Guest post: The Appeals Chamber’s Chastisement of PTC II for its Article 87(7) Referral Gameplaying

[Hemi Mistry is an Assistant Professor in Law at the University of Nottingham, where she researches and teaches in the area of international criminal justice. Her current research focuses upon how judicial procedure before international courts and tribunals affects how those institutions pursue their mandates. She wishes to thank Dov Jacobs for hosting this guest post.]

It’s a been a fiery time in the world of ICL: first, the decision by Pre Trial Chamber (PTC) II not to authorise an investigation into the situation in Afghanistan situation and second, this week’s bumper decision by the Appeals Chamber (AC) in Al Bashir, concerning Jordan’s non-cooperation in the execution of the arrest warrant for Omar Al Bashir and – more widely – the matter of immunities under customary international law. With the judicial division of the ICC seemingly intent upon institutional (self-)immolation, the international legal Blogosphere and Twittersphere have watched on (and fanned the flames) in horror. Amidst the flames, I wish to highlight one important aspect of Monday’s AC decision which represents a positive development in the Court’s caselaw. Specifically, this post concerns the Appeal Chamber’s reversal, by majority, of PTC II’s decision under Article 87(7) to refer Jordan’s non-cooperation in the arrest and surrender Omar Al Bashir to the UN Security Council and ICC Assembly of States Parties. Against the backdrop of the bigger ‘immunities’ question (for first of what will likely be many takes, see Jacobs and Akande) and the post-Afghanistan decision ‘crisis’, the corrective issued by the AC may at first glance appear to be a case of fiddling while Rome burns. However, as I will argue, this aspect of the AC decision represents an important attempt by the Chamber to reign in, or at least draw a line under, one of the most problematic and (excuse the pun) unruly traits in recent PTC decision-making. But first, a lot has happened, so to understand the significance of this week’s decision it is necessary to go back to basics…settle in for a story…

Article 87(7) represents one of the limited number of tools available to the ICC to respond to failures by its States Parties to comply with their obligations under the Rome Statute and to provide cooperation to the Court in the forms requested by its organs. It reads:

Where a State Party fails to comply with a request to cooperate by the Court contrary to the provision of this Statute, thereby preventing the Court from exercising its functions and powers under this Statute, the Court may make a finding to that effect and refer the matter to the Assembly of States Parties or, where the Security Council referred the matter to the Court, to the Security Council.

Article 87(7) determinations of non-cooperation and referrals to the Security Council have become a standard response to the failure by States Parties to arrest and surrender Omar Al Bashir upon his presence within their jurisdictions. The first three ‘referrals’ of non-cooperation by PTC I of Sudan in Harun and Ali Kushayb and of Kenya and Chad in Al Bashir, all issued in 2010, were not technically referrals under Article 87(7), but the exercise rather a summarily-claimed inherent power to ‘inform’ the Security Council of the accused’s presence upon the territory of a State Party, or the situation-State. However since then, PTC II, initially configured as PTC I, made three Article 87(7) referrals of Sudanese non-cooperation in violation of its obligation to cooperate with the Court established under Security Council Resolution 1593: two in Al Bashir (here, and here) and one in Hussein (here). Further to this, Chambers have issued non-cooperation decisions against eight States Parties for their failures to arrest and surrender Al Bashir upon his presence within their respective jurisdiction: Chad (2011, and 2013),  Malawi (2011), Nigeria (2013), the DRC (2014), South Africa (2015), Uganda (2016), Djibouti (2016), and Jordan (2017). In all but two instances (Nigeria (2013) and South Africa (2015)) of these instances of non-cooperation the Chamber went on to refer that non-cooperation to the Security Council and the Assembly of States Parties. Monday’s AC decision, however, reverses the PTC’s decision to refer Jordan’s non-cooperation to the Security Council.

PTC II’s caselaw on Article 87(7) is characterised by two themes: 1) Its growing frustration with the failure of states to engage with – even acknowledge – their obligations to cooperate under the Rome Statue, and 2) the PTC’s growing frustration with the Security Council and its failure to do anything to follow up on the Chamber’s referrals once they had been made. These frustrations manifested themselves in a number of ways. One way has been the gradual escalation in the language and detail with which the Chamber exhorts the responsibility of both States Parties and the Security Council to cooperate with and support the Court. Another way this frustration has manifested is the Chamber’s attempt to leverage whatever coercive tools it has within the Rome Statute toolkit to nudge States Parties towards increasing constructive engagement with the Court around the matter of cooperation, with the ultimate goal being fulfilment of cooperation requests. One such tool is the discretionary component of Article 87(7).

The early jurisprudence in Al Bashir adopted a straightforward interpretation of Article 87(7). The permissive, or discretionary, nature of the power to determine non-cooperation and refer was not acknowledged and, instead, once the Chamber had established the State Party was under an obligation to cooperate and two had failed in fact to provide that cooperation, the referral of that finding of non-cooperation was automatic (see, for example, Chad (2011, and 2013),  Malawi (2011)). In these early decisions the Chamber was presented with quite straightforward non-cooperation: the states in question simply did not engage with or acknowledge their obligations under the Statute, asserting instead the immunity of then-President Al Bashir. With non-cooperation findings and referrals stacking up and gathering dust untouched by the Security Council, in 2013 the Chamber was presented with a new situation when, for the first time, it was presented with non-cooperation by a State Party – Nigeria – which did, at least, acknowledge and accept its obligations under the Statute, which professed its commitment to fulfilling those obligations and which provided explanations for its failure on that occasion meet the Court’s cooperation request. In response, for the first time the PTC acknowledged the permissive ‘may’ in Article 87(7) and in doing so emphasised the discretionary nature of its powers under Article 87(7) (para.10). Although it appeared to implicitly acknowledge that Nigeria’s failure to arrest and surrender had been a breach of its obligations to cooperate, the Chamber did not make a formal finding to that effect. Instead, after ‘taking note’ of the explanation offered by Nigeria for its inaction, it concluded simply that ‘it is not warranted in the present circumstances to refer the matter’ (para.13).

Yet, the following year, when the DRC argued that the Chamber should not refer its non-cooperation in materially similar circumstances to Nigeria and having made similar representations as to its good faith commitment to its obligations under the Rome Statute, the Chamber refused to follow its Nigeria decision. Rather, it did two things. First, it appeared to distinguish DRC’s non-cooperation on the facts: referring in its dispositive paragraphs to the ‘deliberate’ refusal of the DRC to arrest and surrender Al Bashir, and elsewhere in the decision it described the DRC’s explanation for its non-cooperation as ‘unconvincing’ (para.13). With no objective difference between the motives of, and explanations provided by, Nigeria and the DRC respectively, it would seem that this distinction on the facts was made on the basis of a subjective assessment of the good faith credentials of the relevant parties. The exercise of discretion on this basis is not itself problematic – indeed the value of judicial discretion is in its ability to allow judges to be responsive to the circumstances of the decision at hand. However, the Chamber’s refusal to be convinced by the DRC’s explanation for its non-cooperation is itself difficult to accept given that the previous year it was ready to accept on face value that same explanation when offered by the Nigerians. That aside, the most problematic aspect of the Court’s decision in the DRC case is the second thing it did: it introduced a new and additional obligation, the obligation to consult with the Chamber in accordance with Article 97 of the Statute. On the basis of two findings – 1) the ‘deliberate’ failure to arrest Al Bashir (which was no more deliberate than Nigeria’s failure) and 2) the DRC’s failure to engage in Article 97 consultations, the Chamber referred the DRC’s non-cooperation to the Security Council, even though no mention of Article 97 was ever made in the previous Nigeria decision. It is this manoeuvre by the Court – the moving of the goalposts for the exercise of its judicial discretion – that is difficult to accept as legitimate.

Up until now, the structure of the Chamber’s analysis under Article 87(7) had been broadly the same. Notwithstanding the implicit acknowledgement of Nigeria’s factual non-cooperation, in all cases, the Chamber swept together the non-cooperation finding and the referral of that finding. However, in 2015, PTC II once again departed from its caselaw, confronted now by a state – South Africa – whose own Supreme Court had determined that it had failed to provide cooperation to the Court, contrary to its obligations under the Statute. Yet, not only did South Africa itself advance this Supreme Court ruling as evidence of its commitment to its obligations under the Statute and – more broadly to the rule of law – it had gone further, and had engaged the process of consultations under Article 97 which the Chamber had, in its DRC decision, identified as a decisive factor in the exercise of judicial discretion. Thus, the Chamber was forced to uncouple the non-cooperation finding and the referral: it made the finding of fact that South Africa had failed to cooperate in accordance with its obligations under the Statute, but nevertheless it decided not to refer that finding to the Security Council.

Having moved the goalposts once in the DRC decision, and having exhausted the (very limited) panoply of obligations under the Rome Statute that it could add to the justifications for the exercise of its discretion in South Africa and in doing so move the goalposts again, the PTC II had – it appeared – exhausted the leveraging potential of its discretion under Article 87(7). Yet, when it came to address Jordan’s failure to arrest and surrender Al Bashir in 2017 although Jordan had objectively overcome all the hurdles set out by the PTC II in Nigeria, then DRC and South Africa, the PTC refused to follow its decision in South Africa. Instead, it determined that what Jordan claimed was engagement with the consultation process in Article 97 was not, in fact, engagement with the consultation process (para.46-49). It further argued that Jordan could not have engaged in consultations for the purpose of Article 97 because the obstacle impeding cooperation that it advanced as the subject of those consultations – i.e. the question of immunities – had been addressed, resolved and removed in the course of the consultations on the same matter undertaken by South Africa (para.54-55). Now, therefore, it would seem that if Jordan wished to invoke the consultation procedure to qualify for the Chamber’s consideration of not to refer its non-cooperation, it needed to have come up with a novel obstacle to cooperation to consult over. Yet again, the goalposts were moved. This manoeuvre was one step too far and, on Monday, the AC reversed the PTC’s decision to refer Jordan’s non-cooperation. Not only did the AC conclude that the PTC made an ‘error’ in mischaracterising Jordan’s attempts to engage the Article 97 consultation process, but that the differential treatment of Jordan and South Africa constituted ‘an abuse of judicial discretion’ (para 210-211).

This aspect of the AC decision, and especially the strongly worded rebuke of the PTC’s conduct, should be welcomed. On one hand, the Chamber’s experiment with leveraging its discretion to refer non-cooperation to the Council to influence the behaviour – even if it is just the argumentative behaviour – of States Parties was partly successful. The DRC – rather than not engaging with the cooperation obligation – acknowledged, like Nigeria did, its obligation under the Rome Statute, stated its commitment to its Rome Statute obligations, but explained the practical impediments to its ability to discharge those obligations at the case at hand. Similarly, Jordan sought to engage the same consultation process under Article 97. Even though, as the Chamber itself pointedly observed, the Council has consistently failed to act upon its referrals – such that there is little ‘hard’ consequence from such a referral – States Parties, first the DRC and more recently Jordan, have taken steps to align their non-cooperation with that of Nigeria and South Africa to avoid a referral to the Security Council. Whereas PTC I has argued that the referral under Article 87(7) are meant to be ‘value neutral’ and ‘not designed to sanction or criticise the requested State’ (para.33) (in the same breath, however, it appeared to concur with PTC II in characterising the referral power as a tools ‘“to use at a certain point in time as a last resort measure or as part of a comprehensive strategy to promote cooperation (para.24)), it would seem that PTC II has had some success in persuading States Parties that a non-referral is a ‘carrot’ worth striving (or at least arguing) for.

On the other hand, having achieved that, the Chamber’s refusal to treat like cases alike and its frustration of the legitimate expectation of states, simply, reeks of bad faith. One of the fundamental tenants of the rule of law is the expectation that, under the rule of law, like cases will be treated alike. It is one if not the factor that distinguishes the politics of legality from those other – less desirable – forms of politics that legality exists in opposition to (Shklar 1964). It is through the giving of reasons that legitimate expectations are created and stability and predictability in the exercise of judicial discretion are guaranteed. In the face of the considerable political, diplomatic and – indeed – legal controversies arising out of the Al Bashir arrest warrant, it was reasonable for states to look to the PTC’s caselaw for guidance and to rely upon that caselaw. The development of – and adherence to – stable and predictable principles might do little to incentivise the kinds of changes in state behaviour (i.e. the movement towards cooperation) that the strategy of leveraging its discretion has been aimed towards. Arguably, for this reason, the experiment with decoupling of the non-cooperation determination and the referral to leverage the referral discretion was always doomed to folly. However, PTC I has also accepted the decoupling of the non-cooperation determination and the Security Council referral, having characterised non-cooperation findings as value neutral objective assessments of fact (Al Islam 2014 at para.23) and referrals as ‘one of the tools available to the Court “to use at a certain point in time as a last resort measure or as part of a comprehensive strategy to promote cooperation”’ (Al Islam 2014 at para.23). As such, the discretionary nature of the referral determination is likely here to stay and, going forward, Chambers need to develop transparent principles to guide that discretion if they are to rebuild some of the trust of States Parties. In the face of the failure of the Security Council to take political action in response to PTC, the manoeuvring of the goalposts for the Article 87(7) non-referral was clearly a proactive attempt to enhance the effectiveness of the Court. However, judicial chambers are not the Security Council or the Assembly of States Parties; they are judicial bodies, not political bodies. With the composition of PTC II now changed, this week’s decision by the AC reminds the new crop of judges that – going forward – if they too wish to engage in politics, they must play the game of legality and abide by the basic rules of that game.

You have just entered Narnia: ICC Appeals Chamber adopts the worst possible solution on immunities in the Bashir case

This morning, 6 May 2019, the Appeals Chamber issued its Judgment on the Appeal filed by Jordan against an 11 december 2017 decision by Pre-Trial Chamber II whereby it was found that Jordan failed to comply with an order to arrest and transfer Bashir to the ICC because Bashir did not benefit from Head of State immunity in the context of a UNSC Referral. As a consequence, Jordan’s non-cooperation was referred to the UNSC and the ASP.

It was a long-awaited decision, touching upon fascinating issues of public international law, treaty interpretation, customary international law, effects of UNSC resolutions, etc. The process that led to the Judgment was itself fairly novel, the Appeals Chamber having invited and received amicus briefs from the AU, the Arab League and a dozen law professors, who, in addition to their written briefs, were heard and questioned by the Chamber at length over several days in September 2018.

This is a really difficult blog post to write, because the Judgment comes at the conclusion of 8 years of debates (since the 2011 Malawi decision), with many sub-plots and twists. Explaining comprehensively why this is a terrible decision therefore would ideally require some prior knowledge of what the stakes are and would require me to fill way too many pages for a blog.

Therefore, for a comprehensive overview of the issues and explanations of why the reasoning of the Appeals Chamber is not convincing, I simply refer you to the innumerable blog posts I’ve written on the topic (particularly my initial reaction to the Malawi decision back in 2011, which is relevant here given the fact that the Appeals Chamber dug it up from nowhere) and my comprehensive chapter on the issue (an earlier draft of which you can find here).

What I will do here is simply pick and choose some particular problematic or noteworthy aspects of the Chamber’s reasoning.

First of all, at the heart of the Appeals Chamber’s reasoning is the conceptual idea that there exists a concept of “international tribunal” that is not simply the “pooling” of the exercise of jurisdiction by States (see par. 115 of the Judgment). These “international tribunals” would have a different nature and would therefore not be subject to the same rules of international law than States are, which allows the Appeals Chamber to say that while Head of State immunity continues to apply between States, it does not apply before international tribunals. However, this reasoning is very problematic. Indeed, it relies on a moral, rather than legal basis, as seen in the claim that “international tribunals” act in the name of the “international community as whole” rather than individual States. But what is the “international community” as a whole from a legal perspective? It is of course never defined in the Judgment.

Moreover, where does the “pooling” of jurisdiction end and the “international tribunal” start? For example, if France and Belgium, who separately would not be able to arrest and prosecute a foreign head of State, create an “international tribunal” through a bilateral agreement, would that new institution be able to prosecute that same Head of State? I doubt it, because States cannot grant to an IO a power they do not possess. This is basic common sense. You can’t just avail yourself of this simple rule by essentially claiming the moral high ground. Also, what arrogance to claim that you are acting in the name of the “international community as a whole”, especially when you see in practice the opposition that exists to the ICC (whether it is justified or not is another issue).

Given the reasoning adopted by the Appeals Chamber on this point, I find it somewhat ironic that the Judges go on the explain that “The law does not readily condone to be done through the back door something it forbids to be done through the front door” (par. 127). Not only is there a slight Judge Dredd feeling about this sentence (who is the “law” exactly? “I am the law!”), but it is exactly what the Judges did to remove immunity in the first place: treaty rules might be a problem? let’s use the back door of the “international tribunal” and magically go through the cupboard into the magical Narnia world when standards rules of international law disappear under the benevolent gaze of the “international community as a whole”…

The key issue should not so much be what an abstract category of “international tribunal” would look like, but rather, in each specific circumstances how a tribunal was created and what effects it might have on third States. This requires no creativity whatsoever, but simply a basic application of public international law rules. In the case of the ICC, as a treaty body, there is simply no reason why it should bind third States and that should be the end of the discussion (the PTC in the Rohinga decision attempted to justify that the ICC could have an effect on third-States as an IO with objective legal personality, but their reasoning was not very convincing either). 

Second of all, as a consequence of the Judgment, Article 98(1) is rendered mostly meaningless. Of course, Article 98(1) does not actually list what immunities need to be respected by cooperating States, but it is somewhat difficult to imagine that if the drafters of the Rome Statute really thought that Article 27 removed all immunities, even in the horizontal relationship between States, as an established rule of customary international law, it would have bothered to introduce Article 98(1) in the first place.

Third of all, the consequences for third States are quite big.

1) this Judgment means that immunities cannot be claimed by nationals of non-State parties, even when the situation is not referred to the Court by the UNSC. In other words, even if a situation is opened through a State referral or a proprio motu decision by the Prosecutor, State Parties would have to arrest and surrender nationals of non-State parties who would otherwise benefit from immunity.

2) This is probably one of the weirdest consequences of the Judgment: because the Appeals Chamber claims the existence of a rule not just in the Rome Statute, but in customary international law, that there are no immunities before “international tribunals”, one could arguably claim that the obligation to arrest and surrender a person would rest not only on State parties but also on non-State parties, because customary law is binding on all States… this is of course a ridiculous proposition, but it shows the absurdity of the Judgment.

Fourth, I note that the AC makes no mention of the Malabo Protocol, which explicitly provides for Head of State immunity. Presumably, by the AC’s standards, the tribunal constituted by the Malabo Protocol would be an “international tribunal”. Therefore, it would be acting in the name of the “international community as a whole”, in claiming that immunities exist before “international tribunals”. Why would this not equally be evidence of a contrary customary international law?

Fifth, I note that the Appeals Chamber mentions in one paragraph that Jordan’s obligation to cooperate would also stem from the Convention against Genocide. This was the position taken in a separate opinion to the South Africa decision by Judge Brichambaut (see my commentary here). I was not entirely convinced by the argument at the time, but at least the Judge made an effort to analyse the Genocide Convention specifically and explain how it would be related to the Rome Statute. Here, there is no such effort and it is impossible to understand what the relevance of the Genocide convention is in the current proceedings. Maybe it was included as a possibility in a earlier draft of the Judgment and someone forgot to remove it, because as it stands, these few lines, which seem to be added as an afterthought, with no explanations, 1 footnote and no references, are completely useless.

Sixth, I don’t have much to add that I haven’t said in the past to the analysis provided by the Appeals Chamber of the “UNSC route”. Just a few quick thoughts:

1) It’s not entirely clear why the AC bothers with this section of the Judgment at all, given that the questions raised become essentially moot given the customary law avenue taken by the Chamber.

2) I do note that the reasoning provided by the AC is interesting when it comes to determine under what sections of the Rome Statute Soudan would be obliged to cooperate with the Court. I must say that on first reading, I am convinced with the argument according to which, given the language of the UNSC Resolution, Soudan would have to respect the cooperation provisions relating to State parties rather than non-State parties. However, this does not automatically mean that Soudan would be bound by Article 27, and on this the Judgment is less convincing.

3) I also note that the AC did not follow my friendly advice: it did not actually asked those States on the UNSC who wrote the referral if in fact they did intend to remove immunities. This would have been all the more interesting as some of these States have made public Statements to the contrary (see here).

Seventh, in relation to the referral of Jordan to the ASP and the UNSC, I remain a bit lost at what the legal framework is. In such discretionary matters, it seems more of a divination exercise than one of legal reasoning, so I don’t have much to say about that. I just wonder, maybe naively, when a referral will ever be justified. Indeed, when the Jordan cooperation issue comes up, there is 7 years of constant (if not consistent) case-law asking State parties to arrest and surrender Bashir. I have all the sympathy in the world for Jordan’s legal position, which I mostly share, but practically, in those circumstances, it is difficult to see Jordan’s non-cooperation as anything else but a clear refusal to comply with a clear order from the Court. How is that not sufficient for a referral? I’m wondering, but this is just me thinking out loud, whether the real issue is to avoid that this discussion be forced on the ASP/UNSC, bodies which might make statements that the Judges would not want to hear. Indeed, it would not look good if the UNSC or the ASP (under the pressure of the AU States) actually came out in defense of Jordan…

Eighth, while I have yet to read the separate opinion (I might blog on it later), I am not sure what to make of its existence. Indeed what are we taking about?  it is a separate concurring opinion which focuses on written by 4 of the 5 judges in relation to the 2  grounds of appeal for which the decision was unanimous. So basically, did the 5th Judge not agree with the content of the separate opinion? because if she had, it would be signed by all 5 judges, concurring with themselves and then this would just be another part of the Judgment… Also, as also noted by Hemi Mistry, there are regular cross-references between the Judgment and the concurring opinion, which raises the question of the exact status of the document. I hope to have more clarity when I actually read it.

In conclusion (for now), I honestly believed that the AC would play it safe and stir away for the Customary Law route. I was wrong. As noted by Dapo Akande, this radical view justifies even more that the AU move for an ICJ advisory opinion on the matter. They have nothing to lose now.

From the perspective of the ICC, the Appeals Chamber has sadly confirmed I was right when I chose “the frog that wanted to be an ox” title, for my book chapter on immunities. This is again a case of the ICC Judges trying to be more (and to make the Court be more) than it actually is. In the fable, the frog actually exploded at the end… it’s of course just a metaphor, but given recent developments at the Court, it’s increasingly becoming a concrete risk for the institution.

 

Some extra thoughts on why the ICC Pre-Trial Chamber acted ultra vires in using the “interests of justice” to not open an investigation in Afghanistan

In my previous post on the decision by the Pre-Trial Chamber to not allow the opening of an investigation in the Afghanistan decision on the basis that it would not be in the “interests of justice”, I claimed:

First of all, and perhaps most importantly, the exercise that the Pre-Trial Chamber did is most likely ultra vires. Indeed, Article 53(1)(c) is very clear that it is the Prosecutor who can decide to not open an investigation in the “interests of justice”. It is only if the Prosecutor makes such a decision, that a Pre-Trial Chamber can review it (Article 53(3)(b)). The only job of the PTC when the Prosecutor requests the opening of an investigation is to determine jurisdiction and admissibility. And the “interests of justice” fall under neither of these categories.

Kevin Jon Heller, over at opinio juris, takes issue with my claim the the Pre-Trial Chamber acted ultra vires in discussing whether the interests of justice did in fact warrant not opening an investigation in the situation.

This is the heart of Kevin’s reasoning, after quoting Article 15(4) and 53(1):

To find that a reasonable basis exists, in short, the OTP must consider (1) jurisdiction, (2) admissibility, and — critically — (3) interests of justice. If there is no jurisdiction or no admissibility, there is no reasonable basis to open an investigation. And even if there is jurisdiction and admissibility, there is still no reasonable basis to open an investigation if the interests of justice counsel against it.

As noted, Art. 15(4) requires the PTC to agree with the OTP that there is a reasonable basis to open an investigation. If the OTP is wrong about any of the relevant considerations — jurisdiction, admissibility, or interests of justice — the requisite reasonable basis does not exist. So I fail to see how the PTC could fulfil its mandate under Art. 15(4) unless it reviews not only jurisdiction and admissibility, but also the interests of justice.

Kevin’s point seems unassailably logical: if there are 3 criteria listed in article 53(1) that the Prosecutor needs to look at in order to determine that there is a reasonable basis to proceed, then, it should follow that the Pre-Trial Chamber’s “control” over the Prosecutor must cover all 3 criteria. Which includes the interests of justice.

Of course, this argument makes sense. But I still stick to my interpretation for several reasons.

Before I explain these reasons, I’d like to indulge in a small (maybe demagogic) digression: Kevin and I rarely disagree on issues of interpretation of the Rome Statute. The few times it has happened in the past few years (like last year on the interpretation of Article 18 of the Rome Statute), it is often because the Statute is at places terribly drafted, with provisions not being very clear, and their articulation not always very obvious, therefore leading to conflicting interpretations, which can nonetheless all be very reasonable. I think this is clearly one of those cases… and as a result, the fact that I am most certainly right, does not actually mean that Kevin is wrong…

So, moving on to my explanation:

First of all, for me, the phrasing of Article 53(1) is key. As Megan Fairlie aptly pointed out to me, it is “oddly written”. It calls upon the Prosecutor to “consider” (interestingly, this seems to allow some discretion, as opposed to other terms which could have been used, such as “determine”, but that is a different discussion) whether:

(a) The information available to the Prosecutor provides a reasonable basis to believe that a crime within the jurisdiction of the Court has been or is being committed;

(b) The case is or would be admissible under article 17; and

(c) Taking into account the gravity of the crime and the interests of victims, there are nonetheless substantial reasons to believe that an investigation would not serve the interests of justice.

While (a) and (b) are phrased in positive terms (the Prosecutor considers whether the Court has jurisdiction and whether the case would be admissible), (c) is phrased negatively: that “an investigation would not serve the interests of justice” (as opposed to “an investigation would serve the interests of justice”). Kevin simply ignores this difference in phrasing, but the choice of words has to mean something!

Indeed, this clearly means for me that while jurisdiction and admissibility need to be positively established, the fact that (once determined in a first step that the Court has jurisdiction and that the situation would be admissible, as suggested by the term “nonetheless”) an investigation would be in the interests of justice is presumed, unless the Prosecutor finds “substantial reasons” to the contrary. This is turn would suggest that there is no positive obligation on the part of the Prosecutor to even consider the matter in a request, and therefore no autonomous power of the Pre-Trial Chamber to take up the issue proprio motu.

I do note in that respect that the Prosecutor does in fact devote a handful of pages to the matter in her request to open an investigation. But I would argue that her request would have been perfectly legally valid, even if she had not provided those few pages.

Second of all, it should be pointed out that, technically, if the Prosecutor were to decide to not open an investigation based on 53(1)(c), the issue would never come before the PTC as part of an Article 15(4) determination of a reasonable basis to proceed, because there would logically not be a request put to the Chamber. This is why I believe that, contrary to what Kevin says, Article 53(3)(b) is extremely relevant to the discussion. It states that:

the Pre-Trial Chamber may, on its own initiative, review a decision of the Prosecutor not to proceed if it is based solely on paragraph 1 (c) or 2 (c). In such a case, the decision of the Prosecutor shall be effective only if confirmed by the Pre-Trial Chamber.

As an aside, this is yet another “oddly written” provision because I cannot think of a scenario where a decision not to open an investigation based on paragraph 1(c), will not “solely” be based on that paragraph. Indeed, should the Prosecutor find that there is no jurisdiction and/or that a situation would not be admissible under (a) and (b), it will never move on to (c). On the other hand, if there is jurisdiction and the situation is admissible, the decision not to proceed will necessarily have to be based “solely” on (c).

But the main point here is that, clearly there is a specific procedural framework to deal with 53(1)(c) decisions, which is 53(3)(b), which flows from the actual “negative” formulation of Article 53(1)(c), and which therefore exclusively envisions the possibility for the PTC to review decisions not to proceed founded on the fact that an investigation would not be in the interests of justice. This of course makes sense, because that is how Article 53(1)(c) is actually framed.

Considering all these points, I therefore still believe that the procedural framework around “interests of justice” evaluations is constructed around the fact that it requires an actual negative assessment by the Prosecutor, which would then lead to the only procedural route available under the Statute: 53(4)(b) (because again, in such a case, there would, by definition, never be a request under Article 15).

Finally, I note that there is case law supporting my view. Indeed, as rightly pointed out by Nabil Orena on twitter the Pre-Trial Chamber that authorised the opening of an investigation in the Kenya situation stated that (par. 63):

Unlike sub-paragraphs (a) and (b), which require an affirmative finding, sub-paragraph (c) does not require the Prosecutor to establish that an investigation is actually in the interests of justice. Indeed, the Prosecutor does not have to present reasons or supporting material in this respect. Thus, the Chamber considers that a review of this requirement is unwarranted in the present decision, taking into consideration that the Prosecutor has not determined that an investigation “would not serve the interests of justice”, which would prevent him from proceeding with a request for authorization of an investigation. Instead, such a review may take place in accordance with article 53(3)(b) of the Statute if the Prosecutor decided not to proceed with such a request on the basis of this sole factor. It is only when the Prosecutor decides that an investigation would not be in the interests of justice that he or she is under the obligation to notify the Chamber of the reasons for such a decision, thereby triggering the review power of the Chamber.

Obviously it doesn’t necessarily mean I’m right, but it’s interesting to acknowledge this position nonetheless, something that the Afghanistan PTC did not even bother doing.

In sum, I think Kevin asked the wrong question as the title of his post (“Can the PTC Review the Interests of Justice?”) because my point was never that the PTC could not do so. Indeed, my point was questioning when this can be done. And in my humble understanding of the phrasing of the Rome Statute, this can logically only be done after a decision not to open an investigation on the basis of Article 53(1)(c) (because that is the actual phrasing of that provision), and therefore, as an inevitable procedural consequence, never in the context of an Article 15(4) decision because no request to be authorised to open an investigation filed by the OTP would be before a PTC in such a case.

ICC Pre-Trial Chamber rejects OTP request to open an investigation in Afghanistan: some preliminary thoughts on an ultra vires decision

Normally, I wouldn’t post on a Friday evening, but, in the interests of justice, I feel compelled to do so for once.

Indeed, earlier today, Pre-Trial Chamber II rejected the Prosecutor’s request to open an investigation in Afghanistan. This was a much awaited decision which will undoubtedly create a lot of controversy.

The timing of the decision itself is interesting. Compared to other such decisions, it took an incredible amount of time to be issues: nearly 17 months. Reading through the decision, it is difficult to understand why this took so long. Indeed, most of the decision in relation to jurisdiction and admissibility is quite uncontroversial and there does not seem to be any particular practical difficult in dealing with the various issues rapidly. This timing, mere weeks after the ramping up of US hostility towards any investigation which might relate to US nationals, will inevitably raise suspicions as to whether the Judges gave in to outside pressure. Whether this is true or not, the impression that this sends out is bound to be negative in this respect.

Moving on to the decision itself, there are a few noteworthy points I’d like to briefly raise.

1. the scope of the “situation”.

The Pre-Trial Chamber found that:

More specifically, the precise width and breadth of the Prosecutor’s power to investigate are to be determined on the basis of the scope of the Chamber’s authorisation: the Prosecutor can only investigate the incidents that are specifically mentioned in the Request and are authorised by the Chamber, as well as those comprised within the authorisation’s geographical, temporal, and contextual scope, or closely linked to it.

[…]

The filtering and restrictive function of the proceedings under article 15 further implies that the Chamber’s authorisation does not cover the situation as a whole, but rather only those events or categories of events that have been identified by the Prosecution. To conclude otherwise would be tantamount to equating the authorisation to a blank cheque, which would run against the very rationale of article 15 and thus defeat its underlying purpose. The authorisation sets the framework of the probe; investigation on incidents not closely related to those authorised would only be possible on the basis of a new request for authorisation under article 15, with a view to allowing the Chamber to conduct anew its judicial scrutiny on all relevant requirements, including jurisdiction, complementarity, gravity and interests of justice.

This approach by the Pre-Trial Chamber seems extraordinarily restrictive of the powers of the Prosecutor during a formal investigation: if she hasn’t mentioned particular incidents in her request to open an investigation, then she cannot deal with them, even if new evidence comes to light once the formal investigation has started. This is problematic for many reasons.

First, the whole point of a formal investigation is that the OTP will finally have concrete investigative powers (which include obligations for State parties to investigate), which will allow it to seriously collect solid evidence. It is not until the investigation is conducted that the Prosecutor can know what crimes can be proven or not.

Second, it transforms “situations” into a series of investigations in incidents and collapses it into “mini-cases”, which I don’t think is the nature of a situation. I do believe that the Prosecutor should have discretion to select the cases it feels it has the strongest evidence for, within the situation, in order to move forward to the next phase.

Third, this decision creates an unjustified practical distinction between an investigation in a situation following a proprio motu request from the OTP (which is subject to a PTC decision) and an investigation following a referral from a State or the UNSC (which is not subject to a PTC decision). The trigger mechanism should only affect how an investigation is opened, not how it is subsequently conducted, which should be the same for all situations. If the PTC were to be followed, this would not be the case.

2. The relevance of Article 98(2) agreements.

The PTC found that:

Furthermore, as to the Agreement of 30 September 2014 between the United States and Afghanistan pursuant to article 98, requiring the consent of a sending State to surrender a national of that State to the Court, the Chamber concurs with the Prosecution that agreements entered into pursuant to article 98(2) of the Statute do not deprive the Court of its jurisdiction over persons covered by such agreements. Quite to the contrary, article 98(2) operates precisely in cases where the Court’s jurisdiction is already established under articles 11 and 12 and provides for an exception to the obligation of States Parties to arrest and surrender individuals.

I would tend to agree with this point. For me, Article 98(2) agreements are not relevant for the determination of the existence of jurisdiction of the Court, and only become relevant later on in the proceedings, as a possible obstacle to the arrest and surrender of individuals.

I do know that there is a counter-argument out there, to the effect that these agreements are relevant to jurisdiction in the following way: even if Afghanistan is a State party, it cannot be deemed to have transferred to the ICC the capacity to exercise criminal jurisdiction over persons when it could not itself, due to that Article 98(2) agreement, exercise such criminal jurisdiction. I think this argument, while interesting, takes too rigid a view of the nature of the delegation of powers from State parties to the ICC, which is a general delegation, not a specific one in my view.

3. The territorial scope of a situation.

One interesting aspect of the request to open the investigation was that the Prosecutor brought in the question of torture of Afghan nationals at CIA “black sites” in various European countries, all State parties to the Rome Statute. The PTC does not deal with this in a very adequate way (par. 51-55), by claiming that 1) the person must be captured on the territory of Afghanistan and 2) that some of the torture must take place on the territory of Afghanistan. Not only does this seem to be a very restrictive view of the application of IHL rules, but it seems to exclude the possibility of a “multi-territorial” situation. Indeed, even putting the territorial scope of the non-international armed conflict aside, one could make an argument for crimes against humanity having been committed, with a policy implemented over several territories. I don’t see what legal obstacle there would be in doing that.

4. The “interest of justice” argument.

And now we come to the obvious main point of the decision: the fact that the PTC, despite finding that the Court has jurisdiction and that the situation would be admissible, decides to reject the OTP request based on the “interests of justice”. Several remarks on this need to be made.

First of all, and perhaps most importantly, the exercise that the Pre-Trial Chamber did is most likely ultra vires. Indeed, Article 53(1)(c) is very clear that it is the Prosecutor who can decide to not open an investigation in the “interests of justice”. It is only if the Prosecutor makes such a decision, that a Pre-Trial Chamber can review it (Article 53(3)(b)). The only job of the PTC when the Prosecutor requests the opening of an investigation is to determine jurisdiction and admissibility. And the “interests of justice” fall under neither of these categories.

This is extraordinary to say the least, and is one more step in the slow erosion of Prosecutorial discretion that the judges are slowly implementing in the past few years, as in the Comoros and the Bangladesh decisions.

Second of all, the criteria used by the PTC to determine why the opening of an investigation would not be in the “interests of justice” are equally extraordinary. I cannot quote the full 3 pages that the Judges devote to this, but here is just a taste:

subsequent changes within the relevant political landscape both in Afghanistan and in key States (both parties and non-parties to the Statute), coupled with the complexity and volatility of the political climate still surrounding the Afghan scenario, make it extremely difficult to gauge the prospects of securing meaningful cooperation from relevant authorities for the future, whether in respect of investigations or of surrender of suspects; suffice it to say that nothing in the present conjuncture gives any reason to believe such cooperation can be taken for granted.

[…]

Furthermore, the Chamber notes that, in light of the nature of the crimes and the context where they are alleged to have occurred, pursuing an investigation would inevitably require a significant amount of resources. In the foreseeable absence of additional resources for the coming years in the Court’s budget, authorising the investigation would therefore result in the Prosecution having to reallocate its financial and human resources; in light of the limited amount of such resources, this will go to the detriment of other scenarios (be it preliminary examinations, investigations or cases) which appear to have more realistic prospects to lead to trials and thus effectively foster the interests of justice, possibly compromising their chances for success.

Three conclusions can be drawn from the Court’s reasoning :

a) States should be as vocal as possible about their opposition to the Court and their absence of will to ever cooperate with it in order to make sure no investigation is opened in a situation. I usually leave the realist political analysis on the broader impact of the ICC to more competent people, but this is an absolutely terrible message to send, in the context of the opening of an investigation!

b) The PTC have transformed themselves into the financial comptrollers of the budget of the OTP and how it is spent, which is definitely not their role. The Budget of the Court, and more particularly the budget of the OTP, while of course a major issue generally, should not have any concrete impact on actual judicial proceedings. Indeed, it leads to the balancing of legal considerations, which should be the only relevant considerations inside the courtroom, with extra-legal management considerations. This decision is a very visible example of such a balancing exercise, but it has not been absent of the case law in the past on other issues (for example in relation to ordering video-link testimony in lieu of live testimony or in relation to the translation of documents in the proceedings). The Judges seem to confuse their role as managers of the legal proceedings with a role of managers of the funds of the Court.

c) This reasoning transforms the Judges at the PE phase into fortune-tellers about what cases have a more likely chance of going to trial in the future. But of course, they give us absolutely no indication of what criteria they have in mind for that! Of course, the implication of the decision is that only situations where the suspects are easy to catch and the evidence is dumped, gift-wrapped, on the Prosecutor’s lap should be investigated. But this cannot be acceptable legal criteria because, as noted before, it would reward uncooperative States and this would exclude all the difficult cases where the ICC, arguably, is expected to be present.

One final point on this. The Chamber concludes with the following thoughts on the victims:

It is worth recalling that only victims of specific cases brought before the Court could ever have the opportunity of playing a meaningful role in as participants in the relevant proceedings; in the absence of any such cases, this meaningful role will never materialise in spite of the investigation having been authorised; victims’ expectations will not go beyond little more than aspirations. This, far from honouring the victims’ wishes and aspiration that justice be done would result in creating frustration and possibly hostility vis-a-vis the Court and therefore negatively impact its very ability to pursue credibly the objectives it was created to serve.

I actually agree with that statement, in terms of creating unrealistic expectations for victims of a situation, as I’ve said countless times on this blog. However, what the Judges fail to see, is that this argument, if accepted, could arguably lead to never opening an investigation in any situation, ever. Indeed, the ICC is structurally incapable of satisfying the expectations among affected communities, in relation to the recognition of their individual suffering and even less so in terms of actual reparations. This is an extraordinary moment of lucidity by the Judges of the limitations of the ICC, which could logically have led them to drop the mic and leave the stage.

All in all, this is probably one of the most dramatic case of “sawing the branch we’re sitting on” anyone has ever seen in an international tribunal decision…

I would expect the OTP to try and appeal this decision. However (and thank you to my friend and colleague Joe Powderly for pointing it out), because the basis of the decision is arguably neither an issue of jurisdiction nor an issue of admissibility, there is no automatic right to appeal and the Prosecutor will need to seek leave to do so under Article 82(1)(d). I wonder what the odds of the Judges having issued the decision granting leave to appeal it….

[UPDATE: Kevin Jon Heller has a post up already on this question of appeal here. We share the same skepticism on this issue, but I think Kevin is wrong to say that Article 82(1)(d) does not apply because it applies “to specific cases and proceedings, not to situations”. 82(1)(d) applies to all decisions that do not fall under the other three categories, whatever the phase of the proceedings. This has been a consistent practice of the Court to date. For example, quite relevantly, the current appeal procedure in the Comoros situation was triggered following a 82(1)(d) request for leave to appeal (see here and here). This is clearly a comparable situation (no pun intended…) and I don’t see why 82(1)(d) would not apply.].

Gadafi admissibility decision: the ICC probably gets it right on amnesties, but for entirely wrong reasons

Last friday, 5 april 2019, Pre-Trial Chamber I of the International Criminal Court rejected the admissibility challenge filed by the Defence team of Saif Al-Islam Gadafi. I’ll let you read the decision itself to get a full picture of the procedural history, but, in essence, Gadafi is claiming that his case is no longer admissible because he was tried (in abstentia) and convicted for the same conduct which is the subject of the arrest warrant at the ICC. Gadafi also claims he was released on the basis of an 2015 Amnesty law (an English version of the law can be found here).

There are a number of interesting points to discuss in this decision which I’m sure other people will pick up on.

For example, I continue to disagree with the idea that the burden to prove all relevant aspect of the admissibility lies exclusively on the State or person challenging the admissibility of a case. For me, the burden should be split: the challenging party need only prove that the case is being investigated or prosecuted and it is for the Prosecutor to prove the alleged unwillingness or inability of that State.

Another example is the impact of a trial in absentia in relation to the finality of a judgment for the purposes of the ne bis in idem principle. First of all, I find that  the decision is not very clear on this point. On the one hand, the Judges seem to suggest that if the person is finally brought into custody, there is an automatic retrial: “According to the Libyan national law, once the person is arrested, his trial should start anew » (par. 48). On the other hand, the Defence seems to be saying that a retrial is only an option triggered by the Accused himself (see par. 55). Second of all, one can wonder to what extent it is good judicial policy to base the ne bis in idem analysis on hypothetical procedural developments, based on hypothetical future events. The admissibility assessment should be made based on the factual reality at the time of the decision, and the decision can be revisited should this factual reality change. If not, no case will ever be admissible because there will always be a hypothetical risk that something might happen (a revision, a retrial, etc.).

The main point I would like to comment on now is the question of amnesties. I know that academics are often criticised for a sort of egocentric impulse to quote their own work. Generally, I do not understand this criticism. We work hard on our research, on which we are judged professionally. Why should we not refer to it? Moreover, there are times, all too rare, when we develop a theoretical idea which turns out to have a real practical impact. In such cases, referring to our own work becomes even more relevant. And this decision is such a case for me.

Indeed, in 2012, I published a book chapter in a volume edited by Larissa van den Herik and Carsten Stahn (The Diversification and Fragmentation of International Criminal Law) entitled « Puzzling Over Amnesties: Defragmenting the Debate for International Criminal Tribunals » (see here for the first draft on SSRN). In that paper, I analysed the approach taken to amnesties in a number of international tribunals, more particularly international criminal tribunals and proposed two findings which I think are relevant here:

1) Contrary to what a number of commentators were saying at the time and still are today, I claimed that it was highly unlikely that an amnesty would ever be considered a valid reason to find a case inadmissible. Indeed, as a court founded on a criminal justice model, the ICC is for me quasi-ontologically incapable of considering mechanisms other than criminal law mechanisms. Whatever the political and societal benefits that one could see in some amnesties (a position I personally share), the ICC is not a court of political legitimacy (an assessment which the Judges are not equipped to make in any case). This conclusion is confirmed by the actual language of the Statute itself, where complementarity is explicitly defined in relation to « national criminal jurisdictions ». I therefore believe it would take an extraordinary set of circumstances for an amnesty to ever be accepted by the ICC. So, in this sense, I’m not surprised by the outcome of the Gadafi decision itself.

This being said, in my chapter, I had mostly considered amnesties in an isolated manner, not in combination to other procedures, notably the actual conduct of a trial. But I do not think this changes my analysis. Indeed, how is an amnesty following a conviction not literally technically a way of « shielding the person concerned from criminal responsibility »? I’ve seen commentators on twitter lament the fact that the majority confused amnesties and pardons. But I personally don’t see how the technical term used changes anything, so I’d be happy to hear more about this.

2) In my chapter, I explained that international criminal courts are not courts of legality of domestic legislation. They are tasked with determining whether laws enacted by States are in conformity with international law or not. When you apply this to amnesties, it mean that when the issue comes before them, they have to determine to what extent such amnesty is relevant procedurally, not determine whether this amnesty was adopted in contravention to accepted international norms relating to the duty to prosecute. More particularly, in the context of complementarity at the ICC, this means that the only thing that a Chamber needs to do is identify whether an amnesty is opposable to the Court in the context of an admissibility challenge. The advantage here was obvious: have international criminal judges stick to their function. Indeed, a strict finding that such amnesty is not procedurally relevant should require neither a determination that the amnesty was legally problematic neither a finding that the amnesty was politically illegitimate. As I noted in the conclusion of my chapter, it is “entirely possible to ascertain that amnesties for crimes within the jurisdiction of an international tribunal should not be recognized, while still leaving open the question of the validity in international law of that amnesty, or its political legitimacy ».

As a consequence of this, I obviously think that the majority in the Libya decision followed the wrong methodology in assessing the amnesty that Gadafi might have benefited from. They did not need to fill pages upon pages with other decisions that pronounced on the legality of similar amnesties, because that was not relevant for the admissibility assessment.

In the reasoning of the majority, one sentence struck me as problematic: « The Chamber believes that there is a strong, growing, universal tendency that grave and systematic human rights violations – which may amount to crimes against humanity by their very nature – are not subject to amnesties or pardons under international law » (par. 61).

I’ve always been irritated by Judges talking about « tendencies » when it comes to the content of the law (the same goes for other similar expressions, my favourite being that something is a « crystallising » norm). Their job is to determine the content of law as a judicial organ, not comment on a hypothetical law formation process.

You might say that I’m ignoring the very nature of customary law formation by saying this. But the judges themselves never explicitly situate themselves in such a discussion: they never even mention customary law (except in quoted excerpts from other courts and tribunals) and at no point conclude on the customary law status of the alleged prohibition of amnesties.

Instead, they use the very convenient trick of referring to Article 21(3), which asks Judges to interpret the Statute consistently with internationally recognised human rights. However, this is flawed because the Judges are not technically interpreting anything here, let alone the Statute. They are pronouncing on the international legality of a domestic amnesty law. I doubt this is the purpose of Article 21(3).

As a concluding point, I note that the majority’s developements on amnisties would appear all the more theoretical given that there appears to be some confusion about whether Gadafi in fact benefited from the amnesty law at all.

Indeed, the Defence claims he did (admissibility challenge, par. 26). However, the decision itself claims, based on what the Libyan authorities told the ICC, that Gadafi has not in fact benefited from this law because the crimes he was charged with were explicitly excluded from the law. Moreover, nothing seems to indicate in the decision itself that Gadafi benefited from the law for the crimes falling within the ICC arrest warrant. If that is the case, and more importantly, if this is what the majority believes to be true, it’s not entirely clear why they embark in the discussion on the legality of the amnesty at all.

It would not be the first time that Judges at the ICC bring in an irrelevant « sexy » issue in a decision, maybe to make themselves feel like they are contributing to the greater good of international justice rather than being perceived as boring technicians of international criminal procedure. But by refusing to accept what they are, and seeing themselves as something else, some international judges sometimes comes dangerously close to not actually doing what they were elected for, to the detriment of all Parties involved and, ultimately, to the ICC itself.