2.1 Classic Rationalist and Integration Theories

Theories that seek to explain intergovernmental cooperation and organization have long been a mainstay in IR research. Classic rationalist theories are characterized by an understanding of international politics as being driven by states as rational decision-makers in pursuit of national self-interests in an anarchic international system. Rationalist theories have in common that they assume states to pursue some type of public good, such as security or economic welfare (Katzenstein, Keohane, and Krasner 1998), and cooperate with each other if the expected utility outweighs the costs. Rationalist theories which emphasize absolute over relative gains view cooperation as a logical step to solve transnational problems that defy national boundaries (Keohane 1984; Abbott and Snidal 1998). To achieve effective cooperation under conditions of anarchy, states often institutionalize their cooperation by creating international organizations (IOs). Abbott and Snidal (1998) argue that in addition to the functions performed by international regimes, IOs offer cooperative benefits by virtue of being more centralized and independent. Thus, “states are able to achieve goals that they cannot accomplish on a decentralized basis” (1998, 29).

Classic integration theories, such as neofunctionalism (Haas 1977; Schmitter 2005) and intergovernmentalism (Hoffmann 1966; Moravcsik 1993), move down to the regional level and seek to explain regional integration. In line with rational institutionalist theories, these theories take interdependence as the impetus for states to integrate at the regional level even though they differ, most notably, in the actors they emphasize (cf. Börzel 2016). Despite these distinctive features, both theories draw from, and focus on, European integration. Integration theories typically view national sovereignty as a resource that states are prepared to cede, if not enthusiastically, for the actualization of mutual benefits. Moreover, they often imply pluralistic democratic states in and across which the formulation of national interests takes place in order to provide certain public goods. In this regard, classic integration theories imagine national interests and mutual benefits to be primarily economic in nature, and cooperation in matters of ‘high-politics’, such as security, is considered less likely or at least lagging economic cooperation. Especially outside of Europe, however, common security threats among states are salient. Nowadays, scholars with a functionalist perspective increasingly acknowledge that security interdependence may drive regionalism and “prompt states to create or resuscitate regional organizations” (Legrenzi and Lawson 2018, 692). According to this perspective, security is considered one of the salient functional demands that spur regionalism (Kacowicz and Press-Barnathan 2016; Börzel and Risse 2019). This functional demand is undoubtedly present in Africa.

2.2 Sovereignty-/Regime-Boosting Theories

Over time, the dominance of classic rationalist and integration theories became increasingly challenged. With the aim of understanding African security regionalism, the theoretical currents of sovereignty- and regime-boosting regionalism are of particular importance. These currents draw from experiences outside the European context. Thereby, they challenged common Eurocentric perceptions of regionalism and expanded our horizon with respect to the reasons why states choose to pursue regionalism. Importantly, these conceptions of regionalism are neither inherently ‘irrational’, in the sense of actors not being strategic utility maximizers, nor are they necessarily driven by anything but ‘functional’ demands (see Stoddard 2015). Instead, they shift the reference point (in whose interests?) and expand on the purposes behind regionalism (to which ends?). While sovereignty-boosting and regime-boosting has been used interchangeably at times (see Söderbaum 2011, 76), I argue that it makes sense to draw an analytical distinction (see also Vinokurov and Libman 2017, 22 for this distinction, albeit with a slightly different emphasis). Strictly speaking, sovereignty-boosting introduces a new purpose of regionalism, whereas regime-boosting – which is also the focus of the authoritarian regionalism literature – emphasizes a different beneficiary of regionalism. While there is undoubtedly a large intersection in practice, they are nonetheless conceptually distinct.

The term ‘sovereignty-boosting regionalism’ was first introduced by Söderbaum (2004) as one of the modes of governance in Africa. He explicitly challenges the prevalent idea that states must choose between pursuing regional integration and keeping their national sovereignty intact (2004, 425). Instead, national sovereignty is even introduced as a new purpose of regionalism. As Acharya (2016, 117) notes, “the foundational goal of most regional groups in the post-colonial non-Western world was autonomy, defined as the protection and preservation of state sovereignty.” Besides de jure sovereignty, the literature on weak postcolonial states emphasizes their quest for ‘empirical statehood’ (see Jackson and Rosberg 1982). While particularly relevant to the case of Africa, similar concerns about the empirical dimension of sovereignty can also be found in regions such as the Asia-Pacific (Narine 2004) and the Middle East (Barnett and Solingen 2007). I distinguish two central ways in which sovereignty is ‘boosted’. First, sovereignty-boosting regionalism typically entails the commitment to the norms of non-interference and territorial integrity among members and against external interference (passive support). Second, sovereignty-boosting regionalism may entail the commitment to provide support – military or otherwise (financial, logistical, rhetorical etc.) – to recognized national governments who seek to (re-)establish or maintain control over their own territory (active support). While both passive and active support likely have regime-boosting effects, the prime concern under this logic is to protect the region from negative externalities and general instability that frequently emanate from fragile or weak states.

Regime-boosting regionalism denotes a type of regionalism that primarily serves the interests of incumbent regimes. This stands in contrast to classic IR and integration theories which presume the pursuit of broader ‘national’ interests. Thus, regime-boosting theories introduce a new reference point. Leading in this regard is the literature on authoritarian regionalism (Libman and Obydenkova 2018; Obydenkova and Libman 2019; Debre 2021). At its core stands the question why authoritarian regimes push for or partake in regionalist projects. Centrally, the authoritarian regionalism literature challenges the idea that (meaningful) regionalism is limited to regions with democratic and pluralistic states. Instead, the literature contends that regionalism is attractive to authoritarian regimes due to the benefits it produces for them. The impact of ROs may be either direct (‘autocracy promotion’) or indirect (‘autocracy diffusion’) (Obydenkova and Libman 2019, 59; see also Libman 2015, 134–36). Importantly, regime-boosting as a phenomenon is not inherently limited to ROs composed solely of autocracies (cf. Stoddard 2017; Agostinis and Closa 2022). Indeed, the beneficiary must not be an authoritarian regime in the first place. In line with the logic behind sovereignty-boosting regionalism, governments in weak states as well as weak governments – irrespective of the level of democratization in the country – have an inherent incentive to use ROs to pursue their parochial interests (Taylor and Söderbaum 2016, 139–40). This is especially relevant under conditions of neopatrimonialism, in which political power and control over economic resources are intrinsically linked.

2.3 Towards a Synthesized Theory of Intervention

This section draws on the two outlined logics of cooperation and theorizes how their parallel presence impacts the nature and design of AU interventions. I define interventions broadly as diplomatic efforts determined through formal resolutions and decisions to address a crisis or conflict involving at least one member state. I take off from the observation that the AU’s institutional design reflects a type of security regionalism that genuinely seeks to deal with security interdependence and regional instability, and to address transnational security problems. This roughly follows the ideas of classic rationalist and functionalist IR theories. However, theories around sovereignty and regime-boosting flagged the importance of two characteristics found in contemporary regionalism. First, boosting sovereignty can be a purpose of regionalism, and sovereignty is not merely a resource states trade to reap other (often economic) benefits of regional cooperation. Second, incumbent regimes are a key reference point, and boosting their rule may trump the pursuit of broader national interests in regional cooperation. This has implications for the behavior of the AU and the nature and design of its interventions, which has thus far not been properly theorized.

In the following, I outline three main theoretical expectations regarding AU interventions that we should observe in practice in the context of an African hybrid regionalism that continuously balances the two central logics of cooperation.³

First, regarding their ‘prevalence’, we should expect African hybrid regionalism to lead to (1) frequent interventions and (2) the use of diverse diplomatic tools, including coercive ones. Two distinct but interrelated sources contribute to this. On the one hand, there is African security regionalism. African ROs were transformed to be active managers of security matters on the continent. The comparatively high frequency of violent conflicts and the great diversity of security threats implies that numerous situations require the attention of the AU, among others, and spark regional reactions. This requires a certain degree of flexibility to make use of context-appropriate diplomatic tools. Different types of security threats call for different approaches to deal with them. To this end, the AU has modified its foundational treaties and claimed a key role in the prevention, management and resolution of conflicts and crises in Africa (African Union 2002). First and foremost, this entailed the establishment of the PSC with considerable discretion to decide on when and how to intervene (Art. 7r, 9.1 of the PSC Protocol). On the other hand, African security regionalism is complemented by a regime-boosting form of regionalism. This adds to the list of instances that draw in the AU, as even intrastate issues may be considered relevant. The partiality, in favor of the incumbent regime, also means that intervention types beyond mediation become increasingly relevant. This is because ‘antagonistic’ interventions enable the AU to act aggressively towards and exert pressure on challengers of incumbents. This differs from the dominant view within the OAU, held at least for the first decades, which was that “conflicts within States fell within the exclusive competence of the States concerned” as noted by former director of the OAU’s political affairs department, Sam Ibok (quoted in Powell 2005, 10) – not only when the incumbent regime committed atrocities and other violations of international norms, but also when the regime itself was threatened or overthrown by domestic opponents (cf. Udombana 2002, 1208–19; Ibrahim 2012, 37–41). The AU has demonstratively renounced this view. What is more, the threat that intrastate conflicts or crises spill over and produce regional instability is higher in weaker states (cf. Obamamoye 2019), often rendering regional and regime security two sides of the same coin.

Second, regarding the ‘context’, we should expect (1) interventions to be triggered by a wide array of norm violations and (2) the strongest interventions to take place when problem-solving and regime-serving logics converge. African security regionalism is broad in its scope, not least because conflicts and crises in Africa are notoriously prone to exceed state boundaries and become regionalized, which necessitates regional solutions and prompts regional responses (Keller 1997, 300; Söderbaum and Hettne 2010, 19). The African continent is also increasingly confronted with non-traditional security challenges such as failing states, terrorism and transnational crime which broadens the security agenda (cf. Emerson and Solomon 2018). Additionally, regional actors acknowledge the inherent link between security and the adherence to good governance norms such as democratic practices, the rule of law, and respect for human rights (see Art. 3f of the PSC Protocol). Hence, these issues require collective attention and likely become points of emphasis under problem-solving regionalism. At the same time, threats to incumbent regimes and their claims to sovereignty are relevant in accordance with a regime-serving logic of cooperation even if these threats remain rather localized within a country’s internationally recognized borders. This adds to the instances in which the AU feels compelled to intervene, even though it is still incentivized to frame those interventions as responses to violations of widely accepted norms instead of blunt efforts to protect or otherwise benefit national leaders. The strongest and most coercive types of interventions are likely to take place when a regional security concern coincides with a challenge to an incumbent regime or their claims to juridical, or exercise of empirical, sovereignty as both logics of cooperation interact. This includes cases of armed rebellions, especially in weak states, and terrorist groups operating across borders. On the opposite end of the spectrum, interventions are likely to be weaker or not present at all when there is neither a transnational problem to solve (i.e. in primarily domestic contexts) nor a reason to support the incumbent regime (i.e. the regime is not threatened, and might even reject external involvement).

Third, regarding their ‘targets’, we should expect interventions to (1) frequently target groups other than the incumbent regime, and (2) seldom openly target incumbent regimes. The first part refers to coercive interventions being considerably more likely to exert targeted pressure on non-regime actors. This includes rebel groups and criminal or terrorist networks, operating nationally or transnationally, but potentially also external state and non-state actors. The actions of these actors challenge the incumbent regime’s claim to sovereignty or threaten its survival directly. Targeting these actors through interventions largely follows the regime-serving logic as boosting incumbent regimes often implies rallying against its challengers, rhetorically or otherwise. The second part is an expression of hybrid regionalism as it reflects both logics of cooperation. Violations of norms and threats to regional security may compel the AU to intervene, but the supportive environment for incumbent regimes renders a head-on challenge of the regime unlikely even when the regime contributes to those violations or threats. Such reluctance could be seen during the days of the OAU, in which incumbent regimes cooperated for their own benefit and refused to hold each other accountable for even the most egregious norm violations (Ibrahim 2012, 39). As a result, ambiguous target references, in which incumbents are not called out explicitly but potentially implied, are to be expected when such inherently contentious instances come up.

3.1 Data and Method

To this end, I focus on the AU PSC’s coercive interventions. Coercive interventions, as herein defined, seek to alter the behavior of targeted groups through the application of pressure or the use of force. The focus on coercive interventions is justified for two reasons. First, they constitute partial interference in an ongoing conflict. This makes them not only potentially more contentious and intrusive, but also requires the AU to take a clear stand and reveal its priorities. Second, legal provisions for coercive intervention types are part of what separates the AU from the OAU. The shift is exemplary for a new type of regionalism, one with a broader security mandate and the right to become involved in ostensibly domestic affairs, both as a possible check on regimes as well as a boost for their survival and exercise of sovereignty. I differentiate three broad categories, namely military operations, sanctions, and forceful rhetoric. These are the key diplomatic tools that are used to apply pressure on conflict actors. The PSC is chosen due to its primary agency in dealing with security matters on the continent, its high level of activity, and its broad mandate and authority.

This paper employs content analysis. Content analysis refers to “a systematic, replicable technique for compressing many words of text into fewer content categories based on explicit rules of coding” (Stemler 2000, 1). I identify instances of coercive diplomatic interventions with the help of a coding scheme that defines the activation of various diplomatic tools, including – from stronger to weaker – the authorization of military deployment; the imposition and threat of targeted sanctions, suspensions, embargos, and other punitive consequences; and instances of public condemnation, as well as behavioral demands directed at conflict parties. In an in-depth coding exercise, I applied this coding scheme and analyzed close to 1,000 Communiqués and Press Releases issued by the PSC between 2005 (its first full operational year) and 2021. For each instance of coercive intervention, I coded its type, proclaimed subject or trigger, declared target, and additional context information such as the date, the affected countries, and normative justifications where provided. It bears mentioning that this approach only captures positive cases, that is, those in which an intervention takes place. Hence, this paper does not comprehensively analyze the conditions under which an intervention is more or less likely to take place. Instead, it derives insights from the positive cases to illustrate the influence the AU’s hybrid regionalism has on its coercive interventions. This is done by looking at the prevalence, context, and targets of military operations, sanctions, and forceful rhetoric. The following section dives into the AU’s coercive interventions and outlines the results of this analysis along the three main theoretical expectations.

3.2 Prevalence of Coercive Interventions

First, interventions were hypothesized to be frequent and use a diverse set of diplomatic tools under African hybrid regionalism due to the convergence of problem-solving and regime-serving logics of cooperation. I identify three overarching types of coercive interventions by IOs under which individual interventions can be subsumed. Those are military operations, sanctions, and forceful rhetoric. The AU and its PSC have provisions that allow for the use of various coercive intervention types. The AU has the authority to conduct peace and other military operations under Articles 4(h) and 4(j) of the CA, and Article 7.1c of the PSC Protocol. Regarding sanctions, Article 23.2 of the CA provides the AU with considerable discretion to impose sanctions for failure “to comply with the decisions and policies of the Union.” Through the PSC Protocol, the PSC is only explicitly tasked with imposing sanctions in cases of UCGs (Art. 7.1g) but does possess the discretion to “take initiatives and action it deems appropriate” in executing its mandate to prevent and resolve conflicts (Art. 9.1). Forceful rhetoric, as the third type, is not explicitly part of the mandate but practiced by many IOs around the globe (Squatrito, Lundgren, and Sommerer 2019). Importantly, the AU not only established frameworks that allow for various types of interventions but actively uses its coercive diplomatic tools.

The AU has conducted multiple peace operations, including most notably the missions in Somalia (AMISOM, 2007 to 2022), Sudan (AMIS, 2004 to 2007), Central Africa (MISCA/AFISM-CAR, 2013 to 2014) and Burundi (AMIB, 2003 to 2004). In the Comoros, the AU authorized a special military operation and deployment of security forces to the island of Anjouan. Moreover, the AU has undertaken a joint peace operation with the UN to Darfur (UNAMID) which replaced AMIS in 2007. Indeed, the AU has developed into a peace operations partner for the UN with an established division of labor (Williams and Boutellis 2014; de Coning 2017). Lastly, the AU has formally authorized or backed various peace operations by other African ROs as the de jure primary security RO on the continent, including an ECOWAS mission in Mali (AFISMA), an SADC mission in Mozambique (SAMIM), and an IGAD mission in Somalia (IGASOM). This is in addition to repeatedly authorizing cross-border military operations against the LRA in the Great Lakes region (RCI-LRA), Boko Haram in the Lake Chad basin (MNJTF), and violent extremists in the Sahel region (G5-Sahel Force). This level of involvement in multilateral military and peace operations is exceptionally high among ROs (for reference see, for instance, the MILINDA dataset by Jetschke and Schlipphak 2020).

Unlike some other ROs, the AU’s treaties do envision the use of sanctions. The affirmative attitude can be traced back African historical experiences with apartheid (Hellquist 2014). In practice, the AU has suspended members and imposed sanctions against domestic groups on multiple occasions. I identified over 50 instances in which the PSC publicly imposed, extended, or strengthened sanctions – with each type of sanction counted separately – on 15 different member states across 31 country-years. Instances in which the PSC threatened imposition of sanctions, or unspecified punitive measures, are even more frequent. Moreover, the PSC occasionally calls on other actors – mostly the UNSC – to impose or consider imposing sanctions. This further underscores the AU’s propensity to use or threaten coercive measures in reaction to perceived norm violations.

A recent study by Squatrito, Lundgren, and Sommerer (2019) analyzed condemnatory speech acts between 1980 and 2015 across 27 IOs. According to this study, the (O)AU ranks fourth in the use of condemnatory speech acts with a total of 63. However, this relatively high ranking still underestimates the prevalence of naming and shaming in AU documents, as the study only looks at public documents from the AU Assembly and not from the more active PSC. Public condemnations are highly frequent in PSC communiqués. The PSC issued public condemnations more than 400 times between 2005 and 2021, or around 25 condemnations per year. Behavioral demands are less frequent but nonetheless common features in PSC communiqués with slightly over 150 instances. Consequently, forceful rhetoric is by far the most used coercive diplomatic tool of the PSC.

In summary, coercive interventions of various types are used frequently by the AU. This is in line with the corresponding expectation that the AU’s form of hybrid regionalism incentivizes interventions due to being driven by two logics of cooperation. While the level of frequency and diversity in tools alone do not prove that these interventions are driven by the combination of the two logics that underpin the AU’s hybrid regionalism, they provide a first hint to this end that justifies further inquiry.

3.3 Context of Coercive Interventions

Second, interventions were hypothesized to be a response to a wide array of norm violations and the strongest when problem-solving and regime-serving logics converge. The strongest coercive interventions that are at the AU’s disposal are military operations, followed by sanctions and forceful rhetoric.

Military operations under AU leadership occur in some of the worst crises that destabilize the entire region. The conflicts in Somalia, Sudan, South Sudan, and the CAR, which triggered AU-led peace operations, are among those with the highest numbers of refugees since the AU came into being. At the same time, these countries tend to lack empirical sovereignty with a monopoly on the use of force across their territory. These armed conflicts usually challenge both statehood and incumbent regimes. For this reason, the concept of ‘stabilization’ has become a staple of AU peace operations (cf. Dersso 2016; de Coning 2017; Moe and Geis 2020). The fact that the AU intervenes in some of the most severe crises on the continent with a ‘stabilization’ mandate, which frequently includes the restoration of state authority in weaker states (African Union 2007b; 2013c) or, as done in the case of Somalia, very open and explicit support for the incumbent regime (African Union 2013b), underscores that problem-solving and regime-serving logics operate hand in hand. Still, military operations conducted or authorized by the AU also regularly refer to widely accepted humanitarian norms and include mandates to protect civilians, promote human rights, or facilitate humanitarian relief. In one exceptional case, the AU approved military intervention to restore government authority in a conflict with limited negative externalities with reference to democratic norms (African Union 2007c). This mission was in support of the efforts of the central government of the Comoros to regain control over one of its islands, Anjouan, after the island’s regional government declared independence. Regarding transnational conflicts driven by armed groups in the Sahel, the Lake Chad basin and the Great Lakes region, the AU continues to demonstrate its willingness to authorize military operations initiated and led by the incumbent regimes of the involved countries to deal with regional security threats (African Union 2011; 2015; 2017).

The imposition of sanctions is more prevalent in cases of UCGs, especially around coups d’état. Various agreements and documents depict UCGs as a threat to regional security. The strict UCG policy itself has its roots in the 1990s, during which member states increasingly realized that prevalence of coups threatens peace and security in Africa (Organization of African Unity 2000). At the same time, the primary form of UCGs that is of concern to the AU are coups d’état which, by their very nature, threaten incumbent regimes. Even when the AU does not demand that an overthrown regime is reinstated, the strict UCG policy on coups d’état indirectly serves the broader community of incumbent regimes through desired deterrence effects. Typically, the AU seeks a return to what it calls ‘constitutional order’ when it imposes sanctions. This links back to the norms of constitutionality, democracy, good governance, and the rule of law that are enshrined, among others, in the African Charter on Democracy, Elections and Good Governance (ACDEG) (African Union 2007a). Beyond this, sanctions and threats are only sporadically used in lower-intensity conflicts that threaten to (re-)ignite such as in Burundi in 2015 and Guinea Bissau in 2018. In this regard, it is common for the PSC to use the threat – and to a lesser extent imposition – of sanctions in the aftermath of brokered transition, ceasefire, or other types of peace agreements to compel violators into compliance and avoid a resurgence of violence. Warnings of consequences are quite frequent in cases of attacks on civilians or human rights violations. Surprisingly, this is not the case when it comes to the imposition or threat of sanctions, where invocations of humanitarian norms are comparatively rare albeit not entirely absent.

Forceful rhetoric, as the lowest form of coercion, is used in a much broader range of contexts and not largely limited to UCGs or the most severe conflicts. This is unsurprising as it is a comparatively cheap diplomatic tool that can be quickly and effortlessly used to publicly position oneself in a conflict situation and respond to norm violations. The fact that public condemnations can be found in country-specific communiqués dealing with security issues in 29 different countries, or more than half of all AU member states, speaks to this idea as it suggests a great variety of contexts in which this comparatively less coercive tool is used. It is notable that violence against civilians, including attacks and human rights violations, is the most frequent action condemned by the PSC with more than 200 statements. This is followed by condemnations of attacks against the state, including government officials, security forces, and military or civilian infrastructure at around half that rate. Demands also often link to violence against different entities but are, in relative terms, more likely to demand adherence to agreements or constructive engagement with peace processes. Rejection of external interference is common albeit not dominant for either type of forceful rhetoric. Explicit reasons as to why condemned or rejected behavior is harmful vary from civilian suffering to the threat of conflict escalation, national and regional instability, and norm violations. Curiously, sovereignty infringement is rarely referenced as an explicit reason for this type of intervention.

In conclusion, the expectation that interventions are triggered by a wide array of norm violations holds true. At least in terms of public presentation, the AU has intervened due to violations of democratic and constitutionality principles, human rights violations and violence against civilians, violence against the state or international actors, breaches of agreements and external interference. It has also intervened in cases in which national or regional stability and the incumbent regime were threatened. Moreover, the findings support the expectation that the strongest measures – military operations – are more frequently present in conflicts that combine regional security implications with threats to incumbent regimes and their claims to sovereignty. UCGs, for their part, constitute an obvious threat to incumbent regimes but are also perceived and framed as a regional security threat. Hence, the fact that most sanctions are imposed in this context also fits the hybrid regionalism paradigm. The observation that less coercive measures, in turn, are found in a broader variety of contexts only lends further credence to the idea that those crises which threaten regional security as well as incumbent regimes and their claims to sovereignty are most likely to unify member states and produce the strongest responses.

3.4 Targets of Coercive Interventions

Third, interventions were hypothesized to frequently target groups other than the incumbent regime, and seldom openly target incumbent regimes.

Regarding military operations, the AU operates with the consent of the host government. For this reason, they do not target the incumbent regime directly or even challenge it through other means. Article 4(h) of the CA, the only way to intervene without host state consent, has yet to be invoked. The PSC only threatened to recommend invoking Article 4(h) to the Assembly once, in the context of the Burundian crisis in 2015, which sparked bellicose rhetoric from the Burundian authorities in response and concerns among various member states, and ultimately saw the PSC walk back from its threat (cf. McCormick 2016). The Burundian case is exceptional and effectively ended with a strong affirmation of norms that protect incumbent regimes and their claims to sovereignty. Instead, AU military operations are commonly designed to target criminal and terrorist networks, and increasingly do so by name, and at times explicitly mandate their elimination or neutralization such as in the Sahel and the CAR (African Union 2014; 2017). Host states, and thereby incumbent regimes, also benefit from AU efforts to enhance the capabilities of national institutions and train national security forces (African Union 2007b; 2012; 2013c). As such, AU-led and AU-authorized military operations by and large target armed groups opposed to the incumbent regime or established state structures, and benefit incumbent regimes through the efforts to enhance empirical statehood in the short and long term.

Similarly, the AU imposes or threatens sanctions directly against the recognized incumbent regime and its country only in extraordinary circumstances. In fact, the PSC itself has never explicitly imposed sanctions in this context, but merely supported targeted sanctions and an embargo against Eritrea in 2010 (African Union 2010), and an embargo against Côte d’Ivoire in 2005 (African Union 2005). In both cases, the PSC followed the lead of the UNSC. At the same time, sanctions are most frequently imposed on ‘rogue’ actors whose claim to government is not recognized in the region. This includes those that came into power through a coup, to which the PSC refers as ‘de facto authorities’, and their loyal security forces. Concerning armed groups, including rebels and terrorist networks, the PSC is more likely to threaten than impose sanctions, which can be explained by unofficial division of labor practices with the UNSC. This is notwithstanding the observation that most threats issued by the PSC lack specification and address all groups that engage in sanctionable behavior or blanketly refer to all parties to a conflict. This is especially the case when it comes to violations of negotiated peace or truce agreements and, to a lesser extent, violence against civilians.

A lack of specificity continues to be an important theme for forceful rhetoric, but the range of targets broadens. As is the case for other coercive means, incumbent regimes are rarely openly named as the targets of condemnations. For demands, however, there is a clear increase in the number of statements that can be identified as also being directed at the incumbent regime that is part of a conflict. This notwithstanding, condemnations that exclusively target armed groups, by name or generic reference, occur at a ratio of roughly 15 to 1 when compared to condemnations that explicitly and exclusively targets incumbent regimes, including through references to parts of the government as well as generic references to the state. This lends further credence to the notion that government actors are comparatively insulated from public naming and shaming diplomacy. External actors are occasionally condemned for their actions. However, with just slightly more than one such condemnation per year on average, this theme is perhaps less common than one might expect considering Africa’s historical experiences with external interference.

Overall, it becomes strikingly obvious that incumbent regimes are almost fully insulated from being openly targeted through coercive interventions. They may, however, be implied in rhetoric that references ‘all actors’ and ‘the conflict parties’, especially when it comes to non-adherence of agreements or violence against civilians. This reflects a concern for the public good that is not primarily driven by a regime-serving logic even though this logic constrains the public messaging around the misbehavior of regime actors. Conversely, opponents of incumbent regimes are regularly, strongly, and unequivocally targeted even as much of the less coercive rhetorical posturing remains vague or ambiguous. The findings are generally in line with the outlined expectation regarding the targets of interventions under African hybrid regionalism. It bears mentioning, however, that the AU does not target external state actors, or IOs, by name when criticizing external interference. Instead, the language used in communiqués remains either below the threshold for forceful rhetoric or, when such rhetoric is used, the targets remain vague with references to ‘foreign elements’ or ‘non-African states’. This could hint at the AU’s keenness on being on good terms with current or future regional, interregional, or global partners as it strives to solve Africa’s problems.