1. Introduction
Increasingly, formerly domestic policy areas have been absorbed into issue-area specific, sectorally-limited regimes beyond the nation state, representing the fragmentation of social systems at the international level.Footnote 1 Stephen Krasner understood international regimes as ‘a set of implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge in a given area of international relations’.Footnote 2 More specifically, they constitute ‘a union of rules laying down particular rights, duties and powers and rules having to do with the administration of such rules, including in particular rules for reacting to breaches’.Footnote 3 Areas of competence between such regimes are usually not clearly delineated, resulting in overlaps. This can produce parallel action, mutual blockade, or even clashing decisions by the respective institutions.Footnote 4 Because of the absence of a hierarchy of legal norms or courts in the international arena such conflicting competences or ‘regime collisions’ are not being systematically resolved. Given the resulting issues for the effectiveness and legitimacy of regimes involved, theoretical approaches from different disciplines share the assumption that such overlaps constitute a problem calling for a solution (Section 2).
This article examines ‘collision rules’ as a solution for the issues caused by overlapping normative regimes. Ideally a collision rule would either resolve the collision in substance, or it would prescribe the conditions for prioritizing one regime's solution over the other. However, as regimes represent different logics and sets of interests,Footnote 5 such a rule would have to be considered legitimate by the parties involved in order to provide for a successful collision management. In this paper, we build on Thomas Franck's work on the legitimacy of international norms to develop criteria for legitimate collision rules (Section 3).
We exemplify the problem by reference to collisions between the International Criminal Court (ICC) and the UN Security Council (Section 4). The ICC is competent to prosecute ‘the most serious crimes of concern to the international community as a whole’ (Art. 5 of the Rome Statute) even in ongoing situations of armed conflict or crisis that are being dealt with by the Security Council. A collision rule was, therefore, introduced into the Rome Statute (RS) in the form of Article 16, which provides for a (temporally limited) priority of the security logic over the (criminal) justice logic (Section 4.1). However, although the norm production process – the Rome Conference – has been praised as an approximation of rational discourse (Section 4.2),Footnote 6 all instances of the application of Article 16 so far have been highly controversial (Section 4.3). Indeed, conflict may intensify once the new Articles on the crime of aggression enter into force, since this has to be considered the core competence of the Security Council.Footnote 7 While norm application is always controversial to some extent, (external) application control of Article 16 could contribute to its legitimacy (Section 5). In light of this example, we propose criteria for a ‘good’ collision rule that could avoid some of the problems that Article 16 faces (Section 6).
2. Regime collisions – an interdisciplinary approach
The problem of an unco-ordinated plurality of competences of different international institutions has been discussed both in the sub-discipline of international relations as well as in international law. The crucial aspects of the phenomenon will be highlighted by summarizing the main arguments of one relevant approach from each discipline.
2.1. Regime theory: overlaps and interplay management
In political science, the ever growing number of international institutions with area-specific, often overlapping competences has been a topic especially of regime theory, a neo-institutionalist theory of international relations, using terms such as ‘institutional linkage’,Footnote 8 ‘institutional interaction’,Footnote 9 ‘regime overlap’,Footnote 10 ‘regime interplay’,Footnote 11 or ‘regime complexes’.Footnote 12 Starting from the assumption that international institutions influence each other, rather than simply compete for power (as realists would have it), these approaches focus on identifying different types of interaction and their respective positive (synergetic) or negative effects on the efficiency of the interacting institutions.Footnote 13
The ‘interaction’ type of overlapping regimes is one where ‘the functional scope of one regime protrudes into the functional scope of others’.Footnote 14 In cases where the overarching goals and norms of the regimes or the concrete rules on the attainment of these goals diverge or are even mutually exclusive, this tends to lead to conflicts:Footnote 15
Environmentally motivated trade restrictions will be considered as undesired interference with international trade from the perspective of the WTO, which aims at liberalizing international trade and seeks to abolish trade obstacles. The same measures are appreciated as effective means for enforcing international environmental standards from the perspective of international environmental regimes. . . .Footnote 16
This kind of ‘normative interplay’Footnote 17 does not necessarily favour the interests of powerful actors since it also creates opportunities for strategic action for less powerful actors.Footnote 18 What is clear, however, is that it poses a potential risk for the legitimacy of international norms as well as for norm compliance, i.e., their effectiveness.Footnote 19 From the point of view of regime theory, such a conflict therefore requires a conscious ‘interplay management’, understood as ‘deliberate efforts by participants in tributary or recipient regimes to prevent, encourage, or shape the way one regime affects problem solving under another’.Footnote 20 Regime theorists primarily see this as the task of political decision-makers,Footnote 21 by an exchange of information between decision-making institutions, common programme planning, co-ordination of substantive decision-making activities, etc. Others also advocate co-operation between the regimes’ administrative or judicial bodies, e.g., by negotiating division of labour arrangements, or by drafting common guidelines.Footnote 22
In order to enhance legitimacy, regime theorists also emphasize that norms should be determinate and coherent.Footnote 23 In the area of normative interplay, what matters is ‘whether states, when unable to apply the rule consistently in all contexts, can justify the deviation in terms of principles that are capable of generalization’.Footnote 24 Since such generalizable exception principles are often vague and open to different interpretations, a proactive interplay management is required, e.g., through the communal development of codes of conductFootnote 25 or saving clauses.Footnote 26
2.2. Systems theory: collisions and heterarchic networks
International lawyers often approach the issue of overlapping regimes as a challenge to the presumed unity of law,Footnote 27 resulting from a problematic absence of a hierarchy of normsFootnote 28 and courts at the international level. Others question this technical view, in particular legal scholars of systems theory, an overarching theory of society founded by the German sociologist Niklas Luhmann.Footnote 29
Systems theory describes the global society as differentiated into ‘autonomous social systems’ such as the political or the economic system, each with their own logic. Where the legal system interacts with such other systems, it forms legal regimes built around their logics, translating them into legal terms. This also comprises autonomous private regimes such as the lex mercatoria or the lex digitalis.Footnote 30 As a result, global law is necessarily and irreversibly fragmented into different legal regimes which follow different normative logics, such as trade, security, environmental protection, etc. The competences of these diverse regimes tend to overlap and collide, as each seeks to universalize its own logic (Eigenrationalität)Footnote 31 at the expense of the others. The proliferation of international courts, tribunals and quasi-judicial bodies increases the risk of contradictions between individual decisions.Footnote 32 The main problem with such ‘regime collisions’ is the fact that it is not only normative orders that collide but also underlying conflicting societal goals and interests:Footnote 33 ‘At core, the fragmentation of global law is not simply about legal norm collisions or policy-conflicts, but rather has its origin in contradictions between society-wide institutionalized rationalities, which law cannot solve. . . .’Footnote 34 This means that overlapping legal regimes are not a mere technical consistency problem, and that ‘one-dimensional solutions’, either legal or political, would underestimate the legitimacy and effectiveness problems caused by the fragmentation of law.Footnote 35 Indeed, the hope for a normative unity of law is seen as illusory – a paradise lost.Footnote 36
From a systems theory point of view, collisions can only be resolved via ‘heterarchical forms of law that limit themselves to creating loose relationships between the fragments of law’.Footnote 37 Such network structures function by means of mutual observation and mutual irritation as systems interact and react to one another, each in their own logic but responding to the other's logic, in order to establish compatibility.Footnote 38 What is required to resolve regime collisions is therefore a legal form for autonomous regimes to reflect other regimes’ interests, e.g., in terms of a duty to consider or deference clauses.Footnote 39 Given the democratic deficit of such network structures, procedural safeguards are needed to ensure transparency and mutual accessibility, as well as participation and deliberation, alterity orientation,Footnote 40 a fair balance of interests and the neutralization of power.Footnote 41
2.3. Regime collisions: common insights
Despite differing basic assumptions, both approaches discussed share a common starting point: the fragmentation of the international system into issue-area specific regimes with sectorially limited authority. Following this, regime collisions can be defined as conflicts that arise from overlapping competences of two or more regimes with divergent goals and norms or rationalities. Such collisions can involve both state and non-state regimes. They are not just technical, legal problems; they point to a conflict of goals and interests that are being articulated in the language and logic of the respective regime. It is this very clash of logics that makes the resolution of such collisions difficult. In addition, the colliding regimes are brought into being and supported by overlapping or even identical arrangements of more or less powerful actors with different interests, whose options for action are driven by each regime's specific momentum.Footnote 42
Another point of consensus between the approaches discussed is the fact that such collisions may pose a problem both for the legitimacy and the effectiveness of the regimes in play. Neither regime theory nor systems theory consider it a desirable or even feasible solution to establish a hierarchy of norms and institutions modelled on the nation state. This leaves only heterarchical co-ordination efforts between regimes as a means to reduce collisions, for which different mechanisms are discussed. From the point of view of regime theorists, solutions range from division of labour arrangements between competing institutionsFootnote 43 to conscious political action,Footnote 44 for example by devising an overarching framework for the interacting institutions, e.g., in the form of ‘principled priority’ of matters such as the environment or human rights, or through joint management by means of creating horizontal structures for coordination.Footnote 45 Systems theorists focus on network structures between regimes and horizontal co-ordination mechanisms such as default deference.Footnote 46
The third point of consensus is that such efforts, whether legal or political, should meet several requirements, such as transparency, participation rights and a fair balance of interests, in order to be able to preserve the legitimacy and efficiency of both regimes.
This paper will examine collisions between the UN security regime, embodied in the Security Council, and the ICC's international criminal justice regime and focus on a particular type of ex-ante co-ordination, namely collision rules. Such rules serve to avoid or resolve collisions by way of co-ordination and assignment of competence. In the case of the Security Council and the ICC, such a rule can be found in Article 16 of the RS. The following sections will further develop the requirements for successful collision management (Section 3) and then apply them to Article 16 of the RS (Section 4).
3. Ideal conditions for a successful collision management: Thomas Franck reconsidered
In light of our definition of a regime collision, collision rules can be seen as successful if they sustainably resolve the collision problem, i.e., the threat to the legitimacy and effectiveness of the regimes in play. Therefore, they have to couple each regime to the other regime's logic and find an optimal balance between the colliding regime logics.
Since collision rules have to be considered legitimate in order to successfully resolve collisions – in particular when they prescribe conditions for prioritizing one regime logic over the other – we refer to Thomas Franck's work on the legitimacy of international normsFootnote 47 and adapt it to the particular context of collision management.
3.1. Adapting Thomas Franck's legitimacy model to the particular context of collision rules
Following Weber, Franck considers a norm's legitimacy to be central to its prospect of compliance, defined as ‘that quality of a rule which derives from a perception on the part of those to whom it is addressed that it has come into being in accordance with right process’.Footnote 48 The quality of a rule itself, but also the process of rule production and its application are crucial.
According to Franck, first, a norm's legitimacy depends on its determinacy;Footnote 49 it has to ‘convey a clear message’Footnote 50 that limits both its flexibility and its potential for manipulation.Footnote 51 The clearer the instructions of a norm to its addressees, the smaller the room for conflict. Transferred to the regime collision context, this means striking a balance between the regime logics or prioritizing one over the other in clearly defined situations, for specific reasons. Moreover, the collision rule should prescribe the responsiveness between the two regimes as a reflection of the interests and goals pursued by the other regime.
Second, what is important to the legitimacy of a norm is its symbolic validation, which, among others, depends on its pedigree.Footnote 52 Thus, a rule may derive authority from its historical origins or its cultural deep-rootedness.Footnote 53 The pedigree of the rule-making institution and process is also relevant, since in the complex and dynamic process of socialization that leads to compliance, how norms are produced, interpreted and applied is equally important.Footnote 54 In a regime collision context, this would imply involving both regimes equally in developing the collision rule and guaranteeing that the different interests or regime logics are taken into account in the norm production process.
Since the determinacy of a norm is limited, not least by the degree of complexity of the issue it regulates, it follows thirdly that the norm application must be assigned to a forum that balances the remaining substantive elasticity of the norm by making its application predictable by way of process determinacy.Footnote 55 This does not necessarily have to be a court; what matters is that the institution is considered legitimate by those to whom it applies the norm. This is the case if the institution has a clear and coherent mandate, is itself legitimately established (pedigree), and makes decisions based on general fairness principles instead of its own interests.Footnote 56 In a regime collision context, this means in particular that the institution deciding on the application of the collision rule must not unilaterally represent or favour one regime logic, but rather has to be accepted by the participants of both regimes.
Furthermore, following Dworkin's integrity principle, norm application requires that the norms of a system must form a logically coherent body and guarantee the equal treatment of similar cases.Footnote 57 Exceptions and compromise have to be rationally justified by appealing to generally recognized principles. Procedural secondary rules can promote this.Footnote 58 In a regime collision context too, this means establishing clear criteria for the application of a collision rule as well as procedural secondary rules for implementation control.
3.2. Semantic struggles over norm application
Franck himself points out that norm determinacy is an ideal rather than a concrete objective, given the dilemma that ‘clarity is far from simplicity’:Footnote 59 ‘A rule finely calibrated to reflect complex considerations, embodying a textured system of regulatory and exculpatory principles, may suffer legitimacy costs because it invites disputes as to its applicability in any particular case.’Footnote 60
Of course, it will be hardly possible to anticipate all possible constellations of regime collisions and resolve them once and for all through a collision rule that leaves no room for interpretatory struggles. Such struggles around norm application are not, as such, an indicator of a lack of legitimacy. They arise because regimes are not the neutral problem-solvers they pretend to be, but stand for political choicesFootnote 61 – choices that are also in play where collision rules are applied.
While the norm production process engages justification discourses (Begründungsdiskurse) that address the reasons for having the norm, its validity and shape in the abstract, application discourses (Anwendungsdiskurse) pertain to the norm's appropriateness in a concrete situation.Footnote 62 Inevitably, norms will remain unsaturated with respect to the need for interpretation arising from specific constellations of application,Footnote 63 which means that they can continuously be challenged by different interpretations.Footnote 64 Even legitimate norms will therefore engender interest-driven application discourses, in which ‘[s]uccessful interpretations, that is, interpretations that find acceptance, can . . . be conceived as expressions of power.’Footnote 65 Law becomes the ‘battleground’ for semantic struggles between actors trying ‘to implement the use of a particular expression and conception in relation to specific facts that suits their interests or normative convictions’.Footnote 66 Law thus provides a normative framework, the material infrastructure for political debate.Footnote 67
However, legal discourse subjects all participants – strong and weak alike – to its particular logic. All interpretative claims have to take a ‘particular argumentative form’.Footnote 68 Disqualifying arguments based on power alone by requiring a normative justification capable of universalization,Footnote 69 legal discourse opens up opportunities for marginal actors to inscribe meanings furthering their interests, too. While only states can adopt international treaties, interpretations of particular legal rules by non-state actors such as NGOs may lead to directions not intended by the states. Franck's neutral monitoring instances, such as judicial bodies applying generally recognized fairness criteria, can provide a framework for such ‘interpretative struggles’. They can thereby contribute to the mobilization of legitimacy resources by less powerful actors, adding to the legitimacy of the outcome.
The following sections will analyse Article 16 of the RS as an example of a collision rule between a security and a justice regime in light of these criteria.
4. Peace versus (criminal) justice – regime collisions between the UN Security Council and the ICC
Established in 2002, the ICC constitutes the first permanent international organ to deal with the four crimes of the Nuremberg tradition: war crimes, crimes against humanity, genocide, and – presumably from 2017 onwardsFootnote 70 – the crime of aggression (Arts. 1 and 5 of the RS). These crimes usually take place in the context of a threat to or a breach of international peace and security, placing them under the concurrent competence of the Security Council. According to Article 24(1) of the UN Charter (the Charter), this organ bears the main responsibility for the maintenance or restoration of international peace and security,Footnote 71 and has the power to take measures in this respect (Art. 39 of the Charter).Footnote 72 Both institutions can therefore be tackling the same conflict.Footnote 73 For example, the Security Council has passed a great number of resolutions on the situation in Darfur, ordering measures for the restoration of peace and security.Footnote 74 At the same time, the ICC is investigating the criminal responsibility of several individuals, in this case upon referral from the Security Council itself.Footnote 75
In such cases, given the different logics or rationalities of these institutions, the collision potential is high. The Security Council is first and foremost a (power) political organ that addresses conflict with political or even military measures, while the ICC is in charge of legal assessments. In short, it is ‘peace v. justice’Footnote 76 – although this juxtaposition relies on ideal types that do not fully reflect the complexity of either institution.Footnote 77 While the maintenance of international peace and security is primarily a political matter, it can also involve legal assessments, for example in determining whether an act of aggression has been committed (Art. 39 of the Charter). The Security Council, often characterized as the ‘executive’ of the international order,Footnote 78 has also recently taken to passing abstract and general ‘thematic’ resolutions under Chapter VII of the Charter, for example on terrorism or on the proliferation of weapons of mass destruction,Footnote 79 which have been likened to international legislation.Footnote 80 On the other hand, prosecuting ‘the most serious crimes of international concern’ (Art. 1 of the RS) can involve strategic choices that have little to do with simply applying the law, both when it comes to securing the support of the state parties the ICC so dearly relies onFootnote 81 and in managing its institutional relationship with other international bodies, such as the Security Council.
Still, in their relationship with one another, it is the ideal-typical ‘peace v. justice’ that can lead to collisions. Thus, an intervention by the Security Council into ICC matters for political reasons can threaten the Court's independence and hence the good ‘justice’. By contrast, international criminal justice has usually been considered to serve the good ‘peace’,Footnote 82 the rule of law being part of a positive concept of peace.Footnote 83 In exceptional cases, however, criminal investigations against key actors of an ongoing conflict can threaten the peace process by reducing the incentive for negotiations. In such situations, where the goals of one regime can only be realized at the expense of the other’s, mere co-ordination as in the Relationship AgreementFootnote 84 between the UN and the ICC is not enough. There, both parties recognize each other's mandate and responsibilities – the ICC's independence as well as the UN's responsibility for world peace – but establish only general co-operation duties.
A collision rule can be found in Article 16 of the RS, since its aim is to balance the two rationalities. Ideally, such a rule would couple the highly judicialized area of competence of the ICC with the political rationality of the Security Council, while also tying the politically motivated Security Council to the logic of law enforcement by prosecution. However, analysis shows that Article 16 serves as a one-way street for the Security Council to assert its political interests (albeit temporarily) instead of establishing a mechanism for mutual responsiveness. Our analysis will first address deficits in the norm production process and in norm determinacy (Sections 4.1. and 4.2.) and then focus on the first instances of norm application (Section 4.3.), to finally propose avenues for reform, both for an internal and for an external stabilization of its application (Section 5).
4.1. The Rome Statute's collision rule: Article 16
The Security Council has not only the power to refer cases to the ICC in the interest of international peace and security under Chapter VII of the UN Charter, thereby conferring jurisdiction even over non-state parties (Art. 13(b) of the RS). It can also temporarily defer investigations and prosecution under Article 16 of the RS. This latter power takes the form of a ‘request’ to the ICC, also passed as a resolution under Chapter VII of the UN Charter. Where such a request is made to the ICC, ‘[n]o investigation or prosecution may be commenced or proceeded with under this Statute for a period of 12 months. . .’ (Art. 16 of the RS). Such a request is binding upon the ICC and can be renewed under the same conditions.
This collision rule takes account of the fact that, in exceptional cases, criminal prosecution may not serve but threaten international peace and security, whose maintenance is the Security Council's prime task. Under specific conditions, it thus subjects the activity of the ICC to that of the Security Council, establishing a – temporary – priority of the Security Council and its ‘peace logic’ over the ICC and its ‘justice logic’. At first sight, this norm appears to fulfil the determinacy requirement, as it contains clear criteria and consequences. However, the justification controversies of Article 16's norm production process (Section 4.2.) have resurged with force in the first application cases (Section 4.3.) and point to specific deficits of Article 16 that go beyond the usual semantic struggles.
4.2. Contested collisions I: norm production process
Leading up to the Rome Conference, the relationship between the Security Council and the ICC was one of the most controversial issues. The opposing arguments mainly referred to these institutions’ respective logics: the Court's judicial independence as opposed to the Security Council's responsibility for the maintenance of international peace and security.
The original draft of the StatuteFootnote 85 provided that the ICC's jurisdiction would only be engaged by Security Council referrals under Chapter VII. It also barred any ICC activity on cases that the Security Council was dealing with itself under that Chapter. This would have put the ICC entirely at the disposal of the Security Council. Some delegates argued that this reflected the Security Council's responsibility to maintain international peace and security.Footnote 86 Others voiced strong concerns,Footnote 87 arguing that the judicial function of a court should not be subjected to the actions of a political organ and warned of possible instrumentalization.Footnote 88 Moreover, the need for such a rule was questioned, given that there was none in the Statute of the International Court of Justice (ICJ) either.Footnote 89 Progress was finally made by ‘turning’ the rule ‘around’ into an optional request for a deferral.Footnote 90 However, the controversy remained alive, mainly due to the limited legitimacy of the Security Council. Several alternatives were proposed, including a referral power for the UN Human Rights Commission (now Human Rights Council),Footnote 91 or a subsidiary referral power for the General Assembly.Footnote 92 None of these proposals was accepted, and the final text of Article 16 was agreed on the last day of the Conference.
Ideally, a norm production procedure should be considered legitimate by the affected parties. For a collision rule context, this not only requires a good pedigree of the norm producing institution but also taking the different regime logics into account during the production process. It appears that the conditions in Rome were quite auspicious in this respect. The Rome Conference was attended by more than 160 governments from all regions of the world and approximately 200 representatives of civil society organizations. Not only were both regime logics represented, but also, instead of mere cost-benefit calculations, a real persuasion process took place,Footnote 93 superseding the ‘political reality frame’ of the Draft Statute with a ‘public interest frame’ focused on an independent and impartial Court, in an approximation of the ideal of Habermasian rational discourse.Footnote 94
However, an unresolved concern remained: entrusting the application of the collision rule to the Security Council, whose actions many states saw not as a legitimate representation of the interests of all UN member states, but as power-driven, ‘cynical exercises of authority by great powers’.Footnote 95 Moreover, this organ one-sidedly represents the security interests of one regime (or rather, the interests of the most powerful members of this regime), without having to take the justice interests of the other regime into account. Whether or not the ICC's interests will be considered depends entirely on the political choices of the members of the RS represented in the Security Council. Therefore, shortfalls in Article 16's norm content spill over into norm implementation, as the following sections demonstrate.
4.3. Contested collisions II: norm application
4.3.1. Norm application exercised: peacekeepers
The first Article 16 case arose in 2002, the very year the RS came into force. Resolution 1422 (2002) of the Security Council contained a blanket deferral request for 12 months for members of UN peace missions whose home states had not ratified the RS. The US had pushed this through by threatening to discontinue its participation in such missions (the United Nations Mission in Bosnia Herzegovina (UNMIBH), in particular).Footnote 96 In the course of the Security Council debate, in which numerous delegations participated, several state representatives argued that the request was ultra vires,Footnote 97 and that the Security Council was abusing Article 16, which was meant to address concrete individual cases of threats to international peace and security, as a blanket ex ante immunity clause.Footnote 98 Some saw this in substance as a reform of the RS, for which the Security Council had no authority.Footnote 99 Such double standards would undermine the legitimacy and effectiveness not only of the ICC, but also of the Security Council and of the UN as a whole.Footnote 100 In the end, Resolution 1422 was passed unanimously in order to protect the UNMIBH, but without the automatic renewal the US had proposed.
The renewal by Resolution 1487 (2003) was no less contentious.Footnote 101 Critics could advance a new argument: By now, the Court's staff had started working, whose high moral character, integrity, qualifications and competence many delegates emphasized was the best protection against possible politically motivated trials, as feared particularly by the US.Footnote 102 This time three Security Council members abstained from the vote: France, Germany and Syria.
The following year the US withdrew a third draft request, possibly because of the accusations relating to Abu Ghraib,Footnote 103 and embarked upon an alternative strategy: Resolution 1497 (2004), not citing Article 16, places peacekeeping staff in Liberia under the exclusive jurisdiction of their respective home states.Footnote 104 The Security Council has now extended this practice to cases it has explicitly referred to the ICC,Footnote 105 albeit with an explicit reference to the possibility of an Article 16 deferral.
4.3.2. Norm application rejected: Al-Bashir
The Security Council was also sharply criticized for a refusal to request a deferral. After an unsuccessful period of appeasement following the Security Council referral of the situation in Darfur in 2005,Footnote 106 the Office of the Prosecutor (OTP) had, in 2008, extended its investigations to acting Sudanese President Omar Al-Bashir and requested an arrest warrant.Footnote 107 The Council of the Arab League criticized this the following day, and was joined by the African Union (AU) within a week.Footnote 108 The AU demanded a deferral request under Article 16, arguing that criminal proceedings at this point in time would not be in the interest of the victims and of justice, but would threaten the peace process.Footnote 109 Similar arguments were advanced by the Organization of the Islamic Conference and by the Non-Aligned Movement.Footnote 110
The UN Security Council considered this ten days later when discussing the extension of the United Nations African Mission in Darfur (UNAMID) mandate in Darfur. Some states, surprisingly including the US, criticized the opposition created between peace and justiceFootnote 111 and emphasized the need for trials given the scale of the crimes,Footnote 112 while others, such as Russia and China, voiced concern for the detrimental effect on the peace process.Footnote 113 No decision was taken: The Security Council passed the mandate in Resolution 1828 (2008), only briefly ‘taking note of’ the AU's position and of ‘concerns raised by members of the Council regarding potential developments subsequent to the application of the Prosecutor’, and stating its intention to ‘consider these matters further.’Footnote 114 Despite repeated ‘requests for a request’, no further action was taken.
President Al-Bashir reacted by expelling aid organizations from Darfur, thereby gravely exacerbating the humanitarian crisis.Footnote 115 Frustrated by its repeated appeals to the UN, the AU, at its summit at Sirte, decided to ask its members to deny co-operation to the ICC in this case.Footnote 116 Al-Bashir has since been able to travel to several African ICC member states who failed to comply with their obligation to extradite him to the ICC.Footnote 117
The conflict also led to demands for a reform of Article 16. In reference to debates in Rome, the AU suggested empowering the General Assembly to request deferrals ‘in cases where the Security Council has failed to take a decision within a specified time frame, in conformity with UN General Assembly Resolution 377(V)/1950 known as “Uniting for Peace Resolution”. . .’.Footnote 118 This proposal did not make it onto the agenda of the 2010 Review Conference at Kampala, but is being discussed by a working group of the Assembly of State Parties.Footnote 119 Alternative proposals suggested taking away ICC jurisdiction by establishing a hybrid tribunal for Darfur.Footnote 120
Despite these experiences, and concerns in the Security Council about sequencing peace and justice,Footnote 121 the referral to the ICC of the on-going conflict in Libya passed unanimously in February 2011.Footnote 122 Again, the AU called for a deferral request,Footnote 123 but no move was made in the Security Council. Whether this affected the chances of a peaceful resolution is unclear; in any case, Gaddhafi was killed in late 2011, his aides captured. Libya subsequently raised complementarity objections under Article 17 of the RS, and while the ICC rejected these in the case of Saif Al-Islam,Footnote 124 it upheld them in the case against Mohammad Al-Senussi, permitting a trial in Libya.Footnote 125
Meanwhile, the critique by African governments and the AU of peace-threatening actions by the ICC became more entrenched when the ICC brought charges against six Kenyan suspects, including two sitting cabinet members, in its first proprio motu case in 2012.Footnote 126 When the two cabinet members won the elections in April 2013, the AU Assembly decided that no charges should be commenced or continued before any international court or tribunal against any serving AU head of state or government.Footnote 127 Moreover, it condemned the politicization and misuse of indictments against African leaders by the ICC and demanded that the trials of now President Kenyatta and Deputy President Ruto be suspended until they complete their terms of office.Footnote 128 The Security Council declined in November with only seven votes in favour; all of the ICC parties as well as the US abstained.Footnote 129 A parallel AU initiative in the ICC's Assembly of State Parties to preclude prosecutions against sitting heads of states and governments failed as well.Footnote 130
Meanwhile, the ICC has had a hard time accomplishing its work; in December 2014, the OTP had to drop the charges against Kenyatta due to lack of evidence.Footnote 131 Moreover, the Security Council appears to find it easier to refer cases and prevent deferrals than to support the actual investigations. Just days after throwing in the towel in the Kenyatta case, Prosecutor Bensouda announced to the Security Council that, given its lack of back-up for her work in Darfur, she will ‘hibernate’ investigations there and shift resources to more promising cases.Footnote 132
4.3.3. Application of Article 16 of the RS in light of the ideal type
These first application cases prove that the concerns articulated in the norm production process are playing out in practice.
First, Article 16's determinacy is diminished by the fact that the Security Council has wide discretion in determining the existence of a Chapter VII situation. Indeed, given the extraordinary powers of its permanent members, decisions are often selective, driven by their partial interests. Such an institution cannot provide stable procedural expectations, as it does not oblige its members to act coherently and rationally or to follow generally recognized fairness principles. For want of secondary rules, the implementation of Article 16 has remained a matter of discretion to the Security Council, lacking adherence to general principles.
Moreover, the pedigree of the Security Council is in bad shape, claims for reform are widespread. The Security Council has seen its legitimacy wane since the Second World War, as the veto power of the P-5 increasingly fails to reflect their actual power and influence. Article 16 thus allows for possible instrumentalization by an organ with little legitimacy, threatening the independence of the ICC.
Third, the application of Article 16 of the RS was not entrusted to a neutral institution. On the contrary, the Security Council is biased insofar as it is the prime proponent of the security logic. It is free to impose its rationale on the ICC, undermining the legitimacy of the ICC by making it appear as a pawn of the Security Council and not as an independent and impartial institution.Footnote 133 Since Article 16 does not require responsiveness, there is no mechanism securing the consideration of the criminal law regime's justice interests. Although ICC members France and the UK have the power to veto a deferral request that they feel may damage the Court, this is a matter of political choices, as Resolution 1422 (2002) demonstrates.Footnote 134 In otherwise allowing one regime to supersede the other's logic at will, Article 16 does not warrant an optimal balance of the colliding regime logics – as the first application cases prove. The fact that the US finally refrained from calling for an extension of Resolution 1422 (2002) was owed to specific circumstances and is no guarantee against future deferral requests that may harm the ICC's legitimacy.
Moreover, while the priority of the security logic can only be established by the Security Council, it has no obligation to exercise this power. In fact, the Security Council is even free to leave some collisions unresolved by refusing to apply Article 16. If it fails to do so, although on-going criminal investigations present a threat to international peace and security – as it was claimed in several cases – the ICC has no power to desist: A discretion of the OTP exists only ‘in the interests of justice’, not of peace (Art. 53 of the RS). Thus, there is no mutual responsiveness making up for each regime's ‘tunnel vision’.Footnote 135 While, arguably, Security Council referrals under Article 13(b) of the RS can be seen as bolstering the ICC's legitimacy and possibly balancing out Article 16, they have proved to be quite a mixed blessing, as the Darfur case shows.
Due to these shortcomings, Article 16 has not been able to contribute to an effective and sustainable collision management. On the contrary, the semantic struggles in the cases so far have undermined both the ICC's and the Security Council's legitimacy. Although the application and non-application cases are driven by quite different interests, the types of arguments used and the conflict lines are actually quite similar.
This raises the question of how to improve the collision management. Since a reform of Article 16 is unlikely in the near future, particular attention should be devoted to the application control. As mentioned above (at Section 3.2.), neutral institutions such as courts could provide a stable framework for semantic struggles over norm application. Therefore, the following sections analyse whether responsiveness and consistency can be established by the ICC itself, or whether Article 16 requires external stabilization by other actors.
5. Patrolling the border between peace and justice: application control of Article 16?
When asked whether he was becoming a politician at the ICC, Prosecutor Luis Moreno Ocampo answered:
On the contrary. I am putting a legal limit to the politicians. That's my job. I police the borderline and say, if you cross this you’re no longer on the political side, you are on the criminal side. I am the border control.Footnote 136
Adopting this metaphor, is the Security Council alone in policing the border between its regime and the ICC’s? We argue that, based on the normative requirements for a deferral request (Section 5.1.), the ICC has the power to determine whether a deferral request is in line with Article 16, a power which it has yet to use, and which is not entirely satisfactory (Section 5.2.). Other ‘border patrols’ are also possible: the ICJ's power to supervise the Security Council has been widely discussed in other contexts,Footnote 137 and the General Assembly can both parallel Security Council action and step in where it fails to act (Section 5.3).
5.1. The normative framework for deferral requests
Article 16 contains some explicit criteria for deferral requests. As a limited exceptional power, a request has to comply with both the RS and the UN Charter.Footnote 138 Just like an Article 13 referral, a deferral request requires a Chapter VII resolution, which in turn presupposes a threat to international peace and security in accordance with Article 39 of the Charter.Footnote 139 In this respect, it is consented that the Security Council enjoys a wide margin of discretion in order to be able to properly execute its political function.Footnote 140 This, however, does not place it above the law:
It is clear from this text [the text of Article 39 of the UN Charter] that the Security Council plays a pivotal role and exercises a very wide discretion under this Article. But this does not mean that its powers are unlimited. The Security Council is an organ of an international organization, established by a treaty which serves as a constitutional framework for that organization. The Security Council is thus subjected to certain constitutional limitations, however broad its powers under the constitution may be. Those powers cannot, in any case, go beyond the limits of the jurisdiction of the Organization at large, not to mention other specific limitations or those which may derive from the internal division of power within the Organization. In any case, neither the text nor the spirit of the Charter conceives of the Security Council as legibus solutus (unbound by law).Footnote 141
Just as any other UN organ, the Security Council is bound by the UN Charter as well as by jus cogens and general principles of international law:Footnote 142 Article 24 of the Charter obliges all UN organs to act in accordance with the goals and principles of the United Nations. Council decisions are binding on UN member states under Article 25 of the Charter only if they are passed ‘in accordance with the Charter’.Footnote 143
With regard to the RS, the requirement of a Chapter VII resolution in Article 16 suggests that the Security Council has to justify the request as a measure serving the maintenance or restoration of international peace and security.Footnote 144 Its systematic position in relation to Articles 13–15 of the RS shows that Article 16 is meant to cover scenarios where an investigation or prosecution or preparatory measures are already underway.Footnote 145 This also means that there must be a reference to an actual ‘situation’Footnote 146 that would be or is being investigated, as this is what measures under Articles 13–15 would refer to. Therefore, a request must address a specific situation, as it is or would be subject to investigation or prosecution, and can neither be made in an abstract mannerFootnote 147 nor in relation to certain individuals or groups of individuals only.Footnote 148 The reasons given must therefore specifically address the effects of a commencement or proceeding of particular (future) investigation or prosecution measures into a concrete situation.
5.2. Responsiveness through application control by the ICC?
Given that the Security Council has license to impose its regime logic unilaterally, involvement of the ICC in the application of Article 16 might be able to introduce an element of responsiveness, or at least to counteract the Security Council's apparent arbitrariness by subjecting its requests to legal verification.
It is true that Article 16 does not accord discretion to the ICC: in the case of a request passed under Chapter VII, ‘no investigation or prosecution may be commenced or proceeded with’. However, some argue that the ICC's organsFootnote 149 can verify whether the criteria contained in Article 16 (as described in the previous section) have been met.Footnote 150 While Security Council decisions are binding upon UN members, deferrals are requests, since the Charter claims no binding power in relation to international organizations such as the ICC.Footnote 151 Such binding force can only arise from the ICC's own Statute. If the RS establishes criteria for binding requests, the ICC must be able to verify whether they have been fulfilled.Footnote 152 This is obvious for the requirements that the request be limited to a maximum of 12 months, and its reference to a ‘situation’.
Moreover, Article 16 requires that the Security Council act under Chapter VII. Of course, the Security Council enjoys a wide margin of discretion, and it is not for the Court to pass judgment on how it exercises its discretion.Footnote 153 This discretion, however, can only exist within the limits of the Security Council's powers under the Charter. The ICC must therefore be able to determine whether these legal limits have been overstepped.Footnote 154 This means that the ICC can not only examine whether the Security Council has established a threat to international peace and security, but also whether it was entitled to do so, i.e., whether the conditions for such a determination were met – within the Security Council's margin of discretion.Footnote 155
Article 16, therefore, does not leave the ICC entirely at the disposal of the (permanent) Security Council members’ interests. The ICC is able to reject a request for deferral which is ultra vires or otherwise not in accordance with Article 16, however, the potential of this power to enhance the resolution of regime collisions is limited.
First, the ICC may be loath to exercise this power – it actually failed to do so in the peacekeeper cases although arguably the conditions of Article 16 were not met. Of course, no ongoing or planned investigations were actually affected. The question remains whether the ICC would be willing to take it up with the Security Council at all. In practice, it is entirely dependent on the co-operation of its state parties; they alone have the power to arrest and surrender a suspect. The Security Council is an important ally in this; its resolutions have the power to oblige non-member states to co-operate with the ICC. While the OTP has become more confrontational as its conciliatory strategy failed to yield results,Footnote 156 it is doubtful whether this will also play out in relation to the Security Council and its members.
Second, this power is unhelpful where the Security Council declines to request a deferral. Should its investigations or prosecution really threaten international peace and security, the ICC would be unable to resolve the collision itself. The OTP is not supposed to discontinue relevant investigations unless ‘[a] prosecution is not in the interests of justice’ (Art. 53(3)(c) of the RS);Footnote 157 the interests of international peace and security are not elements of its discretion. Once the case has reached the trial stage and the charges have been confirmed, there is no stopping it for reasons external to the trial.
Third, the ICC would also be unable to resolve a strategic blockade where the Security Council legitimately exercises its Article 16 power but does not address the threat itself, so that it is subjected neither to a political nor to a legal solution.
5.3. Application control by third parties?
Given the deficits of Article 16 and the limited use of the ICC's reviewing power, neutral third parties may be able to stabilize the norm and heighten its legitimacy, as well as make up for the Security Council's ‘tunnel vision’.Footnote 158 Within the UN system, the General Assembly and the ICJ have become well-known candidates for second opinions, albeit very different ones: while the ICJ is a highly juridified forum for dispute resolution, the General Assembly is no less a political organ than the Security Council, but without veto powers.
An intervention by the ICJ in disputes over Article 16 would likely contribute to the norm's legitimacy, given that the Court is largely considered independent, neutral and legitimate. Moreover, the Security Council has no power to keep it from pursuing a case. Starting with the Nicaragua case,Footnote 159 the ICJ has repeatedly been active parallel to the Security Council in matters relating to international peace and security, arguing that there was no separation of powers between it and the Security Council, whose main responsibility was not exclusive but complementary.Footnote 160 Its power to examine the legality of Security Council actions has been widely discussed.Footnote 161 The ICJ is competent to rule on ‘any question of international law’ in disputes between states under Article 36(2) of the ICJ Statute, not excluding ones that the Council has already dealt with (possibly differently). Moreover, the General Assembly and other UN organs can request an Opinion ‘on any legal question’ under Article 65 of the ICJ Statute and Article 96 of the Charter, e.g., whether there is an act of aggression. In principle, the ICJ could also examine whether the requirements of an Article 16 request are fulfilled. As indicated above, despite its discretion under Chapter VII, the Security Council is not above the law.
In practice, though, the ICJ has consistently avoided any material disagreement with the Security Council,Footnote 162 emphasizing the priority role of the Council in armed conflict.Footnote 163 Moreover, it is not easy to seize; UN member states can only bring a case if both sides have accepted its jurisdiction (Art. 36(2) of the ICJ Statute). Among the 66 states that have issued a general declaration to this effect – many with reservations relating to armed conflictFootnote 164 – the United Kingdom is the only veto power, after the US revoked theirs in the aftermath of Nicaragua.Footnote 165 A General Assembly request for an Opinion under Article 96 of the Charter requires a majority vote, i.e., substantial backing from the international community. Finally, it would take the ICJ a long time to come to a decision, whereas Article 16 situations typically have an emergency character.
The General Assembly, a political institution like the Security Council, has repeatedly been proposed as a ‘border patrol’ itself because it is considered more ‘democratic’ than the exclusive forum of the Security Council. While the Assembly's resolutions are non-binding, it can be a forum for ‘counter-politicization’ sans veto power.Footnote 166 It can ‘discuss all matters relating to the maintenance of international peace and security’ and issue recommendations, or refer matters to the Security Council for further measures (Art. 11 of the Charter). While Article 12(1) of the Charter suspends the Assembly's recommendation powers while the Security Council deals with a matter, the Assembly has asserted them in situations where the Council fails to address a threat due to a disagreement between its permanent members.Footnote 167 It could, therefore, step in in cases of blockade following a legitimate Article 16 request without further Security Council action. Due to the absence of veto powers, the General Assembly may come to disagree with the Security Council on matters of international law or declare its actions not in accordance with the law. It remains that this, too, requires a majority vote within the Assembly in accordance with Article 18(3) of the Charter.
While putting third parties such as the ICJ or the General Assembly in charge of overseeing the application of the collision rule may contribute to externally stabilizing the collision management, at the same time the potential for collisions increases. It would be difficult – and possibly not desirable – to come up with explicit collision rules for each and every potential collision. Instead, it would be desirable and necessary – in all cases where explicit collision norms are lacking or deficient – that the institutions involved apply the principle of default deference, i.e., respecting the decision of another institution or organ unless there are weighty reasons to differ. This principle is already being applied between courts in the area of fundamental rights in Europe.Footnote 168 Between the different domestic and European courts there is no consented hierarchy – horizontal co-ordination has been working rather well so far.Footnote 169 Such a ‘duty to consider’ would be difficult to codify in the UN system, even more so because there are political bodies involved, however, it can be developed by way of continuous practice and jurisprudence. Even between regimes with different rationales, such network structures are beginning to develop,Footnote 170 however, the practice of the ICJ in relation to the Security Council shows that much is left to be desired. Improving Article 16 of the RS could, therefore, pave the way to a more general culture of deference that goes both ways.
6. Article 16's potential to be developed into a ‘good’ collision rule
The examples discussed demonstrate that Article 16 does not manage to secure both the maintenance of international peace and security by the Security Council and the criminal prosecution of the gravest international crimes by the ICC, while resolving goal conflicts in each individual case in a consistent and balanced way. The semantic struggles over norm application threaten to affect the legitimacy and effectiveness of both the Court and the Security Council. It is to be expected that the proposed crime of aggression will only exacerbate the problem, given that this is part of the core responsibility of the Security Council. While Article 16 has not been applied since Resolutions 1422 (2002) and 1487 (2003), possibly because of its problematic nature, calls for its application, and frustration over the Security Council's disregard or inaction, continue.
The first application cases show that the arrangement in Article 16, unilaterally putting the Security Council in charge of suspending ICC activities if they interfere with the Council's tasks, does not play out satisfactorily. First, the Security Council's broad discretion in matters of peace and security undermines norm determinacy. Second, the Security Council is a body that, due to the veto powers of the ‘P-5’, is considered an exclusive forum of power players who can flout the law if they wish, and not an institution with good pedigree in the sense of Franck's theory. In terms of the application of Article 16 of the RS, it is neither guaranteed that like cases are treated alike, nor is the Security Council required to base its decisions on generally accepted principles. Third, as one of the parties involved, the Security Council promotes only its own regime logic without having to take that of the ICC into account. Moreover, the Security Council may strategically fail to request a deferral despite threats to international peace and security, e.g., if a permanent member vetoes such a request. The prosecutor, however, is not able to react, since she is required to act in the interests of justice alone.
Currently, the possibilities of application control as a possible solution to collision (mis-)management are also not entirely satisfying. The ICC's power to review deferral requests does not cover all problematic situations, and it may not be exercised in practice. These problems can partially be resolved by third-party involvement: the ICJ and the General Assembly are able, to some extent, to step in where the ICC's powers in relation to the Security Council are concerned. However, their ‘external stabilization function’ comes at the price of further collision potential.
Summarizing the lessons learned from the previous remarks, it appears that a ‘good’ collision rule, i.e., one that successfully prevents or resolves threats to the legitimacy and effectiveness of the regimes concerned, would not put one of the regimes in charge of ‘patrolling the border’ unilaterally without demanding that the other regime's logic be taken into account. It is essential that responsiveness is secured, either by installing a neutral institution which is responsible for the application of the collision rule – an institution that does not unilaterally favour one logic – or at least by conferring on the other regime's institution (here, the ICC) an explicit and comprehensive reviewing power.
Coherence is another important element warranted by a collision rule; a code of conduct with more specific criteria can contribute to this. In struggles over the power of interpretation, such codes or secondary rules force actors to articulate their interests in legal terms, limiting the opportunity for power politics manipulation, as in the case of Resolutions 1422 and 1487. In the context of the application of Article 16 of the RS, this could be a code of conduct stipulating the requirement for the Security Council to justify its decision under Chapter VII of the Charter or to specify what exactly constitutes a threat to peace in a particular situation. It would also be helpful to require that requests under Article 16 of the RS be justified by reference to generally recognized principles to guarantee that similar cases are treated equally.
However, at the moment, the main potential to improve collision management via Article 16 of the RS rests with those institutions provided with the authority to control its application. So far, the ICJ and the General Assembly have not exhausted their – albeit limited – competences in this respect. Both institutions should use their respective powers courageously in order to stabilize and legitimate the application of the norm and should not leave the stage to the Security Council alone.
