The other side of the Article 21(3) coin: Human rights in the Rome Statute and the limits of Article 21(3)

Abstract

The drafters of the Rome Statute sought to accord human rights a central place within the legal framework of the International Criminal Court (ICC). This was done not only through numerous provisions on the rights of the accused, victims, and witnesses, but also through the inclusion of the overarching Article 21(3) of the Rome Statute. Article 21(3) Rome Statute requires that the interpretation and application of all ICC law be consistent with internationally recognized human rights. While this provision has been employed on numerous occasions to bolster human rights protection in the ICC legal framework, it is not without its limits. In a series of decisions over the past few years, ICC judges have placed limits on the protections that can be read into the ICC legal framework on the basis of Article 21(3). Beyond stating that the ICC ‘is not a human rights court’, the decisions in question articulate no clear justification for the limitations imposed on Article 21(3). The present article analyses these decisions and identifies the underlying rationale for the Court’s approach: the principle of speciality. However, the picture is further complicated by the judges’ willingness to overlook the principle of speciality when particularly serious violations of human rights are involved. This leaves the precise contours of human rights protection in the ICC legal framework undefined.

1. Introduction

The adoption of the Rome Statute and the subsequent official establishment of the ICC was a victory for the human rights movement: here was a permanent international institution dedicated to the fight against impunity for atrocities and mass human rights violations.¹ Compared with previous international criminal tribunals, the Rome Statute took big steps forward in protecting the rights of those participating in ICC proceedings. Among other things, the Rome Statute provides for the direct participation of victims, establishes a Trust Fund for victims, allows the possibility to claim reparations from perpetrators,² and contains detailed provisions on the rights of the accused aimed at guaranteeing procedural fairness.³

Perhaps the greatest innovation of the Rome Statute, in human rights terms, was Article 21(3). The provision stipulates that the ICC must interpret and apply all law in a manner consistent with internationally recognized human rights. The result, in the words of the ICC Appeals Chamber, is that ‘human rights underpin the Statute; every aspect of it’.⁴ Much has been said about the promise of Article 21(3), and indeed the ICC’s case law has already demonstrated the important role that the Article can play in expanding, sometimes very ambitiously, the human rights protections contained in the Statute. Early predictions that Article 21(3) would ‘have little concrete impact on the implementation of the applicable law’ have proved incorrect.⁵

For the most part, commentary on Article 21(3) to date has focused on the uses to which the provision has been put and the human rights protections that have come about as a result. Recent decisions in certain ICC cases indicate that it may be time to look not only at what Article 21(3) can do but also at what it cannot do. This contribution will comment on a series of decisions from the past few years in which ICC judges have placed limits on the potential of Article 21(3). Following a brief overview of the Article itself, its place in the Rome Statute, and a sample of some of its uses to date, the article will consider three instances where ICC judges have set out limitations to it. In doing so, the article will seek to identify the underlying rationale that justifies the decisions reached by the Court, and reflect on what this tells us about human rights in the ICC legal framework.

2. Article 21(3) so far

Article 21(3) is the final paragraph in a provision that is innovative in its entirety.⁶ Article 21 sets out the sources of law to be applied by the ICC and the relationship between them; its level of detail is attributable to the drafters’ aim of increasing legal certainty by restricting judicial discretion when it came to the sources of law.⁷ As one might expect, Article 21(1)(a) stipulates that ‘in the first place’ the Court must apply the Rome Statute, the Elements of Crimes, and the Rules of Procedure and Evidence. One tier down in the hierarchy, and to be applied ‘in second place’ and ‘where appropriate’, are the sources listed in Article 21(1)(b), namely applicable treaties and the principles and rules of international law. Yet another tier down in the hierarchy, and to be applied only if the sources in subparagraphs (a) and (b) do not yield an answer, Article 21(1)(c) empowers the ICC to apply ‘general principles of law derived by the Court from national laws of legal systems of the world’.⁸

Concluding Article 21 is Article 21(3), which requires that all law listed in the Article be interpreted and applied by the Court in a manner consistent with ‘internationally recognised human rights’. The effect of this provision is not to make human rights a separate and additional source of law to those listed in Article 21(1), but to put in place a ‘mandatory principle of consistency’.⁹ In other words, Article 21(3) establishes a human rights consistency test. Whenever the Court is dealing with a provision or norm from one of the sources in Article 21(1), such as an article of the Rome Statute or a principle of international law, it must check whether its interpretation and application of that article or principle is consistent with human rights. If the proposed interpretation or application would not be consistent, then it must be reconsidered; if the relevant provision or principle is incapable of being interpreted and applied consistently with human rights, then the Court cannot apply it at all.¹⁰ Through Article 21(3) human rights law is placed in the position of lex superior as regards the rest of the sources in Article 21.¹¹

Academics and judges alike agree on the importance of Article 21(3) in the ICC legal framework. By academics it has been described variously as ‘one of the more important provisions of the Rome Statute’,¹² with ‘significant potential in shaping the practice before the ICC’, yet somewhat ‘mind-boggling’.¹³ By the ICC judges it has been employed to great effect, as can be seen from a selection of decisions taken in the course of the Bemba, Lubanga, and Katanga and Ngudjolo cases.

Firstly, in a decision in the Bemba case,¹⁴ the Single Judge held that deceased victims could participate in ICC proceedings (represented of course, by someone else). Even though no mention is made of deceased victims in the Rome Statute or Rules of Procedure and Evidence (RPE), when the relevant Rule¹⁵ is read in light of internationally recognized human rights, it was held to be ‘self-evident’ that a victim does not cease to be a victim because of his or her death.¹⁶ The human rights instruments referred to by Judge Kaul in the decision include, among others, the International Covenant on Civil and Political Rights, the regional human rights treaties from Europe, Africa, and America, the Convention on the Rights of the Child, and soft law instruments.¹⁷

Secondly, during the Lubanga trial, the prosecution created a situation where full disclosure of exculpatory evidence to the defence was impossible. Article 54(3)(e) of the Rome Statute allows the Prosecutor to collect evidence on the basis of confidentiality, but only for the purpose of generating new evidence. Trial Chamber I held that the provision has been misused, ‘with the consequence that a significant body of exculpatory evidence which would otherwise have been disclosed to the accused is to be withheld from him’.¹⁸ Deciding that a fair trial was no longer possible, the judges of Trial Chamber I unconditionally stayed proceedings against the defendant. The Appeals Chamber upheld the Trial Chamber decision, holding that through Article 21(3), the Court was empowered not only to grant a stay where justice could not be done, but that two types of stay were possible, namely permanent and conditional.¹⁹

Lastly, in the course of the Katanga and Ngudjolo case, the Presidency found that the Court was obliged to fund family visits for indigent defendants.²⁰ Reading the legal framework of the ICC in line with Article 21(3), it was found that international human rights instruments and jurisprudence recognize a right to receive family visits, and that for this to be an effective right the Court had an obligation to financially facilitate visits where necessary.²¹ The sources cited by the Presidency in support of this position included, among others, the case law of the European Court of Human Rights,²² case law of other international criminal tribunals,²³ and soft law instruments such as the Standard Minimum Rules for the Treatment of Prisoners.²⁴

Despite its use in case law to date, the precise contours of Article 21(3) remain somewhat unclear. While its inclusion in the Rome Statute demonstrates a commitment on the part of the drafters to making human rights fundamental to the ICC legal framework, the boundaries of this commitment are undefined.²⁵ The questions of which rights are included in the category of ‘internationally recognised human rights’, and what methodology the Court does/should employ when identifying them, remain sources of uncertainty.²⁶ These are important questions that will remain relevant; however, in addition to these questions, there is a new source of uncertainty and ambiguity that requires attention. In a series of decisions over the past few years, the ICC has set limits on the scope of Article 21(3), and it has done so without offering much by way of explanation or justification. These decisions touch upon issues that sit on the boundary line between the realm of the ICC and the realm of the state parties. Understanding the contours of these limits is important to a comprehensive understanding of the role that Article 21(3) plays within the ICC legal framework, and by extension, the extent to which human rights are protected at the ICC.

3. The limits of Article 21(3)

The best way to understand the limits that the Court has placed on Article 21(3) is to set out the instances in which these limits have been imposed. To the author’s knowledge, three such instances exist to date. While three may seem a relatively small number, in each instance the actual or potential harm suffered and the impact on the rights of the individuals concerned were significant. As such, they merit closer examination.

This section will set out the three instances, and the analysis of each instance is composed of three parts. The first part describes the factual background and the circumstances that gave rise to the situation. The second part proposes how the applicable law could have been interpreted, using the terms of Article 21(3), to meet the protection needs of the individual in those circumstances. The third part sets out the interpretation that was actually adopted by the ICC and the limits placed on Article 21(3). In Section 4, the author will seek to extract the underlying, often unarticulated, rationales that may be at work and which may explain the limited approach to Article 21(3) in these instances.

3.1 Detained witnesses and Article 21(3)

The first situation where the ICC placed limits on Article 21(3) involved a particular set of facts which led to four witnesses being detained in the ICC Detention Centre for an extended period of time. The Court found that Article 21(3) could not be used as a basis for the ICC to release the witnesses, even if the detention could be said to have become arbitrary.

In 2011, four witnesses were transferred from a prison in the DRC to give evidence in the Katanga and Lubanga cases. The individuals had been arrested and detained by DRC authorities some years earlier on suspicion of involvement in the death of UN peacekeepers and high treason, but at the time of the transfer had not been formally charged.²⁷ Detained witnesses can be an important source of insider information, especially if the reason for their detention is somehow linked to the crimes the accused is charged with. For this reason, even though they are a small group, special provision and arrangements are made for detained witnesses under Article 93(7) of the Rome Statute.²⁸ Article 93(7) imposes two conditions on the transfer of detained witnesses to the ICC: first, the witnesses must remain detained while at the seat of the Court (Art. 93(7)(a)); and second, they must be returned to the sending state once their testimony is complete (Art. 93(7)(b)). In the case of the four witnesses transferred in 2011, once their testimony was complete and they were set to be returned to the DRC, they claimed that they would be at risk of an unfair trial, physical harm, and even death if sent back. To avoid this outcome, the witnesses communicated an application for asylum to the Dutch authorities.

The first decision that the ICC had to make was whether, in light of the asylum applications, the detained witnesses could still be returned to the DRC, as required by Article 93(7)(b). Trial Chamber II cited numerous international instruments to establish that the right to seek asylum is an important right,²⁹ and held that to immediately return the witnesses would prevent them from effectively exercising this right. If removed from Dutch territory, it would be ‘impossible for them to exercise their right to apply for asylum’.³⁰ Deciding that Article 93(7)(b) could not be applied in a manner consistent with human rights – as required by Article 21(3) – the trial chamber ordered that the return of the witnesses be suspended.³¹ This suspension of return ultimately lasted for three years, during which time different chambers of the ICC, as well as all levels of the domestic courts of the Netherlands, wrestled with the many thorny issues that the situation gave rise to.³² During this period of time, the witnesses remained detained in the ICC Detention Centre, with the ICC requesting that the Netherlands take over custody of the witnesses and the latter refusing to do so.

This state of affairs gave rise to concerns with respect to the witnesses’ right to liberty, and the witnesses made applications to both the Dutch authorities and the ICC in an attempt to secure their release. Under Dutch law, the period of time for which asylum seekers can be held in detention is limited to 18 months, and only in exceptional circumstances.³³ As such, the witnesses stood a good chance of being released while their claims were being decided if they were transferred from ICC custody to Dutch custody. While a lower Dutch court agreed with the witnesses’ argument that their detention had become unlawful and ordered the Dutch state to negotiate their transfer to Dutch custody,³⁴ this was overruled on appeal.³⁵ Meanwhile the witnesses requested that the ICC review their detention, find it to have become unlawful under the circumstances, and order their immediate release regardless of the position of the Dutch authorities.³⁶

The legal issue before the ICC turned on the interpretation of Court’s obligation pursuant to Article 93(7)(b) to keep the witnesses detained.³⁷ In particular, is the Court obliged to maintain the witnesses’ detention no matter what, or is the obligation to detain the witnesses conditional on human rights considerations? Given the already demonstrated potential of Article 21(3) to secure high standards of human rights, it is highly plausible to argue that the Article 93(7)(b) detention obligation should be read as conditional. In a dissenting opinion, Judge van den Wyngaert strongly supported such an interpretation of Article 93(7)(b) and argued that the Trial Chamber had the competence, under certain circumstances, to review the detention of witnesses held in the ICC Detention Centre, and if necessary, to order their release from custody.³⁸ To interpret Article 93(7)(b) otherwise would mean that the ICC could be obliged to maintain a situation of indefinite and arbitrary detention, which would not be an interpretation consistent with human rights.

This conditional interpretation of Article 93(7)(b) was rejected by ICC judges, both at the Trial and Appeals levels.³⁹ It was held that the ICC did not have the power to review the four witnesses’ detention nor to order their release. The reasoning of the majority of Trial Chamber II was that, since the DRC ordered the detention, only the DRC was in a position to review it. The ICC had not itself issued an order for detention, and was instead merely maintaining custody of the witnesses.⁴⁰

Furthermore, in order for the ICC to be able to review the detention of the witnesses, the majority of the Appeals Chamber considered that it would need to examine the domestic basis for the detention, a process that would involve consulting material and orders from the Congolese judicial authorities and gathering observations from the witnesses, their counsel, and the Congolese authorities.⁴¹ In so doing, the ICC would ‘evidently be acting as a court of human rights’ and the ICC ‘was never conceived as such’.⁴² Article 21(3) was said not to require the ICC to come to a different conclusion, as there was no obligation on the Court ‘to ensure that State Parties respect internationally recognized human rights in their domestic proceedings’.⁴³ In other words, reviewing the witnesses’ detention would require an assessment of the DRC’s domestic proceedings, and conducting such an assessment would place the ICC in the position of a human rights court. Nothing in the Rome Statute, including Article 21(3), was designed to place the ICC in such a position. As a result, the ICC was bound to keep the witnesses detained until the DRC informed them otherwise.

3.2 Witness protection and Article 21(3)

The second instance of Article 21(3) being subject to a limitation concerns the scope of witness protection under the Rome Statute. The judges of the ICC held that Article 21(3) could not be used to justify a broad scope of witness protection.

This instance arose out of the same factual background as that discussed in the previous section. The situation of the four detained witnesses gave rise not only to questions concerning the right to liberty, but also to questions concerning the scope of the ICC’s witness protection obligations. In particular, the Court had to determine which types of risk the ICC was obliged to protect witnesses from, and which risks fell outside of its realm of responsibility. While the right to liberty concerns discussed in the previous section are likely to be rare and may not occur again, the question of the scope of witness protection, and the role of Article 21(3) in delimiting this scope, is relevant for all ICC witnesses.

The ICC has an obligation under Article 68(1) of the Rome Statute to take appropriate measures to protect the safety, physical, and psychological well-being of witnesses. Threats to witnesses can take many shapes and forms, and arise for many different reasons. If a witness were threatened by the supporters of an ICC accused against whom he or she testified, this would clearly be a situation in which the ICC should be involved. But where a witness was threatened for speaking out against their government, would the ICC be obliged to intervene?

From a purely textual perspective, Article 68(1) does not favour one way or the other. Rules 17 and 87 of the ICC RPE do link witness protection measures to ‘risk on account of testimony’, however, the RPE must be consistent with the Rome Statute, and in case of conflict, the Statute prevails (Art. 51). As a result, if Article 21(3) requires a broad scope of protection under Article 68(1), resulting in Article 68(1) conflicting with rules of the RPE, then under Article 51 the Rome Statute provision takes precedence.

On the basis of Article 21(3), the argument can be made that consistency with human rights requires the Court to protect witnesses from all risks to their safety, and not only from those risks connected to their involvement with the Court. Given the lack of direction in Article 68(1) itself and the broad formulation of Article 21(3), the former is certainly capable of supporting such an interpretation.⁴⁴ Furthermore, practical considerations support the need for a broad interpretation. It will in practice be very hard to ascertain whether a given risk is or is not connected to a witness’ testimony. For example, the ICC Victims and Witnesses Unit has stated that if the threat to a witness’ safety is linked to domestic violence that is unconnected to the person’s status as a witness, the VWU would not take action.⁴⁵ But can the reasons behind domestic violence always be determined? Perhaps acting as a witness caused the victim to be away from home for an extended period, and it was this that provoked the violence. Thorny questions of causation arise here. This difficulty was acknowledged by ICC judges: ‘in practice it will be impossible to determine whether any attempt to harm the witnesses will be linked to their testimony’.⁴⁶ It is certainly arguable that the difficulties that exist in distinguishing between risks connected with testimony and risks unconnected with testimony may amount to protection being granted or withdrawn arbitrarily. This arbitrariness, whereby witnesses are not treated alike in comparable situations, can lead to practices in which the Court could not always uphold the principle of equal treatment, a key aspect of human rights law.⁴⁷

This broad approach to witness protection was not the one taken by Trial Chamber II when this issue arose during the Katanga and Ngudjolo case. In delimiting the scope of witness protection under Article 68(1), the Chamber distinguished between three types of risk that a witness might face: (i) risk incurred on account of co-operation with the Court; (ii) risk arising from the broader human rights situation in the situation state; and (iii) risk of treatment that would amount to persecution such as would found an asylum claim under the Refugee Convention. It was acknowledged by the Chamber that while the general human rights situation in a country would influence a witness’ risk assessment, the three types of risk should not be conflated. The Chamber held that the ICC’s witness protection role under Article 68(1) was restricted to the first type of risk, and so protective measures only needed to be provided against risks arising specifically from co-operation with the Court, and not against the risk of human rights violations committed by the authorities of the witness’ country of origin more generally.⁴⁸

In its decision, Trial Chamber II explicitly rejected the idea that Article 21(3) would lead to a different conclusion. It held that the scope of witness protection could not be broadened based on Article 21(3) because ‘Article 21(3) does not place an obligation on the Court to ensure that State Parties properly apply internationally recognised human rights in their domestic proceedings’.⁴⁹ In the case at hand, the risk against which the witnesses sought protection was an unfair trial upon their return to their home country, hence the reference to ‘domestic proceedings’. That being said, the Chamber did not signal that the tripartite distinction of risk types was intended only to apply to the facts of the case, and the Chamber’s wording indicates a statement of a more general nature. As such, one can understand Trial Chamber II’s decision as establishing that there will be a range of risks that witnesses may face that will not qualify them for ICC protection, despite the instruction of human rights consistency in Article 21(3).

3.3 Complementarity and Article 21(3)

The third situation in which limits were placed on Article 21(3) concerned the admissibility of cases at the ICC. Admissibility at the ICC centres around the principle of complementarity, which sets up the ICC as a court of last resort, only coming into play when the domestic court system either does not or cannot deal with the commission of crimes listed in Article 5 of the Statute (war crimes, crimes against humanity, genocide, and aggression).⁵⁰

According to Article 17 of the Rome Statute, a case will be admissible before the ICC when the domestic authorities are unwilling or unable to prosecute an individual for the same case. Article 17(2) and (3) respectively set out what to consider when determining unwillingness and inability. A finding of unwillingness can be made: where the domestic proceedings are designed to shield the accused from justice; where there has been an unjustifiable delay in the domestic proceedings; or where those proceedings are not being conducted independently and impartially. In each instance the unwillingness must demonstrate a lack of intention to bring the accused to justice. Inability hinges on there being a total or substantial collapse of the domestic judicial system such that the state is unable to carry out proceedings against the accused.

The first situation to present a concrete complementarity challenge was that of Libya, referred to the ICC by the UN Security Council in 2011.⁵¹ Arrest warrants were issued against Muammar Gaddafi, Saif Al-Islam Gaddafi, and Abdullah Al-Senussi who held positions within the Libyan government at the time of the uprising in 2011. Proceedings against Muammar Gaddafi were terminated upon his death, but continued with respect to his son Saif Gaddafi and Al-Senussi. The Libyan authorities contested the admissibility of the cases before the ICC, arguing that they were willing and able to prosecute the accused in Libya. Both ICC prosecution and defence argued in favour of admissibility. In the case of Saif Gaddafi the Appeals Chamber confirmed the Pre-Trial Chamber finding that Libya was not investigating the same case as that before the ICC, and deemed the case admissible.⁵² Not so with Al-Senussi, with respect to whom the Appeals Chamber determined that domestic proceedings ongoing in Libya prevented the case from progressing at the ICC.⁵³

In terms of examining Article 21(3) and its limits, the decision to cease proceedings against Al-Senussi is the most interesting. The arguments made in that case touched on a question that had already been discussed at length in academic circles, namely the relationship between due process and complementarity.⁵⁴ In essence, the question was whether compliance in domestic proceedings with due process or human rights standards was relevant to admissibility before the ICC. The defence in Al-Senussi argued that the case must be admissible before the ICC because Libya could not guarantee the due process rights of the suspect, and this rendered Libya unwilling or unable.⁵⁵

Debates on this question turn on the interpretation and application of Article 17 of the Rome Statute. Based solely on the text of Article 17, Aksenova argues convincingly that the legal framework contains sufficient ‘wiggle room’ for due process considerations to play a role in admissibility decisions.⁵⁶ When this is combined with the need to adhere to the ‘mandatory principle of consistency’ in Article 21(3), it is very plausible to argue that the legal framework could be interpreted so that a lack of human rights protection in domestic proceedings could render a case admissible before the ICC. Although Heller ultimately dismisses what he terms the ‘due process thesis’ with respect to admissibility proceedings, he does concede that ‘the text of article 21(3) is capable of supporting such an interpretation’.⁵⁷

Once again, this is an instance where Article 21(3) was in theory available as a means to uphold human rights standards, but where the ICC decided on a more limited approach. In Al-Senussi, the Appeals Chamber held that the possibility of human rights violations occurring in domestic proceedings was not per se sufficient to find that a state was unwilling to prosecute an individual under Article 17.⁵⁸ The Appeals Chamber further held that if it accepted the defence’s arguments, the Court would come close to acting as an international court of human rights, which it was not designed to be.⁵⁹ While the Appeals Chamber does not expressly mention Article 21(3) in connection with this particular finding – unlike in the two other situations above, where Article 21(3) is explicitly referred to – it states that the finding is in response to the defence arguments, which in turn were based (to a large extent) on Article 21(3).

Also, unlike the two situations above, the Appeals Chamber did not stop at this point. Later in the decision, the Chamber took note of the human rights consistency requirement in Article 21(3), and while reiterating that human rights concerns were not per se enough to establish unwillingness, held that the Court need not turn a blind eye when national proceedings completely lacked fairness.⁶⁰ A case may be admissible if the domestic proceedings would be ‘contrary to even the most basic understanding of justice’ or when ‘the violations of the rights of the suspect are so egregious that it is clear that the international community would not accept that the accused was being brought to any genuine form of justice’.⁶¹ Interestingly, the Chamber did not attempt to reconcile this finding with its statements earlier in the decision⁶² that the ICC should not sit in judgment on the internal workings of the domestic legal system of states.

This finding by the Appeals Chamber appears to be an (at least partial) endorsement of an argument put forward by the Al-Senussi defence, centred on the term ‘justice’ in Article 17. Article 17(2)(b) and (c) allow a case to be prosecuted at the ICC if the state is behaving in a way that is ‘inconsistent with an intent to bring the person concerned to justice’. This can be read to mean that a case will be admissible if the state is making it harder to prosecute a suspect, thereby attempting to shield him/her from justice.⁶³ The defence, however, argued that the term ‘justice’ must be interpreted to mean a ‘decision which has been based on a fair trial’, and notes how Article 21(3) has been invoked previously by the Appeals Chamber to proclaim that ‘a fair trial is the only means to do justice’.⁶⁴ On this logic, if a domestic trial is not fair, it is not justice, and would not preclude admissibility under the terms of Article 17.

While the Appeals Chamber did not accept this argument in its entirety, it did agree that when the human rights violations reached a certain severity the trial in question could no longer be described as ‘justice’, and it connected this finding to the fact that human rights underpinned the Statute thanks to Article 21(3). In this instance therefore, the protective potential of Article 21(3) is still being limited because it will only be relevant to a narrow subset of violations, namely those that touch on the basic idea of justice. Where this threshold is not reached, Article 21(3) does not play a role. However, its relevance is not entirely dismissed, as it was in the previous two situations.

4. Understanding the limits of Article 21(3)

In each of the instances described above, the ICC judges took a limited approach to Article 21(3) and restricted the extent to which it could be used to include human rights protections in the ICC legal framework. However, also in each of these instances, the wording of both Article 21(3) and of the other articles in question would have supported an interpretation and application that provided more extensive protection to the individuals concerned. These situations stand in contrast to those in which Article 21(3) has been used to add far reaching protections to the ICC legal framework, of which those listed above in Section 2 from the Bemba, Lubanga, and Katanga and Ngudjolo cases are just some examples.⁶⁵ This raises the question: why the difference in approach?

Article 21 of the Rome Statute sets out the sources of law to be applied by the ICC and the hierarchy between them. Placed above the other sources in Article 21’s hierarchy are internationally recognized human rights. However, despite the wording of Article 21(3), human rights are not, it seems, at the very top of the hierarchy, with nothing else above them. There are rules that govern the workings of the international legal order as such that prevail over the entirety of the Rome Statute.

When states create international organizations to pursue co-operation and common goals, they transfer a part of their sovereignty to that organization. The realm of activity in which that organization can operate is dictated by the parameters of this transferred sovereignty. This is known as the principle of speciality or the principle of attributed powers. As set out by the International Court of Justice in the WHO Advisory Opinion:

international organizations are subjects of international law which do not, unlike states, possess a general competence. International organizations are governed by the “principle of speciality”, that is to say, they are invested by the states which create them with powers, the limits of which are a function of the common interests whose promotion those states entrust to them.⁶⁶

The principle of speciality is a reflection of the fact that states are the only actors recognized by international law as having full personality, competence, and powers.⁶⁷ The personality of other international actors, including international organizations, is derived from the original personality of states,⁶⁸ and as such they only have the competence and powers that are attributed to them (explicitly or impliedly) by states.⁶⁹ These legal concepts are fundamental to the way in which the international legal system was set up and operates.

In light of the above, the competence and powers of the ICC, as an international organization,⁷⁰ are limited to those attributed to it explicitly by means of the Rome Statute (and other relevant documents, such as the RPE), or implicitly through the notion of implied powers. The limits of the ICC’s competence and powers are connected to the purpose for which states established it: an international court of criminal jurisdiction, whose aim is to guarantee respect for and enforcement of international justice.⁷¹

With this in mind, it is worth revisiting the refrain repeated by the ICC in all three of the situations discussed above: that the ICC is not a human rights court. While no further elaboration is provided on what the Court means by this, it can be understood as a reflection of the principle of speciality. The arguments made by the parties in the three instances discussed would, according to the ICC judges, require the Court to exercise powers that they consider to be more generally associated with a human rights court than with a criminal court (for example, the power to look in detail at the factual situation in a country, the functioning of the government infrastructure, the content of domestic laws, and the standards followed in judicial proceedings). If these powers have not been attributed to the ICC by states, to take on such a role would contradict the principle of speciality. This is particularly so given the importance that the drafters placed on sovereignty in the Preamble to the Rome Statute.

When considered from this angle, it is unsurprising that the situations where we see limits being placed on Article 21(3) are also those situations that sit on the boundary between the realm of the ICC and the realm of state parties. In drawing this boundary, the ICC is making a determination of just how much sovereignty was transferred to it by states, as they are situations in which greater human rights protection equals greater scrutiny by the Court of a state’s domestic affairs.

For instance, in the situation concerning the detained witnesses, the legal basis for their detention in the ICC Detention Centre was the fact that they were detained in the DRC; the ICC was merely maintaining custody of them for the duration of their time at the Court. As a result, in order to examine whether the right to liberty had been violated, the Court would need to scrutinize the domestic legal proceedings which led to the original detention order. Such scrutiny would be an unjustified intrusion into the realm of the DRC to which the latter did not consent. Turning to the scope of witness protection, assessing the protection needs of a witness requires scrutiny of the conditions within that witness’ home country: what kinds of risk do they face? Are the local police able to offer protection? Will that witness be subject to human rights violations, either by the government or third parties? According to Trial Chamber II, the ICC is only obliged to protect a witness from risks that can be linked to their involvement with the Court. In those circumstances, scrutinizing the affairs of a state is justified because it is part of the role assigned to the Court by states. However, where there is no link between the risk and the testimony, such scrutiny would not be justified and the Court would be unduly intruding in domestic affairs. Finally, in operationalizing the principle of complementarity some probing into the activity of a state will always be necessary, in order to assess whether that state is willing and able to prosecute a case. The extent and type of probing is what has been contested, and the Appeals Chamber ultimately decided that it is not for the ICC to pronounce on the ‘quality’ of domestic proceedings.

So, the ICC can (and indeed must) use Article 21(3) to uphold human rights standards within its own proceedings: ensuring fair trials for accused, ensuring full participation of victims, etc. However, when human rights concerns arise at the boundary between the competence of the ICC and the proper realm of states, Article 21(3) is not a blank cheque. That is not to say that the ICC is precluded from ever scrutinizing state behaviour. Indeed, there are instances where it is accepted that it must do so. For example, the ICC exercises a supervisory role over domestic authorities when they are carrying out arrest proceedings and where they are enforcing sentences on behalf of the Court.⁷² In these instances the ICC must determine whether the rights of arrested and convicted persons have been respected at the state level. However, for these powers there is provision in the Rome Statute; they do not involve using Article 21(3) to confer on the ICC new powers and functions.

If this were the end of the story, then the limits placed on Article 21(3) would seem justified. However, the ICC judges don’t always follow this line of thinking. As was discussed in relation to Article 21(3) and complementarity, there are times when human rights considerations seem to trump concerns about the ICC acting as a ‘court of human rights’. In relation to complementarity determinations, the Appeals Chamber in the Al-Senussi case determined that human rights violations in domestic proceedings would not per se render a case admissible before the ICC, but if the violations reach a given level of severity then the case can be admissible. In other words, it is not appropriate for the ICC to sit in judgment on the quality of national proceedings, even if human rights are violated, unless those violations are very serious.

Another instance where ICC judges followed the same line of thinking as in Al-Senussi was in the Lubanga case. Because the Court’s analysis in this case did not purport to limit the role of Article 21(3), it is not included in the discussion in the previous section. However, the reasoning in relation to serious human rights violations is relevant here. Thomas Lubanga was arrested and detained by authorities in the DRC prior to an ICC arrest warrant being issued for him. Lubanga contended that the ill treatment he received during this time meant that he could not receive a fair trial, and that the ICC should decline to exercise jurisdiction over him on the grounds that it would constitute an abuse of process. Both the Pre-Trial Chamber and Appeals Chamber held that in order for abuse of process to be considered, it would have to be shown that there was ‘concerted action’ between the ICC and DRC authorities – that the ICC had somehow been involved in the violations suffered – and of this there was no proof.⁷³ However, both Chambers agreed that if there was evidence of torture or serious mistreatment, that would suffice for abuse of process to be invoked, regardless of whether or not there was concerted action. The logic applied by both Chambers is similar to that employed with respect to complementarity: the ICC cannot, without more, look into the behaviour of a state, unless the violations in question are of a particular nature.

Given the just discussed decisions in Lubanga and Al-Senussi, an explanation which calls upon the principle of speciality to justify the limits to Article 21(3) is incomplete. Instead, yet another layer must be added to the hierarchy of sources in the ICC legal framework: at the bottom are the sources of law listed in Article 21(1), followed by internationally recognized human rights in Article 21(3). Above internationally recognized human rights are the overarching rules governing the functioning of the international legal order, such as the principle of speciality. It would seem that in the top spot, above all these other sources, are human rights of a fundamental nature. Where such rights are involved, considerations based on speciality take second place. From the ICC’s decisions discussed above, key aspects of fair trial rights (as seen in Al-Senussi) and the prohibition on torture (as seen in Lubanga) would appear to qualify for this category, but beyond this there is no indication.

The two additional layers in the Article 21 hierarchy explain the approach of the ICC judges in the situations addressed in this contribution, and elucidate the underlying rationale for the limitations to Article 21(3) not articulated by the ICC judges themselves. However, given the undefined contours of these additional layers, the scope and potential of Article 21(3) for infusing human rights standards into the ICC legal framework remains unclear.

5. Concluding reflection

There are many more instances in which Article 21(3) has led to greater protection of human rights than where the ICC judges have limited its use. Indeed, the author only succeeded in finding the three discussed above. However, it is also true that the situations where Article 21(3) cannot bolster human rights protection are those that potentially involve quite significant rights violations – the situation of the detained witnesses in particular is evidence of this, given that the individuals were kept in detention for three years.

In light of the above discussed decisions, the following concluding reflection can be offered: that the protective potential of Article 21(3) is greater where the core of the Court’s activities is concerned, but is more limited when dealing with the periphery of the Court’s activities. The core refers to the essence of the ICC’s competence, namely the conduct of criminal proceedings. How witnesses are protected when testifying in Court, the fair trial protections guaranteed to suspects, and technical rules of evidence, are all matters connected to the ICC proceedings as such, and are matters over which the ICC has been given full control and competence by the state parties. The periphery, on the other hand, refers to those areas of the Court’s activities where it must navigate the line between its own competence and a state’s competence. Here human rights give way to what one could call ‘overarching’ rules of international law, such as the principle of speciality or attributed powers. Human rights are not so fundamental that they can override such considerations and extend the ICC’s powers and competence to examine a state’s domestic affairs. That being said, human rights at the periphery sometimes do trump these overarching rules, if the violation in question is sufficiently serious and the right in question is of a fundamental character.