Followers of the International Criminal Court (ICC) interested in theories of criminal liability were looking forward to the Appeals Chamber judgment in the Ntaganda case, in which Bosco Ntaganda was convicted of war crimes and crimes against humanity in Democratic Republic of the Congo (DRC) based on an enigmatic theory of liability: indirect co-perpetration through an Organized Structure of Power (OSP). When the judgment was delivered on March 30, 2021, we were still puzzled over its validity; with two separate opinions and one partly concurring opinion regarding this concept, the theory remains contested.
On July 8, 2019, the Trial Chamber rendered its Conviction Decision, in which it found Ntaganda guilty of five counts of crimes against humanity (murder and attempted murder, rape, sexual slavery, persecution, forcible transfer and deportation) and thirteen counts of war crimes (murder and attempted murder, intentionally directing attacks against civilians, rape, sexual slavery, pillage, displacement of the civilian population, conscripting and enlisting, intentionally directing attacks against protected objects, and destroying the adversary’s property). Ntaganda was sentenced to 30 years imprisonment.
Ntaganda was found guilty as an indirect co-perpetrator for all crimes charged and as a direct perpetrator for one act of murder, constituting a crime against humanity and a war crime, as well as an underlying act of persecution as a crime against humanity. He was found liable for crimes committed by members of the Union des Patriotes Congolais (Union of Congolese Patriots, or UPC) and its military wing, the Forces Patriotiques pour la Libération du Congo (Patriotic Force for the Liberation of Congo, or FPLC) and Hema civilians, who executed a common plan he agreed upon with fellow co-perpetrators, an OSP he controlled.
The majority of the Appeals Chamber in Ntaganda agreed with and endorsed the Trial Chamber rulings on conviction and sentencing. The judges were divided, however, on the theory of liability of indirect co-perpetration.
Indirect Co-perpetration
Indirect co-perpetration combines two modes of liability codified in Art. 25(3)(a) of the Rome Statute of the ICC: joint or co-perpetration (“commits…jointly”) and indirect perpetration (“commits…through another person”). Article 25(3) states in full:
In accordance with this Statute, a person shall be criminally responsible and liable for punishment for a crime within the jurisdiction of the Court if that person:
(a) Commits such a crime, whether as an individual, jointly with another or through another person, regardless of whether that other person is criminally responsible.
(b) Orders, solicits or induces the commission of such a crime which in fact occurs or is attempted;
(c) For the purpose of facilitating the commission of such a crime, aids, abets or otherwise assists in its commission or its attempted commission, including providing the means for its commission;
(d) In any other way contributes to the commission or attempted commission of such a crime by a group of persons acting with a common purpose. Such contribution shall be intentional and shall either:
(i) Be made with the aim of furthering the criminal activity or criminal purpose of the group, where such activity or purpose involves the commission of a crime within the jurisdiction of the Court; or
(ii) Be made in the knowledge of the intention of the group to commit the crime;
Indirect co-perpetration was applied for the first time in the Katanga & Ngudjolo case. Both Katanga and Ngudjolo were accused of using rebel organizations at their disposal and only by combining their forces could they perpetrate the atrocities. Each defendant was not only responsible for the actions of their own troops but also for the actions of the other’s troops (para. 484). This “cross-liability” represented the full force of the indirect co-perpetration doctrine. Leaders become responsible not just for individuals under their command, but also for individuals that their collaborators command.
Indirect co-perpetration is a potent prosecutorial tool. As Ohlin, points out, it allows the conviction of defendants who are substantially removed from the physical perpetrator of the crime along two axes. Not only are such defendants vertically removed from the commission of the crime (by virtue of their indirect perpetration), but they are horizontally removed as well (by virtue of their co-perpetration with other collaborators on the horizontal level). The notion of “control” (discussed below) provides the connection that links such defendants, along the two axes, to the physical perpetrators of the atrocities.
Indirect co-perpetration was thought to represent the future of international criminal prosecutions before the ICC. It is similar to Joint Criminal Enterprise (JCE), which holds individuals liable for crimes committed through, or by way of, a criminal enterprise and which came to define the prosecutorial strategy regarding perpetration at the International Criminal Tribunal for the Former Yugoslavia (ICTY). Indirect co-perpetration already features in many ICC cases, including the Kenyatta and Muthaura, Gbagbo and Blé Goudé, Ongwen and Ntaganda cases.
To fully understand the different opinions in the Ntaganda judgment, we need to briefly reflect on the underlying “control theory.”
Control Theory
The Rome Statute stipulates in Art. 25(3)(a) that perpetrators are those who “[p]hysically carry out the objective elements of the offence, but also …who, in spite of being removed from the scene of the crimes, control or mastermind its commission because they decide whether and how the offence will be committed” (Lubanga, para. 330). They are liable as principals (Katanga & Ngudjolo, para. 498) because they control the commission of the offence and are aware of having such control. It follows, then, that those who control the crime or those committing crimes are at the apex of a hierarchy of blameworthiness; they are considered “most responsible” (hereinafter the “hierarchy thesis”). In contrast, those who order, facilitate, or act with a common purpose, criminalized under Art. 25 in sub-paragraphs (3)(b-d), are accessories. They contribute to crimes and, as such, are lower on the culpability ladder than perpetrators who commit crimes.
The hierarchy thesis can be best illustrated by the – alleged – distinction between ordering and indirect perpetration. Ordering in Art. 25(3)(b) criminalizes contributing to a crime by way or ordering it. It is accessorial liability and should be distinguished from indirect perpetration in Art. 25(3)(a) which is a form of perpetration/commission and concerns those who mastermind and commit crimes through others. It follows from the control theory and its hierarchy thesis that ordering in Art. 25(3)(b) captures liability of those in lower or middle rank positions where an order (to commit crimes) is given by someone other than a mastermind.
While some have welcomed the control theory and indirect co-perpetration as a good alternative to JCE, it has been criticized for its breadth, for being too complex, for being a concept developed by German theorist Claus Roxin, which has no place in international criminal law (ICL), and for not reflecting the reality on the ground. The most broadly voiced critique is that the text of the Rome Statute does not support the control theory. Judge Fulford, in the Lubanga case, disagreed with the distinction between perpetrators and accessories. In his view it was unnecessary and beyond the text of the provision. Similarly, Judge Van den Wyngaert, in her Concurring Opinion to the Ngudjolo case, writes that the control theory is not consistent with the ordinary meaning of Art. 25(3)(a) of the Rome Statute.
Particularly problematic is the import of the hierarchy thesis. There is no suggestion in the text of the Statute, nor in the travaux préparatoires, that perpetrating crimes under Art. 25(3)(a) is more blameworthy – and thus worthy of a more serious sentence – than contributing to crimes under Art. 25(3)(b-d). In other words, the provision itself, listing different ways in which one can engage in criminal activity, does not indicate that one form of criminal responsibility should attract a more serious sentence than another. For instance, there is no reason why indirect perpetrators under Art. 25(3)(a) should be more severely punished than those who order or instigate crimes under subparagraph (b). The only mode of liability that could probably be termed as lesser liability is aiding and abetting in article 25(3)(c). Building on case law at the ICTY, we can track a development in ICL where aiding and abetting crimes has developed into a form of liability that concerns those on the margins rather than at the center of the criminal endeavour.
The hierarchy thesis does not “fit” the sentencing regime of the ICC Statute. According to Articles 77 and 78 of the Rome Statute, the Court can impose any sentence taking into account the gravity of the crime and the individual circumstances of the convicted person. Rule 145 (1)(c) of the Rules of Procedure and Evidence stipulates that judges give consideration to the “degree of participation of the convicted person.” It is at the level of sentencing, not at the conviction level, that the degree of responsibility is expressed.
Those critical of the control theory and indirect co-perpetration have been vindicated. After a number of controversial procedural twists and turns, the Trial Chamber severed the cases of Katanga and Ngudjolo, recharacterized the charges against Katanga as common purpose liability under Art. 25(3)(d) and acquitted Ngudjolo. The takeaway from this case was that squeezing ambitious fact patterns into legal categories to ensure a serious sentence may backfire.
The Ntaganda Appellate Ruling
Three of the five judges in the Ntaganda appellate ruling wrote separate opinions on the issue of indirect co-perpetration. Judge Morrison and Judge Eboe-Osuji separately agree that indirect co-perpetration results from an erroneous interpretation of Art. 25(3)(a). Judge Morrison, in his opinion, recalls the dissents by Judge Fulford in Lubanga and Judge Van den Wyngaert in Ngudjolo, reiterating their view that the text of Art. 25(3) does not support the control theory nor the hierarchy thesis (para. 3). He proposes the ICC abandon indirect co-perpetration,
“if only because the theory’s complexity lends the theory a certain elasticity so that criminal responsibility may be broadly stretched over: (i) groups that are not as tightly organised as the hierarchical power structures envisaged by Roxin; and (ii) figures whose power to control the crimes committed through such groups is dubious” (para.17).
Judge Morrison does, however, endorse the conviction. He is convinced that “Mr Ntaganda was responsible for the crimes of which he was convicted and his sentence reflected the gravity of his conduct.” Moreover, “the jurisprudence was known to the appellant from the outset of the proceedings and therefore part of the case he knew he had, or might have had, to meet” (para. 2).
Judge Eboe-Osuji criticises importing the control theory from German criminal law theory into ICL. This was done in early ICC case law (Pre-Trial Chamber decisions in the Lubanga and Katanga & Ngudjolo cases). In recent years, a more “enlightened view” (para. 96-102) has taken hold. He refers to the Bemba et al ruling, where the hierarchy thesis has been abandoned (paras. 59-60). Judge Eboe-Osuji disagrees with the conviction of Ntaganda as an indirect co-perpetrator. But he does concur with the majority when it comes to convicting Ntaganda as a direct perpetrator.
Peruvian Judge Ibánez wrote a separate opinion supporting indirect co-perpetration, essentially repeating many of the points made in the early ICC case law to justify importation of indirect co-perpetration and the control theory. In her view, the control theory and indirect co-perpetration are the most appropriate way to capture the liability of those who mastermind international crimes.
It is interesting to reflect on the fault lines between Judges Ebeo-Osuji and Judge Morrison, on the one hand, and Judge Ibanez, on the other hand. The latter is familiar with the theory of control. Peruvian criminal law is heavily influenced by German law and has adopted Roxin’s theory of control distinguishing between perpetrators and accessories. In fact, it was at the basis of the conviction of former Peruvian leader Fujimori for crimes against humanity.
The other two judges originate from jurisdictions where no system of moral classification exists. Accessories or accomplices are not classified as “less liable” than perpetrators. They are participants whose liability derives from that of the principal who directly caused the actus reus. In the legal language of complicity theory this makes them “secondary participants.” In moral terms there is nothing secondary about them. Instigators are secondary participants but generally regarded as equally, or sometime more, culpable than the principals they induced or abetted. Moreover, and most importantly, secondary participants are punished as principals; for the crime itself. The weight of their role is considered separately in the sentencing phase of criminal proceedings.
As I have argued elsewhere, Art. 25(3) of the Rome Statute contains both approaches. The control theory, which is premised on the concept of a remote perpetrator, who as an indirect perpetrator uses culpable direct perpetrators as tools to commit crimes, is captured in “perpetration through another person” in Art. 25(3)(a). At the same time, Art. 25(3) comprises the classic complicity theory. The latter is relied upon by Judge Eboe-Osuji and Judge Morrison and is incorporated in (physical) direct perpetration or “commission” under Art. 25(3)(a) and as liability for contributing to crimes as secondary parties, accessories, or accomplices under Art. 25(3)(b-d).
The problem with the control theory as developed in early ICC case law is that it is applied by exclusion to other theories of liability. It is read into the whole of Art. 25(3). This is the fatal flaw of the ICC’s control theory. While it should be seen as a theory of perpetration that recognizes perpetration beyond direct/physical perpetration, the control theory has been made the theoretical grounding for the whole of Art. 25(3) rather than just 25(3)(a). As such it displaces classic complicity theory which is reflected in the wording of subparagraphs (b-d). The other problem of the control theory is that it reduces the modalities in subparagraphs (b-d) to lesser liability. Doing so does not comport with the text nor the drafting history of Article 25. As Per Saland, chairman of the Working Group on General Principles recalls, Art. 25(3) posed great difficulties to negotiate; eventually a near-consensus was reached where there would be one provision covering the responsibility of principals and all other modes of participation. It was to provide the court with a range of modalities from which to choose.
Where does this leave us? Indirect co-perpetration will remain under a cloud. I expect prosecutors to start looking at alternatives when charging suspects. Defence counsel will target this theory every step of the way.
The attraction of the control theory and indirect co-perpetration is its expressive value. Those who mastermind crimes are perpetrators. However, to quote Judge Morrison in his Separate Opinion, the theory has contributed more “to mislabelling facts and events than it does to fair labelling of an individual’s criminal responsibility” (para 39).