I. INTRODUCTION
In February 2017 the Independent International Commission of Inquiry on the Syrian Arab Republic issued its report into events in Aleppo between July and December of the previous year.Footnote 1 The Report detailed widespread violations of international human rights law and international humanitarian law. One of its findings concerned the agreement by parties to the conflict in mid-December of 2016 to allow the evacuation of civilians from the east of the city. On that agreement, the Report concluded in the following terms:
After the Government reached an evacuation agreement with armed groups in mid-December, residents of eastern Aleppo were transported from the city in government buses and private vehicles to Idlib, while others fled to western Aleppo. None had the option to remain in their home. As part of the agreement, more than 1,000 people were evacuated from Foah and Kafraya and went to Aleppo, Tartous, Homs and Latakia governorates. As warring parties agreed to the evacuation of eastern Aleppo for strategic reasons – and not for the security of civilians or imperative military necessity, which permit the displacement of thousands – the Aleppo evacuation agreement amounts to the war crime of forced displacement.Footnote 2
Whatever its implications for the parties to the conflict, the conclusion that the evacuation agreement amounted to the war crime of forced displacement caused problems for the International Committee of the Red Cross (the ICRC).Footnote 3 On one hand, it was the ICRC, an organisation committed at its core to compliance with the laws of war, which facilitated the evacuation of the civilians from Aleppo. On the other hand, it has been recognised since the inception of modern international criminal law that those who knowingly contribute to international crimes may be liable as accomplices.Footnote 4
In the specific case of the evacuation of Aleppo, subsequent commentary cast serious doubt on the Commission's finding that the agreement to evacuate constituted a war crime.Footnote 5 For the ICRC, then, this is exculpatory—with no principal wrong there is nothing for the accomplice to be complicit in.Footnote 6 Nonetheless, the case initially looked like a specific example of a wider dilemma facing humanitarian actors—a dilemma created by the complex ethical environments in which they operate.Footnote 7 Often, humanitarian actors are required to make a difficult judgment about whether to provide assistance in non-ideal conditions.Footnote 8 The non-ideal condition, in the present sense, is that other actors are failing to comply with their legal and moral duties.Footnote 9
This dilemma gives rise to the problem of the virtuous accomplice—a term suggesting that knowingly contributing to another actor's wrong might, all things considered, be the right thing to do. This problem has provoked a rich literature in political theory and medical ethics; the question in this literature concerns the morality of complicity in situations of this kind.Footnote 10 The present article considers its legality under international criminal law.Footnote 11 That is to say, it considers the potential individual criminal responsibility of officials and aid workers under international law.
As to structure and argument, Section II sketches two scenarios to anchor the discussion and shows how humanitarian actors, at first glance, risk liability for complicity. It locates that legal risk primarily in the fact that knowledge Footnote 12 is sufficient for accomplice liability in international criminal law, as opposed to an intention that the crime be committed, and argues that as a general rule, a standard of knowledge is defensible. Section III then provides two clarifications on the scope of the problem. First, paying attention to the doctrinal requirement that an accomplice's assistance must make a substantial or significant contribution to principal's crime diminishes its scope. Second, humanitarian actors may be able to take measures to mitigate the risk of their complicity. Nonetheless, the problem remains. Section IV, drawing on domestic jurisdictions, then discusses three ways that the problem of the virtuous accomplice may be resolved—two doctrinal and one institutional. It argues that the best solution is to accept the potential application of a defence of necessity—of justified complicity—and shows that such an approach is grounded in international criminal law. Section V briefly considers the wider implications of this solution, before Section VI concludes.
II. UNDERSTANDING THE PROBLEM
A. Two Scenarios
Although cataloguing the range of scenarios in which the dilemma arises is not possible, two examples anchor the analysis. These I will term mitigating complicity and general complicity. In addition to providing some colour, these two scenarios are worth distinguishing because they have different implications for the doctrinal analysis to come.
Mitigating complicity refers to cases in which the humanitarian actor's assistance is provided in order to mitigate the harm or risk of harm to the victims of the principal's wrong. A variation on the Aleppo case is emblematic—consider a humanitarian agency that provides safe transport for a population that is being deported from occupied territory in an international armed conflict—the deportation constituting a war crime.Footnote 13 Consider a nurse who sterilises equipment to be used by a captor to amputate the arm of a prisoner as torture,Footnote 14 or a doctor in a refugee camp who provides contraception to soldiers knowing that rape is endemic in the camp.Footnote 15 Cases of mitigating complicity are marked by the putative accomplice's intention to reduce the harm or risk of harm to the victims of the principal wrong.
General complicity refers to cases in which humanitarian actors provide assistance to a vulnerable population where they know that parts of that aid will be siphoned off by perpetrators in order to commit international crimes or, perhaps, where aid must be provided to those perpetrators in order to access vulnerable populations under their de facto control.Footnote 16 Lepora and Goodin give the example of the Rwandan Armed Forces’ (FAR) control of refugee camps established in the aftermath of the Rwandan Genocide:
In July 1994, a cholera epidemic hit the 850,000 refugees of Goma camp in Zaire, causing more than 80,000 deaths in ten days and ravaging the camp for months. The proliferation of aid actors trying to respond to such an overwhelming emergency further strengthened the political and military power of the FAR. Empowered by the recognition of UN camp management and humanitarian organisations, FAR groups took leadership of the newly created refugee camps. Having done so, they set about diverting aid, using camps as military recruitment and training centres, killing opponents and further spreading genocidal propaganda.
When providing aid, organizations were obliged to acknowledge, interact with and contribute to those perpetrators of genocide. Even nongovernmental organizations that intervened on a purely humanitarian basis thus ended up contributing to FAR's power, from a symbolic and sometimes material point of view. All international organizations faced the same dilemma: continue working in the camp, thereby further strengthen the power of genocidal perpetrators over the refugees; or withdraw from the camps, abandoning a population that was in extreme distress.Footnote 17
This is general complicity: cases where a putative accomplice's intention to provide assistance to a vulnerable population is marked by knowledge, or foresight of a serious risk,Footnote 18 that it is thereby contributing to the commission of international crimes by wrongdoers who receive its aid.
B. Knowing Contribution in International Criminal Law
Those, then, are two scenarios—situations of mitigating complicity and general complicity. The next step is to understand, as a matter of law, why the problem of the virtuous accomplice arises. In the first place, aiding and abetting in customary international criminal law is defined by a wide actus reus—any conduct that assists, encourages, or lends moral support to the perpetration of the principal crime.Footnote 19 Beyond this, the key driver of the problem is that under customary international law, knowledge that one is contributing to the principal crime is sufficient for liability as an aider and abettor. That is to say that the classic debate between knowledge and purpose—intention in the narrow sense—in the mens rea of complicity is resolved in favour of knowledge.Footnote 20 To assert this in such bare terms is to skip over reams of judicial and academic debate. But, for present purposes, following the decisions in Taylor, Šainović, and Stanisić, the orthodox view is that knowledge is enough to inculpate an accomplice.Footnote 21
On its face, Article 25(3)(c) of the Rome Statute appears to take a different approach—that aiders and abettors must act ‘[f]or the purpose of facilitating the commission of such a crime’.Footnote 22 This provision has prompted extensive debate.Footnote 23 However the ICC resolves the provision's meaning,Footnote 24 it appears that in many cases the same problem will arise under the provision that follows, Article 25(3)(d). That provision inculpates those who contribute to the commission of international crimes by a group of persons acting with a common purpose. Here, as set out in Article 25(3)(d)(ii), knowledge of the intention of the group to commit the crime is sufficient.Footnote 25 Given that many, if not the majority of, international crimes are committed by groups acting with a common purpose, the inculpation of knowing accomplices under Article 25(3)(d) opens up the possibility of the liability of virtuous accomplices.
Why, then, does this standard—a standard of knowing contribution—create the problem of the virtuous accomplice? It creates the problem because a standard of knowledge, by its very design, inculpates assisters whose reason for acting—whose reason for providing aid—is not to further the wrongful acts of the principal. More specifically, in basing liability on knowledge, rather than purpose, the rule renders irrelevant to the legal inquiry even the most virtuous intentions of the putative accomplice.Footnote 26
Two other doctrinal points constitute the problem. First, and of particular relevance to cases of mitigating complicity, it is no defence for the humanitarian actors to argue that the principal crime would have happened anyway without their contribution. In Blaškić, the Appeals Chamber of the ICTY held that ‘proof of a cause-effect relationship between the conduct of the aider and abettor and the commission of the crime, or proof that such conduct served as a condition precedent to the commission of the crime, is not required’.Footnote 27 Just as international criminal law denies to an arms dealer a potential causal defence based on the idea that ‘if I didn't do it someone else would have’Footnote 28, so it would not allow a humanitarian actor to deny the causal effect of its contribution on this basis.Footnote 29
Second, and of particular relevance to situations of general complicity, international criminal tribunals have sometimes required something less than actual knowledge on the part of the putative accomplice. Stewart argues that ‘international criminal justice here tolerates a type of doublespeak, claiming knowledge but applying recklessness’.Footnote 30 A good example is the Trial Chamber of the Special Court for Sierra Leone in Taylor, which held that aiding and abetting requires that the accused ‘knew that his acts or omissions would assist the commission of the crime, or that he was aware of the substantial likelihood that his acts would assist the commission of the crime’.Footnote 31
As a matter of principle, this slide to foresight of risk ought to be resisted.Footnote 32 As a matter of legal jeopardy for humanitarian actors, this lower threshold has the potential to draw more situations of assistance into the ambit of international criminal liability.
C. Raising the Mental Element?
By way of aside, it is worth pausing on what might seem a tempting solution to the problem. If the problem is that a fault element of knowledge captures even virtuous accomplices, shouldn't the fault element be raised to a standard of purpose, whether through interpretation or amendment? If purpose were the standard, liability would arise only where the accomplice intends, in the narrow sense, that the principal commit the crime. In civilian legal terminology, the fault element would be limited to situations of dolus directus in the first degree—situations in which facilitating the principal's wrong is the accomplice's reason for acting.Footnote 33 Knowing assistance would be insufficient for liability as an accomplice to international crimes.
This approach would certainly solve the problem of the virtuous accomplice, but it would be to throw the baby out with the bathwater. In solving the problem, it would leave the rule too narrow. As a general point, at least, knowingly contributing to another's wrongdoing is blameworthy.Footnote 34 This is not a claim that knowing contribution is either indistinguishable from, or equivalent to, purposeful assistance, but rather a claim that in and of itself knowingly contributing to another's wrongdoing ought to be proscribed.Footnote 35 It ought to be proscribed because to knowingly assist another actor's commission of a wrong is to pay insufficient attention to the ways in which our acts contribute to the acts of others.Footnote 36 Such assistance is marked by, as Gardner and Jung put it, a set of ‘advertence-based omissive deficiencies of practical wisdom, failures to give any or enough weight to known salient considerations in one's practical thinking’.Footnote 37 Moreover, the depth of such a deficiency is increased when the principal act is an international crime—conduct which, by definition, is extremely grave.
In addition, a standard based on knowledge has the potential to play a systemically important role in the international order. The promise is that concern for the wrongdoing to which their assistance contributes will become part of the day-to-day practice of States, international organisations, humanitarian actors, and individuals—turning their indifference or insufficient concern into attention.Footnote 38 Before acting, international actors may obtain legal advice, including advice as to their agents’ potential individual criminal responsibility for complicity. Non-governmental organisations can put before domestic courts and legislative committees reports that show evidence of routine or ongoing wrongdoing by those receiving assistance.Footnote 39 As a corollary, it is plausible that potential recipients—of development aid, or arms, or intelligence information—will be incentivised to act in conformity with the relevant primary rules of international law. In a related context, Lowe has argued that the ‘bureaucratisation of the monitoring of compliance with international law … would make a significant contribution to the entrenchment of the rule of law in the international community’.Footnote 40
This is to say that there are good reasons to prefer, at least as a general rule, a standard of knowing contribution in the fault element of complicity in international criminal law. We are thus left with the problem of the virtuous accomplice.
III. TWO CLARIFICATIONS ON THE SCOPE OF THE PROBLEM
Before turning to potential solutions, two clarifications on the scope of the problem may be helpful. The first is doctrinal; the second is practical. Both are particularly relevant to scenarios of general complicity, and together they minimise, but do not eliminate, the problem.Footnote 41
A. Substantial/Significant Contribution to a Specific Crime
Rules of complicity must specify the required connection between an assister's aid and the principal wrong. In international criminal law, the Appeals Chamber in Taylor held that the customary rule demands that the assistance had ‘a substantial effect on the crime’.Footnote 42 This is consistent with the practice of other tribunals.Footnote 43 As is often noted, the requirement of a certain connection between the assister's aid and the principal wrong is demanded both by the derivative nature of accomplice liability and the bare fact that assistance can capture an extremely broad range of conduct.Footnote 44
For present purposes, there are two relevant points here. On one hand, international criminal law knows no generalised liability for providing aid to a particular group, as opposed to a crime committed by members of that group. This point is evident in contrasting accomplice liability under international criminal law with the federal prohibition on providing material support to designated terrorist organisations in the United States. 18 USC 2339B(a)(1) provides:
Whoever knowingly provides material support or resources to a foreign terrorist organization, or attempts or conspires to do so, shall be fined under this title or imprisoned not more than 20 years, or both, and, if the death of any person results, shall be imprisoned for any term of years or for life.Footnote 45
Rather than demanding a connection between the support and a specific crime, the provision works through a process of organisational designation. Under the Immigration and Nationality Act, the Secretary of State is authorised to designate a foreign group as a terrorist organisation.Footnote 46 On designation, and where there is knowledge of that designation, almostFootnote 47 all forms of assistance are prohibited. In other words, the statute is not really a form of accomplice liability for participation in a specific crime, but rather a proscription of support to a designated group.Footnote 48 That international criminal law does not work in this way alleviates at least some risk for humanitarian actors.
On the other hand, the requirement in customary international law of a substantial contribution, and requirement under Article 25(3)(d) ICC Statute of a significant contribution, might assist in addressing the potential over-breadth problem in situations of general complicity.Footnote 49 In Taylor, the Appeals Chamber held that the ‘fungibility of the means may establish that the accused is not sufficiently connected to the commission of the crime’, but qualified fungibility as relevant only to situations where the principal crime was an isolated act.Footnote 50 Beyond this bare statement, it is probably fair to say that this requirement has not been properly fleshed out in case law,Footnote 51 with the most common refrain simply being that the matter must be addressed on a case-by-case basis.Footnote 52 Nonetheless, this requirement has the potential to exclude certain acts of assistance that have a minimal effect on the principal crime.Footnote 53
B. Practical Measures of Risk Mitigation
If the first clarification is doctrinal, the second is practical. In situations akin to the Rwandan example mentioned above, humanitarian actors may be able to take a range of practical measures that reduce the risk that their aid will contribute to the principal wrong.Footnote 54 It is a mistake to talk in general terms about the provision of aid or assistance in a humanitarian emergency. To do so obscures certain practical choices available to humanitarian organisations. These practical choices include questions as to whom precisely within a camp or territory assistance is given; whether any form of ongoing collaboration or monitoring is offered; and whether measures are put in place for withdrawal of support if the information available to the humanitarian actor changes.Footnote 55
Practical measures of this kind have received increased attention in other areas of international law. For instance, the Arms Trade Treaty requires State parties to assess and mitigate the risk that exports could be used to commit or facilitate serious violations of international law.Footnote 56 In the context of expulsion and deportation, assurances and monitoring of treatment have been used by States to mitigate the risk of ill-treatment by recipient States, thus taking the matter outside the scope of the non-refoulement rule.Footnote 57 In peacekeeping, the United Nations Due Diligence Policy on UN Support to non-UN Security Forces requires a detailed risk assessment before a UN entity provides support to national or regional security forces.Footnote 58 The point is not that humanitarian actors will always be able to take measures to lower the risk that its aid will contribute to an international crime. But, in at least in some cases, practical measures offer a way out of the problem.Footnote 59
IV. POTENTIAL SOLUTIONS
To sum up the position so far: international criminal law imposes accomplice liability where an individual makes a knowing contribution to another's international crime. That is to say, it does not require that accomplices act with the purpose that the crime be committed. This is justifiable as a general position but is potentially overbroad in a narrow set of cases. In those cases, proper attention to the requirement of substantial contribution might minimise the problem, and a range of practical measures of risk mitigation available to humanitarian actors may reduce it further. But the problem remains. If the ICRC provides buses to move people safely out of a city in the course of unlawful deportation, are its members at risk of individual criminal liability as accomplices to the war crime? If the only way for UNHCR to access a refugee camp to treat a cholera outbreak is to provide assistance through (and to) suspected perpetrators in de facto command of the camp, are its members at risk of criminal liability for subsequent crimes to which that assistance contributes?
To answer that question, this section looks to domestic criminal law and practice. Underlying the turn to domestic law is the intuition that whenever complicity rules inculpate accomplices on the basis of knowledge, as well as purpose, there is a risk of overly broad liability. It discusses three ways that legal systems have dealt with such a risk—two doctrinal and one institutional. At the same time, it considers their potential application in international criminal law.
A. Justified Complicity—A Defence
One way to think about the problem is to work from the starting point that humanitarian actors are faced with a genuine dilemma in these situations. We can grant that knowingly contributing to the commission of an international crime is wrong, but condition that judgment with the possibility that such a knowing contribution may, all things considered, be justifiable. This is Goodin and Lepora's argument on the potential morality of complicity in at least some situations of this kind:
Being complicit with the wrongdoing of another, doing something that potentially contributes causally to that wrongdoing, might be the best thing to do in a bad situation. … [I]t would be far better all around if there [were] no situations in which contributing to the wrongdoing of others would actually be the best thing to do to improve the bad situation. But if those bad situations exist whether or not we engage with them—as surely, and sadly, they do—then it is better for us to do what we can to make them better than to stand aside.Footnote 60
To be clear, this is not an argument of pure consequentialism. The argument is not that an accomplice who shares the principal's commitment to the crime is justified in contributing where the good brought about by that wrongdoing outweighs its evil. Rather, it is an argument that looking to that balance might be appropriate where, despite knowingly contributing to the principal wrong, the putative accomplice's reason for action is justifiable.
Turning to domestic law, we can find a doctrinal structure like this in a number of different places.Footnote 61 To give two examples from England and Wales—first, Section 50 of the Serious Crime Act of 2007 establishes a defence of ‘acting reasonably’ to a set of inchoate crimes proscribed by the Act, a set which includes the offence of encouraging or assisting an offence believing it will be committed.Footnote 62 Second, Section 73 of the Sexual Offences Act of 2003 provides that a person is not liable as an accomplice to certain sexual offences against children if he acts for certain specified purposes relating to protecting the child's well-being.Footnote 63 Section 73 removes, for instance, any riskFootnote 64 of liability for a doctor who provides contraception to an underage girl in the knowledge that he is thereby contributing to a sexual offence by another party.
The value of this approach lies in its creation of a doctrinal structure ‘better calibrated to moral nuance’Footnote 65—inculpating many knowing contributors while exculpating a small subset of them. The question, then, is whether international criminal law might accommodate a defence along these lines.Footnote 66 Taking Article 31 of the Rome Statute as orthodoxy,Footnote 67 it appears that it might. Article 31(d) provides:
In addition to other grounds for excluding criminal responsibility provided for in this Statute, a person shall not be criminally responsible if, at the time of that person's conduct:
The conduct which is alleged to constitute a crime within the jurisdiction of the Court has been caused by duress resulting from a threat of imminent death or of continuing or imminent serious bodily harm against that person or another person, and the person acts necessarily and reasonably to avoid this threat, provided that the person does not intend to cause a greater harm than the one sought to be avoided. Such a threat may either be:
(i) Made by other persons; or
(ii) Constituted by other circumstances beyond that person's control.Footnote 68
This provision takes together the often-distinguished defences of duress and necessity.Footnote 69 As Moran explains, key to its application are four elements: (i) the seriousness and (ii) imminence of the threat; (iii) the necessity and reasonableness of the defendant's act; and (iv) the absence of an intention to cause greater harm than the one the defendant sought to avoid.Footnote 70 Let us assume that both (i) seriousness and (iv) the subjective requirement are met, leaving imminence and the necessity and reasonableness of the response.
Although contextual assessment is inescapable, it is possible that in some cases humanitarian actors may be able to meet these requirements and justify their putative complicity. Imminence fits well with the scenarios set out above, though may play an important limiting role in other cases.Footnote 71 Necessity requires that no other course of action be available and, together with reasonableness,Footnote 72 denotes the choice-of-evils requirement in a defence of necessity.Footnote 73 As to mitigating complicity, the humanitarian actor's (knowing) participation in the principal's crime is outweighed by the legal interests served by its acts of assistance. In the deportation case, for instance, the provision of safe transport might be a necessary and reasonable way to avoid the threat to lives that would ensue in a mass deportation without that support. Here, the fact that the humanitarian actor is acting to alleviate harm to the very class of people whose interests are protected by the underlying rule makes the application of the test more straightforward.
Cases of general complicity are trickier, even assuming imminence is shown. Take the Rwandan case mentioned above—here the humanitarian actor knowingly contributes to the commission of international crimes against other individuals in order to enable the provision of relief to a different group.Footnote 74 This is a classical choice-of-evils situation. Key to resolving such a case will be an assessment of the scale and nature of the relevant principal crimes to which the assistance contributes and the underlying interests they protect, compared with the harm threatened by the emergency. In grave situations akin to that set out above—a cholera outbreak threatening the lives of tens of thousands of people—the requirement of the defence may be met. As in domestic law, however, there is always the risk that a claim of necessity is not accepted.
B. Objective Attribution
The previous section discussed a doctrinal solution based on a defence of necessity—the idea of justified assistance. A number of legal systems limit the scope of accomplice liability in different ways. Here, the idea of objective attribution of conduct builds normative considerations into the attribution inquiry. On this approach, it is not that the accomplice's (causally relevant) assistance is justifiable, as in the previous section, but that the conduct does not count as assistance in a legally relevant sense at all.
To take the German approach specifically, and brushing over various complexities, two lines of analysis are relevant.Footnote 75 First, there is a long-standing controversy in German criminal law about so-called ‘neutral acts’—everyday, often commercial, acts that make a (factual) contribution to another actor's crime.Footnote 76 Here, German scholars have proposed, and German courts used, different tests that serve to limit liability.Footnote 77 One of these focuses on the creation of risk on the objective level. Here, as Hefendehl explains, ‘the value of the legally protected interest on the one hand and the immediacy of the risk that exists for this interest on the other are used as parameters for a decision between two conflicting spheres of liberty, the one of the acting person and the one of the victim’.Footnote 78 Second, a limitation on attribution is generally accepted in German law in situations where the assistance reduced the risk of harm to the underlying legally protected interest.Footnote 79 Even if injury were caused to the victim, the defendant would not be liable as an accomplice to the assault on the basis that her conduct reduced the risk of injury.Footnote 80
In either case, these ‘normative, value-based’Footnote 81 considerations operate at the attribution stage to close off the potential liability of the assister. The key question, then, concerns their applicability—separately—in international criminal law. To start with a limitation based on the idea of neutral acts, Ambos, in particular, has argued in favour of the adoption by international criminal tribunals of a model of objective attribution for complicity that excludes certain neutral acts from the ambit of liability.Footnote 82 An argument along these lines has found some support in certain Separate and Minority Opinions at the ICC in respect of liability under Article 25(3)(d) ICC Statute. In her Separate Opinion in Mbarushimana concerning the requirement of assistance in that provision, Judge Fernández de Gurmendi held:
I am not persuaded that such contributions would be adequately addressed by adding the requirement that a contribution be significant. Depending on the circumstances of a case, providing food or utilities to an armed group might be a significant, a substantial or even an essential contribution to the commission of crimes by this group. In my view the real issue is that of the so-called ‘neutral’ contributions. This problem is better addressed by analysing the normative and causal links between the contribution and the crime rather than requiring a minimum level of contribution.Footnote 83
Moreover, in Katanga, Judge van der Wyngaert, concerned with the ‘extremely low’ thresholds in the actus reus and mens rea for liability under Article 25(3)(d) ICC Statute, proposed that the ICC take into account whether the accused's assistance was specifically directed at the criminal part of the principal group's activities.Footnote 84 Here, of course, Judge van der Wyngaert was drawing from the decision of the Appeals Chamber of the ICTY in Perišić, which held specific direction to be part of the actus reus of aiding and abetting.Footnote 85 What unites the opinions of Judges Fernández de Gurmendi and Van der Wyngaert is a concern to limit liability in situations of knowing, but neutral, contribution.
It may be the case that the ICC does, in time, develop a normatively driven limitation on the idea of assistance under Article 25(3)(d) ICC Statute.Footnote 86 However, in respect of aiding and abetting under customary international law, no such limitation exists. To return to Perišić itself and specific direction at the ICTY, the subsequent Appeals Chamber decisions in Šainović, Popović, and Stanišić firmly reject any such element in the actus reus of aiding and abetting as well as, implicitly, any other normative limitation in the attribution of conduct to the defendant based on the generality of her aid.Footnote 87 Moreover, in Taylor at the Special Court of Sierra Leone, whether an additional limitation existed for neutral acts was at issue in the appeal, with the defence explicitly citing the German approach with a view to establishing a distinction between neutral and other forms of assistance.Footnote 88 Here, the Appeals Chamber rejected any such limitation, reasoning that it is simply the effect of the assistance that matters.Footnote 89
This is to say that as customary international criminal law stands, there is no limitation on aiding and abetting liability based on the idea of neutral acts. This point is of much wider significance, but for present purposes it means that there is no solution along these lines to situations of general complicity by humanitarian actors. As to situations of mitigating complicity, this conclusion probably holds, but is less certain for the simple reason that it has never been tested.Footnote 90 It may be that in the right case an international tribunal would adopt a limitation based on the fact that the putative accomplice's assistance actually reduced the harm to the victim. As it stands, however, no such approach exists.
C. An Institutional Solution
The two previous sections considered doctrinal solutions found in domestic criminal law. The present section discusses an institutional solution found in England and Wales. For the most part,Footnote 91 the problem of the virtuous accomplice is not resolved in England and Wales through doctrinal criminal law. Instead, it resolves the problem through the institutions of the criminal justice system. One way to explain the institutional approach is from a slightly obscure angle—the old case of Gillick in England and Wales.Footnote 92 In Gillick, the central issue was whether a doctor could lawfully prescribe contraception to a girl under the age of 16 without the consent of her parents. The decision of the House of Lords is most widely known for its establishment of the standard of Gillick-competence—that the parental right to determine whether a child under 16 ‘will have medical treatment terminates if and when the child achieves a sufficient understanding and intelligence to enable him or her to understand fully what is proposed’.Footnote 93 But the decision also excited close attention amongst criminal lawyers and, in particular, criminal lawyers interested in the mens rea of complicity.
The issue of complicity arose in the plaintiff's submission that the policy of the Department of Health and Social Security would render the doctor liable as an accessory to unlawful sexual intercourse with a girl under the age of 16. More specifically, the plaintiff averred that:
[A] doctor who provides contraceptive advice or treatment knowing that unlawful sexual intercourse will thereby be facilitated, makes himself an accessory to each offence of unlawful sexual intercourse committed under section 6 of the Sexual Offences Act 1956 by providing the means whereby the criminal act is performed … Analogous situations would be where a person supplies poison to a potential murderer knowing that the poison was intended for the commission of murder, or where a person supplies a car knowing that it is intended to be used in making a quick escape from a burglary or robbery.Footnote 94
In the House of Lords, the plaintiff's argument that the policy was unlawful failed. Although the precise elements of the judgment are not relevant for present purposes, contrasting academic responses reveal something interesting about how England and Wales, at that time, dealt with the problem.
On one hand, Dennis argued that Gillick stood for the proposition that accomplice liability only arises where aid or encouragement is provided with the purpose of facilitating the offence. He proposed that the ‘doctor does not prescribe contraceptives in order that intercourse shall take place, but in order that health shall be preserved and unwanted outcomes avoided’.Footnote 95 On this account, knowledge of one's contribution is insufficient for liability.Footnote 96 On the other hand, in his Hamlyn Lectures JC Smith argued that Gillick actually amounted to the acceptance of a concealed defence of necessity. Starting from the general principle that ‘anyone who knowingly assists or encourages the commission of a criminal offence is guilty’,Footnote 97 Smith points to the general policy reasons that underpin the House of Lords’ decision. In other words, ‘the aiding and abetting of the offence is justified as being a lesser evil than the alternative—that the girl will or may have sexual intercourse without contraception and with the risk of pregnancy’.Footnote 98
Neither Dennis nor Smith's reading of the case is quite correct. To be liable in England and Wales, an accomplice must have the intention to assist.Footnote 99 In accordance with the approach to intention more generally, foresight of virtual certainty—that is, knowledge—is not a species of intention but rather a basis on which the jury may find intention.Footnote 100 That the defendant knew she was contributing to the principal crime is not itself sufficient for liability, but may be the basis on which a jury finds that she intended to do so.Footnote 101
To explain that another way, it is the procedural context of Gillick that caused the difficulties—it was a civil application for a declaration rather than a criminal trial. In the case, the House of Lords was required to decide whether a doctor's knowing provision of contraceptives made him an accomplice to the principal crime. But doctrinal criminal law in England and Wales does not answer a question framed in those terms. Instead, juries do. Indeed, the underlying issue in Gillick is sometimes cited as an example of the ‘moral elbow-room’Footnote 102 that is left to juries in making findings as to the existence of intention.Footnote 103 Some knowing assisters will be found guilty; others won't, with a crucial distinguishing feature likely to be the wider morality of the specific individual's reasons for providing assistance.Footnote 104
The institutional solution of a jury is obviously unavailable at the international level, but the case does point us towards the availability of another institutional tool—that of prosecutorial discretion. Amongst other functions, prosecutorial discretion allows legal system to take into account the ‘fine-grained moral evaluations and distinctions’ demanded by justice.Footnote 105 Again focusing on the Rome Statute, Article 53 grants to the Prosecutor discretion not to proceed with a prosecution where it would not be in the interests of justice.Footnote 106 Of course, reliance on an institutional solution of prosecutorial discretion is not entirely satisfactory,Footnote 107 and also opens up the possibility of discriminatory application. Moreover, for responsible organisations, the key question is whether they are complying with their international obligations rather than whether their members will be prosecuted before an international tribunal. But, in exceptional cases, prosecutorial discretion might provide a second-best solution.
V. WIDER IMPLICATIONS
The preceding analysis placed the dilemma in the hands of humanitarian actors—in the ICRC or UNHCR. In addition to illustrating the practical reality of the problem, the focus on actors of this kind also serves to bring home the genuineness of the moral dilemma. For the most part, we have few doubts about the virtue of their intentions, and we are perhaps more easily convinced that some forms of complicity might be, in the end, justifiable. Moreover, to focus exclusively on humanitarian actors as a group diminishes any risk that allowing in a balance-of-evils justification will overwhelm the underlying rule.Footnote 108
Unless, however, we are willing to countenance status or activity-based exemptions from general law,Footnote 109 doctrinal solutions available to one actor are available to others. As noted above, situations of general complicity underlie the controversy around the criterion of specific direction in Perišić. Consider, here, the transfer of arms or financial aid by a State to a non-State group fighting a repressive regime and the potential criminal liability of the transferring State's officials. Moreover, it is not just situations of general complicity that are of potential wider application. Consider the provision of targeting assistance in a situation where the recipient is violating the principle of distinction in the law of armed conflict. Is it possible for the assisting StateFootnote 110 and its officials to argue that though it is knowingly contributing to war crimes, its assistance and training are actually minimising the extent of the indiscriminate attacks?
These are questions beyond the scope of this article. To a large extent, the requirement of imminence, properly interpreted, as well as strict attention to necessity and reasonableness, would minimise the risk that the defence would unduly undermine the value of a complicity rule based on knowing assistance. Moreover, it will remain the case, as always, that a party invoking necessity takes the chance that its claim will be rejected. Nonetheless, here, as is often the case, the acceptance of a justificatory defence comes with risks to the interests protected by the relevant rule.
VI. CONCLUSION
A standard of knowledge in the mens rea of complicity in international criminal law, while generally justifiable, creates a risk of over-inclusivity in some situations of humanitarian assistance. As a backstop, prosecutorial discretion might filter out some of these cases, but a procedural solution of this kind is inferior to resolving the problem doctrinally. In this respect, a defence of necessity provides an appropriate way forward. We ought to accept, albeit in exceptional circumstances, that knowingly contributing to an international crime is the right thing to do.
