ICJ Judge Giorgio Gaja (who was also the special rapporteur on the International Law Commission for the Responsibility of International Organizations) has made the case that International Organizations have a duty to prevent. The context was a talk he gave at the University of Amsterdam in April 2013 on the European Union and the ILC’s Articles on the Responsibility of International Organizations.
If one takes the approach followed by the International Law Commission on the issue of attribution and applies it to the European Union, the Union would be internationally responsible when its organs or agents commit a breach of one of the obligations that the Union has under international law.
Depending on the content of the international obligation, a breach could consist in the failure to comply with a rule requiring the European Union to ensure that Member States do something or in the failure to prevent them from taking certain actions.
This type of obligation does not necessarily consider the conduct of Member States in a specific way. It may be an obligation of result, like arguably those under UNCLOS that were at stake in the Swordfish case between the European Community and Chile. The fact that the European Union does not achieve the required result of the conservation of swordfish stocks would be sufficient to cause a breach, whether the failure is caused by its organs or agents or by its Member States. The WTO agreements may provide further examples of obligations of result that may be breached by the Union because of the conduct of its Member States.
The emphasis on the conduct rather result indicates there is no requirement to suceed. But he goes on to say that this obligation is linked to an IO’s capacity to influence the actions of member states, such that an IO might incur its own responsibility if it fails to prevent a breach of an international obligation.
The simplest, and probably most frequent, scenario of a possible responsibility of the European Union is that the Union is bound not only by the obligation breached by the Member State but also by an ancillary obligation to prevent the relevant wrongful act of member States or at least not to contribute to it. Failure by the European Union to comply with that ancillary obligation would give rise to the Union’s responsibility. The Union would then incur responsibility for the breach of this distinct, though connected, obligation. The responsibility of the European Union would be normally additional to the responsibility incurred by the Member State.
These ideas about an International Organization’s duty to prevent are noteworthy for a few reasons.
- First, they give more context and detail on the duty to prevent than the Draft Articles on the Responsibility of IOs, which refer in general ways to omissions and the Swordfish case, but do not play out scenarios of ancillary IO liability.
- Second, this logic suggests that IOs may be independently liable for the failure to prevent acts by their member states. In other words, states and IOs could be concurrently liable for acts and omissions. For terrific work on shared responsibility generally, see the University of Amsterdam’s SHARES project (where I am spending some of my sabbatical).
- Third, it raises the stakes for IO “supervisory” capacity generally. Indeed, a pivotal question in this regard is what are the circumstances that would trigger the duty to prevent?
An IO must be bound by a relevant primary norm of course, and the acts in question must constitute breaches of those norms by act or omission. These are the two key preconditions for application of the Draft Articles. An additional third factor would be capacity: the ICJ’s Bosnia decision specifies in this regard that the duty to prevent is heavily contextual, in that an actor must use all means available to them, and it will depend on their capacity to influence. For IOs that operate on a consensus basis (ie, NATO) or that work through advice and assistance rather than coercion (ie, the WHO), it might be hard to argue they have much capacity to prevent. On the other hand, IOs like the EU or the UN that can coerce member states, might be facing more liability going forward.
What substantive areas will the duty to prevent be most likely to arise? I would wager to guess that we can expect to see this duty argued in three cases: the use of force, conservation of scarce resources, and situations involving massive human tragedies that trigger the Genocide Convention or other human rights treaties. I would be interested if our readers predict other areas of activity with regards to the duty to prevent.