Legitimacy Struggles in Global Governance: Legitimation and Delegitimation of the Asian Infrastructure Investment Bank

The Chinese-led Asian Infrastructure Investment Bank (AIIB) started operation in January 2016 in the context of severe criticism from the U.S. government and transnational civil society organizations, but with the support from major Asian and European states. The establishment of this new global governance institution (GGI) raises intriguing questions concerning legitimacy and (de)legitimation. What legitimacy challenges has the AIIB as a new institution experienced and how has the Bank tried to boost its legitimacy? What specific legitimation and delegitimation practices have been applied by what actors and with what effects? Contributing to research on legitimacy in global governance, this article develops a framework for analyzing legitimacy struggles. It highlights the dynamic relations between legitimation (practices intended to boost beliefs that the rule of a political institution is exercised appropriately) and delegitimation (practices challenging the appropriateness of a political institution’s exercise of authority). The framework includes a systematic study of different agents of (de)legitimation (including GGIs, states, and nonstate actors), practices of (de)legitimation (categorized as institutional and discursive), and institutional sources of (de)legitimation (related to procedure as well as performance). The case study, based on in-depth qualitative content analysis of documents produced by the AIIB and other actors, demonstrates how legitimizers and delegitimizers respond to each other. Being a new GGI, legitimacy struggles on the AIIB have to a large extent focused on its institutional design referring to (the lack of) democratic procedures. Legitimizers tend to stress technocratic performance, whereas delegitimizers are more concerned with fair performance.


Introduction
Global governance, understood as "the exercise of authority across national borders as well as consented norms and rules beyond the nation state, both of them justified with reference to common goods or transnational problems" (Zürn, 2018, pp. 3-4) is a central feature of contemporary world politics. Global governance institutions (GGIs) can be intergovernmental-such as the World Trade Organization and the World Bank-or include nonstate actors-such as the Forest Stewardship Council and the Internet Corporation of Assigned Names and Numbers. Some GGIs have a genuinely global scope, whereas others-such as the European Union and regional development banks-have a specific regional focus. As these institutions gain more authority, their procedures and performance are more frequently evaluated according to normative standards and criticized based on such evaluations (Bernstein, 2011;Brassett & Tsingou, 2011;Dingwerth, Witt, Lehmann, Reichel, & Weise, 2019;Steffek, 2003;Tallberg, Bäckstrand, & Scholte, 2018;Zaum, 2013;Zürn, 2018). Such critical normative assessments of GGIs may challenge the legitimacy of the institutions. To function well, GGIs need to be perceived as legitimate, at least among influential actors, if not the general public. Therefore, criticized and challenged GGIs have to actively try to boost their legitimacy in the eyes of various audiences. They engage in processes of self-legitimation (Gronau, 2016).
This article suggests that an important aspect of contemporary global governance is the complex interaction 870203S GOXXX10.1177/2158244019870203SAGE OpenUhlin research-article20192019 1 Lund University, Sweden of delegitimation and legitimation, what might be called legitimacy struggles. GGIs, their member states, and possibly other state and nonstate actors too try to make the GGI seem more legitimate. They make statements, initiate institutional reforms, and engage in other practices intended to affect various audiences' legitimacy beliefs concerning the GGI in a positive way. Meanwhile, other GGIs, states, and nonstate actors engage in practices that might affect legitimacy beliefs concerning the GGI in a negative way (Bäckstrand & Söderbaum, 2018). There is often a struggle between (self-)legitimizers and delegitimizers. Extant research has examined the self-legitimation of GGIs (e.g., Gronau, 2016;Gronau & Schmidtke, 2016) and protest against GGIs as delegitimation attempts (cf. Gregoratti & Uhlin, 2018;Haunss, 2007), but the interaction of legitimation and delegitimation has seldom been systematically analyzed (but see Anderl, Deitelhoff, & Hack, 2017). By focusing on both delegitimation and legitimation and how the two processes are related, this article intends to indicate a way toward a more dynamic analysis of (de)legitimation processes in global governance in terms of legitimacy struggles.
Previous research on (de)legitimation in global governance has mainly focused on relatively well-established GGIs (e.g., Gronau & Schmidtke, 2016;Steffek, 2003;Zaum, 2013). The process of establishing a new GGI has not been systematically analyzed from a (de)legitimation perspective. Yet, it is likely that a GGI in the process of establishing itself faces specific legitimacy challenges. According to Suchman (1995), there are three general challenges of legitimation: gaining, maintaining, and repairing legitimacy. Unlike most previous research on (de) legitimation in global governance, which has focused on how GGIs maintain legitimacy or how they try to repair their legitimacy after a legitimacy crisis, this study is concerned with the challenge of gaining legitimacy in the process of creating a new GGI.
One of the most significant recently established GGIs is the Asian Infrastructure Investment Bank (AIIB). Following a Chinese initiative announced in 2013, the new regional multilateral development bank (MDB) started operation in January 2016. With China as its leading member, the establishment of this GGI implies a significant change in the institutional setup for the global and regional governance of development. It has been seen as challenging existing development banks such as the World Bank and the Asian Development Bank (ADB). The Chinese initiative to create a new regional development bank must be understood in the context of China's growing economic power and its dissatisfaction with a limited role in the existing dominant GGIs. The AIIB is part of a wave of new global initiatives that China has promoted, including the G20, the New Development Bank (NDB) of the "BRICS" grouping (Brazil, Russia, India, China, and South Africa), and the Silk Road Economic Belt and 21st-Century Maritime Silk Road (One Belt & One Road; Chin, 2016). The Chinese leadership has been frustrated with its lack of influence in international financial institutions (IFIs; Glaser & Vitello, 2015;Mackintosh, 2016). Governance reforms in the major IFIs have been slow and largely symbolic. The 2010 reform package of the Bretton Woods institutions was blocked by U.S. Congress for many years (Ren, 2016). Hence, China's move to establish the AIIB is reported to reflect "China's open dissatisfaction" with U.S. influence in existing GGIs (Etzioni, 2016, p. 179). As stated by one observer, "/w/hen an emerging power grows but is not embraced or even welcomed by the established powers and the global institutions they dominate, the former trying to create new institutions becomes something inevitable" (Ren, 2016, p. 438).
This development raises a number of intriguing questions concerning legitimacy and (de)legitimation in governance beyond the nation-state. What legitimacy challenges has the AIIB as a new institution experienced and how has the Bank tried to boost its legitimacy? What specific legitimation and delegitimation practices have been applied by what actors and with what effects? The article addresses these questions, first, by developing a framework for analyzing legitimacy struggles in global governance, paying equal attention to legitimation and delegitimation, and distinguishing between different institutional and discursive (de)legitimation practices, and second, by applying this framework to a case study of the AIIB. In doing so, the article aims at contributing to research on legitimacy and (de)legitimation in global governance, particularly through its dynamic conceptualization of (de)legitimation as legitimacy struggles and its focus on the specific case of the establishment of a new GGI. It also aims at contributing to emerging research on the AIIB, which has so far not paid any systematic attention to legitimacy issues. Timewise, the focus is on the process leading to the establishment of the new institution (from 2013) and its first 2 years of operation (2016)(2017).
Following this introduction, the article proceeds with the development of an analytical framework, followed by methodological considerations. Thereafter, the empirical analysis of legitimacy struggles related to the AIIB is divided into two sections, focusing on institutional and discursive dimensions of (de)legitimation, respectively. The article ends with some concluding remarks suggesting new directions within this field of research.

Legitimacy Struggles in Global Governance: A Framework for Analysis
This section clarifies how key concepts are used and outlines an analytical framework for empirical research on legitimation and delegitimation in global governance. Legitimacy in global governance has been studied both from normative and empirical perspectives. Social scientists can be authors as well as observers of legitimacy evaluations (Hurrelmann, 2017). This study follows the second tradition, that is, empirical research observing (de)legitimation practices.
Legitimacy can be defined as "a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (Suchman, 1995, p. 574). Legitimacy is relational. It depends on audiences' beliefs about the exercise of authority (Tallberg & Zürn, 2019). Legitimacy beliefs must be distinguished from other reasons to obey rules, such as coercion and self-interest (Hurd, 1999). Acceptance of an authority because of fear of punishment or because of cost-benefit calculations do not imply legitimacy (Steffek, 2003). Hence, legitimacy requires normative grounding. Legitimacy evaluations consist of "normative validity claims of a generalizable character" (Hurrelmann, 2017, p. 64).
Rather than legitimacy as an attribute of an institution, this study is concerned with processes of (de)legitimation. Legitimacy cannot be empirically observed. What can be observed is legitimation as a process (Hurrelmann, 2017). Legitimation means deliberately boosting beliefs that the rule of a political institution is exercised appropriately, whereas delegitimation implies challenging the appropriateness of a political institution's exercise of authority (Tallberg & Zürn, 2019).
A central claim of this article is that legitimation and delegitimation practices often occur simultaneously. As stated by Bäckstrand and Söderbaum (2018), "legitimation and delegitimation often shape each other and therefore need to be integrated within a single framework" (p. 102). Certain actors try to delegitimize an institution, whereas others attempt to legitimize it. When GGIs gain more authority, they typically face more "politicization" or delegitimation, which forces them to engage in legitimation practices (Ecker-Ehrhardt, 2018b;Zürn, 2018, pp. 89-90;Zürn, Binder, & Ecker-Ehrhardt, 2012). This pattern, where a GGI engages in self-legitimation as a response to protest and criticism, has been clearly observed in the case of the WTO (Anderl et al., 2017). The self-legitimation practices of a GGI may also provoke criticism and be challenged by delegitimation attempts. In interactive processes of legitimation and delegitimation, it is not always clear what the starting point is. Such dynamic interaction of legitimation and delegitimation has been referred to as "legitimacy games" (Van Rooy, 2004) and "legitimation contests" , but I prefer the term legitimacy struggles. This concept highlights the contentious nature of legitimacy dynamics in global governance. A struggle involves at least two combatants with (partly) diverging goals and interests. They fight each other to win advantages. While the vocabulary of "struggles," "combatants," and "fight" might invoke violent connotations, legitimacy struggles are typically fought by peaceful means. Those involved in a legitimacy struggle apply different (de)legitimation practices to influence various audiences' legitimacy beliefs. The advantages of this concept are that it highlights the contested nature of legitimacy in global governance and draws attention to the ways in which legitimation and delegitimation practices are linked to and shape each other.
To systematically analyze legitimacy struggles in global governance, four aspects of (de)legitimation need to be considered: the object of (de)legitimation, the agents of (de) legitimation, the practices of (de)legitimation, and the institutional sources of (de)legitimation claims. A fifth aspectaudiences of (de)legitimation (Bexell & Jönsson, 2018)-will not be included here as this aspect of (de)legitimation would require a study of its own.
First, the object of (de)legitimation in global governance is an institution that has authority and whose right to rule is evaluated. In this study, the object of (de)legitimation is a GGI in the process of being established: the AIIB. Generally speaking, established GGIs can be assumed to be perceived as more legitimate than new ones (Lenz & Viola, 2017). The process of setting-up a new GGI may be a contested process, especially if the new GGI is perceived to challenge and compete with existing GGIs in the field. In a process of "counterinstitutionalization" (Zürn, 2018, p. 142), it is likely that there are both delegitimation attempts and self-legitimation by the GGI, which differ from (de)legitimation of older, more well-established GGIs. Whereas legitimacy concerns related to established institutions tend to focus on decisionmaking procedures (input legitimacy) and performance (output legitimacy), the central legitimacy issues for a new GGI are arguably related to the constitutive justification of its authority, what Oates (2016) calls "constitutional legitimacy." Constitutional legitimacy focuses on the jurisdiction rather than exercise of institutional authority (Oates 2016). It is concerned with "what governance institutions are necessary, whose interests they should serve, and how authority should be distributed within an institutional order" (Oates 2016, p. 200).
Second, there may be many different agents of (de)legitimation. Previous research has examined how GGIs try to legitimize their exercise of authority (Ecker-Ehrhardt, 2018a, 2018bGronau & Schmidtke, 2016). GGI self-legitimation might be carried out not only by the institutions' management but also by different parts of the GGI bureaucracy, not least public communication departments, which have been set up by many GGIs (Ecker-Ehrhardt, 2018b). Moving beyond various administrative parts of GGIs as agents of legitimation, other important legitimation agents are the member states of the GGI (Zaum, 2013). Nonmember states may also play an important role in legitimizing or delegitimizing a GGI as may other GGIs. Whereas nonmember states are of little concern for global GGIs with more or less universal membership, nonmember states and other GGIs may be important agents of (de)legitimation in relation to regional organizations in the global South (Bah, 2013). Much international relations (IR) research on legitimacy has focused exclusively on states and intergovernmental organizations as agents (Gronau & Schmidtke, 2016;Hurd, 1999;Zaum, 2013). However, from the perspective of legitimacy struggles, it is important to also acknowledge the agency of nonstate actors. Business and civil society actors are increasingly seen as important actors in global governance, alongside states and intergovernmental organizations. Such nonstate actors may also be agents of (de)legitimation vis-à-vis GGIs. At least since the 1990s, there has been significant civil society protest against GGIs (Gregoratti & Uhlin, 2018). It might also be important to consider the origin of nonstate actors (whether they come from member states or nonmember states, the global North or the global South, etc.). A focus on the dynamic interplay between a broad range of state and nonstate actors is essential for the understanding of legitimacy struggles in global governance.
While this study is limited to one object of (de)legitimation (the AIIB), it aims at covering a broad set of state and societal actors as agents of (de)legitimation, including the management and staff of the AIIB itself, other GGIs (mainly the World Bank), member states (especially China), other states (mainly the United States), and civil society actors (both from Asia and the global North).
Third, (de)legitimation practices can be institutional (reforms of the institutional design of a GGI intended to affect audiences' legitimacy beliefs), discursive (statements supporting or challenging the rightfulness of a GGI's rule), or behavioral (nonverbal acts that can affect audiences legitimacy beliefs; Bäckstrand & Söderbaum, 2018). Institutional, discursive, and behavioral (de)legitimation practices are ideal types. In practice, they often go together (cf. Gronau & Schmidtke, 2016;Tallberg & Zürn, 2019). It is indeed difficult to determine that a certain institutional or behavioral practice is a (de)legitimation practice if it is not accompanied by a discursive (de)legitimation statement.
While recognizing that the full spectrum of (de)legitimation practices might be important to consider when analyzing legitimacy struggles, this case study of the AIIB will be limited to institutional and discursive practices. The reasons for this are partly that it would be difficult to get sufficient empirical depth across all types of (de)legitimation practices within the limited scope of this article, and partly that behavioral practices seem to be less important in the case of the AIIB. For instance, civil society demonstrations, a major form of behavioral (de)legitimation observed in extant research, have hardly occurred in the AIIB case.
Institutional legitimation consists of (changes in) institutional features of a GGI, accompanied by justification of its exercise of authority. Institutional features that can influence legitimacy beliefs include (a) the formal structure of GGI authority, for example, voting rights; (b) broadened participation allowing new nonmember actors to take part in one or several phases of the GGI's policy process; (c) transparency, implying some kind of public communication policy that regulates the spread of information about the GGI; (d) accountability mechanisms, that is, some kind of retrospective channel for answerability; and (e) cooperation agreements or partnerships with other actors (cf. Bäckstrand & Söderbaum, 2018). Institutional delegitimation refers to criticism against the GGI's implementation of these institutional features.
Discursive (de)legitimation consists of verbal evaluations of a GGI's exercise of authority according to certain normative standards or sources. Discursive practices, found in texts and speech acts, have received most attention in research on (de)legitimation in global governance (Ecker-Ehrhardt, 2018a, 2018bGronau, 2016;Gronau & Schmidtke, 2016;Schneider, Nullmeier, & Hurrelmann, 2007;Steffek, 2003). This study covers the interplay between institutional and discursive legitimation and delegitimation practices by both state and societal actors in relation to the AIIB.
The fourth and final dimension of the analytical framework consists of the institutional sources of legitimacy linked to (de)legitimation practices. There might be many different sources of legitimacy, including Weber's famous categories of rationality, tradition, and charisma (Weber, 1922(Weber, /1978. More adjusted to the context of legitimacy in global governance, Scholte and Tallberg (2018) develop a 2 × 3 matrix typology of institutional sources of legitimacy. The typology takes the well-known distinction between procedure (input) and performance (output) as a point of departure and adds a threefold distinction between democratic, technocratic, and fair as three generic qualities that may apply to both the procedures and the performance of GGIs. Stakeholder participation and accountability are characteristics of a GGI's policy-making process that might enhance its legitimacy. This refers to democratic procedure. Democratic performance as a source of GGI legitimacy is related to increased popular participation and public accountability in wider society as an outcome of the activities of the GGI. Legitimacy derived from technocratic procedure implies efficiency or expertise. Technocratic performance is related to problem-solving capacity resulting in societal benefits. As for fair procedure, this quality has to do with institutional features, such as impartiality and access for those affected by the policies of the GGI. Fair performance, finally, refers to outcomes of GGI activities that secure human dignity and distributive justice.
It might be argued that this typology does not adequately capture the "constitutional legitimacy" (Oates, 2016) referred to above as particularly important in the process of establishing a new GGI. However, I consider the choices of institutional design related to authority and decision-making procedures that are of central concern for a new GGI to be included in the category of democratic procedure. This source of institutional legitimacy should be interpreted as referring to not only the exercise of institutional authority but also how these institutions are designed in the first place. Although not comprehensive, the Scholte/Tallberg typology is still more systematic and precise than earlier efforts. Therefore, I use it in the analysis of (de)legitimation practices related to the AIIB.
To sum up, the framework outlined here intends to analyze the dynamic relations between legitimation and delegitimation in terms of legitimacy struggles. Legitimation-boosting beliefs that the rule of a political institution is exercised appropriately-and delegitimation-challenging the appropriateness of a political institution's exercise of authorityshould be analyzed together. In addition to the object of (de) legitimation-a new GGI-there are three major components of the analytical framework: the agents of (de)legitimation, which might include GGIs, states, and nonstate actors; practices of (de)legitimation, which can be categorized as institutional and discursive; and institutional sources of (de) legitimation. Institutional (de)legitimation refers to such institutional features of a GGI as its formal structure of authority, arrangements for broader participation in policy making, transparency, accountability mechanisms, and cooperation agreements with other actors. Discursive (de) legitimation consists of verbal evaluations of a GGI's exercise of authority. The justification of institutional reforms and discursive (de)legitimation in general can be grounded in the following institutional sources of legitimacy: democratic procedure, democratic performance, technocratic procedure, technocratic performance, fair procedure, and fair performance.

Methodological Considerations
Studying the dynamic relations between delegitimation and legitimation in legitimacy struggles surrounding a GGI requires in-depth analysis of a number of different types of texts. A qualitative content analysis, implying a close reading of documents produced by the GGI and other relevant agents of (de)legitimation, is the most suitable methodological approach. It allows for both a systematic search for the conceptual categories identified in the analytical framework and a contextual understanding of (de)legitimation practices. However, there are many methodological challenges in this kind of empirical study of legitimacy struggles in global governance. What kind of material should be used? How can (de)legitimation be distinguished from other types of discursive and institutional practices? How can the institutional sources of legitimacy be captured in the analyzed texts? Here I consider these challenges and how I deal with them.
Institutional and discursive legitimation practices can be empirically studied in documents produced by the GGI in question. Discursive (de)legitimation practices by other actors (including the legitimation and delegitimation of institutional reforms) can be found in texts produced by these actors. Media reports on a GGI may also contain (de)legitimation practices. To capture self-legitimation by the AIIB, I analyze all material available at the AIIB website during 2016 and 2017, the first 2 years of the AIIB's operation. This includes a number of policy documents (for instance, on public communications, environmental and social framework [ESF], and consultation complaint mechanism), two annual reports, and 34 news items. To capture (de)legitimation by other actors, I analyze documents produced by the civil society organizations (CSOs) most active in relation to the AIIB. These include the NGO Forum on ADB, the Bank Information Center, and Transparency International. An indication of their significance for AIIB (de)legitimation is that they frequently appear in AIIB consultation processes, as referred to on the AIIB website. Extant research on civil society engagement with other MDBs has also identified these CSOs as significant actors. I systematically searched the websites of these organizations for material on the AIIB. There was also an element of snowballing in the sampling process as other relevant actors were found in the documents of the most prominent CSOs. For (de)legitimation by other actors, especially before the AIIB was established, I rely on published research on the AIIB and media reports (found through simple Google searches and references in other texts). All texts were entered into MAXQDA, a software for qualitative data analysis, and coded following the analytical framework developed in the previous section.
The question of how to distinguish (de)legitimation from other forms of public communication and institutional reforms is a major challenge in this field of research. The sociological or empirical approach to legitimacy links (de) legitimation practices to legitimacy beliefs. The most common approach in the literature is to define (de)legitimation practices as practices intended to affect audiences' legitimacy beliefs (Gronau & Schmidtke, 2016;Tallberg & Zürn, 2019). While this approach is helpful in distinguishing (de) legitimation from other practices, it raises the question of how to determine the intentions and motivations behind a practice. Combining text analysis with interviews with representatives of all agents of (de)legitimation might be a possibility, but a very demanding approach in terms of research resources. A more feasible solution is to only code statements that clearly refer to the GGI's exercise of authority as rightful or not. Such statements can be assumed to be intended to influence audiences' legitimacy beliefs. This means that a description of an institutional reform is coded as a legitimation practice only if it is accompanied by a justifying statement referring to the rightfulness of the GGI's exercise of authority. Similarly, a critical statement of an institutional reform is coded as a delegitimation practice only if the criticism is linked to an evaluation of the GGI's exercise of authority. This does not mean that statements necessarily have to explicitly make use of words such as "authority," "rightful," and "legitimate" to qualify as (de)legitimation statements. Rather, a qualitative interpretation of the implicit meaning of evaluative statements within the broader context has to be made.
A further methodological consideration is how to operationalize the institutional sources of legitimacy outlined in the analytical framework. A statement is coded as democratic procedure if it discusses the GGI's decision making in relation to how participatory/inclusive it is, if it is based on deliberations, how accountable it is, and/or how transparent it is. A statement is coded as democratic performance if it evaluates the results that the GGI produces in terms of being positive or negative for democratic accountability and participation in a wider societal context. A statement is coded as technocratic procedure if it assesses the GGI's decision making in terms of how efficient it is (producing a high number of policy decisions in a short period of time and/or at a low cost) and to what extent it is based on expertise relying on specialized and specialist/expert knowledge and skills). A statement is coded as technocratic performance if it evaluates the results that the GGI produces in terms of efficiency (solving a large number of problems in a cost-efficient and timely way) and/or effectiveness (addressing problems in a way that have noticeable positive impact). A statement is coded as fair procedure if it refers to impartiality and access for affected stakeholders when evaluating the decision-making procedures of the GGI. A statement is coded as fair performance if it evaluates the results produced by the GGI in terms of human dignity and distributive justice (cf. .

Legitimacy Struggles on the AIIB
In the initial phase of establishing the AIIB, most delegitimation attempts came from the United States. The establishment of the AIIB was "seen by many in the US as a challenge to existing Western-led institutions, as well as a potential opportunity for China to expand its influence in the region at the US expense" (Glaser & Vitello, 2015, p. 25). The United States refused to join the Bank and reportedly pressured its allies to do the same (Etzioni, 2016). Discursive delegitimation statements from the U.S. government and Congress focused on concerns that the AIIB was unlikely to adopt "best practices" developed by established GGIs, especially related to human rights abuses and environmental risks, as well as transparency, good governance, and anticorruption mechanisms. U.S. politicians expressed worries that the AIIB would steer infrastructure projects to Chinese contractors and use the new Bank to twist the arm of Asian states that need infrastructure funding (Etzioni, 2016). In his State of the Union address in January 2015, President Obama, referring to the establishment of the AIIB and other Chinese initiatives in global governance, said that "China wants to write the rules for the world's fastest-growing region," and added, "We should write those rules. We should level the playing field" (Glaser & Vitello, 2015, p. 25). Treasury Secretary Lew explained the U.S. position on the AIIB at a Congress hearing: Our concern has always been . . . will it adhere to the kinds of high standards that the international financial institutions developed. . . . Will it protect the rights of workers, the environment, deal with the corruption issue appropriately? (Glaser & Vitello, 2015, p. 26) Meanwhile, the Chinese government actively tried to delegitimize the leading role of the United States in the global financial governance structure and framed the creation of a new MDB as part of a democratization of the international economic order. The AIIB was legitimized by China as an institution devoted to Asia's economic growth, reflecting China's responsible leadership in reforming global governance (Park, 2017).
According to media reports, U.S. officials actively discouraged several U.S. allies, including Australia and South Korea, from joining the Bank. These efforts, however, failed. On March 12, 2015, the British government declared its intention to join the AIIB as a founding member, which opened the door for other U.S. allies to follow suit. Expressing frustration, an anonymous representative of the Obama administration accused the U.K. government of "constant accommodation" of China (Glaser & Vitello, 2015, p. 26). Eventually, AIIB founding members numbered 57, including close U.S. allies such as Australia, Israel, South Korea, Germany, France, and Italy. Besides the United States, the only major economic power that chose not to join the AIIB was Japan, the main actor in the ADB. Other GGIs also joined in the legitimation of the new MDB. World Bank President Jim Yong Kim in April 2015 welcomed the AIIB as a "new major player." He labeled the AIIB and the BRICSbacked NDB "potentially strong allies" for the World Bank (Donnan, 2015). The widespread support for the AIIB, in terms of membership and supportive statements from various actors, indicates that the U.S. campaign to delegitimize the new MDB failed. The Economist characterized the AIIB story as one of "a victorious campaign against American-led skepticism" (quoted in Etzioni, 2016, p. 174).
When most of its strongest allies had announced their support for the AIIB, the U.S. government began to soften its rhetoric. Realizing that it had lost the legitimacy struggle on the AIIB, the Obama administration appeared to accept the new Bank. Instead of voicing worries and concerns about the AIIB, the new U.S. message was to propose cofinancing between the AIIB and the World Bank and ADB (Glaser & Vitello, 2015). President Obama voiced conditional support for the AIIB for the first time in a press conference on April 28, 2015. He said that he wanted to "dispel this notion that we were opposed or are opposed to other countries participating [in the AIIB]" and added that if the AIIB "ends up having . . . safeguards, is run in a way that ultimately is actually going to lead to good infrastructure and benefit the borrowing countries, then we're all for it" (Glaser & Vitello, 2015, p. 26).
Hence, when it was clear that the AIIB would indeed be established, and with the support of most Asian and Western governments, more specific criticism replaced the overall delegitimation attempts. Instead of challenging the establishment of the new GGI, critics focused on specific aspects of its governance structure. A number of CSOs have been particularly active in this respect. Civil society networks that had for long criticized other MDBs were quick to also engage the AIIB. For instance, the NGO Forum on ADB started to monitor the AIIB as well.
CSOs have typically not questioned the existence of the new MDB, but raised serious concerns about its exercise of authority in relation to specific governance policies. This criticism, while presumably intended to affect legitimacy beliefs about the AIIB, cannot be seen as a fundamental challenge to the AIIB's legitimacy as the critics do not question the rightfulness of the GGI's exercise of authority at large. Rather, by participating in the Bank's consultation processes, they implicitly contribute to the legitimation of the new institution. Nevertheless, civil society critics have pointed out serious flaws in AIIB's institutional design and policies. Compared with other IFIs, the AIIB is accused by CSOs to have less legitimacy. As noted by Diehl (2017), "/f/ or a long time, environmental and human-rights organisations have been complaining about established IFIs only applying inadequate environmental and social standards. Now they accuse the AIIB of failing to meet even those standards." Hence, self-legitimation attempts by the AIIB have faced delegitimation attempts mainly from the U.S. government and transnational CSOs monitoring IFIs. Legitimation and delegitimation have focused on institutional features of the new GGI and on discursive statements referring to different sources of legitimacy. The rest of this section will address these two dimensions in turn.

Institutional (De)Legitimation
The process of establishing a new GGI is very much about institutional design. Legitimacy concerns are naturally tied to various institutional features of the new GGI. A GGI in the making needs to develop institutional features that are perceived as legitimate by important legitimacy granting audiences. The AIIB was created in a context of already existing IFIs with certain more or less established norms and standards. "Environmental protection, social responsibility, and corruption avoidance have become broadly accepted global norms in the international system" (Chin, 2016, p. 12). The establishment of AIIB must be understood within this broader social structure of established norms and institutions (cf. Bernstein, 2011). The actors behind the creation of the AIIB could hardly ignore this normative context and it was natural to also draw on the experiences of established GGIs. AIIB President Jin Liqun argued that the AIIB is a "new type of development bank," governed according to the "highest possible standards," learning from the experiences of existing institutions (Rosenzweig, 2016). When critics voiced concerns that the new Bank would not meet established standards, the Chinese counterinitiative was to recruit former senior IMF and World Bank staff to the AIIB and give them the task to design the new GGI's institutional structure (Mackintosh, 2016;Park, 2017).
To establish the AIIB as a GGI enjoying widespread international legitimacy, China had to compromise with major Western states (Wilson, 2019). In this process, a number of institutional legitimation practices can be identified. Many legitimation practices, however, were contested in what I call legitimacy struggles. The AIIB's formal structure of authority has been the subject of both positive and negative legitimacy beliefs, depending on audience. In many ways, the AIIB has a similar setup as other IFIs, but one notable exception is that, unlike the World Bank, ADB, and other similar institutions, the AIIB does not have a resident board of directors. AIIB directors are not based at the AIIB headquarters in Beijing, putting them in a weaker position in relation to the AIIB management (Mackintosh, 2016). Moreover, unlike in other similar GGIs, directors and alternate directors are unpaid (Chin, 2016), further indicating their more marginal status. Skeptics have accused Beijing of trying to give the Bank's management more unchecked power over the institution. For instance, a former senior U.S. treasury official remarked that "I understand why the [US] or other advanced countries prefer the resident board . . . . We do not trust the likely management" (Chin, 2016, p. 16). A former U.S. representative on the ADB Board of Directors commented that the lack of a resident board of directors in the AIIB was "a major flaw in the AIIB governance structure" (Orr, 2016). Without a permanent board that has day-to-day oversight, there is a risk that the AIIB will be much more centrally controlled by the Chinese government and there will be less transparency and accountability, he argued. Moreover, according to him, a resident board would provide civil society with easier access and a voice (Orr, 2016).
Others have defended the decision to have nonresident directors, arguing that this makes the institution more efficient. Echoing the arguments of AIIB management, Bin (2017) suggests that having nonresident directors will not only delineate responsibility and increase efficiency; it will be helpful for management to make decisions within its capacity impartially by considering economic factors only, rather than being unduly interfered by directors, as it has been the case in other MDBs with resident directors. (p. 152) This kind of argumentation was obviously convincing from the perspective of many states that were eager to join the new MDB.
Another institutional feature on which a legitimacy struggle can be identified is participation by nonmember actors.
In line with what has become common practices of MDBs and many other GGIs, the AIIB has held consultations with various stakeholders and invited comments on major governance policies, including its ESF, transparency policy, energy strategy, and proposed complaints mechanism. However, many CSOs have criticized these consultation processes for failing to be democratic and inclusive. Already in October 2015, before the AIIB started operations, a number of CSOs led by Article 19 called for the proposed new MDB to engage with civil society in a more transparent and accountable manner (Article 19 et al., 2015). Referring to the AIIB's draft ESF, the CSOs stated, Unfortunately, we are not able to comment on the contents of the draft ESF due to fundamental problems with the ESF consultation process which preclude meaningful participation by civil society, and especially those communities who will be most affected by AIIB's activities. (Article 19 et al., 2015).
The CSOs specified their "serious concerns and critical recommendations" in terms of the too short timeline of the consultation process, the lack of release of relevant documents in local and regional languages, the limitation to online consultations, and the limitation to only one round of consultations. They further demanded that /c/onsultation meetings should ensure accessibility for persons with disabilities, be gender sensitive and culturally appropriate, and allow for anonymous contributions by civil society members who may fear repercussions for participation. (Article 19 et al., 2015) Similar concerns were raised by CSOs concerning consultations on AIIB's public information policy (CEE Bankwatch Network, Both ENDS, & RE: COMMON, 2016) and Energy Strategy (BothEnds et al., 2016). The latter took place in the period 2016 to 2017 and indicated that the AIIB had made some changes in line with some, but far from all, civil society demands. The consultation now had two phases. During the first phase, the Bank received over 40 responses from government agencies, private sector companies, and CSOs, which it claimed to incorporate in a draft energy strategy that was posted on the AIIB website for comments during the second phase of the consultation process (AIIB, 2017a).
Similarly, in April 2017, the AIIB called for public consultations concerning its proposed complaints handling mechanism. The Bank declared its eagerness to receive suggestions "from any individual, organization or stakeholder" (AIIB, 2017c). The consultation process had two phases. After the first phase, a proposal should be drafted and posted on the AIIB website for further comments during the second phase. In addition to welcoming written submissions, a series of video conferences were planned in both phases. The participation of a broad spectrum of stakeholders in these consultation processes was obviously seen by the AIIB leadership as potentially boosting its legitimacy. In his speech during the 2017 Annual Meeting, President Jin Liqun stressed the importance of AIIB outreach activities to include a broad set of stakeholders. "Over this past year, the Senior Management Team has been reaching out to the media, NGOs, CSOs, and other stakeholders for meaningful dialogue and information-sharing about the Bank," he said (AIIB, 2017f). However, CSOs monitoring the AIIB remained critical of the Bank's consultations with CSOs and other stakeholders. In a letter to the Director General of the Compliance, Effectiveness, and Integrity Unit (CEIU) of the AIIB, a number of prominent advocacy CSOs, including the Bank Information Center and the NGO Forum on ADB, welcomed the Bank's intention to have public consultations concerning its proposed complaints mechanism. However, the CSOs criticized the procedures and demanded that the AIIB "conduct a two-phase public consultation process that is consistent with the practices of other international financial institutions (IFIs)" (Accountability Counsel et al., 2016). More specifically, the CSOs criticized the use of English as the only language, the lack of in person meetings, and the short time frames for submission of input. Hence, institutional features enabling genuine and inclusive stakeholder participation in AIIB's policy making has remained an important focus of legitimacy struggles between Bank management and civil society critics.
Another contested institutional feature relates to (the lack of) transparency. When the AIIB began operations in January 2016, the Bank had a provisional public information policy in place. While committing to promote transparency and accountability, the document also stresses "protecting confidentiality" as a guiding principle. The AIIB is said to have a "responsibility to restrict access to information whose disclosure could cause harm to specific parties or interests, particularly its members, clients, and co-financiers" (AIIB, 2016). This statement is likely to boost the Bank's legitimacy in the eyes of certain borrowing states and private company partners, but it might affect the legitimacy beliefs of many other audiences negatively.
Civil society critics argued that AIIB's public information policy is below the standards of other IFIs. The obligations to publish documents are quite limited compared with other IFIs. There are too many vague exceptions and no harm-test. There is a lack of specificity on the timing and content of the Board documents to be released, the interim policy does not say anything about implementation, and there is no provision for an appeal to an independent oversight body (Business and Human Rights Resource Centre, n.d.; CEE Bankwatch Network et al., 2016;Mendel & Summers, 2016). A 40-page detailed report prepared by the Centre for Law and Democracy and the Bank Information Center concludes that /d/espite some strengths, in many areas the Interim Policy fails to meet the minimum information disclosure standards established by international law and the practice of other international financial institutions (IFIs). As a result, there is a clear need for fundamental revision of the rules regarding the disclosure of information at the AIIB. (Mendel & Summers, 2016) Yet another institutional aspect that might enhance the legitimacy of a GGI is cooperation with other GGIs. The AIIB has been keen to cooperate with the World Bank, ADB, and other MDBs. While this might certainly be motivated by functional requirements to coordinate activities for the sake of efficiency, it can also be seen as a way of countering arguments that the new Bank is competing with the established MDBs. In April 2017, World Bank Group President Jim Yong Kim and AIIB President Jin Liqun signed a memorandum of understanding to strengthen cooperation and knowledge sharing between the institutions (AIIB, 2017g). The way AIIB has tried to publicly draw attention to this cooperation indicates that it is seen as boosting the Bank's legitimacy. In his speech during the 2017 Annual Meeting, President Jin Liqun stressed the importance of partnership with other GGIs. He said that "/s/since the Bank's inauguration, we have enjoyed close partnership with our MDB partners-we have co-financed operations with a number of MDBs and are working, collaboratively, with them and others to support global initiatives and platforms" (AIIB, 2017f).
Also relying on the legitimacy of other actors is the creation of institutional bodies recruiting prominent people whose support can boost the legitimacy of the GGI. Such an institutional setup is the AIIB's "International Advisory Panel." Established in October 2016 to "support the President and senior management on the Bank's strategies and policies, as well as on general operational issues," it consists of prominent people with experience of the operation of IFIs, including researchers and former ministers (AIIB, 2017e).
In sum, in the process of establishing the AIIB, the design of specific institutional features has been contested in legitimacy struggles featuring self-legitimation by the AIIB management and delegitimation mainly by U.S. officials and transnational CSOs monitoring IFIs. While choosing a similar governance structure as established IFIs in many respects, the AIIB differs in having nonresident directors. Representatives of the Bank see this as more efficient and something that can boost AIIB's legitimacy, particularly among borrowing member states, but critics have argued that this strengthens the power of AIIB management and decreases transparency and accountability. Similar to many other GGIs, the AIIB has held consultations with CSOs and other stakeholders concerning major policies. This is something that has been promoted as legitimacy boosting by the AIIB management, but the way these consultations were conducted has been severely criticized by CSOs. Institutional mechanisms for transparency and accountability have been set up by the AIIB as part of the GGI's attempt to gain legitimacy. However, these institutional features have also been criticized by CSOs for being flawed and not meeting the minimum standards of other IFIs. Despite such criticism, the AIIB has attracted a large membership of states from all over the world. Partnership with established MDBs such as the World Bank and ADB also indicate that major global authorities view the new MDB as legitimate.

Discursive (De)Legitimation
The AIIB slogan "lean, clean and green" indicates an attempt to legitimize itself in terms of both procedure and performance. While no reference to democratic performance can be found in AIIB self-legitimation statements, democratic procedures feature quite frequently in the Bank's discursive legitimation. On its website, the AIIB declares that "/g/ood governance is our hallmark and we strive to operate at the highest possible standards in governance, transparency and accountability" (AIIB, 2017b). Democratic and fair procedures were prominent themes in President Jin Liqun's speech at the 2017 Annual Meeting. Referring both to accountability and impartiality as central values for the AIIB (2017f), he said, We are clean-good governance is crucial to our credibility and the key to success. No compromise is allowed in this regard. We hold ourselves fully accountable for the Bank's management, and adhere to full compliance with the Bank's rigorous governance standards and principles. We operate transparently and have zero tolerance for corruption.
Meanwhile, in civil society discourses on the AIIB, references to both procedure and performance appear and CSOs point out shortcomings related to democracy and fairness as well as technocracy. Hence, a broad spectrum of legitimacy sources is referred to. The CSOs have been particularly concerned about flaws in democratic procedures of AIIB policy making, as indicated in the previous section's references to civil society criticism against allegedly flawed consultation processes and a general lack of opportunities for meaningful stakeholder participation.
The AIIB sometimes refers to fair performance as a way of legitimizing specific projects funded by the Bank. For example, a loan to build access roads to approximately 4,000 villages in all 33 districts of Gujarat, India, was legitimized in terms of providing . . . approaches to educational institutions, schools and hospitals. The upgraded road access is expected to have a positive impact on women and girls by improving school attendance rates for girls who currently drop out of school due to a lack of access to all weather roads. (AIIB, 2017d) In his speech at the 2017 Annual Meeting, President Jin Liqun said, "We care about those who might otherwise be left behind without our support, hence our rigorous implementation of the safeguards. We care about poor people's access to urban infrastructure, hence the slum upgrading in Indonesia" (AIIB, 2017f). However, CSOs monitoring the AIIB have questioned AIIB's commitment to justice and equality. Concerning the specific project in Indonesia that the AIIB President referred to, CSOs claimed that the human rights of women and indigenous people were violated in this project (Diehl, 2017).
The AIIB's claimed efficiency and expertise-that is, its legitimacy derived from technocratic procedures-are captured in the "lean" part of the slogan, "lean, clean and green." As put by AIIB President Jin Liqun, "/w/e are lean-we operate as an agile and responsive institution, with a focused management team, clearly defined mandate and accountability, and a core of talented, highly experienced, and dedicated staff" (AIIB, 2017f).
However, the most common source of discursive selflegitimation of the AIIB is technocratic performance. The AIIB has been established to "foster sustainable economic development and create wealth and improve infrastructure connectivity in Asia, and to promote regional cooperation and partnership in addressing development challenges" (AIIB, 2016). The AIIB's capacity to solve problems of lack of infrastructure and deliver coordinated infrastructure solutions for the benefit of whole societies is stressed in the presentations of all the Bank's activities. According to the AIIB President, "/t/he importance of quality investment in infrastructure and regional connectivity cannot be overstated" (AIIB, 2017f).
Civil society discourses criticizing the AIIB rarely seem to refer to procedures related to fairness or technocracy. The democratic procedures discourse highlighting participation and accountability is most prevalent. Meanwhile, when it comes to performance, CSOs are most concerned with values that can be related to fairness. Main concerns related to fairness include AIIB's allegedly weak commitment to sustainable energy (BothEnds et al., 2016) and human-rights violations in the context of AIIB-funded projects (Diehl, 2017).
The AIIB's approach to investment in the energy sector has been particularly controversial. The AIIB has not yet approved investment in coal power, but the Bank does not rule out the possibility to fund "efficient and clean" oil and coal-fired power plants (Liu & Tang, 2017). When the Bank refers to "modern energy" in its Energy Strategy, this does not necessarily mean clean and renewable energy, as pointed out by civil society critics. In a response to critics, AIIB President Jin Liqun, in what can be understood as a relegitimation attempt, stated, "there are no coal projects in our pipeline, and we will not consider any proposals if we are concerned about their environmental and reputational impact" (AIIB, 2017f).
Environmental impacts of AIIB-funded projects and other substantial issues are obviously important for CSOs monitoring the Bank, but the relatively strong focus of civil society advocacy on providing input on AIIB governance policies have made CSOs focus more on the lack of democratic procedures than on the actual performance of the new MDB. This is not surprising in the initial phase of establishing the new GGI. As the AIIB engages in more development projects, CSOs are likely to focus more on the performance of the Bank.
One early example of this is a critical evaluation of one of AIIB's first projects, the Tarbela 5 project in Pakistan (Bank Information Center, 2017). In cooperation with the World Bank and the Pakistan government, the AIIB will boost production at an existing hydro-dam and link it to the national grid via new transmission lines. The Tarbela hydropower project, however, has a troubled history. Tens of thousands of people were displaced between the 1960s and 1990s. To this day, many thousands of families remain impoverished and are still seeking fair compensation and redress for their losses. In the new project, the AIIB and the World Bank have promised compensation for previous harm but, according to the critical report, redress will be severely limited. Moreover, consultation and access to information have not followed best practice. For instance, translations of crucial project documents in Urdu were only made public following inquiries to the banks. Hence, the civil society report focuses not only on (the lack of) fair performance but also the (lack of) democratic procedures.
In sum, discursive legitimation and delegitimation have focused a lot on democratic procedure. Issues of stakeholder participation and accountability are arguably of particular importance in the early phase of establishing a new GGI when choices of institutional design are high on the agenda. The most common source of discursive self-legitimation of the AIIB, however, is technocratic performance. Meanwhile, delegitimation discourses among CSOs, in addition to democratic procedure, mainly focus on fair performance. References to fair performance can also be found in AIIB self-legitimation statements, but they do not appear as frequently as technocracy. Hence, while both legitimation and delegitimation refer to most types of legitimacy sources, legitimation tends to rely more on technocratic performance, whereas fair performance is a main theme of delegitimation statements.

Concluding Remarks
This article has proposed a dynamic analysis of legitimation and delegitimation as a way forward for research on legitimacy in global governance. Much previous research has analyzed either legitimation or delegitimation. This case study of the AIIB clearly demonstrates how legitimizers and delegitimizers respond to each other. GGIs, their supporters, and their critics are involved in legitimacy struggles in which delegitimation practices are met by self-legitimation attempts and vice versa. Such legitimacy struggles feature both institutional and discursive (de)legitimation practices.
The AIIB faced serious delegitimation attempts from the United States before it was established and a number of prominent transnational CSOs have continued to critically monitor the new MDB. The AIIB leadership has engaged in self-legitimation, both discursively and through the design of various institutions. However, institutional features intended to boost the legitimacy of the AIIB have been criticized and delegitimized by external actors. The AIIB's decision to have nonresident directors is one example of a contested institutional practice. Similar to many other GGIs, the AIIB has held stakeholder consultations on major policies. While these can be understood as self-legitimation practices, the way these consultations were conducted has been severely criticized by CSOs. Institutional mechanisms for transparency and accountability have been set up by the AIIB as part of the GGI's attempt to gain legitimacy. However, these institutional features have also been criticized by CSOs. Despite such criticism, the large membership and close cooperation with established MDBs indicate that the AIIB is widely considered a legitimate GGI.
Concerning discursive (de)legitimation related to the AIIB, both legitimation and delegitimation statements refer to democratic procedure as a source of legitimacy. Problems of stakeholder participation and accountability are probably especially significant for a new GGI. The most common source of discursive self-legitimation of the AIIB, however, is technocratic performance. Meanwhile, delegitimation discourses among CSOs, in addition to democratic procedure, mainly focus on fair performance. Hence, legitimizers and delegitimizers tend to put different weight on different sources of legitimacy although they share a focus on democratic legitimacy.
The AIIB's self-legitimation has been successful in the sense of securing support from most major states and other GGIs in the field of development. This finding raises the more general question on who is likely to win a legitimacy struggle and why. While a thorough answer to this question is beyond the scope of this article, I offer some tentative ideas, which could be fruitfully developed in future research. First, the relative success of different legitimation and delegitimation practices in influencing legitimacy beliefs, and thus leading to a victory in a legitimacy struggle, obviously depends on the audience. AIIB legitimation has been successful in relation to all the states that have joined the Bank as members, but less successful among civil society actors, many of whom tend to view the AIIB as less legitimate than other MDBs.
Second, agents of (de)legitimation are more or less powerful and this influences the likelihood of winning a legitimacy struggle. States and GGIs (including the AIIB itself) have been most successful in the legitimacy struggles analyzed here, whereas civil society actors appear to have less influence on the legitimacy beliefs of influential audiences. However, a very powerful state-the United States-can be seen as a major loser of the legitimacy struggle surrounding the establishment of the AIIB, suggesting that it is not only about the overall power of the (de)legitimation agent.
Third, the qualities of the practices may influence the outcome of legitimacy struggles. In general, institutional practices are likely to be more effective than discursive practices as speech acts that are not accompanied by any real institutional change are less convincing. Legitimation statements by the AIIB have probably been more successful when they have referred to real institutional features.
Fourth, the outcomes of legitimacy struggles are also likely to be related to the substance of the GGI's activities. If a particular audience has significant grievances concerning what the GGI does, self-legitimation practices without any real transformation are unlikely to change legitimacy beliefs (cf. Hurd, 2018). Those civil society actors who are deeply concerned about the overall activities of MDBs and believe that such GGIs contribute to sustaining a fundamentally unjust world order will not change their minds because of anything the AIIB or its supporters say.
Last, but not least, we need to take broader social structures into account (Bernstein, 2011;Scholte, 2018). As argued by , "the outcomes of the 'legitimation game' are influenced not only by the interests and power of those who take part in that game, but also by the social structures in which legitimation takes place" (p. 38). Legitimation of a GGI is easier if its general institutional features and mission are in line with dominant norms. The AIIB is a new GGI but, in most respects, it is not a new type of GGI. It is mainly modeled after established intergovernmental organizations in the field of development and tend to adhere to dominant norms in the neoliberal global order. This has certainly contributed to its relative success in gaining legitimacy in the view of powerful states, corporations, and GGIs.
The economic power of China and self-interests of states joining the AIIB obviously offer alternative explanations to the relative success of the AIIB, but this article suggests that an understanding of how specific legitimation and delegitimation practices play out in complex legitimacy struggles provides a more nuanced picture. As such, it complements and, to some extent, moves beyond existing geopolitical and political economy accounts of the establishment of the AIIB. Unlike most previous scholarship on the AIIB, this article has also highlighted the role of nonstate actors. Civil society actors are typically more skeptical to the new GGI than are government officials or business actors. Nevertheless, the major transnational CSOs targeting the AIIB do not challenge the overall authority of the GGI. Rather, their willingness to take part in the Bank's consultation processes indicates an implicit acceptance of the AIIB's overall authority as legitimate even if they are very critical of specific policies and projects (cf. Gregoratti & Uhlin, 2018).
Legitimacy struggles are arguably especially prominent in the processes of establishing new GGIs. This case study of the AIIB indicates that gaining legitimacy might be more challenging than maintaining legitimacy. When a new institution is established, fundamental choices have to be made concerning different governance procedures, and so forth. Such institutional choices are open for contestation and debate to an extent that they are not likely to be in wellestablished institutions. When the likelihood of influencing institutional design is larger, it is more likely that there will be more intense (de)legitimation practices. An example of this is the controversy surrounding the AIIB Board of Directors. Other features of new GGIs, as opposed to wellestablished GGIs, is that they can relate to, draw on the experiences of, and be compared with established GGIs in the field. The AIIB has deliberately tried to learn from and collaborate with other MDBs and it is monitored and assessed according to what is considered best practices within the governance of development. A further difference between new and old GGIs is that performance is less of a legitimacy concern in the process of establishing a new GGI. Legitimacy struggles in this early phase tend to focus more on fundamental choices of institutional design and procedures. This pattern is very clear in the case of the AIIB.
The empirical findings in this article are by necessity preliminary. The analysis is limited to the process of establishing the AIIB and its first 2 years of operation. Future research will have to follow up on new developments and challenges of the Bank. The analytical framework has proved useful in demonstrating the broad variety of (de)legitimation practices and highlighting their contentious nature and how legitimation and delegitimation practices are linked. Future comparative research could aim at explaining the variation in (de) legitimation practices across GGIs and other agents of (de) legitimation. More comparative research is also needed on the effectiveness of different (de)legitimation practices in terms of influencing legitimacy beliefs. Moreover, it might be fruitful to systematically compare the legitimacy struggles of the AIIB with legitimacy struggles of established GGIs to further theorize the specific legitimacy challenges related to the creation of new GGIs.