Legitimacy Deficits of International Organizations: design, drift, and decoupling at the UN Security Council

Abstract Despite ubiquitous calls for their reform, international organizations (IOs) often suffer from legitimacy deficits. What explains the emergence of legitimacy deficits and what effects do they produce? This article discusses the gradual emergence of legitimacy deficits through the concept of legitimacy drift. Legitimacy drift occurs when an organization loses legitimacy by failing to adapt itself to a changing environment. It identifies three sources of legitimacy drift: failure to live up to pre-existing standards (broken promises), changes in the standards of legitimacy by which organizations are assessed (shifting standards), and changes in an organization’s relevant public (audience shift). Legitimacy deficits typically prompt organizational responses. These include attempts at re-legitimation through institutional reform and operational adaptation, but also other ‘coping mechanisms’ such as promises of reform, the logic of confidence, and decoupling. Coping mechanisms are especially important where reform is blocked. This model is illustrated by the history of the United Nations Security Council, one of the oldest and most powerful IOs. A conclusion calls for bridging historical and sociological institutionalism to better understand IO legitimacy in time.


Introduction
Since the Second World War, international politics has seen an explosion in the number of international organizations (IOs). While new IOs continue to be created, the demography of IOs is also changing. The oldest extant IO is already more than two centuries old, 1 and while it is clearly an outlier, the IO population continues to age. Of a recent stratified random sample of 50 IOs, the average age is forty-five years (Tallberg et al. 2014). At a time of rapid changes in the international environment, IOs are becoming increasing aged (Heiskanen 2001, 1).

Legitimacy deficits of international organizations 97
Take, for example, the United Nations (UN) Security Council. Few countries are enthusiastic about the institutional design and procedures of the Security Council, and many question its legitimacy. Countries often avoid its formal procedures in favour of informal consultations and other flexibilities in its working methods. Its institutionalization of great power hierarchy conflicts with modern norms of sovereign equality and democratic decision-making. By all accounts, it is an anachronism. Yet it remains central to the international political and legal order, and states persist in seeking its reform, as they have done for decades. It suffers from a chronic legitimacy deficit (Binder and Heupel 2015;Blum 2005;Franck 2006;Hurd 2007).
To existing theories of international institutional legitimacy, IOs with persistent legitimacy deficits are, at first sight, puzzling. 2 States are said to favour acting through multilateral institutions as a means of garnering legitimacy, giving them an interest in ensuring that institutions retain this legitimating function (Claude 1966). Moreover, a lack of legitimacy must be compensated by increased material inducements or coercive imposition, which increases an institution's operating costs (Hurd 1999). Legitimacy deficits are seen as weakening the 'compliance pull' (Franck 1990, 24) of institutions, which in turn imperils the gains that supposedly flow from international cooperation. In extreme cases, members may simply abandon an illegitimate institution. If one assumes that states are able to rationally design institutions in accordance with their wishes (Koremenos, Lipson, and Snidal 2001), why is the legitimacy of so many IOs deeply contested, and why do they seem incapable of undertaking commensurate reforms?
This article addresses this puzzle by developing the concept of legitimacy drift. To do so, it adopts a longitudinal perspective on institutional development and combines the insights of rival institutionalist theories. First, in contrast to strict interpretations of rational choice institutionalism, this article conceives of legitimacy as having an important role in institutional design-legitimacy considerations are never far from the minds of those who design and set up IOs. Second, it develops the concept of legitimacy drift to conceptualize the legitimacy deficits that emerge when an institution stays the same but its normative and political environment changes. The article identifies three sources of legitimacy drift: failure to live up to pre-existing standards (broken promises), changes in the standards of legitimacy by which it is assessed (shifting standards), and changes in an organization's relevant public (audience shift). As the population of IOs continues to age, fertile ground for legitimacy drift to take hold has emerged. Third, the article posits that responses to legitimacy drift on the part of IOs include not only re-legitimation through institutional reform and operational adaptation, but also diverse coping mechanisms, such as promises of reform, the logic of confidence, and decoupling. Coping mechanisms are especially likely where reform is blocked due to vested interests or institutional lock-in.
In developing this model, the article draws on concepts from historical institutionalism, a research tradition deeply familiar with the weight of history and institutional inefficiencies, but which has until now neglected the problem of institutional legitimacy. 3 Moreover, in line with sociological institutionalism, the model emphasizes the role of legitimacy for institutional development, and organizational responses to legitimacy deficits. 4 Theorists from international relations and international law have excelled at exploring the interrelationship of power and legitimacy as they manifest in international institutions (Clark 2005;Claude 1966;Franck 1990;Hurd 2002;Ikenberry 2001;Johnstone 2003;Reus-Smit 2007), but they have not yet utilized the analytical categories or explanatory accounts offered by institutionalist theory. This article aims to bridge these theoretical approaches while highlighting the ways in which rival institutionalisms can complement rather than compete with one another. By avoiding strong claims about the logics of practice or deep socialization, it endeavours to side-step claims of ontological incommensurability (Buhari-Gulmez 2010, 263), and rather build on the assumption that goals-oriented actors operate in a social environment constituted by power-plays and economic gains-seeking as well as the pursuit of social approval and normative legitimacy (Dimaggio and Powell 1991, 33;Hanrieder 2015;Nielson, Tierney, and Weaver 2006).
In a second step, this article illustrates the theoretical model of legitimacy drift via the case of the Security Council. The model of legitimacy drift-categories describing its origins and its effects-helps to explain why the Council has such precarious legitimacy and how it has coped with its persistent legitimacy deficit. The case study draws on secondary literature and a range of primary UN documents recording speeches made in intergovernmental debates at important moments in the Council's history in the 1940s, 1960s, and mid 2000s. The debates in the 1940s relate to the Council's 'founding moment' and come from the first sitting of the General Assembly in Central Hall, Westminster, London, in January 1946. The second emerge from the mid-1960s, when the number of rotating members of the Council was expanded from six to ten. The debates in the 2000s occurred at the height of reform debates in the mid-2000s. The General Assembly, rather than the Council, is chosen as the appropriate source material as it remains the primary legitimacy constituency in whose eyes the Council appeals for legitimacy. The material illustrates the three sources of legitimacy drift developed in the model: the Council's capacity to deliver on its original criteria for legitimacy has eroded (broken promises); over time, it has become subject to a different repertoire of legitimacy demands (shifting standards); and its relevant public has changed due to the expansion in UN membership (audience shift). The absence of re-legitimation of the Council through formal reform has led to different coping mechanisms such as adaptations of its working methods, promises of reform, public relations initiatives, and decoupling.
The first section of this article presents the concepts of legitimacy and legitimacy drift, elucidates legitimacy drift's three sources, and discusses institutional responses. The second section empirically assesses the model's plausibility in the case of the Security Council. Finally, a conclusion summarizes the findings and suggests implications for legitimacy research and institutional theory in relation to IOs.

The Legitimacy of International Organizations in Time
While IOs have traditionally been understood in power-based or functionalist terms, a burgeoning body of scholarship has recognized the significance of legitimacy for IOs and other international institutions (Bexell 2014;Brassett and Tsingou 2011;Buchanan and Keohane 2006;Coicaud and Heiskanen 2001;Grigorescu 2015;Symons 2011;Zaum 2013b;Zürn 2004). The legitimacy of an institution consists of the degree to which its relevant public generally regards its features and behaviour as desirable, correct or appropriate within some socially constructed system of norms, values, beliefs and definitions (Franck 1990, 24;Hurd 1999, 381;Reus-Smit 2007, 159;Suchman 1995, 574;Zaum 2013a, 9). 5 Because an IO's basic structures and principles must resonate with social beliefs in order to be perceived as legitimate, its legitimacy is bound to the perceptions and beliefs of its internal participants and external constituents (Hurd 1999, 387-88;Zaum 2013a, 9;Zürn and Stephen 2010, 95).
Legitimacy, by definition, cannot be achieved through coercion, but legitimacy is deeply related to power. While classical realists emphasized legitimacy as an instrument of state power (Carr 1946, 132-45;Morgenthau 1948, 32), more recent scholarship has focused on the dialectic between power and legitimacy. While legitimate power is more effective than illegitimate power, legitimacy is dependent on recognition, and can be withheld when the exercise of power becomes too blatant-the requirements of legitimacy both enable and constrain power holders (Clark 2005;Claude 1966;Franck 1990;Hurd 2002;Ikenberry 2001;Johnstone 2003;Reus-Smit 2007). International organizations too are made more powerful when they have legitimacy, but they need to conform to the demands of their audiences to retain it (Clark 2011, 150-54;Cox 1983, 172-73;Hurd 2007, 128-31;Ikenberry 2001, 10;Zaum 2013a, 19-22).
Within the sociological tradition, the legitimacy of international institutions is seen to derive from various sources. Legitimacy is often parsed into procedural, performance-based, and norm-or law-based sources (eg Binder and Heupel 2015, 240-42;Franck 1990, 16-18). Procedural legitimacy accrues to an institution by following widely recognized rules of decision-making (Franck 1990). Performance-based legitimacy derives from an institution's delivering on its mandate or being guided by recognized expertise (Lake 2009a;Zürn, Binder, and Ecker-Ehrhardt 2012). These two elements of institutional legitimacy are sometimes simplified into the two categories of 'input' and 'output' legitimacy (Scharpf 1999(Scharpf , 2003. Additionally, norm-and law-based legitimacy emerges from acting 5 Following the empirical-sociological tradition, in this paper an institution is taken to be 'legitimate' quite simply when its subjects believe it to be so-what Beetham calls legitimacy 'for a social scientist ' (1991, 6). A legitimate institution in this sense may be good or bad from one's own normative point of view. For this distinction, see also Buchanan and Keohane (2006), Clark (2005, 18), and Steffek (2003, 253-54). It must be acknowledged, however, that some are sceptical of a neat separation between normative and sociological approaches to legitimacy (Beetham 1991, 11;Hurrell 2002;Zaum 2013a, 10). 100 Matthew D. Stephen in accordance with established, often law-based, norms (Barnett and Finnemore 2004;Clark 2005, 18;Steffek 2003, 251). 6 While these dimensions of legitimacy are widely recognized, no consensus on the 'correct' criteria for legitimacy in global governance yet exists (Bernstein 2011, 22). As Franck (1990, 18) explains, legitimacy is in fact 'a bracketing of many integral factors, which are related but different and which must be investigated by reference to different social data.' Rather than make strong a priori claims about the correct sources of legitimacy, adopting a sociological approach suggests that the question of which elements of legitimacy are most important for an institution ultimately depends on the beliefs and perceptions of its relevant public. These can change over time. Publics are unlikely to speak with one voice either. What makes an institution legitimate in practice is ultimately an empirical question.
While global governance scholarship has already explored these dimensions of IO legitimacy in recent years, with a few notable exceptions, 7 scholars have accorded less explicit attention to changing patterns of legitimacy over time. Moreover, institutional 'coping mechanisms' that emerge under the conditions of a legitimacy deficit-often emphasized in sociological institutionalism-remain underexplored. Attention to legitimacy 'crises' (Morris and Wheeler 2007;Reus-Smit 2007;Seabrooke 2007) may also distract from the incremental emergence of persistent legitimacy deficits. These considerations call for ways of conceptualizing how institutional legitimacy deficits can emerge gradually over time.

Legitimacy by Design
When creating IOs, state representatives seek not only to achieve certain instrumental objectives, but also to achieve institutional legitimacy (Fehl 2004;Hurd 2007, 4;Wendt 2001Wendt , 1025. This imposes limits on the range of 'rational' choices available and is a neglected dimension of institutional design (Wendt 2001(Wendt , 1039. Just as questions of institutional design (Koremenos et al. 2001) and purpose are up for negotiation during founding moments, so are considerations of how to render the institution legitimate. It is difficult, for example to explain the goal of universal membership in many international agreements without acknowledging the legitimacy dividend that emerges from this inclusiveness. Likewise, the selection of 'one-state, one-vote' decision-making rules for most IOs is hard to explain without considering the assumed fairness of such procedures (Wendt 2001(Wendt , 1025. Legitimacy needs may even (partly) explain the creation of some IOs. For example, Fehl (2004, 374-75) has argued that the International Criminal Court (ICC) was set up in part to remedy the deficient legitimacy of ad hoc UN tribunals. The bottom line is this: Designers of IOs have an interest not just in establishing institutions to serve particular functional purposes, but also in ensuring that they are perceived as legitimate, at least by those with the capacity to undermine their activities.
The design phase of an IO typically takes place within 'founding moments', which historical institutionalists suggest are deeply significant in shaping an institution's future trajectory (Fioretos 2011, 369). This is because such moments constitute critical junctures that temporarily open up possibilities for contingency and institutional innovation (Capoccia and Kelemen 2007). From a macro-historical perspective, these founding moments are associated with the aftermaths of great power conflicts (Clark 2005;Ikenberry 2001), but the same logic applies at each instance of institutional creation. Legitimacy will always be imperfect, but the flexibilities available during founding moments provide creators more freedom to accommodate legitimacy demands into the design of an institution than would be available in an already-institutionalized status quo. Especially when institutions are created with strong status quo coalitions or veto players, the decisions reached during founding moments can persist long into the future, setting institutions 'on paths or trajectories, which are then very difficult to alter' (Pierson 2004, 125). Since legitimacy depends on the beliefs of a relevant public, it follows that an institution's legitimacy must be related to the beliefs prevalent during these critical junctures.

Legitimacy Drift
The concept of legitimacy drift stems from the observation that international institutions often encounter challenges to their legitimacy as a result of the aging process (Franck 1970;Heiskanen 2001, 1-2;Zaum 2013a, 7). After their creation, IOs become embedded in changing environments to which they may not perfectly adapt (March and Olsen 1998, 954-56). In other words, 'rates of environmental change frequently outpace rates of organizational adaptation' (Dimaggio and Powell 1991, 33). As a consequence, IOs have a tendency to become less efficient or effective than a purely functionalist approach would suggest. While familiar also to sociological institutionalists, the historical institutionalist concept of 'institutional drift' is one way to capture this mode of incremental change. This describes a process in which an institution remains the same while its impact changes or becomes less effective due to shifts in the broader environment (Mahoney and Thelen 2010, 17;Streeck and Thelen 2005, 24-26). Legitimacy too must be continually maintained (Suchman 1995, 593-94). When legitimacy maintenance does not result in synchronous adaptions to new demands, an institution's legitimacy may thus be undermined, producing legitimacy drift. This refers to a situation in which an institution does not adapt to a changing political and normative environment, leading to a decline in its perceived legitimacy. 8 Legitimacy drift is a process that gives rise to the outcome of a legitimacy deficit. 9 Institutional legitimacy can increase or decrease according to the extent to which an institution meets the criteria for legitimacy in the eyes of its relevant public. This article recognizes three sources of legitimacy drift, (1) disruption in perceptions of whether an institution meets existing standards of legitimacy (broken promises), (2) changes in the standards or criteria for legitimacy applied to an institution (shifting standards), and (3) changes in an institution's relevant public (audience shift). Each of these three sources of legitimacy drift can erode the procedural, performance-based, or norm-or law-based sources of legitimacy upon which IOs are based. These three sources of legitimacy drift are comprehensive, yet they are meant to be abstract categories-their specific meaning or substance needs to be fleshed out in light of concrete cases. They can each be applied to any dimension of an institution's legitimacy. While sometimes hard to separate empirically, these processes are analytically separate and clarify the mechanisms that generate legitimacy drift.
Broken promises: One source of legitimacy drift occurs when an IO's audience perceives it as no longer meeting the standards of legitimacy originally applied to it. When an organization is perceived as not living up to its prior standards for legitimacy, it is seen as breaking its promises. Promise breaking can hurt any dimension of an institution's legitimacy. On the input side, David Beetham has discussed the example of declining proportionality in the British parliament-its first-past-the-post voting system has resulted in a loss of legitimacy as the system has exhibited an increased tendency to produce national parliaments divergent from the proportion of votes cast. This loss in legitimacy did not result from people's changing beliefs, but from the diminished ability of an institution to deliver on its existing criteria of legitimacy (Beetham 1991, 11-12). Similarly, if the legitimacy of an institution such as the Security Council is predicated on representing the countries of principal industrial and military strength, the emergence of new major powers outside the institution will lead to a decline in this aspect of its legitimacy. Regarding output, if an IO's legitimacy was originally predicated on its ability to provide public goods (a form of performance legitimacy), a decline in its capacity to provide these goods is likely to erode its legitimacy. For some observers, the perceived lack of efficacy of the World Trade Organization as a site for negotiated trade liberalization has led to a decline in its original claim to legitimacy. Broken promises thus refer to the decline of an institution's capacity to live up to its initial criteria for legitimacy.
Shifting standards: An institution may or may not live up to its original conditions for its legitimacy; however, the conditions of legitimacy that are applied to an IO by its relevant public may also change. Martha Finnemore has explored how internationally held norms and values change over time (Finnemore 1996, 2), and Ian Clark has explored how legitimacy's content changes over time in co-evolution with a changing international society (Clark 2005, 13). Shifting standards refers to the situation that occurs when normative change alters the criteria for institutional legitimacy and consequently erodes an institution's legitimacy. Failure to adapt to these 'shifting standards' is another source of legitimacy drift. For example, some institutions have been formerly evaluated primarily according to output-oriented, technocratic standards of legitimacy. Over time, however, new demands for transparency or representation may join this technocratic justification (Zürn and Stephen 2010, 96). Alternatively, an institution's entire mandate may change, requiring an overhaul of an institution's claims to legitimacy. This occurred, for example, to the International Monetary Fund (IMF) in the 1970s with the shift to floating exchange rates (Krueger 2006). Shifting standards may also be a question of extent, in which case a previously minor legitimacy criterion becomes more important over time. The importance of giving a voice to transnational societal actors may have increased over time, for example, gradually challenging the legitimacy of relatively 'closed' intergovernmental IOs (Tallberg and Jönsson 2010).

Legitimacy deficits of international organizations 103
Audience shift: Finally, an institution can be exposed to new legitimacy demands through a change in its relevant public (or legitimacy constituency). The link between power and legitimacy is particularly obvious when it comes to defining an institution's relevant public, as some actors' legitimacy perceptions clearly matter more than others' (Symons 2011(Symons , 2558. Most IOs' primary audiences will be their member states, yet an institution's relevant public can potentially widen to include secondary audiences, such as non-member states, domestic publics, NGOs, or other transnational actors. Moreover, IOs may become reliant on these secondary audiences to achieve their goals. This growing reliance constitutes audience shift through an expansion of an IO's relevant public. In these cases, an institution may remain legitimate in the eyes of its founding members, but still lose legitimacy (Symons 2011(Symons , 2562. This can occur, for example, through the accession of states as new members of an IO. New members of an IO's relevant public may in addition hold different ideas about what makes an institution legitimate, which may contribute to shifting standards. Yet audience shift can also have an immediate effect in cases where an institution is thought to require participation or representative legitimacy-in such cases, the addition of new members without changes in the procedures by which interests are aggregated and represented may ipso facto erode its representative legitimacy. Among IOs' member states, however, there is unlikely to be a consensus on the sources of legitimacy germane to the institution; moreover, the Janus-faced nature of IOs as both corporate agents and platforms for their memberships complicates the directionality of legitimation (Zaum 2013a, 13-22). As Christian Reus-Smit (2007, 164) has argued, 'The question of which constituency an actor must establish legitimacy in can be answered only with reference to the political realm in which he or she seeks to act'. Since this political realm is unlikely to remain static over time, institutions will need to recalibrate their legitimacy as their relevant publics evolve. Such cases of audience shift are likely to give rise to calls for institutional reform to keep an IO in line with external developments.
These three sources of legitimacy drift-broken promises, shifting standards, and audience shift-all generate the potential for eroded institutional legitimacy. While these processes often interact, they can also vary independently of one another. An institution's legitimacy can be undermined through audience shift, even if it continues to perform well according to previously established criteria. Alternatively, shifting standards of legitimacy can lead to new challenges for institutional legitimacy, even if the relevant public remains unchanged. Or an IO's legitimacy may erode simply because it becomes less able to fulfil the legitimacy demands originally made of it, and so on.

Organizational Responses
International organizations can respond in a number of ways to legitimacy deficits. A response exists when an organization changes its structure or behaviour in response to legitimacy demands. Each of the responses can be applied to ameliorate any element of an IO's eroded legitimacy.
The most straight-forward forms of organizational responses to legitimacy drift are re-legitimation through institutional reform and operational adaptation. These responses entail changing formal structures or actual practices to adapt in line with 'collectively valued purposes in a proper and adequate manner' (Meyer and Rowan 1977, 349). In response to audience shift, for example, institutional reforms that integrate new members into an IO's activities could recalibrate an institution's design with its new public. Legitimacy erosion could be addressed through compensatory reforms that bolster an IO's performance in weak areas, or through meeting increased expectations in a particular area of performance (output legitimacy). Shifting standards may require more fundamental reforms, such as changing an IO's mandate to cover new areas or changing its decision-making procedures in accordance with new ideas of legitimate procedures.
Re-legitimation through reform and operational adaptation are, however, not always viable responses to legitimacy drift (Bromley and Powell 2012;Meyer and Rowan 1977, 356-59;Oliver 1991). Unlike founding legitimacy, legitimacy repair is a reactive process, often coming late, and encountering encrusted organizational obstacles (Suchman 1995, 597). Reform of IOs also involves navigating powerful vested interests and built-in constraints (Hanrieder 2015). To address this, some institutions have been designed with provisions for future 'updates' already in mind. But these provisions are often stymied in practice. For example, the IMF provides for quinquennial revisions of its weighted voting structure. But obstruction tactics enabled by qualified majority voting rules tend to hinder this updating in practice (Vestergaard and Wade 2015). The International Labour Organization has been relatively more successful in expanding the size of its Governing Body, but has still fallen short of implementing constitutional reform. 10 Reform can also be prevented due to problematic features of legitimacy concerns themselves. Where legitimacy demands are conflicting or even contradictory, re-legitimation finds itself on the horns of a dilemma. In such cases, new legitimacy demands may be internally inconsistent, rendering reform strategies indeterminate (Oliver 1991, 162;Rasche and Gilbert 2015, 246;Suchman 1995, 594).
Persistent legitimacy deficits combined with institutional inertia are more likely to call forth 'coping mechanisms' such as public relations initiatives, promises of reform, the logic of confidence and good faith, and decoupling (Meyer 2009, 51;Meyer and Rowan 1977, 356). 11 Public relations initiatives involve legitimacy-enhancing communications or expressing commitment to legitimate norms ('impression management') (Allen and Caillouet 1994;Bansal and Clelland 2004;Ecker-Ehrhardt 2017). While promises of reform in the future may also mollify critics, they come at the price of conceding illegitimacy today (Meyer and Rowan 1977, 356). But in the case of IOs whose primary legitimacy constituency remains their state members, experienced diplomats are likely to see through simple window-dressing efforts.
Persistent legitimacy deficits of IOs may be alleviated more substantially by what sociological institutionalists know as decoupling and the related logic of confidence and good faith (Meyer and Rowan 1977, 356-59). In contrast to institutional reform or meaningful operational adaptation, these responses to legitimacy 10 An Instrument of Amendment to the ILO Constitution was adopted in 1986 which would greatly expand membership in the Governing Body and abolish permanent seats. However, it has failed to gather the ratifications necessary to come into effect (two-thirds of the ILO member states and at least five of the ten states with permanent seats) (International Labour Organization 2017). 11 This list of alternative organizational responses to legitimacy deficits is not designed to be exhaustive but to include the most salient categories as represented in sociological institutionalism. For an elaboration see Pache and Santos (2010). deficits are ceremonial rather than substantive (Kentikelenis, Stubbs, and King 2016). Both options appear more likely when reform paths to re-legitimation are blocked.
Decoupling has been observed in many organizations (Bromley and Powell 2012;Dick 2015;Maclean and Behnam 2010;Oliver 1991;Rasche and Gilbert 2015), but is still curiously absent from literature on IOs (but see importantly Kentikelenis et al. 2016). It refers to the gaps that tend to emerge between institutions' structures, policies, and practices, and is seen most vividly in the 'famous gaps between norms and behavior' (Meyer 2009, 50). Since satisfying all members of a legitimacy audience is often difficult, institutions often practice a certain level of hypocrisy (Brunsson 1989). Decoupling enables organizations to seek the legitimacy from adaptation to normative criteria while failing to meet them in practice (Boxenbaum and Jonsson 2008, 79;Meyer and Rowan 1977, 357). Decoupling is traditionally seen to proceed by rendering formal structures in line with legitimacy requirements but continuing organizational practices unchanged. While decoupling may be less preferable than institutional reform for relegitimating an IO, in an imperfect world it may play an important role in reconciling legitimacy drift with institutional stability.
Relatedly, the logic of confidence suggests how internal participants and external constituents can collude in avoiding, being discreet about, or overlooking glaring departures from normatively mandated structures or practices (Meyer and Rowan 1977, 358). Thus, rhetorical practices of hypocrisy, euphemizing, and indirect speech can be deployed by those seeking to preserve an institution's legitimacy (Bourdieu 1977;Brunsson 1989;Finnemore 2009;Stephen 2015). These forms of rhetorical decoupling enable speakers to bridge the gap between normative rhetoric and political reality. These discursive practices may not be sustainable in the longer-term, either because actors' preferences converge with their espoused beliefs in order to reduce cognitive dissonance (Dutton and Dukerich 1991), or because outside critics expose decoupling and denounce it as hypocrisy (Finnemore 2009). Alternatively, it may persist as a more or less regularized way of mediating normative and functional or interest-based demands while saving face (Meyer and Rowan 1977, 358;Stephen 2015, 773). Table 1 summarizes these categories of legitimacy drift and institutional responses.
The next section applies this model of legitimacy drift and its organizational responses to the case of the Security Council. The section begins by examining the standards of legitimacy that underpinned the Security Council in the eyes of UN Members at its founding moment and during its earliest months of operation. It Promises of reform then examines subsequent changes in the political and normative environment that combined with institutional stasis to create legitimacy drift. Decolonization and its resulting additional UN membership resulted in audience shift, the Cold War and later power shifts led to legitimacy erosion, and the delegitimation of sovereign inequalities generated shifting standards. This section then shows how the resulting legitimacy deficit prompted (until now unsuccessful) attempts at institutional reform, leading to promises of reform, decoupling and the logic of confidence.

The Legitimacy of the Security Council in Time
The institutional design of the Council began as the Second World War drew to a close-a critical juncture in which powerful states could redraw the institutions of world politics with few constraints. As in prior great conflicts, the victors of war were able to shape the parameters of international legitimacy, but they did so in a way that drew to an unprecedented degree on international institutions (Clark 2011, 123-46;Cox 1987, 211-67;Ikenberry 2001, 146-214). Initially sketched out by American and British policy planners, the basic design for a new IO to replace the League of Nations solidified during the 'Dumbarton Oaks Conversations' from August to October 1944. The war context and the limited number of participants gave the negotiators unparalleled flexibility to set the terms of the new organization (Bosco 2009;Hildebrand 1990;Hurd 2007). While most UN organs were based on one-state, one-vote voting procedures, the Security Council granted special rights and responsibilities to its five permanent veto-wielding members (the P5). 12 The Council would embody Roosevelt's idea for a 'trusteeship of the powerful', whose institutional design would be the result of hard bargaining among the 'Big Three' wartime allies of the United States, United Kingdom, and the Soviet Union. The results of these negotiationsincorporating the Republic of China in the second part of the Dumbarton Oaks conference, and resulting in the addition of France as a final permanent memberwould then be presented for approval to the other wartime allies at the United Nations Conference on International Organization in San Francisco in 1945. Hierarchy was not new to international society at this time (Keene 2007;Lake 2009b), but the Security Council institutionalized it to a new degree. To this day, the veto remains glaringly inconsistent with the norm of sovereign equality in the UN system. Moreover, this departure from the norm and legal principle of sovereign equality appears to conflict with the UN's own commitment to the principle of the sovereign equality of all its members. How was this institution legitimated at this founding moment? 13 12 The question of voting procedures was deferred until agreement was announced at Yalta in February 1945. See United Nations (1947. 13 The legitimacy of the Security Council is made more complicated in that the Council is both an autonomous body and an intergovernmental forum (Welsh and Zaum 2013:67). To some extent, Council legitimacy is a reflection of the members that make it up, yet there are good reasons for treating the Council as an autonomous organization for legitimation purposes. The most compelling is that states act 'as if' the Council were an independent body, and in so doing, constitute it as one (Searle 1995). While its decision-making procedures are power-based and institutionalize great power hierarchy, its resolutions are made in the name of the Council. From a legal perspective, as well, IOs carry legal rights and obligations separate from their memberships (d 'Aspremont 2007), indicating that their legitimacy is independent as well.

Institutional Design and Founding Legitimacy
The Security Council was designed with three primary goals in mind: to ensure control for the victorious 'great Power' permanent members, to prevent a rerun of the failure of the League of Nations, and to secure some legitimacy in the eyes of other UN members. The veto ensured that the organization never took decisions that went against the core interests of the great powers (Bosco 2009, 5;Clark 2011, 169;Hilderbrand 1990, 3;Hurd 2014, 366-67). In this way, the veto system would solve the first two problems, but raise complications for the third.
Existing accounts of the Security Council's founding legitimacy centre on three components. First, pivotal to the legitimation of the Council in the eyes of small and medium states was its promise, despite its hierarchical structure, to maintain international peace and security (Bosco 2009, 35-37;Bourantonis 2005, 6;Hurd 2007, 89-95;Schlesinger 2003;Simpson 2004, 165-93). The privileges of the P5 were to be accepted as a sine qua non of an effective organization (Simpson 2004, 192). This 'output' mode of legitimacy is indispensable in that any institution's legitimacy must depend, to some extent, on its capacity to fulfil its goals.
Second, while the presence of non-P5 states on the Council was always a core element of its representative legitimacy (Bourantonis 2005, 6), this representation was also heavily weighted in favour of states of principal material capabilities (Luck 2006, 11). The representation of lesser powers on the Council as non-permanent members was designed primarily to let them 'blow off steam', in Anthony Eden's words (Bosco 2009, 15). At the San Francisco conference in 1945, the only substantive amendment to the proposed Council was the addition of two criteria for election as rotating members: 'in the first instance' contributions towards maintaining international peace and security, and, second, equitable geographical distribution. 14 The outcome of the conference gave little indication that the Security Council should in some way 'represent' the UN membership generally. Representation in this sense played a relatively minor role in the Council's founding legitimacy.
Third, the great powers chose to open their proposed Charter to a public process of deliberation with lesser powers at the San Francisco Conference in 1945, on the basis of the Dumbarton Oaks 'Proposals' from the previous year. This undoubtedly made it appear more legitimate than if it had been simply imposed after the Dumbarton Oaks negotiations amongst the four major allies (Hurd 2007, 107-8). This deliberation lent the Council at least some semblance of democratic approval, despite the Charter adopted being almost identical to what was originally proposed by the Big Three.
But an account of the Council's founding legitimacy would remain incomplete if it did not take into account two further factors which were practically taken for granted at the critical juncture of its founding, but which have subsequently become increasingly contested. These are (1) the legitimacy accorded to the allied great powers by virtue of their roles in fighting the Second World War, and (2) the widespread assumption that an institutional hierarchy favouring the great powers was a legitimate feature of international society.
A large part of the Council's founding legitimacy derived from the war. All along, the Council had its roots in a wartime alliance, in which the 'big four' powers (the United States, United Kingdom, Soviet Union, and China) were accorded special status (Plesch 2008, 144-48). The UN Charter contained references to the 'enemy states' of the UN, and still does. Indeed, the five permanent members were not selected based on their perceived capabilities alone. Both France and China were added in large part because of their status as major allies in the war, not their power capabilities (Bosco 2009, 24-26;Heimann 2015). 15 When the General Assembly met for the first time in London in January 1946, the legacy of the war still dominated discussions, and the Council's founders repeatedly associated it with the defeat of the Axis powers. Prime Minister Clement Attlee of the United Kingdom reminded delegates that the UN had its origins in plans made 'while our enemies were still in the field against us' (A/PV.1, 39), 16 while the Soviet delegate reflected: The United Nations … was created by the same anti-Hitler coalition that was headed by the United States of America, Great Britain and the Soviet Union, which bore on their shoulders the brunt of the struggle against our common enemy and which were anxious to create an effective international organization for the defense of post-war peace and security (A.PV/42, 836; also A/PV.84, 92).
The USSR also justified its support of the Norwegian Foreign Minister for the office of President of the General Assembly on the basis of Norway's 'resistance to the German invaders' (A/PV.1, 43). Small states also linked the Council and its P5 members to their roles in winning the war. New Zealand's representative stated that '…we, and those who think as we do, recognize and acknowledge at once that the great Powers who played the predominant part in winning the war must similarly play the predominant part in winning the peace' (A.PV/39, 785). Uruguay had also earlier circulated a position document conceiving the great powers' roles in the Security Council as a 'reward' for their role in the war (Simpson 2004, 171). In general, the speeches delivered at the Security Council's first meeting contained numerous references to its members' roles in defeating the Axis powers (Bosco 2009, 41). To a significant extent, the legitimacy of the Security Council had already been prepared at Stalingrad and Normandy.
The legitimacy derived from the war effort was accompanied by a widespread belief among states that the great powers-those states with the greatest material capabilities-had special rights and responsibilities in international politics. Recognition was explicitly tied to (material) power (Brown 2004;Bukovansky et al. 2012;Donnelly 2006). Hierarchy was not foreign to the international system during the Council's founding moment. Sovereign states were accorded superior political rights to dominions and colonies; great powers were accorded greater rights and responsibilities than lesser sovereign states. International organizations like the Council could therefore draw on a degree of stratificatory legitimacy: a conviction 15  that 'institutionalized inequality in the distribution of primary resources-such as power, wealth, and prestige-is essentially right and reasonable' (Fave 1980, 955).
At the San Francisco conference, the great powers repeatedly 'emphasized the naturalness and inevitability of an unequal international system' (Hurd 2007, 101). Of course, the specific extent of great power privilege was hotly contested. Many small and medium states were dissatisfied with the broad sweep of the P5's veto power and saw the San Francisco conference as an opportunity to pressure the great powers into limiting the veto's place in the functioning of the Council (Hurd 2007, 89-95;Simpson 2004, 165-93). But this dispute 'mostly concerned the detailed implementation of special responsibilities, rather than any object to the basic principle itself' (Bukovansky et al. 2012, 30). According to Ian Hurd, proposals by Cuba, Ecuador, and Iran that would have limited the veto even in enforcement actions 'were as unpopular with the rank and file as they were with the Great Powers' (2007, 94 note 37). 17 The fundamental concept of great power hierarchy, as well as the basic concept of the veto, were largely taken for granted (Hurd 2007, 93;Simpson 2004, 170). According to the Chairman of the General Assembly, the great powers, 'by virtue of Articles 24 and 27 of the Charter, and by the very nature of things, will shoulder the chief responsibility for the maintenance of peace and security' (A/PV.1, 38).
To summarize, the Council's founding legitimacy was based not only on its promise to secure international peace and security, the presence of rotating members, and deliberative legitimation at the San Francisco conference. It was also intimately connected to the legitimating perceptions that the permanent Council members had earned their places by fighting the war, and that its hierarchical structure simply reflected the inherent rights and privileges of the great powers in organizing international affairs. As the following section will show, this founding legitimacy may have been effective at this critical juncture, but it would prove increasingly problematic in the long run.

Legitimacy Drift
The Security Council is a paradigmatic case of path-dependence in the institutional design of an IO. Entrenched veto players have strongly limited institutional adaptation. Initial proposals to 'index' the distribution of privilege in the Council were rejected outright by the wartime great powers. 18 While the Security Council remained remarkably static, the world changed rapidly around it. This situation had important implications for the Council's legitimacy.
First, the Council very quickly began to break its promises, eroding the Council's capacity to live up to its initial criteria for legitimacy. The loss of empire rendered Britain and France's great power statuses increasingly honorific, calling into question the legitimacy of the Council's composition as a great power directorate. At the same time, the Council's inability to act, arising from the paralysis of the Cold War, undermined its claim to be able to uphold international peace and security through great power cooperation. This weakened the crucial output dimen-110 Matthew D. Stephen sion of its institutional legitimacy (Wallensteen and Sollenberg 2004). As Thomas Franck (1970, 810) lamented, 'Almost from the moment the San Francisco Charter was signed, this essential prerequisite for U.N. collective enforcement action-the unanimity of the great Powers-was seen to be an illusion.' Many countries began to question why they had agreed to the creation of a Council that appeared racked by Cold War rivalries.
Second, the Council's audience shifted. Soon after the Council's first meeting in 1946, the number of sovereign states quickly began to increase. The anti-colonial struggles led to the gradual universalization of the principles of self-determination and sovereign equality, expanding the number of UN members and leading to new calls from Asian and African states to expand the membership of the Council. From the original 51 founding members in 1945, the membership of the UN increased to 76 by 1955, and to 117 by 1965, mostly coming from newly independent African and Asian states (Bosco 2009, 98). This also upset the Council's geographic apportioning of rotating seats, as new UN members came overwhelmingly from Africa and Asia (Bourantonis 2005, 7-8;Morris 2000, 266). Additionally, many newly independent states had very different legitimacy beliefs, which congealed into a North-South conflict at the UN (Malone and Hagman 2002). The Security Council had difficulties adapting to these changes in its relevant public.
Third, these changes were also accompanied by shifting standards. The Non-Aligned Movement became a significant force challenging inequalities and stratification amongst the society of states by championing norms of national self-determination, racial equality, sovereignty, and democracy. This was accompanied by the gradual liquidation of European empires as part of a broader normative shift that discredited hierarchy and venerated equality as a Grundprinzip of nearly all political institutions. This normative shift manifested itself in various forms-in the rules that determined who could claim self-determination and statehood (liberating colonial peoples), in how domestic political orders were justified (democracy in liberal or 'socialist' variants), and in prevailing cultural values (reason and equality) (Donnelly 2013;Meyer et al. 1997). Landmarks at the UN included the 1960 Declaration on the Granting of Independence to Colonial Countries and Peoples (General Assembly Resolution 1514) and the 1970 Declaration on Principles of International Law Concerning Friendly Relations and Co-operation Among States (Resolution 2625) (Donnelly 2013, 105). This contrasts strongly with the situation that prevailed in 1945, when anti-colonial campaigners failed to have an anti-colonial clause inserted into the UN Charter (Bhagavan 2010). 19 Just as democracy became a way to realize the political equality of individuals, sovereign equality of states came to be associated more strongly with equal representation in IOs. Democratic pressures also made themselves felt in relation to IOs (Grigorescu 2015).
The debates about Council reform in the 1960s show evidence of the early stages of this normative shift. One indicator is a shift in the diplomatic vocabulary. While, in 1945, UN diplomats debated the privileges and responsibilities of the 'great Powers' explicitly, by 1963, they had begun to shift towards alternative terminologies with different connotations. While the term 'great Power' was still widely used in General Assembly debates, 20 many speakers chose to refer instead to 'big Powers', 21 'major Powers',22 'industrial powers',23 and 'nuclear Powers'. 24 Moreover, speakers associated with the Non-Aligned Movement spoke not of great powers but of 'colonizing Powers ',25 'colonial Powers',26 or even 'the so-called great Powers'. 27 In general, international society was becoming less tolerant of sovereign inequalities and changing discursive practices reflected this. By 1965, the permanent members could not avoid Council reform entirely, and four more rotating seats were added.
While often taken for granted today, these shifting standards were a debated matter of public record at the time. In 1972, the UN Secretary General, Kurt Waldheim, reflected upon the meaning of this normative transformation for the Council, Even if the security council were to acquire a new effectiveness through Great Power détente, the idea of maintaining peace and security in the world through a concert of great powers, although these powers obviously have great responsibilities in matters of peace and security, would seem to belong to the nineteenth rather than to the twentieth century, where the process of technological advance and democratization is producing a new form of world society (Waldheim 1972, 19).
By the mid-2000s, when Security Council reform had once again resurfaced in public debates, the term 'great Power' had disappeared completely. 28 While explicit appeals to great power entitlements were commonplace in the immediate post-war period, states today appear reticent to invoke explicit inequalities as a legitimate procedural norm. Likewise, the role of member states in the Second World War is no longer invoked as a legitimacy claim of the Security Council. 29 At the same time, contemporary debates about Council reform indicate that the legitimacy of the Security Council has become increasingly publicly evaluated according to democratic principles such as accountability, transparency, and representativeness (Binder and Heupel 2015). As organizational sociologists have noted, recent decades have witnessed a universal intensification of rationalizing pressures associated with concepts of accountability, assessment and transparency (Bromley and Powell 2012, 484). Representation, in particular, has become so central to the legitimation of the contemporary Security Council that it can sometimes function as 'a proxy for legitimacy' generally (Lowe et al. 2008, 33).

Responses to Legitimacy Drift
Since the Council has not been comprehensively updated in its seventy-year history, and its hierarchical privileges remain in place, how has it maintained legitimacy? The Security Council's internal constituents and external audiences have responded to its drifting legitimacy in a variety of ways. Formal institutional reform occurred for the first and only time in 1965, increasing the number of rotating seats from six to ten. The effect of this expansion on the Council's activities was unremarkable. Due to a combination of vested interests (the P5), veto players (the P5) and inconsistencies in legitimacy demands (contradictory demands from the UN membership), reform of its formal structures has not occurred since then. For over 50 years, all attempts at formal institutional reform of the Council have been famously unsuccessful (for overviews, see Blum 2005;Cox 2009;von Freiesleben 2013).
The Council has been more successful at operational adaptation. It has addressed new demands for Council performance regarding new threats, although in the absence of formal institutional reform, this has largely been dependent on creative reinterpretation of its legal basis (see below). While the formal structure of the Council remains largely unchanged, and its original envisaged purpose of confronting acts of international aggression has been only rarely implemented, it has embarked on new ventures such as numerous peacekeeping operations (Lipson 2007), convening international criminal tribunals (Sandholtz 2008), imposing new 'smart' sanctions (Drezner 2011), and using other diplomatic implements (Luck 2006, 60-61) to fulfil its mandate of ensuring international peace and security. But operational adaptation has not been a silver bullet either. Member states have not universally welcomed this expanding Security Council agenda and the raised the normative ambitions to which it is assessed (Andreopoulos 2008).
Discretionary prerogatives of the Council have also been exploited to informally and provisionally amend its practices, in the absence of institutionalized reform. In the case of the Council's so-called 'Provisional Rules of Procedure', these informal rules have maintained their 'provisional' character since 1946 and have never been substantively altered (Security Council Report 2007, 2). Rather, accommodation to new legitimacy demands has taken place through the evolution of Security Council working methods, which the Charter left up to Council discretion (Hulton 2004;Wood 1996). Many of these ad hoc changes have been made in response to new legitimacy demands for more 'transparency' and 'accountability' in the Council's activities, even though normative demands for greater transparency are often regarded as conflicting with functional effectiveness (Wood 1996, 154-56). The working methods of the Council have been adapted pragmatically, centring on dialogue with non-members, consultations and 'voice' with troop-and police-contributing countries as well as the Secretariat, and informal interactive dialogues with NGOs and so-called 'Arria-formula' meetings (United Nations Secretariat 2002). Since the end of the Cold War, most of the real work began to take place in 'informal consultations of the whole', and the trend of less frequent Council voting emerged, with those votes that do occur more likely to succeed, as they are often 'pre-cooked in informal consultations' (Wood 1996, 155-56). These changes have not been wholly symbolic. This has allowed the Council to express conformity to new legitimacy demands while remaining largely-if not entirely-unchanged.
As re-legitimation through institutional reform has largely been blocked, and operational adaptation can only partly alleviate its legitimacy deficit, the Council Legitimacy deficits of international organizations 113 has witnessed the emergence of other coping mechanisms. Some emerged soon after the critical juncture of Council creation.
In reacting to pressures of legitimation while remaining structurally unchanged, the Security Council has experienced decoupling from its written rules and its actual practices. First, many aspects of the UN Charter have been ignored or creatively reinterpreted to facilitate unforeseen forms of institutional adaptation. Already many years ago the UN Charter was described as bearing as little resemblance to the modern world as a Magellan map (Franck 1970, 810). Examples are numerous. On the procedural side, a standard practice quickly emerged of not regarding abstentions or absence by the permanent members as obstacles to the adoption of Council resolutions (in contrast to the wording of Article 27 of the UN Charter). Departing from the formal rules was found to be functional for organizational practices (Hurd 2014, 371). The Council ignores other anachronisms of the UN Charter in practice, including references to 'enemy states' (Article 53), the (non-existent) armed forces on standby at the Security Council's disposal (Article 43), and the inclusion of states that no longer exist (such as the USSR) (Article 23). 30 Additionally, while the Charter provides that 'a party to a dispute shall abstain from voting' (Article 27), this provision has never been used against a permanent member. On the performance legitimacy side of the ledger, the Council's operational activities have also been adapted to confront new perceived international threats, in a process of increasingly creative interpretation of the Charter and the concept of international threats. Council practice, together with advisory opinions of the International Court of Justice, have developed meanings left 'implicit' in the Charter text, such that human rights violations, civil wars, non-traditional armed conflicts, weapons of mass destruction, and the crime of apartheid have been identified as threats to international peace and security, and therefore come under Council authority (Franck 1970(Franck , 810-12, 2006Sato 2001;Welsh and Zaum 2013, 80-86). 'When there is a willingness to make the Charter work in new circumstances, the dead hand of text has not, in the past, always barred the way to transformative change' (Franck 2006, 603).
Second, the logic of confidence and good faith has helped repackage 'legitimated vocabularies' of Security Council reform talk (Meyer and Rowan 1977, 349;Rasche and Gilbert 2015, 242). While explicit appeals to great power entitlements were seen as more or less appropriate at the Council's founding, states today are reluctant to invoke explicit inequalities as a legitimate procedural norm, even if this is widely understood as a necessary concession to Realpolitik. Thus while 'great Power' talk has disappeared, 'global realities' have become an indirect way of discussing the correspondence of the Council's composition with new members of the great power club (Clark 2011, 153;Stephen 2015, 780). While debates about the Council's design were framed in terms of the rightful privileges and responsibilities of the 'great Powers', today this discussion uses a less direct vocabulary and coexists with stronger norms of sovereign equality and democratic decision-making. Great power principles that legitimize the Security Council as a 'directorate of the most powerful states' have to be softened (Müller 2013). Likewise, aspiring permanent members carefully frame their arguments in terms that affirm state equality; adding new members to the great power club will make the Security Council more representative, transparent, effective, participatory, and inclusive (Stephen 2015, 780). In this way, international hierarchy can be discussed while not being referenced explicitly (Bull 1977, 228-29).
Finally, the Council has also coped with legitimacy drift through promises of reform. The prosecution of a protracted and open-ended Security Council 'reform process' signals acceptance of conformity to new legitimacy demands via the promise of a (permanently deferred) institutionalized re-legitimation in the future. Beginning in 1979 with the addition of the 'question of equitable representation on and increase in the membership of the Security Council' to the General Assembly agenda, the modern reform process has effectively allowed the Council to signal formal commitment to newly strengthened principles of representation, accountability and transparency, while engaging in business as usual. Reform processes in themselves can operate as 'a main currency of organizational legitimacy' (Hanrieder 2015, 2). Promises of reform hold out the prospect of re-legitimation sometime in the future. The cost, however, is that the organization's current structure is tacitly acknowledged to be illegitimate (Meyer and Rowan 1977, 356). Table  2 summarizes the argument.

Conclusion
This article has argued for a historical and theoretically eclectic approach to institutional legitimacy. While institutions are typically designed with their own legitimation in mind, failure to subsequently 'update' an institution to changed circumstances can engender legitimacy drift. Responses to legitimacy drift include re-legitimation through institutional reform and operational adaptation, as well as through other coping mechanisms emphasized by sociological institutionalism. The Security Council exemplifies these processes, while providing a vivid illustration of the changing normative structure of international society over time. The initial legitimacy of the Security Council was not only intimately tied up to perceptions of the great powers' inherent special rights and responsibilities, but also the legitimacy that the allied great powers accrued in fighting the Second World War. While this mixture of legitimation fit the critical juncture of the immediate post-war setting, subsequent political developments and normative shifts subjected this legitimacy to a process of drift. The audience to which the Council must appeal for legitimacy became far larger and more diverse, while the Council's permanent members accounted for a much-diminished share of global power capabilities. The elevation of sovereign equality and norms of transparency, accountability and representation have shifted standards. Since attempts at re-legitimation through institutional reform have been blocked, coping mechanisms such as decoupling, the logic of confidence, and promises of reform in the future have emerged. While the Security Council may be a particularly 'sticky' institution, and thus particularly prone to legitimacy drift, the model developed in this article emphasizes the diversity of processes contributing to legitimacy drift. Analogous processes may be observed in other IOs. The Bretton Woods institutions are prominent examples whose nature as shareholding institutions with weighted voting has also conflicted with the shift towards principles of democracy and sovereign equality between states. The Group of Seven industrialized countries (G7) provides another example, in which an informal club of like-minded capitalist democracies began, over time, to assume competencies that affected all states in the global economy, exposing the group to audience shift and new legitimacy demands from excluded states to broad social movements and civil society groups. Historicizing institutional legitimacy may help explain the legitimation dynamics that other IOs face today.
Finally, this article also suggests that developing persuasive and historically sensitive accounts of IO legitimacy stands to gain from drawing on insights from multiple institutionalist theories. First, scholarship gains a more explicit understanding of the origins and causal processes of institutional legitimacy deficits. Second, scholars can assemble different theoretical insights into establishing persuasive descriptions and explanations of concrete cases. As IOs grow older and the global population of IOs increases, the clean assumptions of functional design and choice-based regime theory may become less rewarding relative to the insights provided by historical and sociological institutionalisms.

Disclosure statement
No potential conflict of interest was reported by the author.

Funding
The author benefited from a Research Fellowship at the Berlin-Potsdam Research Group (KFG) 'International Rule of Law-Rise or Decline?' at the Humboldt University of Berlin in 2016.

Notes on contributor
Matthew D. Stephen (Dr. phil. Freie Universität Berlin) is a Research Fellow in the Global Governance unit at the WZB Berlin Social Science Center. He has recently taught at the Freie Universität Berlin, the University of Potsdam, and the Stanford University Berlin Program. His research interests include the rise of new powers, legitimacy, international institutions, linguistic approaches to IR, and materialist approaches to IR. He recently authored articles in the European Journal of International Relations and Global Governance.