Leiden Journal ofInternational Law (2015), 28, pp. 349-371
© Foundation of the Leiden Journal of International Law 2015   doi:io.ioi7/So92 2156515000114
Current Trends in the Definition of 'Perpetrator' by the International Criminal Court: From the Decision on the Confirmation of Charges in the Lubanga case to the Katanga judgment***
ALICIA GIL GIL* AND ELENA MACULAN**
Abstract
The jurisprudence of the International Criminal Court (ICC) up to the Lubanga judgment showed definite interpretive trends on the modes of principal liability. This article aims first to make a critical assessment of these trends by focusing on methodological and substantive aspects. On the one hand, the practice of having resort to theories derived from Continental legal systems, albeit legitimate, is based on a methodology that raises some concerns as to the selection and (mis)interpretation of such theories. On the other hand, the Court has clearly adopted a wide interpretation of some critical elements in which the different modes of principal liability are grounded. This choice has caused a significant expansion of the scope of principal liability as well as a breach of the principles of legality and of individual criminal responsibility. In our opinion, the underpinning of these interpretations is a flawed understanding of the criteria for distinguishing between principals and accessories.
This perspective has been overturned by the Katanga judgment, on which the second part of this article will focus. This judgment correctly argues that the distinction between perpetrators and accomplices is grounded only on the autonomous or vicarious character of their contribution to the offence. Furthermore, it follows a partly different approach as to both the methodology and the interpretation of the constitutive elements of principal liability. In our view, this approach better fits both the relevant statutory provision and the basic principles of criminal law.
Key words
control over the crime; co-perpetration; indirect perpetration; Katanga; Lubanga
i. Introduction
Thomas Lubanga Dyilo and Germain Katanga are the first people convicted by the International Criminal Court (hereinafter ICC) for the crimes committed during the
* Professor of Criminal Law at the UNED (Universidad Nacional de Education a Dištancia), Madrid (Spain); Subdirector of the Instituto Universitario Gutierrez Mellado, Madrid [agil@der.uned.es].
:* Post-Doc Research Fellow at the Instituto Universitario Gutierrez Mellado (UNED), Madrid; Lecturer in Transitional Justice, University of Trento (Italy) [emaculan@igm.uned.es].
:* This article is the result of research conducted in the framework of the research project 'Criminal responsibility in International Criminal Law. Modes of liability for international crimes' (reference: DER2009-07236) funded by the Ministry of Economy and Competitiveness (Spain) and directed by Prof. Alicia Gil Gil. The results of the project are collected in A. Gil Gil, (dir.), E. Maculan (coord.), Intervención delictiva y Derecho penal internacionál: Reglas de atribución de responsabilidad en crímenes internacionales (2013).
350   ALICIA GIL GIL AND ELENA MACULAN
armed conflict that took place in the region of Ituri in the Democratic Republic of the Congo (DRC).1 Both convicts were leaders of rebel groups: Lubanga Dyilo was the President of the Force Patriotique pour la Liberation du Congo, and Katanga the leader of the Forces de Resistance Patriotique en Ituri.
Leaving aside these common traits between the defendants, the judgments against them by the ICC reflect slightly different approaches to the interpretation and application of modes of liability This is one of the central issues in international criminal law, since the phenomenology of international crimes poses difficult challenges to the identification of the liability doctrines subject to application in this field.2
In this respect, the Court has shown definite trends in interpreting Article 2 5(3) of the Statute - the provision establishing the bases for attributing individual criminal responsibility within the ICC regime. This is demonstrated by the approach adopted by the ICC chambers in the decisions on the confirmation of charges (hereinafter DCC) in both the Lubanga and the Katanga and Ngudjolo Chui cases and confirmed in the Lubanga trial judgment. However, the Katanga trial judgment partially overturns the previous decisions and proposes a novel and, in our view, more convincing approach.
The aim of this article is to make a critical assessment of interpretive trends in the ICC jurisprudence up to the Lubanga trial judgment as regards the modes of principal liability. It focuses on the Court's resort to liability theories derived from Continental legal systems. This choice, albeit legitimate, is based on a deficient methodology and raises concerns related to (mis)interpretation of such theories in the Court's jurisprudence.
The Court has clearly adopted a wide interpretation of some of the critical elements in which the different modes of principal liability are grounded, namely, the 'essential contribution' as a criterion for establishing the control over the crime and the 'common plan' as a requirement for co-perpetration. This choice has caused a significant expansion of the scope of principal liability and breached the basic principles of legality and individual criminal responsibility, as affirmed in Articles 22 and 25(2) of the ICC Statute, respectively.
1 Lubanga was convicted by the ICC as a co-perpetrator of the war crime of conscripting and enlisting children under the age of 15 and using them to participate actively in hostilities: Prosecutor v. Thomas Lubanga Dyilo, Judgment pursuant to Art. 74 of the Statute, ICC-01/04-01/06-2842, T.Ch. I, 14 March 2012. The judgment and the sentencing judgment have been upheld on appeal: Prosecutor v. Thomas Lubanga Dyilo, Judgment on the appeal of Mr. Thomas Lubanga Dyilo against his conviction, ICC-oi/o4-oi/o6-3i2i-Red, A. Ch, and Judgment on the appeals of the Prosecutor and Mr Thomas Lubanga Dyilo against the 'Decision on Sentence pursuant to Article 76 of the Statute', ICC-01/04-01/06-3122, A. Ch., both of 1 December 2014. Katanga was indicted before the ICC for a number of crimes committed during the attack against the village of Bogoro on 24 February 2003 by the Ngiti militia of the rebel group of Walendu-Bindi and Lendu. He was finally convicted for the crime against humanity of murder and for the war crimes of murder, attacking a civilian population, and destruction of property and pillaging; however, he was acquitted of the charges of rape and sexual slavery (both as war crimes and crimes against humanity) as well as of the charge of using children under 15 to participate actively in hostilities: Le Procureur c. Germain Katanga, Jugement rendu en application de 1'article 74 du Statut, ICC-01/04-01/07, T.Ch. II, 7 Mars 2014. He was sentenced to 12 years' imprisonment. The judgment and the sentencing judgment have become final following the discontinuation of the appeals. Katanga's co-defendant, Mathieu Ngudjolo Chui, had already been acquitted of all of the charges after the severing of his case from Katanga's case: see Prosecutor v. Mathieu Ngudjolo Chui, Judgment pursuant to Article 74 of the Statute, ICC-01/04-02/12-3, T. Ch. II, 18 December 2012.
2 T. Weigend, 'Problems of Attribution in International Criminal Law', (2014) 12 JICJ 253.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 35I
In our opinion, these interpretations are underpinned by a flawed understanding of Article 25(3), especially of the criteria for distinguishing between principal and accessory liabilities. This approach has been expressly overturned by the Katanga trial judgment, on which the second part of this article will focus. Whereas this shows the pending disagreement within the Court on the interpretation of Article 25(3), the Katanga judgment arguably follows a more convincing approach, in that it better fits both the relevant statutory provision and the basic principles of criminal law.
2. ICC INTERPRETATION OF 'PERPETRATOR' UP TO THE lubanga TRIAL JUDGMENT
Article 25(3) provides the applicable modes of liability under which a person may be held responsible before the Court. It sets a clear distinction between the modes of principal liability (letter a of the provision) and accessory liability (letters b, c, and d). This distinction is mostly adopted by domestic criminal systems.3 Therefore, the ICC Statute applies a 'differentiation' model.4 Accordingly, it rejects the unitary approach to criminal responsibility that dominated post-Second World War jurisprudence and is still applied in a few domestic systems, such as Italy. This approach makes no normative distinction between principal and accessory liability; it simply states that anyone who in any way contributes to a crime may be held responsible for it. Nonetheless, the different degrees of contribution and blameworthiness might be taken into account at the sentencing phase.5
But Article 25(3) neither provides a criterion to distinguish between principals and accessories, nor does it state explicitly the requirements under which a person may be charged.
In an attempt to elucidate this open-textured provision, the ICC has moved to charge the defendants on the basis of the 'control over the crime' doctrine, as opposed to the theory predominantly applied by ad hoc tribunals, that is, the Joint Criminal Enterprise (hereinafter JCE). The DCC in the Lubanga case explicitly stated that the JCE doctrine cannot be applied under the Rome Statute, where the provision on the modes of liability is drafted differently than in the statutes of the ad hoc tribunals.6 However, the abandonment of JCE by the ICC is probably also due to the abundance of strident criticism to which it has been subject.7
3 See the comparative study of U. Sieber and K. Cornils (eds.), Nationales Strajrecht in rechtsvergleichender Darstellung (2010).
4 G. Werle and B. Burghardt, 'Coautorfa mediata: idesarrollo de la dogmatica juridico penal alemana en el Derecho penal international?', (2011) 28 Revista penal 197, at 198-9.
5 Judge Fulford criticizes the idea of an existing division between the modes of liability encompassed in Art. 2 5(3) ICC Statute; he argues instead that these concepts 'will often be indistinguishable in their application vis-a-vis a particular situation': Separate Opinion of Judge Fulford, attached to the Lubanga judgment, supra note 1, para. 7 [hereinafter Separate Opinion Fulford].
6 Prosecutor v. Thomas Lubanga Dyilo, Decision on the confirmation of charges, ICC-01/04-01/06-803, P.-T. Ch. I, 29 January 2007, paras. 332, 346 et seq. [hereinafter Lubanga DCC].
7 For the main criticisms of this theory, see, inter alia, J. D. Ohlin, 'Three Conceptual Problems with the Doctrine of Joint Criminal Enterprise', (2007) 5 JICJ 69; K. Ambos, Treatise on International Criminal Law (2013), Vol. I, at 17 2-6; M. E. Badar, '"Just Convict Everyone!" - Joint Perpetration: From Tadic to Stakic and Back Again', (2006) 6 ICLR 293; M. Gutierrez Rodriguez, 'Joint Criminal Enterprise iuna especie juridica en vfas de extincidn
352   ALICIA GIL GIL AND ELENA MACULAN
That said, the way the ICC has interpreted the modes of liability up to the Katanga judgment give similar results to those that would have arisen if the JCE doctrine had been applied.8 Both are grounded in a broad understanding of the modes of principal liability as opposed to accessory liability9 Additionally as will be discussed, the broad interpretation by the ICC of the 'common plan' requirement in cases of co-perpetration takes us back to the controversial JCE III.10
Another constant in the ICC case law has been the choice for modes of liability originating in Continental legal systems.11 In particular, the Court has used the theory of the 'control over the crime', according to which the perpetrator is the one who has the final say as to whether the crime will be committed and how it will be carried out.12 This theory allows to charge as perpetrators those who physically carry out the actus reus of the offence (commission of the crime in person, or direct perpetration), those who have control over the will of whoever carries out the objective elements of the offence (commission of the crime through another person, or indirect perpetration),13 and those who possess, along with others, the control over the offence by reason of the essential task assigned to them (commission of the crime jointly with others, or co-perpetration).14 Additionally, the ICC has relied on Roxin's doctrine of control over an organization in order to interpret the concept of perpetration through another person who is criminally responsible.15 According to
en el derecho penal international?', in A. Gil and E. Maculan (eds.), Intervencion delictiva y Derecho penal internacional: Reglas de atribuciön de responsabilidad en crimenes internacionales (2013), at 413. Yet, JCE is still applied by other international and hybrid criminal courts. The ECCC, for example, have accepted JCE in its first two variants (however, rejecting JCE III); see K. Gustafson, 'ECCC Tackles JCE', (2010) 8 JICJ 1323; L. Marsh and M. Ramsden, 'Joint Criminal Enterprise: Cambodia's Reply to Tadid, (2011) 11ICLR 137.
8 A similar view is held by M. Cupido, 'Pluralism in Theories of Liability: Joint Criminal Enterprise versus Joint Perpetration', in E. van Sliedregt and S. Vasiliev (eds.), Pluralism in International Criminal Law (2014). The author argues that the interpretation of both forms of liability in relevant case law ends up making them an autonomous crime, similar to the crime of participation in a criminal organization. In contrast, Ohlin points out the inconsistencies of both the JCE and the control theory as applied by the ICC, and concludes by asserting the need for a new, alternative rationale to ground individual responsibility for collective criminal action. He develops a proposal based on the concept of joint or shared intentions; see: J. D. Ohlin, 'Joint Intentions to Commit International Crimes', Paper 169 Cornell Law Faculty Publications (2011), available at <http://scholarship.law.cornell.edu/facpub/169> (accessed 29 January 2014).
9 This tendency is also followed in the jurisprudence of domestic courts in Latin American countries when dealing with international crimes: see E. Maculan, 'La fertilizaciön cruzada jurisprudencial y los modelos de responsabilidad', in Gil and Maculan, supra note 7,69.
10 See Section 2.2.2., infra.
11 E Jessberger and J. Geneuss, 'On the Application of a Theory of Indirect Perpetration in Al Bashir. German Doctrine at The Hague?', (2008) 6 JICJ 853; H. G. van Der Wilt, 'The Continuous Quest for Proper Modes of Criminal Responsibility', (2009) 7 JICJ 307; T. Weigend, 'Intent, Mistake of Law, and Co-Perpetration in the Lubanga Decision on Confirmation of Charges', (2008) 6 JICJ 471, at 479.
12 The ICC adopts this concept as developed by C. Roxin, Täterschaft und Tatherrschaft (2000); see also C. Roxin, 'Crimes as Part of Organized Power Structures', (2011) 9 JICJ 193.
13 Prosecutor v. Germain Katanga and Mathieu Ngudjolo Chui, Decision on the Confirmation of Charges, ICC-01/04-01/07-717, P.-T.Ch I, 30 September 2008, paras. 500-39, although later modified [hereinafter Katanga and Ngudjolo DCQ; Prosecutor v. Abdullah Al-Senussi, Warrant of Arrest, ICC-01/11-01/11-4, 30 June 2011, at 11.
14 This mode of liability has been applied in the Lubanga case (see Lubanga DCC, paras. 317 et seq.; Lubanga Judgment, para. 978), in the Banda and Jerbo case {Prosecutor v. Abdallah Banda Abakaer Nourain and Saleh Mohammed Jerbo Jamus, Corrigendum of the Decision on the confirmation of charges, ICC-02/05-03/09-121-Corr-Red, 7 March 2011), and in Prosecutor v. Omar Hassan Ahmad Al Bashir, Warrant of Arrest, ICC-02/05-01/09-1, P.-T.Ch. I, 4 March 2009, paras. 210 et seq.
15 Although recalling that doctrine and case law have also developed other theories to define indirect perpetration, the Court has argued that 'the cases most relevant to international criminal law are those in which the perpetrator behind the perpetrator commits the crime through another by means of "control
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 353
that theory, the responsibility of the perpetrator behind the perpetrator is based on the control over the agent's will within the framework of an organization.16
The Court has also developed a new mode of'cross-liability'17 that in effect merges co-perpetration based on the control over the crime with indirect perpetration.18 This form of liability - indirect co-perpetration - appeared for the first time in the Katanga andNgudjolo DCC.19 This construction allows the Court to charge a defendant with crimes committed by individuals who are not their subordinates, but who obey the orders given by another co-perpetrator.20 The rationale for charging the defendant in such cases is the mutual attribution of each co-perpetrator's contribution. In respect of Katanga it was later replaced by a mode of accessory liability given the lack of sufficient evidence as to the constitutive elements of'cross-liability'.21 As the following discussion will show, the choice made by the ICC of these Continental theories is problematic from both a methodological and a substantive perspective.
2.1. Methodological flaws
Given that Article 25(3)^) does not provide specific definitions of the modes of liability, the Court has defined their constitutive elements by extensive reference to domestic literature and jurisprudence.
The importation of theories developed in certain domestic systems is allowed under Article 2i(i)(c), which refers to the 'general principles of law derived by the Court from national laws of legal systems of the world', as a subsidiary source of law to which the Court may have recourse. Domestic case law and literature, along with legislation, are therefore the materials from which the Court may glean the general principles of law by means of a comparative exercise. Article 2i(i)(c) leaves
over an organisation"': Katanga and Ngudjolo DCC, para. 498. The possibility of labelling an individual as an indirect perpetrator even if the direct perpetrator is criminally responsible is expressly envisaged in the ICC Statute, in so far as Art. 25(3)^) states: 'regardless of whether that other person is criminally responsible'.
16 Roxin, Täterschaft, supra note 12, at 242. The constitutive elements on which Roxin bases this construction are: first, an organization whose hierarchical structure and number of members ensure automatic compliance with the orders given by the leader; second, the requirement that the organized power structure as a whole must operate outwith the law; third, the leader's effective power to give orders; fourth - and most importantly - the interchangeability of subordinates with respect to the compliance with orders (and, similarly, with respect to the execution of the crimes). This doctrine has given rise to a great debate in Germany, Spain, and Portugal; for an overview see K. Ambos, Der Allgemeine Teil des Völkerstrafrectcht (2002), at 590; R Faraldo Cabana, Responsabilidad penal del dirigente en estructuras jerárqukas (2003), at 88.
17 E. van Sliedregt, Individual Criminal Responsibility in International Law (2012), at 169.
18 See Weigend, supra note 2, at 258 et seq., welcomes this novel mode of'cross liability'. He also argues that it had already been developed in German jurisprudence and literature, despite being controversial within that system.
19 Katanga andNgudjoloBCC^íľí. 525. Judge Van den Wyngaert rejects this constructionbecause it goes beyond the terms of the Statute, therefore amounting to a violation of the principle of legality enshrined in Art. 22: see Concurring Opinion of Judge Christine Van den Wyngaert, attached to the Ngudjolo Chui Judgment (hereafter Cone. Opinion Van den Wyngaert), paras. 7 and 58-64.
20 On the case law applying this figure, see H. Olásolo, The Criminal Responsibility of Senior Political and Military Leaders as Principals to International Crimes (2009), at 302, and, more recently, H. Olásolo, Tratado de Autoríd y Participación en Derecho Penal Internacionál (2013) at 563.
21 See section 3, infra.
354   ALICIA GIL GIL AND ELENA MACULAN
the judges of the ICC broad discretion in selecting the legal systems on which to rely.22
It would be unfair not to acknowledge the extensive research work done by the chambers of the ICC, given the remarkable level of doctrinal and jurisprudential analysis in their judgments. However, their selection and interpretation of domestic sources has not always been methodologically sound.
First, the Court on occasions refers only to those sources that support the theory it has decided to endorse. This selective approach leads to a potential misstatement of the degree of acceptance of a particular theory in domestic jurisdictions. For example, the Lubanga DCC attempted to demonstrate that the criterion of the control over the crime enjoys wide acceptance in many legal systems, whereas in reality it is recognized almost exclusively in German- and Spanish-speaking literature.23
Similarly, PTC I in the Katanga and Ngudjolo DCC affirmed that many domestic jurisdictions apply the mode of indirect perpetration based on the control over an organization.24 In reality, however, it could only cite five systems25 out of the states parties to the Statute.26 In addition, some of the decisions cited by PTC I have subsequently been modified by higher courts in the same country, thus casting doubt on the effective acceptance of the theory even within those systems.27
Second, the Court has sometimes misinterpreted not only the degree of acceptance of a certain theory, but also the grounds on which it would be applied. Thus, when quoting national judgments in which courts supposedly applied the theory of
22 M. McAuliffe de Guzman, 'Article 21', in 0. Triffterer (ed.), Commentary on the Rome Statute of the International Criminal Court (2008), at 708-11.
23 LuhangaDCC, paras. 330 et seq. The only American scholar mentioned is G. Fletcher, Rethinking Criminal Law (2000), at 639. Fletcher argues that virtually all of the legal systems recognize the principal responsibility of the individual who commits a crime using another as an instrument ('perpetration by means'), but refers only to cases where the use was made of an innocent or non-responsible agent. One may wonder, therefore, if this reference is in itself sufficient to support the notion that the wide concept of control over the crime as employed by the Court (namely, including the use of a criminally responsible agent) finds recognition in common law systems. A similar remark is made by S. Manacorda and C. Meloni, 'Indirect Perpetration versus Joint Criminal Enterprise', (2011) 9 JICJ159, at 170.
24 Katanga and Ngudjolo DCC, para. 504.
25 Ibid., note 666. The Court quotes German, Peruvian, Chilean, Argentinean, and Spanish jurisprudence. For an overview of this jurisprudence, see F. Munoz Conde and H. Olasolo, 'The Application of the Notion of Indirect Perpetration through Organized Structures of Power in Latin America and Spain', (2011) 9 JICJ 113; Maculan, supra note 9. Van Sliedregt points out that, since the Spanish and Latin-American approach is highly influenced by the German one, in the end only the latter system lies at the basis of the control theory: E. van Sliedregt, 'Perpetration and participation in Article 25(3) of the Statute of the International Criminal Court', in C. Stahn (ed.), The Law andPractice of the International Criminal Court (2015), available at: <http://papers.ssrn.com/s0l3/papers.cfm?abstract_id=24927io> (accessed 29 January 2015), at 12.
26 See Sieber and Cornils (eds.), supra note 3, demonstrate that many countries do not recognize indirect perpetration: e.g., the Chinese criminal code applies a wide concept of direct perpetration (at 7); under the French criminal code, the indirect perpetrator is punished as an instigator (at 79). They also note that within several criminal systems indirect perpetration is admitted only with an innocent agent: for instance, in England and Wales (at 55), in the US Model Penal Code, para. 2.o6(2)(a), and in Ivory Coast (at 35 and 362).
27 In Peru, a few months after the judgment convicting Fujimori based on Roxin's theory of indirect perpetration through the control over an organization (see infra, note 30), the Supreme Court asserted that indirect perpetration does not apply in cases in which the direct perpetrator is criminally responsible: Corte Suprema de Justicia de Peru, Sala Permanente, Grupo Colina case (RN N. 4104-2010), Judgment of 13 June 2013. It opted instead for the normative concept of perpetration developed by Jakobs; see G. Jakobs, 'Sobre la autoria del acusado Alberto Fujimori', in K. Ambos and I. Memi (eds.), La autoria mediata. El caso Fujimori (2010), at 110.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 355
indirect perpetration by virtue of an organized power apparatus following Roxin, the Katanga DCC makes reference to specific decisions that do not rely entirely on that theory28 Moreover, the PTC makes reference to domestic precedents that applied a mixture between the theory developed by Roxin and that of F.-C. Schroder.29 The latter theory is based on the idea of the direct perpetrator's predisposition to commit the relevant act (Tatgeneigtheit, predisposition al hecho)}0 However, it fails to take this complementary element into consideration when defining the requirements of indirect perpetration. Instead, the PTC adopts the concept of the interchangeability of direct perpetrators, which is central to Roxin's doctrine.31 Nor does the Katanga and Ngudjolo DCC take into account the requirement of the detachment of the organization from the law, which, according to Roxin, is a constitutive element of this mode of liability.32 This is also so according to many domestic precedents in which his interpretation has been followed.33
Third, the Court follows minority opinions that are highly controversial even within the legal systems from which they stem. In defining co-perpetration, for example, the ICC admits that this label may also apply to a party who is only involved in the preparatory stage of the offence.34 By contrast, the majority view reflected in Continental literature is that whoever is involved in the preparatory stage of the offence leaves it to another actor to decide whether or not the offence will be committed.35 As a result, one may doubt that this interpretation really amounts to a 'general principle of law'.
28 The PTC quotes for example a judgment by the Spanish Supreme Court (Tribunal Supremo) of 2 July 1994 as if it applied indirect perpetration by virtue of an organized power apparatus, although in reality it did not apply this theory. The PTC also quotes another Spanish decision (Juzgado Central de Instruction No. 5 of the Audiencia Nacional (Investigating Judge of the Spanish High Court), Decision of 29 March 2006), in which the judge simply mentions the concept of indirect perpetration as a possible ground for charging the accused, together with the instigation or commission by way of omission. For discussion of these decisions, see A. Gil, 'El caso espanoP, in K. Ambos (ed.), Imputation de crimenes de los subordinados al dirigente (2009), at 87.
29 F.-C. Schröder, Der Täter hinter dem Täter. Ein Beitrag zur Lehre von der mittelbaren Täterschaft (1965).
30 The most relevant case is the conviction of former Peruvian President Alberto Fujimori by the Peruvian Supreme Court: Corte Suprema de Justicia del Peru, Sala Penal Especial, Alberto Fujimori (Exp. No. 19-2001-09-A.V.), Judgment of 7 April 2009, subsequently upheld by Primera Sala Transitoria, Judgment of 30 December 2009. In relation to this case, see K. Ambos, 'The Fujimori Judgment: A President's Responsibility for Crimes against Humanity as Indirect Perpetrator by Virtue of an Organized Power Apparatus', (2011) 9 JICJ 137. This judgment combines the traditional requirements of Roxin's construction with others taken from the doctrine developed by Schröder in order for the judgment to arrive at an 'integrated' position. See C. Roxin, Apuntes sobre la sentencia Fujimori de la Corte Suprema del Peru', in Ambos and Meini (eds.), supra note 27, at 100-1.
31 See Katanga and Ngudjolo DCC, para. 511: 'The organization must be composed of sufficient subordinates to guarantee that superiors' orders will be carried out, if not by one subordinate, then by another. The subordinated are used as "a mere gear in a giant machine'".
32 See Roxin, Täterschaft, supra note 12.
33 See, e.g., the Fujimori Judgment, supra, note 30. On the contrary, the German Federal Supreme Court has not required this element, see G. Werle and B. Burghardt, 'The German Federal Supreme Court (Bundesgerichtshof BGH) on Indirect Perpetration, Introductory Note', (2011) 9 JICJ, at 210. Latin-American jurisprudence does not show a consistent position on the need for this element (see Maculan, supra note 9).
34 See Section 2.2., infra.
35 See Roxin, Täterschaft, supranote 12, at 292; M. Gutierrez Rodriguez, La responsabilidad penal del coautor (2001), at 369, with further citations. Accordingly, co-perpetration in England and Wales is limited to cases where 'both (or more) of the accused act together to produce the actus reus jointly', see R. Cryer, 'Imputation and Complicity in Common Law States', (2014) 12 JICJ 267, at 272-3.The Chamber itself quotes authors in favour and against this idea. However, it does not explain why one solution should be preferred to another.
ALICIA GIL GIL AND ELENA MACULAN
These examples seem to be a manifestation of what Cassese has aptly defined as 'approche sauvagd}6 They are based on a selective and partial use of external references that does not take into account either their representative value or their specific elements.
In our view, when it is not possible to find in comparative law any common principles that offer a satisfactory response to issues raised by the prosecution of international crimes, it is not justified to advocate a particular solution arguing that it has reached the status of a quasi-general principle of law This approach may be described as 'a dubious exercise of jurisprudential comparison'.37 This remark has also been made by Judge Fulford and Judge Van den Wyngaert in their separate opinions on the Lubanga and the Ngudjolo Chui judgments, respectively.38
It is advisable for the Court to apply the ordinary-meaning and systematic methods for interpretation. If the Court applies a theory developed in a domestic system (such as Roxin's doctrine), it should do so because the theory most closely represents Article 2 5(3).39 Such a choice would be an attempt to build an international Dogmatik, but it should not claim to give the theory a spurious status of a general principle of law.
2.2. Expansive interpretation of principal liability
Besides the methodological flaws, the ICC's interpretation of the modes of liability up to the Lubanga trial udgment raises questions as to the concept of perpetration as such. The jurisprudence demonstrates a trend towards the widening of the scope of principal liability, as a result of a broad reading of some of its requirements, namely, the 'essential contribution' to the offence and the 'common plan or agreement'. In our view, this infringes the principles of strict construction and in dubio pro reo under Article 2 2(2). A similar warning was made, as we shall see, in the Katanga judgment.40
36 A. Cassese, 'L'influence de la CEDH sur 1'activite des Tribunaux penaux internationaux', in A. Cassese and M. Delmas-Marty (eds.), Crimes internationaux et juridictions Internationales (2002), at 143. The author contrasts it with the 'approche sage' - the approach based on a rigorous legal approach, which recognizes the predominance of the sources of international criminal law and uses the jurisprudence of other bodies and courts as a simple 'auxiliary means of determination'.
37 This observation is made by Werle and Burghardt, supra note 4, at 201, referring to the efforts of the ICC to assert the international recognition of indirect perpetration by virtue of the control over an organized power apparatus.
38 See Separate Opinion Fulford, para. 10, and Cone. Opinion Van den Wyngaert, para. 5, the latter stating that 'this direct import from the German legal system is problematic. Considering its universalist mission, the Court should refrain from relying on particular national models, however sophisticated they may be'. Both judges advocate, instead, for a plain reading of Art. 25(3), although with somewhat different proposals.
39 This is what the Appeals Chamber claims to have done in the recent judgment on appeals: Prosecutor v. Thomas Lubanga Dyilo, Judgment on the appeal of Mr. Thomas Lubanga Dyilo against his conviction, ICC-01/04-01/06 A 5, A. Ch., 1 December 2014 [hereinafter Lubanga Appeal Judgment]. In assessing the argument made by the defendant, that the control over the crime approach followed by the Trial Chamber 'was first developed in domestic legal doctrine, which is, as such, not applicable at the Court', the Appeals Chamber states 'that it is not proposing to apply a particular legal doctrine or theory as a source of law. Rather, it is interpreting and applying Article 25 (3) (a) of the Statute. In so doing, the Appeals Chamber considers it appropriate to seek guidance from approaches developed in other jurisdictions in order to reach a coherent and persuasive interpretation of the Court's legal texts' (ibid., para. 470).
40 See Section 3.2, infra.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 357
2.2.1. Interpretation of'essential contribution'
In order to identify which contributions may amount to 'control over the crime', as a ground for both co-perpetration and indirect perpetration, the Court in the Lubanga case has opted for the criterion of 'essential contribution'.41
This ambiguous concept has been defined as an element 'whose absence would have frustrated the common plan to commit the crime'.42 However, this interpretation implies a merely 'negative' control over the crime,43 meaning that the person cannot determine whether the crime will be committed but only whether it will not. Such control is not unique for the perpetrator but is also shared by any participant.44
Additionally, in identifying the elements that allow the determination that a contribution has indeed been essential, the Court has not always focused, as it should, on the causal and mental link between the defendant's conduct and the offence. The Lubanga trial judgment has instead inferred the essentiality of the contribution from a number of indications of the defendant's highest-ranking position in the organizational hierarchy.45
Such construction entails the risk of holding the defendant accountable solely based on his/her position in the hierarchy and on the dereliction of duties inherent in this status.46 Nevertheless, these forms of responsibility do not fall within the scope of Article 25(3), which has clearly opted for grounding principal liability in the defendant's contribution to performing the offence by the use of verbs like commit, order, solicit, induce, and assist. Therefore, this interpretation arguably amounts to a violation of the principle of legality enshrined in Article 22 of the Statute.
It is true that the senior position held by the defendant may give a special significance to his/her contribution to the offence. Yet, in our view, this element should be assessed in sentencing, and not as the sole ground for holding him/her liable.
What is more, the ICC has maintained that the relevant 'essential contribution' does not need to be performed at the execution stage of the offence.47 It has
41 See: Lubanga DCC, para. 343; Lubanga Judgment, para. 999. The criterion has also been upheld in the recent Lubanga Appeal Judgment, para. 469. Similarly, J. D. Ohlin, 'Searching for the Hinterman', (2014) 12 JICJ 329. The other requirements for co-perpetration are the existence of an agreement or common plan and the mens rea regarding the implementation of the plan.
42 Lubanga Judgment, para 347; in similar terms, Lubanga Appeal Judgment, para. 473. See also: Prosecutor v. William Samoei Ruto, Henry Kiprono Kosgey, and Joshua Arap Sang, Decision on the Confirmation of Charges Pursuant to Article 6i(7)(a) and (b) of the Rome Statute, ICC-01/09-01/11-373, P.-T.Ch II, 23 January 2012 , para. 308 [hereinafter Ruto, Kosgei and Sang DCC].
43 J. D. Ohlin, E. van Sliedregt, and T. Weigend, Assessing the Control-Theory', (2013) 26 LJIL 725, at 727.
44 Gutierrez, supra note 35, at 414. See also S. Wirth, 'Co-Perpetration in the Lubanga Trial Judgment', JICJ, 10 (2012), 971, at 987, arguing that the test is unhelpful whether it refers to the offence in concreto or in abstracto.
45 Lubanga Judgment, para. 1269. See, in detail, A. Gil, 'Responsabilidad penal individual en la sentencia Lubanga. Coautoria', in K. Ambos and E. Malarino (eds.), El caso Lubanga. Un andlisis de la primera sentencia de la Corte Penal International {201^), 263, at 280. Ohlin et al., supra note 43, at 732, after having criticized the ambiguous 'essential contribution' criterion, suggests to look for a 'cluster of factors' which indicate a 'central' role in a criminal enterprise, and which include both objective and subjective factors.
46 See also K. Ambos, Internationales Strafrecht(20ii), at 159. For a criticism, see E. Fernandez Ibanez, 'iConstituye lafungibilidaddelejecutorinmediatounpresupuesto estructural imprescindible?', (2005) 1 RevistadeDerecho penal 337, at 361.
47 'The Appeals Chamber considers that the most appropriate tool for conducting such an assessment is an evaluation of whether the accused had control over the crime, by virtue of his or her essential contribution to it and the resulting power to frustrate its commission, even if that essential contribution was not made at the execution stage of the crime' (Lubanga Appeal Judgment, para. 473, emphasis added).
ALICIA GIL GIL AND ELENA MACULAN
consequently allowed the charging of an individual as a co-perpetrator even if that individual performs no acts after the preparatory stage. The LubangaDCC asserts: 'although some authors have linked the essential character of a task - and hence ability to exercise joint control over the crime - to its performance at the execution stage of the crime, the Statute does not contain any such restriction'.48 The decision quotes some German and Spanish scholars who have developed the criterion of 'essential contribution' as a ground for the control over the crime. Yet, the Chamber does not fully respect these authors' views, in so far as it does not follow their claim that a contribution may be deemed as essential only if it was made, or at least reinforced, at the execution stage of the offence.49
This limitation is grounded in the consideration that whoever acts only in a preparatory stage does not have control over the crime, since the final decision as to whether and how to commit the offence is left to direct perpetrators.50 For a person to be charged as a co-perpetrator, therefore, he/she has to make a contribution at the execution stage. Additionally, this contribution has to be considered as essential in light of the common plan to commit the crime, that is, from an ex ante perspective.51
It is true that some scholars recognize that when dealing with organized criminal structures, the leaders may be labelled as perpetrators even if they perform acts only in a preparatory stage.52 However, these proposals still represent a minority view,53 and the Lubanga DCC does not explain why it considers this interpretation more convincing than the opposite (and more restrictive) one. The decision simply states that the Statute contains no restrictions on this point.
In relation to this rather weak argument, we may recall that the Statute contains no reference at all to the 'essential contribution' criterion. The literal meaning of the statutory provision is relied upon by both Judge Fulford and Judge Van den Wyngaert, who reject the applicability of the 'essential contribution' criterion.54
In addition, in so far as the Statute allows for a substantial judicial discretion at sentencing, labelling a person who performs an act only in a preparatory stage of the
48 Lubanga DCC, para. 3 48.
49 The decision (ibid., footnote 425) quotes Roxin, Täterschaft und Tatherrschaft (2000), at 294, 299; Mir Puig, Derecho penal Parte general (2000), at 385; Herzberger, Täterschaft und Teilnahme (1977), at 65 et seq.; Köhler, Strafrecht Allgemeiner Teil(iggj), at 518.
50 Roxin, Täterschaft, supra note 12, at 292; Gutierrez, supranote 35, at 369.
51 See also for further references: Gutierrez, supra note 35, at 392.
52 See, e.g., E Munoz Conde, 'iCdmo imputar a titulo de autores a las personas que, sin realizar acciones ejecutivas deciden la realization de un delito en el ämbito de la delincuencia organizada y empresarial?', (2001) Modernas tendencias en la Ciencia del Derecho penal yen la Criminologia, 512 (quoted in the decision under discussion).
53 This is the conclusion reached, after a thorough analysis of different positions, by E. Fernandez Ibänez, La autoria mediata en aparatos organizados depoder(2006), at 283.
54 Separate Opinion Fulford, para. 12, and Cone. Opinion Van den Wyngaert, paras. 41-2. Judge Fulford instead adopts the test of a causal, 'operational' link between the individual's contribution and the crime (ibid., para. 15), whereas Judge Van den Wyngaert proposes the criterion of direct contribution for a person to be labelled as a (co-)perpetrator, i.e. that the contribution has 'an immediate impact on the way in which the material elements of the crimes are realized' (ibid., para. 46) regardless of the physical presence of the joint perpetrator on the scene of the crime.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 359
offence as an accomplice by no means implies the imposition of a lower sentence. The approach of ascribing control over the crime even to those who perform no acts after the preparatory stage emerged in Germany precisely to overcome the lack of a legal provision within that system allowing the punishment of an accomplice who made a very significant contribution with the same penalty as that imposed on the perpetrator.55 Such a provision exists in Spain and in most Latin-American legal systems, which envisage a figure called cooperadornecesario: 'those that provide a contribution to the execution of the crime without which it would not have been committed'.56 An accomplice rather than a perpetrator, the cooperadornecesario does not undertake acts in the execution of the offence. They only make a contribution, whether at the executive or preparatory stage of the offence. Yet, the importance of their contribution allows the imposition of the same sanction as that imposed on the perpetrator.
In our view, this legislative choice allows the matching of a sanction to the participant's contribution while retaining the concept of perpetrator for non-accessory interventions.57 Common law systems, on the other hand, envisage different modes of participation in a crime, although the maximum sentence for accessory liability is generally the same as that for principals.58
The same option might be followed, in our opinion, in the ICC system, in which a participant may receive virtually the same penalty as the perpetrator.59 Although Article 25(3) of the ICC Statute does not encompass a mode of liability similar to the cooperador necesario, such a form of liability may fall within the scope of Article 25(3)(b) or (c). Therefore, there is no need for enlarging the scope of perpetration to cover an essential contribution made at the preparatory stage. The person who performs his/her act at that moment may be labelled as an accessory and still receive the same sentence as the perpetrator.60
2.2.2. Broad definition of the common plan or agreement
Another arena for this expansive interpretation of the modes of principal liability is the concept of common plan or agreement, which is also a requirement for co-perpetration. Although it has shown some inconsistencies in this respect,61 the
55 The same remark is made in Separate Opinion Fulford, para. 11.
56 This is the definition reflected in Art. 28 of the Spanish criminal code.
57 See M. Diaz y Garcia Conlledo, La autoria en Derecho penal (1991), at 677, 689. For a view on the discussion on the German and Spanish literature on this question, see Ambos, supra note 16, at 562; V. Garcia Del Blanco, La coautoria en Derecho penal (2006).
58 See H. Vest, 'Problems of Participation. Unitarian, Differentiated Approach, or Something Else?', (2014) 12 JICJ 295, at 301, with further citations, and Cryer, supra note 35.
59 Rule i45(i)(c) of the ICC Rules of Procedure and Evidence ('Determination of the sentence') provides that the Court should consider every relevant factor, inter alia, 'the degree of participation of the convicted person'. On this point, see Vest, supra note 58, at 300, 303.
60 See further section 2.3. infra.
61 For instance, in the Ruto, Kosgey and Sang DCC, para. 301, the Court required the inclusion of the crime in the common plan: 'The first objective element for indirect co-perpetration is the existence of a common agreement or a plan among those who fulfil the elements of the crime through another person. As established in the jurisprudence of the Court, the agreement or plan must include an element of criminality, meaning that it must involve the commission of a crime with which the suspect is charged (emphasis added).
ALICIA GIL GIL AND ELENA MACULAN
Court has followed a broad reading of this element, by stating that the plan in itself does not need to be criminal.62
Accordingly, the ICC deems it sufficient that the common plan implies an 'element of criminality'. This requirement would be met, as the Lubanga trial judgment affirms, whenever 'its implementation embodied a sufficient risk that, if events follow the ordinary course, a crime will be committed'.63
Such a broad understanding of the 'element of criminality' is hardly compatible with the rule of strict interpretation, because, if the plan does not envisage the commission of a concrete crime, 'there is nothing (agreed) what (sic) could be mutually attributed'.64 Furthermore, this interpretation infringes on the principle of individual criminal responsibility in so far as it implies the possibility of charging the defendant with crimes that are committed in excess of the common plan. When one of the parties commits a crime not initially included, even implicitly, in the common plan, the adopted broad understanding of the common plan allows the extension of responsibility for that excess to all those who agreed to the plan, provided that they foresaw the (excess) crime as a possible occurrence in the ordinary course of events. Thus, Katanga and Ngudjolo Chui were initially charged as co-perpetrators for the alleged crimes of rape and sexual slavery committed by the militia members during the attack against the village of Bogoro, on the ground of their alleged knowledge that those crimes would occur in the ordinary course of the events as a consequence of the common plan shared by them.65 This element of knowledge is not sufficient in itself to charge the defendants with the crimes committed during the attack. The Pre-Trial Chamber should have also identified the respective contribution of each of them to the alleged rapes or acts of sexual slavery committed by their subordinates.
The construction adopted by the Chamber makes the discharge of the burden of proof an easier task for the prosecution, since it is not necessary for it to prove that the crime with which the defendant is charged was part of the agreement. But this solution effectively removes the common intention to commit the crime as an independent requirement, which is both the foundation and the limit for mutual imputation in co-perpetration.
62 Lubanga Judgment, para. 984; Katanga and Ngudjolo DCC, paras. 566-7 and 569.
63 LubangaTnal Judgment, para. 984.
64 K. Ambos, 'The First Judgment of the International Criminal Court (Prosecutor v. Lubanga): A Comprehensive Analysis of the Legal Issues', (2012) 12 ICLR 115, at 140. A similar position can be found in Cone. Opinion Van den Wyngaert, para. 31: 'By focusing on the realization of a common plan, the mens rea and actus reus requirements are now linked to the common plan instead of to the conduct of the actual physical perpetrators of the crime ... the Statute does not contain a form of criminal responsibility that is based on the mere acceptance of a risk that a crime might occur as the consequence of personal or collective conduct (emphasis added).
65 See Katanga and Ngudjolo DCC, paras. 567-9. Judge Usacka, in her Dissenting Opinion to this Decision, found that the prosecution failed to present sufficient evidence to establish that the suspects were really aware that their conduct would certainly result in the commission of those crimes; see Partly Dissenting Opinion of Judge Anita Usacka, paras. 19-23. Both defendants were finally acquitted of these charges. Prosecutor v. Francis Kirimi Muthaura, Uhuru MuigaiKenyattaand Mohammed Hussein A/;, Decision on the Confirmation of Charges Pursuant to Article 6i(7)(a) and (b) of the Rome Statute, ICC-oi/o9-02/n-382-Red, P.-T.Ch II, 23 January 2012, para. 415, found that 'Muthaura and Mr. Kenyatta knew that rape was a virtually certain consequence of the implementation of the common plan', thereby labelling them as co-perpetrators of that offence.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR'
The simple awareness that the plan may lead to the commission of a crime in the ordinary course of events amounts to a lower threshold of mens rea than that required by the Statute. Article 3o(2)(a) defines a person as having intent if 'in relation to conduct, that person means to engage in the conduct'. By following this interpretive path, the ICC considers it sufficient instead that the conduct in which the other participant has engaged, resulting in excess crimes, had been foreseen, and not necessarily intended, by the defendant.66
This construction takes us back to the controversial JCE III.67 Arguably, it amounts to a violation of the principle of culpability. In addition, it is out of step with the theory of control over the crime, since the party who acts within the scope of the agreement cannot have joint control over an excess offence individually perpetrated by another.68 In a similar vein, in situations where indirect perpetration could be applied, the person who masterminded the crimes and had them perpetrated through another individual cannot be held responsible for any excess autonomously committed by the latter.69
This interpretation also goes against the principle of proportionality: it equates the joint commission of the crime with the mere creation of the risk of someone else taking one's contribution in order to commit a crime that has not been agreed.70 The former conduct is obviously more morally reprehensible and blameworthy than the latter whereby the defendant's action creates the risk that another person would take advantage of the situation to commit a crime that has not been agreed to. To prevent these problematic outcomes, co-perpetration liability should cover only the offences that formed part (even if tacitly or implicitly) of the common plan or agreement. Similarly, in order to charge a defendant as an indirect perpetrator, the main party must have given instructions to commit the crime.71
Accordingly, it would suffice that an agreement is reached on the 'extermination of all terrorists', to understand that the common plan includes the individual deaths of all of those who have been classified as such by the leaders. However, as held in the Katanga judgment, overturning the approach followed in the DCC in the same case, formulating the plan to defend or attack a territory is not per se sufficient to impose criminal liability for pillage or rapes committed by soldiers during the attack. Such conduct cannot be deemed as implicit in the agreement or in the order, even if such acts occur frequently or are foreseeable in these situations.
Finally, the interpretation based on a broad understanding of the common plan or agreement purported by the ICC, falls short of complying with the requirement of an objective contribution to the specific crime. As long as all of the members of the
66 For more detail about this problem, see A. Gil, 'Mens Rea in Co-Perpetration and Indirect Perpetration according to Article 30 of the Rome Statute. Arguments against Punishment for Excesses Committed by the Agent or the Co-perpetrator', (2014) 14ICLR 82.
67 The same view is held by Ohlin et al., supra note 43, at 17-8.
68 In the same sense, Ambos, supra note 16, at 558, with further quotations.
69 See Roxin, Täterschaft, supra note 12, § 24, at 244 et seq.; Gil, supra note 66, at 109 et seq.
70 See Gil, supra note 66, at 90. The difference between the two conducts is pointed out also by Ohlin, supra note 8, at 748.
71 See Gil, supra note 66, at 109 et seq.
ALICIA GIL GIL AND ELENA MACULAN
common plan may be held responsible as co-perpetrators for any excesses committed by one of them, all may be charged simply on the basis of a mere contribution to the common plan, which might in itself not be criminal.72 As the majority of the Trial Chamber ruled in the Lubanga judgment, 'the commission of a crime jointly with another person involves two objective requirements:... [one being] that the accused provided an essential contribution to the common plan that resulted in the commission of the relevant crime'.73
This surreptitious change fails to meet the requirement of an objective contribution to the commission of a specific crime, which is essential for charging any individual with that crime. As Judge Van den Wyngaert correctly asserts:
What troubles me in the Pre-Trial Chamber's interpretation is that ... the focus of attention has shifted away from how the conduct of the accused is related to the commission of a crime to what role he/she played in the execution of the common plan. Indeed, under the Pre-Trial Chamber's interpretation, it suffices for an accused to make a contribution to the realization of the common plan, even if this contribution has no direct impact on the coming into being of the material elements of a crime.74
This outcome breaches the principle of individual criminal responsibility. It does not comply with Article 30(1) of the Statute either, since it does not ensure the sufficient linkage of the mens rea to the carrying out of the material elements of a crime. Because the mental element is linked instead to 'a contribution towards a broadly defined common plan, as the control theory does, ... the connection to the crime might be almost entirely lost'.75
2.3. The common underpinning of these trends: a hierarchy of blameworthiness
All the interpretive trends just described are underpinned by the Court's aim of charging the defendant on the basis of principal liability.76 The rationale for this approach is not the necessity of imposing a more severe penalty, since the ICC Statute does not associate different sentences with the conviction as a 'principal' or 'accessory'. This suggests that 'while the modes of participation in Article 25(3) ICC Statute reflect a differentiation model, the statutory provisions on sentencing provide for a unitary range of punishment'.77 Furthermore, the sentencing practice
72 Lubanga Judgment, paras. 1221-2,1267,1270.
73 Ibid., para. 1006 (emphasis added).
74 Cone. Opinion Van den Wyngaert, para. 34.
75 Ibid., para. 35.
76 Recently the ICC Chambers have applied other modes of liability provided for in the Statute. In the Bemba case, the initial charge as a perpetrator was later modified into superior responsibility under Art. 28. In the Ruto and Sang case, the ICC charged the second defendant as a participant based on Art. 2 s(3)(d) (Ruto, Kosgey and Sang DCC, paras. 3 50 et seq.). The same mode of liability is applied in Prosecutor v. Ahmad Muhammad Harun and Ali Muhammad Al Abd-Al-Rahman (Au Kushayb'), Warrants of Arrest against Ahmad Harun, ICC-02/05-01/07-2 (para. 12), and in Warrant of Arrest for Ali Kushayb, ICC-02/o5-oi/o7-3-Corr, both by P.-T.Ch. I, 27 April 2007 (para. 13), together with accessory liability under Art. 25(3)^), in the first case, and with liability as a perpetrator, in the second case.
77 Vest, supra note 58, at 307. Van Sliedregt also points out that this feature contributes to the fading of the distinction between unitary and differentiated systems of criminal participation: Van Sliedregt, supra note 17, at 73-4 and 85.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR'
of international and mixed criminal tribunals shows that other participants have sometimes been imposed harsher penalties than perpetrators, and vice versa.78
Instead, the actual reason for which judges tend to bolster the modes of principal liability appears to be the stigmatizing message that the Court aims to send to the international community as a whole, by virtue of the symbolic and expressive function of its decisions.79
At the root of this approach is the assumption that principal liability is by definition more blameworthy than any forms of accessory liability, and that Article 25(3) establishes a hierarchy of blameworthiness.80 Accordingly, the Lubanga trial judgment argues that the notion of principal liability has the 'capacity to express the blameworthiness of those persons who are the most responsible for the most serious crimes of international concern'.81
Nonetheless, in the previous paragraph of the same judgment the Court apparently focused on a different criterion, rightly asserting that: As such, secondary liability is dependent on whether the perpetrator acts. Conversely, principal liability, which is closer to the violation of the legal interests protected by the norm, is not the subject of such dependence'.82
This statement explains, in our view, the most appropriate criterion for distinguishing between these categories of liability, i.e. the autonomous or derivative nature of responsibility. Thus, the contribution made by the principals is in itself criminally relevant, whether as a committed crime or as an attempt thereof. By contrast, accomplice liability is accessorial, meaning that there must be a perpetrator who commits or at least attempts the commission of the offence for the contribution of the accomplice to be punishable.83
It is therefore contradictory that the Court, while endorsing this notion of a derivative nature of accessorial liability, allows charging individuals as co-perpetrators even if their contribution is made at the preparatory stage only.84 The definition of
78 For example, the former President of Liberia, Charles Taylor, was convicted by the Special Court for Sierra Leone (SCSL) for both aiding and abetting and for planning the commission of a number of war crimes and crimes against humanity. Yet, despite having labelled him as an accomplice, the SCSL imposed upon him a sentence of imprisonment for a term of 50 years, which is much harsher than the sentence imposed by the ICC on Lubanga for co-perpetration. See SCSL, Prosecutor v. Charles Taylor, Judgment, SCSL-03-01-T, T. Ch. II, 18 May 2012, and Sentencing Judgment, SCSL-03-01-T, 30 May 2012. The sentence was upheld by the Appeals Chamber (Judgment, SCSL-03-01-A, A. Ch, 26 September 2013). Similarly, an interesting study on ICTY sentencing practice shows that the higher penalties were imposed on defendants who were convicted for planning or instigating the crimes, which is accomplice liability: see B. Hola, A. Smeulers, C. Bijleveld, 'Is ICTY Sentencing Predictable? An Empirical Analysis of ICTY Sentencing Practice', (2 009) 2 2 LJIL 7 9, at 91.
79 The same view is held by E. van Sliedregt, 'The Curious Case of International Criminal Liability', (2012) 10 JICJ 1172, at 1184-5.
80 However, both Separate Opinion Fulford (paras. 6-9) and Cone. Opinion Van den Wyngaert (paras. 22-4) expressly reject the idea that Art. 25(3) sets a hierarchy of blameworthiness. For an overview of this debate, see Ohlin et al., supra note 43, at 740-5.
81 LubangaTnal Judgment, para. 999. The recent judgment on appeals in the same case has confirmed this view: 'a person who is found to commit a crime him- or herself bears more blameworthiness than a person who contributes to the crime of another person or persons' (Lubanga Appeal Judgment, para. 462).
82 Ibid., para. 998.
83 This criterion has been applied in the Katanga Judgment: see section 3.1., infra. In agreement with this view, see Van Sliedregt, supra note 25, at 21.
84 Scholars have expressed diverging opinions as to what marks the beginning of an attempt in cases of indirect perpetration. For an overview of these different proposals, see C. Roxin, 'Der Anfang des beendeten Versuchs.
ALICIA GIL GIL AND ELENA MACULAN
'attempt' under Article 25(3X1) requires an action 'that commences the execution' of the crime.85 A person who acts only at the preparatory stage does not commence the execution of the crime; therefore, he/she does not reach the statutory threshold of the attempt. Since the perpetrator is the one who commits or at least attempts the crime, a person who makes a contribution only at the preparatory stage cannot be labelled as perpetrator. In order for this contribution to be punishable, someone else has, at least, to commence the execution of the crime. A contribution made at the preparatory stage is therefore accessory in nature. Accordingly, it should only be charged under accessory liability.
3. The Katanga judgment: A change of perspective
The Trial Chamber's judgment against Germain Katanga marks a significant shift in the interpretation of Article 25(3), although the DCC against him initially labelled him as an indirect co-perpetrator and reflected the main interpretative approach to that provision as described above.
Nevertheless, at the trial stage, the prosecution could not provide sufficient evidence to secure the defendant's responsibility under that label. As a consequence, although Trial Chamber II considered that the case against Ngudjolo Chui was ready for judgment, it decided to continue the trial against Katanga separately, subject to a possible change in the mode of liability under Regulation 55.86 It is noteworthy that Judge Van den Wyngaert opposed this decision by arguing that such re-characterization would go beyond the scope of the charges and would therefore infringe upon a number of fair trial standards.87 In the view of the majority, however, Katanga's responsibility would more appropriately fall within Article 25(3)^): participating in a crime committed by a group acting with a common purpose.
In the final judgment, Katanga was indeed convicted as an accessory under Article 2 5(3)(d) for the charges of murder as a crime against humanity and for the war crimes of wilful killing, directing an attack against a civilian population or against individual civilians, destroying property, and pillaging. The Chamber acquitted him of the charges under Article 25(3)^).
Zugleich ein Beitrag zur Abgrenzung von Vorbereitung und Versuch bei den unechten Unterlassungsdelikten', in Festschriftfiir Reinhart Mäurach zum 70. Geburtstag, Karlsruhe (1972), at 213-3. In our opinion, the indirect perpetrator performs the objective elements of the crime (execution stage) from the moment when he/she, by issuing an order, sets in motion a course of events that, from that moment, unfolds automatically, i.e. without further contributions on his/her part.
85 See Katanga and Ngudjolo DCC, para. 460: 'in order for an attempt to commit a crime to be punished, it is necessary to infer the intent to further an action that would cause the result intended by the perpetrator, and the commencement of the execution of the act (emphasis added).
86 Prosecutor v. Germain Katanga and Mathieu Ngudjolo Chui, Decision on the Implementation of Regulation 5 5 of the Regulations of the Court and Severing the Charges against the Accused Persons, ICC-01/04-01/07, 21 November 2012.
87 Minority Opinion of Judge Christine Van den Wyngaert, ICC-01/04-01/07-3436-A11XI, paras. 19, 43-5 and 58. The majority decision was nevertheless upheld by the Appeals Chamber, which confirmed (Judge Tarfusser dissenting) that Regulation 5 5 allowed a change in the applicable mode of liability from principal to accessory, but also requested the Trial Chamber to assess whether this re-characterization prejudiced the defendant's right to a fair trial: Prosecutor v. Germain Katanga, Judgment on the appeal of Mr. Germain Katanga against the decision of Trial Chamber II of 21 November 2012, ICC-01/04-01/07 OA 13, A. Ch., 27 March 2013.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 365
As regards the interpretation of the modes of liability, the Katanga judgment shows some continuity with the previous approach. First, the Chamber recalls that Article 2 5 separates principals (letter a) from accessories (letters b, c, and d), thereby implicitly rejecting the criticisms made by Judge Fulford in his dissenting opinion attached to the Lubanga judgment.88 Second, the judgment rejects both the objective and subjective concept of perpetrator, deeming those to be incompatible with the Statute.89 Third, the Chamber adheres to the theory of the control over the crime.90 Fourth, it relies on the doctrine of the Organisationsherrschaft developed by Roxin when defining indirect perpetration, although it does not apply it in entering the conviction.91
Notwithstanding these common points, the Katanga judgment adopts an interpretation of the modes of liability that is in some respects different from the previous one. In our view, its approach is more convincing.
3.1. The accessorial nature of the contribution as a distinctive criterion
The Katanga Trial Chamber first distances itself from previous case law when it rejects the idea that Article 25(3) establishes a hierarchy of blameworthiness. The judgment clearly asserts that 'the proposed distinction between the responsibility as a perpetrator or an accomplice does not constitute in any case a hierarchie de culpabilite (hierarchy of blame worthiness), nor does it set, even implicitly, a gradation of penalties'.92 The Chamber therefore adopts the opinion that we maintain: in the ICC framework, a conviction by way of principal liability does not need to be more serious per se than a conviction as an accomplice, since the Statute establishes no direct link between a specific mode of liability and the sentence to be imposed.
As stated earlier, the hierarchy of blameworthiness underlies all of the interpretive solutions aimed at giving the modes of principal liability the broadest possible scope. It is unsurprising, therefore, that the abandonment of this view by the Katanga Chamber entails its detachment from the interpretations leading to the expansion of the concept of perpetrator.
According to the Katanga Chamber, the sole criterion for distinction between principals and accessories in a crime lies in that the conduct performed by the former is constituent^ constitutif) for the commission of an offence, whereas the contribution of the latter is only accessorial ('en lien') to the commission of, or the attempt to commit, a crime by another individual.93 As the Chamber stated, '[a]n accomplice cannot be held accountable as long as another person does not commit or attempt to commit a crime over which the Court has jurisdiction'.94 Instead, the responsibility as a perpetrator 'depends on no other individual's intervention',95 since it per se amounts to an offence or to an attempt to commit an offence.
88 Katanga Judgment, paras. 1383 et seq.
89 Ibid., para. 1392.
90 Ibid., para. 1382.
91 Ibid., para. 1404 et seq.
92 Ibid., para. 1386 (authors' translation).
93 Ibid., para. 1384.
94 Ibid., para. 1385 (authors' translation).
95 Ibid.
366   ALICIA GIL GIL AND ELENA MACULAN
The Katanga judgment therefore confirms the criterion in favour of which we have argued, i.e. the accessory or autonomous nature of the contribution made by the defendant.96 Accordingly, the co-perpetrator is always involved in the material execution of a crime. Senior leaders, as long as they generally operate from a remote position, may be better labelled as indirect perpetrators, if the relevant requirements for this form of liability are met.
Otherwise, senior leaders may be charged as instigators or under other modes of participation envisaged in Article 2 5(3)(b)-(d). In such cases, in order to ensure the stigmatizing message of the conviction, we suggest that the judgments should label defendants as 'masterminds' or 'intellectual perpetrators' in the reasoning, although considering them as accessories from a technical point of view. Where proving the requirements necessary to establish these modes of liability is not possible, persons in authority may still be charged under superior responsibility for failing to act (Article 28).
3.2. A new methodological choice
When it comes to finding an operational criterion for delimiting this distinction between principals and accessories in specific cases, the majority of the Katanga Chamber applies the theory of the control over the crime. According to this interpretation, principals are all those who have control over the commission of the offence as well as the knowledge about the factual circumstances on which this control depends. Accessories to the crime, by contrast, can be distinguished from principals because, even if they hold a position of authority, the decision on the execution of the crime still lies with another individual.97
Yet, unlike in previous decisions where this theory was applied, the Katanga judgment does not try to legitimize it with reference to Article 2i(2)(c) (that is, by positing it a 'general principle of law'). According to the Katanga judgment, it is not necessary to prove that a criterion falls within that category for it to be applied. What matters for the Chamber is that 'the principle allows to draw the distinction between principals and accessories operational in practice'.98
By stating that, the Court rejects the notion that Article 25(3) has a gap that must be filled by having recourse to subsidiary legal sources. According to the judgment, control over the crime is a criterion that simply clarifies the meaning of the relevant statutory disposition, in accordance with the interpretive methods provided for in the Vienna Convention on the Law of Treaties.
It is noteworthy that, while applying the rules established by that Convention, the Chamber also points out the specificity of the ICC system.99 The latter, in fact, has recognized some principles that are fundamental to criminal law, and which have to be taken into account when interpreting the relevant provisions. The judgment
96 This criterion was recognized by the Lubanga Trial Judgment, although it also pointed out, albeit not clearly enough, the hierarchy of blameworthiness that exists between principals and accessories.
97 Katanga Judgment, para. 1396.
98 Ibid., para. 1395 (authors' translation).
99 Ibid., paras. 38 et seq.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR'
makes reference to the principle of legality and its corollary in dubio pro reo, both of which are envisaged in Article 22. These principles rule out interpretive methods, such as the teleological one, in so far as they may result in a reading that goes beyond the ordinary meaning of the statutory provisions and is to the detriment of the defendant.100
The methodological approach followed by the Katanga judgment appears more convincing than that unsuccessfully attempting to ground the control over the crime theory in a putative general principle of law.101 Furthermore, the reminder of the primacy of the principle of legality over teleological interpretations sets a limit on the expansive trend that was followed in the previous jurisprudence.
3.3. Refraining from expansive interpretations
As mentioned above, the Katanga judgment accepts both the theory of the control over the crime and its specific form of control over an organization, as elaborated by Roxin. In defining the constitutive elements of this form of indirect perpetration, the Chamber only focuses on two requirements: (i) the automatic functioning of the power apparatus enabled by the interchangeability of the potential direct perpetrators and (ii) the effective control the leader holds over this apparatus.102 The latter implies that the indirect perpetrator uses at least a part of the structure over which he/she has power in order to direct the subordinates towards the commission of a crime.103
In the view of the Chamber, the combination of these criteria allows an accused to be charged as a perpetrator only under the condition that he/she really has an effective control over the course of events leading to a specific crime with which he/she is charged.104 This implicitly overturns the previous understanding of indirect perpetration as encompassing every possible excess committed by a subordinate (or, in case of co-perpetration, by another member in the common plan). The Chamber requires instead a very close control over the commission of the crime, in terms of 'knowledge, supervision and planning at different levels, [and] control of its bringing about'.105
Furthermore, the Katanga judgment rejects the criterion of the essential contribution as a ground for co-perpetration. Its reasoning deals with the definition
100 Ibid., para. 55. Stahn points out that this principle is not really complied with by the Chamber itself, in that it later makes 'far-reaching interpretations' of the modes of liability, thereby following the tradition of 'progressive development of the law' which it itself criticizes. See C. Stahn, 'Justice Delivered or Justice Denied? The Legacy of the Katanga Judgment', (2014) 12 JICJ 809, at 816.
101 The opposite opinion is maintained by Stahn, who argues that it would have been preferable to make a more thorough comparative analysis to check the reception and potential criticism of Roxin's theory in domestic systems: Stahn, supra note 100, at 825.
102 Katanga Judgment, paras. 1408-12. In addition, the Chamber recalls the mens rea requirement, which under this mode of liability includes not only the mental element of the crime with which the defendant is charged, but also the knowledge of the circumstances that allow him/her to have control over the crime, i.e. his/her position in the apparatus and the automatic functioning thereof (paras. 1413-5). The Court does not engage in the analysis of the form of indirect co-perpetration.
103 Ibid., para. 1411.
104 Ibid., para. 1412.
105 Ibid, (authors' translation).
ALICIA GIL GIL AND ELENA MACULAN
of perpetration under Article 25(3)^) as a whole. In analysing that provision, the Chamber makes no reference to the criterion of 'essential contribution'. Nor does it do so when defining the theory of the control over the crime. It is true that the judgment does not specifically consider a possible charge as a (direct) co-perpetrator. In any event, as explained above, the notion of 'essential contribution' made at the preparatory stage of the offence as a ground for perpetration is incompatible with the criterion of non-accessorial contribution, which the Chamber itself applies as the distinctive criterion between principal and accessory liability Had the Chamber applied the criterion of 'essential contribution', it would have probably convicted Katanga as a co-perpetrator by virtue of an essential contribution, rather than - as it did - convicting him under Article 2 5(3)(d).
When applying this reasoning to the factual circumstances of the case, the Chamber concluded that the evidence collected in the trial fell short of showing the presence of these elements beyond any reasonable doubt.106 Not only did the prosecution fail to prove that the Ngiti militia that committed the offences was really an organized power apparatus, it was also unable to demonstrate that Katanga had an effective control over the militia members and their crimes.
Hence, the Chamber re-classified the mode of liability and, in line with the proposal previously made in its decision to sever Katanga's case, it applied the form of accessory responsibility envisaged in Article 25(3)^).
An extensive analysis of the reasoning the Chamber offered regarding this mode of liability goes beyond the scope of this article, which focuses on the interpretation of the forms of perpetration. But it is noteworthy that the Court affirmed that '[t]he crimes that are merely the result of the opportunistic conduct of the members of the group and that do not have any relevance in the common purpose cannot be ascribed to the concerted action of the group'.107 This statement confirms the view that a defendant who took part in a common plan108 cannot be charged with an offence committed by a participant of the common plan, which goes beyond what has been agreed to by other participants.
In addition, the Chamber made a remark about the requirement for a specific contribution to the offence as the basis for charging a defendant under Article 2 5(3)(d).109 If in order to label a defendant as an accessory under this provision the contribution has to refer to the commission of a crime and not only to the generic activities of the group, the same strict criterion should all the more apply to principal liability, in which the contribution is in itself legally relevant.
106 Ibid., para. 1420.
107 Ibid., para. 1630 (authors' translation).
108 Art. 2 5(3)(d) makes reference to the concept of 'common purpose', but the Trial Chamber recognizes the existence of a common plan as an indicator of that element, although it requires something more than a mere plan. The reasoning concerning this element of accessory liability (common purpose) can therefore be applied, mutatis mutandis, to co-perpetration based on the existence of a common plan, which simply sets a higher threshold.
109 See Katanga Judgment, para. 1632.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR'
In conclusion, the majority considered that the factual circumstances established in the case met all of the constitutive elements required by Article 25(3)^) and, therefore, declared the responsibility of Katanga as an accessory to the crimes.110
In her dissenting opinion, Judge Van den Wyngaert stated instead that Katanga's conduct lacked both a specific link to the commission of the offences and the required mens rea in terms of knowing that the contribution would cause the commission of the crimes in the ordinary course of events. Hence, she concluded that Katanga should have been acquitted.111
Such strong opposition not only to the label for the conviction, but also to the conviction itself inevitably lends the majority judgment a certain fragility.112 Furthermore, the permanent disagreement among the ICC judges as to the interpretation of Article 25(3) gives rise to confusion. That Trial Chamber II takes a different position on this provision in the Katanga judgment than Trial Chamber I did in the Lubanga judgment raises questions as to the consistency in the ICC jurisprudence.
Although the Statute does not encompass the rule of binding precedent as such, Article 21(2) recognizes that ICC precedents maybe applied in subsequent decisions. The two approaches compared in this article show a profound disagreement that still exists on the fundamental issue of the modes of liability before the ICC. The Chambers of the Court should endeavour to arrive at a common understanding of basic issues. The inconsistency in their legal findings seriously undermines the predictability of the ICC law.
4. Concluding remarks
The ICC interpretive approaches to the modes of principal liability have relied mostly on theories imported from the Continental legal systems: the theory of the control over the crime and the doctrine of control over an organization when it comes to indirect perpetration. Although resort to domestic sources of law has been grounded, at least prior to the Katanga judgment, in Article 2i(i)(c), the Court's analysis was often flawed from a methodological point of view.
Until the issuance of the Lubanga judgment, the Court relied upon a broad interpretation of the essential contribution requirement and the concept of a common plan as the constitutive elements of the theories it applied. This expansive trend, which assumes the notion that labelling an individual as a perpetrator reflects a greater degree of blameworthiness than an accessory liability charge, arguably infringes upon the principles of legality and individual criminal responsibility.
110 On the other hand, concerning the charge as a direct co-perpetrator of the war crime of using children under the age of fifteen years to actively participate in hostilities, the Chamber focused only on the performance of the objective elements of this crime by the defendant and concluded that there was no sufficient evidence thereof. Ibid., para. 1086 et seq.
in See Katanga Judgment, paras. 317-20. Judge Van den Wyngaert already reached this conclusion in her Minority Opinion appended to the decision severing charges, supra note 87.
112 See Stahn, supra note 100, at 815.
370   ALICIA GIL GIL AND ELENA MACULAN
The Katanga judgment partially overturns this approach in that it rejects the interpretation that the distinction between principals and accessories is indeed based on different degrees of blameworthiness. In contrast, it holds (and we agree) that the only valid criterion for distinguishing between perpetrators and accomplices is the autonomous or derivative character of their respective contributions to the crime. This means that a contribution by principals is in itself relevant from the standpoint of criminal law, that is, it amounts per se to a committed or attempted crime. By contrast, accomplice liability is accessorial, meaning that there must be a perpetrator who commits or at least attempts the offence for the contribution of the accomplice to be punishable.
This new approach coincides with a stricter reading of the constitutive elements of principal liability, in that a close and effective control over the commission of the crime is required before an individual can be charged as a perpetrator. This criterion helps avoid the undue expansion of this category of liability that has occurred in the Court's previous practice.
Besides, the Katanga judgment reflects a different interpretive method, which obviates the need for the theory that is subject to application to amount to a 'general principle of law'. This method focuses on the wording of the statutory provisions and on a systematic reading thereof. It simultaneously rejects (at least in principle) any teleological interpretation that may cause an expansion of criminal accountability contrary to the principles of strict construction and in dubio pro reo.113
The ICC system is far from evincing a shared understanding of Article 25(3) and the criteria for principal liability. However, in our opinion, the Katanga judgment offers an important step in the right direction. It is hoped that the future decisions by the ICC will follow this path and further clarify a number of issues.
In this respect, we would advise to limit the application of indirect perpetration through an organized power apparatus to those cases in which there is no doubt that the organization operates as an automatic mechanism pursuant to the orders of the superior.114 Second, the Court should refrain from charging as a co-perpetrator an individual who has only made an essential contribution at a preparatory stage of the offence but took no part in the commission of the crime. Last, we strongly recommend that a thorough assessment be made of the objective and subjective links between the contribution made by the defendant and the specific crime with which he/she is charged.
Therefore, senior leaders, who normally operate from a remote position, would be better charged as indirect perpetrators, provided that the elements of this form of liability are met. If none of the essential requirements for principal liability is met, or where there is insufficient evidence to establish such elements, Article 2 5(3)(b)-(d) of
113 In this respect, we cannot but agree with Judge Van den Wyngaert's statement that: 'Courts of criminal justice cannot claim to protect an accused's fundamental rights to a fair trial while making expansive interpretations of articles that define modes of liability'. Cone. Opinion Van den Wyngaert, para. 68.
114 Van Sliedregt argues that the previous Katanga and Ngudjolo DCC was instead 'blinded by the beauty of Dogmatikand lost sight of the African reality': van Sliedregt, supra note 25, at 23.
CURRENT TRENDS IN THE DEFINITION OF 'PERPETRATOR' 37I
the Rome Statute envisages a wide range of modes of accessory liability Liability for instigating or ordering the crimes may suit the typical position of senior leaders and their contribution.115 In order to keep the stigmatizing message of the conviction, the judgment could label them as 'masterminds' or 'intellectual perpetrators' but at the same time consider them as accessories from a technical point of view In other words, not all defendants must necessarily be considered perpetrators.
115 Additionally, Art. 2 8 of the Statute envisages a mode of liability based on a failure to act, which is intended for military and civilian superiors.
Copyright of Leiden Journal of International Law is the property of Cambridge University Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.