Archive of posts for category
Africa

Why Art. 4(h) of the AU’s Constitutive Act Does Not Support UHI

by Kevin Jon Heller

I like much of what Jennifer Trahan says in her recent post about the permanent veto. But I would take issue — unsurprisingly — with her claim that “there is in fact more past precedent for reliance on the doctrine of humanitarian intervention than is often recognized, including not only NATO’s 1999 intervention in Kosovo, but also African and Arab practice.” There is no need to belabour the Kosovo argument, which I have already criticised in my response to Harold Koh. Instead, I want to focus in this post on the idea that African practice provides support for a customary right to unilateral humanitarian intervention (UHI). That idea seems to be popping up with more regularity these days, as those who believe UHI should be legal cast about for a plausible argument to that effect. Marc Weller, for example, recently made a similar claim in his intemperate response to Marko Milanovic’s recent post at EJIL: Talk! on the illegality of the Syria attack:

Moreover, the African Union has in fact formally committed itself to the doctrine in its own constitutive treaties–an act that destroys the presumption that the use of force for humanitarian purposes would necessarily violate the prohibition of the use of force.

Weller is referring to Art. 4(h) of the Constitutive Act of the African Union, which has been ratified by all 53 African states have signed the Act. Art. 4(h) permits the AU’s Assembly of Heads of State and Government of the Union to authorise the use of force against a Member State that is responsible for the commission of international crimes:

The Union shall function in accordance with the following principles… (h) the right of the Union to intervene in a Member State pursuant to a decision of the Assembly in respect of grave circumstances, namely: war crimes, genocide and crimes against humanity.

Properly understood, however, Art. 4(h) of the Constitutive Act provides no support whatsoever for the legality of UHI. To see why, it is useful to begin by explaining why the case for a customary rule permitting UHI is not supported by humanitarian uses of force authorised by the Security Council under Chapter VII of the UN Charter. The reason is straightforward: the essence of UHI is that the state being attacked does not consent to force being used on its territory, and all states consent to the possibility of the Security Council authorising the use of force on their territory when they ratify the Charter. The relevant provisions are Art. 24(1) and Art. 25:

Article 24

  1. In order to ensure prompt and effective action by the United Nations, its Members confer on the Security Council primary responsibility for the maintenance of international peace and security, and agree that in carrying out its duties under this responsibility the Security Council acts on their behalf.

Article 25

The Members of the United Nations agree to accept and carry out the decisions of the Security Council in accordance with the present Charter.

Because of Art. 24(1) and Art. 25, humanitarian intervention authorised by the Security Council under Chapter VII is by definition legally consensual — even if the state being attacked is politically opposed to the use of force on its territory.

For similar reasons, humanitarian intervention authorised by the AU’s Assembly of Heads of State and Government of the Union pursuant to Art. 4(h) of the Constitutive Act cannot be non-consensual. By ratifying the Constitutive Act, all AU members have empowered the Assembly to authorise the use force on their territory to stop the commission of international crimes. They may not like that use of force, but they have consented to it.

Humanitarian intervention authorised by the AU, therefore, no more supports the customary legality of UHI than humanitarian intervention authorised by the Security Council. In each case, the use of force in question is authorised by a supranational organisation acting on the basis of the consent of the parties that have ratified its founding treaty. That is the antithesis of the nonconsensual use of force — the constitutive feature of UHI.

Escalation of Violence in Bangui: Has MINUSCA Become Party to a Conflict in CAR, and What Would That Mean?

by Bianca Maganza

[Bianca Maganza is a PhD candidate in International Law and a Teaching Assistant at the Graduate Institute of International and Development Studies in Geneva.]

Some days ago, the United Nations Multidimensional Integrated Stabilisation Mission in the CAR (hereinafter, MINUSCA) got involved in heavy fire exchange with an armed group known as KM5 in Bangui, the capital of the Central African Republic (hereinafter, CAR). Security in CAR is rapidly deteriorating since some weeks, and accidents such as the one just mentioned are becoming, if not the norm, for sure no longer the exception. Indeed, it is not the first time that a situation of direct confrontation between the UN mission and the rebels arises. Although facts are not fully confirmed and the situation is still unfolding at the time of writing, it seems worth considering the consequences that the episode may entail under international humanitarian law (hereinafter, IHL).

The issue of the role played by IHL in the context of so-called peace operations is a much-debated and complex one (see, for example, the recently published Leuven Manual on the International Law Applicable to Peace Operations). Crucial aspects such as those related to the consequences of the application of IHL on issues of targeting and on the international prosecution of crimes committed in relation to the situation have been brilliantly covered elsewhere, and I will not discuss them here. The same is true for jus ad bellum considerations vis-à-vis the mandate and the evolution of traditional peacekeeping into “stabilisation” missions. While the main case study in this respect has so far been that of MONUSCO in the DRC, in particular as regards its Intervention Brigade, it has been rightly claimed that MINUSCA is the new frontier for such discussion.

In this post, I want to focus on a specific aspect of the situation, namely the possibility that – provided that some factual circumstances are reached – a multinational mission spearheaded by the UN becomes as such party to a conflict with the armed group against which it is fighting. This case is different from the one following the so-called support-based approach developed by Ferraro, which I am not discussing here. Although the fact that IHL may apply to situations in which a multinational operation is involved is nowadays accepted, states and international organisations are generally reluctant to admit that a peace operation has become party to an armed conflict. Such possibility has been often discussed but frequently let sink through the lines and rarely tackled head on. If, on the one hand, the capacity of a multinational operation to become party to the conflict is often taken for granted, the concretisation of such a possibility is on the other hand never explicitly acknowledged in terms of its consequences.

The facts that recently took place in Bangui provide the occasion to discuss the legal significance of such a case.

The facts

After the failure of a first attempt undertaken some days before, in the night between 7 and 8 April 2018, what was announced as joint law enforcement operation of MINUSCA, the Central African Armed Defense Forces (FACA) and the Internal Security Forces (ISF) aimed at disarming and arresting some criminal groups in the PK5 area of Bangui turned into violent fire confrontation between the armed groups and the mission.

According to the press statement released by MINUSCA, at a given moment, “heavily armed elements of criminal groups in the area deliberately fired on international forces, prompting a return of fire to push these elements back”. According to other sources, it is the MINUSCA contingent that entered the area shooting in the first place. As a result, eight peacekeepers were wounded, and one died. Some civilian casualties are also reported. In any case, it is quite clear that the operation was not a success for MINUSCA: for the second time in a matter of few days, the contingent had to leave the area and had been the object of subsequent attacks.

A conflict between MINUSCA and the armed group?

Regardless of who started the fire exchange, it is more than legitimate to ask whether the threshold for IHL applicability has been reached with the incident of 8 April – if not even before. In fact, despite the careful employment of the language of law enforcement in MINUSCA’s press releases – referring to “criminal groups”, “gangs” and “disarmament operations” – the reality on the ground seems to point to another direction.

Working on the hypothesis that a conflict, if any, ought to be a NIAC due to the involvement of non-state armed groups against the contingent spearheaded by the UN, the situation has to be analysed under the two criteria of intensity of hostilities and organisation of the parties as developed by the ICTY in the Boškoski case.

Intensity

Applying the first prong of the test to the single incident under analysis, video footage clearly shows MINUSCA troops engaging the area through the use of heavy weaponry. The presence of civilian casualties and fleeing civilians, if confirmed, is a further factor to consider when assessing the intensity of the armed confrontation. The episode must be assessed against the broader background of the situation in CAR. The general secrecy regarding the number, details and nature of MINUSCA operations cannot hide the fact that incidents like the one discussed here are not isolated but had already happened and keep happening in the country. The recent security deterioration in the area and the repeated, unsuccessful attempts by MINUSCA to seize the neighbourhood testify to the seriousness of the situation. The situation remains unchanged after the incident of 8 April, and following the main incident the mission has been the object of subsequent attacks in another area of the city. It is highly probable that the mission will attempt yet another operation to gain control of the neighbourhood – possibly, deploying more force to be able to finally overcome the enemy.

In light of the above, it seems at least plausible to argue that the intensity level has been reached either regarding the single attack under analysis – in fact, a NIAC can be a matter of hours or days (see, in addition to the case law of the ICTY, Abella v. Argentina, known as “La Tablada” case, before the Inter-American Commission of Human Rights) – or as a result of the series of episodes involving MINUSCA and the armed group since some time. Assuming that this is the case, for the sake of argument, the next step is an assessment of the organisation of the parties involved.

Organisation of the parties

That of organisation of the parties contains in itself the two other interdependent criteria of control over the troops and capacity to comply with the rules of war – both in abstract, through a disciplinary system and chain of command, and through the concrete possibility of respecting the most basic principles such as those of distinction and precaution. These seem, to me, the features that differentiate a collective entity having the capacity to become party to a conflict from a random group of individuals involved in armed violence.

Most of the analyses of situations of multinational operations potentially involved in an armed conflict stop at the assessment of the organisation of the armed group for the sake of establishing the existence of a NIAC. However, it seems to me that the only way to fully satisfy the test is to apply the analysis to both sides. As evident as the result may be, to be able to claim that MINUSCA has become party to a conflict with KM5, one shouldn’t stop at whether the armed group is sufficiently organised to reach the threshold of a NIAC but also ought to address the features of the mission itself.

In the case at stake, it seems undeniable that KM5 shows and deploys a considerable amount of organisation. To a certain extent, it can be argued that it even controls the neighbourhood known as PK5: it was indeed following continuous requests by the residents of the area that MINUSCA decided to act to disarm the armed group and try to “liberate” PK5. As we have seen, MINUSCA attempted to take the neighbourhood back more than once, without success. As for elements of control, KM5 is organised in auto-defence patrols, therefore showing capacity to coordinate action and logistics, and acts in execution of the orders given by its leader, Nimery Matar (known as “Force”), who even recently released a video declaring to be ready to fight in case MINUSCA would have dared entering the area he controls. While we would need more facts to weigh the group’s potential compliance with IHL, the existence of a command structure can arguably point to such abstract capacity.

When it comes to the UN mission, all three accounts seem equally warranted. The mission has by definition a clear organisational structure and a chain of command. Moreover, respect for IHL is part of its mandate and generally warranted by the 1999 UN Secretary General’s Bulletin. Even though the Status of Forces Agreement (hereinafter, SOFA) between MINUSCA and CAR is not publicly available, respect for IHL is included in the model UN SOFA and usually reflected in single SOFAs concluded with host countries.

MINUSCA as party to a conflict

The purpose of this post is not directly that of taking a position on whether or not the situation in CAR amounts to an armed conflict between MINUSCA and the armed group against which it is fighting. Even though the facts reported above seems to point to a considerable level of intensity, full confirmation of what happened is still awaited and sources are scattered and not always reliable. My point is rather that, in case the required intensity threshold is reached, both sides have the capacity to become parties to a conflict.

The main consequence of considering MINUSCA as party to an armed conflict – a NIAC, in this case – would be that the mission as a whole is bound to respect at least the provisions of CA3 plus any other customary international law rule deemed applicable to the situation. More specifically, MINUSCA would have to respect and ensure respect for IHL by its armed forces and other persons or groups de facto acting on its instructions, or under its direction or control (Rule 139) and train its armed forces in IHL (Rule 142). These obligations seem particularly relevant in light of the problems often posed by issues of legal interoperability in the context of multinational operations. Seeing the mission as one single entity for the sake of IHL would allow deeming it responsible for the training and subsequent conduct of its peacekeepers independently of – but not in contradiction with – the repartition of command and control between the international organisation and troop contributing countries. Concretely, this would mean that, in addition to the obligations of each troop contributing country, the UN should take the responsibility of substantively training the contingent and making sure it respects and applies the same rules of IHL as whole, instead of a patchwork of obligations deriving from the law applicable to single member states. Arguably, such a scenario does not differ too much from what already happens in practice through the joint training of peacekeeping missions before deployment, and the respect of a basic, uniform set of IHL rules is the aim of the Secretary General’s Bulletin. However, the idea that those obligations might directly derive from the status of party to the conflict under IHL sounds very unlikely to be accepted for the consequences it might entail when it comes to issues of attribution and allocation of responsibility for wrongdoings between the member states and the international organisation. For all that, it seems undeniable that acting under a unified legal framework has the potential for being a game changer at least in terms of prevention of violations of IHL.

No less important consequences derive from the potential role of MINUSCA as a party to the conflict under international criminal law. As aptly noted by Labuda, although the UN keeps denying the existence of an armed conflict between the mission and the armed groups in CAR, it at the same time reiterates that any attack against peacekeepers may amount to a war crime. Leaving aside the contradictory nature of statements of this kind, what is interesting is that incidents such as the one of 8 April are likely to fall under the jurisdiction of the recently established Special Criminal Court for CAR. Further judicial pronouncement is needed to shed some light on the features of the crime of attacking peacekeepers in a situation of armed conflict. The premise that because of the status of a peacekeeping mission as party to the conflict members of its military branch could be targeted as forming part of the armed forces thereof has never, so far, been accepted in the case law of international criminal tribunals. Will the Special Court choose to do so, indirectly saying the unsaid? We will need to wait and see if the whispered possibility of considering a UN mission as a party to the conflict will be finally made explicit in such a way.

A Letter to Israel About Its Plans to Forcibly Deport Africans

by Kevin Jon Heller

Opinio Juris readers might be interested in this letter from GLAN Legal — the Global Legal Action Network — to the Presidents and Attorneys General of Israel and Uganda. It was written by Itamar Mann, Yannis Kalpouzos, and Omer Shatz, with input from me. Here is the introduction:

The Global Legal Action Network (GLAN) is an organization of lawyers initiating transnational human rights litigation around the world. Our focus is on cases in which “developed” countries are responsible for violations occurring in “developing” countries. We write to respectfully warn both Israel and Uganda that the forcible deportation plan currently being discussed for Eritrean and Sudanese asylum seekers living in Israel may amount to a crime within the jurisdiction of the International Criminal Court (ICC).

The letter then proceeds to lay out the basics of the argument — which seems pretty unimpeachable to me. Note the reliance on the OPT’s recent argument concerning deportation from Myanmar to Bangladesh!

Implications of the Rohingya Argument for Libya and Syria (and Jordan)

by Kevin Jon Heller

In my previous post, I offered three cautionary thoughts about the OTP’s decision to ask the Pre-Trial Division to hold that the ICC has jurisdiction over Myanmar’s deportation of the Rohingya to Bangladesh. In this post, I want to offer a few thoughts on what a successful outcome would mean for refugee crises elsewhere — particularly in Libya and Syria.

We can start with the relatively easy case: Libya. Because of UNSC Res. 1970, the ICC has jurisdiction over any international crime committed on the territory of Libya from 2011 onwards. If the OTP’s theory of deportation is correct, it would be able to prosecute both (1) deportations from Libya committed by the Libyan government and armed forces, and (2) deportations into Libya from neighbouring states that are not members of the ICC — namely, Egypt, Sudan, and Algeria. Fatou Bensouda has already signalled interest in prosecuting “migrant-related crimes” committed by Libyans in Libya. A positive outcome to the OTP’s request in the Rohingya situation would make it possible for her to prosecute Egyptian, Sudanese, and Algerian nationals as well.

The more interesting case is Syria, given that Syria is not a member of the ICC and there is little chance the Security Council will refer the situation in Syria to the Court anytime soon. There are two scenarios worth discussing here: (1) deportations by the Syrian government or by various Syrian rebel groups into a neighbouring state; and (2) deportations by the Syrian government or by various Syrian rebel groups onto the high seas.

Going in reverse order, the ICC would have no jurisdiction over any deportation from Syria onto the high seas. There are two possible situations here: where the deported civilians drown on the high seas, and where the deported civilians make it to an ICC member-state such as Italy or Germany. In the first (profoundly sad) situation, no essential element of deportation (or of any other international crime) would have taken place on the the territory of a state party. And in the second situation, the civilians would not have been deported “directly” into the territory of a member-state — they would have been directly deported onto the high seas and only “indirectly” deported into the territory of a state party. Deportation’s “essential element” of crossing an international border would thus have taken place in Syria and on the high seas — not on the territory of a member state. This is the importance of the OTP’s repeated insistence in its request (see paras. 4, 13, and 28) that only direct deportations activate the Court’s jurisdiction.

In the second scenario, where the civilians are directly deported into a neighbouring state’s territory, a positive outcome to the OTP’s request in the Rohingya situation would mean that the ICC would have jurisdiction over any deportation from Syria into the territory of a state party. There is, in fact, only one ICC member-state that borders Syria: Jordan. Jordan would be a particularly attractive application of the OTP’s theory of deportation, given that more than 650,000 Syrians are currently taking refuge there — almost exactly the same number as the Rohingya in Bangladesh. (Though not all of the Syrian refugees in Jordan have been the victim of deportation.)

I wonder, though, what Jordan would think of the OTP opening an investigation into deportations from Syria into Jordan. Presumably, that situation would not be limited to deportations, but would also include Jordan’s own treatment of Syrian refugees. (A more tailored situation would be immediately seen for what it was — a sop to Jordan.) As Human Rights Watch has ably documented, Jordan has been summarily returning hundreds of Syrian refugees back to Syria each month, a clear violation of international law. Those actions very likely qualify as the crime against humanity of persecution, especially when Jordanian authorities specifically target for expulsion a national group such as Palestinians. So it is not difficult to imagine the OTP bringing a case involving Jordan’s expulsions as part of an investigation into Syrian deportations. Indeed, the OTP would likely find it much easier to prosecute the expulsions, given that Jordan is not only obligated to cooperate with the ICC but has long been one of its most vocal supporters. Ironically, then, an investigation into Syrian deportations might mean that a Jordanian ends up in the dock before a Syrian!

NOTE: I don’t think the Jordan prosecution I discuss above is likely to happen. I’m just teasing out the possible implications of the ICC having jurisdiction over deportations from non-member states to member-states.

A Problematic Take on the Lubanga Trial

by Kevin Jon Heller

Justice in Conflict has a guest post today from a scholar who has written a book about the Lubanga trial. I think the post makes some excellent points about the problems with the trial. But I have serious reservations — acknowledging that I have not read the book — about the author’s take on why the trial did not focus on sexual violence:

Another [serious flaw] was the Chamber’s embargo on sexual violence. The matter of sexual violence loomed large in the trial not by its presence but by its absence. It became the trial’s trademark shame, a conspicuous token of the Chamber’s failure to place the substance of the Ituri province’s tragedy above the Chamber’s perpetual legal jousting. For most of the trial the Chamber did what it could to hear as little as possible about how frequently young women were raped and enslaved.

This is both unfair and mistaken. There is one reason, and one reason only, that sexual violence did not figure more prominently in the trial: Luis Moreno-Ocampo decided not to charge Lubanga with the relevant war crimes or crimes against humanity, choosing instead to focus exclusively on the war crime of conscripting or enlisting child soldiers. Here, for example, is what Patricia Viseur Sellers, a former Legal Advisor for Gender and prosecutor at the ICTY has to say:

Crimes of sexual violence were not charged. Such accusations were certainly within the purview of the Prosecutor. The Prosecutor could have brought charges related to sexual violence. Under the ICC Statute, enslavement, rape, torture, sexual slavery and inhuman acts are defined as crimes against humanity. In the Lubanga case, charges were brought under Article 8, war crimes, and as such could have included charges of torture, rape, sexual slavery or outrages upon personal dignity.

The Trial Chamber noted that they chose not to amend the charges. The Prosecutor could have amended the indictment at anytime prior to trial or even at a reasonable moment during the presentation of the prosecution case [to include charges for crimes of sexual violence]. The Prosecutor has suggested that to do so would have been detrimental to the due process rights of the accused. However, in the event of granting the Prosecutor’s move to amend, the Trial Chamber could have allowed the accused whatever time he needed to prepare his case in light of additional charges. That is a fairly standard procedure at other international tribunals.

Given Moreno-Ocampo’s decision to charge Lubanga solely with conscripting or enlisting child soldiers, the Trial Chamber had no choice but to limit the amount of testimony the prosecution could introduce regarding sexual violence. The Chamber explained why in paras. 629 and 630 of its judgment:

629. Notwithstanding the conclusions set out above, and given the submissions made at various stages of the proceedings, the Chamber needs finally to address how the issue of sexual violence is to be treated in the context of Article 8(2)(e)(vii) of the Statute. It is to be noted that although the prosecution referred to sexual violence in its opening and closing submissions, it has not requested any relevant amendment to the charges. During the trial the legal representatives of victims requested the Chamber to include this conduct in its consideration of the charges, and their joint request led to Decisions on the issue by the Trial Chamber and the Appeals Chamber (viz. whether it was permissible the change the legal characterisation of the facts to include crimes associated with sexual violence). Not only did the prosecution fail to apply to include rape and sexual enslavement at the relevant procedural stages, in essence it opposed this step. It submitted that it would cause unfairness to the accused if he was tried and convicted on this basis.

630. In accordance with the jurisprudence of the Appeals Chamber, the Trial Chamber’s Article 74 Decision shall not exceed the facts and circumstances (i.e. the factual allegations) described in the charges and any amendments to them. The Trial Chamber has earlier pointed out that “[f]actual allegations potentially supporting sexual slavery are simply not referred to at any stage in the Decision on the Confirmation of Charges”.1810 Regardless of whether sexual violence may properly be included within the scope of “using [children under the age of 15] to participate actively in hostilities” as a matter of law,1811 because facts relating to sexual violence were not included in the Decision on the Confirmation of Charges, it would be impermissible for the Chamber to base its Decision pursuant to Article 74(2) on the evidence introduced during the trial that is relevant to this issue.

Moreover, I think the author’s claim that “[t]he matter of sexual violence loomed large in the trial not by its presence but by its absence” is considerably overstated. Not only did sexual violence figure prominently in both the prosecution’s opening and closing arguments, as the Trial Chamber notes in its judgment, there was also considerable testimony concerning sexual violence during trial. The judgment points out in a footnote (n. 54) that 30 different witnesses, 18 female, 12 male, “referred to acts of sexual violence which they either suffered or witnessed.” And it discusses testimony given by one witness, P-0046, at length. Here is just a snippet of P-0046’s testimony:

890. According to the evidence of P-0046, all the girls she met at the demobilisation centres, except for a few who had been protected by certain women in the camps, told the witness that they had been sexually abused, most frequently by their commanders but also by other soldiers. Some fell pregnant, resulting in abortions; and there were instances of multiple abortions. The witness gave evidence that the psychological and physical state of some of these young girls was catastrophic.

891. The youngest victim of this sexual abuse interviewed by P-0046 was 12 years old. The witness stated that some of those who became pregnant were thrown out of the armed group and ended up on the streets of Bunia. Others went to join their relatives, and although they may have felt they remained part of the UPC, the latter failed to provide them with support. It was difficult to reintegrate them into their families because the girls were stigmatised, and significant mediation was necessary. The witness stated that the children provided her with a clear account of systematic sexual violence in the camps.

Should the Lubanga trial have included specific crimes of sexual violence? Absolutely. But the absence of those charges and the (relatively) limited testimony concerning sexual violence cannot be attributed to the Trial Chamber. If you are looking for someone to blame — and you should be — blame Luis Moreno-Ocampo.

NOTE: I have not addressed the victims’ efforts to add sexual-violence charges in the middle of trial. If you want to blame the Chamber for rejecting that request, fair enough. But I have already explained why I think the Chamber was correct.

Vargas Niño’s Mistaken Critique of My Position on Burundi

by Kevin Jon Heller

Spreading the Jam has a guest post today from Santiago Vargas Niño criticising my argument that the OTP was required to notify Burundi as soon as it decided to ask the OTP to authorize the investigation. Here is what he says:

Professor Heller cites Article 15(6) to argue that, by receiving information under articles 15(1) and 15(2) of the Statute, the Prosecution has initiated an investigation. An equally plain reading of Article 18 would suggest that a parallel duty to notify concerned States would arise as soon as a situation caught the Prosecutor’s eye. Yet he acknowledges that “notification cannot be required every time the OTP decides to advance a preliminary examination (…) The better interpretation of Art. 18 is that notification is required once the OTP has decided to ask the PTC to authorize an investigation.”

Not only is that moment different to the “initiation” of an investigation, both under articles 15 and 18, thus rendering any claims of “natural” interpretation of the Statute inane, but professor Heller’s amalgamation of preliminary examination and investigation flies in the face of Article 15(3). This provision orders the Prosecution to submit a request for authorisation if it concludes that there is a reasonable basis to proceed with an investigation. Years of unchallenged practice have led to the understanding that such conclusion can only be reached through the preliminary examination, a stage that precedes the opening of an investigation and that is described by Article 15(2) – not by Article 15(1). Professor Heller’s argument also discounts the significance of Article 15(4), which squarely attributes the power to authorise the “commencement” (i.e. “initiation”) of an investigation to the PTC, and which conditions it upon the Prosecution’s demonstration that there is a reasonable basis to proceed under Article 53(1).

Furthermore, equating the launch of a preliminary examination with the artificial “initiation” of an investigation under Article 15(1) is extremely risky. If that were the case, the Prosecution should not have rushed to apply for authorisation to commence an investigation in Burundi before 25 October 2017 because its preliminary examination would have constituted a “criminal [investigation] (…) which [was] commenced prior to the date on which the withdrawal became effective” under Article 127. Such interpretation would also allow the Prosecution to exercise its powers under Article 54, as professor Jacobs puts it, since the moment: “an OTP investigator sitting in front of his computer in The Hague [starts] downloading HRW and Amnesty International reports.”

According to Vargas Niño, my argument “stems solely from [my] peculiar approach to Article 15.” Alas, it is his approach that is peculiar. And not just peculiar — wrong…

A Response to Dov Jacobs on the Burundi Investigation

by Kevin Jon Heller

At Spreading the Jam, Dov Jacobs defends the Pre-Trial Chamber’s conclusion in the Burundi situation that the OTP is not required to notify a state until after the PTC has authorized an investigation. Here are the critical paragraphs from his post:

Note the different language used [in Art. 18] depending on whether there is a referral under 13(a) (state referral) or 13(b) (proprio motu): in the former case, the notification must come when “the Prosecutor has determined that there would be a reasonable basis to commence an investigation”, in the latter the notification must come when “the Prosecutor initiates an investigation” pursuant to Article 15. This seems to mean that the initiation of an investigation is something different, in a proprio motu context, that the fact that the OTP considers that there is a reasonable basis to proceed with an investigation. For me, this means that all procedural steps of Article 15 need to have been followed (including the formal authorisation) before the notification obligation of Article 18 kicks in. The determination by the Prosecutor that “there is a reasonable basis to proceed with an investigation” (Article 15(3)), which is sufficient to initiate an investigation under 13(a) is only one step of the procedure under article 15. Kevin seems to equate the authorisation under 15(4) and authorisation under 18(2). But I think these are two different “authorisations”. Under 15(4), a PTC authorises the initiation of a proprio motu investigation, while under 18(2), it is an authorisation to investigate despite the request for deferral by a State, irrespective of whether the investigation was initiated initially under 13(a) or 13(c).

I think the confusion comes from a possible misunderstanding on the scope of Article 18. Article 18 provides for a limited procedure to be followed for a preliminary ruling on admissibility which I think is self-contained within Article 18. Which means that in my view the notification requirement under Article 18 cannot be read in as a condition for the validity of the Article 15 procedure. I should add also that Article 18 does not lead to a formal challenge to admissibility, which will fall under Article 19. In this sense, I do not think Kevin is right (whether one agrees with his interpretation of Article 18 or not) in saying that Burundi will not be able to challenge admissibility before a case is brought. They will be able to do so at any time (especially given the ICC’s case law that “case” in the Rome Statute does not really mean “case” in the context of admissibility questions, which is why everybody assesses admissibility as early as the PE phase. I think that doesn’t make sense, but that is a different debate…).

In my view, though clever, Dov’s argument is problematic. The first problem concerns his claim that “all procedural steps of Article 15 need to have been followed (including the formal authorisation) before the notification obligation of Article 18 kicks in.” That position is irreconcilable with Art. 15. As I pointed out in my previous post, Art. 15 not only specifically distinguishes between the OTP initiating an investigation proprio motu (paragraph 1) and the PTC authorizing the commencement of that investigation (paragraph 4), it specifically deems the former but not the latter part of the preliminary-examination process (paragraph 6). Paragraph 6 makes no sense if “initiates” in paragraph 1 refers to all of the steps in Art. 15, including authorization.

The only way Dov can avoid that critique is to assert that “initiates” in Art. 15(1) does not mean the same thing as “initiates” in Art. 18(1). If they mean the same thing, Art. 18(1)’s notification requirement necessarily kicks in — as I previously argued — prior to the PTC authorizing the proprio motu investigation (because Art. 15(6) says initiating is part of the preliminary-examination process and authorization is not). Dov provides no evidence that “initiates” means different things in Art. 18(1) and Art. 15(1), and any such argument is difficult to reconcile with the fact that Art. 18(1) specifically refers to “the Prosecutor initiat[ing] an investigation pursuant to articles 13 (c) and 15,” thereby using “initiates” in Art. 15(1) to give meaning to Art. 18(1)’s notification requirement. Moreover, if the drafters of Art. 18 wanted the proprio motu notification requirement to kick in only after all of the steps in Art. 15 had been completed, why would they not simply have written “or the Court authorizes the commencement of the investigation” instead of “or the Prosecutor initiates an investigation”?

Equally problematic is Dov’s insistence that Burundi will be able to challenge the validity of the proprio motu investigation even though the PTC has already formally authorized it. Dov’s argument to that effect is strangely devoid of any reference to the actual language of Art. 19; he simply says that “’case’ in the Rome Statute does not really mean ‘case’ in the context of admissibility questions.” If Dov believes that Art. 19 allows a state to shut down an already-authorized proprio motu investigation by invoking complementarity, he should make the argument. In my view, nothing in Art. 19 permits such a challenge, given that the Article is limited — both in name and in terms of its specific provisions — to cases. Art. 19(2) is particularly revealing in that regard, as it specifically limits jurisdiction and admissibility challenges to “[a]n accused or a person for whom a warrant of arrest or a summons to appear has been issued under article 58” (subparagraph a) or “[a] State which has jurisdiction over a case, on the ground that it is investigating or prosecuting the case or has investigated or prosecuted” (subparagraph b).

Dov’s reading also makes a mishmash of the relationship between Art. 19 and Art. 15. If Dov is right, a state can use a complementarity challenge under Art. 19 to shut down a proprio motu investigation that has already been authorized by the PTC under Art. 15. Yet the PTC has to consider issues of complementarity in order to authorize a proprio motu investigation in the first place, because it has to find the OTP’s contemplated cases admissible in order to conclude that there is a “reasonable basis to proceed with [the[ investigation.” Dov’s position thus requires the PTC to consider complementarity twice in Burundi-like situations: once when the OTP asks it to authorize a proprio motu investigation ex parte (under Art. 15), and again when the affected state asks it  to defer the investigation (under Art. 18). When deciding to authorize the investigation, the PTC will hear only from the OTP; when deciding to defer the investigation, the PTC will hear from both the OTP and the PTC. Why would the drafters of the Rome Statute adopted such a duplicative and cumbersome process? My (textually sound) interpretation of Art. 18’s notification process makes much more sense, because it means that the PTC will only address complementarity once, before it authorizes a proprio motu investigation.

My interpretation is also superior to Dov’s in terms of the politics of proprio motu investigations. If Dov’s interpretation of the Rome Statute is correct, a state facing referral by another state can use Art. 18 to prevent the PTC from ever formally approving the OTP’s belief that an investigation is warranted, while a state facing proprio motu investigation cannot invoke Art. 18 until after the PTC has formally approved a similar belief. States are thus better off being referred by another state than being investigated proprio motu whenever the OTP can convince the PTC to grant the latter ex parte. It goes without saying, however, that states at the Rome Conference were far more concerned by proprio motu investigations than state referrals.

Dov’s defense of the PTC’s interpretation of Art. 18’s notification requirement is very clever. But I think it’s also clearly incorrect.

How the PTC Botched the Ex Parte Request to Investigate Burundi

by Kevin Jon Heller

Last week I argued that the OTP’s failure to ask the Pre-Trial Chamber to authorize an investigation prior to Burundi’s withdrawal from the ICC becoming effective — 28 October 2017 — meant that the Court no longer had jurisdiction over crimes committed on Burundi’s territory prior to that date. I still think my legal analysis is correct, but my factual assumption was clearly not. As it turns out, the OTP filed an authorization request with the PTC on September 15, but did so ex parte and under seal — a possibility the ever-brilliant Sergey Vasiliev discussed a few days ago here at Opinio Juris. The PTC authorized the investigation on October 25, three days before Burundi’s withdrawal became effective, but only released a public redacted version of its decision yesterday, November 9. As it stands now, therefore, the ICC retains jurisdiction over crimes committed in Burundi prior to 28 October 2017.

Unfortunately, the PTC’s decision contains a critical legal flaw — one whose importance cannot be overstated. Because the OTP filed its request to open an investigation ex parte and under seal, Burundi was not informed that the request existed until after the PTC had already decided to grant the request and authorize the investigation. The PTC makes this clear in paragraph 11 of its decision:

11. In sum, the Chamber finds that, on the basis of a combined reading of articles 15(3), 18 and 68(1) of the Statute and rule 50(1) of the Rules, a procedure pertaining to a request for authorization of an investigation may, under certain circumstances, be conducted under seal, ex parte, with the Prosecutor only.

In fact, the OTP did not even inform Burundi about the investigation immediately after the PTC authorized it, because the PTC accepted the OTP’s argument that it needed 10 additional days to ensure that victims and witnesses were protected. (See paragraphs 16-19.)

Here is the problem: Art. 18 of the Rome Statute required the OTP to notify Burundi when it initiated the investigation into the situation there, not when the PTC authorized the investigation. Here is what the PTC says in paragraph 17 (emphasis mine)…

A Dissenting Opinion on the ICC and Burundi

by Kevin Jon Heller

As has been widely reported, Burundi has just become the first state to formally withdraw from the ICC. The OTP has been examining the situation in Burundi since April 2016, but it did not formally ask the Pre-Trial Chamber (PTC) to authorize an investigation prior to Burundi’s withdrawal becoming effective. So what does Burundi’s withdrawal mean for the OTP’s preliminary examination (PE)? Can the OTP still ask the PTC to authorize an investigation into crimes committed in Burundi prior to withdrawal? Or does Burundi’s withdrawal divest the Court of jurisdiction over the situation?

The relevant provision is Art. 127(2) of the Rome Statute (my emphasis):

A State shall not be discharged, by reason of its withdrawal, from the obligations arising from this Statute while it was a Party to the Statute, including any financial obligations which may have accrued. Its withdrawal shall not affect any cooperation with the Court in connection with criminal investigations and proceedings in relation to which the withdrawing State had a duty to cooperate and which were commenced prior to the date on which the withdrawal became effective, nor shall it prejudice in any way the continued consideration of any matter which was already under consideration by the Court prior to the date on which the withdrawal became effective.

The ICC is taking the position that Art. 127(2)’s bolded language means Burundi’s withdrawal does not affect the Court’s jurisdiction over crimes committed prior to the date the withdrawal became effective — 28 October 2017. It does not explain why, but the argument is relatively straightforward: (1) the PE in Burundi began prior to 28 October 2017; (2) a PE qualifies as a “matter”; (3) the OTP is part of the Court. Hence (4) the Burundi PE “was already under consideration by the Court prior to the date on which the withdrawal became effective” and the Court continues to have jurisdiction over (“consider”) the situation.

A number of commentators agree with the ICC’s position, including Amnesty International and Beitel van der Merwe. The only dissenting voice is Dov Jacobs, who is skeptical about point (2) — whether a PE really qualifies as a “matter” for purposes of Art. 127(2). Here is what he says:

The key issue is what is covered by the expression “any matter already under consideration by the Court”. Alex Whiting makes the argument that this expression is broad enough to cover preliminary examinations by the OTP. Possibly, he is right from a linguistic point of view. However, I have a difficulty with the idea that such an informal phase as a preliminary examination (which might simply involve an OTP investigator sitting in front of his computer in The Hague downloading HRW and Amnesty International reports) might have such massive consequences as trumping the decision of a State to withdraw from the Rome Statute.

I agree with Dov. As is well known, the OTP divides the preliminary-examination process into four phases: (1) determining whether a situation falls “manifestly outside” of the ICC’s jurisdiction; (2) determining whether there is a reasonable basis to believe an international crime was committed in the situation; (3) assessing admissibility; (4) assessing the interests of justice. According to the OTP, a “formal” PE begins with Phase 2 (emphasis mine):

80. Phase 2, which represents the formal commencement of a preliminary examination of a given situation, focuses on whether the preconditions to the exercise of jurisdiction under article 12 are satisfied and whether there is a reasonable basis to believe that the alleged crimes fall within the subject-matter jurisdiction of the Court. Phase 2 analysis is conducted in respect of all article 15 communications that were not rejected in Phase 1, as well as of information arising from referrals by a State Party or the Security Council, declarations lodged pursuant to article 12(3), open source information, and testimony received at the seat of the Court.

Dov’s example of the OTP investigator downloading HRW or AI reports about a situation is thus spot on. Because such reports are “open source information,” the very act of looking at it means that a PE is at Phase 2 and a “formal” PE has commenced. Which means in turn that — according to the ICC’s interpretation of Art. 127(2) — the Court retains jurisdiction over the situation in the report. (And retains it in perpetuity, because there is no time limit on an OTP decision to advance a PE to a full investigation, as the 13 year-old Colombia PE indicates.)

Like Dov, I am not sure “matter” can or should be interpreted to include any formal PE, even one triggered by an OTP investigator (or even an intern?) downloading an NGO report (or even just reading it on the screen?). But I think there is a more important question about the ICC’s interpretation of Art. 127(2): whether a situation is under “consideration by the Court” simply by virtue of the OTP preliminarily examining it. Alex Whiting believes that it is (emphasis mine):

There is a decent but far from certain argument that jurisdiction should survive at least for any crimes that are the subject of a preliminary examination by the Office of the Prosecutor before the date of a State Party’s effective withdrawal. Following the broad first sentence of Article 127(2), the provision addresses two specific situations: (1) when an investigation or proceeding is underway before effective withdrawal, the departing State Party continues to have a legal duty to cooperate with the Court’s inquiry even after the State Party has left the Court, and (2) the State Party’s departure cannot prejudice the Court’s “consideration of any matter” that was already underway before departure. The “Court” in the Rome Statute refers to the entire ICC, including the Prosecutor, and not just the judges.

I disagree. There is no question that “the Court” sometimes refers to “the entire ICC,” such as when the Rome Statute is referring generically to the ICC’s location or international legal personality. Indeed, Art. 34 says that “the Court” is composed of the Presidency, the judiciary, the OTP, and the Registry.

But the Rome Statute also uses “the Court” in a more restrictive fashion — to refer specifically to the judiciary, excluding the OTP. Here are some examples:

[1] Art. 19(3) provides that “[t]he Prosecutor may seek a ruling from the Court regarding a question of jurisdiction or admissibility.”

[2] Art. 19(7) provides that “[i]f a challenge is made by a State referred to in paragraph 2 (b) or (c), the Prosecutor shall suspend the investigation until such time as the Court makes a determination in accordance with article 17.”

[3] Art. 19(10) provides that “[i]f the Court has decided that a case is inadmissible under article 17, the Prosecutor may submit a request for a review of the decision.”

[4] Art. 21(2) provides that “[t]he Court may apply principles and rules of law as interpreted in its previous decisions.” The OTP doesn’t issue decisions.

[5] Art. 65(5) provides that “[a]ny discussions between the Prosecutor and the defence regarding modification of the charges, the admission of guilt or the penalty to be imposed shall not be binding on the Court.”

[6] Art. 66(3) provides that, “[i]n order to convict the accused, the Court must be convinced of the guilt of the accused beyond reasonable doubt.”

[7] Art. 67(2) provides that, with regard to the rights of the defendant, “[i]n case of doubt as to the application of this paragraph, the Court shall decide.”

I could go on. The point is that, contra Alex, we cannot simply assume that Art. 127(2)’s reference to “the Court” includes both the judiciary and the OTP. It may well be that Art. 127(2) refers only to the judiciary. The distinction, of course, is critical in the context of Burundi’s withdrawal: if a matter must be “under consideration by” the judiciary for Art. 127(2) to apply, then the OTP’s failure to open an investigation into the situation means that the Court (writ large) no longer has jurisdiction over any crimes committed in Burundi — not even over those committed prior to the date Burundi’s withdrawal became effective.

I cannot claim with absolute certainty that the more restrictive reading of Art. 127(2) is correct, especially as Amnesty International says that the travaux preparatoires do not shed any light on the issue. But it seems like the much stronger position. Most importantly, the precise expression “under consideration by the Court” also appears in Art. 95, which deals with the postponement of requests in connection with admissibility challenges (emphasis mine):

Where there is an admissibility challenge under consideration by the Court pursuant to article 18 or 19, the requested State may postpone the execution of a request under this Part pending a determination by the Court, unless the Court has specifically ordered that the Prosecutor may pursue the collection of such evidence pursuant to article 18 or 19.

“Under consideration by the Court” in Art. 95 clearly means “under consideration by the judiciary.” Normal rules of treaty interpretation thus suggest that the same expression in Art. 127(2) is similarly restrictive.

This interpretation of Art. 127(2) is also supported by the problem Dov identifies — that deeming any PE a “matter” gives the OTP so much power that the withdrawal provision is a virtual nullity. A PE does indeed seem like a “matter,” but that does not mean a PE is a matter “under consideration by the Court.” The better view is that a situation is only “under consideration by the Court” once the OTP asks the PTC to authorize an investigation into that situation.

The upshot of all of this is that, in my view, the Court no longer has jurisdiction over crimes committed on the territory of Burundi prior to the state’s withdrawal from the ICC. If the OTP had wanted to keep alive the situation, it needed to ask the PTC before 28 October 2017 for permission to open an investigation. And it failed to do so.

This Is Why People Think the ICC Is Unfairly Targeting Africa

by Kevin Jon Heller

Snapshot of two days in the life of the ICC.

On Tuesday, the ICC issued a new arrest warrant in the Libya situation — for Mahmoud al-Werfalli, a commander in the so-called Libyan National Army (LNA), which defected from the Libyan army during the revolution and is currently vying for power with the UN-backed Government of National Accord (GNA). The arrest warrant represents a new phase in the ICC’s completely unsuccessful investigation in Libya, as it is the first to focus on events that happened after the revolution. There is no reason to believe, however, that the warrant for al-Werfalli will be any more successful than the ones for Gaddafi and al-Senussi: the LNA has already made clear they will not surrender him to the ICC, and the GNA has zero prospect at present of capturing him.

On Wednesday, Rodrigo Duterte, the President of the Philippines, instructed his police to shoot human-rights activists who are “obstructing justice” by investigating his war against (alleged) drug dealers. That war has involved at least 7,000 extrajudicial killings in the past 13 months and has featured Duterte openly admitting not only that he has ordered the extrajudicial kilings, but that he has personally committed themHuman-rights groups and even a Philippine senator have called for the ICC to open an investigation into the situation.

There seems to be little question that al-Werfalli is guilty of ordering and participating in more than two dozen summary executions of captured soldiers — remarkably, there is video to that effect. But al-Werfalli is one military commander among hundreds responsible for horrific crimes in Libya. Duterte, by contrast, is the President of one of the only states in Southeast Asia that has ratified the Rome Statute. Even if he never ended up in the ICC’s dock, a formal investigation of the situation that he has almost single-handedly created in the Philippines would do more to deter the commission of international crimes than 500 arrest warrants for thugs like al-Werfalli. Yet despite issuing a strong statement making clear that the Court has jurisdiction over the situation and could prosecute individuals responsible for international crimes, there is no indication that the OTP has seriously contemplated opening a formal investigation in the Philippines.

The ICC fiddles in Benghazi while Manila burns. And yet the ICC claims not to understand why so many people think it’s obsessed with Africa.

ICC Appeals Chamber Says A War Crime Does Not Have to Violate IHL

by Kevin Jon Heller

One of the most basic assumption of ICL is that an act cannot be a war crime unless it violates a rule of international humanitarian law (IHL). Article 6(b) of the London Charter criminalised “War Crimes: namely, violations of the laws or customs of war.” Article 3 of the ICTY Statute provides that “[t]he International Tribunal shall have the power to prosecute persons violating the laws or customs of war,” while Article 4 of the ICTR Statute provides that “[t]he International Tribunal for Rwanda shall have the power to prosecute persons committing or ordering to be committed serious violations of Article 3 common to the Geneva Conventions of 12 August 1949 for the Protection of War Victims, and of Additional Protocol II thereto of 8 June 1977.” And Article 8 of the Rome Statute criminalises “[g]rave breaches of the Geneva Conventions of 12 August 1949”; “[o]ther serious violations of the laws and customs applicable in international armed conflict”; [i]n the case of an armed conflict not of an international character, serious violations of article 3 common to the four Geneva Conventions of 12 August 1949″; and “[o]ther serious violations of the laws and customs applicable in armed conflicts not of an international character.” In each and every case, war crimes are limited to violations of IHL.

No more. The Appeals Chamber (AC) at the ICC has just unanimously held in Ntaganda that a perpetrator can be convicted of a war crime even if his act does not violate IHL. That decision is not simply “unprecedented,” as the AC openly acknowledges. It is simply incorrect — as this post will demonstrate.

The judgement itself addresses allegations that Ntaganda is criminally responsible for two war crimes — rape and sexual slavery — involving children forcibly recruited into his organised armed group, the UPC/FPLC. Ntaganda challenged that allegation, arguing that “crimes committed by members of armed forces on members of the same armed force do not come within the jurisdiction of international humanitarian law nor within international criminal law.” The Trial Chamber (TC) disagreed, in a judgment ably discussed and critiqued by Yvonne McDermott. Ntaganda appealed, giving rise to this judgment. Here is the AC’s “key finding”:

2. Having regard to the established framework of international law, members of an armed force or group are not categorically excluded from protection against the war crimes of rape and sexual slavery under article 8 (2) (b) (xxii) and (2) (e) (vi) of the Statute when committed by members of the same armed force or group.

Before turning to the logic of the judgment, it is important to be very precise about the terms of my quarrel with the AC. I completely agree with the AC that there are situations in which a member of an armed force can, in fact, commit the war crime of rape or the war crime of sexual slavery against a member of the same armed force. As the AC rightly notes, although the Third and Fourth Geneva Conventions do not apply to acts committed by a combatant against someone from the same side of the conflict — whether by virtue of membership in that same armed force (GC III) or by nationality (GC IV) — the First and Second Geneva Conventions contain no such limitation:

59. In contrast, Geneva Conventions I and II, which protect the wounded and sick on land and the wounded, sick and shipwrecked at sea respectively, provide protection “in all circumstances […] without any adverse distinction founded on sex, race, nationality” and prohibit violence against them. Importantly, such protected status is not limited to persons belonging to enemy armed forces, but includes wounded, sick or shipwrecked members of a party’s own armed forces, a rule that corresponds to the understanding of the scope of protection since the first Geneva Convention was adopted in 1864. It follows from the above that the notion of grave breaches under Geneva Conventions I and II includes violations committed against the wounded, sick or shipwrecked committed by members of their own armed force.

Nothing in GC I or GC II suggests, however, that IHL protects all members of the armed forces against member-on-member violence. On the contrary, let’s take a look at the AC’s statement again, with the critical language in bold:

59. In contrast, Geneva Conventions I and II, which protect the wounded and sick on land and the wounded, sick and shipwrecked at sea respectively, provide protection “in all circumstances […] without any adverse distinction founded on sex, race, nationality” and prohibit violence against them. Importantly, such protected status is not limited to persons belonging to enemy armed forces, but includes wounded, sick or shipwrecked members of a party’s own armed forces, a rule that corresponds to the understanding of the scope of protection since the first Geneva Convention was adopted in 1864. It follows from the above that the notion of grave breaches under Geneva Conventions I and II includes violations committed against the wounded, sick or shipwrecked committed by members of their own armed force.

Under GC I and GC II, in other words, member-against-member violence violates IHL only if the victim is wounded, sick, or shipwrecked. If the victim is none of those things — if he or she is not hors de combat — that violence may well violate a state’s domestic criminal law, but it does not violate IHL.

If the AC had limited the scope of its judgment to rape and sexual slavery committed against child soldiers who were hors de combatdefined by the ICRC, in relevant part, as “anyone who is defenceless because of unconsciousness, shipwreck, wounds or sickness” — it would have been on firm ground. But that is not what it has done. On the contrary, the AC goes to great lengths to make clear that member-against-member rape and sexual slavery are war crimes even if the victim is an active combatant –– ie, one who is not hors de combat. Here is the relevant paragraph (emphasis mine):

64. With regard to the second issue – namely whether Status Requirements exist in international humanitarian law specifically for the war crimes of rape and sexual slavery – the Appeals Chamber observes that the prohibitions of rape and sexual slavery in armed conflict are without a doubt well established under international humanitarian law. As noted by the Trial Chamber, protection under international humanitarian law against such conduct generally “appear[s] in contexts protecting civilians and persons hors de combat in the power of a party to the conflict”. In this regard, the question arising before the Appeals Chamber is whether such explicit protection under international humanitarian law suggests any limits on who may be victims of such conduct. In the view of the Appeals Chamber, there is no conceivable reason for reaching such a conclusion.

Notice the bold language, because it’s critical — and wrong. IHL protection does not “generally” apply only to civilians and combatants hors de combat. On the contrary, each and every IHL convention applies only to those two categories of individuals. As we have seen, the AC itself acknowledges that limitation with regard to all four of the Geneva Conventions. It cites no other source of IHL, instead simply noting that the ICRC states in its new commentary to GC I “that Common Article 3 protects members of armed forces against violations committed by the armed force to which they belong.” But that statement is incomplete and misleading, because the ICRC makes unequivocally clear that CA3’s prohibitions apply only to individuals who are hors de combat:

518  Subparagraph (1) covers all ‘[p]ersons taking no active part in the hostilities, including members of armed forces who have laid down their arms and those placed hors de combat by sickness, wounds, detention, or any other cause’. The article does not expand on these notions and this part of the article did not give rise to much discussion at the 1949 Diplomatic Conference. The protection afforded under this subparagraph requires that the person be in the power of a Party to the conflict (see section E.4).
519  The protection of persons not or no longer participating in hostilities is at the heart of humanitarian law. The persons protected by common Article 3 are accordingly described by way of explicit delimitations: ‘persons taking no active part in the hostilities, including members of armed forces who have laid down their arms and those placed hors de combat by sickness, wounds, detention, or any other cause’ (emphasis added). Parties to a non-international armed conflict are under the categorical obligation to treat these persons humanely, in all circumstances and without any adverse distinction.

The Trial Chamber’s judgment is no better. The TC rests its conclusion that member-against-member rape is a war crime even when the victim is an active combatant solely on two things: the Martens Clause and Art. 75 of the First Additional Protocol (AP I). Here is paragraph 47:

While most of the express prohibitions of rape and sexual slavery under international humanitarian law appear in contexts protecting civilians and persons hors de combat in the power of a party to the conflict, the Chamber does not consider those explicit protections to exhaustively define, or indeed limit, the scope of the protection against such conduct. In this regard, the Chamber recalls the Martens clause, which mandates that in situations not covered by specific agreements, ‘civilians and combatants remain under the protection and authority of the principles of international law derived from established custom, from the principles of humanity and from the dictates of public conscience’. The Chamber additionally notes that the fundamental guarantees provisions [in Art. 75] refer to acts that ‘are and shall remain prohibited at any time and in any place whatsoever’ and as such apply to, and protect, all persons in the power of a Party to the conflict.

I don’t have time to get into a detailed discussion of the Martens Clause. Suffice it say here that it is very unlikely that the Clause can ever be relied upon to expand IHL not only beyond conventional law, but even beyond customary IHL — and as the AC itself acknowledges (para. 60), there is literally zero state practice indicating that member-against-member mistreatment is a war crime even when the victim is an active combatant. Even Antonio Cassese, no stranger to judicial activism, dismisses this “norm-creating” reading of the Martens Clause as “radical.” As he says, “[s]urely the Clause does not envisage — nor has it brought about the birth of — two autonomous sources of international law, distinct from the customary process.”

As for Art. 75 of AP I, the Protocol’s “fundamental guarantees” provision, the TC’s position is deeply problematic. Here is n. 111:

Article 75 of Additional Protocol I refers to ‘a Party to the conflict’ (emphasis added) and therefore does not limit the fundamental guarantees to persons in the power of the opposing party.

The TC conveniently fails to note that Art. 75 applies only to international armed conflict — and that Art. 4 of AP II, the “fundamental guarantees” provision in the NIAC Protocol, is specifically limited to “persons who do not take a direct part or who have ceased to take part in hostilities” (ie, civilians and combatants hors de combat).

Given that conventional IHL uniformly requires the victim of member-against-member mistreatment to be hors de combat, on what basis does the AC hold that the status of the victim is irrelevant? The answer comes from this paragraph (emphasis mine):

65. The Appeals Chamber agrees with the Trial Chamber’s finding that “there is never a justification to engage in sexual violence against any person; irrespective of whether or not this person may be liable to be targeted and killed under international humanitarian law”. Accordingly, in the absence of any general rule excluding members of armed forces from protection against violations by members of the same armed force, there is no ground for assuming the existence of such a rule specifically for the crimes of rape or sexual slavery.

This is simply incorrect. To begin with, there is a specific rule excluding active combatants from the war crimes of rape and sexual slavery in member-against-member situations: namely, the rule that says violence in member-against-member situations violates IHL only when the victim is hors de combat. The AC’s judgment suggests that states not only had to specify that rule in the various IHL conventions, they also had to add: “oh, and by the way, this limit means that mistreating active combatants doesn’t violate IHL.” But that’s silly: the former implies the latter. After all, expressio unius est exclusio alterius is a basic rule of treaty interpretation.

But even if that was not the case, there would still be a general rule excluding active combatants from the war crimes of rape and sexual slavery in member-against-member situations: the rule that says a war crime must involve a violation of IHL. As noted at the beginning of this post, that is one of the most basic assumptions of IHL. Not all violations of IHL are war crimes, but all war crimes are violations of IHL. So the burden of proof was not on Ntaganda to show that rape and sexual slavery cannot be war crimes in member-against-member situations if the victim is an active combatant. The burden was on the prosecution to prove that such acts actually violate IHL. Because if they don’t — and they don’t, as we have seen — the Court has no jurisdiction whatsoever over Ntaganda’s acts, at least insofar as they are legally characterised as war crimes.

In the end, the AC’s decision in Ntaganda is little more than the latest iteration of the Court’s willingness to rely on teleological reasoning when the Rome Statute does not protect victims as much as the judges think it should. No one is in favour of raping and sexually enslaving child soldiers. But the solution isn’t to detach the law of war crimes from its moorings in IHL by holding — if only implicitly — that an act can be a war crime even if it does not violate IHL. To do so is not only legally indefensible, it risks delegitimising both the Court and the law of war crimes itself.

Emailing Does Not Pass the Kiobel Test: US Court Dismisses ATS Case Against Anti-Gay Pastor

by Julian Ku

Distracted by #ComeyDay and other international crises, I missed this recent U.S. federal court decision in Sexual Minorities of Uganda v. Livelydismissing an Alien Tort Statute lawsuit on Kiobel extra-territoriality grounds.  While using unusually critical language to denounce U.S. pastor-defendant Scott Lively’s involvement in Uganda’s anti-homosexual laws and actions, the U.S. District Court for Massachusetts held:

…Defendant’s status as an American citizen and his physical presence in the United States is clearly not enough under controlling authority to support ATS extraterritorial jurisdiction. The sporadic trail of emails sent by Defendant to Uganda does not add enough to the record to demonstrate that Plaintiff’s claims “touch and concern the territory of the United States . . with sufficient force to displace the presumption against extraterritorial application.” Kiobel, 133 S. Ct. at 1669.

What is notable about this case is that the same court and judge refused to dismiss this case on Kiobel grounds back in 2013 with largely the same allegations. The main difference with the result in 2017 seems to be that discovery revealed that Lively, the U.S. pastor, did not provide any

financial backing to the detestable campaign in Uganda, he directed no physical violence, he hired no employees, and he provided no supplies or other material support. His most significant efforts on behalf of the campaign occurred within Uganda: itself, when he appeared at conferences, meetings, and media events.

On these facts, this seems like the right result.  Kiobel requires something more than communications from the United States to “displace the presumption against extraterritoriality.” But caselaw continues to be a little muddy and I fully expect this to be appealed.