Lubanga Decision Roundtable: More on Co-Perpetration

I’ll have much to say about various legal aspects of the Lubanga judgment in the days to come, but I wanted to start by discussing the relatively narrow — though critically important — point that Jens addressed in his post: the dispute between the majority and Judge Fulford concerning the correct interpretation of co-perpetration in Article 25(3)(a) of the Rome Statute, the sole mode of participation at issue in the case.  I think Judge Fulford correctly rejects the majority’s interpretation, but I disagree with his reasoning.

At the outset, it should be noted that Article 25(3)(a) is maddeningly vague: it simply provides that a person shall be criminally responsible for an international crime if he “[c]ommits such a crime, whether as an individual, jointly with another or through another person, regardless of whether that other person is criminally responsible” (emphasis mine).  Lubanga was charged with responsibility for the conscription and enlistment of child soldiers by committing that crime “jointly with another” — co-perpetration.  But what qualifies as co-perpetration?

The obvious solution for the Trial Chamber would have been to hold that co-perpetration under Article 25(3)(a) is equivalent to JCE I at the ICTY, which is what the legal representatives of the victims advocated at the pre-trial stage.  Instead, the Trial Chamber followed the Pre-Trial Chamber (in Lubanga and in other cases) and held that co-perpetration differs from JCE I in that it requires the perpetrator to make an “essential” contribution to the common plan to commit a crime, whereas JCE I is satisfied as long as the perpetrator contributes in any way to the common plan:

1006. The Majority therefore concludes that the commission of a crime jointly with another person involves two objective requirements: (i) the existence of an agreement or common plan between two or more persons that, if implemented, will result in the commission of a crime; and (ii) that the accused provided an essential contribution to the common plan that resulted in the commission of the relevant crime.

As Judge Fulford (and Jens) notes, this “essential contribution” requirement reflects the “control theory of co-perpetration” pioneered in the 1960s by Claude Roxin, a leading German criminal-law theorist, and largely accepted by German criminal law.  Judge Fulford, however, rejects the control theory in his separate judgment.  First, he argues that the Court should not have uncritically accepted it, because it reflects a need to distinguish between principals and accessories for purposes of sentencing that is absent from the Rome Statute, which determines sentence more holistically:

10. The control of the crime theory has its origins in the post-war German legal system, where particular domestic considerations – which do not exist at the ICC – have made it appropriate to apply this principle. In adopting this theory, the Pre-Trial Chamber focused substantially on a minority view from the ad hoc tribunals, in that it cited the judgment or the ICTY Trial Chamber in the Stakic case when it held that the accused was responsible as a co-perpetrator (the conviction on this basis was set aside on appeal) and Judge Schomburg’s separate opinion in the ICTR Appeals Chamber’s judgment in the Gacumbitsi case. In these two instances, the judges relied heavily on the scholarship of the German academic Claus Roxin as the primary authority for the control theory of co-perpetration, and in the result, this approach was imported directly from the German legal system. While Article 21(l)(c) of the Rome Statute permits the Court to draw upon “general principles of law” derived from national legal systems, in my view before taking this step, a Chamber should  undertake a careful  assessment as to whether the policy considerations underlying the domestic legal doctrine are applicable at this Court, and it should investigate the doctrine’s compatibility with the Rome Statute framework. This applies regardless of whether the domestic and the ICC provisions mirror each other in their formulation. It would be dangerous to apply a national statutory interpretation simply because of similarities  of  language, given the overall  context is likely to significantly different.

11. This case demonstrates why a detailed assessment of this kind is necessary. Under the German legal system, the sentencing range is determined by the mode of liability under which an individual is convicted, and it is therefore necessary to draw clear distinctions between principals on the one hand and accessories on the other. As set out above, these considerations do not apply at the ICC, where sentencing is not restricted in this way, and this example of the differences that exist is of significance in this context.

Second, and perhaps logically prior, Judge Fulford believes that the control theory is inconsistent with the text of Article 25(3)(a):

15. When establishing joint perpetrator liability, the prosecution must prove that an individual committed the crime jointly with another. The latter words (“jointly with another”) clearly indicate the involvement of at least two people, whilst the expression “commits […] jointly” denotes coordination between the individuals involved. This self-evidently necessitates a sufficient meeting of minds, by way of an agreement, common plan or joint understanding….  Finally, the verb “commits” requires a contribution to the commission of the crime. Nothing in the Statute requires that the contribution must involve direct, physical participation at the execution stage of the crime, and, instead, an absent perpetrator may be involved. Either way, the use of the word “‘commits” simply requires an operative link between the individual’s contribution and the commission of the crime. Additionally as regards causation, the plain text of Article 25(3) does not require proof that the crime would not have been committed absent the accused’s involvement (viz. that his role was essential). Rather, the prosecution must simply demonstrate that the individual contributed to the crime by committing it with another or others.

Judge Fulford thus concludes that, correctly interpreted, acting “jointly with another” is equivalent to JCE I in terms of the actus reus of co-perpetration.  In particular, he believes that any contribution to carrying out a common plan suffices for criminal responsibility; an “essential contribution” is not required (para. 16).

I agree with Judge Fulford that the majority gets Article 25(3)(a) wrong.  But I disagree with his explanation of why Article 25(3)(a)’s “jointly with another” provision should be read to adopt JCE I’s contribution requirement instead of the control theory’s.  To begin with, I find his textual analysis unconvincing.  I agree that the “commits… jointly with another” language does not naturally support the majority’s “essential contribution” requirement — but I don’t think it supports his “any contribution” requirement, either.  At best, we can infer that co-perpetration requires a common plan, whether express or implied; that is the plain meaning of acting “jointly with another.”  Judge Fulford tries to infer the “any contribution” requirement from “commits,” but that interpretation is hardly self-evident.  “Commits” indicates that the three modes of participation in Article 25(3)(a) should be considered forms of principal liability; it does not specify the substantive content of any of those modes.

By contrast, a contextual interpretation of “jointly with another” does support Judge Fulford’s conclusion — but not his reasoning.  The majority adopts the “essential contribution” requirement because they believe it is necessary to distinguish between principals and accessories — they say (para. 999) that “lowering that threshold would deprive the notion of principal liability of its capacity to express the blameworthiness of those persons who are most responsible for the most serious crimes of international concern.”  Judge Fulford rejects that argument, insisting that there is no distinction between principals and accessories in Article 25(3) and that, as a result, there is no problem with different modes of participation overlapping.

I don’t think Judge Fulford’s position is correct.  Although his point (for example) about ordering not necessarily being less serious than acting jointly with another is well taken, that does not mean Article 25(3) does not distinguish between principals and accessories.  What it means is that the distinction itself is problematic in the context of international crimes, where the most responsible individual will often — perhaps normally — not be the individual who commits the actus reus of the offence.  I don’t know any scholar who does not think Article 25(3)(a) specifies forms of principal liability (“commits”) while Article 25(3)(b)-(d) specify forms of accomplice liability.  Moreover, the principal/accessory distinction is deeply entrenched in international jurisprudence dating back to Nuremberg.

That said, I’m also skeptical of the majority’s insistence that we need the “essential contribution” requirement to distinguish between co-perpetration as principal liability and forms of accomplice liability such as aiding and abetting (Article 25(3)(c)) and contributing to a group crime (Article 25(3)(d)).  It seems to me that the difference between principal and accomplice liability in that context is more obvious: co-perpetration requires an agreement to commit a crime — a common plan — whereas modes of participation such as instigation, aiding and abetting, and contributing to a group crime do not.  In light of that fundamental difference, it would in no way blur the lines between principal and accomplice liability to adopt the same contribution requirement for both.

My view, it should be noted, assumes that Article 25(3)(d) is limited to contributions to a group crime that are made by individuals who are not members of the group. If the provision applies to group members, as well, the majority’s argument makes more sense: because both co-perpetration and group contribution would then involve individuals who agree to commit a crime, it would be necessary to distinguish between them at the level of actus reus in terms of the required degree of contribution.

Scholars are divided concerning the correct interpretation of Article 25(3)(d).  Cassese, Eser, and Werle agree with me; Jens and both the Pre-Trial Chamber and the Trial Chamber do not, insisting that Article 25(3)(d) is functionally equivalent to JCE I.  This post is not the place to argue the issue at length, and readers interested in Jens’ position should read his excellent article in a special issue of New Criminal Law Review that I edited; you can find it on SSRN.  Suffice it to say here that I find it difficult to understand why, if Article 25(3)(d) includes situations in which an individual contributes to the commission of a crime he agreed to commit, the drafters would have either (1) created two provisions that covered contributions to a common plan without specifically distinguishing between them in terms of the contribution they require, or (2) placed a mode of participation that is considered a form of principal liability by every other international criminal tribunal and by the overwhelming majority of domestic legal systems after modes of participation (such as soliciting and aiding and abetting) that are clearly forms of accessorial liability.  Moreover, unlike Jens, I find revealing the fact that Article 25(3)(d) is borrowed almost verbatim from Article 2(3) of the International Convention on the Suppression of Terrorist Bombing, a provision that (as Jens acknowledges) clearly specifies forms of accomplice liability.

What, then, is the correct interpretation of Article 25(3)(a)’s “jointly with another” provision?  In my view, the legal advocates for the victims in Lubanga were right: co-perpetration in Article 25(3)(a) is essentially JCE I — the only difference being mens rea: JCE I requires the intent to commit the agreed-upon crime, whereas Article 30’s default mens rea means that co-perpetration is satisfied by intent or knowledge that the crime will be committed in the ordinary course of events.  Both the Pre-Trial Chamber and the Trial Chamber have always rejected that interpretation, but they have never articulated a convincing reason for doing so.  Neither a textual nor a contextual analysis supports the control theory, as Judge Fulford points out.  Yet I do not think it is methodologically adequate to argue, as Judge Fulford does, that Article 25(3)(a) embraces JCE I because “policy considerations” (para. 10) support such an interpretation. What about, for example, the intent of the drafters?  If the drafters of Article 25(3)(a) intended to base co-perpetration on the control theory instead of on JCE I, it would be inappropriate for the judges of the ICC to adopt JCE I simply because they considered it pragmatically superior.

Nothing in the travaux preparatoires, however, indicates that the drafters of Article 25(3)(a) intended to adopt the control theory.  Indeed, such an intent would be surprising, given how few domestic criminal-law systems predicate responsibility for a group crime on an individual making an essential contribution to carrying out the common plan.  To the best of my knowledge, no common-law system requires an essential contribution.  Moreover, German criminal law is itself far from uniform in its adoption of the control theory.  (Judge Fulford also notes in n. 20 that Roxin himself seems to have advocated a “substantial importance” contribution requirement, not an “essential contribution” requirement.)

In light of the absence of a clear textual, contextual, or preparatory indication that Article 25(3)(a) adopts the control theory, a coherent methodological approach to interpreting that provision — one based on Article 21 of the Rome Statute — would look not to policy considerations but to customary international law or “general principles of law derived by the Court from national laws of legal systems of the world.”  Such an analysis, I think it is safe to say, would support reading Article 25(3)(a) to embrace JCE I, not the control theory.  No analogue to the control theory exists in the jurisprudence of any previous international criminal tribunal, and the ICTY Appeals Chamber specifically rejected that theory in favor of JCE I in Stakic.  Moreover, unlike the control theory, some version of JCE I exists in nearly all domestic criminal-law systems.  On what basis, therefore, does the ICC continue to insist that Article 25(3)(a) is based on the control theory of co-perpetration?