The scope of ICC jurisdiction over the crime of aggression: a different perspective

In his post of 26 June 2017 Dapo Akande asks:

“Are nationals of states that do not ratify or accept the Kampala amendments, and which also do not opt out of ICC jurisdiction as provided for in those amendments, subject to ICC jurisdiction over aggression in cases where the situation is referred to the Court by a state, or the prosecutor takes up the matter proprio motu?”

Why does the answer to this question matter? “No” means that an ICC state party that has ratified the amendments will enjoy the Court’s judicial protection only if it falls victim to aggression by one of the other (currently) 33 ratifying states. It would be an opt-in regime for potential aggressor states, and in fact, they could at any time later opt-out again (opt-in-opt-out). “Yes” means that such protection extends to aggression committed by any of the 123 other ICC states parties – of course with the significant caveat they can still opt out. That would be an opt-out regime. All of this of course only in the absence of a referral by the UN Security Council, which would make state consent a moot point.

The issue is currently discussed by ICC states parties in view of the activation decision to be taken in December 2017. I am therefore happy to explain why I think the answer is “yes”, even though Dapo gave a thoughtful argument for “no”. More importantly though, I agree that these discussions should not lead to a re-opening of the Kampala negotiations. Indeed, Dapo made some excellent suggestions on how to move forward. This is what states should focus on.

Full disclosure: I am guilty of having contributed to the drafting of the amendments as legal adviser to the lead negotiator from 2003 to 2010, Ambassador Christian Wenaweser (Liechtenstein). I also admit that Dapo’s first post on the issue (“What Exactly Was Agreed in Kampala on the Crime of Aggression?”) sent shivers down my spine, as I returned from Kampala and read how the result of years of hard work left some very smart people confused. In my defense, I subsequently tried to promote better understanding of the amendments, including by publishing, with Claus Kress, the Travaux Préparatoires of the Crime of Aggression and the Commentary on the Crime of Aggression. In the latter, Niels Blokker and I take an in-depth look at the jurisdictional regime, with a special focus this question. I also published an article with Christian Wenaweser explaining the sequence of events, again with much emphasis on Dapo’s question (Essays in Honour of Roger S. Clark).

The logic of the deal in Kampala

As discussed more fully below, Dapo’s interpretation has no support in the drafting history of Kampala. For years, states had been haggling over the question whether aggressor state consent should be required for jurisdiction. They were split in the middle: One half (“Camp Consent”) wanted an opt-in system. The other half (“Camp Protection”) wanted a no-consent system, as would be the default rule under article 12(2)(a) of the Statute (for an indicative list of states belonging to each camp, see footnote 23 here) The compromise reached in Kampala excludes non-states parties from jurisdiction altogether and establishes an opt-out system for states parties. Camp Consent got almost everything, namely a fully consent-based system. Their only concession was that, by default, jurisdiction regarding ICC states parties was given (provided the amendments had entered into force for the victim state) – but could easily be revoked by a declaration. Camp Protection gave up almost everything. But that was the deal.

An opt-in-opt-out regime, however, was not the deal. That is not a compromise. It is not even within the spectrum of positions debated during and prior to Kampala! No delegation from Camp Consent ever asked for such an extreme version of consent; all they wanted was an opt-in. Why would Camp Protection have conceded an opt-in, and on top of that agreed to add an opt-out? An opt-in-opt-out regime would pull Camp Protection all the way over to the other side, and then beyond. Conversely, it would give Camp Consent more than it ever dared to ask for. There is no evidence in the formal and informal record that this was contemplated in Kampala, let alone that this was so agreed.

Remembering what the Rome Statute is all about: fighting impunity

While the drafting history is the simplest response, we must of course also look at other elements. Before doing so, one more general observation: those supporting Dapo’s interpretation often argue with the perspective of states in mind that want to be protected from the Court’s jurisdiction. The arguments focus strongly on consent and sovereignty. That is fair enough. But we should not forget the perspective of states with the opposite interest: to be protected from crimes through the Court’s jurisdiction. Given the object and purpose of the Rome Statute, this should really not seem too far-fetched.

Entry into force vs. exercise of jurisdiction

I agree with Dapo that the original Rome Statute is the first place to look for answers, (though Kampala also plays an important role). Yet the Rome Statute’s provisions dealing with the crime of aggression are not exactly clear. They are incoherent in more than one way. As a consequence, the question how the amendments would enter into force, and what the consequences would be for jurisdiction, was among the most polarizing in the negotiations from 2003 to 2010. It was also the subject of intensive debate in Kampala (see the account of events in my article authored with Leena Grover for AJIL, here, pages 523 to 526).

To my mind, Dapo’s strongest argument is that in Kampala, states parties explicitly stated that the amendments “shall enter into force in accordance with article 121, paragraph 5” (emphasis added). Said paragraph consists of two sentences: The first addresses how amendments enter into force – namely only for those states parties ratifying or accepting them, after one year. The second sentence, however, does not address how the amendment enters into force, but deals with exercise of jurisdiction: It excludes the nationals and the territory of states parties that have not ratified the amendment from jurisdiction. This constitutes an exception to article 12(2), which allows for jurisdiction over nationals or the territory of states that have not consented to the Court’s jurisdiction.

Dapo says that “the effects provided for in the second sentence of that article are to follow for any amendments”. Admittedly, that is certainly the first instinct that most lawyers would have when reading article 121(5). But under the Rome Statute, entry into force and exercise of jurisdiction are separate concepts. Look no further than article 12(2) itself, which allows the exercise of jurisdiction over nationals of states for which the Rome Statute has not entered into force,.

Dapo also rightly notes that states parties explicitly confirmed that the link between the two sentences fully applies in the context of the war crimes amendments (resolution here, page 13, preamble). This is not surprising: there was never any doubt among states parties that this link could not be broken for “normal” amendments under article 121(5). These would be either crimes newly added to the Statute (amending the list of crimes in article 5), or amendments to crimes already defined (amending either articles 6, 7 or 8). But aggression does not fall into either category. It was a crime already included in the Court’s jurisdiction, yet undefined. The reference cited by Dapo thus cannot simply be transposed to aggression.

Indeed, no explicit reference to the second sentence of article 121(5), or to its content, can be found in the amendments, nor in any resolution adopted by the ASP later on – because it was always carefully avoided! Instead, references to article 121(5) are always limited to the statement that the aggression amendments “enter into force” in accordance with this paragraph: meaning, they enter into force for each ratifying state individually. This is because in the context of aggression, the link between the first and the second sentence is less obvious. This very issue was discussed by states parties for years.

Arguments from Rome

1. There are two important arguments – at the level of the Rome Statute itself – as to why the second sentence applies to any amendment except aggression.The second sentence of article 121(5) stands in contradiction to article 12(1). ICC states parties have already accepted the ICC’s jurisdiction over the crime of aggression. This is literally what article 12 says: “A State which becomes a Party to this Statute thereby accepts the jurisdiction of the Court with respect to the crimes referred to in article 5 [which includes the crime of aggression]” If one assumes that consent of the aggressor state is needed for ICC jurisdiction, this certainly sounds like consent. The ordinary meaning of this sentence could not be clearer. One counter-argument often heard is that states parties could not possibly have meant it this way, since aggression was not yet defined. This argument holds true for states from “Camp Consent”, but not for “Camp Protection”. In Rome, many of these states seeking protection insisted on including aggression in the Court’s jurisdiction, and they thought that Art. 12(1) meant what it said. What about their treaty rights? Another counter-argument is that the exercise of jurisdiction was relegated under Art. 5(2) to a future “provision”. This is true, but not incompatible with states parties expressing their acceptance, in principle, of such future jurisdiction. Otherwise, the reference to aggression in Art. 12(1) would have been reduced to having absolutely no meaning – which is a rather radical reduction of a provision that is both clearly worded and a central provision of the Rome Statute.
2. The second sentence must also be read together with article 5(2). The latter mandates states parties to set out the conditions for the exercise of jurisdiction. By virtue of this provision, States parties in Kampala thus had the competence to deviate from the Statute’s default conditions for the exercise of jurisdiction. And deviate they did: Driven mostly by the concerns of “Camp Consent”, they packed into article 15 bis six (!) aggression-specific hurdles to jurisdiction that are nowhere to be found in the Rome Statute: 1) the requirement of 30 ratifications; 2) the activation procedure; 3) the opt-out for states parties; 4) the exclusion of non-states parties; 5) the need to allow the Security Council 6 months for a determination of aggression; 6) the requirement to have the investigation authorized not just by a Pre-Trial Chamber, but the Pre-Trial Division. All these jurisdictional hurdles derogate from various provisions of the Statute – including articles 12, 14 and 15. Why would states parties then not have been allowed to derogate from the second sentence of article 121(5) – incidentally the only deviation that that benefits Camp Protection and promotes the fight against impunity?

The argument made here is neither particularly complicated nor novel in its approach. These two provisions (article 12(1) and article 5(2)), both clearly worded in their relevant parts and specific to the crime of aggression, are lex specialis in relation to the second sentence of article 121(5). Their effect is limited to the crime of aggression and does not extend to the above-mentioned “normal” amendments under article 121(5). Consequently, this interpretation does not make the second sentence redundant, as claimed by Dapo. The second sentence applies fully to “normal” amendments of crimes– just not to aggression.

The opposite interpretation, by contrast, does make a provision redundant: namely the acceptance by states parties of the Court’s jurisdiction over the crime of aggression (article 12(1) in conjunction with article 5(1)). Absolutely no normative meaning would be left to this central provision of the 1998 Rome Statute.

Arguments from Kampala

The above arguments confirm that Rome allowed Kampala to foresee jurisdiction over nationals of states parties that have not ratified the amendments. The question remains though whether Kampala actually did do so. Many elements speak in favor of such a view:

1. Article 15 bis (4) states that the Court may exercise jurisdiction “in accordance with article 12”, and therefore refers to the default jurisdictional regime of article 12(2). It does not refer to the general exception to that rule contained in the second sentence of article 121(5).
2. In describing the opt-out regime, article 15 bis (4) refers to “aggression committed by a State Party” rather than to a state party that has ratified the amendments.
3. The resolution adopting the amendments ‘notes that any State Party may lodge a declaration referred to in article 15 bis prior to ratification or acceptance’. Mentioning this possibility only makes sense if one assumes that there can be jurisdiction in the absence of (i.e. “prior to”) ratification.
4. The resolution also recalls, in its first preambular paragraph, article 12(1) of the Statute – and thus the pre-existing acceptance of the Court’s jurisdiction over the crime of aggression.
5. The resolution further recalls article 5(2) of the Statute, thereby recalling the mandate to establish aggression-specific conditions for the exercise of jurisdiction.
6. The resolution, the amendments and the understandings don’t contain any reference, directly or indirectly, to a limitation of jurisdiction in the sense of second sentence of article 121(5), even though there would have been several logical places to do so. In fact, an existing draft reference, which was part of the negotiation text until almost the end of the conference (the so-called “negative understanding”) was deleted. (see next point)
7. And finally as mentioned, the drafting history in Kampala, accessible in the official record and more fully reflected, step-by-step, in the Travaux Préparatoires, supports only one interpretation. As mentioned above, one half argued for an opt-in system (i.e. 121(5) with full application of the second sentence), while the other half wanted a no-consent system (such as 121(5) without application of the second sentence, or 121(4)). Both camps had their positions reflected in the official negotiating text – in two alternative draft understandings (the “negative” vs. the “positive” understanding, see here, page 39). Once Canada and Argentina/Brazil/Switzerland had come up with a common position paper (Travaux, page 772), the opt-out clause was added in the official negotiating text (see the President’s proposal here, starting at page 40). At the same time, both draft understandings were deleted, since the opt-out clause made it clear that it was an opt-out system – and therefore neither an opt-in system (“negative understanding”), nor a no-consent system (“positive understanding”). In presenting the opt-out system, both in plenary and in countless informal meetings, President Wenaweser emphasized that it was based on the unique position of the crime of aggression in the Statute and on states parties’ “acceptance already given under article 12(1)” (transcript in Travaux, page 780). Furthermore, practically all of the above-mentioned elements (1 to 5) were included around the same time, precisely to highlight the legal underpinning for the opt-out system.

And finally, some significance should also be attached to the fact that the amendments were adopted by consensus. This strengthens – as an additional argument to those above – their relevance as subsequence practice among states parties that could straighten out some of the inconsistencies of the Rome Statute provisions regarding the future aggression amendments.

The way forward

In the ongoing activation discussions, some states parties have called for “clarity”. Of course, if it was just about clarity, then that could be achieved by an ASP statement re-confirming the logic of the opt-out regime. 

But what these states really want is assurance that there will be no jurisdiction over their nationals in the absence of their ratification, and without having to resort to an opt-out declaration. Obviously, this will be unacceptable to those states that fought hard for a protective regime, and which in 2010 already conceded the above-mentioned six jurisdictional hurdles in article 15 bis, including the opt-out. Stating in 2017 that the “negative understanding”, which was deleted in 2010, actually applies after all – that will be one hurdle too many.

Realistically, it will not be possible for the ASP to agree on an explicit statement expressing that the second sentence of article 121(5) does apply to the crime of aggression. Both sides therefore need to give up on something:

Camp Protection will have to live with the risk that the Court might at some point rule on the matter and agree with Dapo and others, handing Camp Consent a belated victory and further restricting the already very limited scope of jurisdiction.

States from Camp Consent will have to decide whether to make a declaration as foreseen in article 15 bis (4). In doing so, they would not even have to agree with this reading of the Rome Statute and acknowledge that they are “opting out” – quite to the contrary. They only have to inform the Registrar (as they are already now informing all other states parties) that they do “not accept” the Court’s jurisdiction over the crime of aggression in the absence of their own ratification. They could even use the declaration to explicitly reject this interpretation of article 15 bis, while safeguarding the legal benefit of the opt-out system, just in case the Court would see things differently. In other words, they could have their cake and eat it, too.

From my personal perspective, asking Camp Consent to do the latter is not asking for much. Consider also how narrow the scope of the question actually is: Is active consent (ratification) required or does passive consent (not opting out) suffice? Consent remains decisive, and the state concerned thus remains in control. By comparison, much bigger problems have already been overcome in the negotiation process. And most importantly, consider what is at stake overall: The activation of the Court’s jurisdiction over the “supreme crime” under international law, thereby finally – and with serious jurisdictional restrictions – reviving the promise of Nuremberg. I hope delegations at the ASP will keep things in perspective.