Don’t Blame IHL for Attacks on “Hospital Shields”

Just Security published a very interesting post yesterday entitled “Military Attacks on ‘Hospital Shields’: The Law Itself is Partly to Blame,” which seeks to explain why deliberate attacks on hospitals are becoming increasingly common — in Syria, in Yemen, and elsewhere. The authors acknowledge that deliberate attacks on hospitals are almost always unlawful under IHL, because they violate the principle of distinction. But they nevertheless insist that “[t]he tragic irony is that international humanitarian law itself offers the legal toolkit for these regimes to justify the bombing of hospitals.”

The argument is provocative, but it is also misguided. And the authors’ solution is, I think, worse than the problem.

Let’s take a look at the how the authors believe IHL itself helps justify the bombing of hospitals. They cite two interrelated rules. Here is the first:

First, the logic of the human shields clauses can, it seems, quite easily be transferred to medical facilities. International law prohibits the use of civilians as human shields to protect military targets, but it also permits the attacking forces to kill human shields as long as they abide by the principle of proportionality. In this instance, then, international law ceases to protect civilians and actually becomes a weapon of the strong, protecting those who kill non-combatants. By extension, if hospitals are used as shields, they too can be bombed provided the principle of proportionality is not breached.

This argument is absolutely correct. Yes, human shields can be killed as long as they are not directly targeted and as long as an attack on the legitimate military objective they are shielding does not kill them disproportionately. And yes, the same goes for civilians objects, including hospitals, that are being used to shield military targets. But it is not clear what is problematic about either of those IHL rules, which are straightforward expressions of distinction and proportionality — unless you think that combatants should be able to render themselves completely immune from attack by hiding in hospitals. (Which is precisely what the authors think, as discussed below.) You don’t have to be soft on the military to suggest that combatants should not be able to use the protection IHL offers hospitals to gain a military advantage over their enemy.

Moreover, the authors fail to note that even though IHL permits “hospital shields” to be attacked, it imposes significant restrictions on such attacks. Most importantly, Art.19 of the Fourth Geneva Convention provides that hospitals being misused lose their protection against attack “only after due warning has been given, naming, in all appropriate cases, a reasonable time limit and after such warning has remained unheeded.” Assuming that the attacking military complies with Art. 19 (and it is hardly IHL’s fault if it doesn’t), no hospital will be attacked that has not had an opportunity to expel the military objective it is shielding. IHL thus puts so much emphasis on protecting hospitals that it would rather require an attacker to let the enemy escape unharmed rather than apply the normal targeting regime of distinction and proportionality.

A similar problem undermines the authors’ argument concerning the second rule:

Second, international law affirms that the protection to which hospitals are entitled is revoked when they are “used to commit, outside their humanitarian duties, acts harmful to the enemy.” This extremely vague formulation lends itself to those who target hospitals. Unlike IHL clauses prohibiting torture, which are absolute, applying at all times and in all circumstances, the articles relating to the bombardment of hospitals are conditional. Therefore, in certain “exceptional” situations medical facilities do lose their protected status.

This argument is vastly overstated. Is the “acts harmful to the enemy” formulation “extremely vague”? Sometimes, perhaps. But certainly not always. Art. 19 makes clear, for example, that “[t]he fact that sick or wounded members of the armed forces are nursed in these hospitals, or the presence of small arms and ammunition taken from such combatants and not yet been handed to the proper service, shall not be considered to be acts harmful to the enemy.” IHL can thus hardly be blamed when a military justifies attacking a hospital by claiming that the hospital was treating enemy soldiers. (A claim made by Afghanistan regarding the MSF hospital in Kunduz.)  Any such attack categorically violates IHL.

There is also widespread agreement about what kinds of acts do qualify as “acts harmful to the enemy.”  The ICRC’s commentary to Art. 21 of the First Geneva Convention is typical:

Such harmful acts would, for example, include the use of a hospital as a shelter for able-bodied combatants or fugitives, as an arms or ammunition dump, or as a military observation post; another instance would be the deliberate siting of a medical unit in a position where it would impede an enemy attack.

It is possible to criticise this understanding of harmful acts as being overbroad and in need of revision. I, for one, have a problem with the idea that a hospital can be attacked simply because combatants are using it as “an arms or ammunition dump.” Given the importance IHL puts on protecting medical units, that doesn’t strike me as enough to justify a hospital forfeiting its protected status. I might even be convinced that the mere presence of unwounded combatants in a hospital shouldn’t justify a deliberate attack.

But the authors are not arguing for a tightening of the rules that govern when hospitals can be deliberately attacked. No: because they believe it is too easy for an attacker to claim that a hospital is shielding a military objective (permitting attack after a warning), they want a categorical rule that prohibits attacking hospitals no matter how they are being used:

The only way to overcome this travesty is if IHL clauses pertaining to the protection of hospitals are reformulated in a way that categorically prohibits the use of lethal force against them. Currently, IHL provides the necessary protections for hospitals, and all that is really needed is to erase the clauses stipulating exceptions since these in essence hand out militaries a license to bomb medical facilities.

In other words, the authors believe that a hospital should be immune from attack even when combatants are using it to attack the enemy. What the attacked forces are supposed to do in such a situation, the authors never explain. Apparently, they are simply supposed to either tolerate casualties or somehow avoid coming within the hospital’s field of fire. (Which may well be impossible, depending on the kinds of weapons the holed-up combatants possess.)

That will never happen, of course. Not even the most professional military will tolerate being fired upon from a civilian hospital — or at least it won’t tolerate it indefinitely, which is what the authors expect. IHL recognises this basic reality of combat, which is why it attempts to strike a balance between humanitarian concerns and military necessity by permitting a hospital that is being used as a shield to be deliberately attacked, but requiring a warning and a reasonable time to comply.

As with the definition of “acts harmful to the enemy,” it is possible to quibble with the procedural requirements for attacking a hospital that is being used as a shield. Perhaps IHL should avoid using a mushy “reasonableness” test for the length of the warning a military must give before attacking a hospital, imposing a minimum amount of time instead. I’d be open to that. But again: the authors are not trying to strike an appropriate balance between humanity and necessity. They simply want a categorical ban on attacking hospitals — even those that are genuinely being used to shield military objectives.

I understand (and am sympathetic to) the reasoning behind the authors’ position. They don’t want innocent civilians, especially medical personnel and the wounded, to pay the price for the actions of combatants who have no respect for IHL. Hospitals don’t choose to be misused by combatants, and there will almost certainly be situations where combatants simply refuse to leave a hospital that has been warned of an impending attack. So hospitals may well find themselves in an impossible situation: subject to attack because their premises are being misused, but unable to do anything about it.

The answer, however, is not to categorically prohibit attacking hospitals. As noted, militaries would never comply with such a prohibition. And, of course, such a prohibition would ensure that combatants who don’t respect IHL will use hospitals as a shield as often as possible. The authors begrudgingly acknowledge that possibility, writing that “[s]ome might argue that such norms will produce the perverse incentive among certain belligerents to use hospitals as shields.” But that is a serious understatement. Some might argue? What IHL scholar wouldn’t? Why would a military that doesn’t respect IHL not use a well-placed hospital as a shield if doing so renders its forces absolutely immune from attack?

The authors are correct that anything short of a categorical prohibition on attacking hospitals will leave open the possibility of militaries inventing facts to justify attacks. That is the nature of IHL rules that are not categorical — and a reflection of the fact that IHL is neither purely humanitarian nor purely war-enabling. The authors’ solution, however, is worse than the problem. A categorical prohibition will not prevent IHL from being misused; it will simply ensure that IHL is ignored — resulting in far more “incidental” deaths than under the current IHL rules. The better solution (absent a tightening of the rules as discussed above) is to searchingly examine the legitimacy of each and every attack on a hospital and hold militaries to account when they use the concept of the “hospital shield” to justify an unlawful attack.

And that, of course, is exactly what IHL requires.