1. Introduction
On 14 May 2013, the ‘Mavi Marmara’ situation was referred to the International Criminal Court (the ‘ICC’ or ‘Court’) by the Republic of the Comoros.Footnote 1 The referral asks the Prosecutor to investigate the possible commission of crimes within the jurisdiction of the Court allegedly committed by Israeli military forces on board the vessels of the humanitarian flotilla bound for the Gaza Strip on 31 May 2010.Footnote 2 The primary basis of the Court's jurisdiction is that the alleged crimes were committed on board vessels having the nationality of states parties, i.e. the Comoros, Greece, and Cambodia.Footnote 3
Consider the following example: the Prosecutor succeeds in having an arrest warrant issued for war crimes against Mr. X, an Israeli national and member of the Israeli Defence Forces participating in the mission. The arrest warrant alleges the commission of crimes on board a vessel flying the flag of Greece. The defence files a challenge to the jurisdiction of the Court. It argues that the Court does not have jurisdiction because the signature of the Greek official on the registration certificate of the ship was forged. How would the Court proceed to consider the issue? Would the defence have a burden of proof? If so, at which point would this burden be considered discharged, thus triggering the Prosecutor's obligation to disprove?
This example raises the question of proof of jurisdictional facts before the ICC. This topic received no legislative foresight in the instruments of the Rome Statute's system. The Rules of Procedure and EvidenceFootnote 4 in particular are largely silent on the applicable legal regime of preliminary proceedings on jurisdiction.Footnote 5 Similarly, few provisions on evidence were finally agreed upon.Footnote 6
In principle, the Rules afford discretion to the judges to arrange the procedure for both challenges to jurisdiction and the examination of evidence,Footnote 7 subject to certain limitations.Footnote 8 As regards jurisdiction, the Preparatory Commission was aware of the sensitivity of the issue. Therefore, it was reluctant to adopt rules for this Part that would go beyond the ‘clarification’ of the relationship between Articles 15, 18, and 19 of the Statute.Footnote 9 For evidence, this discretionary regime is justified partly on the presumption that, since judges are professional lawyers, it was unnecessary to adopt extensive rules on evidence, used primarily to protect a jury of laymen from inadmissible or irrelevant evidence.Footnote 10 At the same time, this solution arguably facilitated the successful conclusion of the negotiations on the Rules by avoiding a decision on the contest between the common and civil law approaches to the admission of evidence.Footnote 11 However, there is no evidence in the preparatory works to suggest that this divide was influential as regards the adoption of rules on the assessment of evidence and the standard of ‘free evaluation’,Footnote 12 since formal rules on this question are generally absent in both legal systems.Footnote 13
The judges have in principle put this discretion to good use. However, as in most cases of judicial discretion, a certain degree of fluctuation on the particular point of the burden and standard of proof of jurisdictional facts seems to be emerging.
The purpose of the present article is to examine the Court's fluctuating practice on these points in the context of defence challenges to jurisdiction. Accordingly, the question of burden of proof of jurisdiction will be tackled first, followed by the standard of proof. At the end, certain concluding observations will critically reflect on the state of the law and make tentative suggestions on its future development.
2. Burden of proof
The Prosecutor has the initial burden to prove the Court's jurisdiction. The manifestation of this obligation in the courtroom depends on the triggering mechanism and the type of referral at hand. The earliest moment is the Prosecutor's request for authorization of an investigation under Article 15(3) to the relevant Pre-Trial Chamber (proprio motu action under Article 13(c)).Footnote 14 Ordinarily, in other types of referral, the judges review the Prosecutor's argument in favour of the Court's jurisdiction later, when assessing the Prosecutor's application for an arrest warrant or a summons to appear under Article 58.Footnote 15 As a result, the assessment of jurisdiction will also depend on the triggering mechanism through which the Court obtains the situation.
In proceedings relating to challenges to the Court's jurisdiction, the burden of proof shifts to the party claiming that the Court lacks jurisdiction to hear a case. In its first decision on a defence challenge to jurisdiction, the Court stated: ‘pursuant to rule 58(1) of the Rules, ‘a request or application made under article 19 shall be in writing and contain the basis for it’ (emphasis added). It is therefore the responsibility of the defence to set out the basis for its jurisdictional challenge’.Footnote 16 The Court went on to highlight that this is a manifestation of the ‘widely-accepted legal principle that the party raising a motion before a court should provide the proof upon which his/her motion is based’.Footnote 17
This position has been criticized on the grounds that it constitutes an unlawful reversal of the burden of proof.Footnote 18 From this perspective, it is argued that the defence enjoys, expressly under Article 67(1)(i), the right not to bear the burden of proof, and that jurisdictional elements form part of the elements of the crime. Therefore, by reversing the burden of proof on jurisdiction, it is contended that the defence is forced to disprove a case that the Prosecution should have the burden of proving at trial ‘beyond reasonable doubt’.Footnote 19
While technically accurate, this argumentation takes a rather isolated view of Mbarushimana's decision on the allocation of the burden of proof. To begin with, the challenging party does not bear the entire burden of proof for all the aspects of the Court's jurisdiction. It has to prove solely the specific elements of its challenge and only after the Prosecutor has at least initially satisfied the Court's jurisdictional requirements. The Pre-Trial Chamber made a clear pronouncement on that part under Rule 58(1), invoking also the generally accepted principle of actori probatio incumbit, which entails that the party making an allegation has the burden of proving it.Footnote 20 Indeed, to hold otherwise would probably encourage frivolous or ill-based jurisdictional challenges.
Moreover, it is not clear whether Article 67(1)(i) applies to jurisdiction, or whether the purported dual nature of such elements as both jurisdictional and substantive, is valid or relevant, if at all true.
As regards Article 67(1)(i), the provision explicitly prohibits imposing on the accused a reversal of the burden of proof, or onus of rebuttal. However, both the wording of the provision (‘accused’) and its positioning in the Rome Statute (‘Part 6 – The Trial’) indicate that this guarantee is specific to the question of guilt or innocence, and does not necessarily apply to pre-trial jurisdictional proceedings.Footnote 21 Thus, although the matter is far from settled, there are strong indications in the text of the treaty that Article 67(1)(i) specifically is not applicable in pre-trial jurisdictional proceedings. This is in line with human rights jurisprudence, which seems to suggest that the presumption of innocence protects mainly from presumptions of guilt, as opposed to assumptions of jurisdiction. The latter have generally been subjected to different treatment, usually under fair trial rights analysis and the requirement of trial by a ‘competent’ court ‘established by law’. Footnote 22
Finally, it is unclear whether the purported ‘dual nature’ of such elements is valid or relevant. The Appeals Chamber's decisions in the Kenyan appeals are a clear indication that this is not a critical consideration in pre-trial jurisdictional proceedings, as will be elaborated later on.Footnote 23
It seems reasonably clear therefore that the Prosecutor bears the initial burden of proving that the Court has jurisdiction, and can exercise it over a situation or a case. If the Pre-Trial Chambers are satisfied at that stage of the proceedings that the Prosecutor's submissions are well grounded, the burden then shifts to the party raising the challenge.
3. Standard of proof and jurisdiction
3.1. Standard(s) of proof at the ICC – basic definitions and classifications
In the context of the ICC Statute, the concept of standard of proof generally refers to the overall question of guilt or innocence of the accused.Footnote 24 It seeks to answer the question whether there is enough evidence to convince the Court that a specific individual committed a certain crime. The standards are progressive. ‘Reasonable grounds to believe’ is the required threshold for issuing an arrest warrant or summons to appear,Footnote 25 ‘substantial grounds to believe’ is required in order to confirm charges,Footnote 26 and ‘proof beyond reasonable doubt’ for a finding of guilt.Footnote 27
From a narrower perspective, the standard of proof may relate to proof of a specific fact relating to the objective or subjective element of the crime.Footnote 28 This is the more technical subject regulated by the law of evidence. Here, the crucial question is to decide whether a fact can be considered as ‘proven’ or ‘not proven’ on the basis of reliable evidence, such as documents or witness statements. The Rome Statute system provides for a different process of analysis of such questions. In short, the existence of a fact shall be proven by the collection and production of evidence that is relevant, admissible, and has probative value.Footnote 29 If the admissibility or value of a piece of evidence is challenged, the burden rests with the party introducing the evidence.Footnote 30 The standard differs depending on the type of facts and the party challenging its existence.Footnote 31 Exceptionally, it is also possible that a certain fact will not need to be proven, such as facts agreed upon by the Prosecutor and the defence under certain conditions,Footnote 32 or facts of common knowledge.Footnote 33
The existence or application of rules on the standard of proof at the stage of jurisdiction is unclear. Therefore, the judges have addressed this question mostly through jurisprudential constructions, as the next sections will demonstrate.
3.2. Between prima facie proof and avoiding the question of a standard of proof in jurisdictional matters – Mbarushimana
The Mbarushimana case is the first case involving a challenge to the jurisdiction of the ICC. In 2010, Pre-Trial Chamber I issued a warrant for the arrest of Mr. Mbarushimana.Footnote 34 The suspect was charged with participation in war crimes and crimes against humanity, allegedly committed by the FDLR (Forces Démocratiques de Libération du Rwanda) in the Kivus, in his capacity as the spokesperson of that group.Footnote 35 Jurisdictionally speaking, the problem in Mbarushimana was the time lapse between the referral (2004) and the time of the alleged commission of the crimes (2009). The Court needed to prove the connection between the case and the temporal limits of the situation.Footnote 36
Instead of resolving the matter fully when it issued the arrest warrant, the Pre-Trial Chamber adopted a prima facie standard of proof of jurisdiction. It briefly explained that there is a connection between the alleged crimes and the DRC situation on the basis of Security Council resolutions and Presidential statements on the DRC situation at the time.Footnote 37 It concluded that ‘the Chamber is therefore satisfied, on a prima facie basis, that the case against Mr. Mbarushimana falls within the context of the DRC situation of crisis …’.Footnote 38
Almost nine months later, the defence filed a challenge to the Court's jurisdiction, primarily contesting the Pre-Trial Chamber's earlier ruling on the connection of the case to the situation.Footnote 39 The Pre-Trial Chamber dismissed the challenge.Footnote 40 As a preliminary matter, the Chamber explained that the defence had the burden of proving ‘the basis for its jurisdictional challenge’,Footnote 41 but avoided stipulating the requisite standard, which the defence would need to satisfy in order to successfully discharge its burden. The Chamber stated that:
the present challenge is based, and its determination will thus depend, much more on issues of a legal nature than on whether a given fact can or cannot be considered by the Chamber as properly established. There is thus no need for the Chamber to address the issue as to the nature of the “standard of proof” to be satisfied by the party bringing the jurisdictional challenge.Footnote 42
This unhappy formulation leaves the Chamber's decision vulnerable to criticism.
Starting from the obvious, the Chamber candidly admitted that the challenge is ‘much more’ legal than factual.Footnote 43 ‘Much more’, however, does not mean ‘exclusively’. To be precise, the defence argued extensively that the Prosecutor failed to produce enough evidence to prove that the FDLR committed atrocities that contributed to the ‘situation of crisis’ in the DRC prior to the date of the DRC referral (3 March 2004).Footnote 44 In support of its claims, the defence invoked Security Council Resolutions and Presidential Statements, as well as a report by Human Rights Watch.Footnote 45 It further asked for a clear finding on the applicable standard of proof for the Prosecution and the defence.Footnote 46
For its part, the Chamber decided that there was sufficient evidence to prove FDLR involvement – again – due to certain Security Council resolutions and Presidential Statements.Footnote 47 The Chamber took a narrow view to the challenge. It restricted itself to responding to the defence submissions. However, the ‘much more’ construction is particularly problematic, since the Chamber also tackled questions of fact. This is evidenced in the Chamber's elaborate effort to establish the existence of FDLR action in the Kivus by recourse to UN documents.Footnote 48 Beyond the more limited confines of the case, however, this approach begs the following question: is the production of Security Council resolutions and Presidential Statements sufficient to prove or disprove facts in proceedings relating to jurisdiction? Mbarushimana does not explain conclusively the degree to which defence will have to prove the factual basis of its challenge on jurisdiction.
3.3. Making law: the ‘degree of certainty’ standard
Pre-Trial Chamber II, on the other hand, stipulated a standard of proof on jurisdiction – although without any significant jurisprudential basis and, in the beginning at least, as an obiter dictum.
In 2009, the Chamber issued its confirmation of charges decision in the Bemba case.Footnote 49 The defence had hinted that it considered challenging jurisdiction,Footnote 50 although in the end it never did so. However, this did not stop the Chamber from stipulating a standard of proof as regards jurisdiction. The Chamber first noted that Article 19(1) of the Statute provides that ‘[t]he Court shall satisfy itself that it has jurisdiction in any case brought before it’. From there, it proceeded to make the following determination:
The Chamber considers that the phrase “satisfy itself that it has jurisdiction” also ‘implies’ that the Court must ‘attain the degree of certainty’ that the jurisdictional parameters set out in the Statute have been met. Thus, the Chamber's determination as to whether it has jurisdiction over the case against Mr. Jean-Pierre Bemba is certainly a prerequisite for the issuance of the present decision under article 61(7)(a) and (b) of the Statute.Footnote 51
The Chamber made no references to any authority in this obiter dictum.Footnote 52 Nonetheless, it proceeded to apply this standard in subsequent cases, where questions of jurisdiction actually emerged.
The situation in Kenya was the first to be initiated by the Prosecutor proprio motu. In deciding on the Prosecutor's request for authorization of an investigation under Article 15, Pre-Trial Chamber II debated for the first time whether the crimes in question fell within the Court's subject-matter jurisdiction.Footnote 53 In particular, the judges were divided on the meaning of ‘organizational policy’ and ‘organization’ under Article 7(1)(a) of the Rome Statute. Judge Kaul considered that these elements were absent and, consequently, that the alleged crimes did not amount to crimes against humanity.Footnote 54 That said, the judges avoided any reference to the standard at that early stage of the process.Footnote 55
The Chamber reverted to the ‘degree of certainty’ standard in its pre-trial decisions on the summons to appearFootnote 56 and the confirmation of chargesFootnote 57 as established case law, without providing reasons concerning its origins or its effects.
The defence challenges to jurisdiction were discussed at the confirmation hearing. They effectively reiterated Judge Kaul's arguments.Footnote 58 The majority of the Pre-Trial Chamber rejected the objections, on the argument that they raised questions of the merits and not jurisdictional ones.Footnote 59
In his dissenting opinion, Judge Kaul dedicated a few lines to the ‘degree of certainty’ standard.Footnote 60 The judge explained that the ‘degree of certainty’ standard means that the Statute's jurisdictional parameters must be proven to have been met.Footnote 61 Accordingly, he drew two conclusions from this standard:
Firstly, the answer to the question of whether the Court has such jurisdiction is, in principle, not subject to the progressively higher evidentiary thresholds which apply at the different stages of the proceedings. Secondly, an affirmative answer to that question is a pre-condition to the Court's consideration of the merits. Consequently, the question cannot be deferred to the merits but must be ruled upon definitively ab initio. In other words, the Court does not have limited jurisdiction when issuing a warrant of arrest or a summons to appear; slightly more jurisdiction at the confirmation of charges stage; and jurisdiction “beyond reasonable doubt” at trial, after the merits have been fully adjudged. The Court either has jurisdiction or does not.Footnote 62
This is to date the only standard of proof of jurisdiction available at the ICC and Judge Kaul's is the only explanation. As such, it calls for certain critical observations.
First of all, the decision suggests that the question of jurisdiction is sufficiently mature for a final determination regarding the ‘degree of certainty’ threshold at that early stage of the proceedings, i.e. the issuance of the arrest warrant or the confirmation of charges. However, it would seem that this position is open to criticism on at least three separate grounds.
To begin with, it is not supported by the practice of the Assembly of States Parties (‘ASP’). As a matter of principle, the view that jurisdiction and admissibility should be addressed and finally decided upon as early as possible was presented by the French proposal in the Preparatory Commission tasked with the promulgation of the Court's Rules and was ultimately defeated.Footnote 63 While this does not have any effect on the time limits imposed under the Statute, it serves as an indication that the precise timing for assessing jurisdiction may be subject to uncertainty.
Moreover, the rules seem to require a certain maturity of the judicial process for the final determination of these questions. The Prosecutor's collection of evidence may well continue until the confirmation of charges and exceptionally even thereafter.Footnote 64 Jurisdiction, much in the same way as guilt or innocence, depends also on the evidence produced by the Prosecutor. A final decision on jurisdiction made at the pre-trial stage would likely be premature, if the picture is not yet complete. It would probably make more sense to reserve the final determination for a procedural moment after the evidence gathering process has been completed. In principle, this moment is after the conclusion of the confirmation of charges hearing, barring exceptional circumstances.Footnote 65 Judge Kaul himself seems to have been aware of this difficulty and therefore insisted on the early completion of the Prosecutor's investigation.Footnote 66
Additionally, this approach is further corroborated by Article 19(4). It provides that challenges to jurisdiction may be heard by the Trial Chamber and, in any event, until ‘the commencement of the trial’.Footnote 67 This suggests, at least in theory, that the question of jurisdiction is open to debate up until and including the trial.Footnote 68 Certainly, this solution may seem inefficient. However, a final disposition of jurisdiction at an early stage in the process would likely render the entire system of Article 19 largely ineffective – particularly insofar as the right of the accused to challenge jurisdiction is concerned.Footnote 69 Jurisdiction has to remain an open question, while the defence retains the right to challenge it under the Statute.
Second, the Chamber was right to reject the ‘gradation of standards’ approach, since these standards exist to answer a different question, not the issue of jurisdiction.Footnote 70
Third, and conclusively, the Pre-Trial Chamber created a separate standard for jurisdiction. The ‘degree of certainty’ seems to require a definitive affirmative answer to the question of jurisdiction. As Judge Kaul explained, the Statute's jurisdictional parameters must be proven to have been met: ‘the Court either has jurisdiction or does not’.Footnote 71 This standard appears to resemble a threshold similar to the ‘beyond reasonable doubt’ standard for guilt. The reason is that the latter requires that ‘the existence of a certain fact is the only reasonable conclusion’, which entails that the Prosecutor disproves any other reasonable conclusion and eliminates any reasonable doubt;Footnote 72 it denotes ‘certainty as to the commission of the crime’.Footnote 73 In the same spirit, the ‘degree of certainty’ for jurisdiction suggests that the Court's jurisdiction exists, if that is the only possibility, to the elimination of all others. Therefore, the two standards appear to impose similar thresholds.
One of the main problems with the ‘degree of certainty’ appears to be that it constitutes an incomplete transposition or interpretation of the degree of certainty standard adopted by the International Court of Justice. The Pre-Trial Chamber, in the paragraph just before the stipulation of the ‘degree of certainty’ explicitly referred to the World Court's Nicaragua ruling.Footnote 74 This gives reason to believe that the Chamber was strongly influenced by the Nicaragua standard, when adopting its own ‘degree of certainty’. Interestingly, however, the ICJ formulates this standard in a somewhat different manner than Pre-Trial Chamber II:
The use of the term “satisfy itself” … implies that the Court must attain the same degree of certainty as in any other case where the claim of the party … is sound in law, and … the facts on which it is based are supported by convincing evidence.Footnote 75
Evidently, Pre-Trial Chamber in its own definition has omitted certain crucial aspects of this test, such as that the proof of facts must be supported by ‘convincing evidence’.
It would appear therefore that the matter requires further elaboration with convincing legal reasoning premised on both the system of the Rome Statute and general international law. Presently, beyond the Chamber's undoubted authority to interpret the Statute and the Court's jurisdiction, the ‘degree of certainty’ standard remains, at its core, an obiter dictum interpretation of Article 19(1).
It is perhaps this lack of legal reasoning that has not permitted the general adoption of the ‘degree of certainty’ standard by other chambers of the Court. Pre-Trial Chamber I did not adopt it in its October 2011 decision in Mbarushimana. Moreover, as we will see in the next section, the Appeals Chamber has also not endorsed it in its recent decisions on jurisdiction.
3.4. Is a standard of proof even necessary in jurisdictional proceedings? The decisions of the Appeals Chamber in the Kenyan cases
The Appeals Chamber's recent jurisprudence on jurisdiction seems to indicate that the questions concerning the standard and burden of proof in proceedings relating to jurisdiction are, in reality, moot.
On 24 May 2012, the Chamber issued its substantively identical decisions in the appeals raised by the defendants in the Kenyan cases.Footnote 76 The subject of the appeal was the defendants’ contention that the Pre-Trial Chamber erred in its dismissal of their challenge to the Court's subject-matter jurisdiction.Footnote 77 The defendants espoused Judge Kaul's dissenting opinion in the confirmation of charges decisions.Footnote 78 The crux of the argument was that the Prosecutor had failed to prove the existence of an organizational policy, which is one of the contextual elements of crimes against humanity under Article 7(2)(a) of the Rome Statute.Footnote 79 Thus, the defence argued on two levels; first, that the interpretation of ‘organizational policy’ adopted by the Pre-Trial Chamber was wrong, whereas Judge Kaul's definition was correct; and second that, irrespective of the preferred interpretation, its existence was in any event not borne out by the facts of the case as presented by the Prosecutor.
The Appeals Chamber rejected the appeals. Essentially, it agreed with the Pre-Trial Chamber. It held that the issues raised by the defence are not issues of subject-matter jurisdiction and therefore should be dealt with as part of the merits of the confirmation of charges decision.Footnote 80 However, since leave to appeal was not granted for that part of the confirmation decision, the appeal was rejected as inadmissible.Footnote 81
The Appeals Chamber formulated the question before it as follows: ‘whether, in the context of this case, the interpretation and existence of an “organizational policy” are matters relating to subject-matter jurisdiction and are therefore appropriately before the Appeals Chamber pursuant to articles 19(6) and 82(1)(a) of the Statute’.Footnote 82
The Chamber reasoned in the first place that this is a question that pertains to the merits of the case and the decision whether or not to confirm charges.Footnote 83 The defendants’ arguments could be made as part of the case under Articles 61(6) and (7) of the Statute, as indeed they were.
The Chamber then stated that, if one were to accept that these are jurisdictional questions, the procedural structure of the pre-trial process would be confounded. The Chamber explained that under Rule 58(2), questions of jurisdiction need to be addressed at the confirmation hearing prior to the merits. Therefore, if one accepted the argument of the defence, there would be a duplication and conflation of the process. ‘Duplication’ would take place, because the same issue that belongs properly to the merits of the confirmation stage would need to be tackled at the jurisdictional stage as well. ‘Conflation’ on the other hand would occur because, under Rule 58(2) the question would be treated first as a jurisdictional matter, before the merits are actually heard. This would fuse the otherwise separate processes of jurisdiction and confirmation, while it is the objective of only the latter to ‘consider the matters raised in these appeals and filter unmeritorious cases’.Footnote 84
The Appeals Chamber then turned to the ‘dual nature’ argument of Judge Kaul for the contextual elements of crimes; namely that such elements are both substantive and jurisdictional, at least since in their absence, the crimes do not fall within the Court's jurisdiction.
The Appeals Chamber did not answer this point. It simply stated that these arguments ‘do not affect’ the Chamber's conclusion. The reason is simple: ‘[a]s the Prosecutor has expressly alleged crimes against humanity, including the existence of an organizational policy, the Appeals Chamber finds that the Court has subject-matter jurisdiction over the crimes with which Mr Muthaura and Mr Kenyatta have been charged’.Footnote 85 Proving the existence of such charges depends on the facts of the case and belongs to the merits of the case. The Chamber explained that, if no organizational policy is proven to exist, that means that crimes against humanity have not been committed, rather than that the Court does not have jurisdiction.Footnote 86
The Chamber considered the jurisprudence of other courts and tribunals on this issue. While it was mindful of the important structural differences separating the ICC and the ICTY/R systems, the Appeals Chamber approvingly noted that ‘the general approach taken in the ICTY and ICTR jurisprudence has been that factual and evidentiary issues are to be considered as trial, not as part of pre-trial jurisdictional challenges’.Footnote 87 As an example, the Chamber mentions that, while the existence in customary law of a mode of liability is a jurisdictional issue, its ‘contours’ or proof in a specific case is a question for the merits.Footnote 88 Thus, the Chamber concluded that the issue raised was not a jurisdictional one, and the appeal therefore was rejected. The Court's reasoning may be criticized from a number of perspectives.
First of all, the Chamber starts by drawing a conclusion, namely that these questions belong to the merits. It does not explain convincingly why this is the case; there is certainly no list of issues that belong exclusively to jurisdiction or to the merits of the case in Articles 61(6) or 19 of the Rome Statute. It is possible that such issues may well be raised under both provisions; this is the gist of the problem. Starting from a premise that is actually the conclusion, the Court goes on to underline that, if one accepts the appellants’ contentions and addresses such questions as jurisdictional, then they will be treated as a matter of priority in the process. This would consequently lead to duplication, because the same issues will need to be considered later on as part of the confirmation of charges as well. In this reasoning, the foundation is that these questions belong to the merits of the confirmation. However, if one considered that they also belong to jurisdiction, then the problem would be not with the jurisdictional stage, but rather with the merits of the confirmation hearing. If the issue of organizational policy was dealt with first, why would a Pre-Trial Chamber need to revisit it at confirmation and duplicate, fuse, or obfuscate the process? If anything, its earlier treatment would likely speed up the process and – if the challenge was meritorious – save the Court's scant resources for crimes that actually belong in its dockets.
The Chamber's true motivations behind this decision are therefore difficult to decipher. One could hazard some guesses. The Chamber may have been concerned with invoking two different standards within the same procedure to prove the same fact. On the one hand, the higher threshold of ‘degree of certainty’ when the issue is considered as a (priority) question of jurisdictionFootnote 89 and, on the other hand, the lower threshold of ‘substantial grounds’ when the same issue is considered later within the merits of the confirmation of charges under Article 61 of the Rome Statute. As an example, if Judge Kaul's ‘dual nature’ argument would be applied, the existence of the ‘organizational policy’ would need to be proven with a degree of certainty for the purposes of jurisdiction, but only to the standard of substantial grounds for the merits of the confirmation. Arguably, this would defeat the purpose of Article 61 which limits the threshold to ‘substantial grounds’ in the pre-trial phase. In addition, the Chamber may have likely intended to avoid the conflation of a higher judge-made standardFootnote 90 with a lower standard imposed by the Statute. Moreover, the Appeals Chamber may have been keen to avoid delays due to the transformation of the pre-trial process to a full trial. In Mbarushimana, for example, the defence explicitly invoked its intended jurisdictional challenge as a means to acquire documents and other evidence that it deemed necessary for its substantiation.Footnote 91 This led to a ‘trial-like’ process that lasted for more than half a year.Footnote 92
Furthermore, the Appeals Chamber may have well been concerned about the use and abuse of Article 82(1)(a). This appears to be the most convincing reason. As Friman has suggested, the Court's restrictive approach to requests for leave to appeal pre-trial issues has led the defence to rely on an expansive interpretation of the notion of ‘jurisdiction’, in order to have access to the Appeals Chamber without prior leave.Footnote 93 The Appeals Chamber ruling appears to indicate that questions such as assessment of evidence proving a fact, standards of assessment or any sort of factual question are not jurisdictional in nature. Therefore, an appeal on such issues would not fall within the scope of Article 82(1)(a).Footnote 94
Finally, this approach may also be justified by human rights considerations. Since the threshold for determining culpability is lower at the confirmation hearing, the corresponding rules on evidence differ between the pre-trial and trial stages.Footnote 95 Evidentiary findings at the pre-trial stage cannot prejudice the right of the defence to full examination of the evidence at the trial stageFootnote 96 or substitute the final determination on evidence reserved for the Trial Chamber.Footnote 97 Therefore, such issues should be ultimately reserved for trial. Finally, the Prosecutor may continue collecting evidence during the pre-trial stage and even after the confirmation of charges hearing;Footnote 98 the list of evidence may be complete only at the trial stage.
There are therefore good reasons justifying the Appeals Chamber's restrictive view. Questions of jurisdiction and their interconnection with the merits are not easy to answer. The lines between jurisdiction and merits tend to get blurred – and intelligent lawyers may wreck procedural havoc absent judicial vigilance.
However, it is one matter to steer clear of such complicated situations; it is quite another to decide in a wholesale fashion that facts and their proof do not belong to the stage of jurisdiction. After all, one cannot fully divorce facts from jurisdiction; some facts will always be pertinent. Therefore, in order to avoid turning this ruling into something other or more than what it seems to be, one could read caveats in its future application.
For one, the Appeals Chamber decision may be read restrictively. As such, it may be relevant only for questions of subject-matter jurisdiction, since the Court did not deal in its decision with any other type of challenge. The language used by the Appeals Chamber in its ruling is precise enough to allow challenges to jurisdiction ratione temporis, loci or personae to escape its reach. Ultimately, this argument could also be the result of the latter three jurisdictional facets’ more preliminary or objective nature, which would arguably pose less of a duplication risk. Questions of subject-matter jurisdiction, on the other hand, appear to be more closely intertwined with the merits of the criminal case and therefore present a higher risk of conflation.
For another, it may be just a cautionary approach to the right of pre-trial access to the Appeals Chamber, a restrictive interpretation of Article 82(1)–(a). The Appeals Chamber is keen to avoid delays in the process by the submission of manifestly inadmissible challenges to the Court's jurisdiction. Procedural efficiency and effectiveness are critical. The ruling does not seem to prevent Pre-Trial Chambers from making reasoned use of their discretion under Rule 58 if the need arises.
Ultimately, the Court never said that these questions should not be litigated. The only issue is the procedural stage where such litigation should take place. The Appeals Chamber has authoritatively pointed to the direction of the merits of the confirmation process – and the Trial Chamber. Judge Kaul has raised some valid points in that regard. The Appeals Chamber did not say that the judge was wrong when he wrote on the dual nature of the contextual elements. The Appeals Chamber was equally silent on the valid point made by the Judge on the effective interpretation of Articles 19(1)–(3) and 82(1)(a).Footnote 99
In particular, the Judge argued that, if one accepted the Prosecutor's contention that the Court has jurisdiction simply because a suspect is charged with a crime, it would mean that ‘the Prosecutor could label any crime as a crime within the jurisdiction of the Court thus removing the subject-matter jurisdiction (ratione materiae) from the scope of article 19(1)[…]’.Footnote 100 This would effectively divest the parties of the right to challenge the jurisdiction of the Court on this point as jurisdiction would be finally fixed on the basis of the charges presented by the Prosecutor.Footnote 101 In this light, it is probably fair to say that the Appeals Chamber was more concerned with two issues: first, avoiding the need to prove the existence of the same fact against two different standards within one procedure, and second, closing the door under Article 82(1)(a).
That said, the Appeals Chamber's decision makes two points clear. First, the classification of the crime by the Prosecutor in the charges is decisive as regards subject-matter jurisdiction. If an individual is charged with a crime falling within the jurisdiction of the Court ratione materiae, the Court cannot depart from the Prosecutor's classification at least in that regard. This seems to be the case with almost all crimes, save very few exceptions, where recourse to customary law would be required, such as Article 7(1)(k).Footnote 102
Second, as a matter of fact and evidence, the Appeals Chamber has indicated that ‘factual and evidentiary issues are to be considered at trial, not as part of pre-trial jurisdictional challenges’.Footnote 103 With this statement, the Appeals Chamber seems to respond to Judge Kaul, who convincingly explained that some facts always form part of jurisdictional challenges, and will need some proof at the pre-trial stage. The Chamber's reply is that, in tandem with international criminal practice, questions of fact do not belong to the jurisdictional stage of the ICC system. This finding, coupled with the decisive value attached to the Prosecutor's charges, means that, at least insofar as subject-matter jurisdiction is concerned, questions of burden and standard of proof are effectively without object for the vast majority of crimes within the Court's jurisdiction.
3.5. From admissibility to jurisdiction – time for reverse engineering?
One last point relates to the possibility of applying by analogy the relevant standard of proof applicable in admissibility proceedings to jurisdiction. Both processes share certain common attributes in terms of procedural treatment,Footnote 104 and the Court has already applied admissibility case law in proceedings concerning challenges to jurisdiction.Footnote 105
Admissibility involves an assessment of complementarity and gravity.Footnote 106 In light of the principle of complementarity, the main question for the judges in admissibility proceedings is whether the challenging party could prove that the suspect before the ICC was or is the subject of a national investigation or prosecution, which ‘cover[s] the same individual and substantially the same conduct as alleged in the proceedings before the Court’.Footnote 107 As regards the burden of proof, the Court has ruled that the party challenging admissibility has the burden of proving its contentions.Footnote 108
For defence challenges to admissibility, the standard of proof is the ‘civil standard’, namely ‘balance of probabilities’.Footnote 109 Trial Chamber III has specified however that, while the defence bears such burden, this does not imply that, once it is discharged, the Court will have no option but to declare the case inadmissible. It has emphasized that an assessment on admissibility will be the result of weighing the competing arguments. As such, while the defence bears a burden of proving on a balance of probabilities its own contentions, this is only one part of the equation. The Court's function is to consider also the merits of the other competing submissions, which seems to suggest correspondingly that the Prosecution also bears a burden of proving its own submissions on admissibility.Footnote 110
Would it be possible to transplant these approaches from admissibility to jurisdiction? From the perspective of the burden of proof, it would seem that for both jurisdiction and admissibility, actori probatio incumbit applies. As regards the standard of proof, however, the analogical application of the ‘balance of probabilities’ to questions of jurisdiction is more challenging. The critical issue is whether the underlying reasoning for the adoption of this standard in admissibility is equally applicable to jurisdiction. In this context, the perceived points of convergence and difference between the two processes, and particularly the function or nature of the proceedings compared, are important.
The first stop is the soundness of the Bemba reasoning on admissibility and its correlation to jurisdiction. The Bemba Trial Chamber was confronted with three alternative proposals as regards the standard of proof in admissibility. The defence argued for the balance of probabilities,Footnote 111 the Prosecution for clear and convincing evidence,Footnote 112 whereas the OPCV (victims) argued for proof through prima facie evidence.Footnote 113 In the end, The Trial Chamber adopted the balance of probabilities standard. It admitted at the outset that the Statute and the Rules provide no guidance on the matter. Then it reasoned that, ‘the overwhelming preponderance of national and international legal systems apply … the “civil standard” of proof (a balance of probabilities) when the burden lies upon the defence in criminal proceedings’.Footnote 114 Moreover, the Chamber considered that the Prosecution's proposed standard was not sufficiently established in law and concluded that, since the proceedings at hand are criminal,
it would be unjust to impose a variable standard of proof on the accused, depending on the seriousness of the application that he is making. His ability to defend himself should not be made more difficult simply because he is challenging the prosecution's right to continue with the prosecution at this Court.Footnote 115
The Chamber said nothing on the standard proposed by the victims. The reasoning of the Trial Chamber is admittedly not fully developed and, at times, unclear.
For one thing, it starts from a general observation on the standard of proof in criminal proceedings as regards the accused. The Chamber stated that, when the burden rests with the defence in criminal proceedings, the applicable standard of proof in most national and international criminal justice systems is balance of probabilities.Footnote 116 With respect, however, the Chamber offered absolutely no evidence to justify its statement. There is no reference to national practice in the decision, let alone any reasoning justifying how such national proceedings bear any resemblance to ICC admissibility processes.Footnote 117 The Chamber seems to take liberties here, that the Appeals Chamber has strenuously refused to the parties in the past, insofar as the sufficiency of state practice is concerned as evidence of the existence of a rule or general principle.Footnote 118
Correspondingly, from a normative perspective, the Chamber's analysis does not seem to consider the structure of Article 21 of the Rome Statute. To avoid instances where interpretation becomes legislation, Article 21 provides for a hierarchy of norms.Footnote 119 In the present case, the Chamber's argument would be stronger if it explained that it has recourse to the ‘civil standard’ in admissibility proceedings because it constitutes a rule of customary law under Article 21(1)(b) or a general principle of law under Article 21(1)(c) of the Rome Statute. Again, however, there is no reference to state practice or opinio juris on point.
Beyond these considerations, however, the reasoning is confusing. The Chamber imposed the burden on the defence and adopted the ‘balance of probabilities’ standard, largely because in its view, it would be ‘unjust’ to impose on the accused a standard of proof different and more onerous compared to that relating to the accused's ability to defend himself. However, the standard relating to the accused's ability to defend him/herself is, first of all, variable throughout the ICC process, and secondly, at trial level, proof ‘beyond reasonable doubt’. In order to avoid a finding of guilt, the accused has to simply create a ‘reasonable doubt’ at trial. By establishing, however the ‘balance of probabilities’ on admissibility and imposing the burden on the accused, the Trial Chamber imposes a burden higher than that applicable to the accused's ‘ability to defend’ him/herself compared to ‘reasonable doubt’. In any event, if one does follow the standards of the pre-trial process in the ICC system, then the Chamber should have explained the relationship between ‘balance of probabilities’, ‘reasonable grounds to believe’, and ‘substantial grounds to believe’. The Chamber seems to circumvent these all but subtle differences. However, even if one was convinced that the Chamber adopted the correct standard, without perhaps fully convincing reasoning, its analogical application to jurisdiction seems difficult. The main reasons are the language of Article 19(1) and the nature of the two sets of proceedings.
Article 19(1) makes it clear that the Court has the duty to decide on its jurisdiction, whereas findings of admissibility are discretionary and would ordinarily be reviewable upon a party's application.Footnote 120 The Court's obligation in this context is not only one of procedural initiative, but has substantive content as well; the Court's obligation to decide whether or not, on the basis of the case before it, it has jurisdiction. This seems to impose a ‘residual obligation’ on the Chamber to fully satisfy itself on jurisdiction, irrespective of the likelihood that a party's case for or against jurisdiction is true.Footnote 121
Additionally, as the Appeals Chamber itself noted recently,Footnote 122 the two sets of proceedings answer two different questions. If the Court does not have jurisdiction over the case, then proceedings must be terminated and the suspect or accused must be released. If, on the other hand, the case is inadmissible, then it should be investigated or tried at the national level – unless this has already happened. Admissibility operates as a mechanism to resolve jurisdictional conflict; in the end, a judge somewhere will examine the case (unless this has already happened). Evidently, in light of the Court's purpose to end impunity, a decision of lack of jurisdiction will be likely more problematic than a ruling of inadmissibility. A teleological approach to jurisdiction might require a more stringent test than ‘balance of probabilities’.
Be that as it may, the Bemba Chamber's conclusion seems to be that, if the defence argument of inadmissibility is more likely true than not, this will be duly considered in its judicial assessment; it does not mean automatically that the case is inadmissible. The application of this test by analogy to questions of jurisdiction has not been rejected by the Court to date.
4. Conclusion
The burden and standard of proof of facts at the jurisdictional stage of proceedings pose difficult questions. Mature national systems have struggled to find answers for many years, with mixed results.Footnote 123 It could not be any different in the young ICC system. In that context, judges need to solve a riddle: how to decide that certain facts exist to such a degree that the Court can proceed with a case, without predisposing matters that should be heard at trial or jeopardizing the right of the accused to a fair trial? On the one hand, the need to avoid pre-disposition and to respect the rights of the defence points towards lower evidentiary thresholds, where ‘probability’ or ‘likelihood’ are keywords. On the other hand, procedural economy and efficiency of process dictate finality of jurisdictional determinations as early as possible. Finally, the Court's objective to end impunity would arguably militate against stringent self-imposed limitations to the reach of the Court's jurisdiction. Legal certainty hangs in the balance between these competing concepts.
This article has followed the practice of the Court as regards burden and standard of proof in jurisdiction. Section 2 demonstrated that the burden of proof is largely settled in case law; the defence has the burden of proving the basis of its challenge. Section 3, however, showed that the Chambers are not in agreement as regards the precise standard of proof that the defence has to meet. In particular, this section demonstrated that Mbarushimana avoided to further its own earlier mention of a prima facie standard, on the basis that the defence challenge to jurisdiction in casu depended ‘much more’ on law than on facts. Pre-Trial Chamber II adopted a ‘degree of certainty’ standard, without however explaining its legal foundation or propriety. Notwithstanding Judge Kaul's efforts to offer some reasoning on the rejection of a gradation of standards and the effect of the ‘degree of certainty’, it remains to date a standard that has not been endorsed by the Appeals Chamber. On the contrary, in 2012, the Appeals Chamber effectively ruled that facts should be properly litigated at the merits stage of a hearing, be it at confirmation or trial, at least as regards challenges to jurisdiction ratione materiae. Finally, this section explored the analogical application of the admissibility standard of ‘balance of probabilities’ to jurisdiction, a possibility that has yet to be excluded.
A ruling by the Appeals Chamber seems therefore necessary, considering that facts will have to be litigated to one extent or another soon in the process of raising a challenge to jurisdiction before a Pre-Trial Chamber. In adopting such final solution, the Court could be guided by certain important considerations.
The first is that the differences between the Court's institutional framework and other national or international courts cannot be underestimated.Footnote 124 Analogies cannot be drawn lightly from other national or international systems.Footnote 125 On the other hand, however, the Court is still part of the international system. The Statute requires recourse to other sources in Article 21, when Article 21(1)(a) reaches its limits.Footnote 126 The jurisdiction of the Court is not excluded from this hierarchy.
Essentially, the Court could more rigorously seek guidance from international jurisprudence on possible solutions in deciding questions of burden and standard of proof as regards jurisdiction. After all, it is not the first court in international history to deal with these issues. It could consider some lessons learned from other institutions and tailor the best available practices to its framework.Footnote 127
The application of solutions adopted by international courts on the question of jurisdiction in the ICC context can be illustrated through our earlier example. If the suspect raised a challenge to jurisdiction alleging forgery in the registration of a ship, it would seem extremely inefficient to defer the matter until the merits stage of the confirmation hearing or until trial, as the 2012 Appeals Chamber ruling might be read to imply. A better approach would be to follow an international law approach and address the questions raised at the jurisdictional stage, as the challenge raises factual questions that, if established, would affect the jurisdiction of the Court to decide on the case.Footnote 128 It would be for the challenging party to explain whether and how the facts alleged, if true, would affect the Court's jurisdiction to decide on the alleged crimes.Footnote 129
As regards the standard of proof, the Court has many options available. It might reject the degree of certainty and opt for an internationalized civil law approach, requiring conviction of the court, without reference to a more specific standard.Footnote 130 Such approach may be all the more necessary considering that the drafters have decided to afford the judges considerable discretion in making such determinations.Footnote 131
In the alternative, the Court could maintain and fine-tune the degree of certainty. In doing so, it could engage in a teleological interpretation of Article 19(1) of the Statute. Since one of the Court's main objectives is to end impunity, the standard of proof for a defence challenge to jurisdiction should arguably be very high in order to avoid fostering the very impunity gap the Court was created to combat. At the same time, however, considering that evidence relating to jurisdiction would in all likelihood be at the hands of a state and not the defence, the standard could be somewhat moderated, particularly if a state does not comply with the Court's requests for co-operation in that regard.Footnote 132 Otherwise, the burden upon the defence could be unjust and a reversal might be in order.Footnote 133
In this spirit, the Court could stipulate this standard more clearly, in tune with the overall approach of the ICJ on this point. The latter has explained that the ‘degree of certainty standard’ does not operate in a vacuum, but rather requires ‘certainty’ supported by ‘convincing evidence’.Footnote 134 Obviously, that is not to say that the ICJ itself has addressed all questions concerning the degree of certainty in its system.Footnote 135 However, its recent case law appears to suggest that ‘convincing evidence’ may be seen as a half way standard between ‘balance of probabilities’ and ‘fully conclusive evidence’.Footnote 136
It is therefore not so much a question of judges becoming legislators and overstepping their bounds. Judges should decide these issues. In doing so, however, the Court could examine more carefully other international solutions on questions of jurisdiction. Thus far, the Court has been very restrictive in its approach; it has paid lip-service to the inapposite case law of the ICTYFootnote 137 or invoked without support in admissibility decisions the purported legal regulation in ‘the overwhelming preponderance’ of legal systems in the world.Footnote 138
Understandably, a more serious engagement with international jurisprudence may be beyond the judges’ zone of judicial comfort – or time availability. However, this course of action seems necessary in order to protect the Court from the adoption of decisions ill-based in law and corresponding accusations of judicial activism.
