Crime of Aggression
Equipo Nizkor
        Bookshop | Donate
Derechos | Equipo Nizkor       

11Jul11


The Uncertain Legal Status of the Aggression Understandings


Kevin Jon Heller |1|

I. Introduction

Annex III of Resolution RC/Res.6, adopted by consensus at Kampala on 12 June 2010, contains seven "Understandings regarding the amendments to the Rome Statute of the International Criminal Court on the crime of aggression." Understandings 1-3, which address Security Council referrals and the Court's jurisdiction ratione temporis, |2| were first discussed in the final meeting of the Special Working Group on the Crime of Aggression (SWG) in February 2009. |3| Understandings 4 and 5, which concern domestic jurisdiction, |4| were first proposed by the Chairman of the SWG on June 4. |5| And Understandings 6 and 7, which concern the meaning of a "manifest" violation of the UN Charter, |6| emerged out of a proposal offered by the United States at the Review Conference on June 7 and were discussed for the first time on the afternoon of June 9. |7|

As Claus Kress et al. have noted, "[t]he precise legal significance of what were to become the Understandings was neither debated nor decided upon in the course of the negotiations. There was, for example, no debate as to whether and where the Understandings are to be situated within the legal framework of article 31 of the Vienna Convention on the Law of Treaties." |8| The absence of such debate has led scholars to acknowledge that it is "unclear what legal value" the Understandings have |9| and that their "uncertain legal authority" makes it difficult to determine "whether and how the adopted Understandings will... influence prosecutions before the Court." |10| Indeed, no scholar to date has attempted to systematically analyze the legal status of the Understandings.

This essay attempts to fill that lacuna in the burgeoning literature on the aggression amendments. |11| It considers four possible interpretations of the Understandings: (1) that they are amendments to the Rome Statute; (2) that they are a primary means of interpreting the aggression amendments under Article 31 of the Vienna Convention on the Law of Treaties |12| (VCLT); (3) that they are an agreement to modify by one or more parties to the Rome Statute under Article 41 of the VCLT; and (4) that they are supplementary means of interpreting the aggression amendments under Article 32 of the VCLT. It concludes that, unless they are adopted by all of the States Parties to the Rome Statute at the 2017 Review Conference, the Understandings are nothing more than supplementary means of interpretation that the Court would have the right to ignore once the aggression amendments enter into force.

II. Are the Understandings Substantive?

The "precise legal significance" of the Understandings is irrelevant, of course, unless they would have a substantive effect on the Court's application of the crime of aggression. If the Court would decide each and every hypothetical aggression case the same way regardless of whether it took the Understandings into account, their legal status is irrelevant. By contrast, if the Court would decide even one hypothetical aggression case differently depending on whether it applied the Understandings, determining their legal significance is critically important.

It seems unlikely that Understandings 1-4 would have any substantive effect. Understanding 1 simply reiterates Article 15ter, which provides that the Court cannot exercise jurisdiction on the basis of a Security Council referral until after the aggression amendments are ratified by at least 30 States Parties and formally adopted at the 2017 Review Conference. |13| Understanding 3 similarly parallels Article 15bis, which governs state referrals and proprio motu investigations of acts of aggression. |14| Understanding 2 makes it clear that states do not have to accept the Court's jurisdiction with regard to Security Council referrals, which is the default position under the Rome Statute. |15| And Understanding 4 mirrors Article 10 of the Rome Statute, which provides that the definitions of crimes in Part II do not limit or prejudice "existing or developing rules of international law for purposes other than this Statute."

Understanding 5, by contrast, may have substantive effect. As Van Schaack notes, the Review Conference "ultimately decided to leave the default complementarity regime intact vis-a-vis the crime of aggression." |16| Understanding 5's insistence that "the amendments shall not be interpreted as creating the right... to exercise domestic jurisdiction with respect to an act of aggression committed by another State" has nevertheless led one scholar, Carrie McDougall, to express concern that the Understanding could be interpreted to exempt aggression from Article 17's complementarity regime, thus prohibiting any state other than the aggressor from challenging the admissibility of a prosecution on the ground that it was investigating or prosecuting the act of aggression itself. |17| Van Schaack rejects that interpretation, |18| but she acknowledges that Understanding 5 "expresses a subtle preference that states not incorporate the crime into their domestic codes." |19| McDougall's interpretation clearly gives Understanding 5 substantive effect, and even Van Schaack's means that the Court's decision to recognize Understanding 5 would affect the development of the crime of aggression.

Understanding 7 may also have substantive effect. Read together, the two sentences that comprise the Understanding indicate that an act of aggression |20| cannot qualify as a "manifest" violation of the UN Charter unless at least two of the three criteria specified in Article 8bis(1) - character, gravity, and scale - are satisfied. Although that interpretation of Understanding 7 is arguably consistent with the conjunctive ("and") language of Article 8bis(1), |21| Van Schaack believes that the Understanding "slightly" raises the threshold for the finding of a manifest violation. |22| Trahan also acknowledges that possibility. |23|

Finally, it is clear that Understanding 6 has significant potential to affect how the Court applies the crime of aggression. There is a critical textual difference between Article 8bis(1) and Understanding 6 concerning whether an aggressive act qualifies as a manifest violation of the UN Charter: whereas Article 8bis(1) focuses on the "character, gravity, and scale" of the aggressive act, Understanding 6 focuses on "the gravity of the acts concerned and their consequences." |24| That difference is not accidental: Understanding 6's emphasis on consequences reflects the United States' desire to ensure that the Court will not consider unilateral humanitarian intervention - intervention not authorized by the Security Council - to be criminal. The U.S. originally proposed an Understanding that would have specifically excluded unilateral humanitarian intervention from the crime of aggression. |25| When that proposal was (soundly) rejected, |26| the U.S. proposed the "consequences" language instead, thereby attempting to accomplish indirectly what it could not accomplish directly. As Van Schaack says:

    The focus on "consequences" in the Understandings allows for an opening to argue that a military operation that may have violated Article 2(4) of the U.N. Charter as a technical matter might not be deemed to constitute an act of aggression by virtue of the fact that it ultimately improved the situation on the ground by protecting civilians and vulnerable groups from further attack. |27|

It is an open question, of course, whether the Court would give substantive effect to Understanding 6. It is possible that it would exclude unilateral humanitarian intervention from the crime of aggression even in the absence of the Understanding. Trahan argues, for example, that the "character, gravity, and scale" criteria are designed to exclude legally debatable cases and that "truly humanitarian" unilateral interventions fall within that category. |28| But it seems more likely that the Court would not exclude unilateral humanitarian intervention solely on the basis of Article 8bis(1); after all, the UN Charter does not expressly permit such unilateral intervention, and both the Secretary General's High-Level Panel on Threats, Challenges and Change |29| and General Assembly Resolution 60/1 |30| specifically condition humanitarian intervention on Security Council authorization. |31| Regardless of who has the better of the argument - and a comprehensive analysis of the issue is obviously beyond the scope of this essay - it is nevertheless conceivable that the Court would exclude unilateral humanitarian intervention from the crime of aggression solely on the basis of Understanding 6.

As this analysis indicates, at least three Understandings could have substantive effect on how the Court applies the crime of aggression. It is thus critically important to determine what the legal status of the Understandings might be.

III. Do the Understandings Amend the Rome Statute?

The first possibility is that the Understandings in Annex III qualify as amendments to the Rome Statute. A Review Conference has the authority to adopt an amendment to the Rome Statute either by consensus or by a 2/3 majority of all States Parties. |32| Indeed, because of the poor drafting of Article 121(3), it appears that a Review Conference can amend the Rome Statute by consensus even if fewer than 2/3 of the States Parties are present. |33| Such a procedure is inconsistent with Article 9 of the VCLT, which requires either the consent of all parties to the original treaty or a 2/3 vote, but Article 39 of the VCLT permits a treaty to contract out of the requirements of Article 9. |34|

There is little evidence, however, that the Understandings are amendments. Most obviously, unlike the provisions in the other two Annexes, they do not actually amend anything. Annex I amends the Rome Statute by deleting Article 5(2); adding new Articles 8bis, 15bis, and 15ter; and modifying Articles 25, 9, and 20. Annex II amends the Elements of Crimes - a source of law for the Court pursuant to Article 21(1)(a) of the Rome Statute - by adding the elements of new Article 8bis. |35| Annex III, by contrast, simply recounts how the Review Conference believes the new provisions in the Rome Statute and the new Elements should be interpreted.

The text of Resolution RC/Res.6 also indicates that the Understandings are not amendments. The Preamble specifically distinguishes between "amendments" and the "understandings," noting that the Review Conference "also decides to adopt the understandings regarding the interpretation of the [aggression] amendments contained in annex III of the present resolution." |36| That distinction is then reinforced by the titles of the various annexes: whereas Annexes I and II are styled "Amendments to the Rome Statute" and "Amendments to the Elements of Crimes," respectively, Annex III is styled "Understandings regarding the amendments."

Finally, the drafting history of Resolution RC/Res.6 indicates that the Understandings were not viewed as amendments to the Rome Statute. The first significant discussion of the Understandings came in the final meeting of the SWG in February 2009, when the Chairman suggested, in response to a number of issues concerning the definition of aggression, that "[t]he Review Conference could adopt language on these issues simultaneously with the amendments on aggression, in an appropriate format that is yet to be discussed." |37| Some of those issues were eventually addressed by Understandings 1-3; others, however, "were instead addressed directly in the text of the amendments." |38|

Although it is clear that the Review Conference did not explicitly amend the Rome Statute to recognize the Understandings, it might still be possible to argue that the Conference implicitly amended it. The most logical interpretation would be that Resolution RC/Res.6 amended Article 21 ("Applicable Law") to include the Understandings as a source of law that the Court must apply, thereby placing the Understandings on the same plane as the Rome Statute, the Elements of Crimes, and the Rules of Procedure and Evidence. |39|

This interpretation, of course, would raise the Article 121(4)/121(5) issue that so vexed the delegates to the Review Conference. An analysis of which amendment regime applies is beyond the scope of this essay, but it is interesting to note that Roger Clark speaks for nearly all of the scholars who have written about the issue when he claims that "the argument in favor of the seven-eighths solution is stronger on the basis of the plain language and it is consistent with the complex preparatory history." |40| Suffice it to say here that if it is questionable whether adding three new Articles to the Rome Statute (Articles 8bis, 15bis, and 15ter) and modifying three others (Articles 25, 9, and 20) qualifies as an "amendment to articles 5, 6, 7 and 8" -a view that Manson describes as "simply an outright abuse of the current language of Article 121(5)" |41| - it is even more questionable whether the implicit addition of a new source of law to Article 21 could be accomplished in such a fashion.

IV. Are the Understandings A Primary Means of Interpretation?

To date, no scholar has argued that the Review Conference implicitly amended Article 21 when it adopted the Understandings. |42| By contrast, although the Conference did not address the issue, at least two scholars have taken the position that the Understandings qualify as primary means of interpretation under Article 31 of the VCLT. Joanna Harrington, for example, contends that "[a] connection can be drawn between these understandings and article 31... and more specifically the use of any subsequent agreement between the parties regarding the interpretation of a treaty or the application of its provisions." |43| Similarly, although Marko Milanovic acknowledges that "the Statute itself doesn't mention recourse to any understandings," he insists that "the Statute is a treaty, and it is a general rule of treaty interpretation that recourse must be had to any agreement between the parties as to the treaty's interpretation, either as part of the treaty's 'context', or subsequent to its conclusion - see Art. 31(2) and (3) VCLT." |44|

This is a plausible argument. Upon examination, however, it becomes clear that the Understandings cannot - at least on the basis of the consensus achieved at the Review Conference - qualify as a primary means of interpretation.

A. Article 31(2)

Article 31(2) of the VCLT provides, in relevant part:

    2. The context for the purpose of the interpretation of a treaty shall comprise, in addition to the text, including its preamble and annexes:

      (a) any agreement relating to the treaty which was made between all the parties in connection with the conclusion of the treaty;
      (b) any instrument which was made by one or more parties in connection with the conclusion of the treaty and accepted by the other parties as an instrument related to the treaty.

The Understandings do not qualify as "context" for the Rome Statute under these provisions. Subparagraphs (a) and (b) of Article 31(2) specifically require the relevant agreement or instrument to be made "in connection with the conclusion of the treaty." Similarly, although Article 31(2) does not expressly impose such a temporal restriction on annexes, the requirement is nevertheless implicit in the paragraph. As the ILC pointed out in its commentary on what became Article 31, although both context and subsequent agreements qualify as primary means of interpretation, the division between paragraphs (2) and (3) is a temporal one: Article 31(2) includes all text, Preambular provisions, annexes, and other agreements/instruments that were part of the treaty as originally concluded, while Article 31(3) - discussed below - includes all agreements concluded subsequent to the adoption of the original treaty that affect how the original treaty should be interpreted. |45| The Understandings were obviously not adopted in connection with the conclusion of the Rome Statute itself; they were adopted more than a decade later, in connection with the conclusion of Resolution RC/Res.6, which amended the Rome Statute. Their status as primary means of interpretation is thus governed by Article 31(3), not Article 31(2).

It could still be argued, of course, that Article 31(2) applies not only to agreements made in connection with the conclusion of a treaty, but also to agreements made in connection with the conclusion of an amendment to a treaty. Such an interpretation finds no support in the text of the provision or in the ILC commentary - but even if it did, it would not save the Understandings. The Understandings would still have to satisfy the requirements of either subparagraph (a) or (b), and they cannot. Article 31(2)(a) requires the interpretative agreement be "made between all the parties" to the treaty - an expression that requires each party to be involved in creating an agreement, as opposed to simply accepting an agreement created by other parties. |46| There is some uncertainty about how many voting-eligible States Parties were present at the Review Conference when the Understandings were adopted on June 12 as part of Resolution RC/Res.6: Van Schaack estimates as many as 85 |47|; Schabas estimates 80, while acknowledging that some might have left by the time of the "vote" |48|; Blokker and Kress estimate a "few more than 74." |49| Regardless, a minimum of 26 States Parties were not involved in "making" the Understandings.

The absence of so many States Parties from the Review Conference also dooms any attempt to justify the Understandings under Article 31(2)(b). That subparagraph does not require all of the parties to "make" an interpretative instrument, but it does require them to "accept" that the instrument is related to the treaty - a requirement that is satisfied only if all of the parties (1) accept that the instrument concerns the interpretation of the treaty, and (2) agree with "the content of the particular instrument. concerning the treaty term to be interpreted." |50| It is possible for States Parties to signal their acceptance of an interpretative instrument through acquiescence; a positive act is not required. |51| Indeed, that possibility seems to underpin Harrington's defense of the Understandings: namely, that "States [we]re aware" that the amendment provisions in the Rome Statute, "albeit poorly drafted, require presence for voting... but some chose not to be present. One cannot argue afterwards that they did not know or that there was lack of notice of the meeting, etc." |52|

That argument, however, is both factually and legally problematic. It is factually problematic because it assumes (1) that all of the States Parties who did not attend the Review Conference knew precisely what the delegates intended to adopt by consensus; and (2) that, if the absent States Parties disagreed with the Understandings, they could have registered their disagreement by breaking consensus and forcing a vote. Neither assumption is tenable. Understandings 6 and 7 were not formally introduced at the Review Conference until June 7, the second week of negotiations, and the Understandings as a whole were not finalized until the morning of June 10. |53| Absent States Parties thus had less than 36 hours to obtain the final draft of the Understandings, decide that they were opposed to them, and appear at the Review Conference to register their disagreement. Under those circumstances, it strains credulity to claim that the failure to disagree was tantamount to acquiescence.

That claim is also legally problematic. Nothing in the VCLT, the ILC commentary, or in the scholarly literature suggests that a State Party who chooses not to attend a review conference should be presumed to accept through acquiescence any agreement that the conference adopts. Indeed, given the variety of reasons why States Parties might have chosen not to attend Kampala - time, resources, competing obligations, etc. - any such presumption would contravene the obligation to determine acquiescence in good faith, |54| a duty that requires States Parties "to act honestly, fairly and reasonably, and to refrain from taking unfair advantage." |55| At a minimum, States Parties would need to be given a reasonable amount of time to indicate their disagreement with the meaning of the Understandings before they could properly be considered "context" under Article 31(2)(b).

More than a year after Kampala, of course, no absent State Party has objected to the Understandings. It could thus be argued that those States Parties have already had a reasonable amount of time to register their disagreement. Given the uncertainty that surrounds the meaning of the Understandings, however, any such inference would be unwarranted. The reason that Article 31(2)(b) instruments represent primary means of interpretation is precisely that "all parties themselves agree on (or at least accept) the interpretation of treaty terms by means which are extrinsic to the treaty." |56| States Parties that do not know how the Understandings will affect the crime of aggression cannot be said to have accepted a particular interpretation of the aggression amendments by failing to object to them. |57|

B. Article 31(3)

As noted above, the more logical interpretation of the Understandings is that they are a subsequent agreement between the parties to the Rome Statute. Article 31(3) of the VCLT provides, in relevant part:

    3. There shall be taken into account, together with the context:

      (a) any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions;
      (b) any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation;
      (c) any relevant rules of international law applicable in the relations between the parties.

To qualify as primary means of interpretation under Article 31(3)(a), an agreement must satisfy three requirements: (1) it must be a "subsequent agreement"; (2) it must be "between the parties"; and (3) it must concern "the interpretation of the treaty or the application of its provisions." |58| The Understandings clearly satisfy the first and third requirements. But they do not satisfy the second. Although subparagraph (a) does not specifically state that the subsequent agreement must involve "all" of the parties to the treaty, the unanimity requirement is implicit in it. Most importantly, Article 31(3) does not contain an equivalent to Article 31(2)(b), which specifically deems context an instrument that was "made by one or more parties" to a treaty and then "accepted" by all of the other parties. The absence of an equivalent provision indicates that Article 31(3) requires subsequent agreements to actually be "made" by all of the parties to the original treaty - thus ruling out the acquiescence argument available for contemporaneous agreements. |59|

That interpretation is reinforced by the internal structure of Article 31(3). The paragraph uses "the parties" in subparagraph (b), as well, and it is clear from the text of subparagraph (b) that subsequent practice concerning the interpretation of a treaty qualifies as a primary means of interpretation only if all of the parties to the original treaty agree on the meaning of the subsequent practice. Indeed, as the ILC points out in its commentary on the provision, Article 31(3)(b) originally did make use of "all the parties":

    The text provisionally adopted in 1964 spoke of a practice which "establishes the understanding of all the parties." By omitting the word "all" the Commission did not intend to change the rule. It considered that the phrase "the understanding of the parties" necessarily means "the parties as a whole." It omitted the word "all" merely to avoid any possible misconception that every party must individually have engaged in the practice where it suffices that it should have accepted the practice. |60|

Finally, it is worth noting that scholarly opinion uniformly agrees that Article 31 uses "the parties" as a metonym for "all the parties." According to Linderfalk, "[b]y 'the parties' all parties are meant. In the legal regime established by Articles 31-33 of the Vienna Convention, a recurring theme is that normally, a phenomenon cannot be included in the context for the interpretation of a treaty, unless all parties have accepted it." |61| Paulwelyn agrees, noting that "[e]xamining the preparatory work of these provisions it seems, indeed, that the 'subsequent agreements' and 'subsequent practice' referred to are only agreements and practice reflecting the common intentions of all the parties to the treaty." |62| Paulwelyn also quotes Sir Humphrey Waldock's observations on the 1964 draft of the VCLT to that effect with regard to the "subsequent practice" prong of what is now Article 31(3):

    [I]t is only subsequent practice which clearly establishes the understandings of all the parties regarding the meaning of the treaty which is recognized as an equivalent to an interpretative agreement and the reason is, of course, that two parties or even a group of parties cannot, by their interpretation of the treaty, bind the other parties as to its correct interpretation. |63|

Given the unavailability of the acquiescence argument, the unanimity requirement implicit in Article 31(3) means that the Understandings cannot qualify as a subsequent agreement. The agreement that included the Understandings, Resolution RC/Res.6, did not involve all of the States Parties to the Rome Statute; at most, it involved 85, slightly more than 2/3 of them. Nor can it be argued that the States Parties present at Kampala were somehow empowered to adopt the Understandings on behalf of the absent States Parties. Although Articles 39 |64| and 40(1) |65| of the VCLT permit States Parties to adopt treaties that modify the rules contained in Part II of the VCLT ("Conclusion and Entry into Force of Treaties") and in Article 40 itself ("Amendment of Multilateral Treaties"), nothing in the VCLT allows States Parties to contract out of Article 31(3)(a)'s unanimity requirement, which is contained in Part III ("Observance, Application, and Interpretation of Treaties"). In any case, the Rome Statute does not actually contain a provision that gives the Review Conference the authority to adopt a "subsequent agreement... regarding the interpretation of the treaty or the application of its provisions" by consensus; Article 121 applies only to "amendments" to the Rome Statute, and we have already seen that it is difficult, if not impossible, to view the Understandings as amendments.

Indeed, it is worth noting that the United States - the most enthusiastic proponent of the Understandings - has traditionally taken the position that a minority of States Parties to a treaty cannot adopt a subsequent agreement under Article 31(3)(a) either through consensus or by assuming the acquiescence of absent States Parties. In response to a dispute over the interpretation of the entry-into-force provision of the Basel Convention on the Control of Transboundary Movements of Hazardous Wastes and their Disposal, the U.S. argued the following in 2007:

    The United States further notes that if the Parties choose to develop such a subsequent interpretative agreement regarding Article 17 pursuant to VCLT Article 31(3)(a), it will need to be adopted by consensus at a Conference of the Parties, rather than by majority voting. In addition, there will need to be a process whereby Parties not present at the relevant Conference of the Parties have an opportunity to express their views - even if that means breaking consensus. Consensus is imperative because entry into force provisions provide the quintessential case in which there must be one rule for all of the Parties. It would not work for a phrase such as "who accepted them" to have different meanings for different Parties, depending on whether they join in a subsequent agreement interpreting the clause. The potential consequences for entry into force of Basel Convention amendments would be entirely too mischievous. |66|

The same logic applies to the Rome Statute, and the same conclusion flows inexorably from the VCLT. Until such time as all of the States Parties to the Rome Statute have "made" the Understandings by expressly adopting them - not simply those who were present at the Review Conference when they were adopted by consensus - the Understandings cannot be considered a subsequent agreement under Article 31(3).

V. Are the Understandings An Agreement to Modify?

Another possibility would be to view the Understandings as an agreement to modify a multilateral treaty "between certain of the parties only." Article 41 of the VCLT provides, in relevant part:

    1. Two or more of the parties to a multilateral treaty may conclude an agreement to modify the treaty as between them alone if:

      (a) the possibility of such a modification is provided for by the treaty; or
      (b) the modification in question is not prohibited by the treaty and:

        i. does not affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations;
        ii. does not relate to a provision, derogation from which is incompatible with the effective execution of the object and purpose of the treaty as a whole.

It seems clear that, in general, the possibility of an inter se agreement to modify the Rome Statute is "provided for by the treaty." After all, Article 121(5) specifically permits the ASP or a Review Conference to amend the core crimes in ways that are binding only on those States Parties who accept the changes. The Review Conference specifically relied on Article 121(5) when it adopted Resolution RC/Res.6; it could thus be argued that the aggression amendments and the Understandings fall under Article 41 of the VCLT.

There are, however, a number of difficulties with this view. Most obviously, the Rome Statute only permits certain kinds of inter se agreements to modify: namely, those that relate to Articles 5-8. If the adoption of Resolution RC/Res.6 is governed by Article 121(4) instead of Article 121(5) - an issue that the Court will ultimately have to decide |67| - neither the aggression amendments nor the Understandings will fall under the purview of Article 41 of the VCLT, and neither will have any operative force (assuming that they are ultimately activated in 2017) until 7/8 of the States Parties ratify them.

Even if the Review Conference properly relied on Article 121(5) when it adopted the aggression amendments, it is still unlikely that Article 41 would justify the Understandings. As Rigaux et al. have pointed out, "amendment" and "modification" are "one part of what may suitably be called a process of "revision" of international conventions. and international practice often utilizes these terms as synonyms." |68| It is difficult to see how the Understandings "revise" the Rome Statute, for all of the reasons discussed above. And if they do not, the Understandings cannot be considered an inter se agreement to modify under Article 41. At best, they could qualify as "context" for the interpretation of the aggression amendments under Article 31(2) or (3) of the VCLT - which is very unlikely, as noted earlier.

Assume, however, that Article 41 is not limited to "revisions" - amendments and modifications - of a multilateral treaty. The Rome Statute does not expressly prohibit the adoption of Understandings, even if its provision expressly permitting amendments to the core crimes seems to implicitly prohibit any other kind of inter se agreement. It is thus worth considering whether the Understandings are consistent with Article 41(1) (b) of the VCLT.

A strong case can be made that they are not. The first condition of that provision is that the unilateral agreement must not "affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations." As discussed above, the Understandings have three potentially important substantive effects: (1) narrowing Article 17's complementarity regime to prohibit any states other than the aggressor from challenging the admissibility of an case on the ground that it is investigating or prosecuting the act of aggression domestically; (2) raising, if only slightly, the threshold for a "manifest" violation of the UN Charter; and (3) excluding unilateral humanitarian intervention from the crime of aggression. If the Understandings represent a valid inter se agreement to modify the crime of aggression, the rights of States Parties under the Rome Statute will be affected in important ways. First, a number of states will lose the right to pre-empt an ICC aggression prosecution by investigating or prosecuting the case themselves. |69| Second, they will all have to choose between not ratifying the aggression amendments at all, thereby losing protection against any act of aggression committed by States Parties who have ratified the amendments but not opted out, and accepting an aggression regime that underprotects them against acts of aggression relative to the non-modified aggression regime with a lower threshold for a "manifest" violation or - worse still - that fails to protect them against "humanitarian" interventions that are not authorized by the Security Council. |70|

The second condition of Article 41(1)(b) is that the inter se agreement must not "relate to a provision, derogation from which is incompatible with the effective execution of the object and purpose of the treaty as a whole." The Understandings likely run afoul of that condition, as well. The primary object and purpose of the Rome Statute is, of course, to combat impunity. |71| That object and purpose would be undermined by excluding the crime of aggression from the complementarity regime in Article 17, |72| particularly given the limited number of cases the Court can investigate and prosecute. And it would at least arguably be undermined by making it more difficult to find that an act of aggression qualified as a manifest violation of the UN Charter or by excluding unilateral humanitarian intervention from the crime of aggression entirely.

These effects could be avoided, of course, if States had the right either to ratify the aggression amendments without the Understandings or to ratify Resolution RC/Res.6 as a whole and then opt-out of only the Understandings. |73| Nothing in the text of the Resolution or in the travaux preparatoires suggests, however, that either option is possible. As a result, even if Article 41 extends to the Understandings, they are likely invalid under Article 41(1)(b).

VI. Are the Understandings A Supplementary Means of Interpretation?

The final interpretation of the Understandings is that they are a supplementary means of interpreting the Rome Statute. Article 32 of the VCLT provides:

    1. Recourse may be had to supplementary means of interpretation, including the preparatory work of the treaty and the circumstances of its conclusion, in order to confirm the meaning resulting from the application of article 31, or to determine the meaning when the interpretation according to article 31:

      (a) leaves the meaning ambiguous or obscure; or
      (b) leads to a result which is manifestly absurd or unreasonable.

At present, this seems to be the most persuasive interpretation of the Understandings. As we have seen, it is very difficult to view the Understandings as an amendment to the Rome Statute, as a contemporaneous agreement, as a subsequent agreement, or as an inter se agreement. By default, then, they can be nothing more than travaux preparatoires, |74| useful for the Court only as permitted by Article 32.

This view, of course, would greatly limit the significance of the Understandings. Most importantly, the Court would not have to rely on them at all; as the ILC commentary notes, the application of a supplementary means of interpretation "is not automatic but depends on the conviction of the interpreter that it is appropriate in the particular circumstances of the case. In other words, recourse to. these principles is discretionary rather than obligatory." |75|

Even if the Court did have recourse to the Understandings, it would almost certainly apply them in a way that would drain them of any substantive effect. To begin with, the Court would be very unlikely to rely on the Understandings to clarify an "ambiguous or obscure" interpretation the aggression amendments. "In practice," according to Le Bouthillier, "one finds very few cases where international tribunals have concluded that the rule of interpretation in Article 31 leads to an ambiguous or obscure result." |76| There is no reason that the ICC would be any different. There is nothing ambiguous about the Review Conference's decision to apply the normal complementarity regime to the crime of aggression, and although there is clearly some ambiguity concerning what qualifies as a "manifest" violation of the UN Charter, |77| the Understandings are designed to make it more difficult to find a manifest violation, not to clarify the definition of "manifest."

For similar reasons, the Court would also be unlikely to rely on the Understandings to avoid a "manifestly absurd or unreasonable" interpretation of the aggression amendments. Even if there are legitimate policy reasons for excluding aggression from the complementarity regime |78| or for excluding unilateral humanitarian intervention from the crime entirely, that does not mean it would be absurd or unreasonable - much less "manifestly" so |79| - for the Court to reach the opposite conclusion, particularly given that absurdity or unreasonableness normally reveals itself by reference to the object and purpose of the treaty. |80| As discussed above, the object and purpose of the Rome Statute support rejecting, not accepting, interpretations of the aggression amendments that give the Understandings substantive effect.

In practice, then, the Court would be likely to rely on the Understandings only to "confirm the meaning resulting from the application of article 31." But that would mean, almost by definition, that the Understandings would have no substantive effect - after all the Understandings can have such an effect only if they change the meaning of the aggression amendments.

VII. Conclusion

Because Resolution RC/Res.6 has yet to become operative, it is impossible to determine with any certainty whether the Understandings will have any substantive effect on how the Court applies the crime of aggression. The Court may simply reject any interpretation of the Understandings that differs from what is already in the aggression amendments. If so, however, the United States will be bitterly disappointed. It believes - passionately - that the Understandings are a critical supplement to the crime of aggression. In the words of Harold Koh, the State Department's Legal Adviser, "[t]he United States considered the definition of aggression flawed, but a number of important safeguards were adopted. Understandings were adopted to make the definition more precise, to ensure that the crime will be applied only to the most egregious circumstances." |81| If there is even the possibility of the Understandings having such a substantive effect, it is critical to determine their precise legal status.

As adopted at the 2010 Review Conference, it seems clear that the Understandings would be nothing more than a supplementary means of interpretation once the aggression amendments entered into force. It is difficult to see how the Understandings can be viewed as amendments to the Rome Statute. They are not "context" under Article 31(2) of the VCLT, because they were not made in connection with the conclusion of the treaty - and even if Article 31(2) applies, they were neither made nor acquiesced in by all of the States Parties to the Rome Statute. The lack of unanimous adoption means that they are not a subsequent agreement regarding the interpretation of the treaty. And their potential impact on the rights of States Parties means that they do not qualify as an inter se agreement to modify the Rome Statute under Article 41 of the VCLT.

The current status of the Understandings, however, is not immutable. The aggression amendments will not be activated until they are approved "by the same majority of States Parties as is required for the adoption of an amendment to the Statute," which cannot occur prior to 1 January 2017. |82| If the 2017 Review Conference is committed to the Understandings - even with the possibility that the Court would give them substantive effect - it could adopt them in a manner that would make them more than simply a supplementary means of interpretation. The first option would be for all of the States Parties to adopt the Understandings, whether by vote, by consensus, or - if the Review Conference believes that Article 31(2)(b) of the VCLT applies to amendments to a treaty - by giving absent States Parties a reasonable opportunity to object to the Understandings post-adoption and treating their failure to object as acquiescence. The Understandings would then qualify as a primary means of interpreting the aggression amendments. That said, as the U.S. has itself acknowledged, even one dissenting State Party would suffice to prevent the Understandings from satisfying the requirements of Article 31.

The second option would be for the Review Conference to explicitly amend Article 21 of the Rome Statute to recognize Understandings as a new source of law. That option would be far less procedurally onerous, given that Article 121 of the Rome Statute does not require all States Parties to agree to an amendment. The Understandings would then apply either to all of the States Parties that ratify the aggression amendments (if Article 121(5) applies) or to all States Parties once 7/8 ratify them (if Article 121(4) applies).

There is, of course, a third option: maintain the status quo. If the 2017 Review Conference follows the path established by the 2010 Review Conference and simply adopts the aggression amendments non-unanimously - whether by vote or by consensus - the Understandings will remain nothing more than travaux preparatoires, for all the reasons discussed in this essay. That may well be the most desirable outcome, given the uncertain meaning of the Understandings and the Review Conference's significant opposition to the U.S. proposal to exclude unilateral humanitarian intervention from the crime of aggression. There is no guarantee, of course, that the Court would give the Understandings substantive effect. But there is also no guarantee that it would not.

[Source: By Heller, Kevin Jon, The Uncertain Legal Status of the Aggression Understandings (July 11, 2011). Journal of International Criminal Justice, Vol. 9, 2011. Available at SSRN: http://ssrn.com/abstract=1883351]

Notes

1. Senior Lecturer, Melbourne Law School. Thanks are due to Claus Kress, Marko Milanovic, Roger Clark, Philippa Webb, and Beth van Schaack for comments on earlier drafts of this essay. [Back]

2. Understandings 1-3 read as follows:

    1. It is understood that the Court may exercise jurisdiction on the basis of a Security Council referral in accordance with article 13, paragraph (b), of the Statute only with respect to crimes of aggression committed after a decision in accordance with article 15 ter, paragraph 3, is taken, and one year after the ratification or acceptance of the amendments by thirty States Parties, whichever is later.

    2. It is understood that the Court shall exercise jurisdiction over the crime of aggression on the basis of a Security Council referral in accordance with article 13, paragraph (b), of the Statute irrespective of whether the State concerned has accepted the Court's jurisdiction in this regard.

    3. It is understood that in case of article 13, paragraph (a) or (c), the Court may exercise its jurisdiction only with respect to crimes of aggression committed after a decision in accordance with article 15 bis, paragraph 3, is taken, and one year after the ratification or acceptance of the amendments by thirty States Parties, whichever is later. [Back]

3. C. Kress et al., 'Negotiating the Understandings on the Crime of Aggression',____, at 1. [Back]

4. Understandings 4 and 5 read as follows:

    4. It is understood that the amendments that address the definition of the act of aggression and the crime of aggression do so for the purpose of this Statute only. The amendments shall, in accordance with article 10 of the Rome Statute, not be interpreted as limiting or prejudicing in any way existing or developing rules of international law for purposes other than this Statute.

    5. It is understood that the amendments shall not be interpreted as creating the right or obligation to exercise domestic jurisdiction with respect to an act of aggression committed by another State. [Back]

5. Ibid., 10. [Back]

6. Understandings 6 and 7 read as follows:

    6. It is understood that aggression is the most serious and dangerous form of the illegal use of force; and that a determination whether an act of aggression has been committed requires consideration of all the circumstances of each particular case, including the gravity of the acts concerned and their consequences, in accordance with the Charter of the United Nations.

    7. It is understood that in establishing whether an act of aggression constitutes a manifest violation of the Charter of the United Nations, the three components of character, gravity and scale must be sufficient to justify a "manifest" determination. No one component can be significant enough to satisfy the manifest standard by itself. [Back]

7. Kress et al., supra note 3, at 11 [Back]

8. Ibid., 2. [Back]

9. C. Stahn, 'The "End," the "Beginning of the End" or the "End of the Beginning"? Introducing Debates and Voices on the Definition of "Aggression"', 9 Leiden J. Int'l Just. (2010) 875-882, at 879. [Back]

10. B. Van Schaack, 'The Crime of Aggression and Humanitarian Intervention on Behalf of Women', 11 Int'l Crim. L. Rev. (2011) 477-493, at 487. [Back]

11. See, e.g., J. Trahan, 'The Rome Statute's Amendment on the Crime of Aggression: Negotiations at the Kampala Review Conference', 11 Int'l Crim. L. Rev. (2011) 49-104; B. Van Schaack, 'Negotiating at the Interface of Power & Law: The Crime of Aggression', __ Colum. J. Transn'l L. (2011)__, at__; N. Blokker & C. Kress, 'A Consensus Agreement on the Crime of Aggression: Impressions from Kampala', 23 Leiden J. Int'l L. (2010) 889-895. [Back]

12. Vienna Convention on the Law of Treaties, 1155 U.N.T.S. 331 (1969). [Back]

13. See art. 15ter(2),(3) ICCSt. [Back]

14. See art. 15bis(2),(3) ICCSt. [Back]

15. See art. 13(b) ICCSt. [Back]

16. B. Van Schaack, 'In Parem Imperium Non Habet: Complementarity and the Crime of Aggression', this volume, at 1. [Back]

17. Cited in Trahan, supra note 11, at 77. [Back]

18. Van Schaack, Complementarity, supra note 16, at 13. [Back]

19. Ibid., 1. [Back]

20. As determined by reference to Article 8bis(2) ICCSt. [Back]

21. See Kress et al., supra note 3, at 15. [Back]

22. Van Schaack, Humanitarian, supra note 10, at 484. [Back]

23. See Trahan, supra note 11, at 78. [Back]

24. Emphasis added. [Back]

25. The proposed Understanding read as follows: "It is understood that, for purposes of the Statute, an act cannot be considered to be a manifest violation of the United Nations Charter unless it would be objectively evident to any State conducting itself in the matter in accordance with normal practice and in good faith, and thus an act undertaken in connection with an effort to prevent the commission of any of the crimes contained in Articles 6, 7 or 8 of the Statute would not constitute an act of aggression." [Back]

26. See Kress et al., supra note 3, at 14. [Back]

27. Van Schaack, Humanitarian, supra note 10, at 485. [Back]

28. Trahan, supra note 11, at 58 & 58 n. 43. [Back]

29. 'A More Secure World: Our Shared Responsibility', Report of the High-Level Panel on Threats, Challenges and Change. U.N.G.A. Doc. A/59/565, 2 Dec. 2004, para. 203. [Back]

30. 2005 World Summit Outcome. G.A. Resolution 60/1, 24 Oct. 2005, A/RES/60/1, para. 139. [Back]

31. The Report of the International Commission on Intervention and State Sovereignty does not completely exclude the possibility of humanitarian intervention not authorized by the Security Council, but it also does not claim that regional organizations or states are legally entitled to act unilaterally when the Security Council fails to act. The Responsibility to Protect: Report of the International Commission on Intervention and State Sovereignty (2001), paras. 6.36-6.40. [Back]

32. Art. 121(3) ICCSt. [Back]

33. "The adoption of an amendment at a meeting of the Assembly of States Parties or at a Review Conference on which consensus cannot be reached shall require a two-thirds majority of States Parties." Read literally, the 2/3 requirement only applies to amendments that cannot be adopted by consensus. [Back]

34. According to Article 9, "[t]he rules laid down in Part II" - which include Article 9 - "apply to such an agreement except insofar as the treaty may otherwise provide." [Back]

35. As an aside, an argument could be made that the elements of new Article 8bis were not properly adopted at Kampala. Although Article 121 permits the ASP to amend the Rome Statute by consensus, Article 9 provides that new Elements "shall be adopted by a two-thirds majority of the members of the Assembly of States Parties"; consensus adoption is not mentioned. It could perhaps be argued that Article 9 was satisfied as long as the Kampala consensus included more than 2/3 of the States Parties - especially as Article 9 does not textually require a vote by 2/3 of the ASP. As discussed below, however, consensus adoption of the aggression amendments meant that the ASP did not have to resolve the difficult issue of whether a sufficient number of voting-eligible States Parties were present at the Review Conference. [Back]

36. Res. RC/Res.6, Preamble, para. 3. [Back]

37. 2009 Note on the Work Programme, para. 5. [Back]

38. Kress et al., supra note 3, at 1. [Back]

39. See art. 21(1)(a) ICCSt. ("The Court shall apply... (a) In the first place, this Statute, Elements of Crimes and its Rules of Procedure and Evidence."). [Back]

40. R.S. Clark, 'Ambiguities in Articles 5(2), 121 and 123 of the Rome Statute', 41 Case W. Res. J. Int'l L. (2009) 413-427, at 426; see also Trahan, supra note 11, at 86 n. 148; D. Scheffer, 'The Complex Crime of Aggression under the Rome Statute', 23 Leiden J. Int'l L. (2010) 897-904, at 903; Van Schaack, Negotiating, supra note 11, at 18; R. Manson, 'Smoothing out the Rough Edges of the Kampala Compromise,' available at iccreviewconference.blogspot.com, at 8; G. Gordon, 'A View of the Aggression Amendments from Kampala', available at www.opiniojuris.org. But cf. D. Akande, 'What Exactly Was Agreed in Kampala on the Crime of Aggression?', available at www.ejiltalk.org. [Back]

41. Manson, supra note 40, at 8. [Back]

42. Marko Milanovic has taken the contrary position. See K. J. Heller, 'Marko Milanovic on Understandings', available at www.opiniojuris.org (arguing that "the understandings are, unlike the amendments, not law themselves"). [Back]

43. J. Harrington, 'The Aggression Negotiations at the ICC Review Conference', available at www.ejiltalk.org. [Back]

44. Quoted in K. J. Heller, 'Are the Aggression "Understandings" Valid?', available at www.opiniojuris.org. [Back]

45. Report of the International Law Commission on the Work of the Second Part of Its Seventh Session, U.N. Doc. A/6309/Rev. 1, 3-28 Jan. 1966, at 221 ("ILC Commentary"); see also M.E. Villiger, Commentary on the 1969 Vienna Convention on the Law of Treaties (Leiden: Brill Academic Publishers, 2009), 431. [Back]

46. U. Linderfalk, On the Interpretation of Treaties: The Modern International Law as Expressed in the 1969 Vienna Convention on the Law of Treaties (New York: Springer Publishing, 2007), 148; Villiger, supra note 45, at 430. [Back]

47. Van Schaack, Negotiating, supra note 11, at 9-10. [Back]

48. W. Schabas, 'Put It to a Vote !' available at iccreviewconference.blogspot.com. [Back]

49. Blokker & Kress, supra note 11, at 891. [Back]

50. Villiger, supra note 45, at 431. [Back]

51. Sir Humphrey Waldock, Sixth Report on the Law of Treaties, U.N. Doc. A/CN.4/186 & Add.1-7, 14 June 1966, at 98; Villiger, supra note 45, at 430. [Back]

52. Quoted in Heller, Understandings, supra note 44. [Back]

53. W. Schabas, 'Kampala Diary, 9/6/10', available at iccreviewconference.blogspot.com. [Back]

54. Villiger, supra note 45, at 430. [Back]

55. Ibid., 425. [Back]

56. Ibid. 429. [Back]

57. This problem, of course, is not unique to the Understandings. There is never any guarantee that an international tribunal will interpret an agreement in the manner intended by the parties that drafted it. That said, there is a significant difference between the typical Article 31(2)(b) agreement, in which absent States Parties are faced with a carefully negotiated document whose intended interpretation is clear, and the Understandings, which were negotiated quickly (particularly 6 and 7), do not have a clear interpretation, and were specifically intended by their motive force - the U.S.- to alter, not interpret, the meaning of the aggression amendments. Given the unusual drafting history of the Understandings, it would be fundamentally unfair to equate the silence of absent States Parties with acquiescence. [Back]

58. See Linderfalk, supra note 46, at 162. [Back]

59. The text of Article 31(3)(a) does not use the term "made," but the nature of the agreement is the same for both that Article and Article 31(2)(a). Villiger, supra note 45, at 431 n. 66. In any case, even if a subsequent agreement could be formed through acquiescence by some of the parties, it would still be unwarranted to find that the States Parties who were not at Kampala acquiesced to the Understandings through their silence, for all of the reasons discussed in the previous section. [Back]

60. ILC Commentary, supra note 45, at 221. As Villiger notes, "[t]his provision resembles subpara. 2(b) in that it requires active practice of some parties to the treaty," in that it distinguishes between parties that create the new interpretation and the parties that accept the new interpretation. Villiger, supra note 45, at 431. Article 31(3)(b) thus provides additional support for the idea that Article 31(3)(a) is not satisfied by a State Party "accepting" an agreement made by other States Parties. [Back]

61. Linderfalk, supra note 46, at 163. [Back]

62. J. Pauwelyn, Conflict of Norms in Public International Law: How WTO Law Relates to Other Rules of Law (Cambridge: Cambridge University Press, 2003), 258. [Back]

63. Ibid. (emphasis added); see also M. Pieth et al., The OECD Convention on Bribery: A Commentary (Cambridge: Cambridge University Press, 2007), 503 (noting, with regard to the Bribery Convention, that "[f]or all practical purposes. interpretative notes should be adopted by consensus of all the parties"). [Back]

64. "A treaty may be amended by agreement between the parties. The rules laid down in Part II apply to such an agreement except insofar as the treaty may otherwise provide." [Back]

65. "Unless the treaty otherwise provides, the amendment of multilateral treaties shall be governed by the following paragraphs." [Back]

66. Submission by the United States Regarding Interpretation of Article 17(5) of the Basel Convention, 16 April 2007, at 4 (emphasis added). [Back]

67. See, e.g., M. Milanovic, 'Aggression and Legality: Custom at Kampala', this volume, at 15. [Back]

68. A. Rigaux et al., 'Article 41', in O. Corten & P. Klein (eds), The Vienna Convention on the Law of Treaties: A Commentary (Oxford: Oxford University Press, 2011) 986-1008, at 989. [Back]

69. It is unclear how many states would be affected; the number would depend on whether there is universal jurisdiction over the crime of aggression - a debatable issue. See, e.g., S. Macedo, Universal Jurisdiction: National Courts and the Prosecution of Serious Crimes Under International Law (Philadelphia: University of Pennsylvania Press, 2006), 30. If there is universal jurisdiction over aggression, all states other than the aggressor would lose the protection of Article 17. If not, only the victim state would lose that protection. [Back]

70. See A. Aust. Modern Treaty Law and Practice (Cambridge: Cambridge University Press, 2d ed., 2007), 274. (noting that inter se agreements can overprotect, but not underprotect, states not party to the agreement). [Back]

71. See Preamble ICCSt., paras. 4 ("Affirming that the most serious crimes of concern to the international community as a whole must not go unpunished and that their effective prosecution must be ensured by taking measures at the national level and by enhancing international cooperation") & 5 ("Determined to put an end to impunity for the perpetrators of these crimes and thus to contribute to the prevention of such crimes."). [Back]

72. See Preamble ICCSt., para. 6 ("Recalling that it is the duty of every State to exercise its criminal jurisdiction over those responsible for international crimes."). [Back]

73. The latter scenario is by no means fanciful. As a number of scholars have noted, a State Party that ratifies the aggression amendments and then opts-out remains protected against acts of aggression even though its own nationals cannot be prosecuted for committing aggression. See, e.g., Milanovic, Legality, supra note 67, at 16. A rational State Party would thus want to opt-out of an aggression regime that maximizes its protection against acts of aggression - which would be a regime that did not include the Understandings. [Back]

74. See, e.g., Villiger, supra note 45, at 446. [Back]

75. ILC Commentary, supra note 45, at 218; see also Villiger, supra note 45, at 446; [Back]

76. Y. Le Bouthiller, 'Article 32', in Corten & Klein, supra note 68, at 850. [Back]

77. See Milanovic, Legality, supra note 67, at 5. [Back]

78. See, e.g., Van Schaack, Complementarity, supra note 16, at 3-7. [Back]

79. The qualifier is important, because "the exception must be strictly limited, if it is not to weaken unduly the authority of the ordinary meaning of the terms." ILC Commentary, supra note 45, at 223. [Back]

80. See Le Bouthillier, supra note 76, at 851. [Back]

81. Quoted in Heller, Understandings, supra note 44. [Back]

82. Art. 15bis(3) ICCSt. [Back]


Tienda de Libros Radio Nizkor On-Line Donations

Crime of Aggression
small logoThis document has been published on 24Jan13 by the Equipo Nizkor and Derechos Human Rights. In accordance with Title 17 U.S.C. Section 107, this material is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes.