminate; irrespective of any expected incidental damage (see more in detail, my discussion on pp 285-286).

Then, thirdly, there is the Gotovina case, in which the Trial Chamber more or less applied the principle to the evidence resulting in an evaluation of the balancing test. This case, in which accused were charged, inter alia, with “the shelling of civilians and cruel treatment [and] unlawful attacks on civilians and civilian objects”, was the first and only time that the principle of proportionality was a key issue in an ICTY case. However, the Trial Chamber’s findings on targeting proved to be so problematic for the Appeals Chamber that it quashed the Trial Judgement and fully acquitted the two remaining accused.

The Trial Chamber made a proportionality determination with respect to attacks on the political leader of the Krajina and supreme commander of the Serbian Army of Krajina, Milan Martić. It was satisfied that Martić’s residence constituted a military target (Gotovina TJ, para. 1899) and noted that both this building and another location where the HV believed Martić to be present and fired at were situated in a residential area (paras. 1191, 1198, and 1910). Whilst having regard to expert evidence on accuracy and blast and fragmentation radius of artillery weapons, the Trial Chamber held that:

At the times of firing, […] civilians could have reasonably been expected to be present [near the locations fired at]. Firing twelve shells of 130 millimetres at Martić’s apartment and an unknown number of shells of the same calibre at the area marked R […], from a distance of approximately 25 kilometres, created a significant risk of a high number of civilian casualties and injuries, as well as of damage to civilian objects. The Trial Chamber considers that this risk was excessive in relation to the anticipated military advantage of firing at the two locations where the HV believed Martić to have been present (para. 1910).

Curiously, this was the only proportionality analysis made by the Trial Chamber, whilst the circumstances regarding the rest of the strikes were not substantially different. The Trial Chamber merely used this finding as an “indicative example” to support its conclusion that “the HV paid little or no regard to the risk of civilian casualties and injuries and damage to civilian objects when firing artillery at a military target on at least three occasions on 4 August 1995” (para. 1910).

The Appeals Chamber criticised the finding on the attacks on Martić, but did not overturn it. By majority, it held that “[t]he Trial Chamber’s analysis of the attacks on Martić involved a lawful military target, was not based on a concrete assessment of comparative military advantage, and did not make any findings on resulting damages or casualties”, but that it did not need to consider whether the Trial Chamber erred in finding that the attack on Martić was disproportionate (Gotovina AJ, para. 82). No clarification was thus given of how the proportionality assessment should have been better applied. Naturally, the Majority was correct in observing the Trial Chamber’s failure to examine the military advantage (at the time of attack) of taking out Martić, and that no findings were made on “resulting damages or casualties”. However, the latter would not have assisted the Trial Chamber. Besides – arguably – strengthening the finding that “civilians could have reasonably been expected to be present”, it should be clear that an evaluation of the resulting damage is by no means necessary for the assessment of the proportionality of an attack (for which the determination needs to be made by the attacker prior to launching a strike) during a criminal trial.

In addition, the Majority found that the proportionality finding was “of limited value in demonstrating a broader indiscriminate attack on civilians in Knin” (para. 82). Indeed, the attacks on Martić could not, without more, change the nature of the operation as a whole into an indiscriminate attack. However, the Trial Chamber could have applied a proportionality analysis to more strikes. Rather than determining the accused’s intent to attack civilians by looking at the location where the projectiles had landed (using the “200 metre” rule/standard that afterwards received fierce criticism, see e.g. here), the Trial Chamber could have examined the proportionality of the attacks by assessing the likelihood that the shells and rockets would impact on objects other than legitimate targets, and could thus be expected to cause incidental damage. Using the reasonable commander standard as described in Galić (para. 58), it could then have considered whether the risk of causing excessive collateral damage was too high when compared to the anticipated military advantage. Unfortunately, the opportunity to clarify the use of the principle of proportionality in ICL was not seized either by the Trial Chamber or by the Appeals Chamber.

So where does the Prlić finding on the attack on the Old Bridge fit into? Well, actually it doesn’t fit into any of the three categories discussed above. The Judgement Summary (both in French and in English), which is the text that was read out by the presiding judge when rendering the judgement, states that “the bridge was used by the ABiH and thus constituted a military target” and that the bridge’s destruction “caused disproportionate damage to the Muslim civilian population”. However, when one reads the 80 (!) paragraphs of the (French) Trial Judgement (paras. 1286-1366) on the bridge’s destruction, one can derive that the Majority considered the bridge to qualify as a military object pursuant to Article 52(2) of Additional Protocol I (paras. 1290, 1357 and 1363), but no explicit and clear finding on the status of the bridge as a military target was made. This is unfortunate as questions on proportionality when attacking dual use objects (see my discussion on pp 305-306), and whether the threshold for (important) cultural property to lose its protected status when used for military purposes is different from a regular civilian object, are still open to debate. These questions thus remain unanswered.

Furthermore, the Chamber did not balance the anticipated military advantage that the HVO would achieve by destroying the bridge and the expected damage to the bridge itself and/or the civilians. As discussed above, it would not be the first time that a chamber makes conclusions as to the disproportionate nature of an attack without actually having applied the proportionality principle to the facts of the case. Yet, this is the most striking thing: the Trial Chamber in Prlić did not conclude that the destruction of the bridge was disproportionate! Whereas interim conclusions were made on the military use of the bridge, which was the only one left in the area, and the direct effect its destruction had on the humanitarian situation on the right bank of the river (para. 1293), the Trial Chamber neither assessed whether the HVO failed to take the effects on the civilian population into account or that these would have been excessive in relation to the anticipated military advantage, nor did it anywhere call the attack disproportionate. The Chamber’s conclusions were as follows:

The Chamber is satisfied that before 8 November 1993, the Old Bridge […] was used not only by the ABiH to supply its troops and transport military supplies to the front line, but also by the population of Mostar for maintaining contact between the two banks [of the river] and for the provision of food and medication. In addition, the Chamber is satisfied that the bridge had enormous symbolic value to (mainly) the Muslims. The armed forces of the HVO had a military interest in this object being destroyed, since its destruction would cut off virtually all options for the ABiH to resupply its operations. However, the loss of the Old Bridge resulted in almost full isolation of the residents of the Muslim enclave on the right bank of the Neretva river. The Chamber finds that on 8 November 1993, […] a tank HVO fired throughout the day on the Old Bridge, making it unusable and on the verge of collapse by the evening of 8 November 1993. (“Conclusions générales de la Chambre sur la destruction du Vieux Pont”, paras. 1364-1366; my translation.)

From the above, it is clear that the Majority only made a factual finding and did not – as opposed to the summary – state that the attack on the bridge, a military object, was unlawful. So why would the Judgement Summary include such a legal finding? One can only guess. Perhaps because the person drafting the summary thought it was a proper summary of the findings in paras. 1286-1366? Or maybe because an important (historic) event, such as the collapse of the Old Bridge, needed special attention in the summary? Something similar was done in the Judgement Summary in Đorđević. Although there was no need for the Đorđević Trial Chamber to make a finding on the nature of the 1999 armed conflict between NATO and Serbia, it included a paragraph in its judgement stating that it was an international armed conflict (see also Dov Jacobs’ post at the time). Despite the fact that this obiter dictum was of little relevance for the Chamber’s conclusions, the four lines of this paragraph (of a 2231 paras/969 pages judgement) found their way – almost verbatim – into the seven page Judgement Summary. Earlier this year, Marko Divac Öberg discussed the mysterious art of judgement drafting. It seems that the drafting of Judgement Summaries is just as mysterious…

When reading the relevant part of the Prlić judgement, one wonders why no explicit legal finding was made on the proportionality of the attack on the Old Bridge; the law (see paras. 124, 169, and 189) was ready to be applied to the facts. Whether this would have resulted in an adequate application of the balancing test is uncertain. The Majority noted, for example, that the bridge was an important cultural object but its main criticism of the bridge’s destruction was the resulting isolation of the Muslims on the right river bank. However, the judgement sets out that the isolation was only really effectuated by the destruction of the Kamenica bridge a few days after the Old Bridge collapsed (para. 1355). Notwithstanding arguments that if the military advantage of an attack should be that of an “attack considered as a whole and not from isolated or particular parts of the attack” (see, e.g., Canada’s reservations to Articles 51 and 57 of Additional Protocol I), the incidental damage also should take into account the attack “as a whole”, the summary only deals with the specific strike on the Old Bridge. It thus seems that without the later destruction of the other bridge yet having taken place, the correct conclusion – based on the Majority’s own findings – would have been that the strike on the Old Bridge was proportionate at the time it was being launched. In any case, since the summary as it mentions itself is “not an official document”, the Tribunal’s case law is thus – quite unfortunately – still lacking satisfactory application of the proportionality principle. Whereas the ICTY has clarified many IHL matters, the military and academia will have to do, for now, without any clarification on this issue.