The Kampala Amendments on the Crime of Aggression Before Activation: Evaluating the Legal Framework of a Political Compromise

[Astrid Reisinger Coracini is is Lecturer at the University of Salzburg and Director of the Salzburg Law School on International Criminal Law, Humanitarian Law and Human Rights Law (SLS). This is the first of a two-part post on the subject.] 

In December 2017, the Assembly of States Parties of the Rome Statute of the International Criminal Court will hold its sixteenth session with the ‘activation of the Court’s jurisdiction over the crime of aggression’ on its agenda. Almost twenty years after the crime of aggression was included within the subject-matter jurisdiction of the Court as one of the four core crimes and seven years after the adoption of a ‘provision on the crime of aggression’ in accordance with article 5(2) of the Rome Statute (hereinafter all articles refer to the Rome Statute, unless otherwise clarified), the Review Conference’s determination to activate this jurisdiction ‘as early as possible’ (preambular paragraph 6 of Resolution 6) will be put to a test. Once the ‘decision to be taken after 1 January 2017’ (common para. 3 of Arts. 15bis and 15ter) will have been reached, the Court will be able to exercise jurisdiction ‘with respect to crimes of aggression committed one year after the ratification or acceptance of the amendments by thirty States Parties’ (common para. 2 of Arts. 15bis and 15ter), a temporal condition that was met on 26 June 2017.

In preparation of its upcoming session, the Assembly established ‘a facilitation, based in New York, open only to States Parties, to discuss activation of the Court’s jurisdiction over the crime of aggression’ (ICC-ASP/15/Res.5, annex I, para. 18(b)). The activation decision is a mere procedural step. However, given the continuing unease voiced by a few States, the facilitation seems to also provide a platform for discussions on substance. This post will address arguments that were presented during the facilitation process by academic experts, Prof. Dapo Akande, Prof. Roger Clark, Prof. Kevin Heller, and Prof. Noah Weisbord. Considering that the current discussion has narrowed down to the question of the Court’s jurisdictional reach, so will this post.

1. The Kampala compromise

Notwithstanding the two temporal conditions for the exercise of jurisdiction over the crime of aggression (entry into force of a minimum number of ratifications and the activation decision), the Kampala compromise is embedded in the jurisdictional regime of the Rome Statute. Following a referral by the Security Council of the United Nations, the Court may exercise its jurisdiction in accordance with Art. 13(b) (Art. 15ter). Following a referral by a State party to the Rome Statute or a proprio motu investigation by the prosecutor, the preconditions for the exercise of jurisdiction as defined in Art. 12 apply (Art. 15bis referring to Art. 13(a) and (c)), albeit with two restrictions.

First, Art. 12 only applies in situations involving ‘a crime of aggression, arising from an act of aggression committed by a State Party, unless that State Party has previously declared that it does not accept such jurisdiction by lodging a declaration with the Registrar’ (Art. 15bis(4)). In other words, the Court may not exercise jurisdiction over an act of aggression committed by a non-State party or committed by a State party that has previously opted-out.

Second, the Court shall not exercise its jurisdiction over the crime of aggression when committed by nationals or on the territory of a non-State party (Art. 15bis(5)). Accordingly, the exercise of jurisdiction is further excluded over any crime of aggression arising from an act of aggression committed against a non-State party as well as over any non-State party national contributing to a crime of aggression over which the Court otherwise could exercise its jurisdiction. As part of the compromise reached in Kampala, States parties further decided that the provision on aggression shall enter into force only for those States that ratify or accept the amendments (in accordance with Art. 121(5)). A more detailed analysis can be found here, here, and in this Chart.

The conditions for the Court’s exercise of jurisdiction over the crime of aggression allow for broad exceptions from the Court’s default jurisdictional regime. Those regarding non-States parties are absolute exceptions and unprecedented in the general jurisdictional regime of the Statute. Those regarding States parties are less expansive but still represent serious deviations from Part 2 of the Statute. The exceptions were strongly criticised, for being too far-reaching as well as for not being sufficiently far-reaching, for establishing a different jurisdictional regime relating to States parties that do not accept the amendments on the one hand and to non-States parties on the other hand, or for establishing a third jurisdictional regime different from Part 2 of the Statute and from Art. 121(5). Still, they represent the compromise that was negotiated in good faith and adopted by consensus in Kampala in 2010. It is a compromise that determines who should be covered by the Court’s jurisdiction over the crime of aggression, which has so far been ratified by 34 States parties, and on the basis of which an opt-out declaration has once been lodged.

2. Does the political compromise of Kampala include application of Art. 121(5) second sentence?

A preliminary note on the meaning of Art. 121(5) second sentence: the overwhelming view holds that Art. 121(5) second sentence establishes a distinct jurisdictional regime for crimes covered by an amendment when committed by nationals or on the territory of a State party that does not accept that amendment. Art. 121(5) second sentence is not directed at non-States parties. The Court’s jurisdictional reach over nationals and the territory of non-States parties remains governed by Part 2 of the Statute and any deviation thereto would require a formal amendment (States should therefore refrain from eroding Part 2 through an enabling resolution). But Art. 121(5) second sentence provides a privileged position for States parties regarding the Court’s jurisdiction over amended crimes. I have argued elsewhere that this provision should be interpreted systematically and in light of the object and purpose of the Statute to not have such a wide scope. The subject-matter jurisdiction of the Court, which is ‘limited to the most serious crimes of concern to the international community as a whole’ (Art. 5(1)), sets a high threshold for amendments. Either a crime is generally accepted to fulfil this criterion and warrants inclusion in the Statute, or it does not and should consequently not be the subject of an amendment. If all crimes equally fulfil this criterion, it is debateable why the sole fact of a later inclusion should submit ‘amended most serious crimes’ to a different jurisdictional regime. Applying two different jurisdictional regimes leads to particularly odd results in the interpretation of the ‘Belgian amendments’ that expand war crimes in the context of a non-international armed conflict. The Court’s jurisdictional reach over the same conduct includes nationals and the territory of States parties when it is committed in the context of an international armed conflict. When it is committed in the context of a non-international conduct, the Court’s reach over nationals or the territory of States parties that have not (yet) accepted the amendment is excluded. The precluded reach over ‘territory’ deprives a non-accepting State party from the protection it otherwise enjoys in the context of an international armed conflict. The precluded reach over ‘nationals’ seems to privilege foreign fighters that join an organized armed group in a fight against governmental authorities. They would be exempt from the Court’s reach, whereas members of the regular armed forces (which under the same constellation would internationalize the conflict) could be prosecuted by the Court. For these reasons, I uphold my previous position, but this interpretation remains a minority view and was not further discussed or acted upon during the negotiations on the crime of aggression. The Kampala compromise is without doubt based on the generally accepted interpretation of Art. 121(5) second sentence. This post will therefore proceed on this basis.

The main controversial issue is currently whether Art. 121(5) second sentence applies to the aggression amendments. To answer this question, it is fundamental to first establish whether the application or non-application of that provision is part of the compromise adopted in Kampala.

A textual interpretation of Art. 15bis and 15ter suggests that the jurisdictional regime established for the crime of aggression differs and partly directly contradicts Art. 121(5) second sentence. First, Art. 15bis and 15ter foresee different conditions for the exercise of jurisdiction for different trigger mechanisms. Art. 121(5) second sentence applies to all trigger mechanisms; at least in the ordinary meaning of its words. Second, Art. 15bis defines an absolute exception to the exercise of jurisdiction over crimes committed by nationals or on the territory of non-States parties. Art. 121(5) second sentence does not exclude the exercise of jurisdiction with regard to non-States parties and it was argued that the main purpose of the provision was not to open such an exception, but instead to privilege States parties and thus provide an incentive for non-States parties to join the Statute. Third, Art. 15bis is based on the assumption that nationals and the territory of States parties that do not accept an amendment are within the Court’s jurisdictional reach. Art. 121(5) second sentence is generally interpreted to provide the contrary. Fourth, a State party that opts-out in accordance with Art. 15bis is still be protected by the Court’s jurisdictional reach as a victim of an act of aggression. Art. 121(5) second sentence on the other hand does not concern itself with State acts and would provide for the same limited jurisdictional regime for aggressor States and their victims.

The aggression amendments do not only constitute a compromise in substance. A major component of the Kampala compromise was to establish agreement on the question of which of the three available amendment mechanisms was to be applied. This also included the question of how the conditions for the exercise of jurisdiction were to be consolidated with the consequences that these mechanisms may provide. The possibility that a compromise may not include Art. 121(5) second sentence had been discussed in detail prior to Kampala. The travaux préparatoires of Kampala provide clear evidence that the non-applicability of Art. 121(5) second sentence was part of the negotiations leading to the compromise. This is particularly underlined by the explanation of position by the delegation of Japan, criticizing ‘cherry picking’ from the relevant provisions of the Statute from a legal point of view.

Finally, it would be incomprehensible why States should have invested in lengthy and arduous negotiations to find a compromise if its essence would subsequently be reversed by way of an amendment procedure. It is therefore understood that the reference to article 121(5) in operative para. 1 of Resolution 6, in the context of the expressions ‘adoption, in accordance with Art. 5(2)’ and ‘shall enter into force in accordance with Art. 121(5)’, must be read as referring to the entry into force leg of article 121(5), namely its first sentence.

3. Does the non-application of Art. 121(5) second sentence violate the Statute?

Having established that the non-applicability of Art. 121(5) second sentence was part of the political compromise reached in Kampala, the key question remains whether this was lawfully so.

In order to address this question, it is necessary to recall the relevant provisions of the Rome Statute, which were themselves the result of a compromise that allowed removing the brackets around the crime of aggression during the final days of the Rome Conference. Accordingly, the crime of aggression falls within the subject-matter jurisdiction of the Court, but the Court shall only exercise its jurisdiction once a provision would be adopted, in accordance with articles 121 and 123, defining the crime and setting out the conditions for the exercise of jurisdiction thereover (Art. 5(1)(d) and Art. 5(2)).

The mandate of Art. 5(2) has generally been interpreted as giving the negotiators wide discretion in order to agree upon a provision on the crime of aggression. That includes that the conditions for the exercise of jurisdiction over the crime of aggression might differ from the general jurisdictional regime established in the Statute. On the procedural level, Art. 5(2) refers to ‘adoption’ in accordance with Art. 121 at a review conference (Art. 123). This reference is indisputably imprecise and it has therefore generated three readings: (i) that a provision on the crime of aggression merely required adoption in accordance with Art. 121(3); (ii) that it required adoption and entry into force in accordance with Art. 121(4); and (iii) that it required adoption and entry into force in accordance with Art. 121(5).

It is inherent in the vagueness of Art. 5(2) that all three readings may be legally substantiated or criticised. However, the decision taken in Kampala renders these competing arguments obsolete. The question today is not anymore, which argument is the strongest one (and therefore merits reflection in the compromise decision). The question today is rather whether the decision taken in Kampala can be reasonably argued within the legal framework of the Statute. Against this background, the provisions of the Statute may support different justifications, which are not mutually exclusive, on the non-applicability of Art. 121(5) second sentence to the provision on the crime of aggression; a position that was expressed in academia and that was supported by a majority of States in the negotiation process.

The Kampala compromise has its foundation in the Art. 5(1) and Art. 12(1) ‘jurisdiction’ argument. Art. 5(1) clearly provides that the crime of aggression falls within the jurisdiction of the Court. This understanding is emphasised by Art. 12(1), which provides that States parties accept the Court’s subject-matter jurisdiction, including over the crime of aggression, upon acceptance or ratification of the Statute. The crime of aggression has a specific position, insofar, as it was expressly listed in the Statute at the time of its adoption. Since the Court’s jurisdiction was already accepted by all States parties, it may be argued that the provision on the crime of aggression does not require further acceptance.

The Art. 5(2) ‘adoption’ argument comes to a similar result. It contends that the aggression amendments are based on Art. 5(2) and require mere adoption in accordance with Art. 121(3). This view was supported by some States during the negotiation process. In Kampala, this minority view was joined by a large number of States that shared arguments based on the specific position of the crime of aggression within the Statute but would not accept a solution without an entry into force mechanism. Under the premise not to be bound by either Art. 121(4) or 121(5), the Kampala conference consequently agreed on an individual entry into force of the amendments in accordance with (and as provided by) the first sentence of Art. 121(5).

According to the Art. 5(2) ‘conditions’ argument, Art. 121(5) may in principle apply to the aggression amendments, albeit subject to the mandate provided by Art. 5(2). Given the broad authority to define specifically the conditions for the exercise of jurisdiction of the crime of aggression in the mandate of Art. 5(2), it is not convincing that the drafters would have been mandated to override conditions prescribed in Part 2 with regard to the crime of aggression but would be limited by conditions foreseen in Art. 121(5) second sentence.

All these elements were part of debates before and in Kampala and helped pave the way for the compromise. They were reflected in the decision of the review conference ‘to adopt’ the provision on aggression ‘in accordance with article 5, paragraph 2’.

1. Does the non-application of Art. 121(5) second sentence violate the law of treaties?

Article 40(4) of the Vienna Convention in the Law of Treaties (VCLT) stipulates that unless the treaty provides otherwise, ‘[t]he amending agreement does not bind any State already a party to the treaty which does not become a party to the amending agreement’. The rule reflects the consent principle, which is also the basis of the general rule regarding third States formulated in Art. 34 VCLT, ‘[a] treaty does not create either obligations or rights for a third State without its consent’.

The Rome Statute does provide otherwise than the VCLT in its general amendment procedure. Art. 121(4) of the Statute foresees an entry into force erga omnes partes after acceptance of an amendment by seven-eighths of the States parties. The Statute also provides otherwise for amendments to provisions of an institutional nature, which are expressly listed in Art. 122. The amendment procedure according to Art. 121(5), on the other hand, follows the default rule. The first sentence of Art. 121(5) guarantees that an amendment (to articles 5 to 8 of the Statute) only enters into force for those States parties that have accepted it. Differently, Art. 121(5) second sentence does not deal with questions of treaty law but with the consequences of exercising (international) criminal jurisdiction over individuals. It defines conditions for the exercise of jurisdiction over crimes covered by an amendment and it thereby establishes a separate jurisdictional regime, different from Part 2 of the Statute.

Art. 15bis equally establishes a distinct jurisdictional regime, which differs from Part 2 of the Statute as well as from Art. 121(5). The common denominator, however, is that all three jurisdictional regimes, the one under Part 2, the one under Art. 121(5), and the one based on Art. 5(2), may affect nationals or the territory of a State that has not accepted the treaty or an amendment and that is consequently not bound by them. Whether the consequences of the exercise of international criminal jurisdiction over individuals (the Court’s jurisdictional reach) are binding or in any way create obligations or rights for third States was extensively discussed in the aftermath of the Rome Conference. The general view has been that consequences of the jurisdictional reach do not affect treaty relations and that there was no requirement of State consent for the exercise of (international) criminal jurisdiction. Consequently, the Court’s jurisdictional reach does not establish a new treaty regime, let alone one with obligations erga omnes. The requirement of 30 ratifications remains a condition for the Court’s exercise of jurisdiction over the crime of aggression.

The aggression amendments also do not establish cooperation obligations for States parties that do not adhere to them. It has been argued that a general obligation to cooperate fully with the Court in its investigation and prosecution of the crime of aggression already stems from Art. 86, which, at the time of adoption, acceptance or ratification of the Statute, referred to all ‘crimes within the jurisdiction of the Court’. If such a general duty to cooperate in the investigation and prosecution of crime of aggression did not exist in the Statute prior to the aggression amendments, but was created by them, it would only create obligations for States that accept them.

What remains, is the claim that the non-application of Art. 121(5) second sentence violates a treaty right that was established by the Statute for States parties to shield their nationals and their territory from the Court’s exercise of jurisdiction of the crime of aggression. Still, the same right is granted to States parties under the aggression amendments. The dispute is consequently reduced to the modification that while Art. 121(5) second sentence grants this right unconditionally, under the Kampala compromise it is granted upon declaration. However, this claim remains a theoretical one. If Art. 121(5) does not apply to the provision on aggression in the first place, the aggression amendments cannot affect any obligations or rights of States parties enshrined therein. The same is true, if the application of Art. 121(5) second sentence is subjected to the mandate of setting out conditions for the exercise of jurisdiction under Art. 5(2). In that case, all obligations and rights would lawfully be subjected to possible alterations within the mandate of Art. 5(2).

2. Does the non-application of Art. 121(5) second sentence violate general international law?

It has been argued that the crime of aggression is different from the other three core crimes because it requires the Court to determine a question of State responsibility as a precondition for the exercise of jurisdiction and that the determination of an act of aggression may require consent of the States concerned [for a detailed discussion, see Astrid Reisinger Coracini & Pal Wrange, ‘Is the Crime of Aggression Different from the other Crimes under International Law?’ in: Claus Kreß/Stefan Barriga (eds.), The Crime of Aggression – A Commentary (Cambridge University Press, 2016) p. 307-350].

The discussion is somewhat moot as the jurisdictional regime established by the aggression amendments is factually a consent-based regime. The Kampala compromise imposes an even stricter regime than Art. 121(5) second sentence. The requirement of a double consent (of the State where the conduct occurred and of the State of nationality of the accused) for the exercise of jurisdiction is required with regard to States parties and non-States parties. But independent of the Kampala compromise, is there a consent requirement under international law for the exercise of jurisdiction of the crime of aggression?

A noteworthy case was made in that regard on the basis of the ICJ’s jurisprudence based on Monetary Gold Removed from Rome in 1943 (Italy v. France, United Kingdom of Great Britain and Northern Ireland and United States of America). The consent principle is without doubt a fundamental principle under international law. It also underlies the jurisdiction of international courts adjudicating disputes between States and was therefore applied by the ICJ and other consent based jurisdictions in proceedings against States. It is however contended whether the principle is directly applicable in the context of international criminal jurisdictions, the consequences of which may affect nationals or the territory of non-consenting States, in proceedings against individuals.

Even under the ICJ’s consent-based jurisdictional regime, an indirect determination of acts of non-consenting States may be possible, if it remains without legal consequences for these States. In Certain Phosphate Lands in Nauru (Nauru v. Australia), the ICJ held that it was in no way precluded from adjudicating upon the claims against one out of three States that shared responsibility in the administration of the territory in question, provided the legal interests of the third States which might possibly be affected did not form the actual subject-matter of the requested decision.

It is undisputed that an act of aggression by a State is one of the elements of the definition of the crime of aggression that needs to be established for individual criminal responsibility to arise. Assessing the elements of a crime is not a formal determination of State responsibility. It requires the establishment of facts and their legal evaluation as a precondition for reaching a verdict. There are no consequences for a State that follow from the Court’s determination of an act of aggression. The ‘very subject-matter’ of the Court’s decision remains the innocence or guilt of an individual.

The ‘requirement’ to determine an act of State may be unique to the crime of aggression, however, this distinction is merely semantic. A consent requirement cannot be based on the frequency of the determination of an act of State but it is a matter of principle. Either the determination of an act of State by an international criminal jurisdiction requires State consent or not. In that regard, all four core crimes can be (and frequently are) committed by State organs. Each determination of guilt or innocence of an organ of a State or any other person whose conduct is attributable to a State, comprises ipso facto a determination of an act of State. This indirect determination of State responsibility has not been an obstacle to establishing a Court with jurisdiction over nationals of non-consenting States.

In conclusion

1. The political compromise that was reached in Kampala excludes the application of Art. 121(5) second sentence to the crime of aggression. This non-applicability can be reasonably argued on the basis of the relevant provisions of the Rome Statute.
2. Art. 121(5) is not the starting point but one out of several possible answers to the question which procedure applies to the provision of aggression. The mandate of Art. 5(2) to adopt such a provision is vague and allows for different interpretations. While there were competing claims as to which procedure was the best position under international law, a full application of Art. 121(5) was only one of them. The compromise decision taken in Kampala is based on a different legal position, which acknowledges the special position of the crime of aggression under the Statute, and which was overwhelmingly supported by a majority of States.
3. The majority understanding of Art. 121(5) second sentence is that it establishes a specific jurisdictional regime for crimes covered by an amendment, which differs from Part 2 of the Statute. The Kampala compromise does not establish agreement on a different reading of Art. 121(5) second sentence. The Kampala compromise establishes agreement on a reading of Art. 5(2) and Art. 12(1) that leads to the non-applicability of Art. 121(5) second sentence to the crime of aggression.
4. If Art. 121(5) second sentence does not apply, the aggression amendments cannot infringe upon a right, enshrined in this sentence. The agreement reached at the Review Conference, which is based on one of several possible and reasonable interpretations of the applicable law, cannot violate rights a State may have enjoyed if an alternative interpretation had been agreed upon. This simply falls beyond the scope of Art. 34 and Art. 40(4) VCLT.
5. Consent to be bound by a treaty and consent to the jurisdictional reach by the Court are two separate matters. Only States that adhere to the aggression amendments are bound by them (as a matter of treaty law). The fact that nationals or the territory of a State may be affected by the exercise of the Court’s criminal jurisdiction over individuals does not qualify as a binding effect for that State.
6. It is difficult to comprehend how a view that the Court cannot exercise its jurisdiction regarding the crime of aggression over nationals of a State or committed on the territory of a States unless that State accepts or ratifies the aggression amendments could be ‘clarified’ without re-opening the text of the amendments. Any subsequent agreement in that regard would clearly reverse the contents of the compromise of Kampala and refute the negotiations process. It would also pose difficulties for the 35 States that have taken legal steps on the basis of the Kampala compromise.
7. Reopening the compromise, which was a package deal, would not only affect the exercise of jurisdiction over a crime of aggression, arising from an act of aggression by a State party that does not adhere to the aggression amendments. It would equally affect other components of the compromise. If Art. 5(2) was not the legal basis to establish conditions for the exercise of jurisdiction over the crime of aggression that might differ from Part 2 and from Art. 121(5) second sentence, Art. 15bis and Art. 15ter would need to be interpreted through the jurisdictional regime of 121(5). This would undermine compromise decisions relating to the opt-out regime, the preferential treatment of opting-out States parties as victims of an act of aggression, referrals by the Security Council, and ultimately the exercise of jurisdiction over nationals and the territory of non-States parties.
8. It was suggested that the opt-out regime of Art.15bis would not be deprived of its meaning if interpreted in light of Art. 121(5) second sentence, because it could allow States parties to opt out in order to fall within the Court’s jurisdictional protection as a victim of aggression. But there is no rational consolidated reading of these two provisions. An interpretation that suggests an opt-out from a jurisdictional regime that States do not consider themselves ‘in’ is unreasonable.
9. Furthermore, the possibility to opt out of the Court’s jurisdictional reach was a concession to those States parties that believed they had acquired a right under Art. 121(5) second sentence to be exempt from the effects of the Court’s exercise of jurisdiction over the crime of aggression. The drafters transposed such a right to the aggression amendments. In order to counterweigh the requisite of depositing a (low impact, yet public) declaration, they even added another right. States parties that opt out in accordance with Art. 15bis, remain under the Court’s protection should they become a victim of an act of aggression. This additional privilege was not given to States parties without a cause. It was a trade-off in order to establish consensus on the basis of the non-applicability of Art. 121(5) second sentence.

At the upcoming session of the Assembly of States Parties, it will be the responsibility of States to defend the compromise they have reached at the Kampala Review Conference. Discussions on the crime of aggression and the jurisdiction of an international criminal court over this crime have started at the time of the League of Nations and they could be easily continued ad infinitum; but it is time for a closure. It lies in the nature of a compromise decision that it cannot reflect all positions. States which supported a meaningful exercise of the Court’s jurisdiction over the crime of aggression under the existing framework of the Statute have succeeded in upholding some structural principles, but have made major concessions when it comes to the factual exercise of jurisdiction by the Court. In that regard, the Kampala compromise was particularly sensitive towards the concerns of some States. Eventually, the aggression amendments are not all that was wished for, but evidently, they are what lies on the table and what will be subjected to an activation decision in December.

Against the never-ending fascination to discuss legal aspects of one sentence of one article of the Rome Statute, the larger issues at stake must not be let out of sight. The Court’s jurisdiction over the crime of aggression should not be considered a mere nuisance when it comes to a decision on the use of armed force. Even under a narrow definition and with limited jurisdiction, the crime of aggression still protects one of the fundamental principles of international law, the prohibition of the use of force. Seven years after Kampala, the significance and urgency of strengthening the protecting scope of this principle unfortunately seems to have increased rather than declined.