Hospital Bombings: Empirical and Theoretical Fallacies of Those Rejecting a Ban

The argument we advance in our recent EJIL Article, ‘‘Hospital Shields’ and the Limits of International Law’, emerged from analysis of empirical data showing how, during the past several years, hospitals were being bombed on a daily basis. Comparing these attacks with official statements released by actors suspected of bombing hospitals, we discovered that one of the recurrent arguments used to legitimise the strikes was that the facilities had been transformed into ‘hospital shields’ and used to conceal military targets. We then decided to reconstruct the history of hospital bombings and found that since 1911 — the first time medical units were bombed from the air — belligerents have consistently justified aerial strikes by claiming that the medical units were being used to hide combatants or harbour weapons.  

This revelation led us to examine in detail the historical development of the legal clauses dealing with the protection of medical units in armed conflicts. Our analysis revealed that the clauses include a number of exceptions that have allowed belligerents to assert that the bombing was carried out in accordance with IHL. We argue that belligerents can do this since hospitals occupy a spatial and legal threshold during armed conflict, and that IHL, which is informed by the rigid distinction between combatants and noncombatants, does not have the vocabulary to deal with liminal people and objects. This, we maintain, enables belligerents to use the law to justify the attacks.  

Our assumption throughout the paper is that IHL is subject to constant interpretation and reinterpretation, and that the way states interpret the law — even if we disagree with their interpretation — helps to establish the law’s meaning. International law is, after all, shaped by states, and through their practices, manuals and utterances they help determine the interpretation of its clauses. Hence, the fact that for over a century many states, among them the most powerful ones, have justified the bombing of hospitals by claiming that they were used as shields is not something we can dismiss by simply claiming that they are misinterpreting the law. After all, those very states introduced the hospital shields exception. 

Taking into account the history of hospital bombings and the way state actors have interpreted international law over the years, we conclude that the only way to protect hospitals from the belligerents that bomb them is by introducing a ban, not unlike the ban on torture. In his response to our piece, Yishai Beer, a former IDF general and president of Israel’s military court of appeal, who is currently a law professor at the Herzliya Interdisciplinary Center, rejects our call to reform IHL by imposing a ban on bombing hospitals. He introduces two main claims. First, he argues that our approach ‘does not appear to be consistent with the empirical data’. Second, that we do not ‘convincingly address the practical effect that absolute immunity might have on the functioning of hospitals and the safety of their patients and staff if they become safe havens for belligerents’. He concludes that ‘the absolute immunity suggested is neither feasible nor normatively desired’. 

We concur with Beer that empirical data is of vital significance, and indeed only after reconstructing the history of hospital bombings did we decide to propose a reform to IHL. Beer intimates, however, that we overstate the problem. Bringing evidence from an ICRC report that monitored violent attacks against health facilities during the 24-months period of January 2012 to December 2014, he notes that ‘only 17 per cent’ (italics added) of the 2,398 incidents were direct attacks. By so doing, Beer understates the problem. Indeed, this 17 per cent corroborates our claim, since it reveals that during those two years every 43 hours on average a hospital was directly attacked. More recent data discloses that the epidemic has not subsided. According to the World Health Organization, in 2016 every day and a half on average a hospital was bombed, and in 2017 and 2018 a hospital was bombed every two days. Taken together, the data suggests that hospital bombings are neither sporadic nor a series of isolated events, but rather a strategy of warfare aimed at weakening the enemy’s infrastructure of existence. This also suggests that most strikes have little to do with the use of hospitals as shields — even though, at times, hospitals are used as shields.

Only by brushing aside the gravity of the empirical data, its massive scale, and the horrifying consequences can Beer argue that the existing protections offered by IHL — which he describes as a ‘wide legal rubicon that is very difficult to cross’ — are adequate and that improvements are needed only in ‘law enforcement and compliance’. Indeed, he maintains that due to the ‘damage multiplier of the sick and wounded’ during attacks against hospitals, ‘only in a very limited number of cases would the potential collateral damage to both persons and infrastructure from targeting hospitals – which, by definition, would be very substantial – not be considered ‘excessive’ vis-à–vis the ‘concrete and direct military advantage anticipated’. ‘Legally’, he adds, ‘the general rule is that hospitals, even shielding ones, are civilian objects. The shielding per se does not justify targeting a hospital. It might only remove the special protection, which is only one of its protective layers’.

If only this was the case. Beer offers no empirical evidence that states interpret the law in this way. Yet, his legal analysis suggests that the vast majority if not all of the attacks on medical facilities that we describe in the article were illegal according to international law and constitute war crimes. If state actors agreed with Beer’s interpretation, then it would seem that the central issue might indeed be enforcement and compliance. But, as it emerges from a 2017 Safeguarding Health in Conflict Coalition’s report (cited by Beer), ‘attempts to accountability have largely failed’. And as our historical reconstruction of hospital bombings – including the bombing of 17 hospitals, 56 primary health care facilities, and 45 ambulances in Gaza during the 2014 war – indicates, states tend to interpret the law as if it provides them with a green light to carry out the attack. Indeed, we specifically mention the Gaza strikes since the Israeli Ministry of Foreign Affairs published a long legal report claiming that its 2014 attacks on medical facilities were carried out in accordance with IHL, and while spokespeople of other governments provide comparable justifications for such attacks we have not come across similar reports. An EJIL blog questioning the justifications provided by Israel and other states that have systematically bombed health facilities would be extremely important, but we fear such an intervention is unlikely to change the way states interpret the law.  

Yet, the fallacy of Beer’s reply is not only based on the fact that he does not appreciate the enormity of the empirical evidence regarding hospital bombings, the power of state actors to interpret the law, and the way in which they systematically interpret it to justify the use of lethal violence against health facilities. It also stems from the misrepresentation of our argument. 

He writes that ‘granting absolute protection to a shielding hospital, whatever its harm may be, seems to undermine the distinction rule and its threat-based rationale’. We, however, are not advocating ‘granting absolute protection to shielding hospitals’, but rather promoting absolute protection to all hospitals. To be sure, absolute protection comes with a price. One way to conceptualize this price is by inverting the logic of ‘collateral damage’ which refers to a situation where civilians are sacrificed in order to kill a high value target. Our assumption is that hospitals are high value sites, and we suggest that in order to guarantee their protection in some instances (when they are actually used as shields) enemy combatants will have to be spared, even though they are legitimate targets according to international law. The presence of a legitimate military target in or near a hospital, cannot, in other words, supersede its inviolability. 

Beer’s fundamental criticism is that we undermine the principle of distinction because a ban on hospital bombing would force state actors to abide by absolute protection while incentivising non-state actors to hide in or shoot from hospitals and exploit their protected status. This argument presents two main problems. The first is empirical, and the second has to do with Beer’s conception of distinction. First, as we pointed out in the article, the claim that a ban would incentivise the use of hospitals as a shield is merely hypothetical, while our argument about the epidemic of hospital bombings and the use of IHL to justify them is based on empirical evidence. Second, Beer’s claim uncritically rehearses one of the foundational myths of the principle of distinction: that of the law-abiding state attacker tricked by the lawless non-state perfidious enemy. Even though this state-centric conception of distinction has been empirically validated in several cases, as we show elsewhere, much more often states have mobilized it to justify egregious war crimes against protected persons, most often in asymmetric conflicts with non-state actors. 

The difference between Beer’s and our position is thus straightforward. We consider hospitals as high value assets, as sanctuaries, and their bombing as cruel and inhumane, and put the protections offered to them before the right of militaries to attack, while Beer, wittingly or not, ends up defending those who attack hospitals. We make the case for a ban with no exceptions regardless of the identity of the adversary, while he accepts the existing exceptions, knowing that states have repeatedly used them to legitimise hospital bombings. We believe the law should provide hospitals with absolute protection from the threat of being bombed, while he believes that the threat of bombing hospitals protects them. 

Finally, Beer also notes that our proposal is not practical, because according to the existing international legal system states must agree to a total ban on bombing hospitals, and it is highly unlikely that currently many states, particularly the powerful ones, would be interested in accepting such a ban. We agree with this observation, but we are also aware that it took several years before states joined the bandwagon and signed the conventions prohibiting torture or anti-personal land mines. The campaign to ban hospital bombings will begin with concerned individuals, civil society and NGOs, and we can only hope that more scholars will follow suit.