Contested multilateralism

“Contested multilateralism” describes the situation that results from the pursuit of strategies by states, multilateral organizations, and non-state actors to use multilateral institutions, existing or newly created, to challenge the rules, practices, or missions of existing multilateral institutions. It occurs when coalitions dissatisfied with existing institutions combine threats of exit, voice, and the creation of alternative institutions to pursue policies and practices different from those of existing institutions. Contested multilateralism takes two principal forms: regime shifting and competitive regime creation. It can be observed across issue areas. It shapes patterns of international cooperation and discord on key security concerns such as combating terrorist financing, halting the proliferation of weapons of mass destruction, and banning certain conventional weapons. It is also evident on economic issues involving intellectual property, on environmental and energy issues, and in the realm of global public health. The sources of dissatisfaction are primarily exogenous, and the institutions used to challenge the status quo range from traditional treaties or intergovernmental organizations to informal networks, some of which include non-state actors. Some institutions are winners from the process of contested multilateralism; others may lose authority or status. Although we do not propose an explanatory theory of contested multilateralism, we do suggest that this concept provides a useful framework for understanding changes in regime complexes and the strategies that generate such changes.

discord as a way to generate international regimes that achieve goals of states through reducing the costs of transactions and providing relevant information (Keohane 1984). A newer strand of literature points out that international regimes are often not integrated-indeed, that some erosion of integrated regimes has occurred in recent years-and that substantial contention over the terms of multilateralism has created more fragmented regime complexes in place of cohesive international regimes (Raustiala and Victor 2004). 1 We argue in this paper that reconciling these insights is aided by a conception of contested multilateralism, which emphasizes that contemporary multilateralism is characterized by competing coalitions and shifting institutional arrangements, informal as well as formal. Multilateralism is not essentially cooperative and characterized by integrated rules. Nor is the alternative to established multilateralism, simply unilateralism or bilateralism. Frequently, multilateral institutions are challenged through the use of other multilateral institutions, either without resort to unilateralism or bilateralism or in conjunction with those strategies. Although states and non-state actors are often committed to multilateral strategies, they may disagree about the policies that multilateral institutions should pursue.
Contested multilateralism describes both strategies and situations. Strategies of contested multilateralism, pursued by states, multilateral institutions, and non-state actors, use formal and informal multilateral practices to challenge established multilateral institutions. They provide ways to promote policy and institutional change that are quite distinct from unilateral or bilateral policies designed to challenge and change the status quo.
As a description of situations, contested multilateralism refers to situations involving contestation in which the contending parties propose multilateral responses that are to some degree incompatible. In this connection, the concept of contested multilateralism helps us to understand that a variety of phrases such as "regime complexity," "forum shopping," or "regime complexes," essentially describe different aspects of the same generic phenomenon. 2 They all refer to the creation or use of formal or informal institutions or networks that disrupt the status quo. 3 The purpose of this paper is to demonstrate the value of the concept of contested multilateralism as a way of describing both strategies and situations. In the current political context, characterized by increasing diffusion of power resources away from the democratic West toward the East and South, contested multilateralism may be particularly relevant. 4 Joseph Jupille, Walter Mattli, and Duncan Snidal have recently developed an impressive analytical framework to analyze issues of institutional choice, focused on (in their terms), usage, selection, change, and creation. They argue, as others have before them, that leaders often exhibit bounded rather than synoptic rationality in assessing alternative institutional designs, and they use their analytical framework to "account for the remarkable stability of the status quo" in a varied set of situations in which institutions had to cope with major changes in their environments (Jupille et al. 2013: 213). 5 Our discussion here is complementary to their analysis. When change in existing multilateral institutions is difficult, building coalitions to contest the status quo is likely to be more promising than unilateral resistance.
Contested multilateralism involves the use of different multilateral institutions to challenge the rules, practices, or missions of existing multilateral institutions. More precisely, the phenomenon of contested multilateralism occurs when states and/or nonstate actors either shift their focus from one existing institution to another or create an alternative multilateral institution to compete with existing ones. Such challenges may or may not entail the creation of new multilateral institutions, but they always involve conflict between the rules, institutionalized practices, or missions of two different institutions. When these challenges to dominant institutions are successful, they typically increase the complexity of an international regime by adding elements to it or strengthening formerly weaker institutions with it. These challenges may constitute reactions to actions of multilateral institutions, but they may also stem from anticipation that established institutions will not respond in ways that the challenging actors view as satisfactory. In this paper we discuss a variety of instances of contested multilateralism, in issue areas ranging from counter-terrorism to global health.
In using the terminology of "contested multilateralism," it is important to emphasize that what is contested is not the institutional form of multilateralism as such-to the contrary, this generic form is accepted-but rather specific institutional forms of multilateralism, challenged by the contesting coalition.
Contested multilateralism can be observed across many different issue areas in international relations, including issue areas such as counter-terrorism and nonproliferation that are central to state security. Contested multilateralism is so common in contemporary world politics that the strategies of dissatisfied powerful states on core issues of national security often run through multilateralism rather than in stark opposition to it. Observers emphasizing power politics might expect that powerful dissatisfied states will resort to unilateral or bilateral options, while institutional scholars such as Randall Stone (2011) have argued that powerful states can use informal means to control existing multilateral institutions. We argue that even powerful states may be stymied by established institutional practices, yet when this situation arises, they often resort to multilateral rather than unilateral or bilateral practices.
Section 1 defines contested multilateralism more precisely; lays out essential conditions for contested multilateralism to exist; describes the pathways through which we expect it to appear; and distinguishes two basic types, regime shifting and competitive regime creation. Section 2 then describes some examples of regime shifting and Section 3 presents cases of competitive regime creation. Our cases are drawn from issue areas that range from intellectual property, energy, and health to security issues involving conventional weapons, counter-terrorism, and nuclear non-proliferation. Under existing IR theory, these examples do not constitute any single theoretical phenomenon but rather represent a range of different institutional processes that scholars have observed. This framework is intended to show what these processes have in common, and suggest that they may be analyzed collectively through the lens of contested multilateralism. 6

Contested multilateralism defined
This section begins by elaborating our definitions. It then continues by specifying some key baseline conditions for contested multilateralism, describing pathways for its emergence, and distinguishing two basic types of contested multilateralism. Finally, it concludes by highlighting how strategies of contested multilateralism can expand regime complexes.
Three criteria define a situation as involving contested multilateralism: 1. A multilateral institution exists within a defined issue area and with a mission and a set of established rules and institutionalized practices. 2. Dissatisfied with the status quo institution, a coalition of actors-whether members of the existing institution or not-shifts the focus of its activity to a challenging institution with different rules and practices. This challenging institution can be either pre-existing or new. 3. The rules and institutionalized practices of the challenging institution conflict with or significantly modify the rules and institutionalized practices of the status quo institution.
Dissatisfied intergovernmental organizations, civil society actors, and weak states are likely only to be able to act effectively to counter the policies of established multilateral institutions with the aid of states, so for them, multilateralism is often the only way to contest effectively such policies. Powerful states, however, may have bilateral or unilateral options. But even they often have incentives to act multilaterally, both to mobilize support and attendant resources, and to gain legitimacy for their contestation of established multilateral policy.
Although contested multilateralism always involves the use of either an established or new multilateral institution to challenge a status quo institution, the outcome of such a challenge is open-ended. Challenges can collapse without long-term impact, leaving a regime complex effectively unchanged. Often, however, these challenges create or expand a regime complex in a way that leads to fundamental changes in institutional practices or changes the distribution of power between institutions. Some challenges may have an impact only over a long time horizon. A dissatisfied coalition composed primarily of weak states, for example, may only be able to mount a symbolic challenge, critiquing an existing institutional practice but being unable to force immediate change. Over time, however, such a challenge may change actor preferences, ideas, and values in a way that delegitimizes an institution, forcing concomitant change or institutional exit. 7 Contested multilateralism refers to conflict between, not within, multilateral institutions. Conflicting resolutions by the UN Security Council and the UN General Assembly do not constitute contested multilateralism, nor do conflicting actions by the European Parliament and the European Council, or adverse judgments on the legitimacy of such actions by the European Court of Justice. Conflicting actions by these bodies represent internal disagreements, not contested multilateralism, and do not change the nature of regime complexes. Nor does contested multilateralism describe "emanations," institutions created from other institutions (Shanks et al. 1996), or situations in which one multilateral institution has already granted authority in a particular domain to another such institution, which then makes decisions invalidating the first institution's actions. For instance, if the WTO Appellate Body rules against the EU in a trade case, this is not contested multilateralism. Contested multilateralism occurs when states, and sometimes also non-state actors, are dissatisfied with an existing institution, find pathways to intra-institutional reform blocked, and decide that it is worthwhile either to shift their focus to other institutions or to create a new one.
Our case studies will illustrate that contested multilateralism is common in world politics and that this concept provides a new way to interpret key events and describe more coherently the processes leading to more fragmented regime complexes. Viewed through the lens of contested multilateralism, significant events such as the North Atlantic Treaty Organization's (NATO) use of force against Serbia in the Kosovo conflict take on new meaning. Such action was significant not just because NATO countries acted without a UN Security Council resolution authorizing use of force, but also because by doing so, countries expanded the scope of NATO's mission and challenged the UN Security Council's monopoly as a legitimizing authority for the use of force. Ultimately, NATO's action changed the institutional hierarchy governing the use of force, creating a nascent regime complex. When NATO partners contemplate future actions, such as intervention in Syria, there is now a precedent for circumventing the Security Council. Such actions can be part of long-term strategy to force institutional change rather than simply efforts to undermine institutional authority. The concept of contested multilateralism therefore emphasizes that the central strategic question for states is rarely "multilateralism vs. unilateralism," but rather what kind of multilateralism will best achieve long-term objectives.
In terms made famous by Albert O. Hirschman, contested multilateralism seems to combine "exit" and "voice" (Hirschman 1970). Hirschman emphasizes that the 7 For example, in 1976 the IMF approved three loans to South Africa, leading the UN General Assembly to issue a disapproving resolution calling on it to refrain from dispensing money to South Africa. Such action constituted symbolic contested multilateralism, demonstrating widespread dissatisfaction with the IMF's approach to apartheid but posing no real prospect of institutional change. Yet the General Assembly's action may have contributed to a larger ideational shift within developed countries, culminating in restrictions on IMF loans to South Africa (International Policy Report 1984). More generally, on the role of ideas in world politics see Finnemore (1996), Goldstein and Keohane (1993) and Katzenstein (1996). prospect of exit may amplify voice, an observation that Laurence Helfer applies to exit from treaties by pointing out that "withdrawing from a treaty (or threatening to withdraw) can give a denouncing state additional voice" (Helfer 2005(Helfer : 1588. However, strategies of contested multilateralism are not limited to exit and voice, since they often involve withdrawing support, or doing so partially, rather than actually leaving an organization with which a coalition is dissatisfied. And some forms of contestation can actually be designed to strengthen an existing institution, at least from the standpoint of some participants. 8

Pathways to contested multilateralism
Typically, dissatisfaction of states or other actors with multilateral institutions is generated by exogenous changes in the global environment or in state preferences. Several of the instances of dissatisfaction to be discussed below resulted from pressure on participating states by domestic interests, international institutions, or transnational activists, or from unexpected events such as the terrorist attacks of September 11, 2001. Occasionally, however, dissatisfaction can be generated endogenously by engrained practices of established multilateral institutions.
When dissatisfied actors are unable to change the status quo there is a possibility of contested multilateralism. Whether the dissatisfied coalition opts for a strategy that generates contested multilateralism will depend on the availability of an outside option, the degree to which communication is impeded by lack of credibility, and the existence of institutional or domestic constraints. The possible pathways to contested multilateralism are illustrated in Fig. 1 and discussed below.
When a dissatisfied coalition-composed of states or a combination of states and non-state actors-seeks to change a blocked institution, the coalition's ability to pursue outside options is a necessary condition for successful contested multilateralism. 9 Outside options include switching to an already-existing alternative multilateral institution or creating a new institution more in line with the coalition's preferences, with respect either to substantive policy or institutional form-formal or informal, hierarchical or networked, integrated or loosely coupled. The key is that the challenging coalition must have an alternative to the existing institution that would serve its interests, and the coalition must be able to credibly threaten to use this alternative organization or practice. State power is a major determinant of whether coalitions have outside options. A group of dissatisfied actors that includes states with significant resources and institutional leverage will have an easier time identifying credible outside options than a coalition of weaker actors.
If dissatisfied actors have an outside option, we should normally expect adaptation by the existing institution, since its authority and the scope of its impact will be adversely affected by the establishment of alternative organizations or practices.
However, existing institutions may fail to adapt. We posit that such failure to adapt occurs through one of two pathways.
First, since dissatisfied actors have incentives to claim that they have outside options even when they do not, their incentives to misrepresent can generate problems of credibility. If the dissatisfied coalition is unable to make credible threats and promises about either its willingness or ability to leverage outside options to force change, it may take actions that result in contested multilateralism. This is one of the pathways in Fig. 1. 10 Even in situations where the dissatisfied coalition is credible, policy adjustment may not occur if there are diverging state interests, or ideational or institutional constraints. Veto players-states, organizational bureaucracies, or other crucial actors-may block changes desired by the challenging coalition because they see their interests-on substantive policy or with respect to institutional authority-threatened by change. Established conceptions of institutional roles and purpose are often persistent. This situation, in which power comes to the fore, is represented by the other pathway to contested multilateralism in Fig. 1.  The most familiar form of contested multilateralism is regime shifting, which occurs when challengers to a set of rules and practices shift to an alternative multilateral forum with a more favorable mandate and decision rules, and then use this new forum to challenge standards in the original institution or reduce the authority of that institution. We borrow this term from Laurence Helfer, who suggests that regime shifting works by "broadening the policy spaces within which decisions are made and rules are adopted" (Helfer 2009: 39). This process may then generate feedback effects. Regime shifting can be found across issue areas, and includes the effort by developing countries and UN agencies to challenge the TRIPS agreement on intellectual property, the creation of the Basel Committee and other innovations in global financial regulation, and the World Bank's entry into global health governance.
Helfer's discussion of regime shifting is focused on state-led action; however, we find that such challenges can also be led by institutions. In such cases, autonomous multilateral institutions issue rules or policies that directly conflict with international standards established by other multilateral institutions. Institution-led regime shifting is a significant development for international politics because institutions are not subject to the same political pressures as states, and therefore may be willing to go further in their challenges. In our example of institution-led regime shifting, the ECJ directly challenged the UN Security Council's counter-terrorism policy, despite the fact that the policy had been endorsed by France and the United Kingdom and was a key component of the global fight against terrorism.
The second form of contested multilateralism is competitive regime creation, which occurs when the coalition of dissatisfied actors creates a new institution or establishes a new informal form of multilateral cooperation to challenge the existing institutional status quo. In this form of contested multilateralism, the challenging coalition will first create a new multilateral institution or forum that more closely represents its interests. This may be achieved by limiting membership to a select group of states with similar preferences (Downs et al. 1998), by establishing informal channels to influence institutional policy over areas of significant concern (Stone 2011), or by creating new transgovernmental networks (Slaughter 2004). The coalition will then use the new institution to challenge existing organizations or networks, generating discord that may or may not lead to inter-institutional cooperation. Examples of competitive regime creation discussed below include the Proliferation Security Initiative (PSI), the International Renewable Energy Association (IRENA), and decisions by donor countries, IGOs, and private foundations to challenge the WHO's monopoly of authority over health issues by creating the Global Alliance for Vaccines and Immunization (GAVI), the Global Fund to Fight AIDS, Tuberculosis and Malaria, and UNAIDS. 11 When a dissatisfied coalition successfully engages in contestation, it creates, reinforces, or expands a regime complex. If the regime is currently integrated and hierarchical, then such a challenge creates a regime complex. Whether the challenge to the status quo institution involves creating a new institution or shifting responsibility for a specific issue to an institution that previously had no authority in such matters, the regime will become less integrated. If a regime complex already exists, then successful challenges will reinforce regime complexity either by strengthening competitor institutions or by creating new institutions.

Regime shifting
When a coalition is dissatisfied with institutional rules or policy and there are alternative institutional forums available in a given issue area, it can choose to shift to a multilateral venue with a more favorable mandate or decision rules. By creating alternative rules or institutionalized practices that conflict with the standards in the original institution or reduce its authority, actors may initiate a process that changes the institutional status quo (Helfer 2004(Helfer , 2009Helfer and Austin 2011). Helfer, who developed the concept of regime shifting, focuses on actions initiated by states in conjunction with non-state actors; we term such cases 'state-led' regime shifting. We expand Helfer's concept to include 'institution-led' regime shifting in which IGOs initiate strategies that contest established multilateral institutions and practices.

State-led regime shifting
Developed countries' efforts to shift authority from the World Intellectual Property Organization (WIPO) to the GATT/WTO negotiating forum provide a clear example of regime shifting. Developed countries grew dissatisfied with WIPO as a negotiating forum for intellectual property after a series of events in the 1970s and 1980s. In the late 1970s, US companies became concerned about the lack of intellectual property protection in developing countries, and began to push for new international standards. When US companies went to WIPO, however, to advocate for updates to the Paris Convention on patent protection, they encountered significant resistance from developing countries (Santoro 1995). By the early 1980s, US companies had convinced the US government and leaders of other developed countries of the need for more stringent patent protection, but they were unable to bring about significant change within the WIPO forum. During patent negotiations in the 1980s, developing countries demanded a revision of the Paris Convention to lower, rather than increase, the minimum standards of intellectual property protection applied to developing countries (Yu 2005). With countries pushing in opposite directions and a one-country, one-vote forum in WIPO, developed countries concluded that they needed a different venue to strengthen the intellectual property regime. 12 Developing countries did not anticipate the full effects of the shift to the GATT/WTO forum for intellectual property negotiations. In advance of the Uruguay Round, countries set out objectives for negotiations and stipulated their intention to create a multilateral intellectual property agreement "without prejudice to other complementary initiatives" such as WIPO. 13 Developing countries misread these objectives and thought they would be able to limit the scope of negotiations primarily to trade in counterfeit goods (Watal 2001). Instead, the United States was able to convince countries through a combination of multilateral and unilateral actions (such as the threat of trade sanctions) to support broad-based intellectual property negotiations within the GATT/WTO. Developing countries were so concerned about being excluded from the new free trade system that they acceded to the shift of venue (Watal 2001).
The Agreement on Trade Related Aspects of Intellectual Property Rights (TRIPS) was adopted in 1994. The final agreement was almost identical to a draft put forward by the GATT Secretariat, which had prepared a "take it or leave" it document. In effect, the developed countries that had negotiated the WTO forced developing countries to accept this draft or be excluded from the benefits of tariff reductions (Steinberg 2002;Yu 2005). TRIPS sets minimum standards for intellectual property to be implemented by all WTO members and allows enforcement through the WTO dispute settlement mechanism.
Since the establishment of TRIPS, WIPO has struggled to remain relevant. In 1994 and 1995, the WIPO General Assembly adopted resolutions requiring WIPO to assist its members with fulfilling TRIPS obligations and offering non-WIPO members compliance assistance as well (May and Sell 2006). In recent years, WIPO has attempted to regain control of this issue area and to convince OECD countries that it is the right forum for intellectual property regulations (May and Sell 2006). WIPO has been somewhat successful at reestablishing itself as the primary venue for negotiating intellectual property treaties (May 2010), as the stalled WTO negotiations no longer offer a viable alternative forum for developed countries.
In this case, the developed countries' challenge to WIPO occurred due to a strong outside option-the WTO negotiations-and an informational gap: developing countries did not accurately understand the motivations and intentions of developed countries. Developed countries were thus able to shift intellectual property negotiations to the WTO, enhancing their bargaining leverage and creating an intellectual property regime complex (Helfer 2004). 14 3.2 Institution-led regime shifting IGOs can also play an essential role in regime shifting, taking the lead in shaping the content of an institutional challenge. Institution-led regime shifting illustrates that IGOs may be able to create outcomes that would be difficult or impossible for a coalition of states to achieve.
Our example of institution-led regime shifting is the challenge by the European Court of Justice (ECJ) to the UN Security Council sanctions regime regulating terrorist financing. This case study shows that autonomous international institutions such as courts may engage in contested multilateralism without guidance or initiative from states. The UN Security Council's failure to anticipate the ECJ's challenge illustrates how credibility can be a key determinant of contested multilateralism.

UN anti-terrorism policy and domestic due process: the Kadi Case
The UN Security Council has made targeted sanctions a core component of its global anti-terrorism policy. Under the "1267 regime," named for Security Council Resolution 1267 that first established it, UN member states are required to freeze the assets of all listed individuals and entities affiliated with Al-Qaida. Although the 1267 regime was initially focused on a small number of Taliban officials and Osama bin Laden, in the weeks following 9/11 the Security Council added close to 100 names to the sanctions list. 15 By January 2002, the Security Council had broadened the sanctions criteria so that countries could propose for inclusion on the targeted sanctions list any individuals or entities "associated with" Osama bin Laden, Al-Qaida, or the Taliban (UNSCR 2002, 1390. From its earliest days, the 1267 regime suffered from incomplete implementation by member states. Few countries had legal and administrative structures in place to actually enforce the measures (Cortright and Lopez 2002). The Council provided states with only small amounts of identifying information, making it difficult to differentiate the individuals in question from others with similar names or birthplaces. Moreover, some countries remained completely ignorant of their obligations under the regime (1267 Monitoring Group Report 1 2002: 5).
For many established democracies, however, the biggest implementation challenge was the fact that the Security Council obligations conflicted with domestic rule-of-law principles. The regime's lack of due process became apparent in 2002 when Sweden pushed for and failed to obtain the removal of three Somali-born Swedish citizens who were added to the sanctions list immediately following 9/11. 16 Arguments about due process in the 1267 regime became part of a broader discussion in the UN about the need to incorporate human rights principles into global counterterrorism cooperation. 17 The Security Council, however, did not modify the 1267 regime to address these concerns in any significant way for many years. Court challenges were few in number and generally affected only a handful of states. Although a UN body concluded in 2005 that "the many legal challenges to the measures, in particular in Europe and the United States, pose a serious impediment to the success of the sanctions regime" (1267 Monitoring Team Report 2005: 16), the Security Council made only minor adjustments to the regime.

Outside options and credibility problems
Although some states were dissatisfied with the UN Security Council's counterterrorism policy as early as 2002, these actors did not have a viable outside option and 15 Interview by Julia Morse of a member of the 1267 Monitoring Team, 15 March 2012. 16 The Swedish government's request was initially blocked by three permanent members of the Security Council. Sweden then entered into bilateral negotiations with the United States and eventually all three names were removed from the list (Heupel 2009: 310, citing Cramer 2003 therefore were unable to push through major change. This situation changed when European courts acted as an alternative pathway for individuals seeking to challenge the legitimacy of the regime. In December 2001, Saudi Arabian national Yassin Abdullah Kadi filed suit in the European Court of First Instance (CFI) to request annulment of the European regulation implementing the sanctions. The CFI ruled against Mr. Kadi in 2005, finding it did not have the jurisdiction to review indirectly the validity of the UN Security Council regulation. But Kadi appealed the judgment to the ECJ, which had the authority to invalidate the European Union's implementing regulation and thus affect the sanctions policy of 27 countries. In January 2008, an ECJ Advocate General issued an advisory opinion suggesting that the European regulation implementing the 1267 sanctions measures infringed on the right to be heard, the right to judicial review, and the right to property (Maduro 2008). 18 Given this strong signal of an impending challenge, why did the Security Council fail to respond adequately? With France and Great Britain as permanent members of the Council, it may seem surprising that the Council did not prevent an adverse ECJ judgment by modifying the 1267 measures and the concomitant EU regulations. We infer that the credibility of an adverse ECJ decision, which would be supported by EU members, was not sufficiently high to make a pre-emptive response seem desirable.
In December 2008, the Court struck down the European Union regulation implementing the Al-Qaida sanctions measures, ruling that Yassin Abdullah Kadi and co-complainant the Al Barakaat Foundation, were denied the right to defense, the right to a legal remedy, and the right to property (Kadi 2008). Although the decision was stayed and the European Union passed a new resolution to implement the measures against Kadi and Al Barakaat, it now became clear that the whole sanctions regime was at risk.
In light of the ECJ's adverse judgment and the prospect of challenges to the revised EU rules from Kadi and Al Barakaat, the Security Council looked for new ways to improve due process. Member states had begun discussing proposals for an outside review mechanism in 2005, when Denmark proposed the creation of an Ombudsperson to receive delisting requests directly from listed individuals and entities and to make recommendations on delisting to the Council. At the time of the proposal, few within the UN system believed it would ever be adopted. 19  the European Union's 2008 regulation implementing the 1267 sanctions, holding that it violated Kadi's rights to defense, judicial review, and property. The Court acknowledged that judicial review of Security Council measures at the national or regional level could potentially encroach on the Security Council's prerogatives under Chapter VII and Article 103 of the UN Charter, 21 but followed the precedent set out by the ECJ in 2008 (Ginsborg and Scheinin 2011: 15). The European Commission appealed the General Court's judgment to the ECJ.
The Security Council now had received multiple signals from the European courts-if the Council did not make significant changes to the 1267 regime, the ECJ would likely make it impossible for European states to implement the sanctions. In other words, the issue of credibility had been resolved. In June 2011, the Security Council expanded the powers of the Ombudsperson by allowing her to make recommendations, rather than simply observations. More significantly, if the Ombudsperson reviews a delisting request and recommends delisting to the 1267 Committee, the Committee must have consensus to overrule the delisting request (UNSCR 2011(UNSCR : 1989. Prior to this point, the Committee needed consensus to remove an individual from the sanctions list; thus by changing the procedure, the Security Council shifted the burden of proof onto states seeking to retain an individual listing. The ECJ has successfully forced significant change in Security Council procedure. Since the appointment of the Ombudsperson in 2010, 26 individuals have been delisted through this process, and only 3 requests have been denied (Office of the Ombudsperson 2013). In October 2012, the Security Council removed Kadi from the 1267 sanctions list. In March 2013, an ECJ Advocate General issued an advisory opinion, recommending the ECJ set aside the judgment of the General Court and dismiss the action brought by Kadi. In the opinion, the Advocate General writes, "The judgment of the Court of Justice in Kadi led to the establishment of the Office of the Ombudsperson, which has made it possible to raise the quality of the list considerably. It would be paradoxical if the Court failed to take account the improvements to which it has directly contributed" (Bot 2013). 22 The ECJ, however, did not follow the advice of the Advocate General, and instead has launched its most direct challenge yet. In July 2013, the ECJ issued its decision on the Kadi II appeal, upholding the European General Court's decision to strike down the EU regulation implementing sanctions against Kadi. Not only did the ECJ affirm its right to review all EU measures implementing UN Security Council resolutions, but it also suggested that nothing short of an external court will provide sufficient due process for listed individuals. 23 By proposing an outside body to review UN Security 21 Chapter VII of the UN Charter gives the Security Council responsibility for the maintenance of international peace and security. Article 103 of the Charter establishes the supremacy of the Charter over all other international agreements. 22 Full opinion available at: http://curia.europa.eu/juris/document/document.jsf?text=&docid= 135223&pageIndex=0&doclang=EN&mode=req&dir=&occ=first&part=1&cid=1765372. Accessed on 13 August 2013. 23 According to the ECJ, "The essence of effective judicial protection must be that it should enable the person concerned to obtain a declaration from a court, by means of a judgment ordering annulment whereby the contested measure is retroactively erased from the legal order and is deemed never to have existed, that the listing of his name, or the continued listing of his name, on the list concerned was vitiated by illegality…" (Paragraph 134). The full text of this decision is available at: http://curia.europa.eu/juris/document/document. jsf?text=&docid=139745&pageIndex=0&doclang=en&mode=req&dir=&occ=first&part=1&cid=1914364. Accessed on 13 February 2014.
Council sanctions decisions, the ECJ is advocating a regime complex on targeted sanctions. 24

Competitive regime creation
When a dissatisfied coalition cannot turn readily to an established multilateral institution to challenge the practices that it opposes, it may create a new multilateral forum with different rules and practices that more closely align with its interests. To ensure that the new forum has a more favorable policy orientation, the coalition may opt for direct control strategies, such as limiting membership to like-minded states, establishing informal channels of control, or structuring voting in a favorable manner. Alternatively, the coalition may create an institution with a mandate more consistent with the interests of its members. By challenging the institutional status quo, the dissatisfied coalition creates additional bargaining leverage vis-à-vis the status quo institution.

State-led competitive regime creation
Although competitive regime creation is a new concept, existing work has documented many cases where states take the lead in creating new multilateral forums to affect the rules or mission of existing institutions. In the area of migration, for example, wealthy states have sought to change their obligations to accept migrants as refugees by creating a migration regime complex that limits the impact of their obligations under the 1951 UN refugee convention and the practices of the UN High Commissioner on Human Rights (Betts 2009). In the area of biodiversity, developing countries wanted to preserve more autonomy to regulate genetically modified organisms (GMOs) in the face of the creation of the WTO in the 1990s. These countries took the lead in negotiating the Cartagena Protocol on Biosafety in 2000, which had a "disruptive" effect on the WTO-based regime (Oberthur and Gehring 2006: 14). 25 Our principal illustrations of state-led competitive regime creation are the creation of the International Renewable Energy Agency (IRENA) and the development of the Proliferation Security Initiative (PSI), both initiated by powerful states; and the reaction to the Convention on Certain Conventional Weapons by a large European-led coalition, leading to the conventions on land mines and cluster munitions. 24 In proposing this standard, the ECJ may have pushed too far with its challenge. The Security Council may instead opt for creative ways around the legal challenges, such as making sanctions regimes less targeted. For more on the ECJ case and the UN Security Council's potential response, see: http://www.ejiltalk.org/kadishowdown/. More generally, we accept the comment of one referee that this case is somewhat different from our others in that the ECJ can be seen as at a "lower level" of operation than the Security Council, rather than strictly parallel to it. It therefore borders on non-compliance, but a particular kind of institutionalized and multilateral non-compliance. Moreover, this form of non-compliance was strategic, designed to push the UN Security Council to change its procedures, as evidenced by the comments of Advocate General Bot (cited above). 25 On earlier related events, see Raustiala and Victor (2004).

Contested multilateralism in the global energy regime complex
There is no integrated global energy regime; rather, there is a regime complex comprising a set of partially overlapping and non-hierarchical institutions (Raustiala and Victor 2004;Colgan et al. 2012). The recent creation of the International Renewable Energy Association (IRENA) to challenge the International Energy Agency (IEA) exemplifies competitive regime creation within a regime complex.
Although the IEA's mandate includes environmental protection and economic development, it has been criticized for its lack of commitment to renewable energy sources. In the early 2000s, several international meetings were held to discuss the increasing demand for cooperation on renewable energy. 26 By 2009, dissatisfied states had gathered enough support to found IRENA, despite strong resistance from the IEA Secretariat and the United States. Seventy-five states were signatories at the founding conference in Bonn in January of that year; by January 2013 IRENA had well over 100 signatories, including the United States and all but two other members of the IEA (Van de Graaf 2013a, b).
The pathway to IRENA was due to institutional constraints that weakened the IEA's ability to adapt to change. Founded in 1974 in response to the first oil shock, and focused from the outset on fossil fuels, IEA membership rules and procedures have been "frozen in time" (Colgan 2009: 3). Germany, supported especially by Denmark and Spain, viewed the IEA as unable effectively to support renewable energy because of its emphasis on fossil fuels and therefore took leadership in establishing IRENA. Van de Graaf writes: "Berlin was convinced that the IEA's mandate was too broad to ensure that the agency could provide a visionary leadership on renewable energy. The German government judged that the IEA's analyses tended to focus on large-scale energy supply, 'without offering advice on adapting markets to more decentralized forms of energy' and 'without fully reflecting the potential of renewable energy" (Van de Graaf 2013b: 163).
The IEA seems to have responded to IRENA's creation by being more positive about renewable sources of energy, and there are signs of a rapprochement between these two potentially rival multilateral institutions (Van de Graaf 2013b: 170). While the shortterm effect of IRENA's creation has been to expand the global energy regime complex, harmonization between IRENA and the IEA will affect the nature of this complex. Whatever happens between these two organizations in the future, IRENA's creation is clearly a case of the form of contested multilateralism that we designate as competitive regime creation.  Keohane (1984), which emphasize the tendency of multilateral institutions to expand, or to have additional institutions grafted onto them, in order to reduce transaction costs.

Contested multilateralism in the global nonproliferation regime complex
A second example of state-led competitive regime creation is the US-led creation of the Proliferation Security Initiative (PSI), which we interpret as a challenge to the common interdiction practices put in place by the UN Convention on the Law of the Sea (UNCLOS). 28 UNCLOS rules, not those of the Nuclear Nonproliferation Treaty (NPT), were at issue since UNCLOS governs maritime conduct and the sea is the main mode of transport for illicit nuclear and ballistic missile technology. Under Article 110 of the UN Convention, a country that suspects proliferation activities has little ability to act against a vessel suspected of transporting illegal weapons without the consent of the state whose flag the vessel flies. 29 Countries trafficking in weapons of mass destruction (WMD) 30 materials can take advantage of this provision by registering their vessels in states the provide little regulation or oversight, or can use vessels flagged by states that continually refuse consent to the exercise of high seas jurisdiction by others (Byers 2004: 527).
The United States became concerned about rules inhibiting search and seizure on the high seas in 2002. Following a tip from US intelligence, a Spanish warship interdicted a North Korean vessel, the So San, and upon boarding, Spanish and US officials found Scud-like missiles and warheads. 31 But the ship was released two days later because there was no lawful basis for seizing the cargo. 32 For the United States, the incident highlighted the disparity between US strategic objectives and the existing UNCLOS rules.
US officials drew two major lessons from the So San incident: that, despite implementation challenges, multilateral cooperation to counter proliferation was feasible, and that the existing international legal and institutional structure for dealing with proliferation threats was ineffective. The United States did not want to undermine UNCLOS, since the regime is essential for the projection of US power worldwide (Kraska 2012), and modifying UNCLOS was impractical and infeasible since the US Senate has never ratified the treaty. But the United States needed to promote a new operational norm, one that would allow it to work around the flag-state consent requirement to impede global proliferation. To address this perceived deficiency in the operation of the UNCLOS regime, the United States began developing plans for a new informal institution: the Proliferation Security Initiative. 28 It is important to note that the PSI was a challenge to UNCLOS, not the NPT. 29 Article 110 stipulates that ships cannot be boarded unless the ship is unregistered (no country-specific flag), the ship is registered in the state that wishes to board it, or the ship is engaged in piracy, the slave trade, or unauthorized broadcasting. 30 We define WMD as "atomic explosive weapons, radioactive material weapons, lethal chemical and biological weapons," and any other weapons that have "characteristics comparable in destructive effect to those of the atomic bomb or other weapons mentioned above" (UN Commission on Conventional Armaments, as quoted in Carus 2012). 31 Spanish authorities were able to seize the ship temporarily because of maritime irregularities, including nationality and papers not being in order, a false manifest, and the vessel's refusal to submit to inspection (Karon 2002). 32 Some legal experts have argued that under existing case law, the US could have been allowed to detain the ship and perhaps even confiscate the cargo as "contraband of war" (Rivkin Jr and Casey 2003).

The path to contested multilateralism in the PSI
As in the middle pathway in Fig. 1, there was a dissatisfied coalition and the established institution was constrained from adjusting sufficiently to avoid contestation. Furthermore, the United States and its allies had outside options due to superior intelligence and naval capabilities that allowed more direct anti-proliferation action in cooperation with smaller countries. Rather than negotiating a change to international law, the United States hoped to force change by transforming these smaller countries' national legal authorities (Koch 2012).
The United States announced the PSI in May 2003 with a group of ten partner countries: Australia, Italy, France, Germany, Japan, the Netherlands, Poland, Portugal, Spain, and the United Kingdom. In September, these eleven countries adopted a PSI mission statement centered on stopping shipments of WMD and related materials in a manner consistent with existing national and international legal frameworks. The states committed to interdicting transfers to/from states of proliferation concern, improving information exchange, strengthening national legal authorities for interdiction, and taking specific action in support of interdiction (PSI Statement of Interdiction Principles 2003). 33 The PSI is designed to promote a broader shift in the multilateral nonproliferation infrastructure, changing UNCLOS practices rather than its principles. PSI participants have found new ways to combat proliferation while acting in accordance with the law. For example, the PSI calls upon participating states to conclude ship-boarding agreements with other states, especially those which are known to act as flag-of-convenience states (Malirsch and Prill 2007). These agreements facilitate the process of interdiction and make it significantly easier for powerful countries to board flagged vessels suspected of carrying illicit materials.
From the outset, the US attempted to institutionalize some of these changes in other forums. Koch suggests that the Bush administration rushed to reach agreement on core principles by September 2003 so that President Bush could bolster his proposal to the UN General Assembly for a new UN Security Council resolution on proliferation (Koch 2012). Indeed, in his speech, President Bush pointed to the PSI as part of US efforts to improve interdiction capabilities and called for "the broadest possible cooperation" to stop would-be proliferators. 34

Institutional responses to the PSI
After months of contentious negotiations, the Security Council adopted Resolution 1540 on nuclear proliferation in April 2004. Although the United States may have originally intended the resolution to provide a more explicit legal basis for the PSI, the final resolution contains no mention of the initiative: at China's insistence, all references to "interdictions" were removed from the final text (Holmes 2005). 35 Despite these gaps, Resolution 1540 was a significant step toward institutionalizing the operations of the PSI. By passing a Chapter VII resolution on proliferation, the Council officially recognized proliferation as constituting a threat to international peace and security. The resolution incorporates language calling upon states, "to take cooperative action to prevent illicit trafficking in nuclear, chemical or biological weapons, their means of delivery, and related materials" (UNSCR 2004, 1540. Although legal scholars suggest this does not constitute a legal obligation but rather an invitation (Joyner 2005), it pushes the legal status quo closer to the PSI. 36 Over the last decade, the legal basis for the PSI's activities has also expanded as the nature of the proliferation regime complex has changed. The 2005 protocol to the Convention for the Suppression of Unlawful Acts Against the Safety of Maritime Navigation prohibits shipping of WMD and allows for the boarding of a ship suspected of proliferation (but only with flag state consent). In September 2010, the International Civil Aviation Organization adopted a treaty focused on proliferation via airplanes. Most importantly, the Security Council has adopted two resolutions (1874 and 1929) that include important interdiction components aimed toward stopping North Korean and Iranian proliferation activities respectively.
Today, the PSI's legality and legitimacy seem secure, but its institutionalization remains incomplete. As of November 2012, the PSI had more than 100 participants, including strategically important states like Russia, South Korea, and Yemen. 37 The PSI has conducted close to 50 interdiction exercises since 2003 and has reportedly had dozens of successful interdictions (Koch 2012). But the PSI still lacks a formal institutional structure and an explicit legal mandate. Support is widespread, but it does not include key countries such as China, Brazil, India, and South Africa. Although the PSI has not been formally incorporated into the UN-centered regime, its creation expanded the regime complex on proliferation, clearly altering the practices of interdiction at sea.

Contested multilateralism on conventional weapons issues
The Convention on Certain Conventional Weapons (CCW) was, in the mid-1990s, the only international treaty that regulated landmines, but its provisions were extremely weak. Despite strong support among European countries for more stringent standards, it proved difficult, in a series of review conferences between 1995 and 2006, to make any significant changes. In the mid-1990s, NGOs and states interested in banning landmines began separate fast-track negotiations. The Ottawa Convention On Land- 35 China has raised concerns about the legitimacy of the PSI from its inception. In December 2003, an official from the Chinese Foreign Ministry said in a press conference: "there are many concerns in the international community about the legitimacy and effectiveness of PSI interdictions…China consistently holds the view that proliferation issues should be resolved within the international legal frameworks." As quoted in Ye Ru'an and Zhao Qinghai (2004). "The PSI: Chinese Thinking and Concern." The Monitor 10.1: 22. 36 Indeed, in December 2003, US Under-Secretary of State John Bolton argued that if a scenario occurred where PSI countries have ambiguous authority to act, a Security Council resolution could provide additional authority (Boese 2003). 37 A full list of participants is available at: http://www.state.gov/t/isn/c27732.htm. Accessed on 13 August 2013.
Mines was signed in December 1997 and entered into force less than two years latermore quickly than any treaty of its kind in history (Goose and Williams 2006). A similar process, led by Norway, culminated in the signing of the Oslo Convention on Cluster Munitions in December 2008 (Bolton and Nash 2010). These new treaties created regime complexes governing certain types of conventional weapons, in which the primary institutional body-the CCW-had standards that conflicted with those of the more comprehensive land mine and cluster munitions conventions. Interestingly, both cases illustrate the use of contested multilateralism against the dominant military powers (the United States, China, and Russia all opposed these new conventions).

Institution-led competitive regime creation
We have seen, with the Kadi case, that intergovernmental organizations can play key roles in regime shifting, and they can also be pivotal in the creation of new institutions to challenge a regime complex. Our case studies of what we call "institution-led competitive regime creation" focus on how coalitions of states and non-state actors have challenged the World Health Organization (WHO) over the last thirty years, leading to a proliferation of regime complexes in global health. 38 Since the San Francisco Conference of 1945, the WHO has been the central UN organization dealing with health issues, and for the first thirty years of existence its preeminence was largely unchallenged. But this began to change in the late 1970s and 1980s when the G8 began to consider health issues, marking the start of a period in which the WHO faced a series of challenges to its authority. Coalitions of states and non-state actors initiated challenges to WHO pre-eminence on three major issues in the 1990s, each of which we discuss below.

The global alliance on vaccines and immunization
Beginning in the early 1980s, many public health officials became committed to the idea of harnessing advances in biotechnology to produce revolutionary vaccines for debilitating diseases in the developing world. As the central multilateral institution for global health issues, WHO was the natural place to coordinate the effort. WHO established two separate vaccine programs, one focused on basic research, the other on delivering vaccines to children in the developing world. The flaws in this bifurcated approach to vaccine development became apparent as the programs grew. Coordination between these programs was weak, and neither had a mandate to facilitate the commercial production of usable vaccines.
By the late 1980s, health advocates and non-governmental organizations were calling for a more holistic mechanism for vaccine development. In 1990, a coalition of five non-state organizations (UNICEF, WHO, UNDP, the World Bank, and the Rockefeller Foundation) reached an agreement to form a new Children's Vaccine Initiative (CVI), designed to advance the full continuum of vaccine development. WHO at first resisted the CVI, but once it became clear that the initiative was inevitable, WHO managed to make the CVI a subordinate entity within the WHO (Muraskin 1998). At this point, therefore, contested multilateralism was avoided. 38 We thank Tyler Pratt especially for his excellent research assistance on the issues discussed in this section.
But the CVI did not thrive within the WHO, which prioritized its own vaccine programs, constrained the CVI's attempts to partner with private industry, and appeared satisfied to let the CVI flounder. Once again, external concern grew about mismanagement of the vaccine effort within the WHO. In the late 1990s, a coalition of concerned parties made a high profile attempt to wrest control of the vaccine initiative from WHO and move it to the World Bank. WHO leaders, perhaps underestimating the extent of dissatisfaction with the organization, responded by trying to defeat the challenge and reassert WHO dominance over vaccines. In 1999, "WHO shockingly announced that the CVI would be dismantled and terminated by the end of the year" (Muraskin 2002: 152).
Civil society organizations, public health advocates, and some elements of the vaccine industry responded vigorously to the WHO response. Not long after the termination of the CVI, in 2000, the Global Alliance on Vaccines and Initiative (GAVI) was created outside the structure of WHO, with significant support from the Gates Foundation (Muraskin 2002). GAVI has been much more successful than CVI was in working with industry and civil society; it receives funding from capital markets, donor countries, and its partners, which include WHO as well as the Gates Foundation and the World Bank (Harman 2012).
GAVI emerged, in terms of the flow-chart in Fig. 1, because of organizational resistance by the WHO and incomplete information that prevented the WHO from accurately gauging the seriousness of the challenge. WHO's resistance to the CVI in the face of a powerful supportive coalition helped generate contested multilateralism-the creation of GAVI, which challenged WHO's dominance of vaccine development and delivery. WHO has continued to function as a partner in GAVI, but the challenge created a regime complex that significantly reduced WHO's authority over vaccine coordination.

The creation of UNAIDS
WHO was slow to recognize and respond to the threat posed by HIV/AIDS. It initially viewed the disease as a problem for the richest countries and dismissed the need to become involved. 39 Only in 1986, five years after the first Note on AIDS by the US Center for Disease Control, did WHO seek funding to develop activities on AIDS. In 1987 WHO was designated the "lead agency" in the fight against AIDS and shortly thereafter, it established the Global Programme on AIDS (GPA).
WHO's leadership role in addressing the AIDS epidemic led to an incomplete and ineffective policy response by the international community, partly because WHO focused on the medical rather than social aspects of the disease. WHO's medical approach was shaped by its organizational culture and extensive experience combating outbreaks of infectious disease. Notably, this medical approach endured despite efforts by the first GPA Director Jonathan Mann to employ a human rights perspective in addressing AIDS (Knight 2008: 15). WHO was simply unable or unwilling to work collaboratively with other UN agencies to address the full impact of AIDS on societies.
By the early 1990s, WHO's inability to effectively manage the global response to AIDS was apparent to donor countries, which began to shift resources away from WHO toward bilateral channels. The GPA attempted to address coordination problems with other agencies, proposing a yearly forum open to all stakeholders, but donor countries did not think this idea went far enough (Mann et al. 1992;Knight 2008: 20). Sweden proposed the creation of a Task Force representing all key stakeholdersdonor countries, developing countries, UN agencies, and civil society organizationsto design a new body for coordinating the UN's efforts on AIDS. The Task Force proposed replacing the WHO's leadership role with a joint, cosponsored program, although UN agencies on the task force may have only accepted this outcome because donors threatened to divert AIDS funding to bilateral channels (Knight 2008: 23).
Despite resistance from UN agencies, particularly WHO, UNAIDS became operational in January 1996. Many of WHO's AIDS-related activities, such as the collection and dissemination of statistical information, were immediately shifted to the new organization. This shift in institutional authority was accompanied by a change in policy priorities. UNAIDS placed a heavy emphasis on building national capacity to respond to HIV/AIDS, and its staff began a political advocacy campaign to build national support for HIV/AIDS programs. At the country level, UNAIDS reorganized the delivery of UN assistance programs into collaborative "Theme Groups" and promoted a broad multi-sectoral approach to HIV prevention. Though UNAIDS lacked the technical assistance resources of WHO, it made a concerted effort to build partnerships with a wide range of entities, encouraging regional technical resource networks and adding several cosponsors during its first five years (Knight 2008).
UNAIDS emerged because of growing dissatisfaction with WHO's narrow medical approach to AIDS and the presence of organizational constraints which prevented WHO from successfully accommodating demands for change. The creation of UNAIDS has been described as "a significant blow to the organization's leadership over a global health issue," but UNAIDS has not been entirely successful at gaining control over AIDS policy in the UN system (Lee 2009: 62). Following the creation of UNAIDS, WHO took steps to reassert its role in global HIV/AIDS efforts, broadening its framework for AIDS response and launching a new initiative focused on the provision of antiretroviral drugs in low-and middle-income countries (Lee 2009). WHO maintains a relatively large HIV/AIDS department, and other institutions like the World Bank and UNDP continue to implement their own AIDS-related programs (Kohlmorgen 2007). Creating UNAIDS effectively just added one more actor to the already broad regime complex on HIV/AIDS. Perhaps for this reason, Jonsson suggests that UNAIDS biggest success may be not in the realm of policy coordination but rather as a norm entrepreneur (Jonsson 2012). Through outreach and education, UNAIDS has been able to change attitudes among stakeholders and reduce the stigmatization of vulnerable communities.

The global fund to fight AIDS, tuberculosis, and malaria
In the second half of the 1990s, the G8 increased its focus on health issues, promoting a shift away from the WHO-centered medical approach and toward a broader socioeconomic conception of global health (Harman 2012). Several G8 members, especially the United States and Japan, supported the goal of combating infectious diseases but sought to circumvent the UN/WHO system, which they considered inefficient, overly bureaucratic, and unable to partner with private industry. The idea for a new Global Fund-a public-private partnership to finance the fight against three major diseases (AIDS, tuberculosis, and malaria)-first made it onto the agenda at the 2000 G8 Summit in Okinawa. The official Global Fund to Fight AIDS, Tuberculosis, and Malaria was established the following year (Bartsch 2007).
The Global Fund is designed to be only a financing mechanism; it provides financial backing for projects and then partners with other bodies such as UNAIDS and WHO (as well as NGOs, civil society, and national governments) to implement them. Still, Lee describes the decision to establish the Global Fund as a separate entity from WHO as "notable" and "clearly influenced by a lack of donor confidence that the former could be an efficient means of rapidly disbursing large sums of funding" (Lee 2009: 52). Kohlmorgen adds that the creation of the Global Fund can be seen "as a challenge to the WHO in terms of direct cooperation and coordination" (Kohlmorgen 2007: 131).
The construction of the Global Fund is a clear example of contested multilateralism, since it was designed not simply to substitute for WHO but to change WHO's authority and practices-as its partnering with WHO suggests. As a condition of providing financing, the Global Fund requires projects to be implemented by multi-stakeholder partnerships called Country Coordinating Mechanisms, which include representation from the private sector. This condition has forced WHO to adjust the manner in which it implements programs, and it has reportedly been a source of conflict between the two organizations (Kohlmorgen 2007).
Organizational constraints within WHO prevented it from adapting sufficiently quickly to forestall contestation in the form of competitive regime creation. The Global Fund now dominates multilateral funding for work on AIDS, Tuberculosis, and malaria; the WHO experienced another significantly loss in authority and has adapted its practices so that it can work with the Global Fund.

Conclusions
Having developed the umbrella concept of "contested multilateralism," we now see a wide variety of examples of it, even apart from those discussed above. Control over multilateral development finance has long been contested between the World Bank, with weighted voting, and various vehicles controlled on a one-state, one-vote basis within the UN structure, such as the UN Special Fund. The Nuclear Non-Proliferation Treaty (NPT) has been supplemented by more or less informal institutions such as the Nuclear Suppliers Group or the International Nuclear Fuel Cycle Evaluation, which shifted the effects of nuclear proliferation policy in the direction favored by the United States and other established nuclear powers. Such challenges to existing multilateral institutions are often multilateral rather than unilateral or bilateral. Two dimensions of such challenges can be distinguished: situations characterized by active contestation and the strategies that states and non-state actors pursue in such situations.
As noted above, some of these challenges focus on the rules governing a set of activities-for example, more stringent intellectual property or due process standards; other challenges, such as the PSI, focus on the actual operation of a regime complex; and still others focus on organizational missions, as in the cases of IEA/IRENA and the WHO. States or coalitions of states, intergovernmental organizations, and non-state actors may support or create multilateral institutions with rules and practices that contradict those of established multilateral institutions to which they belong, or with different missions. Such attempts at contested multilateralism are remarkably common, and they are most interesting and surprising when they are led by powerful states. When successful, they typically create, reinforce, or expand regime complexes. 40 Contested multilateralism may create losers as well as winners. In the cases discussed above, losers included the World Intellectual Property Organization (WIPO) in its contestation with the World Trade Organization; the UN Security Council with respect to authority over rules governing financial controls to limit financing of terrorists; the International Energy Agency when IRENA was established; and the World Health Organization as a result of struggles over institutions with authority over measures to encourage the production and distribution of vaccines, and to combat AIDS and other diseases. 41 We have distinguished two forms of contested multilateralism-regime shifting and competitive regime creation-and have discussed several examples of each type. Our case studies illustrate that contested multilateralism is found across issue areas ranging from intellectual property to renewable energy, from health to security. In the security area we provide case studies of the US-led creation of the Proliferation Security Initiative, which transforms the implementation of the UN Convention on the Law of Sea, and of the European Court of Justice's challenge of UN Security Council rules governing terrorist financing. Table 1 on the next page illustrates how the concept of contested multilateralism helps us analyze seemingly dissimilar cases and understand their effects on the international system. The table summarizes our cases, the sources of dissatisfaction, and the implications of each case for regime complexes. In our example, the challenging coalitions all consist of developed countries, but this is not a general rule: if we had selected developing countries' contestation of World Bank dominance over multilateral development issues, or of US dominance of internet governance arrangements, we would have cases of contestation by developing countries. Most sources of dissatisfaction that we have chronicled are exogenous, with the exception of the WHO cases, in which organizational barriers prevented the institution from fulfilling the policy objectives of states. Four of our cases focus on contestation over rules; two focus on contestation over the missions or practices of established organizations. The newly selected institutions range from international legal conventions and formal organizations to informal networks or organizations including civil society organizations, such as GAVI and UNAIDS. Significantly, in each case, there were implications for the regime complex, redistributing authority or creating a regime complex when there had previously been a dominant treaty or organization. Contested multilateralism thus 40 Two of our referees pointed out differences between Europe and Asia that are relevant to contested multilateralism. Asian states have been notably less willing to embrace multilateralism, and when they do, they tend to create a new institution rather than bundle new functions into existing ones such as the European Union. The lesser Asian reliance on multilateral solutions suggests contradictory expectations about whether Asian states would be likely to rely on contested multilateralism as a strategy. On one hand, skepticism of multilateral solutions should make states more likely to pursue unilateral or bilateral approaches. On the other hand, the informal nature of institutions in Asia and the absence of any overarching body like the European Union create more opportunities for contested multilateralism. 41 We are particularly indebted to Professor Georg Nolte for focusing our attention on the issue of losers from contested multilateralism. Creation of regime complex provides a framework for understanding the growth of regime complexes across a variety of different issue areas.
What we have not done in this article is systematically to test propositions about contested multilateralism, but we think that such propositions could be formulated and tested. 42 Crises may lead to actions, led by great powers, that seem precipitous and unwise to other governments, particularly after the crises have passed. Subsequent reactions by Member States against the UN Security Council's imposition of widespread targeted sanctions implemented in the wake of 9/11 illustrate this point. Second, the existence of institutionalized centers of authority with overlapping mandates seems likely to enhance contestation, as the Kadi case suggests. Third, requirements of unanimity for change in existing multilateral institutions should be conducive to contested multilateralism by making major changes very difficult, especially if they have distributional implications-as the Proliferation Security Initiative case suggests. Finally, delegation of policy to relatively autonomous international bureaucracies-as in the health cases-creates opportunities for contestation when states are dissatisfied with bureaucratic performance but cannot readily change the offending institution. Contestation should therefore be associated with prior delegation to autonomous international organizations.
The observable implications of these hypotheses are fairly clear. Contested multilateralism should be observed in the wake of major crises; there should be more incidences of contested multilateralism in issue areas with multiple centers of authority than in issue areas dominated by a single center of authority; unanimity requirements to change institutional rules and practices should generate contested multilateralism; and delegation of authority to autonomous bureaucracies should also be associated with contested multilateralism.
We study the phenomenon of contested multilateralism because it seems to be a significant and common process, and also to throw light on issues of institutional change. When interdependence is high and power is widely dispersed, multilateral institutions are essential tools for states and non-state actors. When coalitions are dissatisfied with existing institutions, they seek to shift the focus of rule-making to other institutions or to construct new institutions for that purpose. States, intergovernmental organizations, and other non-state actors act strategically, using alternative multilateral organizations to pursue their interests when established institutions fail to do so. The result is a change in the configuration of multilateral institutions, often leading to, or expanding and reinforcing, regime complexes as opposed to integrated international regimes. The phrase, "contested multilateralism," helps us to understand both the persistence of contemporary multilateralism and the major changes that it is currently undergoing.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.