Hostname: page-component-797576ffbb-tx785 Total loading time: 0 Render date: 2023-12-04T23:09:09.572Z Has data issue: false Feature Flags: { "corePageComponentGetUserInfoFromSharedSession": true, "coreDisableEcommerce": false, "useRatesEcommerce": true } hasContentIssue false

African resistance to the International Criminal Court: Halting the advance of the anti-impunity norm

Published online by Cambridge University Press:  23 August 2017

Kurt Mills*
Professor, International Relations and Human Rights, University of Dundee
Alan Bloomfield*
Lecturer, University of Western Australia
* Correspondence to: Kurt Mills, School of Social Sciences, Scrymgeour Building, University of Dundee, Dundee DD1 4HN, UK. Author’s email:
** Correspondence to: Alan Bloomfield, Room 2.41, School of Social Sciences, The University of Western Australia, 35 Stirling Highway, Crawley, Western Australia 6009, Australia. Author’s email:
Rights & Permissions [Opens in a new window]


The creation of the International Criminal Court (ICC) in 1998 marked a substantial advance in the effort to ensure all perpetrators of mass atrocities can be brought to justice. Yet significant resistance to the anti-impunity norm, and the ICC as the implementing institution, has arisen in Africa. The ICC has primarily operated in Africa, and since it sought to indict the sitting Sudanese President Omar al-Bashir in 2008 resistance from both individual African states and the African Union has increased substantially. We draw on the concept of ‘norm antipreneurs’, and the broader norm dynamics literature, to analyse how resistance has developed and manifested itself, as well as the potential effects of this resistance on the anti-impunity norm. We conclude that the antipreneur concept helped us structure and organise analysis of the case – suggesting it could be usefully deployed in other similar cases – but that this case also suggests that antipreneurs do not always enjoy substantial defensive advantages. We also conclude that African resistance to the ICC has substantially stalled the advance of the anti-impunity norm, a finding that has significant implications for the wider effort to reduce mass atrocity crimes in the contemporary era.

© British International Studies Association 2017 


After having discontinued its case against Kenya’s President Uhuru Kenyatta in December 2014, in April 2016 the International Criminal Court (ICC) also terminated its case against Vice-President William Ruto, effectively admitting defeat in its attempt to hold the most senior Kenyan officials accountable for election violence in 2007/08. Once a keen supporter of the ICC, Kenya was now a foe and had ‘giv[en] the world a rule book on how to beat the ICC’.Footnote 1 But this was just the latest setback the ICC has suffered in Africa, with the ongoing controversy over the 2009 decision to issue an arrest warrant for Sudan’s President Omar al-Bashir also framing an evolving conflict between Africa and the ICC and eventually leading to three African states announcing in late 2016 that they would leave the ICC.

The ICC’s mission is fundamentally defined or informed by the anti-impunity norm. And the ICC is dedicated to enforcing a ‘strong’ version of this norm – meaning all perpetrators of mass atrocity crimes, including states’ highest officials, can be held accountable before competent courts. This article’s primary purpose is to explore the dynamics of African resistance to the effort to entrench anti-impunity, and what effect this has had on the norm (and, by implication, the ICC). Because we focus on African resistance we cannot presume to offer a definitive assessment of the anti-impunity norm’s progress towards becoming the normative status quo. Instead, our findings are limited to just one piece of this wider empirical puzzle, although this is a particularly important piece given most of the active situations under investigation in the ICC’s brief concern Africa. To clarify the scope of our argument, we readily acknowledge that international actors have resisted the ICC in various ways and for various reasons. Most obviously, as of 1 July 2017, 124 states had joined,Footnote 2 meaning one third of states are not members. Declining to join thus remains a powerful form of resistance. And many motivations animate these decisions: for example, the relatively rushed negotiation and drafting process of the Rome Statute contributed to some state’s decisions to not join.Footnote 3 Washington’s position has also been justified by David Scheffer with reference to the fact it ‘has special responsibilities and special exposure to political controversy’.Footnote 4 Further, Jack Goldsmith has argued that Washington, Beijing, and Moscow were all concerned that the ICC might intrude upon the Security Council’s responsibility for international peace and security.Footnote 5 More generally, many states were, and remain, concerned that the ICC threatens their sovereignty; for example, India has expressed this concern repeatedly.Footnote 6 Nevertheless, we focus on resistance to the ICC and the version of the anti-impunity norm that envisages indicting sitting Heads of State and Government from African member-states because this sort of ‘insider’ resistance is both very prominent recently and it is potentially very damaging to the wider effort to establish the ICC as an effective institution and to entrench the anti-impunity norm.

We structure our analysis by deploying an agent-centric framework that distinguishes overtly between norm entrepreneurs and norm antipreneurs. The ‘antipreneur’ term was introduced recently by Alan BloomfieldFootnote 7 to highlight that those who resist entrepreneurs’ efforts to challenge the existing normative status quo often enjoy under-appreciated defensive advantages. Accordingly, a secondary purpose of this article is to test this framework’s utility. The article proceeds as follows. We first explain how the entrepreneur-antipreneur framework is both deployed and tested. The next section explains why we treat the anti-impunity norm as the ‘challenger norm’ before the key sites of contestation and the key African actors are examined. The longest sections of the article examine how, when, and why African actors resisted the anti-impunity norm, focusing on how the African Union (AU) itself, and several key states, shifted from entrepreneurial towards antipreneurial roles over time. Finally, in the conclusion we explain why the phenomenon of African resistance to the ICC poses a serious obstacle to the effort to advance and entrench the anti-impunity norm and is responsible for stalling the advancement of the norm, discuss why we find the antipreneur concept has some, but limited, utility in this case, and consider a few potential directions for future research.

The entrepreneur-antipreneur framework

Bloomfield offered the notion of the ‘norm antipreneur’ as a corrective to a norm dynamics research agenda ‘beset with selection biases’ towards the ‘more noticeable’ cases in which entrepreneurs succeeded in changing the normative status in an issue-area, meaning the phenomenon of resistance had been undertheorised.Footnote 8 He argued further that antipreneurs often enjoyed under-appreciated ‘defensive advantages’Footnote 9 given the socio-psychological biases by which most human collectivities, most of the time, prefer the status quo,Footnote 10 and that antipreneurs also often benefit from opportunities to delay or frustrate, or even entirely block, entrepreneurs’ discrete initiatives.Footnote 11 Ultimately, Bloomfield’s argument rests on the insight that entrepreneurs must take the initiative – they must first generate, then sustain, momentum for change – while antipreneurs will often have to expend less political capital and/or take fewer risks to blunt or derail entrepreneurs’ efforts (which is not to suggest antipreneurs will always win; these advantages are arguably power and/or skill ‘multipliers’).Footnote 12

This study deploys the strategic/tactical resistance dichotomy Bloomfield developed to theorise the conceptual relationship between various types of resistance practised by African states to anti-impunity. Specifically, we treat justifications for resisting as ‘strategic’ resistance, and discrete moves by actors to resist as ‘tactical’ resistance.Footnote 13 This distinction is not necessarily a sharp one – both categories of resistance are, in effect, mutually constitutive – but it is nevertheless useful analytically. But, and crucially, Bloomfield conceded that antipreneurs may only enjoy inherent defensive advantages when the normative status quo is deeply entrenched in practice, and/or when it is deeply institutionalised.Footnote 14 This study tests this assumption and asks whether this is one such case.

Bloomfield also offered an agent-centric framework for analysing the dynamics prevailing in particular norm contestation contexts. Specifically, overtly distinguishing entrepreneurs from antipreneurs enables the creation of a spectrum of ‘roles’ that agents might play (and these are treated as labels or categories that describe motives and behaviour, not normative judgements; antipreneurs are not always ‘bad’ and entrepreneurs always ‘good’).Footnote 15 As Figure 1 illustrates, ‘pure entrepreneurs’ seek substantial normative change while ‘pure antipreneurs’ implacably defend the normative status quo. But other actors may have somewhat different intentions vis-à-vis the normative status quo: ‘competitor entrepreneurs’, for example, might seek different or less-radical change, while ‘creative resisters’ may concede to some degree of normative change while still primarily defending the status quo.Footnote 16

Figure 1 The norm dynamics role-spectrum.

Most actors probably fall into the intermediate zone, and they might shift along the role-spectrum over time. This latter insight is particularly important given that the meta-narrative of this study is that several African actors have recently moved towards the antipreneurial end – although in nuanced and complex ways.

The insight that actors’ roles can change can in turn be utilised to determine which of three norm dynamics models offered by Amitav Acharya might best explain the ‘prevailing dynamics’ in this case at particular points in time. Acharya’s ‘norm localisation’ model described how when local actors respond to external pressure from ‘the centre’ – that is, Western states and the IGOs those states dominate – they sometimes essentially adapted new norms to suit their local circumstances.Footnote 17 But he later offered the ‘norm subsidiarity’ model to explain how local actors resisted new norms by invoking ‘existing common global norms … vital to preserving their autonomy … like sovereignty [and] … self-determination’.Footnote 18 Eventually Acharya combined these into a ‘norm circulation’ model that described local actors at first resisting, but then feeding-back a ‘reworked’ version of the new norm to the centre, demanding refinements to the norm’s scope and content before they accept it.Footnote 19 Thus identifying whether key actors are shifting into different roles – adapting, resisting, or seeking to renegotiate – offers clues regarding likely outcomes of norm contestation cases, which in turn hints at the strategies norm entrepreneurs might pursue to successfully entrench the norm they are promoting.

The normative contest, and key sites and actors

We follow convention and define a norm as a ‘standard of appropriate behavior for actors of a given identity’.Footnote 20 More specifically, we treat the anti-impunity norm (detailed below) as the challenger norm even though it has already been formally codified in the Rome StatuteFootnote 21 and a formal institution, the ICC, has been established to implement it. We do so for two interconnected reasons. First, the absence of world government means ‘the authority of international law resides … in States recognis[ing] it as binding upon them’.Footnote 22 Constructivists explain these ‘internal requirements’ for compliance as ‘felt’ effects: when a law’s or norm’s legitimacy is accepted by actors they ‘feel’ it is ‘right’ to comply.Footnote 23 In other words, to be effective, international norms/laws must constitute actors, becoming part of their identities, and thereby informing their interests and affecting their behaviour.Footnote 24 Second, and more specifically, the ICC relies on states’ active assistance to perform many of its functions. Article 86 of the Rome Statute provides a general duty for parties to ‘cooperate fully’ and Articles 87 to 111 provide specific details. But we show below that many African actors’ behaviour suggests that they are not ‘fully constituted by’ many of the demands inherent in the anti-impunity norm, as prescribed in the Rome Statute.

Thus, the fact a norm has been codified in a treaty does not by itself mean that it represents the normative status quo. State-practice should also generally conform with it too, which is not to say that there can be no violations at all; indeed, ‘no single [violation] refutes a norm. Not even many such occurrences necessarily do.’Footnote 25 But the word ‘norm’ itself describes patterns of behaviour that are, in effect, ‘normal’,Footnote 26 implying there should be more compliance than not. Reactions to violations matter too: strong criticism, and especially punishment for violation, also suggests a norm is relatively strong/entrenched.Footnote 27

Our analysis is therefore broadly situated in the constructivist paradigm, meaning we do not treat states’ interests as givens. The relationship between norms and interests is a complex one, but we accept that norms constitute actor’s identities, and that interests flow from identities; in effect, interests ultimately derive from norms.Footnote 28 More specifically, we apply Thomas Risse’s notion of the ‘logic of argument’ and treat actors as not necessarily strongly-constituted by a particular norm – which could imply they follow it ‘mindlessly’Footnote 29 – but instead actors argue about ‘which norms apply under given circumstances’.Footnote 30 This implies states have a sort of ‘menu’ of competing norms from which to choose when formulating responses in discrete circumstances. Accordingly, we must examine the status quo and other norms that African actors invoke to justify their resistance, before we examine the precise normative features (and evolution of) the anti-impunity challenger norm.

The status-quo norm: Sovereign immunity

The most fundamental status quo norm is state sovereignty. It is the constitutive norm of the international system and it is deeply entrenched: the principle of sovereign equality (Article 2(1)) and the rights of non-interference (Article 2(7)) and self-determination (Article 1(2)) appear in the UN Charter.

A closely-related norm – which existed in customary international law well before the UN Charter – is sovereign immunity. Strictly speaking, this establishes that a state cannot be subject to external courts’ jurisdiction, but two other closely related and ‘personal’ immunities flow from it: diplomatic immunity and Head of State immunity. Both are primarily based in functional considerations, namely, the need for a state’s representatives to travel freely to conduct diplomacy, although Head of State immunity is also informed by wider symbolic concerns like ‘respect for’ a state’s sovereign independence.Footnote 31 These latter two norms are nested within the sovereign immunity norm, which, in turn, flows directly from sovereignty itself. Sovereign immunity is deeply-entrenched: it was recently declared jus cogens by the International Court of Justice (ICJ);Footnote 32 diplomatic immunity is also codified in Article 29 of the Vienna Convention; and while Head of State immunity is not codified, in 2000 and 2001 both the ICJFootnote 33 and the US Supreme CourtFootnote 34 confirmed it applied to states’ leaders and other high-ranking officials.

African states have practised strategic resistance by invoking these norms as justifications for resisting anti-impunity, but they have also appealed to other related norms in the wider ‘web of meanings’,Footnote 35 which constitutes the discursive terrain in which they operate. Specifically, they invoked anti-imperialism and Afrocentrism (or ‘African solidarity’ or ‘Pan-Africanism’), the idea of ‘African solutions for African problems’ and the ‘African Renaissance’ agenda, as well as a norm which privileges peace over justice.

The challenger norm: Anti-impunity

The first tentative steps to overturn sovereign immunity were taken in the immediate aftermath of the Second World War at Nuremberg and Tokyo.Footnote 36 In the 1990s the International Criminal Tribunals for the Former Yugoslavia and Rwanda in 1993 and 1994, and other ad hoc, hybrid international courts operated in Sierra Leone, Cambodia, and Lebanon. Then, in 1998, former Chilean President General Augusto Pinochet was arrested in the UK under the principle of universal jurisdiction. The British courts stripped Pinochet of immunity although his extradition to Spain was eventually waived for health reasons.Footnote 37 The principle of universal jurisdiction remains hotly contested. Nevertheless, the Pinochet case signaled that a general norm against impunity was emerging, at least vis-à-vis former leaders.

But the establishment of the ICC, also in 1998, marked the single-most significant development in the effort to entrench anti-impunity; David Bosco called it a ‘remarkable transfer of authority from sovereign states to an international institution’.Footnote 38 A majority of states have joined,Footnote 39 although key great powers like the United States, China, Russia, and India remain aloof, which contributes to our decision to treat anti-impunity as an ‘emerging’, challenger norm. Nevertheless, its codification in the Rome Statute means, at minimum, that it has gone beyond being ‘merely aspirational’.

Importantly, the ICC does not exercise universal jurisdiction per se. Instead, member states must prosecute citizens and others directly related to member states who commit mass atrocities (including senior officials) and the ICC’s jurisdiction only becomes activated if they fail to do so (that is, the principle of complementarity).Footnote 40 The United Nations Security Council (UNSC) can also refer a situation in a non-party to the ICC. Finally, as noted earlier, the Rome Statute requires parties – including those not directly involved in an investigation or prosecution – to assist the ICC.

As was the case vis-à-vis sovereign immunity norms, the anti-impunity norm is also related to other norms, most notably norms underpinning international humanitarian law and norms like the protection of civilians and the responsibility to protect (R2P). Promotion of these – and anti-impunity – should be seen as part of the wider effort underway since the 1990s to reconceptualise sovereignty and establish that how states treats their citizens is a matter of international concern.Footnote 41

Finally, the ICC has been trying to implement an anti-impunity norm that includes sitting Heads of State and Government (and other senior officials). It is obvious why the Rome Statute was framed this way: excluding them would be nonsensical given they are often responsible for ordering atrocity crimes. Further, to allow them to avoid prosecution while in office would allow them to continue to commit atrocity crimes, provide incentives for them to stay in office beyond their terms to avoid prosecution, and allow them to use a ‘circumscribed’ anti-impunity norm against their enemies. This issue has become a focal point for resistance in Africa; we provide several examples below of African actors claiming to ‘support anti-impunity’, but only in a circumscribed (and therefore inherently problematic) form.

Sites of contestation, and key African actors

We have just examined the competing norms in this issue-area that states can pick and choose from when determining how to respond, for example, when the ICC issues an arrest warrant. But where they do so, and who the key actors are, requires explanation before the analysis of African resistance proper can begin.

The ICC is frequently conflated with the prosecutor, but this office is less a site of contestation than an actor, a ‘norm implementer’ that determines when to apply the anti-impunity norm (and probably also a norm entrepreneur, expanding the practice and understanding of anti-impunity). Contestation takes place in the Registry and pre-trial, trial, and appeals divisions: the ICC is a court after all. But the primary site of contestation that interests us – because states contest there – is the ICC’s Assembly of State Parties (ASP). All members have a seat and therefore a vote, and it meets annually. Importantly, a seven-eighths vote is required to amend the Rome Statute, a significant impediment.

The UNSC is another important site of contestation. Its resolutions may contain provisions that reflect various related normative urges (that is, R2P, humanitarian access, anti-impunity, etc.) making it difficult at times to disentangle these norms.Footnote 42 But the Rome Statute empowers the Council to both refer a situation in a non-party to the Prosecutor (Article 15) and to also defer any ICC investigation or prosecution for a renewable term of 12 months (Article 16). The latter power has attracted fierce contestation from African actors. The Council must be ‘acting under Chapter VII’ of the UN Charter, implying it must be attempting to remedy a threat to international peace and security.

Contestation also takes place within the AU. Thirty-four of its fifty-four members are parties to the Rome Statute, almost exactly the same proportion of members worldwide (63 per cent). But we primarily treat the AU as an actor because it has become the primary ‘vehicle’ for African states to organise resistance to anti-impunity; efforts to forge African solidarity take place and can be vested with legitimacy there. We focus mainly on the formal statements and actions of the AU’s supreme governing body, the Assembly (comprised of members’ Heads of State and Government) and its Peace and Security Council (PSC). The AU is not, however, a member of the ICC’s ASP, nor does it have formal standing in the UNSC: to influence deliberations in these forums the AU must induce a member-state to raise a matter and then lobby other African states to support such initiatives.

While we treat the AU mainly as an actor, we do not ignore intra-AU dynamics. Practical limitations prevent us from exploring these exhaustively, but we do sketch how these dynamics have changed over time by examining how several key African states – ICC members Burundi, Gambia, Kenya, Namibia, South Africa, and Uganda – have shifted from the entrepreneur towards the antipreneur end of the role-spectrum recently. We also provide a few examples of contestation over the anti-impunity norm between domestic institutions; we cannot examine all 34 African ICC members intensively, but we nevertheless want to acknowledge that African states are not necessarily unitary actors.

The following sections examine how African resistance to anti-impunity has developed and increased over time. The first considers the period before Sudan’s President Omar al-Bashir was indicted by the ICC, while the next and much longer section examines resistance thereafter. This organisation reflects how the request for an arrest warrant for al-Bashir became a critical turning point after which latent uneasiness towards the ICC blazed into open hostility and sustained resistance.

‘Before al-Bashir’: Adapting to anti-impunity, 2002–8

About a dozen African states were influential players in the drafting process leading to the Rome Statute and they ‘generally advanced progressive positions’.Footnote 43 After it came into force in 2002, Botswana became the most prominent ICC supporter, with Zambia also providing significant support. South Africa and Kenya were also prominent entrepreneurs. Some African states always opposed the ICC and the anti-impunity norm. These typically authoritarian countries never joined and therefore feel threatened by anti-impunity and other human rights norms making them, in effect, pure antipreneurs. Egypt and Eritrea are prominent examples, but Libya under Muammar Gaddafi – who eventually became the target of an ICC investigation via a 2011 UNSC referral – arguably took the lead. They tried to frame the issue within the AU by invoking status quo norms like sovereignty and sovereign immunity, as well as anti-imperialism and local African solidarity norms. We pay these states little attention; we are primarily interested in states that have shifted from the entrepreneurial towards the antipreneurial end of the role-spectrum presented earlier.

In the 2002–8 period the African ICC members, and the AU, generally cooperated with the newly-operational ICC. The ICC’s first two cases were referred to the Prosecutor by Uganda in 2003 and the Democratic Republic of Congo in 2004, and they attracted little adverse comment in Africa. In 2003 Côte d’Ivoire also formally granted the ICC limited jurisdiction. Notably, in 2004, an AU resolution urged its members to sign and/or ratify the Rome Statute,Footnote 44 and in 2005 the UNSC’s referral of the situation in Darfur to the ICC also attracted little controversy.

Then two events took place that were more suggestive of, in terms of Acharya’s multiple frameworks, ‘adaptation to’ (that is, suggestive of his norm localisation model)Footnote 45 rather than simple ‘acceptance of’ anti-impunity. First, in 2003 Interpol issued an arrest warrant for the former president of Liberia Charles Taylor. Nigeria refused to action it without a request from Liberia, but when such came in 2006 Taylor was extradited to the ad hoc Special Court for Sierra Leone. Second, Senegalese courts had refused to try the former President of Chad, Hissène Habré, in the early 2000s, citing jurisdictional limitations, and Senegal’s government also refused to extradite him to Belgium pursuant to claims of universal jurisdiction. But in 2006 the AU set up a ‘Committee of Eminent African Jurists’ to determine whether, and if so, how and where, he should be tried. It determined he should be tried, although it took considerable pressure from the AU, the Economic Community of the West African States, and several African states, especially Chad, to bring Habré to trial (which eventually commenced in 2015, in the AU-created ‘Extraordinary African Chamber’).

These examples suggest that the anti-impunity norm was shaping the behaviour of many African actors in the 2002–8 period. Some operated like pure entrepreneurs by essentially cooperating fully with the ICC while others operated more like competitor entrepreneurs, pursing normative change – by advancing anti-impunity – but somewhat differently. Having said this, there were rumblings of discontent. These did not manifest themselves as direct resistance to how the ICC was applying the anti-impunity norm, but instead as resistance to the related principle of universal jurisdiction.

To clarify, universal jurisdiction is an alternative method of applying anti-impunity: it envisages foreign domestic courts trying ‘alien’ citizens for crimes deemed so terrible they warrant being treated as ‘international’ crimes, while the Rome Statute provides for national courts to try citizens, and failing this, for an international court to do so. Rwanda emerged as the strongest resister of universal jurisdiction. It has long had a contentious relationship with international criminal justice mechanisms: after calling for the International Criminal Tribunal for Rwanda, Kigali was subsequently very critical of its operationsFootnote 46 and Rwanda resisted Western states’ attempts to prosecute members of the Tutsi-dominated government for crimes committed during and after the Hutu-led genocide (while implementing anti-impunity domestically against Hutu genocidaires).Footnote 47

On 1 July 2008 the AU Assembly passed a resolution which, while expressing general support for the principle of universal jurisdiction, also alleged ‘abuse’ (especially against members of the Rwandan government) and said attempts to invoke it against African leaders violated ‘the sovereignty and territorial integrity of these states’.Footnote 48 The AU’s PSC issued a similar statement on 11 July.Footnote 49 The ICC’s investigation of Omar al-Bashir had already begun, so there seemed to be a blurring of the general opposition to universal jurisdiction and the ICC investigation, suggestive of ‘simmering’ disenchantment with anti-impunity. But the ICC’s attempt to arrest al-Bashir, a sitting leader, galvanised what had been sporadic and relatively unfocused resistance. The next section examines how and why this occurred.

‘After al-Bashir’: African resistance to the anti-impunity norm, 2008–16

Just days after the AU Assembly and PSC declarations on universal jurisdiction, the Prosecutor, Luis Moreno-Ocampo, requested that an arrest warrant be issued against Omar al-Bashir for war crimes and crimes against humanity.Footnote 50 In response, the PSC called on the UNSC to defer the matter and to justify this first example of tactical resistance it offered two forms of strategic resistance. First, and echoing Rwanda, it argued that universal jurisdiction/anti-impunity was being selectively applied against African states and leaders, and thereby abused (and it invoked a litany of grievances to establish its case) – this represented some of the first assertions of what was to become a widespread perception that the ICC was ‘Western tool’. Second, it argued that an ICC prosecution could undermine the peace process in Darfur.Footnote 51 This episode therefore provides early evidence for the AU beginning to shift towards resisting the ICC and anti-impunity. To be clear, it had not definitively crossed onto the antipreneurial side of the spectrum yet; it was probably still a competitor entrepreneur (that is, wanting to advance anti-impunity, but not exactly in the same manner as the ICC), because the PSC also recognised the seriousness of the situation in Darfur and called for an investigation into how the AU might address it.

The AU Assembly later also called for deferral, which was ignored,Footnote 52 and the PSC reiterated its position after the ICC issued the arrest warrant against al-Bashir.Footnote 53 This was despite the fact that many African leaders found al-Bashir troublesome. Soon after, three African ICC members – Senegal (the first country to sign the Rome Statute), Djibouti and Comoros – and Libya, a non-member – called on African states to withdraw from the ICC. This was rejected by a meeting of African ICC members, although most supported the deferral request.Footnote 54 This was the first instance of a second tactical move – threat of withdrawal – and signalled that several African ICC members were shifting from entrepreneurial towards more antipreneurial stances.

In July 2009, the AU Assembly made several decisions that would structure future African resistance and empower African antipreneurs. First, it called for an African court to be created to try mass atrocity crimes, which represented a third tactical move. Second, the AU Assembly called on African states parties not to cooperate with ICC arrest and surrender orders – a clear call to violate core Rome Statute obligations.Footnote 55 This latter (and fourth) example of tactical resistance was particularly shocking given a majority of African ICC members could have blocked it. Only Chad officially dissented,Footnote 56 although Botswana, South Africa, Benin, and Uganda indicated unease and declared they would arrest al-Bashir if given the opportunity.Footnote 57

The AU’s High Level Panel on Darfur (HLPD), established by the PSC, then entered the fray. It argued that justice was a key element of addressing the conflict in Darfur, that Sudan must deal with the crimes committed in Darfur, and called for a removal of immunity for ‘State actors’. Its decision therefore seemingly reflected the anti-impunity norm. But it also added detail to the third tactical resistance move by suggesting a hybrid court with Sudanese and non-Sudanese African judges be created to ‘Africanise’ international criminal justice; this constitutes a third form of strategic resistance, namely, the notion that African problems require African solutions.Footnote 58 It was clear that al-Bashir would never be tried by such a court,Footnote 59 so we therefore lean towards characterising the HLPD as a creative resister.

It was a notable development; for the first time the peak regional organisation – or at least a panel empowered by it – had shifted into an antipreneurial role, even as it rhetorically supported anti-impunity. Alternatively, this incident is suggestive of the sorts of dynamics described by Acharya’s subsidiarity model;Footnote 60 the HLDP was invoking local non-interference norms to resist the application of the ostensibly globally-applicable anti-impunity norm.

The AU calls for the Rome Statute’s amendment

Before mid-2009 the AU’s resistance had mainly taken place vis-à-vis the UNSC (and in the media). But thereafter it shifted to the ICC’s ASP. However, as noted earlier, AU initiatives can only be brought before the ASP by dual AU/ICC members, and only they could engage directly in discussions, limiting the scope of some African actors to resist.

Before the ASP meeting in November 2009, 26 African ICC members and 15 non-members met in Addis Ababa. Four main positions – essentially, demands for reform – emerged:

  1. 1. The interests of peace be considered alongside the interests of justice in Prosecutorial guidelines for when to investigate, or not;

  2. 2. The power of the UNSC to refer cases should remain;

  3. 3. The UN General Assembly should be empowered to defer ICC proceedings when the UNSC fails to make a decision;

  4. 4. There should be a discussion regarding whether or not the leaders of non-parties had their immunity removed by the Rome Statute.Footnote 61

Four points are relevant. First, a norm of peace was framed in opposition to justice. Second, the role of UNSC was deemed as legitimate, which is noteworthy given broader African dissatisfaction with its undemocratic and unrepresentative nature. Third, the AU was creatively seeking to empower the General Assembly to defer; because it was dominated by developing countries, the Assembly would hopefully be a friendlier place for deferral discussions (if perhaps more unpredictable). Fourth, the AU was not seeking to re-assert ‘blanket’ sovereign immunity; instead, it wanted to challenge the UNSC’s power to strip leaders of non-ICC members of sovereign immunity.

Ultimately only points 1 and 3 were put to the ASP. Enthusiasm to ‘action’ them was, however, limited. South Africa agreed to formally submit them but made it clear that it was up to each African ICC member to decide whether to support or not.Footnote 62 Indeed, South Africa may have only agreed to take the lead to head off more drastic measures, such as non-cooperation or mass withdrawal.Footnote 63 But only four African ICC members supported point 1 – Burkina Faso, Namibia, Senegal, and South Africa – and only Namibia and Senegal supported point 3.Footnote 64 Nevertheless, they were presented to the ASP. Point 1 represented a fifth tactical move designed to weaken anti-impunity by providing a new reason not to prosecute, backed by the second strategic justification (that is, applying anti-impunity might endanger peace). Point 3 – the attempt to enable ‘forum-shopping’ by empowering the General Assembly to defer matters – represented a sixth tactical move. But both proposals were rejected by the ASP.Footnote 65

This incident highlighted both the inherent difficulties the AU encounters when it has no formal standing in important contestation sites like the ASP, and that formal AU statements can mask complicated internal dynamics. In other words, non-ICC members – unapologetic antipreneurs – can use peer-pressure in AU meetings to secure anti-ICC resolutions, but because they cannot appear in the ASP their ability to ‘follow through’ is limited. The AU seems to have been playing a competitor entrepreneur role at his time: it was not rejecting anti-impunity outright but was instead ‘feeding-back’ a revised version of the norm, seeking changes to its scope and content in a manner reminiscent of Acharya’s circulation model.Footnote 66 But the effort was brusquely rebuffed, which no doubt contributed to the subsequent stiffening of resistance.

Back to the AU

After the setback in the ASP, the AU’s Assembly called again for deferral of the al-Bashir matter and tried to muster a common African position.Footnote 67 Then after a separate arrest warrant was issued for al-Bashir (for genocide) in July 2010, more strenuous resistance began. At the AU summit that month, the Assembly called again for AU members to practice non-cooperation and rejected a proposal made at the November 2009 ASP meeting to open an ICC-AU liaison office in Addis Ababa.Footnote 68 The AU Commission Chairperson Jean Ping claimed this proposal was part of a ‘plot’ against Africa.Footnote 69 We therefore treat the rejection of the liaison office as a seventh tactical move. These formal positions did mask cracks in the perceived unanimity of the African position. ICC members like South Africa, Ghana, and Botswana argued strongly against efforts by non-members like Libya, Eritrea and Egypt to critique anti-impunity and the ICC.Footnote 70 But other ICC members supported the critics. A trend was becoming discernible: the entrepreneurs seemed to have lost momentum, they were fighting rear-guard actions and their numbers were dwindling.

At the July 2010 AU summit the President of Malawi Bingu wa Mutharika, who also Chaired the AU, stated that Heads of State and Government should not face ICC prosecution and should only be tried by African courts, enabling him to claim he was ‘not condoning impunity’.Footnote 71 This position added nuance to the debate: it did not completely repudiate anti-impunity but instead it drew on African solidarity norms to modify and circumscribe it. It reasserted sovereignty, but an updated version of African sovereignty where African problems are addressed with African solutions in a similar manner to how the HLPD had recommended handling the al-Bashir matter a year earlier. Indeed, trying al-Bashir in Africa was discussed at the July summit, although the debate went nowhere after it became clear no existing court could do so. Then, in October 2013, the AU reiterated that international courts should have no jurisdiction over sitting African heads of state and governmentFootnote 72 and it began to move forward on creating an African Court of Justice and Human Rights (ACJHR) by merging the existing African Court of Justice and the African Court on Human and Peoples’ Rights.Footnote 73 There is no provision in the Rome Statute for a regional court to substitute for the ICC because the principle of complementarity only applies to states (although Kenya has made such a proposal to the ASPFootnote 74 ). Yet the move to create the ACJHR obviously reflects the influence of the ‘African solutions for African problems’ and African solidarity norms;Footnote 75 it represents the evolution of what we called earlier the third example of tactical resistance.

The HLPD’s call for an ad hoc hybrid court had morphed into moves to create a permanent, standing African competitor to the ICC. It is therefore tempting to conclude that the AU was still acting as a competitor entrepreneur: it had issued multiple statements supporting anti-impunity, only qualifying its support by claiming this norm should be implemented by a local, not a global, institution. Yet in June 2014 the nuanced 2010 position – that an anti-impunity norm should be implemented by African courts – was abandoned. While an AU summit voted to amend the ACJHR protocol via the Malabo ProtocolFootnote 76 to expand the jurisdiction of the court to include international crimes, such as genocide and crimes against humanity, it also voted to exclude sitting Heads of State and Government from the new court’s jurisdiction.Footnote 77 The AU had clearly become a creative resister: this eighth tactical move – calling for immunity for sitting Heads of State and Government – was now a formal demand that directly challenges a core precept of the anti-impunity norm as enumerated in the Rome Statute – that no one is immune to prosecution – and amounts to a determined defence of the status quo sovereign immunity norm garbed in the cloak of rhetorical commitment to anti-impunity.Footnote 78

We therefore pause briefly to consider which of Acharya’s models best describe the prevailing dynamics between African actors and the ICC in mid-2014. On the face of it one might conclude norm localisation was taking place: African actors were adapting the anti-impunity norm to fit ‘local normative priors’. African actors certainly claimed they were doing this. But we said earlier that a ‘circumscribed’ anti-impunity norm is nonsensical given senior officials have authority to order mass atrocity crimes (and granting them immunity creates other problematic incentives). Accordingly, we conclude that by the middle of 2014 Acharya’s subsidiarity model best describes the prevailing dynamics; local actors were invoking non-interference norms to protect themselves against a norm from the global centre.

Post July 2014: Mass withdrawal?

By mid-2014 key African actors had shifted decisively into ‘creative resistance mode’. Since then little has changed; indeed, the AU’s resistance to anti-impunity has stiffened further. For example, in January 2016, the AU Assembly passed a resolution that called for the preparation of a roadmap for mass African withdrawal from the ICC ‘if necessary’ (that is, if AU demands for ‘reforms’, including recognising Head of State and Government immunity, were rejected).Footnote 79 It is highly unlikely this demand will be met given how profoundly it undermines the anti-impunity norm, and also given many African states continue to support the ICC.

Then, in January 2017, the AU Assembly adopted an ICC ‘Withdrawal Strategy’, which represented a more nuanced approach to the ICC. It notes that the goals of the AU are to:

  1. a) Ensure that international justice is conducted in a fair and transparent manner devoid of any perception of double standards;

  2. b) Institution of legal and administrative reforms of the ICC;

  3. c) Enhance the regionalisation of international criminal law;

  4. d) Encourage the adoption of African Solutions for African problems;

  5. e) Preserve the dignity, sovereignty and integrity of Member States.Footnote 80

This strategy does not actually call for mass withdrawal. Instead, it examines various legal issues related to potential withdrawal. It discusses proposed amendments to the Rome Statute (identified as ‘preconditions’ for non-withdrawal) the AU would like to see, including, most importantly, deferring prosecution for sitting Heads of State or Government, while also noting they ‘will not exempt them from criminal liability’.Footnote 81

Thus, the strategy implicitly argues for reform from within, driven by African states. And on the face of it Heads of States and Governments should be held accountable, thus seemingly reaffirming support for the anti-impunity norm. Yet, by demanding that accountability should be deferred, the Withdrawal Strategy continues to fundamentally repudiate a key element of the anti-impunity norm – that is, that no one is above the law and exempt from prosecution.

This represents an unresolved paradoxical position: paradoxical because anti-impunity as embodied in the Rome Statute makes little sense unless all are subject to the same laws; unresolved because fundamental disagreements remain among African ICC members. Nigeria, Senegal, Cape Verde, and Liberia entered formal reservations about the strategy, while Malawi, Tanzania, Tunisia, and Zambia wanted more time to study it.Footnote 82 Resistance to the ICC was not overwhelming, but neither is there overwhelming support for it. The decision should ‘be understood as a decision taken by individual states with competing views, rather than a unitary collective body’.Footnote 83 Nonetheless, it represents a serious challenge to anti-impunity.

Accordingly, while mass withdrawal has not happened, we discuss below how three African countries have formally moved to withdraw (with two reversing their decisions), and others have actively considered doing so.

To briefly summarise this long section, the AU was an entrepreneur supporting the anti-impunity norm (as it appears in the Rome Statute) for a decade after 1998; ‘adaptive’ dynamics of the sort described in Acharya’s localisation model prevailed. But events since the request for an al-Bashir arrest warrant in July 2008 tell a very different story. In late 2009 the AU shifted to a competitor entrepreneur role, trying to leverage its earlier resistance to amend how the anti-impunity norm should be implemented by requiring the ICC to weigh peace against justice and by empowering the UNGA to defer cases. Neither initiative directly challenged the core of anti-impunity (that is, that no one should be immune), but the prevailing dynamics had changed to ‘feed-back and seek change’ mode, as described by Acharya’s circulation model. But after these initiatives failed, the AU’s behaviour – especially the moves to create the ACJHR and ‘prepare for’ withdrawal if heads of State and Government immunity is not restored – suggest it has become a creative resister; it makes cosmetic concessions to normative change, but ultimately it defends the normative status quo. Accordingly, we conclude that since 2010 Acharya’s subsidiarity model, which describes resistance by local actors to a norm promoted by ‘central actors’ (in this case an international institution),Footnote 84 best describes the prevailing dynamics. The recent Withdrawal Strategy is, again, suggestive of a return to circulation-model dynamics, but the demands for reform are quite radical, and given that less-ambitious reform proposals were rejected by the ASP in 2009 it seems likely the latest initiative will also fail. We therefore see little prospect that resistance will wane any time soon.

Antipreneurial African states

We have focused on the AU-as-actor, but because it is also a site of contestation we now examine several key African states that have either indicated their intent to withdraw from the ICC (South Africa, Burundi, and Gambia) or who have otherwise become resisters (Kenya, Uganda, and Namibia). These states have all shifted towards the antipreneurial end of the role-spectrum, although their exact role is somewhat unclear given the mixed messages emanating from several of them. We also do not suggest that there are no African entrepreneurs anymore (see below).

Further, we note that by mid-2016 al-Bashir had travelled to the following African ICC-member states without being arrested: Chad (2010, 2011, 2013 (twice), 2014); Kenya (2010); Djibouti (2011, 2016); Malawi (2011); Nigeria (2013); Democratic Republic of the Congo (2014); South Africa (2015); and Uganda (2016). Earlier we called the AU’s decision to call for non-cooperation with the ICC in July 2009 the fourth instance of tactical resistance. Obviously these specific decisions by states to allow al-Bashir to travel with impunity are linked to it and, together, they constitute a particularly effective act of resistance.


Burundi was the first African state to formally indicate its intention to withdraw (in October 2016). The government asserted that the ICC was an instrument for powerful countries to punish the weak that do not do their bidding. Senior government figures were facing an ICC investigation for post-election violence in 2015,Footnote 85 and thus may have been trying to avoid the potential repercussions of an investigation, demonstrating the weakness of its commitment to anti-impunity and the self-interested nature of the withdrawal decision.

South Africa

Despite initially being a strong ICC-supporter, the question of whether to action the al-Bashir arrest warrant proved troubling for South Africa. Pretoria stated in 2009 that it would arrest al-Bashir if he came to South Africa (if a bit reluctantly),Footnote 86 so he avoided travelling there for several years. But as we saw above, in 2009 South Africa – somewhat reluctantly – did present the AU’s proposals for amending the Rome Statute to the ASP. And by March 2014 Pretoria was shifting towards the antipreneurial end of the role-spectrum. For example, Deputy President Kgalema Motlanthe supported creating the ACJHR, saying doing so responded to ‘the yearnings of ordinary Africans for justice whilst being sensitive to the unique nature of the Africa context’.Footnote 87

Then in June 2015 al-Bashir travelled to South Africa to attend an AU summit. The South Gauteng High Court immediately ordered that he be prevented from leaving until it could determine whether or not he should be arrested. The court subsequently issued an arrest warrant, but before it could be executed al-Bashir was spirited away from the conference to an air force base and flown home. Pretoria argued that al-Bashir held ‘special’ immunity because he was a Head of State attending an AU summit.Footnote 88 South Africa had privileged African solidarity over human rights before,Footnote 89 and this incident was a very clear example of the fourth resistance tactic – non-cooperation – in action. The South African courts, however, ruled that the government had violated its international obligations, which had been enshrined in domestic law in 2002.Footnote 90 Despite these domestic legal challenges, Pretoria then stunned the world when it announced on 21 October 2016 it would withdraw from the ICC.

South Africa’s position towards anti-impunity has therefore changed significantly. It claims that it has not rejected it because in its instrument of withdrawal the government noted the country’s commitment to fighting impunity. Yet, this commitment appears to be in conflict with its commitment to recognise diplomatic immunity (that is, failing to do so could allegedly lead to ‘regime change’) and its commitment to the peaceful resolution of conflicts, the arguments it offered to justify withdrawing.Footnote 91 Nevertheless, domestic opposition to withdrawal intensified, and on 22 February 2017, in response to the Democratic Alliance’s petition, the High Court found that the government did not have the authority to withdraw without the consent of parliament and ordered the withdrawal be revoked,Footnote 92 which occurred on 7 March.Footnote 93 The government also revoked a bill that would have repealed national laws that outlawed genocide, crimes against humanity, and war crimes.Footnote 94

While domestic proceedings were occurring, the ICC decided in December 2016 that it would rule on whether South Africa had acted unlawfully when it did not arrest Bashir.Footnote 95 A hearing was held on 7 April 2017, and the Court published its decision on 6 July. It found that South Africa had violated its obligations under the Rome Statute. At the same time, it decided not to refer South Africa to the UNSC. The Pre-Trial Chamber noted that it had made six other referrals to the UNSC for non-cooperation in the Bashir case, with no effect (indeed there had been a total of 13 prior findings of noncooperation and/or referrals for action).Footnote 96 It also noted that the government had withdrawn its appeal against the decision by the Supreme Court of Appeal of South Africa that its actions were illegal, thus leading to the conclusion that the government has presumably accepted its obligations to cooperate with the Court.Footnote 97

It is unclear when or if the South African government might attempt to gain parliamentary approval for a withdrawal, with the Justice Minister Michael Masutha having withdrawn a relevant bill on 14 March 2017Footnote 98 and other domestic priorities crowding out the matter.Footnote 99 Indeed, while a discussion document still frames ‘manipulation’ of the ICC in terms of anti-imperialism, reiterates grievances about the UNSC, and argues that the ACJHR should deal with mass atrocity crimes,Footnote 100 this document is noticeably more restrained than a similar 2015 document.Footnote 101 Yet, President Jacob Zuma stated in June 2017 ‘that the decision to withdraw is a principle matter and the principle still stands’, and that the government was working ‘to rectify the procedural challenges’. At the same time, he also noted the reluctance of African states to engage in a mass withdrawal.Footnote 102 South Africa is therefore a highly conflicted country which has traversed almost the full range of the norm dynamics role-spectrum; we conclude that it is currently sitting at a tipping point between the entrepreneurial and antipreneurial sides, demonstrating the fluidity of these identities and correlated interests.


In late October 2016, Gambia also officially announced it intended to withdraw. It repeated the accusation that the ICC was targeting Africans – calling it the ‘International Caucasian Court for the persecution and humiliation of people of colour, especially Africans’.Footnote 103 It also argued that Western war criminals had not been prosecuted, and it had tried to get the ICC to prosecute EU states for migrants drowned in the Mediterranean. But it had also been accused of election-related repression.Footnote 104 Thus, while voicing broad normative themes – for example, discrimination against Africans – the main explanation for the withdrawal appears to be the government’s naked self-interest.

To complicate matters, after the withdrawal was announced the incumbent President Yahya Jammeh lost an election, and the President-elect Adama Barrow vowed to reverse the withdrawal decision.Footnote 105 Jammeh subsequently rejected the results and engaged in further repression, but he was eventually forced to step down and Barrow revoked the withdrawal on 10 February 2017,Footnote 106 noting Gambia’s commitment to human rights and the ‘principles enshrined in the Rome Statute of the International Criminal Court’.Footnote 107 This incident demonstrates how unstable some African states’ interests and normative commitments can be.


Until the recent moves to withdrawal, Kenya had moved the furthest towards the antipreneurial end of the role-spectrum. In late 2009 the ICC opened an investigation into postelection violence in 2007, resulting in the indictment of several Kenyans, including persons (Kenyatta and Ruto) who would later become President and Vice-President. They both sought to delay proceedings and repeatedly denounced the ICC at political rallies (as a threat to Kenyan sovereignty, a destabilising force, and an insulter of African pride) and allegedly began to ‘eliminate, intimidate or bribe’ witnesses’.Footnote 108 As we saw at the outset, they eventually succeeded and charges were dropped (but ‘without prejudice’, meaning they could be refiled).Footnote 109

Kenya has also taken actions contrary to the anti-impunity norm which are technically independent of – but are obviously connected to – its direct dealings with the ICC. In August 2010 al-Bashir travelled to Kenya.Footnote 110 In January 2011 the AU Assembly supported Kenya’s decision to not arrest al-Bashir and called for deferral of the ICC proceedings in KenyaFootnote 111 – although in November 2011 a Kenyan court directed that al-Bashir be arrested if he travelled to Kenya again.Footnote 112 In 2013 the Kenyan parliament voted to withdraw from the ICC, although President Kenyatta did not act on this. Kenya has also been leading efforts to amend Article 27 of the Rome Statute to give sitting Heads of State and Government immunity and has also backed preparations in the AU to secure a mass African withdrawal from the ICC if the amendment push fails.Footnote 113 Thus, while Kenya rhetorically supports anti-impunity, it has actually worked hard to undermine fundamental aspects of the norm.


Uganda was the first state to refer a matter to the ICC Prosecutor, which suggests it was a norm entrepreneur in 2003. But the reality is much more complicated since Uganda attempted to use the ICC as a weapon against the rebel Lord’s Resistance Army (LRA).Footnote 114 But by 2008 Kampala began to consider the ICC as an obstacle, not a resource, on the basis that the ICC investigation was obstructing peace negotiations with the LRA.Footnote 115

President Museveni therefore began promoting an alternative norm to anti-impunity, the interests of peace over justice (an argument also deployed vis-à-vis al-Bashir case). But the shift was not immediately a comprehensive one because Kampala argued for watering down the July 2010 AU statement on non-cooperation,Footnote 116 and in 2009 it revoked an invitation to al-Bashir to attend a summit in Uganda. But over time Uganda’s resistance hardened: Museveni subsequently called the ICC a Western ‘tool that is out to punish Africa’;Footnote 117 in May 2016 al-Bashir visited Uganda to attend Museveni’s fifth Presidential inauguration and was not arrested; and Museveni denounced the ICC as a ‘bunch of useless people’Footnote 118 – even though a year earlier he allowed the LRA commander Dominic Ongwen to be transferred to the ICC.Footnote 119

Uganda has thus become an outspoken critic of the ICC. Its government is seemingly not ‘committed to’ or truly ‘constituted by’ anti-impunity; it used the norm instrumentally when its interests suited doing so, but it invoked alternative norms – peace over justice and African solidarity – when doing so became expedient. And if that instrumentality is no longer compelling enough, it might leave the court: in October 2016 a Ugandan cabinet minister suggested that the withdrawal process had already begun,Footnote 120 although in April 2017 the Ugandan attorney general, William Byaruhanga, stated that while Uganda had concerns about the Court, it had ‘not considered withdrawing’.Footnote 121


The Namibian government announced in 2015Footnote 122 and 2016Footnote 123 that it intended to withdraw from the ICC, although it hasn’t yet. Its reasons are similar to the other states’, namely, that the ICC is biased against Africa and is essentially pursues regime change in Africa. In February 2017, the government indicated that it supported what it described as the AU Assembly’s decision for a collective withdrawal from the ICC (thus miscasting the actual decision). It cited the demand for sitting presidents to be allowed to serve their terms in office before being tried by the ICC, tying this directly to the issue of peace and stabilityFootnote 124 (although this contradicts a subsequent statement).

Thus these six states have moved towards the antipreneurial side of the role-spectrum, although not all have done so decisively. A somewhat confused and chaotic situation is presented. While not necessarily ‘pure’ antipreneurs – none have completely repudiated the anti-impunity norm – they are certainly creative resisters (with perhaps the exception of Gambia after its government changed), attempting to at least partially hollow out the anti-impunity norm, including by putting forth alternative proposals for how it might be implemented, while essentially defending the status quo sovereign immunity norm. While we recognise ongoing support for other aspects of anti-impunity, five of these six states continue to call for an exemption from prosecution for sitting Heads of State and Government, representing a rejection of the core of the anti-impunity norm – that no one is above the law.

The supporters

There is still significant support for the ICC in Africa. Four African countries have joined since 2010 (Seychelles, Tunisia, Cape Verde, and Côte d’Ivoire).Footnote 125 Botswana remains an outspoken ICC-supporter, challenging anti-ICC statements from the AU, declaring that it would arrest al-Bashir if given an opportunity, and stating in 2010 – somewhat ironically given the AU’s penchant for invoking sovereignty concerns – that ‘we have not surrendered [our] sovereignty … to the AU’.Footnote 126 In July 2017 it formally domesticated the Rome Statute, which, the Minister for Defense, Justice, and Security Shaw Kgathi said ‘lifts’ diplomatic immunity.Footnote 127 And in July 2016 Botswana, Nigeria, Senegal, Tunisia, Côte d’Ivoire, and Algeria pushed back against calls for mass withdrawal at the 27th African Union Summit, preventing the proposal from being included on the agenda.Footnote 128

After the withdrawal announcements in late 2016 numerous African civil society organisations protested,Footnote 129 and Botswana, Côte d’Ivoire, Malawi, Nigeria, Senegal, Sierra Leone, Tanzania, and Zambia reiterated their support for the ICC. At the 2016 ASP in December, other African states reiterated their support for the ICC (including both Namibia and Uganda, seemingly contradicting their statements about withdrawing). Even after it had initiated its withdrawal, South Africa had ‘expressed hope for dialogue that could forestall … withdrawal’Footnote 130 and also indicated that it would continue to cooperate with the ICC until its withdrawal was completed.Footnote 131 Senegal has also seemingly changed its stance (recall it had called for withdrawal in 2009, after having been a strong early supporter). And after the AU approved the Withdrawal Strategy in early 2017 a number of countries expressed significant reservations. Therefore, recent events demonstrate that the entrepreneurial camp is not just ‘bleeding members’ – and Jacob Zuma’s June 2017 statement reinforces this perception.Footnote 132 And the moves towards antipreneurialism by several states have proven to be less decisive than they initially appeared, even though significant tensions and push factors in favour of withdrawal remain, and non-cooperation remains a serious challenge to the viability of the ICC.


African actors have been resisting anti-impunity in the form it takes in the Rome Statute, and the efforts of the ICC to implement it, since the ICC Prosecutor requested an arrest warrant for Omar al-Bashir. There were some ‘rumblings of discontent’ before then, but this event galvanised African resistance, morphing into full-blown resistance against the ICC on the part of some African ICC members, and while two countries – for different reasons – turned back from the brink of withdrawal – there is little evidence that resistance will wane in the foreseeable future.

We canvassed a number of ways African actors practised resistance, organised according to a conceptual distinction between strategic (that is, justifications) and tactical (that is, discrete moves) resistance. There were, in effect, three of the former. First, African actors justified resisting by invoking the sovereignty norm, arguing that anti-impunity was being abused and that African states’ sovereignty was imperiled. Second, they argued that sometimes the interests of peace should be prioritised over, or at least weighed against, the pursuit of justice. Third, they invoked local norms like ‘African solidarity’ and ‘African solutions to African problems’ (which are related to wider norms like anti-imperialism). These justifications then informed or supported eight types of tactical resistance: African actors called on the UNSC to defer ICC proceedings; threatened mass African withdrawal from the ICC; attempted to create alternate African judicial structures; called for non-cooperation with the ICC; demanded prosecutorial guidelines be changed; sought to empower the UN General Assembly to defer ICC proceedings; prevented the ICC establishing an AU liaison office; and called for the Rome Statue’s amendment to recognise Head of State and Government immunity.

Interestingly, the use of this strategic/tactical distinction suggests that while the two types of resistance are related, they serve somewhat different functions or are directed towards somewhat different audiences. Specifically, strategic resistance is perhaps directed more towards other African actors, to achieve African unity, especially in AU forums. It is of course directed towards non-African actors too; after all, supporting one’s case by presenting purely self-interested arguments is generally an unpersuasive strategy, so these justifications at minimum provide ‘higher-purpose’ reasons for resisting. African actors failed to convince many non-Africans to support or acquiesce to their demands, but one outcome is clear: enough African actors are now convinced-enough that anti-impunity (and the ICC) threatens their interests for the AU to become an important vehicle for practising tactical resistance. Thus tactical resistance is directed more towards ‘global-level opponents’ (or it takes place in global-level contestation-sites).

And many of these tactical moves have essentially failed: the Council has not deferred any ICC proceedings; no mass African withdrawal has taken place; alternative judicial structures have not yet been implemented; prosecutorial guidelines have not been changed; the General Assembly cannot defer proceedings; and the Rome Statute does not confer immunity to Heads of State and Government. The only outright successes have been the decision to not establish an AU-ICC liaison office and the call for noncooperation. The three recent withdrawals (with two reversals) represent only a partially successful resistance tactic, although this tactic does have potential for significant disruption in the future.

These findings therefore have several important implications. The withdrawal of three African states seemed – as 2017 dawned – to have dealt the effort to entrench the anti-impunity norm – and by extension, the ICC itself – a severe below. Yet the two reversals demonstrate that the resistance is by no means monolithic. The Gambian reversal was accompanied by strong rhetorical support for the ICC and anti-impunity by the new President. The reversal by South Africa demonstrates the power of domestic legal and civil society organisations to push back against governments when they seek to undermine anti-impunity. Yet, the development of the norm has certainly stalled, and could lose further ground if additional states move to withdraw, or if South Africa reinstates its withdrawal. The fact that non-cooperation practices – most notable regarding the matter of arresting al-Bashir – have become very common is also very troubling given how heavily the ICC relies on members for assistance. Non-cooperation has, at minimum, significantly undermined the ICC’s ability to implement the anti-impunity norm.

Yet the ICC does not operate in Africa alone; indeed, the Prosecutor is currently preliminarily examining five ‘situations’ outside AfricaFootnote 133 and one, Georgia, has been upgraded to a formal investigation.Footnote 134 But no arrest warrants have been issued in these cases, let alone warrants against senior political figures. Executing warrants of this sort would constitute strong evidence that the anti-impunity norm was advancing, so we caution against finding the effort to entrench anti-impunity has made major strides forward until such occurs or, at minimum, until issuing such does not provoke the sort of reaction that the al-Bashir warrant precipitated in Africa. And as an aside, while the ICC’s Pre-Trial Chamber has repeatedly admonished African states for not arresting al-Bashir, the ASP has not yet punished any non-cooperating states and while the ICC has referred several to the UNSC,Footnote 135 the Council has also taken no action against them. The ICC’s decision not to refer South Africa to the UNSC is perhaps partly due to recognition of the futility of such referrals. As noted earlier, the reactions by third parties to violations of norms, rules or laws profoundly affects judgements about their efficacy or strength; as Jutta Brunnée and Stephen Toope put it, ‘when posited rules are consistently evaded or undermined without legal consequence, the rules themselves are compromised’.Footnote 136 They had the R2P norm in mind, but the same logic applies to the legal requirement for member states to assist the ICC. This buttresses our finding that the effort to entrench the anti-impunity norm has at minimum stalled – and could go backwards if non-cooperation continues.

Having said this, resistance is not uniform in Africa: while the resisters seem to have in effect ‘captured’ the AU, enabling them to deploy it to organise and legitimise resistance, many African states remain ICC supporters. And while the resisters seem to have the upper hand, most seem to actually want to make the anti-impunity a permissive norm; they are not happy with it being a strong prescriptive Footnote 137 norm, but they do not want to ‘roll-back’ the normative status quo to the strict-sovereign immunity position. Instead they want more discretion over when and how to apply it in African contexts – albeit in a manner which does provide immunity to senior political figures. This does not of course bode well for the pure version of anti-impunity – which brooks no exceptions – which the ICC has been seeking to entrench since 2002. But it is nevertheless apparent that the normative status quo has shifted substantially since then, meaning it seems unlikely we will see a return of traditional, strict-sovereign immunity, and the practice of indicting former leaders has arguably become quite well entrenched (although this creates its own problems, especially incentives for sitting leaders to remain in office, a troubling recent trend in Africa).Footnote 138

Several important theoretical findings also flow from this analysis. The first is that, in effect, ‘practice matters’ (profoundly). In a system of law without strong enforcement mechanisms – like the contemporary international system – many states must be constituted by a law/norm for it to be very effective. Thus while one might presume that because five of eight tactical resistance moves failed – and one has so far had limited effect – resistance overall failed. But one tactic – non-cooperation – was especially potent. This might be a unique feature of this case given how reliant the ICC is on members’ cooperation. But it nevertheless demonstrates the limitations of consent-based systems of global governance and the importance of norm-conforming practice. Withdrawing may become an even more potent resistance tactic. At minimum withdrawing states will, by definition, not be cooperating with the ICC. Withdrawals also provide a precedent for additional withdrawals, further undermining prospects for cooperation. More generally, each withdrawal deals a blow to the ICC’s credibility as an inclusive international organisation, and its aspirations for universality.

Further theoretical findings can be made vis-à-vis the entrepreneur-antipreneur framework itself. Most obviously, resisters in this case did not enjoy ‘overwhelming’, or even ‘substantial’ advantages. They did enjoy advantages in the realm of practice as we have just seen: in particular non-cooperation with an emerging norm is a potent resistance tactic. But they did not enjoy clear advantages in the realm of institutionalisation because the anti-impunity norm has already been codified in the Rome Statute and the ICC has been established. This has allowed pro-anti-impunity entrepreneurs to block African states’ demands for the Rome Statute to be amended; securing an 88 per cent majority in the ASP is extremely difficult. And two ICC members – France and Britain – can veto attempts to defer ICC proceedings at the UNSC. Thus some ‘frustrating’ and ‘blocking’ opportunities – which Bloomfield argued typically accrue to antipreneurs – are actually enjoyed by entrepreneurs in this case. Having said that, China and Russia (non-ICC members) can also veto attempts to punish violators and thereby undermine efforts to entrench anti-impunity. Thus we find that this case does suggest that antipreneurs do on balance enjoy some, but not ‘strong’, defensive advantages simply because entrepreneurs must initiate and sustain momentum for normative change. In short, stalemate constitutes a ‘win’ for antipreneurs – not a resounding one, but a win nonetheless – which reinforces how much practice matters in consent-based governance systems.

In a more abstract sense, this case clearly demonstrates Risse’s logic of argument at work. ICC-resisters have invoked alternative norms and framings to justify their actions: they invoked the status quo sovereignty norm to defend Africa from perceived bias and abuse by the West; they invoked an alternative norm – peace – to argue for immunity for sitting Heads of State and Government; and they deployed African solidarity norms to argue for regional over global justice mechanisms. These are arguments over which norms to implement in a given situation, as well as how to implement the anti-impunity norm. Some of the argumentation is within a truly ‘open frame’ where actors’ preferences are open to discursive challenge.Footnote 139 There may thus be real disagreements about whether or not it is truly best to allow a brutal dictator to continue their atrocities unimpeded to facilitate a broader peace settlement. However, much of the ‘arguing’ seems to have been ‘strategic’, intended to convince others of a fixed position rather than suggesting that actor was open to persuasion.Footnote 140

The question of how open the arguing has been is beyond the scope of this article, as is the question of definitively determining the motivations of all of the different actors. We have no doubt that there are real disagreements about normative priorities, while at the same time these disagreements may also mask more self-interested motivations. Our task is not to evaluate the validity of these disagreements and normative positions, but rather to examine the dynamics of contestation over these positions. And it incontestable that the AU and some African states have used these arguments to call into question and undermine a fundamental element of the anti-impunity norm – that is, that nobody, regardless of status, is exempt from prosecution for atrocity crimes. One might argue that such actors are norm entrepreneurs, supporting norms of peace or African solidarity. This may be true – again, a deep investigation of the motivations is beyond the scope of this article – but that does not forestall identification of antipreneurial behaviour vis-à-vis the anti-impunity norm.

Finally, we noted earlier that identifying whether and how far actors shift along the role-spectrum helps determine which of Acharya’s three models best describes prevailing dynamics in norm contestation contexts at particular points in time. The utility of doing so lies in the fact that these models describe outcomes: adaptation leading to successful norm diffusion; resistance, meaning the norm does not successfully diffuse (at least not to all regions); and feed-back, which implies the norm’s scope and content might change. Identifying which dynamics prevail has two potential uses, although we only present these as potential directions for future research.

First, it may aid research design in that hypotheses can be derived from each model and then tested. For example, evidence of actors invoking competing ‘normative priors’ might suggest resistance is stiffening, but arguably two outcomes are possible; these actors may be trying to defeat the norm outright (or maybe just its relevance in their region) – making them ‘pure’ antipreneurs – or they may be preparing to feed it back to the global centre for renegotiation. Scholars could therefore examine key actors closely; if they were shifting into antipreneurial roles it would suggest diffusion was faltering, but if they were behaving more like competitor entrepreneurs then it would suggest further negotiations were looming.

Second (and we make this point even more tentatively) being aware that key actors may shift along the spectrum – changing the prevailing dynamics as they do so – could have policy implications. Specifically, it might alert norm entrepreneurs to the need to alter their promotional activities. They may decide they need to devote more resources to lobbying an actor directly to ensure it remains a pure entrepreneur; or they may need to focus on ‘exposing’ a creative resister as a ‘Trojan Horse’, an actor which rhetorically supports the challenger norm but who, in actuality, primarily seeks to defend the normative status quo.

Identifying the emergence of more competitor entrepreneurs might be the most interesting scenario. It may suggest the original champions of the norm may have to begin deciding where their ‘red lines’ lie. For example, and in the context of this case, would champions of anti-impunity be prepared to include recognition of Head of State and Government immunity in the Rome Statue? Probably not, because doing so would strike at the core of anti-impunity, although they might be prepared to make less-substantial concessions, like extending the principle of complementarity to encompass regional courts. Nevertheless, the strength of contemporary African resistance makes it difficult to envisage concessions adequate-enough to mollify resisters; would, for example, ICC-supporters agree to extend complementarity to an ACJHR which could not try Heads of State and Government? Again, probably not, yet many African states envisage the ACJHR being limited in this way. Accordingly, the prospects of the anti-impunity norm emerging from the current impasse and resuming its advance towards becoming the normative status quo may, for the moment at least, largely depend on the success of the ICC’s investigations in regions other than Africa.


The authors would like to thank the anonymous reviewers, as well as Antonio Arcudi, Andrea Birdsall, Christian Downie, B. Welling Hall, Ramesh Thakur, participants at the Human Rights in an Age of Ambiguity Conference, and participants at the 2016 annual meeting of the Academic Council on the United Nations System for their helpful comments.


1 Rashid Abdi, quoted in Pilling, David, ‘Bungled Kenyan cases spotlight ICC’s image problem’, Business Day (16 April 2016)Google Scholar.

2 Thirty others had signed but not ratified.

3 Bassiouni, M. Cherif, ‘Negotiating the Treaty of Rome on the establishment of an International Criminal Court’, Cornell International Law Journal, 32 (1999), pp. 443469 Google Scholar.

4 Scheffer, David J., ‘The United States and the International Criminal Court’, American Journal of International Law, 93:1 (1999), p. 12 CrossRefGoogle Scholar.

5 Goldsmith, Jack, ‘The self-defeating International Criminal Court’, University of Chicago Law Review, 70:1 (2003), pp. 9091 CrossRefGoogle Scholar.

6 Ramanathan, Usha, ‘India and the ICC’, Journal of International Criminal Justice, 3 (2005), pp. 627634 CrossRefGoogle Scholar.

7 Bloomfield, Alan, ‘Norm antipreneurs and theorising resistance to normative change’, Review of International Studies, 42:2 (2016), pp. 310333 CrossRefGoogle Scholar.

8 Ibid., p. 312.

9 Ibid., pp. 322–6.

10 Legro, Jeffrey W., ‘The transformation of policy ideas’, American Journal of Political Science, 44:3 (2000), pp. 426429 CrossRefGoogle Scholar; Mahoney, James, ‘Path dependence in historical sociology’, Theory and Society, 29:4 (2000), pp. 507548 Google Scholar.

11 Tsebelis, George, ‘Decision making in political systems: Veto players in presidentialism, parliamentarism, multicameralism and multipartyism’, British Journal of Political Science, 25:3 (1995), pp. 289325 CrossRefGoogle Scholar.

12 Bloomfield, ‘Norm antipreneurs’, p. 326.

13 Ibid., pp. 322–3.

14 Ibid., p. 321.

15 Ibid., p. 313, and Acharya, Amitav, ‘The R2P and norm diffusion: Towards a framework of norm circulation’, Global Responsibility to Protect, 5:1 (2013), pp. 466479 CrossRefGoogle Scholar; Epstein, Charlotte, ‘Stop telling us how to behave: Socialization and infantilization?’, International Studies Perspectives, 13:2 (2013), pp. 135145 Google Scholar have argued that much of the norms literature reveals ‘liberal’ bias by implicitly presenting a picture of ‘enlightened/good’ – and usually Western – entrepreneurs promoting, for example, human rights norms against resistance by ‘unenlightened/bad’ actors.

16 Bloomfield, ‘Norm antipreneurs’, pp. 329–31.

17 Acharya, Amitav, ‘How ideas spread’, International Organization, 58:2 (2004), pp. 239275 CrossRefGoogle Scholar.

18 Acharya, Amitav, ‘Norm subsidiarity and regional orders: Sovereignty, regionalism, and rule-making in the Third World’, International Studies Quarterly, 55:1 (2011), pp. 96102 CrossRefGoogle Scholar.

19 Acharya, Amitav, ‘The R2P and norm diffusion: Towards a framework of norm circulation’, Global Responsibility to Protect, 5:1 (2013), pp. 466479 CrossRefGoogle Scholar.

20 Finnemore, Martha and Sikkink, Kathryn, ‘International norm dynamics and political change’, International Organization, 52:4 (1998), p. 891 CrossRefGoogle Scholar.

21 Rome Statute of the International Criminal Court (1998) UNTS vol. 2187 no. 38544, in force 1 July 2002. Article 25 establishes that the ICC has ‘jurisdiction over natural persons’ (who have allegedly committed mass atrocity crimes (Articles 5 to 8)), while Article 27 specifies that the ‘Statute shall apply equally to all persons without any distinction based on official capacity’ and that ‘[i]mmunities or special procedural rules which may attach to the official capacity of a person, whether under national or international law, shall not bar the Court from exercising its jurisdiction’.

22 Fitzmaurice, Gerald, ‘The foundations of the authority of international law and the problem of enforcement’, Modern Law Review, 19:1 (1956), p. 8 CrossRefGoogle Scholar.

23 Finnemore, Martha and Toope, Stephen J., ‘Alternatives to “legalization”: Richer views of law and politics’, International Organization, 55:3 (2001), p. 749 CrossRefGoogle Scholar.

24 Hoffman, Matthew J., ‘Norms and social constructivism in international relations’, in Robert A. Denemark (ed.), The International Studies Encyclopedia (Blackwell Reference Online, 2014), pp. 13 Google Scholar.

25 Kratochwil, Friedrich and Ruggie, John G., ‘A state of the art on an art of the state’, International Organization, 40:4 (1986), p. 767 CrossRefGoogle Scholar.

26 Hurrell, Andrew, ‘Norms and ethics in international relations’, in Walter E. Carlsnaes, Thomas Risse, and Beth A. Simmons (eds), Handbook of International Relations (New York: Sage, 2002), p. 143 Google Scholar.

27 Richard Price treats three factors – ‘support’, ‘compliance’, and ‘third party reactions’ – as relevant to assessing norm-strength: ‘Detecting ideas and their effects’, in Goodwin, Robert E. and Tilly, Charles (eds), The Oxford Handbook of Contextual Political Analysis (Oxford: Oxford University Press, 2006)CrossRefGoogle Scholar.

28 For the general theory, see Giddens, Anthony, The Constitution of Society: Outline of the Theory of Structuration (Cambridge: Polity Press; 1984)Google Scholar. Also see Weldes, Jutta, ‘Constructing national interests’, European Journal of International Relations, 2:3 (1996), pp. 275318 CrossRefGoogle Scholar; and Hoffmann, ‘Norms and social constructivism’, pp. 1–3.

29 Hoffman, ‘Norms and social constructivism’, pp. 7–8.

30 Risse, Thomas, ‘Let’s argue: Communicative action in world politics’, International Organization, 54:1 (2000), p. 7 CrossRefGoogle Scholar.

31 Tunks, Michael A., ‘Diplomats or defendants? Defining the future of head-of-state immunity’, Duke Law Journal, 52:3 (2002), pp. 652655 CrossRefGoogle Scholar.

32 Germany v. Italy: Greece Intervening, Judgement, ICJ Reports (2012), pp. 99–156.

33 Congo v. Belgium, Judgement, ICJ Reports (2002), pp. 3–34.

34 Tachiona v. Mugabe, US District Court for the Southern District of New York (SDNY), 169 F. Supp. 2nd (2001).

35 Wittgenstein, Ludwig, Philosophical Investigations, Volume II (Oxford: Blackwell, 1953), p. 226 Google Scholar.

36 Schiff, Benjamin N., Building the International Criminal Court (Cambridge: Cambridge University Press, 2008), pp. 2425 CrossRefGoogle Scholar.

37 Roht-Arriaza, Naomi, The Pinochet Effect: Transnational Justice (Philadelphia: University of Pennsylvania Press, 2006)Google Scholar.

38 Bosco, David, Rough Justice: The International Criminal Court’s Battle to Fix the World, One Prosecution at a Time (Oxford: Oxford University Press, 2014), p. 2 Google Scholar.

39 By 1 January 2017, 124 states had ratified the Rome Statute. Thirty others had signed but not ratified.

40 Schabas, William A., An Introduction to the International Criminal Court (4 th edn, Cambridge: Cambridge University Press, 2011), pp. 190199 CrossRefGoogle Scholar.

41 Annan, Kofi, ‘Two concepts of sovereignty’, The Economist, 352 (1999), pp. 4950 Google Scholar; Deng, Francis, Sovereignty as Responsibility: Crisis Management in Africa (Washington, DC: Brookings Institution, 1996)Google Scholar; Mills, Kurt, Human Rights in the Emerging Global Order: A New Sovereignty? (Basingstoke: Macmillan, 1998)Google Scholar.

42 Mills, Kurt, International Responses to Mass Atrocities in Africa: Responsibility to Protect, Prosecute, and Palliate (Philadelphia: University of Pennsylvania Press, 2015)CrossRefGoogle Scholar.

43 Jalloh, Charles C., Akande, Dapo, and du Plessis, Max, ‘Assessing the African Union concerns about Article 16 of the Rome Statue of the International Criminal Court’, African Journal of Legal Studies, 4:1 (2011), p. 14 CrossRefGoogle Scholar.

44 African Union, ‘Statement by Mr. Ben Kioko, Legal Counsel of the African Union Commission on Behalf of the AU Commission’, Review Conference of the Rome Statute of the International Criminal Court (ICC), Kampala, Uganda, 31 May–11 June 2010.

45 Acharya, ‘How ideas spread’.

46 Peskin, Victor, International Justice in Rwanda and the Balkans: Virtual Trials and the Struggle for State Cooperation (Cambridge: Cambridge University Press, 2008)CrossRefGoogle Scholar.

47 Mills, Kurt, ‘“Bashir is dividing us”: Africa and the International Criminal Court’, Human Rights Quarterly, 34:2 (2012), p. 419 CrossRefGoogle Scholar.

48 African Union, AU Doc. Assembly/AU/14(XI), ‘Decision on the Report of the Commission on the Abuse of the Principle of Universal Jurisdiction’, 1 July 2008, AU Assembly, 11th Ord. Session.

49 African Union, AU Doc. PSC/PR/BR(CXLI), Press Statement, 11 July 2008, AU Peace and Security Council, 141st Meeting.

50 International Criminal Court, Press Release, ‘ICC Prosecutor Presents Case Against Sudanese President, Hassan Ahmad Al Bashir, for Genocide, Crimes Against Humanity and War Crimes in Darfur’ (14 July 2008), available at: {}.

51 African Union, AU Doc. PSC/MIN/Comm(CXLII)Rev.1, ‘Communiqué of the 142nd Meeting of the Peace and Security Council’, 21 July 2008, AU Peace and Security Council, 142nd meeting, para. 9.

52 African Union, AU Doc. Assembly/AU/Dec.221(XII), ‘Decision on the Application by the International Criminal Court (ICC) Prosecutor for the Indictment of the President of the Republic of Sudan’, adopted 3 February 2009, AU Assembly, 12th Ord. Session.

53 African Union, AU Doc. PSC/PR/Comm(CLXXV), ‘Communiqué of the 175th Meeting of the Peace and Security Council’, 5 March 2009, AU Peace and Security Council, 175th meeting, para. 4.

54 ‘African countries back away from ICC withdrawal demand’, Sudan Tribune (8 June 2009), available at: {}.

55 African Union, AU Doc. Assembly/AU/13(XIII), ‘Decision on the Meeting of African States Parties to the Rome Statute of the International Criminal Court (ICC)’, adopted 3 July 2009, AU Assembly 13th Ord. Session, para. 5.

56 Ibid.

57 Mills, ‘“Bashir is dividing us”’, pp. 425–6.

58 African Union, AU Doc. PSC/AHG/2(CCVII), ‘Report of the African Union High-Level Panel on Darfur, Darfur: The Quest for Peace, Justice, and Reconciliation’, 29 October 2009, AU Peace and Security Council, 207th meeting, para. 215.

59 Sudanese judges could not impartially assess the matter of their President’s culpability given ‘judicial independence’ is not possible in dictatorial regimes.

60 Acharya, ‘Norm subsidiarity’.

61 African Union, AU Doc. Min/ICC/legal/Rpt.(II), ‘Report of the 2nd Ministerial Meeting on the Rome Statute of the International Criminal Court (ICC)’, 6 November 2009.

62 African Union, AU Doc. EX.CL/568(XVI), ‘Report of the Commission on the Outcome and Deliberations of the 8th Session of the Assembly of States Parties to the Rome Statute of the ICC Held at The Hague, Netherlands from 16 to 26 November’, 26 November 2009, AU Exec. Council, 16th Ord. Session, Annex 2.

63 Mills, ‘“Bashir is dividing us”’, pp. 430–1.

64 African Union, ‘Report of the Commission’.

65 International Criminal Court, ICC-ASP/8/20, Assembly of State Parties, ‘Report of the Credentials Committee’, 2009, pp. 47–81.

66 Acharya, ‘The R2P’.

67 African Union, AU Doc. Assembly/AU/8(XIV), ‘Decision on the Report of the Second Meeting of States Parties to the Rome Statute on the International Court (ICC)’, 2010.

68 African Union, AU Doc. Assembly/AU/Dec.296(XV), ‘Decision on the Progress Report of the Commission on the Implementation of Decision Assembly/AU/Dec.270(XIV) on the Second Ministerial Meeting on the Rome Statute of the International Criminal Court (ICC)’, adopted 27 July 2010, AU Assembly, 15th Ord. Session, para. 6.

69 ‘African Union moves aggressively to shield Bashir from prosecution’, Sudan Tribune (29 July 2010), available at: {}.

70 ‘African Union drops resolution barring arrest of Sudanese president in continent’, Sudan Tribune (26 July 2010), available at: {}.

71 ‘African Union Moves’.

72 African Union, ‘Decision on Africa’s Relationship with the International Criminal Court (ICC)’, Ext/Assembly/AU/Dec.1(Oct.2013), adopted 12 October 2013, available at: {}.

73 Amnesty International, Malabo Protocol: Legal and Institutional Implications of the Merged and Expanded African Court (2016), available at: {}.

74 Nzau Musua and Simon Jennings, ‘Kenya Continues to Push for ICC Changes’, Institute for War Reporting (4 June 2014), available at: {}.

75 Garth Abraham, ‘Africa’s Evolving Continental Court Structures: At the Crossroads?’, South African Institute of International Affairs, Occasional Paper No. 209 (January 2015), available at: {}.

76 African Union, AU Doc. EX.CL/846(XXV) Annex 5, ‘Draft Protocol on Amendments to the Protocol on the Statute of the African Court of Justice and Human Rights’, 27 June 2014, AU Assemb. 25th Ord. Session.

77 Amnesty International, Malabo Protocol.

78 But the Malabo Protocol requires 15 ratifications to come into effect and none have been lodged.

79 Peter Fabricius, ‘Follow Me, I’m Right Behind You, Says Kenyatta’, Institute for Security Studies (4 February 2016), available at: {}.

80 African Union, ‘Withdrawal Strategy Document – Draft 2’ (12 January 2017), p. 2, available at: {}.

81 African Union, ‘Withdrawal Strategy Document’, p. 10.

82 Human Rights Watch, ‘AU’s “ICC Withdrawal Strategy” Less than Meets the Eye’ (1 February 2017), available at: {}.

83 Maddalena Procopio, ‘Reforms or Withdrawal? The Evolving Mosaic of Africa’s ICC Strategies’, Istituto Per Gli Studi Di Politica Internationale, n.d., available at: {}.

84 Acharya, ‘Norm subsidiarity’, 97.

85 Human Rights Watch, ‘Burundi: ICC Withdrawal Major Loss to Victims’ (27 October 2016), available at: {}.

86 Gabel, Katy, ‘Sudan: Facing Arrest, Bashir Stays Away from South Africa’, AllAfrica (7 May 2009), available at: { Google Scholar}.

87 Abraham, ‘Africa’s Evolving’.

88 Mills, Kurt, ‘Bashir in South Africa: Defeat, Victory or Both for International Criminal Justice?’, Justice in Conflict (2015), available at: { Google Scholar}. At a subsequent hearing before the ICC, it changed its rationale for not arresting Bashir. Allan Ngari, ‘Clutching at Straws: SA’s Reasons for not Arresting al-Bashir’ (18 April 2017), available at: {}.

89 Anne Borer, Tristan and Mills, Kurt, ‘Explaining post-apartheid South African human rights foreign policy: Unsettled identity and conflicting interests’, Journal of Human Rights, 10:1 (2011), pp. 7698 CrossRefGoogle Scholar.

90 Coalition for the International Criminal Court, ‘South Africa’s Shapeshifting Position on International Justice’ (21 March 2016), available at: {}.

91 ‘South Africa to Withdraw From War Crimes Court’, BBC News (21 October 2016), available at: {}.

92 Democratic Alliance v Minister of International Relations and Cooperation and Others (83145/2016) [2017] ZAGPPHC 53; 2017 (3) SA 212 (GP); [2017] 2 All SA 123 (GP) (22 February 2017), available at: {}.

93 ‘South Africa revokes ICC withdrawal’, Mail & Guardian (8 March 2017), available at: {}.

94 Coalition for the International Criminal Court, ‘South Africa Reverses ICC Withdrawal: Now Make International Justice Work For All’, (15 March 2017), available at: {}.

95 International Criminal Court, ‘Al Bashir Case: ICC Pre-Trial Chamber II Schedules a Hearing on South Africa’s Cooperation on 7 April 2017’ (8 December 2017), available at: {}.

96 Ngari, ‘Clutching at Straws’.

97 International Criminal Court, ICC-02/05-01/09, ‘Decision under article 87(7) of the Rome Statute of the International Criminal Court on the Non-Compliance by South Africa with the Request by the Court for the Arrest and Surrender of Omar Al-Bashir’ (6 July 2017).

98 du Plessis, Carien, ‘Justice minister withdraws bill that would take South Africa out of International Criminal Court’, Huffington Post South Africa (14 March 2017), available at: { Google Scholar}.

99 du Plessis, Carien, ‘ICC: At hearings, South Africa seeks clarity on Rome Statute’, Daily Maverick (11 April 2017), available at: { Google Scholar}.

100 African National Congress, International Relations Discussion Document, 5th National Policy Conference, 30 June–5 July 2017, pp. 4, 9, 12, available at: {}.

101 du Plessis, ‘ICC: At hearings, South Africa seeks clarity on Rome Statute’.

102 Zuma, Jacob, ‘The ANC must and will emerge from this policy conference stronger’, Daily Maverick (30 June 2017), available at: { Google Scholar}.

103 ‘Gambia Withdraws from International Criminal Court’, al Jazeera (26 October 2016), available at: {}.

104 ‘Amnesty Accuses Gambia Government of Brutal Repression’, al Jazeera (3 June 2016), available at: {}.

105 York, Geoffrey, ‘Gambian dictator suffers shocking defeat in election stunner’, Globe and Mail (2 December 2016), available at: { Google Scholar}.

106 Coalition for the International Criminal Court, ‘A Win For Justice: Gambia to Stay With the ICC’ (20 February 2017), available at: {}.

107 Kennedy, Merrit, ‘Under New Leader, Gambia Cancels Withdrawal From International Criminal Court’, NPR (14 February 2017), available at: { Google Scholar}.

108 Mueller, Susanne D., ‘Kenya and the International Criminal Court (ICC): Politics, the election and the law’, Journal of Eastern African Studies, 8:1 (2014), pp. 3233 CrossRefGoogle Scholar.

109 International Criminal Court, ‘Press Release: Ruto and Sang Case: ICC Trial Chamber V(A) Terminates the Case Without Prejudice to Re-Prosecution in Future’ (5 April 2016), available at: {}.

110 ‘Kenyan PM says Bashir must stand before ICC, wants apology made to int’l community’, Sudan Tribune (30 August 2010), available at: {}.

111 African Union, AU Doc. Assembly/AU/Dec.334(XVI), ‘Decision on the Implementation of the Decisions on the International Criminal Court, Doc. EX.CL/639(XVIII)’, adopted 31 Jan. 2011, AU Assembly, 16th Ord. Session.

112 Bashir Watch (undated), available at: {}.

113 Fabricius, ‘Follow Me’.

114 Nouwen, Sarah M. H. and Werner, Wouter G., ‘Doing justice to the political: the International Criminal Court in Uganda and Sudan’, European Journal of International Law, 21:4 (2011), pp. 947950 Google Scholar.

115 Mills, International Responses.

116 ‘African Union Moves’.

117 Bernard Momanyi, ‘Uganda Leader Lashes Out at ICC’, Institute for War Reporting (20 April 2013), available at: {}.

118 ‘Western Envoys in Uganda Walk Out of Museveni Swearing-in’, BBC News (12 May 2016), available at: {}.

119 Museveni was forced to accept that since Ongwen had committed crimes outside of Uganda, he needed to face international justice. ‘LRA Commander Dominic Ongwen “in Ugandan custody”’, BBC News (14 January 2015), available at: {}.

120 ‘Uganda Starts Process to Withdraw From ICC’, The Insider (7 October 2016), avaialable at: {}.

121 ‘Africa: Uganda Not Withdrawing from ICC – Attorney General’, AllAfrica (4 April 2017), available at: {}.

122 ‘Namibia pulls out of ICC’, The Herald (25 November 2015), available at: {}.

123 ‘Namibia to Withdraw from the International Criminal Court’, The Habari Network (18 March 2016), available at: {}.

124 ‘Namibia: Nam Supports Collective Withdrawal From ICC’, AllAfrica (9 February 2017), available at: {}.

125 Jeangène Vilmer, Jena-Baptiste, ‘The African Union and the International Criminal Court: Counteracting the crisis’, International Affairs, 92:6 (2016), p. 1323 Google Scholar.

126 Nompumelelo Sibalukhulu and Antoinette Louw, ‘The AU and the ICC Still Not the Best of Friends’, Institute for Security Studies (20 August 2010), available at: {}.

127 Lute, Aubrey, ‘Botswana adopts universal jurisdiction’, Weekend Post (17 July 2017), available at: { Google Scholar}.

128 Vilmer, ‘The African Union’, p. 1340.

129 Ibid., p. 1337.

130 Coalition for the ICC, ‘Annual ICC Assembly: States Hold Ground on ICC, but Serious Challenges Remain’ (5 December 2016), available at: {}.

131 Maupas, Stéphanie, ‘ICC African Protest Continues but Does not Spread’, JusticeInfo.Net (21 November 2016), available at: { Google Scholar}.

132 Zuma, ‘The ANC’.

133 Afghanistan, Columbia, Iraq, Palestine, and Ukraine.

134 Vilmer ‘The African Union’, p. 1320.

135 International Criminal Court, ‘Report of the Bureau on non-cooperation’ (8 November 2016), available at: {}.

136 Brunnée, Jutta and Toope, Stephen, ‘The Responsibility to Protect and the use of force: Binding legality?’, Global Responsibility to Protect, 2:3 (2010), p. 193 CrossRefGoogle Scholar.

137 For a broad typology of norms – as ‘proscribing’ (that is, ruling actions out), ‘prescribing’ (that is, requiring actions) and ‘permissive’ (allowing discretion) – see Wayne Sandholtz, ‘Multiple Paths to Norm Replacement’, paper presented at Unsteady Lives: the Dynamics of Norm Robustness workshop, Annual Meeting of the International Studies Association (Atlanta, 15 March 2016), p. 11.

138 McEvoy, Janet, ‘The staying power of African Presidents’, Africa Review (27 October 2015), available at: { Google Scholar}.

139 Risse, ‘“Let’s argue!”’, p. 7.

140 Ibid., pp. 8–9.

Figure 0

Figure 1 The norm dynamics role-spectrum.