On 25 September 2009, the Appeals Chamber of the International Criminal Court (ICC) issued a seminal decision on the subject of complementarity in the case Prosecutor v. Germain Katanga.Footnote 1 The notion of complementarity – which generally provides that where a case is genuinely being prosecuted by a national judicial system, that case will be inadmissible before the Court – is one of the fundamental principles on which the ICC was founded. This article will explore the jurisprudence on this principle in the Katanga case, including the recent Appeals Chamber decision. It will then discuss the impact of the Appeals Chamber's decision on the work of the Court going forward. As explained in detail below, the outcome of the Chamber's decision is that, even if a state has initiated an investigation or prosecution against an individual, the ICC may prosecute that individual for the same crimes or even a more selective range of crimes, so long as the state is willing to close the ongoing investigation or prosecution at the request of the ICC Prosecutor. While this decision is defensible under the language of the Rome Statute and consistent with the notion of state sovereignty in that the Court cannot force states to prosecute cases, this article concludes that, absent vigilance on the part of the ICC's Office of the Prosecutor, the decision could produce consequences inconsistent with the principle that the ICC is intended to act as a court of last resort. This is an important principle for several reasons: (i) there are benefits to national prosecutions, such as the fact that evidence and witnesses would be more readily available; (ii) states should be encouraged, as the Rome Statute preamble affirms, to exercise their jurisdiction to punish egregious international crimes; and (iii) the ICC has limited resources that should be reserved for trying individuals who face no possibility of domestic prosecution. These potential consequences, in turn, suggest that the Prosecution's policy of ‘positive complementarity’ – that is, encouraging genuine national proceedings whenever possible – should be at the core of its case selection strategy. At the same time, in those instances when the ICC prosecution determines that, despite activity by a national system with respect to a particular case, it is appropriate for the ICC to take over the case, the prosecution should clearly and publicly explain the factors that led to its decision.
1. The Katanga background and relevant jurisprudence relating to the complementarity challenge
1.1. Background
On 25 June 2007, the prosecution of the ICC submitted a request to Pre-Trial Chamber I seeking an arrest warrant for Germain Katanga, alleging that Katanga is responsible for a number of war crimes and crimes against humanity committed on or about 24 February 2003 in the village of Bogoro in the Democratic Republic of the Congo (DRC).Footnote 2 On 2 July 2007, the Pre-Trial Chamber issued a decision granting the requested arrest warrant.Footnote 3 In that decision, the Chamber noted that Katanga had been arrested in the DRC in March 2005 pursuant to a Congolese arrest warrant and had since that time been detained in the DRC on charges including alleged crimes against humanity.Footnote 4 Nevertheless, the Chamber found that the case was admissible before the ICC because the proceedings against Katanga in the DRC did not ‘encompass the same conduct’ that was the subject of the Prosecution's arrest warrant application.Footnote 5 Thus the Chamber concluded that there was no bar to the admissibility of Katanga's case pursuant to Article 17 of the Rome Statute.Footnote 6
While the Pre-Trial Chamber provided no further explanation of its decision regarding the admissibility of the Katanga case in its 2 July 2007 decision, the same chamber had taken a similar approach in the case against Thomas Lubanga Dyilo, the first accused to be arrested by the ICC.Footnote 7 Lubanga, like Katanga, was already in the custody of Congolese authorities at the time that the ICC Prosecutor sought an arrest warrant from the Pre-Trial Chamber. Specifically, Lubanga had been arrested in the DRC in March 2005 and was being held by Congolese authorities on several charges, including genocide and crimes against humanity.Footnote 8 On 13 January 2006, the ICC prosecution submitted an application for an arrest warrant to the Pre-Trial Chamber on the basis of allegations that Lubanga bears responsibility for the war crimes of recruiting, conscripting, and enlisting child soldiers from September 2002 to June 2003.Footnote 9 In its application, the prosecution acknowledged the proceedings against Lubanga in the DRC, but argued that there was no bar to admissibility because when the Congolese government referred the situation in the DRC to the ICC in March 2004, the government had stated that it was not able to prosecute crimes falling within the jurisdiction of the ICC, and thus the DRC was ‘unable’ to investigate or prosecute the Lubanga case within the meaning of Article 17 of the Rome Statute.Footnote 10 The Pre-Trial Chamber rejected this argument, explaining that since March 2004, the DRC national judicial system had ‘undergone certain changes’ and therefore could no longer be considered ‘unable’ within the meaning of Article 17.Footnote 11 Nonetheless, the Chamber determined that the prosecution's case against Lubanga was admissible before the ICC because, ‘for a case arising from the investigation of a situation to be inadmissible, national proceedings must encompass both the person and the conduct which is the subject of the case before the Court’.Footnote 12 The Chamber arrived at this conclusion because Article 17(1)(a) provides that ‘a case is inadmissible where . . . [t]he case is being investigated or prosecuted’,Footnote 13 and the word ‘case’ refers to ‘specific incidents during which one or more crimes within the jurisdiction of the Court seem to have been committed by one or more identified suspects’.Footnote 14 Because Lubanga was charged by the DRC with crimes other than war crimes relating to the recruitment and use of child soldiers, the Chamber reasoned, the DRC was not in fact conducting any proceedings relating to the case presented by the ICC prosecution and Article 17 was, therefore, not a bar to admissibility.Footnote 15 Similarly, because the Pre-Trial Chamber found that Germain Katanga was not being investigated or prosecuted by the DRC in relation to the events in the village of Bogoro on or about 24 February 2003, the Chamber found the case admissible and issued an arrest warrant for Katanga.Footnote 16
Following the issuance of the ICC's arrest warrant for Katanga, the Congolese authorities agreed to surrender him to the Court, and Katanga was transferred to The Hague on 17 October 2007.Footnote 17 The charges against Katanga were confirmed on 26 September 2008 and the case was submitted to Trial Chamber II.Footnote 18
1.2. Germain Katanga's admissibility challenge
On 11 March 2009, the defence for Katanga submitted a motion to Trial Chamber II seeking dismissal of the case against Germain Katanga on the ground that the case is inadmissible pursuant to Article 17 of the Rome Statute.Footnote 19 The motion set forth several arguments, including that Pre-Trial Chamber I's so-called ‘same-conduct test’, under which the Chamber held that a case will be inadmissible before the ICC only if a national system was investigating or prosecuting the exact charges that the ICC Prosecution sought to bring, constitutes ‘flawed’ precedent.Footnote 20 According to the defence, the same-conduct test not only lacks authority in the Rome Statute, but also ‘departs from the natural and proper interpretation of Article 17 if viewed in the light of its object and purpose’.Footnote 21 To illustrate this latter point, the defence offered a hypothetical situation where an individual allegedly has participated in crimes against humanity in ten villages, and is under investigation or prosecution at the national level for the attacks in nine of the villages.Footnote 22 The ICC prosecution could then investigate the attack in the tenth village, which would be an admissible case under Article 17 under the same-conduct test. The prosecution, then, is ‘in a position to put an end to serious investigations and prosecution at the national level’, even where the ICC proceedings would pursue more selective charges.Footnote 23 According to the defence, a more appropriate test would be one that compared either the relative gravity of the relevant charges or the comprehensive conduct at issue in each case. Under these tests, a case would only be admissible before the ICC if the case to be prosecuted at the Court involved crimes of greater gravity, or a greater range of alleged illegal conduct, than the crimes being investigated at the national level.Footnote 24 Because the DRC was prosecuting Katanga for crimes of comparable gravity and based on comparable conduct, the defence argued, the ICC should find the case against Katanga inadmissible.Footnote 25
1.3. The Trial Chamber decision
The Trial Chamber issued its decision on 16 June 2009, rejecting Katanga's motion under Article 17.Footnote 26 Interestingly, although the defence motion had primarily been framed around the validity of the same-conduct test, the Trial Chamber held that it need not look at whether or not the test was valid or even at whether it had been satisfied in the case of Katanga. Rather, the Trial Chamber held that, regardless of whether Katanga was being tried for the same or different conduct by the DRC, the Congolese authorities had willingly surrendered him to the ICC, and therefore the national system must be deemed ‘unwilling’ to try the case within the meaning of Article 17.Footnote 27
In its decision, the Trial Chamber recognized that each of the three subparagraphs of Article 17(2), which defines ‘unwillingness’ for purposes of determining the admissibility of a case, relates to ‘unwillingness motivated by the desire to obstruct the course of justice’.Footnote 28 Nevertheless, it then went on to state that there is also
the case of a State which may not want to protect an individual, but, for a variety of reasons, may not wish to exercise its jurisdiction over him or her. This second form of ‘unwillingness’, which is not expressly provided for in Article 17 of the Statute, aims to see the person brought to justice, but not before national courts. The Chamber considers that a State which chooses not to investigate or prosecute a person before its own courts, but has nevertheless every intention of seeing that justice is done, must be considered as lacking the will referred to in Article 17.Footnote 29
Katanga promptly appealed against the Trial Chamber's decision, arguing inter alia that the scenarios listed in Article 17(2) of the Rome Statute are the ‘exhaustive list [of] circumstances from which unwillingness on the part of the State to bring genuinely a person to justice can be detected’.Footnote 30 Therefore, according to the defence, the Trial Chamber was wrong to broaden the list.Footnote 31
1.4. The Appeals Chamber decision
In its decision of 25 September 2009, the Appeals Chamber, like the Pre-Trial Chamber and the Trial Chamber, rejected the argument that Katanga's case was inadmissible before the ICC on account of the proceedings against Germain Katanga in the DRC.Footnote 32 However, rather than applying the ‘same-conduct test’ applied by the Pre-Trial Chamber, or the ‘unwillingness’ test applied by the Trial Chamber, the Appeals Chamber came up with yet a third way to find that the Katanga case could proceed at the ICC. This time, the Court held that Article 17 did not bar the case against Katanga because the DRC, prior to the filing of Katanga's motion challenging the admissibility of the case, had closed the domestic proceedings against Katanga.Footnote 33 Because the relevant provision of Article 17 states that a case shall be inadmissible if ‘the case is being investigated or prosecuted by a State which has jurisdiction over it’,Footnote 34 the Appeals Chamber reasoned, a successful admissibility challenge would require the domestic proceedings to remain ongoing even after a suspect had been transferred to the Court.Footnote 35 Thus the Appeals Chamber reached the same result as the Trial Chamber – namely, that as long as a state is willing to surrender a suspect to the ICC, it is irrelevant whether that person was being genuinely prosecuted in a domestic jurisdiction prior to the Court's intervention – but in a manner that conforms more closely with the plain language of the Rome Statute.
2. Analysis
2.1. While consistent with the language of the Rome Statute, the Appeals Chamber's decision may have consequences inconsistent with a fundamental principle underlying the founding of the ICC
As discussed directly above, the Appeals Chamber's decision on complementarity in the Katanga case means that any time a state chooses to co-operate with a request from the ICC to surrender an individual sought by the Court's Prosecutor, the Court will be able to try a person who was, prior to the ICC's intervention, being prosecuted in a domestic system. Importantly, this will be the case even if the domestic system was trying the person for the same crimes as the ICC, and even where the domestic case involved a broader range of serious crimes. As a matter of statutory application, the Appeals Chamber's decision is completely defensible, as the Rome Statute only bars a case on the grounds of complementarity where it is actively and genuinely being investigated or prosecuted by a domestic judicial system.Footnote 36 This approach makes sense because, as the Appeals Chamber explains, the investigative or prosecutorial actions of a state vis-à-vis a given case may change over time.Footnote 37 If the Court were to examine the status of domestic proceedings as at the time that an arrest warrant is issued, it would be simple for a state wishing to shield an individual from the ICC to commence ‘an investigation’ at the time that the Court issues an arrest warrant and then, following a successful admissibility challenge, drop the investigation. Furthermore, it is defensible as a practical matter, as there may be instances where a state has commenced an investigation but becomes unable to follow through with a case, for example because the state becomes engulfed in an armed conflict or the system finds itself unable to guarantee the safety of witnesses. Because the Court cannot force a state to reopen a closed investigation or prosecution, conducting an admissibility review as of the time an arrest warrant is issued could result in the release of a suspect who no longer faces the prospect of investigation or prosecution domestically.
Nevertheless, it is important to stress that – as a matter of policy – the Chamber's decision effectively authorizes the prosecution to engage in practices that may be inconsistent with one of the fundamental principles underlying the founding of the ICC, at least to the extent that a state closes national proceedings simply because it would prefer that the ICC bear the cost of prosecuting the case, rather than because of any real inability on the part of the state's judicial system. That principle, which is that the ICC should be a court of last resort, was continually stressed during the drafting of the Rome Statute. For instance, in its commentary to the 1994 draft statute that became the text from which the Rome Statute was negotiated, the International Law Commission (ILC) emphasized that the Court ‘is intended to operate in cases where there is no prospect of those persons being duly tried in national courts’.Footnote 38 This sentiment was repeated verbatim in the report of the International Commission of Jurists commenting on the ILC draft statute.Footnote 39 Of course, states were partly concerned with protecting their own sovereignty, and thus wanted to limit the exercise of the Court to exceptional circumstances.Footnote 40 Yet this was not the only reason behind the importance attached by the drafters to ensuring that the ICC would be a court of last resort.
First, as reflected in the report of the 1995 Ad Hoc Committee on the Establishment of an International Criminal Court, states recognized that prosecuting cases domestically carried certain benefits relative to international prosecutions. In particular, it was noted that in domestic proceedings:
(a) all those involved would be working within the context of an established legal system, including existing bilateral and multilateral arrangements; (b) the applicable law would be more certain and more developed; (c) the prosecution would be less complicated, because it would be based on familiar precedents and rules; (d) both prosecution and defence were likely to be less expensive; (e) evidence and witnesses would normally be more readily available; (f) language problems would be minimized; (g) local courts would apply established means for obtaining evidence and testimony, including application of rules relating to perjury; and (h) penalties would be clearly defined and readily enforceable.Footnote 41
The 1995 report also recognized that states ‘had a vital interest in remaining responsible and accountable for prosecuting violations of their laws – which also served the interest of the international community, inasmuch as national systems would be expected to maintain and enforce adherence to international standards of behaviour within their own jurisdiction’.Footnote 42 Interestingly, during the meetings of the Ad Hoc Committee, it was ‘suggested that the draft statute should provide for the possibility that a State might voluntarily decide to relinquish its jurisdiction in favour of the international criminal court in respect of crimes expressly provided for under its statute’.Footnote 43 Yet, according to the 1995 report, ‘[t]his suggestion gave rise to reservations on the ground that it was not consistent with some delegations’ view of the principle of complementarity’, and it was once again reiterated that the Court ‘should only be resorted to in exceptional cases’.Footnote 44 Consistent with this theme, delegates to the Preparatory Committee for the Establishment of an International Criminal Court, which met the following year, expressed the view that the Statute should ‘reiterate the obligation of States . . . to investigate vigorously and prosecute criminal cases’.Footnote 45 This suggestion was ultimately implemented with the inclusion in the preamble to the Rome Statute of provisions ‘[a]ffirming that the most serious crimes of concern to the international community as a whole must not go unpunished and that their effective prosecution must be ensured by taking measures at the national level and by enhancing international cooperation’ and ‘[r]ecalling that it is the duty of every State to exercise its criminal jurisdiction over those responsible for international crimes’.Footnote 46
Finally, delegates stressed that ‘the limited resources of the Court should not be exhausted by taking up the prosecution of cases which could easily and effectively be dealt with by national courts’.Footnote 47 In other words, as communicated in a 1996 press release issued by the Preparatory Committee, the Court was not designed to be ‘a “garbage can” into which national court systems could dump criminals that they should be punishing at the national level’, but rather was designed to ‘fill in the gaps of impunity’.Footnote 48
Commentary published after the adoption of the Rome Statute further supports the notion of the ICC as a court of last resort. For example, the first president of the ICC, Philippe Kirsch, observed that ‘[i]t is the essence of the principle of complementarity that if a national judicial system functions properly, there is no reason for the ICC to assume jurisdiction’.Footnote 49 Similarly, the group of experts convened by the ICC Prosecutor in April 2003 to study ‘complementarity in practice’ observed that the ‘principle of complementarity is based both on respect for the primary jurisdiction of States and on considerations of efficiency and effectiveness, since States will generally have the best access to evidence and witnesses and the resources to carry out proceedings’.Footnote 50 This group of experts further observed that ‘there are limits on the number of prosecutions the ICC, a single institution, can feasibly conduct’,Footnote 51 and that it was ‘important to ensure that the Court does not become overburdened as a result of States shirking their responsibilities to help end impunity’.Footnote 52
Hence the principle of complementarity was not included in the Rome Statute purely to protect state sovereignty. Rather, the Statute's drafters saw benefits to prosecuting cases domestically where possible, wanted to encourage states to take responsibility for punishing egregious crimes, and wanted to safeguard the ICC's resources for prosecuting those individuals who would otherwise go unpunished. In the aftermath of the Appeals Chamber's decision in Katanga, however, the Court is free to take up cases that would otherwise have been prosecuted domestically so long as a state agrees to co-operate with a request from the ICC Prosecutor to allow the Court to take over the case.
It has been suggested that the potentially problematic consequences stemming from the complementarity decision in the Katanga case exist only because the situation out of which the case arose was a ‘self-referral’ – that is, referred to the Court by the DRC itself.Footnote 53 Yet the fact is that the Court would be faced with the same issues even if the situation had been triggered in some other way, as long as the state in which the investigations are occurring is willing to co-operate with the Court. For instance, if the Prosecutor had invoked his proprio motu authority under Article 15 to initiate the situation in the DRC, nothing would prevent the government of the DRC from, at some point, choosing to co-operate with the ICC. Thus, as discussed immediately below, the problems that potentially arise as a result of the ICC Prosecutor choosing to pursue cases that are being investigated or prosecuted by a state must be addressed by a change in the Prosecutor's case selection strategy, not necessarily by a change in the Court's policy of accepting self-referrals.
2.2. Given the relevant policy considerations, the ICC prosecution should encourage domestic prosecutions whenever possible and publicly explain its actions when taking over a case being investigated or prosecuted domestically
Based on the foregoing, it seems logical that, as a policy matter, the ICC prosecution should only be pursuing cases where there is no possibility of those cases being tried domestically. Indeed, the ICC prosecution itself committed to this principle early in its operations. Specifically, the Office of the Prosecutor stated in a September 2003 paper that ‘[a]s a general rule, the policy of [the Office] will be to undertake investigations only where there is a clear case of failure to act by the State or States concerned’.Footnote 54 While the prosecution's policy paper goes on to provide that the Office of the Prosecutor and a state may agree to a ‘consensual division of labour’,Footnote 55 the envisioned use of such consensual labour-sharing agreements is limited. For instance, the paper suggests that the ICC may step in where ‘[g]roups [that are] bitterly divided by conflict’ agree that a prosecution by the ICC would be ‘neutral and impartial’, or where ‘a third State has extra-territorial jurisdiction, but all interested parties agree that the Court has developed superior evidence and expertise relating to that situation, making the Court the more effective forum’.Footnote 56 Yet, in practice, two of the four individuals for whom the prosecution has sought arrest warrants in the situation in the DRC to date – Thomas Lubanga Dyilo and Germain Katanga – were the subjects of domestic proceedings at the time that the ICC issued the arrest warrant, and there is no evidence that the Court took over these cases due to the types of scenario described in the prosecution's policy paper.
Importantly, the practice of pursuing individuals who could have been prosecuted domestically not only seems to run contrary to the principles underlying the Rome Statute's drafters’ desire for the ICC to act as a court of last resort, but may carry additional unintended negative consequences. For example, according to one source, the ICC Prosecutor's decision to take the cases against Lubanga and Katanga out of the DRC ‘undermines the confidence of domestic judiciaries; it sends a message that they might be trying to reform themselves and might be trying to deal with very complicated justice questions, but that's not necessarily going to stop an international body from intervening’.Footnote 57 In addition, where the ICC prosecutes a case that is narrower in scope than the domestic case that is interrupted, large numbers of victims may be denied the benefits associated with a trial of the crimes committed against them.
Given these considerations, as well as those discussed in the previous section, we recommend that the prosecution fully implement its policy of pursuing individuals only in the case of a clear failure by the state to prosecute. As part of this process, the prosecution could consider formalizing its policy of ‘positive complementarity’, under which the Office of the Prosecutor has committed itself to ‘encourag[ing] genuine national proceedings where possible’.Footnote 58 For instance, following the example of the International Criminal Tribunal for the former Yugoslavia (ICTY), which has transferred developed cases as well as investigative files to national prosecutors in Bosnia and Herzegovina, the ICC prosecution could work on developing the ‘internal infrastructure’ necessary to transfer information to domestic judicial authorities.Footnote 59 According to a former Deputy Prosecutor of the ICTY, David Tolbert, this would involve the creation of a specialized unit within the Office of the Prosecutor that would ‘focus on providing information to national judicial authorities, and, where possible and appropriate, work with other actors (e.g., development agencies, [non-governmental organizations]) to assist in building capacity in the relevant national systems’.Footnote 60 Tolbert further recommends implementation of some of the specific tools developed by the ICTY for the purpose of information sharing, such as electronic databases consisting of ‘the non-confidential portions of certain of its data and evidence collection’, which could be accessed by national prosecutors, and a ‘specialized request unit designed to respond to [requests for assistance]’.Footnote 61 The ICC may also want to include in such a programme methods for helping national prosecutors and courts with substantive issues of international law.Footnote 62 With such processes in place, the ICC Prosecution may feel more secure in allowing those cases that are being tried in domestic courts to stay within the national system, leaving itself free to pursue those individuals who are not being investigated by any state.
Of course, even with a formalized ‘positive complementarity’ strategy, it may well be that the ICC Prosecutor finds it necessary to take over cases from national judicial systems in the future due to issues such as the national system's inability to protect witnesses or judges effectively, where there are serious questions concerning the state's ability to deliver a verdict that will be accepted as legitimate across rival factions, or where a state is not genuinely pursuing a supposedly ongoing investigation. In such instances the Office of the Prosecutor should clearly communicate the reasons behind its decision to take over the case, even if such explanation is not required in the context of a formal complementarity challenge at the Court. For instance, the prosecution suggested that no action was being taken in relation to any of the domestic charges against Katanga for the first time, at least publicly, in its brief opposing Katanga's admissibility challenge and, even in that brief, it did not argue that it pursued the case against Katanga because it did not believe that the DRC was genuinely investigating any charges against him.Footnote 63 If the prosecution acted against Germain Katanga because it had no faith that Katanga would ever actually face prosecution in a Congolese court, this fact should have been publicized, perhaps in one of the prosecution's press releases relating to the initial arrest of Katanga.Footnote 64 Such explanations will be helpful in mitigating the potential unintended negative consequences, discussed above, of pursuing individuals who could have been prosecuted domestically and thus contribute to positive public perceptions about the Court.
3. Concluding remarks
In sum, we agree with the conclusion reached by the ICC Appeals Chamber in Katanga regarding the application of Article 17, but have potential concerns about the approach of the prosecution towards complementarity. To address these concerns, we recommend that the prosecution fully implement its policy of ‘positive complementarity’ and ensure effective communication of its reasons behind pursuing individuals that were presumably being prosecuted domestically.