1. Introduction
The UN ad hoc Tribunals and the International Criminal Court (ICC) have construed two distinct theories of joint perpetration – respectively, joint criminal enterprise (JCE) and joint control over the crime – which are nowadays cited as a paradigmatic example of fragmentation and pluralism in international criminal law.Footnote 1 Indeed, one can plausibly argue that no other topic in this field has polarized the commentariat as much as these two legal constructs. Traditionally, comparative research on them has focused on contrasting their hallmark requirements, i.e., the ‘shared intent’ element of the JCE theory and the ‘essential contribution’ element of the control theory, studying their merits and deficiencies, and ultimately endorsing one or the other criterion as a normatively better tool for distinguishing between co-perpetrators and accessories to a collective crime.Footnote 2 This juxtaposition of their underlying rationale has largely shaped our understanding of the difference between the above two theories and, respectively, of the schism in the international jurisprudence on co-perpetration responsibility.Footnote 3 This article, however, will shift the focus to another feature of JCE and the joint control theory, which has caused further fragmentation in the tribunals’ topical case law, yet has attracted considerably less attention. Specifically, the diverging scope and outer limits of the ‘common plan’ element in the law on co-perpetration based on JCE and joint control over the crime will be nuanced.
This article will first briefly define the framework of co-perpetration responsibility under both JCE and the joint control theory, in the course of which it will explain the doctrinal function of the ‘common plan’ element. This will then be used to introduce and examine two examples of further dissonance in the international jurisprudence on co-perpetration, namely: i) the courts and tribunals’ conflicting formulations of the ‘common plan’ element under JCE and the joint control theory; and ii) the resulting disparity in the attribution of liability for offences that fall outside the scope of the ‘common plan’ (i.e., ‘excess’ crimes). These matters have created a chasm between the case law of the ICC, the UN ad hoc Tribunals and two of the so-called hybrid tribunals. More specifically, while the International Criminal Tribunal for the former Yugoslavia (ICTY) and the International Criminal Tribunal for Rwanda (ICTR) have established that the JCE theory requires a common plan that is specifically directed at the commission of a certain crime, the ICC's joint control theory is built on a more diluted definition of this element which requires a common plan that just contains ‘an element of criminality’. Remarkably, although they formally apply the JCE theory, the Special Court for Sierra Leone (SCSL) and the Extraordinary Chambers in the Courts of Cambodia (ECCC) have come to adopt a definition of the ‘common plan’ element that bears a closer resemblance to the ICC's approach to this requirement. The material differences between these formulations will be analyzed, arguing that they have in fact resulted in the existence of not two, but three distinct constructions of co-perpetration liability in modern international criminal law.
Depending on how one defines the precise parameters of the ‘common plan’ requirement, our understanding of when a crime falls outside the scope of the common plan, and how it can be attributed to the confederates in the said agreement, would greatly vary. In international criminal law, this paradigm is normally associated with the so-called ‘extended’ variant of JCE (JCE III): a concept that has long raised doctrinal concerns among scholars, partly on account of its alleged incompatibility with the principle of nulla poena sine culpa. After reflecting on the merits of this line of criticism, the present article will explain why the academic vitriol that JCE III liability has sustained over the years regarding its performance vis-à-vis the culpability principle can now be used to also raise valid questions about the manner in which the ICC has defined its theory of co-perpetration based on joint control, particularly with respect to its ‘common plan’ element.
2. Theories of co-perpetration and the role of the ‘common plan’ element
When the Lubanga Pre-Trial Chamber first defined the legal framework of co-perpetration under the ICC Rome Statute, it started by providing the following generic definition of this concept:
[T]he concept of co-perpetration is originally rooted in the idea that when the sum of co-ordinated individual contributions of a plurality of persons results in the realisation of all the objective elements of a crime, any person making a contribution can be held vicariously responsible for the contributions of all the others and, as a result, can be considered as a principal to the whole crime.Footnote 4
This idea of a reciprocal attribution of acts is indeed the core, underlying feature of any theory of co-perpetration, be it JCE, joint control over the crime, or any other doctrine that is used to dress this form of responsibility in a concrete legal framework. The difference between these theories lies chiefly in the distinct underlying criterion that they endorse to draw the line between primary (principals) and secondary (accessories) parties to a group offence.Footnote 5 For the joint control theory, this is the ‘essential contribution’ element, pursuant to which co-perpetrators in the above-stated paradigm are only those persons whose acts were indispensable for the commission of the crime in that without their participation the common plan would have collapsed.Footnote 6 They are considered to possess joint control over the group crime because each one of them has the power to frustrate its commission by withdrawing his contribution to it.Footnote 7 Thus, in ICC case law, the concept of co-perpetration is based on the joint control theory and has the following constituent elements:
1) A common plan or agreement between two or more persons;
2) Co-ordinated essential contribution made by each co-perpetrator;
3) The accused fulfils the subjective elements of the concerted crime;
4) The accused and the other co-perpetrators are mutually aware and mutually accept that implementing the common plan will result in the realization of the objective elements of the crime; and
5) The accused is aware of the factual circumstances enabling him to jointly control the crime.Footnote 8
The JCE theory, on the other hand, is built around the ‘shared intent’ element as a litmus test for distinguishing between co-perpetrators and accessories to a collective crime. It states that when a crime is committed by persons acting in pursuance of a common plan, co-perpetrators are only those who shared the common purpose to commit the group crime.Footnote 9 The crucial criterion is thus the accused's mental disposition towards the collective crime, rather than the intensity of his particular contribution to it: if he shared the direct intent to commit the crime, and thus identified himself with the criminal purpose, he is a co-perpetrator, even if his contribution to the plan was, at hindsight, less than essential. As one former ICTY judge explained:
In performing the act, [JCE members] were of course carrying out their own will, but they were also carrying out [the accused's] will under that understanding. The perspective is important. The focus is not on whether he had power to prevent them from acting as they did; the focus is on whether, even if he could not prevent them from acting as they did, he could have withheld his will and thereby prevented their act from being regarded as having been done pursuant to his own will also.Footnote 10
The UN ad hoc Tribunals have distinguished between three forms of JCE, commonly referred to as the ‘basic’ (JCE I), the ‘systemic’ (JCE II) and the ‘extended’ (JCE III) variant,Footnote 11 which have the same objective requirements, namely: i) the existence of a plurality of persons; ii) a common plan aiming at or involving the commission of a crime; and iii) the accused's contribution to the plan.Footnote 12 The subjective requirements of the three categories of JCE are said to mark the difference in their legal framework. JCE I requires that the accused shares the common intent to commit the concerted crime.Footnote 13 JCE II applies to cases dealing with systems of ill-treatment, such as detention or concentration camps, and requires that the accused has knowledge of the nature of this system and intends to further its criminal purpose.Footnote 14 Finally, JCE III allows holding an accused liable as a co-perpetrator for crimes that fell outside the scope of the common plan but were nonetheless a natural and foreseeable consequence of its execution. The accused would be held responsible for the un-concerted offence when: ‘(i) it was foreseeable that such a crime might be perpetrated by one or other members of the group and (ii) the accused willingly took that risk’.Footnote 15 For the present purposes, it bears noting that JCE III's mens rea standard has been equated to dolus eventualis.Footnote 16 This notion is commonly associated with the idea of risk-takingFootnote 17 and the ICTY has interpreted it to require that the JCE III accused foresaw and accepted the commission of the ‘excess’ crime(s) as a possible result of executing the common plan: a degree of risk-awareness that is lower than a probability standard, yet requires more than awareness of a minimal risk.Footnote 18
As seen from the above concise overview of their definitional elements, both JCE and the joint control over the crime theory require the existence of a ‘common plan’ between the accused and his confederates. This is only natural since the very concept of co-perpetration is intrinsically rooted in the idea of co-ordinated actions and such co-ordination can only arise when there is some form of underlying plan or agreement.Footnote 19 Aside from this practical aspect, however, there is a crucial doctrinal function that the ‘common plan’ element has in any theory of co-perpetration: it provides the legal basis on which the acts of each participant in the said plan can be reciprocally attributed to the other parties to it.Footnote 20 Put simply, the ‘common plan’ element is what enables us to hold a co-accused who did not himself physically carry out the actus reus of the concerted crime, liable as a joint perpetrator of that crime. As the ICC Appeals Chamber confirmed in Lubanga:
It is this very agreement – express or implied, previously arranged or materialising extemporaneously – that ties the co-perpetrators together and that justifies the reciprocal imputation of their respective acts. This agreement may take the form of a ‘common plan’.Footnote 21
This recognition of the doctrinal function of the ‘common plan’ requirement in any theory of co-perpetration has been consistently upheld in the ICC's jurisprudence, as seen most recently in the Bemba et al. Trial Judgment,Footnote 22 but also in a series of confirmation of charges decisions.Footnote 23 Indeed, it has also been acknowledged in the case law of the ICTY.Footnote 24
This is an important consideration to keep in mind since, in accordance with the principle of culpability, the general rule is that individuals may be held liable only for their own conduct.Footnote 25 When the direct perpetrator of a crime is a fully responsible, autonomous individual, his actions are his own and they cannot be imputed to other individuals. Accordingly, as Van Sliedregt aptly pointed out, in criminal justice systems that strictly adhere to the differentiated model of criminal participation, only the person who physically commits the crime is held responsible for the crime proper, whereas the accessories who otherwise participate in its commission are held responsible for their contribution relative to the said crime, i.e., for their own conduct and its influence on the principal's crime.Footnote 26 The crime itself is thus considered to ‘belong’ to the perpetrator, whereas the accessory's liability is derivative of/‘borrowed from’ that of the principal.Footnote 27 The construct of co-perpetration transcends this paradigm by using the common plan element as a vehicle that unites the contributions of the confederates into one whole and allows treating the thus combined group conduct – i.e., the resulting crime – as, in law, the personal act of each participant in the criminal enterprise. It thus expands the classic boundaries of principal liability and enables us to define as (co-)perpetrators persons who did not physically carry out the actus reus element(s) of the group concerted crime.
The above analysis can help to explain the interplay between the ‘common plan’ element and the different definitional criteria that the JCE and joint control theories rely on to distinguish between co-perpetrators and accessories to an offence. Both theories use the ‘common plan’ as a basis for mutually attributing the individual contributions to a group crime, yet both theories also accept that the scope of this mutual attribution should be limited so that not every participant in a ‘common plan’ is a co-perpetrator: this is where the said definitional criterion comes into play. For the JCE doctrine, only those participants in the common plan who share its criminal purpose (i.e., share a direct intent/purpose to commit the concerted offence) are co-perpetrators, whereas those who contribute merely with knowledge of the said criminal purpose are accessories.Footnote 28 The ICTY Kvočka et al. Trial Chamber explained this rationale by offering the following example:
For instance, an accountant hired to work for a film company that produces child pornography may initially manage accounts without awareness of the criminal nature of the company. Eventually, however, he comes to know that the company produces child pornography, which he knows to be illegal. If the accountant continues to work for the company despite this knowledge, he could be said to aid or abet the criminal enterprise. Even if it was also shown that the accountant detested child pornography, criminal liability would still attach. At some point, moreover, if the accountant continues to work at the company long enough and performs his job in a competent and efficient manner with only an occasional protest regarding the despicable goals of the company, it would be reasonable to infer that he shares the criminal intent of the enterprise and thus becomes a co-perpetrator.Footnote 29
Naturally, the accused must still contribute to the commission of the concerted crime, but it is not the degree of that contribution that is determinative of his responsibility as a co-perpetrator, i.e., it suffices if the accused's contribution was ‘significant’, though not sine qua non for the execution of the plan.Footnote 30 This balancing exercise is reversed in the joint control theory, where the ‘common plan’ to produce child pornography may still be used to mutually attribute the criminal acts to the accountant, though only if his contribution to the enterprise was ‘essential’.Footnote 31 This approach does not require that the accused shared a purpose/direct intent to commit the collective crime: mutual awareness and acceptance that the crime would be committed suffices.Footnote 32 The ‘shared intent’ and ‘essential contribution’ criteria are thus central to the legal frameworks of, respectively, JCE and the joint control theory, yet – as a point of departure – it is the ‘common plan’ that defines which acts can be subject to mutual attribution amongst the confederates to begin with. Put simply, it is how we define the scope of the ‘common plan’ that determines what crimes fall within it (and are thus mutually attributable to the co-perpetrators) and what crimes fall outside the common plan's scope (and are thus un-concerted, ‘excess’ crimes).
In view of its centrality to the concept of co-perpetration, the definition of the exact scope and outer limits of the ‘common plan’ requirement may greatly affect the overall structure of this type of criminal responsibility. The following section will reveal how the international courts and tribunals have varied in their formulations of this element under JCE and the joint control theory.
3. Between common plans to commit a crime and common plans to maybe also commit a crime
In its recent Stanišić & Župljanin Appeal Judgement, the ICTY Appeals Chamber was seized with a submission by one of the defendants, alleging that the Trial Chamber had improperly convicted him as a co-perpetrator in a JCE. Specifically, he argued that instead of establishing the existence of a common plan to commit a concrete crime (and his participation therein), the judges provided ‘loose definitions of the common purpose that merely involve an objective where it is probable that a crime will be committed in pursuit of the objective’.Footnote 33 The Appeals Chamber rejected this allegation and emphasized instead that the trial judges correctly held that the ‘common criminal purpose of the JCE was the permanent removal of Bosnian Muslims and Bosnian Croats through the commission of crimes provided for in the Statute’.Footnote 34
At first look, the difference between the accused's argument and the Appeals Chamber's finding above may be lost to the reader, or perhaps seem insignificant, yet – as shown below – it is a difference that has important practical implications and raises concerns about the nulla poena sine culpa principle. Looking at the co-perpetration jurisprudence of the ICC, the UN Tribunals and the hybrid tribunals, specifically of the ECCC and the SCSL, one can identify three distinct constructions of the ‘common plan’ element.
3.1. The ‘common plan’ element under the ICTY/R case law on JCE
When the Tadić Appeals Chamber first defined the legal framework of JCE liability, it found that this form of responsibility requires inter alia the ‘existence of a common plan, design or purpose which amounts to or involves the commission of a crime’.Footnote 35 The main idea behind this ‘either/or’ formulation is to confirm that JCE applies in cases where the confederates share a common plan that has a criminal objective (e.g., a common plan to kill the members of a certain ethnic group), as well as in cases where the common plan has a non-criminal objective, which is to be achieved through criminal means (e.g., a common plan to establish political control over a given country, by deporting unwanted segments of its population). Here lies a very important point: in the latter scenarios, JCE law still requires that the confederates specifically agreed (and, thus, shared direct intent) to commit a certain crime as the necessary means to achieve their otherwise non-criminal objective.Footnote 36 In either case, thus, the bottom line is that the said common plan must be specifically directed at the commission of an offence: whether as the goal by itself, or as the agreed means to achieve an overall non-criminal objective. Pursuant to this interpretation, under the JCE doctrine, proving that a plurality of persons agreed to pursue a legitimate goal by taking a series of actions that were not inherently criminal would not satisfy the ‘common plan’ element, even if executing the agreed plan entailed an objective risk of the commission of a crime. In the words of the ICTY Vasiljević Trial Chamber, the prosecution has to prove that ‘two or more persons [agreed] that a particular crime will be committed’.Footnote 37 This construction is then respectively matched by the core subjective element of the JCE doctrine: namely, that the accused must share a direct intent/dolus directus in the first degree to commit the core crime(s) of the enterprise.Footnote 38
To illustrate this point, consider the Šainović et al. case, where the prosecution submitted that the accused were co-perpetrators in a JCE that did not have a strictly criminal goal: it sought ‘the modification of the ethnic balance in Kosovo in order to ensure continued Serbian control over the province’.Footnote 39 The Prosecutor then specified that this objective was meant ‘to be achieved by criminal means . . . that included deportations, murders, forcible transfers and persecutions directed at the Kosovo Albanian population’.Footnote 40 In its judgment, the Trial Chamber stressed that JCE law requires proof that ‘the accused and at least one other person . . . came to an express or implied agreement that a particular crime or underlying offence would be committed’.Footnote 41 Upon examining the evidence, it then found that:
the common purpose of the joint criminal enterprise was to ensure continued control . . . over Kosovo and that it was to be achieved by criminal means. Through a widespread and systematic campaign of terror and violence, the Kosovo Albanian population was to be forcibly displaced both within and without Kosovo. The members of the joint criminal enterprise were aware that it was unrealistic to expect to be able to displace each and every Kosovo Albanian from Kosovo, so the common purpose was to displace a number of them sufficient to tip the demographic balance more toward ethnic equality and in order to cow the Kosovo Albanians into submission.Footnote 42
It was only after the Chamber found that the crimes of deportation (Count 1) and forcible transfer (Count 2) were the necessary means which the JCE participants agreed to use in order to achieve their non-criminal aim, that the judges concluded that the ‘common plan’ element was fulfilled.Footnote 43 These two crimes thus formed the scope of the ‘basic’ JCE in the case, whereas the other crimes charged in the indictment – murder (Counts 3–4) and persecution (Count 5) – were considered to fall outside the said common plan. Responsibility for the latter could thus only be assessed under the ‘extended’ category of JCE,Footnote 44 i.e., under the category that allows holding an accused liable for offences that were not part of the plan, provided that he foresaw the risk of their commission and yet continued to participate in the JCE. In sum, the judges’ approach was to require proof that the accused agreed to commit a certain crime and then limit the scope of the ‘common plan’ element strictly to it, while other crimes that could be construed as a foreseeable consequence of this plan were treated as external to it: a rationale that was subsequently upheld by the Appeals ChamberFootnote 45 and that has indeed been standardly applied in the ICTY jurisprudence.Footnote 46
Before contrasting how the other international tribunals have constructed the scope of the ‘common plan’ element, there is one final consideration that should be addressed at this juncture. One could argue that the ICTY/R's restrictive definition of this element is practically immaterial because the ‘extended’ type of JCE allows judges to also hold the accused responsible for crimes that fall outside this narrowly defined scope of the ‘common plan’. In other words, does it really matter that JCE requires a common plan that is specifically directed at the commission of a crime when the liability of the JCE participants is not really confined within the scope of the plan? This question is addressed in detail further below, yet it bears noting at this point that JCE III liability arises only when the accused's responsibility under the ‘basic’ or ‘systemic’ category of JCE has been established, i.e., in order to be held liable for a foreseeable, ‘excess’ crime, the accused must first be shown to have shared the direct intent and contributed to the commission of the common plan's core, ‘original’ crime(s).Footnote 47 As Cassese pointed out, JCE III does not concern cases where a group of persons pursue a lawful enterprise that incidentally leads to the commission of a crime, but is rather about cases where the confederates share a common purpose to commit crime X, yet in the course of effecting the plan crime Y is also additionally committed.Footnote 48 In this sense, JCE III is best described as an add-on notion to the ‘basic’ and ‘systemic’ categories of the theory.
The above analysis affirms the conclusion that without a common plan that is specifically directed at the commission of a particular crime, there can be neither JCE I, nor JCE II, nor JCE III responsibility. Accordingly, if the accused is acquitted under the underlying JCE I/II charges, any charges brought under the ‘extended’ variant of JCE become legally unsustainable.Footnote 49 For the purposes of this article, this is a crucial point to keep in mind: by requiring that the common plan be specifically directed at the commission of a particular crime, the JCE doctrine ensures that co-perpetration liability is not ascribed for participating in common plans that simply create a crime-conducive environment. At least not under the ICTY/R's case law on this mode of liability.
3.2. The ‘common plan’ element under the SCSL and ECCC's case law on JCE
While the SCSL and the ECCC both endorsed in their jurisprudence the theory of co-perpetration based on JCE,Footnote 50 they drastically altered the scope of its ‘common plan’ element, prompting some scholars to argue that these two tribunals in fact created ‘a new, unfounded form of JCE’.Footnote 51
The finding that JCE responsibility requires a common plan that is specifically directed at the commission of a particular crime lost its way in the SCSL and ECCC's case law. Both courts came to adopt a notably different interpretation of this element, according to which the common plan that the JCE participants share ‘must either have as its objective a crime or contemplate the crimes as the means of achieving this objective’.Footnote 52 Although, at first glance, this might seem to be a quite trivial semantic difference from the phraseology used by the ICTY/R – i.e., that the common plan must ‘amount to or involve’ the commission of a given crime – in practice, it has completely redrawn the scope and nature of JCE's ‘common plan’ element.
Changing the ‘or involve’ formula with ‘or contemplate’ has led the SCSL and the ECCC to ascribe JCE responsibility on the basis of common plans that neither have a criminal objective, nor incorporate the commission of a crime as the agreed means to an end, but merely entail a risk of the commission of a crime.Footnote 53 The origins of this distortion of the JCE doctrine can be found in the SCSL AFRC case, in which the prosecution submitted that the three accused participated in a common plan/agreement ‘to take any actions necessary to gain and exercise political power and control over the territory of Sierra Leone’.Footnote 54 The crimes that were charged in the indictment were then defined as falling ‘within [the] joint criminal enterprise in which each Accused participated or were a reasonably foreseeable consequence of the joint criminal enterprise’.Footnote 55 The Prosecutor, thus, did not identify any one particular crime that the accused had specifically agreed to commit as the necessary means to achieve the said non-criminal objective: indeed, pursuant to the above pleading of the common plan, it could be that all the crimes charged in the indictment were only a foreseeable risk of executing the alleged common plan. This amounted to using JCE III liability in the absence of an underlying JCE I, which sharply conflicted with the ICTY/R's case law and practically eviscerated the border between the ‘basic’ and the ‘extended’ type of JCE.Footnote 56 The Trial Chamber rightly dismissed the prosecution's JCE case, stressing as it did that this notion requires a common plan that, from its inception, pursues the commission of a particular crime and is thus inherently criminal.Footnote 57 On appeal, however, the AFRC Appeals Chamber rejected this finding and adopted instead the prosecution's approach, holding that:
the requirement that the common plan, design or purpose of a joint criminal enterprise is inherently criminal means that it must either have as its objective a crime within the Statute, or contemplate crimes within the Statute as the means of achieving its objective.Footnote 58
The Appeals Chamber's conclusion on this point started a new string of JCE case law: in fact, it created a new notion of co-perpetration that was JCE in name, but not in content. Finding that the ‘common plan’ element can comprise crimes that the accused either specifically intended to commit, or that were reasonable foreseeable to them, or indeed both, merges JCE I and JCE III liability into one and effectively jettisons the ‘shared intent’ element, i.e., the requirement that the accused and the other JCE members must share a dolus directus in the first degree to commit the core crime(s) of the enterprise.Footnote 59 Under this formulation, it could be that none of the participants directly intended the commission of a specific crime when forming their common plan, but were merely aware of a possibility that crimes might be committed in the course of its execution. This re-interpretation of the ‘common plan’ element effectively caused what Meisenberg described as a ‘judicial meltdown of the JCE doctrine’.Footnote 60
Other than deviating from ICTY/R's JCE jurisprudence and, thus, raising questions about its basis in customary international law,Footnote 61 this expansive definition of the ‘common plan’ element is problematic because it makes the JCE theory:
incapable of delineating between the collective pursuit of a war, and concerted action in furtherance of a crime. Contributions to a non-criminal objective are deemed contributions to a common criminal purpose, with mere criminal ‘contemplation’ linking the accused into an alleged criminal enterprise.Footnote 62
Indeed, when two or more persons agree on a plan to wage a war, there is always an inherent risk that this course of action, legitimate as it may be, may possibly lead to the commission of crimes. If the scope of JCE's ‘common plan’ element is defined to comprise ‘contemplated’ crimes, then it can be that the only way in which a person in such a context can avoid co-perpetration liability for resulting crime(s) is by not entering common plans to wage wars: a conclusion that hardly fits in the established legal framework of jus ad bellum.Footnote 63 Consider, for instance, the situation where in 2011, acting with the authorization of the UN Security Council, the leaders of various states in a NATO coalition agreed to conduct a military intervention in Libya in order ‘to protect civilians and civilian populated areas under threat of attack in the Libyan Arab Jamahiriya’Footnote 64 (legitimate, non-criminal objective) by using armed force (non-criminal means to achieve the said objective). Under a broad definition of the ‘common plan’ element, the above-defined plan could be used as a basis to ascribe co-perpetration liability to its participants for crimes against civilians that were allegedly committed by NATO forces during its execution,Footnote 65 pending a finding that these crimes were a foreseeable ‘possible consequence’ of executing the common plan. Since armed conflicts are historically known to create a crime-conducive environment,Footnote 66 the dangers of an overly broad definition of the ‘common plan’ element of co-perpetration liability become evident. In this vein of thought, Judge Afande recently stressed in his dissenting opinion to the Stanišić and Simatović Appeal Judgement the need to ‘clearly distingui[sh] JCE liability from another enterprise such as “Joint Warfare Enterprise” [which involves] a plurality of persons . . . who have the intent to further not a criminal purpose, but rather a legal “warfare purpose” which is common to them’.Footnote 67
The SCSL's broad definition of JCE's ‘common plan’ element was subsequently adopted in the ECCC's case law, a recent example of which is the Case 002 Trial Judgement.Footnote 68 The Closing Order alleged that the accused committed a number of crimes through participating in a JCE, which had the following underlying common plan:
The common purpose of the [Communist Party of Kampuchea's] leaders was to implement rapid socialist revolution . . . in Cambodia through a “great leap forward” and to defend the Party against internal and external enemies, by whatever means necessary. The purpose itself was not entirely criminal in nature but its implementation resulted in and/or involved the commission of crimes within the jurisdiction of the ECCC.Footnote 69
When the accused complained that he was impermissibly charged with ‘participat[ing] in a non-criminal common plan that merely resulted in the commission of crimes’,Footnote 70 the trial judges cited the above-mentioned SCSL cases to conclude that the common plan of a JCE ‘must either have as its objective a crime or contemplate crimes as the means of achieving its objective’.Footnote 71 Thereafter, upon examining the evidence, the Chamber found that there was a common plan to implement a rapid socialist revolution in Cambodia, which was effected through policies – i.e., the population movement policy and the targeting policy – that ‘resulted in and/or involved crimes’.Footnote 72 Pursuant to the former policy, many people were forcibly relocated from their villages and cities, and then made to work in agricultural production co-operatives in order to rebuild Cambodia's economy.Footnote 73 The judges found that this policy ‘did not make any provision for the well-being or the health of those being moved, in particular the vulnerable’Footnote 74 and thereby concluded that:
the forced transfers committed by Khmer Rouge officials and soldiers during movement of population (phases one and two) were undertaken pursuant to the Party leadership's express instructions, decisions and policy. Further, they were carried out as part of a pattern of forced transfers, under inhumane conditions and without regard for the well-being or the health of the people being moved. Murders and attacks against human dignity resulted from the inhumane conditions of the transfers, terror-inducing acts of Khmer Rouge cadres and the exercise of force. Party policy intended that such suffering and sacrifice would reeducate the “New People” and attack the class system. On this basis, “New People” were persecuted on political grounds.Footnote 75
In line with the aforementioned broad formulation of the ‘common plan’ requirement, the crimes listed in this paragraph – i.e., murder, political persecution and the other inhumane acts of forced transfer and attacks against human dignity – were all lumped together as falling within the scope of the common plan and the accused were held guilty of them under the ‘basic’ form of JCE.Footnote 76
Under the ICTY/R's JCE jurisprudence, the common plan would have been limited to the crime of forcible transfer because this is the crime that the accused specifically agreed to commit as the necessary means to achieve their broad, non-criminal objective, while the other crimes that ‘resulted from the inhumane conditions of the transfer’ would fall outside its scope and could be charged under the ‘extended’ JCE form.Footnote 77 However, since the ECCC had rejected the use of JCE III for lacking legal basis in customary international law,Footnote 78 this approach was not available to the judges. Their solution was simple: use the SCSL's expansive construction of the ‘common plan’ element that collapses the border between JCE I and JCE III – viz. the ‘aims at or contemplates’ formula – and thereby conclude that all the said crimes fell within the scope of the common plan. In other words, the ‘common plan’ element was constructed to incorporate both crimes that were specifically agreed on and crimes that were a reasonably foreseeable risk of effecting the plan. In turn, a person who shared and contributed to the broad, non-criminal objective of implementing a rapid socialist revolution in Cambodia could be held liable as a co-perpetrator of the said crimes, provided that he either directly intended the commission of the said crimes, or was aware of the risk that they will be committed in the execution of the plan.
Using this very construction, the accused Khieu Samphan was convicted of all the above-said crimes under the ‘basic’ JCE form, even though his mens rea in relation to these crimes was defined in the following terms:
From 1969, when he joined the Party, he participated in meetings, congresses and conferences, where the common purpose was affirmed, developed and the policies to implement it were decided upon . . . KHIEU Samphan knew of the substantial likelihood that crimes would result from implementation of these policies. He knew that these policies did in fact result in and/or involve the crimes committed in the course of phases one and two of population movements and at Tuol Po Chrey. He also had further notice of the crimes after their commission. Despite this knowledge, he continued to contribute to and approve the progress of the democratic and socialist revolutions.Footnote 79
Khieu Samphan was thus found guilty under the ‘basic’ type of JCE for crimes that he did not, at least not initially,Footnote 80 specifically agree to commit (i.e., share a direct intent/purpose to commit) but only ‘knew of the substantial likelihood’ of their commission. This was possible due to the Trial Chamber's broad definition of the ‘common plan’ element, pursuant to which its scope included crimes that were specifically agreed on (the confederates share a direct intent to commit), and/or crimes that were a probable/possible outcome of the plan's execution (the confederates are aware of the substantial likelihood that they may result from effecting the common plan). The ECCC's definition of the ‘common plan’ element in the JCE doctrine is thus analogous to that adopted by the SCSL and is, therefore, subject to the same criticism.Footnote 81
3.3. The ‘common plan’ element under the ICC case law on co-perpetration
When it first construed the doctrine of co-perpetration based on joint control over the crime, the Lubanga Pre-Trial Chamber held that the common plan ‘does not need to be specifically directed at the commission of a crime’ since it suffices if, at the minimum, the plan involves ‘an element of criminality’.Footnote 82 The judges further explained that a common plan with a non-criminal objective will be considered to have ‘an element of criminality’, if it can be established that:
i. . . .the co-perpetrators have agreed (a) to start the implementation of the common plan to achieve a non-criminal goal, and (b) to only commit the crime if certain conditions are met; or
ii. that the co-perpetrators (a) are aware of the risk that implementing the common plan (which is specifically directed at the achievement of a non-criminal goal) will result in the commission of the crime, and (b) accept such an outcome.Footnote 83
This finding on the scope of the ‘common plan’ requirement under the ICC's joint control theory marked a material difference with the ICTY/R's definition of this element under the JCE theory, which requires that the common plan must be specifically directed at the commission of a crime: either as the end goal of the enterprise, or as the agreed means to achieve a non-criminal goal.Footnote 84 By contrast, the Lubanga Pre-Trial Chamber asserted that a plan that has a non-criminal goal and that the confederates agree to pursue by non-criminal (neutral) means may still serve as the basis for co-perpetration responsibility, if its effectuation entails a risk of the commission of crimes.Footnote 85 The Chamber further indicated the level of the said risk by concluding that plan has to involve ‘a substantial risk of bringing about the objective elements of the crimes’,Footnote 86 that the confederates in it must be aware ‘of the substantial likelihood that implementing the common plan would result in [the charged crime]’,Footnote 87 and also that they must be aware and accept that effecting the common plan ‘may result in the realization of the objective elements of the crime’.Footnote 88 This construction of the ‘common plan’ is practically identical to the SCSL and ECCC's definition of this element,Footnote 89 seeing as both include within the scope of the plan crimes that are not specifically agreed on, yet are a probable/possible consequence (risk) of the plan's execution.
The Lubanga Pre-Trial Chamber's conclusion that the ‘common plan’ element is satisfied if the plan contains ‘an element of criminality’ has been affirmed in the subsequent jurisprudence of the Court,Footnote 90 though it has been revised in one important aspect: the level of risk that suffices to bring an offence within the scope of the plan and make it mutually attributable to its participants. The issue was squarely raised by the Lubanga defence when it argued on trial that the concept of co-perpetration liability must require a common plan that is ‘intrinsically criminal’, not one that is ‘merely capable of creating conditions conducive to the commission of criminal acts’.Footnote 91 While the Trial Chamber rejected this submission, maintaining instead that the common plan need only have ‘an element of criminality’, the judges did narrow down the scope of this test by stating that it is satisfied when ‘[the plan's] implementation embodied a sufficient risk that, if events follow the ordinary course, a crime will be committed’.Footnote 92 Thus, even though it found that the common plan in the present case was not inherently criminal (viz. ‘to build an effective army to ensure the UPC/FPLC's domination of Ituri’),Footnote 93 the Chamber held that:
[t]his plan resulted in the conscription, enlistment and use of children under the age of 15 to participate actively in hostilities, a consequence which occurred in the ordinary course of events. This conclusion satisfies the common-plan requirement under Article 25(3)(a).Footnote 94
The confederates did not specifically agree to target children under the age of 15 for the purpose of building an effective army and gaining control over Ituri: rather, the charged crime was treated as part of the common plan since it was ‘a consequence which occurred in the ordinary course of events’.
The Trial Chamber's analysis thus seemed to exclude from the scope of the common plan crimes that are merely a possible consequence of the plan's execution, i.e., a low level of risk that is usually associated with the concept of dolus eventualis.Footnote 95 Instead, the common plan must be of such nature that its execution will result in the commission of the charged crime ‘in the ordinary course of events’. This was further affirmed by the Lubanga Appeals Chamber, which explained that the commission of the charged crime has to be ‘a virtual certainty’ of effecting the common plan in order to determine that the said plan had ‘an element of criminality’.Footnote 96 The appeal judges even pointed out that the Trial Chamber's use of the word ‘risk’ in this context was inappropriate because ‘risk-taking’ is usually associated with the notion of dolus eventualis and awareness of a ‘possibility’ (or a ‘probability’) of a crime occurring, which is not what the requisite ‘element of criminality’ stands for.Footnote 97 Accordingly, under the ICC's concept of co-perpetration based on joint control, the scope of the ‘common plan’ requirement is now construed to incorporate crimes that the confederates specifically agree to commit and/or crimes that are a ‘virtually certain’ result of the execution of the said common plan.
The above construction of the ‘common plan or agreement’ requirement is liable to cause confusion and has, indeed, attracted several types of criticism both within and beyond the ICC.Footnote 98 The idea that co-perpetration liability may be based on a common plan that does not specifically incorporate the commission of a particular crime but simply contains ‘an element of criminality’ has prompted criticism that it tampers with the core requirement that the confederates must agree to commit the charged offence in order to mutually attribute their co-ordinated actions towards its commission.Footnote 99 As already explained above, in any theory of co-perpetration, the ‘common plan’ element provides the doctrinal basis for the mutual attribution of acts among the confederates.Footnote 100 Accordingly, a crime which they do not, expressly or implicitly, agree to commit – whether as an end goal, or as a means to achieve it – is not an agreed part of the plan and, therefore, cannot be mutually attributed to them. It is doubtful whether objectively qualifying an offence that was not specifically agreed upon as a reasonably foreseeable result of executing a common plan does, by itself, establish that the confederates formed an agreement to commit this crime, thereby bringing it within the scope of the common plan. After all, as the UN Tribunals have also stressed, what is natural and foreseeable to one confederate in a common plan may not be natural and foreseeable to another,Footnote 101 which is why it is contentious to treat such crimes as an agreed part of the common plan.
3.4. Varying scopes and serious implications: Concluding remarks on the ‘common plan’ element
In view of the above analysis, it is argued that the manner in which judges define the scope of the ‘common plan’ element of co-perpetration responsibility has at least two important implications: one practical and one doctrinal. On the practical side, the more expansive the construction of the ‘common plan’ is, the more impossible it becomes to distinguish between common plans to wage war and common criminal plans: a consideration that is of exceptional importance for the field of international criminal law.Footnote 102 As a doctrinal problem, defining broadly the scope of the ‘common plan’ element, so as to incorporate in it offences that were not concretely agreed upon but were a foreseeable risk of executing the plan, raises concerns as to whether there is in fact an agreement to commit a crime, i.e., whether there is a valid legal basis for the mutual attribution of acts.Footnote 103
To highlight the differences in the definitions that the ICTY/R, the SCSL/ECCC, and the ICC have given to the ‘common plan’ element, consider again the example of the 2011 NATO military intervention in Libya.Footnote 104 Under the ICTY/R approach, the common plan to wage war in Libya, as defined above, cannot serve as a basis to ascribe co-perpetration liability under the JCE doctrine for the alleged crimes committed against civilians, irrespective of how foreseeable a risk these crimes were.Footnote 105 This is because, as explained above, JCE III (under which crimes that are a foreseeable possible risk of effecting a common plan) is applicable only as an add-on to JCE I/II, i.e., only if there is a common plan that is specifically directed at the commission of a crime.Footnote 106 If the SCSL/ECCC approach to the ‘common plan’ requirement is followed, the said crimes can be incorporated as a mutually attributable part of the common plan if it is established that they were a reasonably foreseeable risk of executing the plan: i.e., the above-analyzed amalgamation of JCE I and JCE III liability.Footnote 107 Finally, under the ICC approach, mutual attribution of the said offences as a part of the common plan would be warranted if it could be proved that they were a ‘virtually certain’ result – as opposed to merely a possible/probable risk – to effecting the common plan.Footnote 108
It is important to emphasize that the definitional criterion that any particular theory of co-perpetration adopts to distinguish between principals and accessories to the collective crime (e.g., ‘shared intent/purpose’, ‘essential contribution’ etc.) has no bearing whatsoever on the scope of the ‘common plan’ element: i.e., these elements do not in any way balance each other. If a certain crime does not fall within the common plan/agreement, then there is no doctrinal basis on which it can be imputed to a confederate who did not physically commit it, irrespective of whether that person wanted/directly intended the crime's commission, or provided an essential contribution to this end.Footnote 109 Moreover, narrowing or broadening the scope of the common plan element cannot be contingent on, e.g., decreasing or increasing the requisite degree of contribution for incurring co-perpetration liability. For instance, it is neither doctrinally coherent, nor logically sound to conclude that broadly constructing a common plan to wage war, so as to include in its scope any marginally possible offence that may result from its execution, is justified by requiring that the accused's participation was ‘essential’ for the success of the said plan.
One should also distinguish the above criticism on the broad formulation of the ‘common plan’ element,Footnote 110 from the situation of ‘single’ perpetration where it has been generally accepted that a person can commit an offence either with direct intent, or with awareness of the substantial likelihood that the said crime will result from his actions.Footnote 111 It is one thing to hold an individual liable as a perpetrator for their own conduct and its foreseeable consequences, which is perfectly in conformity with the individual culpability principle, and it is a wholly different thing to hold a person liable as a co-perpetrator of the acts of another person based solely on such foreseeability. Unless the said acts can be construed as forming part of an agreement between these persons (i.e., falling within the scope of a common plan between them), these acts ‘belong’ solely to the direct perpetrator and cannot be imputed on the accused who foresaw their commission: at the most, he can be an accessory to them.Footnote 112
4. Liability for ‘excess’ crimes of a common plan: What is criminal ‘excess’?
The above analysis of the ICTY/R's, the SCSL/ECCC's and the ICC's varying definitions of the ‘common plan’ element under the JCE and joint control over the crime theory adds another layer to our understanding of the fragmentation of co-perpetration liability under international criminal law. What is more, however, this comparative review can provide us with a fresh perspective and raise new questions on the concept of ascribing liability for crimes that fall outside the scope of a common plan: viz. the so-called ‘excess’ crimes.
As explained above, JCE is a theory of co-perpetrationFootnote 113 that relies on the existence of a common criminal plan/design/purpose as a doctrinal basis for mutual attribution of acts among the participants in a criminal enterprise.Footnote 114 This element ensures that the principle of individual culpability is respected when a JCE member is held responsible as a principal for a crime that he did not himself physically/directly commit.Footnote 115 While this reasoning holds true for the ‘basic’ and ‘systemic’ forms of JCE, both of which ascribe co-perpetration responsibility for crimes that are an agreed part of the common plan,Footnote 116 it runs into troubled waters with this theory's ‘extended’ form. Indeed, aside from its debatable status under customary law, the relentless criticism against JCE III liability has also been based on a critical doctrinal objection:
As to the principle of culpability, the conflict with JCE III is even more evident . . . [I]f, according to this doctrine, all members of a criminal enterprise incur criminal responsibility even for criminal acts performed by only some members and which have not been agreed upon by all members before the actual commission but are, nonetheless, attributed to all of them on the basis of mere foreseeability, the previous agreement or plan of the participants as the only legitimate basis of reciprocal attribution has been given up. On what basis can a member of the original JCE who behaved in full compliance with the original plan then be blamed for the excess crimes?Footnote 117
There is much force in this submission, considering that – in relation to the ‘excess’ crime – JCE III lacks the doctrinal basis for mutually attributing to all participants in a criminal enterprise the deviatory, un-agreed crime that one of them additionally perpetrates. Although it is certainly true that JCE III always requires the existence of a common criminal plan,Footnote 118 the plan exists only vis-à-vis the ‘core’ crime of the enterprise.Footnote 119 The ‘excess’ crimes, which are the focus of JCE III, are considered by the UN Tribunals to fall outside the scope of the common plan exactly because they are not specifically agreed on by the confederates. The fact that such a crime is a natural and foreseeable possible consequence of effecting the original plan does not bring it within the plan's scope, at least not under the ICTY/R's definition of the ‘common plan’ element.Footnote 120 Furthermore, while the accused's participation in the original common plan may provide a causal link with the commission of the ‘excess’ offence,Footnote 121 such a link is not by itself a doctrinal basis for the mutual attribution of the un-agreed act.Footnote 122 To accept that a causal connection between a person's conduct and another person's crime is enough to assign principal responsibility to the former (as opposed to accessorial liability, which is what the participant in another person's crime would normally incur), is to collapse the differentiated model of criminal participation into a unitary one.Footnote 123
Considering the above research on the different definitions that the international tribunals have given to the ‘common plan’ element and its exact scope, one cannot help but notice that the problem of treating JCE III as a form of co-perpetration is closely linked to the ICTY/R's narrow construction of the ‘common plan’ element. Since crimes that are not specifically agreed on – but are still a foreseeable consequence of effecting the JCE – are qualified as external to the common plan, it is rightly observed that there is no doctrinal basis on which these crimes may be mutually attributed to each JCE participant. However, this raises an important normative question: are the un-concerted but reasonably foreseeable crimes treated as ‘external’ simply because the ICTY/R labels them as such, or are they ‘external’ because, in principle, only crimes that the confederates specifically agree to commit can fall within the scope of the common plan?
If it is the latter, the strong criticism that the ‘extended’ JCE type has faced over the years for assigning co-perpetration liability for un-concerted but foreseeable crimes applies with equal force to the ICC's joint control theory, which also allows holding an accused responsible as a co-perpetrator for crimes that were not a specifically agreed part of the common plan/agreement: i.e., the ‘element of criminality’ formula.Footnote 124 Indeed, a broad interpretation of the common plan, which allows including in it crimes that the co-perpetrators did not, expressly or implicitly, specifically agree to commit, is eerily reminiscent of the idea of ascribing responsibility for ‘excess’ crimes, thus raising concerns about the joint control theory's conformity with the culpability principle.Footnote 125 Endorsing the view that the plan must be intrinsically criminal – i.e., that the confederates have to agree to commit a concrete offence – cancels the possibility of assigning co-perpetration liability to individuals who provide non-criminal contributions (e.g., mobilizing armed troops and sending them to a given territory) to achieve a common non-criminal objective (e.g., maintaining political control over that region), based on the mere awareness that this plan may lead to the commission of crimes. This point is especially poignant in the field of international criminal law, which often deals with situations where individuals agree on legitimate common plans to wage war, such as a peace-enforcing operation, or a war of secession, yet, as one SCSL chamber pointed out, history teaches us that such ‘joint warfare enterprises’Footnote 126 are always likely to result in the commission of crimes by some members of the armed forces.Footnote 127 It should also be emphasized here that treating crimes that are not specifically agreed by the confederates, yet are a reasonably foreseeable result of the execution of their plan, as external to the common plan does not necessarily mean that the said individuals can incur no criminal responsibility for the commission of these offences. It only means that they cannot be held liable as co-perpetrators of the ‘excess’ crime, yet they may very well qualify as accessories to these crimes, depending on the facts of the case.
Alternatively, one may subscribe to the view, as the ICC and the SCSL/ECCC have done, that individuals who agree to pursue a non-criminal goal through non-criminal means, ipso facto also agree to the commission of any reasonably foreseeable crime resulting from the execution of this plan, thus including it within the common plan's scope and making it mutually attributable to each confederate.Footnote 128 Those who are comfortable with such a broad construction of the ‘common plan’ element, however, should also not be critical of JCE III liability, because they would accept that the ‘excess’ crimes it deals with are only perceived as such by the ICTY/R, yet in substance they actually fall within the adopted broad definition of the ‘common plan’ element. Considering the above analysis, however, there are good reasons to be cautious about using such an expansive definition to assign co-perpetration liability.Footnote 129 In any case, accepting the view that the ‘common plan’ does not have to be specifically directed at the commission of a given crime, would require also agreement on some minimum threshold of crime-conduciveness that the said plan has to entail in order to be a ‘common plan’ for the purposes of co-perpetration liability.Footnote 130 It would clearly be excessive to state that any crime, for which there was even a minimal risk of being committed in the execution of a common plan (to, e.g., wage war), is an agreed part of the plan that falls within its scope. In this respect, the ICC Lubanga Appeals Chamber's decision to limit the scope of the ‘common plan’ element strictly to those crimes that, even though not specifically agreed upon by the confederates, were ‘a virtually certain’ (rather than merely ‘possible’ or ‘probable’) result of the executing their plan presents a middle ground: it is wider than the ICTY/R definition, yet not as broad as that of the SCSL/ECCC.
5. Conclusion
The international criminal law on co-perpetration liability is in a state of flux and this article has submitted that the division between JCE-aligned and joint control-aligned tribunals has started to run deeper than the commonly reported ‘shared intent’ versus ‘essential contribution’ dichotomy for distinguishing between principals and accessories to collective crime. On a more rudimentary level, the very manner in which the international courts and tribunals have come to formulate the scope of the ‘common plan’ element under their respective theories of co-perpetration has caused further partitioning in their jurisprudence on this mode of liability.
While the existence of a common plan or agreement between the accused and his partners in crime is a constituent element of both the JCE and the joint control concept of co-perpetration, the ICTY/R, the ICC and the various hybrid courts have in fact endorsed three materially distinct definitions of this requirement, thus spawning three different formulations of this type of liability in their case law. The ad hoc Tribunals have adopted the ‘aim at or involve’ formula, pursuant to which the underlying common plan of a JCE has to be specifically directed at the commission of a crime, i.e., the crime, which the JCE participants are required to share a direct intent to commit. It was explained that this is a common requirement for all three variants of this doctrine, which is why, under the ICTY/R jurisprudence, a common plan that has a non-criminal objective and that does not necessarily involve the commission of a crime, as the agreed means to achieve this goal, cannot satisfy the ‘common plan’ element of the JCE theory. The tribunals have stressed that the ‘extended’ type of JCE, which allows ascribing co-perpetration liability for foreseeable crimes of the enterprise, requires that, as a point of departure, there must be an underlying common plan to commit a specific crime. The SCSL and the ECCC have branched away from this reasoning, and have instead endorsed the more expansive ‘aim at or contemplate’ formula, according to which a common plan could also consist of crimes that were just a reasonably foreseeable consequence of its execution. This interpretation effectively conflates the classic boundaries of JCE I and JCE III into one new mode of co-perpetration liability that, when applied in practice, has little to do with the ICTY/R's JCE theory. Aside from challenges to its basis under customary law, this definition of the common plan element raises practical concerns about the legitimacy of using the resulting broad construction of co-perpetration in the context of armed conflicts. As for the ICC, the Court has defined the outer limits of the ‘common plan’ requirement under its theory of co-perpetration based on joint control in a way that, at least initially, resembled more closely the ECCC's and the SCSL's formulation of this element. It was only recently that the ICC appeal judges narrowed its scope by confirming that a common plan has to contain a ‘critical element of criminality’, which means that the plan, although not specifically directed at the commission of the charged crime, is ‘virtually certain’ to result in it. Thus, on a continuum of how intrinsically criminal the common plan has to be, one could place the ICC's definition somewhere between that of the ICTY/R and the SCSL/ECCC, i.e., less crime-oriented than the former and more so than the latter.
The manner in which we define the outer limits of the ‘common plan’ element – viz. what acts can be treated as an agreed part of the plan and are, thereby, mutually attributable to the co-perpetrators – naturally also affects the scope of the crimes that would fall outside the plan. This article studied the doctrinal criticism that has plagued the JCE III theory for years and submitted that much the same objections can be raised against the ICC's theory of co-perpetration based on joint control. In particular, it was pointed out that the latter theory also ascribes principal liability for crimes that the confederates in the common plan did not specifically agree to commit, as long as they can be qualified as a reasonably foreseeable (‘virtually certain’) consequence of effecting the said plan, i.e., the ‘element of criminality’ formula. Since the ICC treats such crimes as falling within the agreed scope of the common plan, and the ICTY/R views them as ‘excess’ crimes that are outside the plan (and subject to JCE III), a normative question arises whether a common plan can ever be said to exist in relation to a crime that a group of persons do not specifically agree to commit, yet its commission is foreseeable as a possible/probable/virtually certain consequence of effecting their agreement to pursue a common goal. The answer to this question should occupy a more central place in the international debates on co-perpetration responsibility: not only because of what it can tell us about the perceived doctrinal deficiencies of JCE III, but also because of the risks involved in adopting broad definitions of the ‘common plan’ element in the specific context of international criminal law.
