Legitimation struggles in international organizations: the case of the African Union

ABSTRACT How do international organizations (IOs) and their proponents claim legitimacy, and how do their opponents undermine such legitimacy? This article develops a framework that accounts for the links between legitimation and delegitimation strategies and how they regularly produce ‘legitimation struggles’. Drawing on the case of the African Union between 2015–2020, the study goes beyond existing research in three ways. First, legitimation struggles are not simply related to input and output legitimacy but are deeply related to the social purpose of the organization. Second, legitimation struggles do not only involve IO representatives and member-states but are strengthened by a range of other non-state agents. Third, while discursive strategies are essential, legitimation struggles are reinforced when they are combined with behavioural or institutional legitimation strategies. Future research would do well to go beyond the current Western-centric bias and draw on our findings to investigate legitimation struggles under different conditions around the world.


Introduction
It is widely believed within the scholarly community that legitimacy matters for international political institutions (Zürn, 2018).An influential team of authors argue that global governance institutions 'are more likely to operate smoothly if they enjoy legitimacy … Without legitimacy, authority is likely undermined or must depend on coercion, secrecy, and trickery to obtain sway and governance is often less effective as a result' (Tallberg et al., 2018a, p. 3).
While there is a rich literature on the perceptions of audiences that an IO's authority is normatively appropriate (i.e.legitimacy), this study shifts the focus to the processes and strategies of justification and critique intended to shape such beliefs (i.e.legitimation and delegitimation) (Bexell et al., 2022;Lenz & Söderbaum, 2023;Tallberg & Zürn, 2019, p. 583).Although the literature on legitimation has grown in recent years, existing scholarship focuses heavily on a rather restricted set of normative standards and discursive strategies as enacted by a specific set of agents, especially IO staff and their member-states (see Ecker-Ehrhardt, 2018;Gronau & Schmidtke, 2016;von Billerbeck, 2020).As a result, scholarship on legitimation strategies remains characterized by a certain selectiveness that prevents scholars from recognizing how different legitimation and delegitimation standards and strategies often reinforce each other and develop into 'legitimation struggles'i.e. the varied, competitive and sometimes contradictory strategies that different agents employ to justify or contest the authority of IOs vis-à-vis different target audiences (Tallberg et al., 2018a, p. 5).
Apart from the selective focus on discursive self-legitimation by IO staff and member-states, the IR literature is also often characterized by a narrow empirical selection.In fact, most IR research remains concentrated to a rather limited number of well-established IOs dominated by the North and West, such as the UN, WTO, World Bank and the EU.Even if studies of IOs from the non-Western world certainly exist, they are often treated as unique or exceptional, and hardly ever impact theorizing in the field.Hence, there are relatively few studies within IR about how legitimation struggles play out in different non-Western contexts, and whether generalizable mechanisms are at work.
Our general interest in legitimation struggles is matched by a specific interest in the African Union (AU).What is puzzling with the AU from a legitimation perspective is that when the long-term pan-African development plan known as Agenda 2063 was agreed upon in the mid-2010s, it was not only associated with a range of overlapping (self-)legitimation attempts by leading proponents, but it was also met with contestation from a wide range of actors even from the AU's own staff and several influential member-states.Although there is a comprehensive literature on the politics of the AU (Tieku, 2021a) only a handful deal with legitimation (see Gelot, 2012;Lotze, 2013;Mickler & Sturman, 2021;Witt, 2019), and none focus on what we refer to as legitimation struggles.
The case selection of the AU helps to moving towards a post-Western research agenda on IO legitimation strategies, which takes us beyond the narrow set of established IOs dominated by the West that receives most emphasis in existing research (e.g.EU, UN, WTO, the World Bank).As the continental IO with most far-reaching authority, the AU is arguably the most relevant case in Africa.In contributing towards a more 'global' and generally applicable research agenda, we classify the AU in the same category as other general-purpose regional IOs, such as the EU, ASEAN, and Mercosur (see Dingwerth et al., 2019;Hooghe et al., 2019).Although somewhat more complicated from a methodological point of view (Döring & Herpolsheimer, 2021), we also see scope for comparison with task-specific IOs, particularly those that are entrusted with a certain degree of authority, such as the International Criminal Court, the WHO, and the WTO.
The AU has some traits that distinguishes it from other comparable IOs, at least outside the African continent.For instance, the AU's legitimation pattern is influenced by Africa's place in an unequal world, post-colonial relationships that shape the African state-society complex as well as the AU's contentious performance (Mickler & Sturman, 2021).For these reasons, external powers, agencies and donors have been more important within the AU compared to 'global' IOs with a more universal membership or 'regional' IOs in the West.However, in line with recent literature on IOs within IR, we reject generalized claims about the AU's 'uniqueness', 'exceptionalism', or indeed a purported lesser importance of non-Western IOs for the advancement of general theory in the field of IO legitimacy and legitimation (Bexell et al., 2022;Lenz & Schmidtke, 2023;Tallberg et al., 2018b).
In order to grasp the legitimation struggle surrounding the pursuit of institutional reforms in the AU between 2015 and 2020, we synthesize existing strands of (de)legitimation research and propose a set of analytical tools that we then use for empirical purposes.This study departs from a recent insight within the research field, namely that legitimation and delegitimation are interlinked and therefore need to be brought together within the same framework (Bexell et al., 2022;Zürn, 2018).Without delegitimation, resistance and critique, there is no genuine 'struggle', only a series of overlapping (self-)legitimation attempts that may go in different directions.More specifically, we transcend the narrow focus of previous research on a selective set of legitimation standards, agents and strategies.Drawing on different strands within previous research we develop a framework that allows us to analyze how a rather diverse set of agents legitimate and delegitimate the AU for different and often competiting purposes and with various types of discursive as well as non-discursive strategies.These insights can inform future investigation of legitimation struggles under different conditions and in both Western and non-Western IOs.
The article is organized as follows.In the next section we review and synthesize previous research into our three-tiered framework (i.e.normative standards, agents, and strategies).The third section provides additional details about our research design and methodological considerations.Thereafter, the empirical analysis is structured in two parts, Pan-African Unity and AU Fit-For-Purpose, in order to trace and analyze the legitimation struggle in the AU during the time period 2015-2020.A conclusion summarizes the main findings and makes the case for future research to investigate legitimation struggles in other IOs and under different conditions.

Previous research and framework
In spite of a rich literature on the sources and effects of both normative and sociological conceptions of the legitimacy of political institutions, much less is known about legitimation strategies.To the extent that legitimation is considered at all, most IR literature focuses predominantly on the discursive (self-)legitimation strategies by IO representatives and their member-states in a rather limited set of IOs dominated by the West (Gronau & Schmidtke, 2016;von Billerbeck, 2020).
We concur with recent studies that argue for the need to go beyond predominantly discursive legitimation strategies and also consider the fact that many IOs are facing increasing critique and contestation, which creates pressure on them to legitimate their rule with different types of discursive, institutional as well as behavorial strategies and against increasingly diverse audiences (Bexell et al., 2022;Dingwerth et al., 2019).We also take inspiration from recent literature that there is, rarely, if ever, widespread agreement about what constitutes appropriate rule for an IO, and claims to authority are typically met with resistance, protest and critique (Anderl et al., 2019;Zürn, 2018).We need therefore to analyze how legitimizers and delegitimizers respond to each other (Uhlin, 2019, p. 10).Differently expressed, legitimation and delegitimation need to be integrated within the same framework.The symbiotic relationship between legitimation and delegitimation is seen in a growing literature on legitimation struggles under diverse conditions, as seen by recent concepts such as 'legitimation contests' (Dingwerth et al., 2019), 'legitimacy struggles' (Uhlin, 2019) and performative 'battles for legitimacy' (Wajner, 2019).
It can be mentioned that the discussion of 'struggle' may also speak to questions around stigmatization.Insofar as delegitimation concerns actors responding to or challenging claims about what counts as appropriate community norms, this research inquiry resonates with work on stigmatization processes in international politics. 1In fact stigmatization can be considered as a particularly severe mode of delegitimation that denotes an onset of legitimacy crises in that stigmatized actors are isolated or estranged from a political community (Adler-Nissen, 2014;Zarakol, 2011).
Although the aforementioned studies on assorted legimitation struggles are both theoretically and methodologically advanced, previous research usually suffer from one or several delimitations, which our framework seeks to overcome.First, several previous studies usually focus on overlapping legitimation strategies and much less on delegitimation strategies, which means that they do not focus as much on how legitimation and delegitimation coevolve.Such analysis require a consideration of the different normative standards of legitimation and legitimation.Second, previous research on legitimation often rely on a rather restrictive selection of agents, particularly IO representatives or member-states.Unless non-state actors are included in the analysis, these studies offer few insights to how non-state actors enact both legitimation and delegitimation strategies (for exceptions, see Gregoratti and Uhlin, 2018;Bexell et al., 2022).Third, there is still a strong emphasis in this literature on discursive rather than non-discursive (de)legitimation strategies.In response to these weaknesses in previous reseach, we develop a framework consisting of three components that are relevant for all legitimation struggles under diverse conditions: (i) the normative foundations of struggle, (ii) agents, and (iii) the types of strategies.

Normative standards of legitimation and delegitimation
The first component of the framework concerns the normative substance upon which agents form views about standards of legitimation and delegitimation.IOs are held up to a diversity of legitimation standards or what is sometimes referred to as the sources of legitimacy (Scholte & Tallberg, 2018). 2 The literature on the sources of legitimacy is heavily indebted to Sharpf's distinction between input and output legitimacy (Scharpf, 1999).Scholarship in IR has built on this distinction and focused on normative standards and justifications related to the quality of decision-making procedures and the outputs of policy-making and institutional performance.Scholte and Tallberg (2018, p. 62) add complexity to the discussion by distinguishing three additional generic qualititesdemocratic credentials, technocratic standards, and fairness of a governing institutionthat may apply to both the procedure and performance of IOs.
The literature on the normative standards suggests furthermore that the dynamics of legitimation varies across different institutional set-ups as well as the nature of policy field.For instance, in multistakeholder institutions in the field of climate change, we would expect a more varied combination of agents and (de)legitimation strategies (discursive, institutional, and behavioural practices) compared to state-centric and intergovernmental IOs, such as the AU (Bäckstrand & Söderbaum, 2018).As a general-purpose IO the AU covers a range of policy fields, which in itself is expected to trigger a diversity of legitimation and delegitimation strategies (Lenz & Schmidtke, 2023).A rich literature emphasizes the importance of democratic legitimation standards in the case of general-purpose regional IOs such as the AU, ASEAN and EU (Dingwerth et al., 2019;Schimmelfennig et al., 2021).However, these types of regional IOs are generally politicized and they also have distinct community-building ambitions, which tends to reinforce the diversity of legitimation standards beyond the democratic standards as such (Lenz & Schmidtke, 2023).One persuasive argument is that communitarian legitimation standards are important for such general-purpose IOs while they are expected to be less significant in other types of institutional setups such as task-specific and specialized IOs (Krösche et al., 2021).Referring back to the distinctive yet not exceptional traits of African regionalism: a rich diversity of legitimation politics surrounding the AU would also be expected because post-colonialismhence post-colonialism has impacted on a vibrant identity politics that manifests in the normative contestation of African regional institutions (Grilli & Gerits, 2020;Mbembe, 2001).

Agents
We adopt an inclusive approach regarding the agents of legitimation and delegtimation.We respond in particular to the fact that many previous frameworks have focused predominantly on IOs and their bureaucracies (Gronau & Schmidtke, 2016;von Billerbeck, 2020) or the most important member-states (Zaum, 2013).Our framework includes the conventional IO representatives and their member-states, but also 'other' (de)legitimation agents, which predominantly refers to civil society and business actors and other IOs (Bäckstrand & Söderbaum, 2018).While previous scholarship conceptualize these as stakeholders or audiences, on the receiving end of legitimation efforts, they may also be (de)legitimation agents in their own right; themselves trying to affect audiences' perceptions of IO legitimacy (Bexell et al., 2022).
Our conceptualization has at least two implications.First, the range of agents that are involved in both legitimation and delegitimation imply that these are pluralistic, even if interaction takes place under conditions of power asymmetry.Second, a number of these agents are simultaneously producers and consumers of legitimation claims.Taken together, both the pluralism and the double role of several actors (as both agents and audiences) reinforce legitimation struggles.

Types of strategies
The final component of the framework concerns the type of strategies enacted to engage in legitimation struggles, which we classify as discursive, institutional and behavioural.An overwhelming majority of scholars in the field have emphasized that establishing, maintaining and countering legitimacy is first and foremost a discursive phenomenon (Halliday et al., 2010;Steffek, 2003;von Billerbeck, 2020).Discursive strategies emerge when governors and critics seek to establish or contest legitimacy through public justification of claims-making and other communicative messages around IO legitimacy.Discursive strategies operate in a number of ways and can usually be identified through language and other means of communication manifested in texts and speech acts, such as mission statements, constitutional documents, annual reports, speeches, policy papers, press releases, public relations communications, protest slogans, etc (Ecker-Ehrhardt, 2018;Steffek, 2003).
Our framework is based on that sometimes 'actions may speak louder than words' in the contestation and legitimation of global governance (Stimmer & Wisken, 2019;Wajner, 2019).There are two non-discursive strategies.Institutional strategies concern the institutional features of an IO that are used to strengthen or undermine the legitimacy of that institution.Examples include constitutional reforms, administrative reform, transparency initiatives, and so on (Scholte & Tallberg, 2018, pp. 56-74;Zaum, 2013).
The third type of strategy is behavioural, and it includes a broader variety of actions and performances that have the communicative or symbolic intent to shape audiences' perceptions of legitimacy (Wajner, 2019).In previous research these strategies have usually been overlooked, downplayed or subsumed under discursive strategies.Examples include a range of symbols, rituals and informal rules, performance reviews, external partner compacts that interact with materialbased practices to signify consent or approval.They involve practices considered meaningful by IO officials, and other relevant actors, enacting norms at variance with the ostensibly shared standard in a larger governance system (Stimmer & Wisken, 2019, p. 521).They also include 'repertoires of contention' to use Tilly's microsociological term, including overt challenges such as protests, strikes, obstruction and delaying tactics, walkouts from meetings, or other forms of non-compliance and resistance (Gregoratti & Uhlin, 2018).Such strategies are available both to IO proponents and opponents.While behavioural strategies mostly are purposive, and closely resemble practices of contestation (Wiener, 2018), they involve a greater element of socialization and symbolism, ritual and performativity, which means that intentionality may not be clearly detected but becomes an important empirical question.

Methodological considerations
As elaborated in the introduction, the case of the AU is intended to help us transcend the conventional focus on established IOs dominated by the West (e.g.EU, UN, WTO, the World Bank) and move towards a post-Western research agenda on IO legitimation struggles.Our case study deals with the different strategies to justify or contest the AU's rule vis-à-vis relevant 'legitimacy-granting' audiences that has surrounded the AU institutional reforms in the period 2015-2020.The institutional reform process guided by the strategic framework of Agenda 2063 commenced in 2015 and the first continental progress report of Agenda 2063 was approved in 2020.Since the institutional reform within the AU entered a new phase in 2020, we limit the analysis to the aforementioned period.
In this time period, we identify two main institutional logics of social action, setting an empirical baseline for our analysis of both ideational and material sense-making strategies: Pan-African Unity and AU Fit for Purpose, respectively.The distinction is made to analytically trace and illustrate the competing normative (de)legitimation standards and the discursive and non-discursive strategies used by leading agents.The important implication that we draw from these nuanced beliefs is that they can (and do) vary and coevolve across actors, groupings and over time.
In this context, it needs to be pointed out that our study is concerned with the drivers and dynamics of legitimation struggles, and not their effects.While we draw on a rich literature showing that legitimation and delegitimation are (usually) consequential, there are significant methodological difficulties to assess and study the effects of legitimation struggles as opposed to how they emerge and are maintained (cf.Bexell et al., 2022).
Furthermore, it should also be mentioned that the legitimacy struggle in the AU is not determined by its level of legitimacy.While considerable attention in the past has been given to an authority-legitimacy link (Tallberg & Zürn, 2019;Zürn, 2018), recent research emphasizes that also lower-authority IOs need to legitimate their rule (Lenz & Söderbaum, 2023).While the AU is one of the IOs in Africa with the highest levels of authority, its legitimacy tends to fluctuate between different time periods and also policy fields (Lotze, 2013;Witt, 2019).It is also complicated to assess the level of legitimacy of the AU.With this said, the AU enjoyed considerable legitimacy after its establishment in the early 2000s, but then it decreased around 2008-2010, especially when Muammar Gaddaffi served as the Chairperson of the AU.Thereafter, the AU's legitimacy has increased again, inter alia due to the adoption of Agenda 2063.
We have used several data-collection techniques: qualitative content analysis of primary materials such as official documents and reports, secondary literature, participant observation, and nearly fifty semi-structured interviews with AU Commission officials, African member-states representatives, civil society and think tank representatives, analysts and consultants in the wider diplomatic circles around the African Union, donor and EU representatives, at various points in time also before and after the time period that we analyze here. 3Interviews with AU representatives and other actors help obtain legitimation agent's perspectives on (de)legitimation.Their limitation is that sometimes interview situations take on the form of legitimation attempts.Building relations with practitioners requires an empirical commitment akin to organizational sociology if one wishes to understand the tacit knowledge and the practice-based aspects of legitimation (Nicolini, 2009).Therefore, one of the benefits of combining participant observation with semi-structured interviews of both AU representatives and other actors, along with other supplementary techniques, is that it enables both the cross validation of data, and the integration of contextual and temporal observations with the perceptional and attitudinal data gathered from interviews (Dawson, 1997).Participant observation is a useful method to grasp more of the nuances and mixes of strategies by which agents engage in behavioural legitimation and delegitimation.In sum, using several methodological techniques can enhance the study's internal validity by overcoming a tendency in interview transcription to overinterpret how actors justify what they feel, think and do.This grounded interpretivist approach offers a rich and systematic way to identify and trace connections between the different (de)legitimation strategies enacted by different agents and the unfolding struggle dynamics.

Legitimation struggles over institutional reforms in the AU 2015-2020
Since its establishment in 2002, the AU has debated and implemented several reform initiatives, inspired by the common concern with governance and democratic deficits across the continent, external interference in African affairs, and the goal to strengthen continental authority and unity.The Agenda 2063 was launched at the jubilee celebrations in 2013 and formally adopted by the AU in 2015 (AU Commission, 2015a).This 50-year plan was developed by the office of former chairperson of the AU Commission Nkosazana Dlamini-Zuma, with significant assistance from both the UN Economic Commission for Africa (UNECA) and the African Development Bank (AfDB).
As it was soon discovered, the realization of Agenda 2063 necessitated a structural reform process of the AU (AU Assembly, 2016).Rwanda's President Paul Kagame and a team of advisors were entrusted to initiate and propose a complementary institutional structural reform strategy, the first of its kind since the AU's establishment (AU Executive Council, 2016).The institutional reform process revealed significant normative cleavages within the AU; we place these along a scale moving from a solidarity-oriented pan-African governance vision referred to as Pan-African Unity to a performance-oriented plan to create an AU Fit-for-Purpose (Confidential interview, AU observer and scholar, email communication March 2021; cf.Grilli & Gerits, 2020; Turianskyi & Gruzd, 2019).At the heart of this period of normative contestation was a struggle about the bedrock standards informing (de)legitimation of the AU that to this day challenges the implementation of the institutional reforms.
Our interest in the most recent period of institutional reforms (2015-2020) derives from the observation that the legitimating and delegitimating standards as well as strategies evolved in response to one another.We also observed that non-state actors played important roles in this normative contestation, and these actors were thus more important (de)legitimation agents than what is conventionally believed in existing research.In this way, the intense legitimation struggle surrounding the AU at this time challenges another influential argument in the study of African governance, namely that AU decision-making is characterized by consensus (Tieku, 2021b).Hence, we should not expect legitimation struggles to arise or intensify under conditions of normative consensus, or in the absence of resistance and critique.
While our framework is based on the interaction between competing normative standards and between legitimation and delegitimation more broadly, our empirical analysis is structured into two subsections, where we aim to locate the normative standards as well as the legitimation and delegitimation agents and strategies under two interactive socialization processes referred to above: Pan-African Unity and AU Fit-For-Purpose.The analysis focuses in particular on how legitimation and delegitimation develop in relation to each other.

Pan-African unitylegitimation and delegitimation
The AU's legitimacy has traditionally been based on African solidarity norms, liberation-struggle credentials and an anti-imperialist tradition (Bareebe, 2018).A wide range of AU member-states and leading representatives of the AU Commission invoke notions of African culture, indigenous values and specific socio-political African diplomatic traditions, such as consensus politics and African solidarity (AU Assembly, 2013;Gelot & Tieku, 2017).
These discursive legitimation strategies draw connections between pan-African values and the AU's institutional identity as a continuation of past continental initiatives such as the Charter of the Organization of African Unity (OAU), the Lagos Plan of Action, the Abuja Treaty, the African Renaissance, the New Partnership for Africa's Development (NEPAD) and the AU's Constitutive Act.Some AU officials draw on Agenda 2063 to empower such discursive strategies (Tieku, 2021a; Confidential interview, AU Commission staff, Addis Ababa, 18 October 2018).For instance, as then AU Commission Chairperson Dlamini-Zuma declared: The strength of Africa lies in its unity and its Pan-Africanism.When our forebears united, Africa was able to win its independence from colonialism against countries with armies and economies far bigger than ours.We should, therefore, never allow ourselves to be divided by anyone, or by anything.We should, as much as possible, coordinate our positions and adopt common positions.Let us stand or fall together!But we will not fall.Forward ever, backward never!(Dlamini-Zuma, 2017) In response to speeches as strongly normative as this one, we see that a range of non-state agents have engaged in discursive delegitimation to publicly expose the elitism of contemporary pan-Africanism, claiming that the notion of a shared collective African history and identity is a disingenuous 'myth' (Gwet, 2016).From their point of view, transnational elites and Africa's present-day leaders are so out of touch with any authentic pan-Africanist social movement that any normative legitimation standard based on values such as 'inclusivity' and 'we, Africans' appears dishonest (Confidential interview, AU observer and scholar, email communication March 2021; Allison, 2013).This is the real reason, according to a wide range of delegitimizers, for the AU's low level of legitimacy among rural people or the working-class (Confidential interview, AU Commission staff, Addis Ababa, 18 October 2018; Chawapiwa, 2015;Gwet, 2016).
Institutional measures and specialized departments or positions have been developed to advance popular and civil society participation within the AU, as seen by various participatory mechanisms such as African governance platforms and other committees including non-state actors tasked with benchmarking or monitoring implementation of Agenda 2063.Bringing the AU closer to the African citizenry is an essential legitimation standard for Pan-African Unity.In fact, attempts to legitimate the AU as a people-driven organization is one of the strongest normative pillars of Pan-African Unity.However, it is at the same time clear that these institutional legitimation efforts still face significant constraints.Civil society organizations challenge what they view as attempts by the AU and African leaders to co-opt civil society organizations and make them loyal AU 'legitimating' agents (Confidential interview, African civil society organization staff Addis Ababa, April 2018; Turianskyi & Gruzd, 2019).Some civil society representatives, who have been invited to AUcivil society cooperation frameworks, make use of their insider position to challenge or rejuvenate the AU's pan-African unity norms (Confidential interview, civil society organization founder, Addis Ababa, April 2018).
The discursive and institutional legitimation of the AU is combined with a rich and growing machinery of behavioural legitimation claims.Sometimes these intend to brand the pan-African character of the AU's public profile and public relations among counterparts and audiences within and beyond Africa (AU Assembly, 2013).Such behavioural legitimation may also be important for the AU's identity-and community-building, for instance through the AU flag, the AU anthem, and other symbols and rituals.AU officials and member-states employ these to legitimate pan-African governance in Africa and to some extent also among outside audiences, especially donors and external powers.Other similar examples include a wide range of symbols such as the Agenda 2063 slogan, 'The Africa We Want' hashtag, along with the performative functions of partnerships, workshops and forums with pro-AU civil society organizations that publicly subscribe to and embody the AU's pan-Africanism. 4These strategies are designed to legitimate the AU across diverse internal and external audiences, such as AU member-states, civil society and business actors, citizenries as well as external powers, donors and even other IOs.
Putting in place a more transparent tracking and monitoring of the implementation of Agenda 2063 is an institutional legitimation strategy (AUSA-NEPAD, 2020).Yet, it can simultaneously be understood as a behavioural legitimation practice that is enacted to symbolize the AU's advancement of pan-Africanism (AU Commission, 2015a, 2015b).The twelve 'flagship projects' of the first ten-year plan for Agenda 2063 includes numerous projects that seem to be more important as symbolic legitimation devices compared to their impact on development outcomes.For instance, the AU Commission's Director of Political Affairs, Khabele Matlosa, called the AU Passportone of the 'flagship projects'a 'test of our Pan-Africanism, the doctrine which underpins the African Union's existence' (Monks, 2016).While the official flagship goal is to remove restrictions and allow free movement of people within the continent, the debate among member-states concerned visa-free access and using the AU's logo on passports.Similarly, while certain other flagship projects may arguably be relevant for long-term development, their realization lie far into the future, and their symbolic legitimating function appears as more importantrelevant examples include the African outer space programme, the establishment of the Great Museum of Africa (in Algiers, Algeria), the creation of an African Virtual University (AU Commission, 2015b).With their pan-African character and coverage, these projects allow heads of state and AU officials to emphasize what the AU 'is' and what shared identity it represents, rather than what it 'does'.The emphasis on pan-Africanism 'in action' persuades some target audiences, including donors, of the AU's soft power, despite the fact that these flagship projects lag behind in terms of implementation (Adebajo, 2021;Tella, 2018).
The legitimation standard of pan-African unity has been delegitimated by a range of agents, who have criticized, among other things, its discursive and performative identity-based logic.Hence, the 'delegitimizers' depart from the argument that the pan-Africanist standard exaggerates what the AU represents at the expense what it does.We find that the behavioural delegitimation response has been stronger than expected compared to previous research on legitimation.Delegitimating actors claim for instance that 'Pan-African Unity' has lagged behind in terms of institutional reform, because it relies too much on symbolic politics and quixotic and ineffective proposals (Confidential interview, African think-tank staff, Bahir Dar, Ethiopia 2018; Adebajo, 2021).Some opponents act outside of the AU framework and 'outperform' the AU.For instance, some non-state actors have skilfully created independent digital technology services to connect with people, such as the example of polling African citizenry on their awareness of the AU or Agenda 2063 or advancing open data sources for citizen's monitoring of national and regional-level democratic governance (Chawapiwa, 2015).Rather, the institutional reforms motioned by proponents of Pan-African Unity, are viewed instrumentally as allowing African heads of state to self-legitimate as Pan-African representativesmainly among their peers but also towards external donors and partnersapproaching the AU as a state-led, often authoritarian, political order.

AU fit-for-purposelegitimation and delegitimation
Setting the stage for the structural reform proposed by Kagame and his team, Rwanda's President Paul Kagame harshly delegitimated the AU's past performance: The Assembly has adopted more than 1,500 resolutions.Yet there is no easy way to determine how many of those have actually been implemented.By consistently failing to follow up on the implementation of the decisions we have made, the signal has been sent that they don't matter … The African Union['s] work lacks clear focus.This makes it difficult to channel resources strategically and results in a fragmented and ineffective organization.(AU Assembly, 2017, p. 5,7) This was an unusually explicit (discursive) delegitimation attempt from an African head of state.In a widely discussed passage of the report, Kagame criticized the AU for being a 'dysfunctional organization in which member-states see limited value, global partners find little credibility, and our citizens have no trust' (AU Assembly, 2017, p. 8).Kagame and the advisory committee's discursive legitimation strategy emphasized that the AU Commission had taken on too many normative functions and that it was time to hand back power and responsibility to the member-states (AU Assembly, 2017, p. 8).This was supported among the member-states who wished to roll back the AU Commission's supranational elements.The Fit-for-Purpose process was thus rooted in a strikingly different normative legitimation standard which intensified the normative contestation of AU governance.
The Kagame-led group of delegitimizers gradually grew in size.The main agents of the competing normative standard included a growing number of influential AU member-states and leaders, leading representatives from the AU Commission, and other key African partner institutions, in particular the African Development Bank (AfDB) and the UN Economic Commission for Africa (ECA).Indicating their agency, leading representatives from the AU Commission also publically contested both the direction and the weak implementation of AU decisions and programs by the member-states (AU Assembly, 2013, p. 5).Importantly, non-state actors, intellectuals, and media representatives have also delegitimated the AU on the basis of the poor record of ratification and implementation of AU instruments (Murithi, 2021;SOTU, 2016).
The institutional reform programme that took shape, and that we associate with the AU Fit-for-Purpose socialization process, set out a number of objectives: (i) slimming down the AU's key priorities with a continental scope, to political, peace and security issues, economic integration, and Africa's global representation and voice; (ii) realigning AU institutions to ensure delivery on the key priorities; (iii) ensuring the efficient and effective management of the AU's activities at both political and operational levels; (iv) securing sustainable financing for the AU in ways that would ensure full ownership on the part of member-states (AU Assembly, 2017).
The normative legitimation standard here is about an AU urgently in need of improved delivery and performancean AU Fit-for-Purpose.This espouses a managerial-technocratic approach.As one senior AU official noted, the AU has in the past been overly concerned with normative legitimacy at the expense of technocratic-utilitarian problem-solving (Confidential interview, senior AU official 25 October 2018).The AU Commission Chairperson, Moussa Faki Mahamat, publicly voiced his determination to speed up the process of institutional reform (Faki Mahamat, 2018).Similarly, the Deputy Chairperson of the AU Commission at the time, Thomas Kwesi Quartey, also officially expressed strong support for the reform program.When a competing legitimation standard was brought out in the open, by a head of state and leading representatives of the organization, this intensified the interaction of legitimation and delegitimation strategies at play since more actors began to position themselves as part of the AU legitimation struggle.
A clear institutional legitimation strategy by the AU Commission leadership was to establish an institutional reforms unit within the Office of the Chairperson of the AU Commission.Legitimating this was done on the grounds of ensuring a structured monitoring and implementation process for the reforms (Duarte, 2018).Getting this unit in place in such a strategic location in the AU could further contribute to legitimation through a number of wide-reaching structural institutional reforms, such as (i) reducing the numbers of continental summits and items on the AU agenda; (ii) establishing sanction mechanisms to enforce binding AU decisions; (iii) clarifying the division of labour between the AU and other mechanisms and organizations; (iv) identifying ways of ensuring gender participation and the private sector's participation; and (v) a set of financial management and accountability principles (AU Assembly, 2017).Although these institutional reforms are believed to be important in order to improve performance and decision-making within the AU, they were also tools for institutional self-legitimation for the AU Commission (Fagbayibo, 2018).The progress in institutional reform implementation connected through its implementation matrix to previous reform proposals, rooting it in an authoritative diagnosis of what worked and what did not in AU governance (Okeke, 2018). 5 One ingredient in the Fit-for-Purpose was how to strengthen the AU's autonomy and self-funding.These proposals became important ingredients of the legitimation struggle, and the revised funding formula was legitimated by a fierce critique of the AU's donors and so-called international collaborating partners.The AU's heavy reliance on donor funding was defined as resulting from post-colonial legacies and donor intrusiveness which undermined both African ownership and the AU's performance (Duarte, 2018;Glas, 2018).The AU Peace Fund and the 0.2 levy were designed to reduce dependence on foreign funding by increasing financial commitments by AU member-states.In addition to institutional legitimation, proponents of the Fit-for-Purpose package also used sets of behavioural legitimation strategies, for instance, public disclosure of payment records by member-states in accordance with the AU Peace Fund decision as well as novel cooperative strategies towards financial and budgetary reform in the committee of finance ministers (Turianskyi, 2019).
Employing institutional delegitimation strategies, a number of African heads of state requested a review of the implications that the proposed institutional amendments would have for the AU's Constitutive Act.Another slightly different example was the claim that the institutional reforms would disempower the AU Commission and centralize power in the director of the AU Commission, at the same time as it would bypass AU checks and balances that were intended to ensure democratic decision-making (Louw-Vaudran, 2017).
AU self-funding and the proposal of a Peace Fund faced delegitimation by a range of African leaders, civil society actors, and at times also by the AU's own staff (Confidential interview, AU accredited African diplomat, Addis Ababa 18 October 2018).Hesitant AU member-states, for example, predominantly used an institutional delegitimation strategy and demanded assurances about sovereign control over the funds, the modalities for implementationsuch as the oversight mechanism to ensure transparency and accountabilityand the principles that would govern the sanctioning of member-states for underpayment of fees (Confidential interview, AU Commission official, Addis Ababa 18 October 2018).
The shift of legitimation standard to a performance-oriented and technocratic one has also been legitimated by less conventional agents.Acha Leke, a senior partner with McKinsey & Co. and a member of Kagame's reform committee, stated that performance-oriented reforms are good for business (McKinsey Global Institute, 2016), adding that 'our own experience working with public-sector institutions across Africa persuades us that practical measures to ensure effective implementation should be the priority' (Leke, 2017, p. 2).African banking, business and philanthropic actors, have publicly called for efficient institutions conducive to improving the investment climate and the roll-out of financial management reforms as part of their assessments of the AU (Mo Ibrahim Foundation, 2018).The support by big business has been deeply contested by certain parts of civil society endorsing Pan-African Unity.According to a pan-Africanist civil society actor: 'The day reform is thought out by the likes of McKinsey, our political union is dead' (Confidential interview, think-tank director, Addis Ababa March 2020).Other civil society actors claim that the spirit of the AU Fit-for-Purpose is equivalent to a 'gross betrayal' of true pan-Africanism (Makori, 2017).
Several observers have noticed the increasing enactment of a type of behavioural delegitimation practice in the AU's institutional culture, namely a repertoire of deliberate tactic of delaying and withholding compliance with the reform requirements until it 'fizzles out' (Confidential interview, former civil society organization staff, Addis Ababa April 2018).Several AU member-states and sub-regional organizations have demanded reviews, transitional periods and clarifications, and have therefore cast in doubt or delayed the various implementation chains.The AU institutional reform unit has engaged in negotiation to meet such requests while attempting to avoid delays, for example by negotiating directly with Southern African Development Community memberstates over a public request they had filed in which they listed several concerns (SADC, 2018).
A variety of civil society actors share the analysis that the Fit-for-Purpose process symbolizes a results-based and outcome-oriented management culture that hinders rather than facilitates longer-term socio-economic transformation (Confidential interview, former civil society organization founder, Addis Ababa April 2018; Witt, 2019, p. 156).Many such civil society actors, thinktanks and higher education institutions have engaged in behavioural delegitimation by providing 'watchdog' and independent advisory functions that ensure adherence on AU rules and regulations (Amani Africa, 2017).Some of them organized public events on 'inclusivity' or the 'people's AU', providing the practical means to stimulate an open conversation about the claim that the reform process undermines people-centredness (Confidential interview, AU accredited African diplomat, Addis Ababa 18 October 2018).Critics have also used political satire and other protest actions to contest Kagame's credentials as a pan-Africanist statesman and legitimate AU reformer.Thus, some delegitimation strategies try to discredit the institutional reform by explicitly tying them to Paul Kagame's controversial leadership style.These are diffuse strategies and it is difficult to determine whether the critique targets the substance of reform, how the AU governs, or whether it is best explained as subjective 'anti-Kagame' sentiments (Turianskyi, 2019).
Epitomizing the symbolism of the legitimation struggle, and neatly capturing how the two processes are located along a normative continuum, we refer to statements by Thabo Mbeki, former South African head of state and a particularly forceful representative of pan-African visions.Mbeki fiercely rejected shifting legitimacy standards towards the AU Fit-for-Purpose position because normatively and ideologically this represented a clash with pan-African values.Mbeki claimed that 'the AU reform process weakens, not strengthens, it' (Mbeki, 2018), and argued that 'Africa does not need a more technocratic, UN-like organization, but an AU that, while it may experience failures and be shambolic, fights to give political protection to Africa globally' (quoted in Onyango-Obbo, 2019).Although Mbeki's delegitimation could be said to be discursive, we emphasize here its symbolic and performative power.His pan-African legacy endures in terms of the 'African renaissance', which represents one side on the legitimation struggle between Pan-African Unity and AU Fit-For-Purpose.

Summary of the AU legitimation struggle
Our analysis highlights the two most significant normative standards in the AU legitimation struggle in the time period 2015-2020.Within both socialization processes, we find that a number of African leaders, civil society actors, business actors, development agencies and high-profile African intellectuals enact a variety of strategies to position themselves in the legitimacy struggle.
When agents legitimate the AU in line with 'Pan-African Unity' they emphasize the shared history, identity, culture and destiny of the continent.Pan-African and 'moral' norms and values provide a symbolic legitimation repertoire.A wide range of AU officials and a significant number of political leaders and governments as well as civil society actors carry out their legitimation activities through a predominantly discursive narrative of Pan-African Unity, which is also combined with certain strategic behavioural legitimation activities.Here the AU is legitimated on the basis of what it is and represents rather than what it does, which is intended to build the AU's legitimacy regardless of the performance and achievements of the organization.Soon after the adoption of Agenda 2063, however, a number of influential AU member-states and leaders, and surprisingly even leading representatives of the AU Commission began to contest and devalue the declatory type of legitimation centred on pan-Africanism and African solidarity.A range of 'other' actorscivil society actors, intellectuals, other African-based institutions (such as the AfDB and UNECA)joined forces and also voiced strong concerns about prevailing declatory practices.
When agents instead legitimate the AU along the lines of the AU Fit-For-Purpose, the AU's rule is appraised on the basis what it does rather than what it is, with emphasis on performance, effectiveness and delivery.The most influential legitimation agents include a number of political leaders and AU officials, such as Rwanda's President Paul Kagame, Donald Kaberuka, former President of the African Development Bank Group (2005Group ( -2015)), the Chairperson of the AU Commission, Moussa Faki Mahamat, and some other leading state representatives.The AU Fit-for-Purpose reform proposal is based on an intimate link between legitimation and delegitimation, because its proponents try to gain support from relevant audiences by publicly shaming the AU and its member-states on the basis of the underperformance and weak political commitment to the organization.With regard to strategies, they draw heavily on institutional strategies, which is seen through the quick implementation of a far-reaching institutional reform package.Its proponents also employ discursive and to a lesser extent behavioural strategies.Several influential AU member countries and civil society actors, especially the proponents of pan-African unity, pushed back and used a variety of discursive and behavioural strategies to delegitimate AU Fit-for-Purpose.

Conclusion
In spite of a growing recognition of contestation in general terms, there are rather few analytical frameworks available for coming to grips with how legitimation and delegitimation feed on each other and develop into 'legitimation struggles'.Drawing on recent insigths, we have developed a framework for the study of legitimation struggles based on three components: normative standards, agents, and strategies.
The case selection is motivated by the view that the AU warrants more consideration in the literature on IO legitimacy and legitimation.The AU is arguably the most important IO on the African continent, and we take it as a crucial case for moving towards a post-Western research agenda on IO legitimation struggles.The AU can be compared to a range of other regional general-purpose IOs, such as ASEAN, ECOWAS, EU, Mercosur, SAARC, and so on (Hooghe et al., 2019).We also find it relevant to compare the AU with a range of task-specific IOs, especially those that enjoy a certain degree of authority and involve a sufficient number of actors that disagree about the standards of appropriate rule (e.g.WHO, WTO, UNFCCC) (see Bexell et al., 2022).
While our empirical findings of the AU offer several important new insights, we believe that the case study also has broader implications for a comparative and post-Western research agenda on legitimation struggles in IOs and multistakeholder institutions.We summarize our main findings in relation to the three parts of our analytical framework.
First, legitimation struggles are deeply related to the social purpose of the organization and competiting normative standards.In our analysis, (de)legitimation of the AU was a means for key agents to approach the deeper stakes in the IO.Indeed, the legitimation struggle was at heart about the social identity of the organization.Significantly, this means that the conventional distinction between input and output legitimacy, or procedure and performance, is not sufficient for understanding the dynamics of legitimation struggles, at least not in the case of the AU.Much of previous research emphasizes democratic legitimation standards in the case of multi-purpose IOs such as the AU.According to our analysis, however, democratic standards had less impact because Pan-African Unity was first and foremost based on community-building and communitarian legitimation standards (i.e.what the AU represents) whereas AU Fit-for-Purpose was strongly legitimated with regard to performance and effectiveness more than anything else.Recent research by Lenz and Schmidtke (2023) reveal that IOs diversify their legitimation discourse by invoking a pluralism of normative standards.Our study shows that the legitimacy struggle does not derive its strength from a pluralism of normative standards, but through the contestation over two specific standards as captured by 'Pan-African Unity' and 'AU Fit-for-Purpose'.Our analysis underlines the need to go beyond a focus on legitimation and also consider delegitimation.In fact, legitimation and delegitimation strategies alike were drawing on these two bedrock pan-African standards.
Second, while IO officials and (the most powerful) IO member-states are generally seen as the most prominent (self-)legitimation agents, the case of the AU underlines the need for a broader and more inclusive approach.Our analysis emphasize that 'other' agentssuch as civil society actors, business actors, as well as other IOsare actively contributing to the legitimation struggle.These other and non-state agents tend to liaise and group together with state actors on both sides of the struggle, which further intensify the legitimation struggle.Furthermore, we found it noteworthy that leading officials from the AU Commission, depending on their location across the normative continuum, were actively involved also in the delegitimation of their own organization.Future research ought to compare our results with other cases in order to develop more sophisticated theoretical propositions about legitimizers and delegitimizers and how they impact each other.
Third, legitimation and legitimation struggles are not simply a discursive phenomena.Clearly, discursive strategies are essential but future research need to consider that legitimation struggles are reinforced when agents use different mixes and combinations of discursive and non-discursive (i.e.institutional and behavioural) legitimation strategies.In the case of the AU, discursive strategies were much more important for the proponents of Pan-African Unity, whereas the proponents of the AU Fit-for-Purpose relied heavily on a combination of institutional and discursive strategies.Furthermore, behavioural delegitimation strategies were even more important to non-state actors compared to what we expected on the basis of our reading of previous research.This means that sources and strategies of discontent surrounding contemporary IOs might generally be understudied and undervalued, which implies a knowledge gap in our understanding of the logic of legitimation struggles.A plausible avenue of inquiry for future research would therefore be to