The latest episode in the selection of the United Nations (UN) Secretary-General (SG), which led to Antonio Guterres's appointment in October 2016, presented several unprecedented features. The UN organized public hearings and town hall meetings with the dozen candidates and launched a website to allow private individuals from anywhere in the world to ask questions directly to the candidates. The presidents of the Security Council (SC) and the General Assembly (GA) also signed a rare joint letter to determine the selection procedure in advance. While the Permanent Five (P5) ultimately decided on the nomination behind closed doors, these innovations remain significant. A prominent UN observer, Edward Luck spoke of a “new approach to electing the secretary general that really is revolutionary … It's historic change that I would not have predicted a year ago would happen.”Footnote 1
The renewal of the SG selection process is a textbook case of what historical institutionalism (HI) calls “layering”—“the introduction of new rules on top of or alongside existing ones.”Footnote 2 As a form of institutional change, layering is particularly prevalent inside international organizations (IOs) whenever outsiders push for reforms that insiders are reluctant to launch.Footnote 3 In this case, the challengers—comprised of activist UN member states and civil society organizations—were unable to formally amend the selection procedure, which incumbents—primarily SC members—were reluctant to open. Instead, reform advocates grafted new practices onto existing procedures, including informal dialogues with candidates and the publication of selection criteria and mission statements.
The selection process of the UN SG forms a compelling case for exploring the nexus between institutions and practices in world politics because the relevant formal provisions are rather loose and incomplete. To fill these gaps, diplomats have historically experimented with various ways of doing things, such as “straw polls,” that emerged incrementally and remain largely uncodified to this day. As Wiseman observes, “If informal practices help explain the Secretary-General's growing role in world affairs, so do they at least equally help explain changes in the Secretary-General selection process.”Footnote 4 To better understand how practices and institutions interact in instances of incremental institutional change, I set up a dialogue between HI and practice theory, showing how each framework stands to benefit from such cross-fertilization.
In this article I pursue two objectives. First, at the theoretical level I develop a “pulling” theory of agency. Distinct from but complementary to the soft rational choice that informs most versions of HI, this practice-centered framework helps capture the socially emergent nature of institutional change, as well as the grounding of agency in its social environment. The theory argues that social action is not only pushed by interests, as per conventional wisdom, but also elicited by the web of practices within which it takes place. Second, at the empirical level, I shed light on a salient case of institutional change in contemporary world politics. Rivlin argues that thanks to his “status of a secular pope … the Secretary-General sits at the apex of world governance.”Footnote 5 The way in which this critical individual is selected has significant import for world politics: after all, a candidate representing the lowest common denominator among P5 is likely to pursue a different political agenda than one who triumphed via the court of worldwide public opinion. Furthermore, some of the new selection practices have already started to diffuse to other IOs, altering the political landscape of global governance.
Using this important case of institutional change, I develop, operationalize, and illustrate three causal mechanisms derived from the pulling theory of agency, concerning the timing, form, and magnitude of layering. What best explains the institutional outcome at hand, I argue, is a set of new or recently imported practices at the UN, which generated conditions conducive to the 2016 renewal of the SG selection process: (1) relational crossover between challengers and incumbents; (2) competence transfers to legitimize heterodox practices; and (3) pushback tools to mitigate crossover and transfers. I use these mechanisms to resolve three empirical puzzles. First, why did the change occur in 2016, while similar calls for reform had already been made in earlier episodes? Second, why did the change take the form it did instead of others? And third, what explains that change was circumscribed in the final stages of the selection process? With the goal of theory development, I build on a critical case of institutional change to climb the ladder of abstraction and coin portable concepts that may later be embedded in other singular causal analyses.
Making Sense of Layering: Toward a Pulling Theory of Agency
The pulling theory of agency aligns with HI's ontology of path dependence and helps resolve problems in intentionalist accounts of institutional change. While it is clear that actors pursue their interests, it is also the social environment within which they evolve, notably the web of practices that surrounds them, that accounts for the dynamics and outcomes of layering.
The pushing and pulling metaphor derives from Max Weber who famously argued that action is “pushed by the dynamic of interest.”Footnote 6 This common understanding, according to which actors’ inner motives explain their choices, dominates social sciences today, primarily in the form of rational choice theory. I argue that the notion that interests push actors is accurate but incomplete: action is also “pulled” by the social environment within which agents find themselves. In being drawn by available practices, agents do not contradict their interests but simply act on them in contextually enabled ways. The pulling theory of agency, then, helps account for the social determinants of interest-driven action. Paraphrasing Adler, we can say that actors are simultaneously pushed and pulled.Footnote 7
Before proceeding, I should distinguish between institutions and practices, which are similarly located at the “meso” level,Footnote 8 “suspended” as they are between individuals and macrostructures.Footnote 9 The semantic proximity between the two concepts shows in standard definitions. According to North, institutions are “the rules of the game in a society,”Footnote 10 whereas practices may be defined, building on Adler and Pouliot, as socially organized and meaningful patterns of action.Footnote 11 Helmke and Levitsky use the terms interchangeably: “Informal institutions do change—and often quite quickly. The centuries-old Chinese practice of foot-binding disappeared within a generation.”Footnote 12 Related as the concepts may be, I maintain a conceptual distinction for two reasons. First, practices generally form more granular units of analysis than institutions; and second, practices point to the dynamic performance of institutions—their “legwork,” so to speak. For instance, a wedding is the practice that anchors the institution of marriage, together with other practices including everyday life habits, housing arrangements, and family parties. While marriage may be described as a set of rules, then, its constitutive practices are concrete ways of doing things and meaningful patterns of action. Thus, bundles of practices compose institutions, set them in motion, and sustain their existence at the level of action.
Intentionalist Theories of Agency and Their Limits
How do institutions gradually transform from within? This is a key question for students of world politics where incremental change is particularly prevalent.Footnote 13 In recent decades, HI scholars have developed a vibrant research program with the goal of “explaining the more gradual evolution of institutions once they have been established.”Footnote 14 Major inroads have been made to theorize the structural conditions of incremental change. Layering, for example, occurs specifically in cases of (1) strong veto possibilities, that is, “access to institutional or extrainstitutional means of blocking change”Footnote 15 and (2) low level of discretion in rule interpretation and enforcement.Footnote 16 Useful as it is, however, this structural framework has less to say about the dynamics of change, especially the role of agency. As Zürn complains, “HI lacks a theory of action and can therefore not provide a full explanation of institutional change or development.”Footnote 17 Indeed, when structural conditions remain constant, as in the case study that follows, the turn to agency becomes an analytical necessity.
Recent years have seen several calls for the development of microfoundations for HI.Footnote 18 To explain agency, scholars have resorted to different versions of rational choice institutionalism. Most seminally, Ikenberry thinks of international institutions as reflections of hegemonic interests, which the dominant actor rationally devises based on the distribution of power.Footnote 19 Alternatively, other authors combine soft rationalism with insights drawn from psychology and cognitive science. Pierson relies on loss aversion theory to explain why agents avoid sunk costs in a series of “nondecisions” that render them more conservative than pure strategic choice would suggest.Footnote 20 Meanwhile, Fioretos builds on prospect theory to argue that institutional change occurs “only when the benefits of a prospective alternative outweigh the losses associated with giving up access to past designs.”Footnote 21 For their part, Jupille, Mattli, and Snidal turn to the notion of bounded rationality to suggest that actors often “stick with the institutional ‘devil they know.’”Footnote 22 These amendments to rational choice theory are important because they seek to account for the micro drivers of path dependence. Overall, they help explain why there is more stasis in institutions than what pure rationality would predict.
Critically, at the theoretical level each of these accounts equates agency with intentionality. In line with Ryle's “ghost in the machine,”Footnote 23 a theory of agency amounts to a theory of the mind. From a methodologically individualist perspective, to be an agent is to make choices. These choices stem from the human mind and its inner dynamics of preference formation and cognitive biases. The brain pushes agents to act in certain ways, based on interests, beliefs, and perceptions. While there is nothing inherently wrong with intentionalist theories of agency, they also have two major blind spots. Theoretically, equating agency with intentionality makes it difficult to capture the socially emergent nature of politics, including institutional change. Empirically, pushing theories like rational choice often ignore the social determinants that influence the specific ways in which actors pursue their interests. I discuss each of these problems in turn.
First, intentionalist theories of agency risk missing the deeply interactional dynamics of institutional change. As by-products of social struggles,Footnote 24 institutions rarely if ever mirror the interests of any single agent, however powerful. By implication, identifying individual preferences as the proximate cause of a given institutional outcome, as in the theory of rational design, seems mistaken. As Wendt puts it, “there will always be some intentionality in the process by which institutions are created. However, this does not mean we can automatically conclude that institutions are intended.”Footnote 25 Of course, players try to push institutional change along their favored course, but in doing so they enter into conflicts whose turbulent effects shape outcomes much more directly than any individual preference. As a socially emergent artifact, then, institutional change owes more to relational processes than to individual volition. Intentions eventually “dissolve” in a stream of cooperative and conflictual relations, such that it becomes impossible to draw a direct connection between the individual preferences that sustain micro struggles and macro outcomes such as institutional change.
To equate change with intentional agency is to take a rather limited view on social transformation. It requires identifying specific actors, inferring their preferences, and linking their actions to given outcomes. People's preferences explain what they want; what they want explains what they do; and what they do explains what happens collectively. Someone, somewhere, wills the change into being; others follow suit either because of power differentials or because doing so allows them to collectively pull closer to Pareto optimality. The analyst presumes a link between intentionality and social outcome, often inferring the former from the latter through backwards inference. Such an approach hardly accounts for the fact—often mentioned in HI—that institutions generally are the result of messy and disorderly generative processes in which a variety of forces (including individual intentional states) collide to produce patchwork outcomes. For instance, incumbents often push back and blunt transformative attempts away from their original design.Footnote 26
The second limitation of pushing accounts of agency is their difficulty in specifying the form of institutional change. In a nutshell, intentionalist theories have an easier time explaining why agents want change than how they want to pursue and implement it. Interests are not fully determinate: there often exist various ways of fulfilling the same preference. For instance, in the case that follows, challengers clearly share an interest in fostering transparency: but how exactly will they go about it? What specific changes do they want, and by what means will they seek to obtain them? Reciprocally, incumbents prefer to maintain the status quo but in practice, there are myriad ways to resist the winds of change. Institutions provide rules, but these are full of interstices and ambiguities that form the basis of agency. Pushing theories of agency need to be complemented with a framework explaining why agents make specific circumstantial moves instead of others that would equally foster their interests.
Empirically, then, intentional theories of action have little to say about how institutional choices derive from the existing social environment—and not just cost-benefit calculations. Social action never takes place on a tabula rasa, but rather within a preexisting social environment made of a variety of institutions and other collective realities.Footnote 27 Cognitive variables such as bounded rationality, for instance, cannot account for the menu of options that frames a particular strategic choice. As Keohane puts it, “the institutional processes that generate endogenous change are not theorized. Bounded rationality creates space for institutional change, but it does not tell us how it happens.”Footnote 28 Likewise, prospect theory explains how cognitive constraints inform choices, but it does not integrate the workings of the social environment itself. Yet it is critical to endogenize social forces as part of agency itself if HI is to honor its commitment to path dependence. For instance, when cognitively limited individuals “consider alternatives”Footnote 29 to the status quo, just where do these alternatives come from? What exactly defines the limited range of options that agents face at a given point?
The twin limitations of intentionalist theories of agency lead to three generic puzzles in the study of institutional change. First is the issue of timing: why does change happen at a particular moment? Without looking at interactional effects and the social environment, rationality and cognition remain indeterminate, especially under stable structural conditions. Second, concerning form, why does change take its particular shape? Here as well, broader social constraints and resources play a major role in defining the options pursued. Finally, with regard to the magnitude of change, why does incremental change unfold up to a point and not further? The struggles between incumbents and challengers, which play out in dense social relations, is central for the expression of interests. I apply these insights after elaborating a practice-based theory of agency.
Accounting for the Pull: The Web of Practices and Its Affordances
A pulling theory of agency is complementary to the pushing approach in that it explains social action without resorting to intentionality. To repeat, I do not conceive of pushing and pulling theories of agency as rival hypotheses: both dynamics are plausibly at work in any given action. Yet because of rampant methodological individualism in social science, we know much more about the former than the latter. The pulling theory of agency is meant to complement, not replace, existing accounts centered on beliefs and interests by endogenizing the social environment within which actors find themselves as part of agency itself.
My account deemphasizes what is going on inside people's minds to instead focus on the social environment within which they find themselves.Footnote 30 To facilitate this move, I categorize agents in two broad categories, challengers and incumbents, and assume attendant preferences for change and status quo.Footnote 31 In the case at hand, P5 countries are incumbents, hoping to preserve their hold on the nomination process, while activist member states and NGOs are the main challengers, striving to introduce more transparency. Important as these clashing preferences may be, they cannot specify, in and of themselves, the exact changes that agents seek (e.g., there exist many different ways of fostering transparency), and how they plan to go about obtaining or containing them (the formal rules being rather loose and underspecified). In sum, interests do some of the pushing, but action still needs to be pulled in certain directions over others.
Paraphrasing Geertz, let us imagine that human beings are suspended in webs of practices that they themselves have spun. To better integrate the social forces that drive agency, I argue that agents are drawn to certain forms of action by the social environment within which they find themselves. People do make choices but in so doing, they are suspended in a web of practices that render certain options more likely than others. As such, agents are pulled toward available ways of doing things: their agency is elicited by the context of action. Recall that practices are collective patterns of action over which no single actor has complete control. Because it is so tightly spun, the web of practices usually transforms slowly and locally, as actors maneuver within it. Thus the web of practices is both enabling and constraining. Revisited like this, action originates not only within agents but also in their social environment. Rationality and practicality become two faces of the same coin.Footnote 32
Perhaps the easiest way to understand the difference that a pulling theory of agency makes is to refer to Gibson's notion of “affordance.” For the American psychologist, there exists a “complementarity”Footnote 33 between agents and their environment. The agent's mind is only one half of the story of agency; the environment is also an active and directing force: “Within limits, the human animal can alter the affordances of the environment but is still the creature of his or her situation.”Footnote 34 In world politics, for instance, the “action potentials” of new media technologies help explain emerging diplomatic practices in the twenty-first century.Footnote 35 Most of the time, affordances appear self-evident to the competent practitioner. This acquired disposition is precisely what explains the environment's power in enabling action: “If the affordances of a thing are perceived correctly, we say that it looks like what it is. But we must, of course, learn to see what things really are [and] this can be very difficult.”Footnote 36 Affordance is a form of complicity between the social environment and the well-attuned player.
At any moment, the web of practices within which actors evolve is full of affordances, making some moves more likely than others. Agents are disposed toward certain ways of doing things, and their familiarity with these practices structures social interaction. In this sense, my approach to agency is dispositional instead of volitional. Searle provides a useful example of this reasoning: “We should not say that the experienced baseball player runs to first base because he wants to follow the rules of baseball, but we should say that because the rules require that he run to first base, he acquires a set of background habits, skills, dispositions that are such that when he hits the ball, he runs to first base.”Footnote 37 Here, knowing the player's preference for scoring a run is less useful for making sense of her specific moves than the practical skills she enacts dispositionally, as part of the baseball game.
At the theoretical level, the key added value of the pulling theory of agency is that it endogenizes social influences as part of agency itself. This is critical for HI and its notion of path dependence, which needs to account for institutional constraints at the level of action.Footnote 38 In the pulling theory, institutions become part of agency via practices. This argument resonates with the concept of “bricolage,” which describes the combinatorial logic by which practitioners experiment with ways of doing things by mixing and putting them to new purposes.Footnote 39 As Campbell observes, “institutions provide a repertoire of already existing institutional principles and practices that actors can use to innovate. The key is to recognize that actors often craft new institutional solutions by recombining elements in their repertoire through an innovative process of bricolage whereby new institutions differ from but resemble old ones.”Footnote 40 In the study of institutional design, the notion helps capture how “change typically occurs through the grafting of modular components rather than the de nova invention of individual institutional features.”Footnote 41
Since the web of practices affords certain actions based on dispositions, its effects are never fully determinate. Meaningful patterns of action form a social infrastructure that defines the scope of agency without precluding deviation. This is because practices are always partly improvisatory. According to Cornut, skilled improvisation rests on two more or less embodied processes: recognizing that the current situation shares traits with past experiences and adjusting to shifting circumstances on the fly.Footnote 42 Micro-deviations from established patterns are the ordinary condition of social life because no two social situations are exactly the same. This is what practice theorists call the “principle of indexicality,” by which “any new situation requires adjusting and re-arranging the practice within it.”Footnote 43 To engage in practice is, by nature, to cope with a reactive and fast-changing world. This is precisely why practices differ from mere habits, in requiring constant adjustment to changing circumstances. Micro-adaptations are a standard part of competent performance.Footnote 44
As collective possessions, practices adjust not only to the situation at hand but also to social expectations about competent performance.Footnote 45 For example, what does it mean to be a good diplomat? What does skillful multilateral negotiation entail? Based on common understandings as they are, such collectively held standards are inherently open to competing interpretations. Here we find Wittgenstein's argument about the indeterminacy of social rules, which allows for different forms of behavior to possibly fall in line with any given practice.Footnote 46 Social practices are always subject to conflicting applications. These may derive from competing ways of doing things in the web of practices, but they may also be discursively constructed. Indeed, innovations and new combinations usually require some form of narration or justification: “story-telling … authorize[s] an unprecedented practice.”Footnote 47 The affordances of the web of practices, then, hinge not only on immediately available modes of action but also on narrative work.
By implication, a key political challenge in transforming an institution is to portray deviant practices as conforming to expectations. I call such narrative work competence transfers. In Hansen's words, “as ‘practitioners practice,’ they perform specific practices that are asserted as though they belong to a general—perhaps ‘self-evident’—practice.”Footnote 48 This kind of story telling transforms standards of competence by skillfully connecting heterodox performances with similar ways of doing things. Remember that any practice is subject to a never-ending struggle over what constitutes competent performance.Footnote 49 And since competence is not self-interpreting, the same standard potentially contains the possibility for orthodox as well as heterodox practices (latent as the latter possibility may remain). In a practice transfer, challengers perform skillfully though in unusual ways, thereby shifting the standards of competence and weaving foreign practices into the community's web.
To sum up, if we conceive of practices as affordances, we obtain a theory of agency that thoroughly endogenizes the social environment without resorting to cognitions. Strategic as they may be, practitioners find themselves suspended in webs of practices that pull them in certain directions and not others. Generally, the more readily available the practice, the stronger its dispositional effects because agents reproduce past patterns of action by default. Even when actors resort to heterodox ways of doing things, they do so on the basis of available practices in their environment. Within limits, narrative work may allow certain transfers between webs of practice, often at the cost of hard political work as the following case demonstrates.
Operationalization: Puzzles, Mechanisms, and Methodology
To operationalize the pulling theory of agency I infer three causal mechanisms that seek to resolve the puzzles of layering: timing, form, and magnitude. The explanatory logic is simple: given a relatively constant structure of power and preferences, the play of practices that circulate in a given configuration enables and constrains the process of institutional change. The analytical challenge is to show how the web of practices affords actions that are determinant in obtaining the institutional outcome. Empirically, I identify three sets of practices that proved key to the 2015–2016 renewal of the UN SG selection process (see Table 1).
TABLE 1. Pulling theory of agency: Mechanisms and empirical manifestations
The first puzzle is timing: how do challengers, given a stable distribution of preferences and power, manage to spark the process of institutional change? The pulling theory suggests that the proverbial window of opportunity opens when surrounding practices afford relational crossover (the generation of social ties across units otherwise separated) between challengers and incumbents. Recall that while institutions are particularly good at anchoring social relations, practices tend to generate them. Social ties are the result of joint accomplishments: communities coalesce around practices.Footnote 50 In a structurally blocked situation, challengers make use of available practices to generate new connections among them, and possibly get a foot in the incumbents’ door.
The first pulling mechanism (timing) then suggests that institutional change is more likely when available practices in the social environment facilitate relational crossover between politically significant groups. This makes a useful addition to the pushing story, which indicates that challengers push for institutional change when shifting structural conditions—especially the distribution of power between incumbents and challengers—induce new cost-benefit calculations on their part.
Second, in terms of form, why do challengers resort to specific modes of action in their struggle for institutional change? In the pulling theory, the particular shape that incremental change takes stems from available ways of doing things. By building on practices in surrounding webs, actors also facilitate the justification and implementation of heterodox ways of doing things through narrative work. Practitioners always start from somewhere in effecting transformation: the stock of existing ways of doing things, within and around their social configuration. Change takes shape on the drawing board of existing practices, which delimits the array of possibilities in line with the HI assumption of path dependence.
The second pulling mechanism (form) entails that in order to effect institutional change, actors resort to, and aim for, existing practices in their social environment or—thanks to narrative work and competence transfer—in its surroundings. This mechanism complements the pushing story, according to which actors pursue changes that are in line with their preferences, by specifying a range of contextually plausible options (the “what” question) and by defining the scope of instrumentally rational tactics (the “how” question).
Third, regarding magnitude,Footnote 51 what explains the extent of transformation? The pulling theory of agency suggests that incremental change is limited by the practices available to incumbents to push back against challengers. By holding fast to established ways of doing things, incumbents seek to close windows of opportunity and undermine competence transfers. Just like challengers, incumbents—who have a vested interest in preserving the status quo—mobilize practices available in the web. “Defenders of the institutional status quo are not just reactive ‘veto points’ standing in the way of reform; they typically take initiatives to slow down, channel, or stop change. To produce a compelling theoretical image of the politics of institutional change, we should theorize these strategies more explicitly.”Footnote 52 Even when someone mounts a successful challenge, then, the web of practices supplies materials for incumbents to resist.
The third pulling mechanism (magnitude) therefore indicates that the extent of institutional change depends on how existing practices in the social environment may be mobilized by incumbents in defending the status quo. This is an important addition to the pushing story, which suggests that the magnitude of institutional change turns on the balance of power between incumbents and challengers. Zooming in on how (that is, by what specific means) the political struggle plays out seems especially important when the distribution of power and institutional rules remain stable.
Table 1 lists the three mechanisms followed by key empirical manifestations from the case. First, in terms of timing, why did the changes occur in 2016, when similar calls (including in GA resolutions) had been made in the earlier two episodes? Here, an intentionalist account of agency faces two problems: first, the institutional terrain, including veto points and formal rules, changed little over time; and second, the challengers’ preference for change was already present in the past. I show how a set of newly available practices generated unprecedented connections among challengers and incumbents. Second, concerning form, why did the changes take the specific configuration they did, including public hearings and web surveys? A preference for opening up the selection process could accommodate a variety of practices other than those enacted. I demonstrate that three available practices in the UN web helped determine why certain options prevailed over others. Third and finally, with regard to magnitude, why did the changes stop where they did, with the SC maintaining its opaque practices at the nomination stage? I document three modes of operation by which the council pushed back on change despite external pressure.
Methodologically, I demonstrate the validity and usefulness of the pulling theory of agency by showing how certain practices, available in the UN web in 2015–2016 but not earlier, generated effects linked to the institutional outcome. For instance, the practice of hybrid panels, which is quite recent in Manhattan, allowed challengers to develop significant political ties in 2015–2016. I show that the nine practices listed in Table 1 played a key role in orienting agency, on the part of both challengers and incumbents. Counterfactually, my claim is that given constant structural conditions, “but for” these practices, the selection process of the UN SG would not have transformed in the way that it did. Recall that, from a pragmatist perspective, a theory—a set of portable and interconnected concepts—is useful, and hence valid, if it sheds light on key social processes, such that “we cannot imagine the outcome having occurred in [their] absence.”Footnote 53 The demonstration benefits from this being a “hard case”: despite the reluctance of veto players and the high political stakes involved, institutional change did occur—largely through the play of practice and its affordances.
My goal here is theory development, not hypothesis testing: concepts were developed abductively, moving back and forth between theory and empirics. Building on Fioretos, I suggest that my framework succeeds in “revealing novel facts,” “accounting for empirical anomalies,” and “fostering awareness of a series of temporal phenomena.”Footnote 54 Given the exacting demand in terms of empirical materials, space is too short here to apply the theory to other cases. Building on a hard and salient case, I seek to develop portable concepts and propositions that may later be embedded in other singular causal analyses. After all, as Keohane observes: “It is not reasonable to demand testable, ex ante hypotheses, much less predictive power, from an orientation that emphasizes the contingency of history and its dependence on complex sets of contextual factors.”Footnote 55
In tune with “practice tracing,”Footnote 56 the research was conducted inductively, gathering empirical materials about the case from various public sources, including the UN website, documents produced by Security Council Report (a New York-based NGO), media reports, and scholarly articles. Two background interviews were also conducted with senior policymakers from the president of the general assembly's office. Based on these data, which allow me to reconstruct the process in detail, I then climb the ladder of abstraction to identify analytically general mechanisms of institutional change. Essentially, the demonstration hinges on the mobilization of practices that, by being available in the 2015–2016 UN web, helped generate the three social effects of interest in shaping this case of layering: relational crossover, competence transfer, and pushback.
Since I do not conceive of pushing and pulling theories of agency as mutually exclusive, my demonstration is not meant to invalidate intentionalist accounts. After all, most social actions are simultaneously pushed by interests and pulled by the web of practices. Empirically, the practices that actors mobilize in my account do not go against their preferences. Yet they do specify how and why their interests got enacted in certain ways instead of others, which helps resolve the three puzzles of layering. To illustrate the added value of my theory, consider Terlingen's account, according to which three factors explain the 2016 renewal: first, challengers started early; second, the president of the General Assembly (PGA) exerted strong leadership; and third, civil society mobilized.Footnote 57 Without a doubt, all three factors mattered but it is striking how underspecified they are at the analytical level. Indeed, starting early does not, in and of itself, change much unless challengers also do certain things differently than in the past; strong leadership matters, but it is not sufficient in and of itself to force veto points such as the P5 to concede ground; and civil society activism often fails to shake jealously guarded intergovernmental processes. My practice account goes much further by specifying how, analytically, these factors actually mattered.
Renewing the Selection Process of the UN Secretary-General
The political significance of the case can hardly be overstated. First, the UN SG is one of the most important political actors in global governance, sitting “at the nerve center of a sensitive communications network, often speaking directly to governments, civil society representatives, and business leaders.”Footnote 58 The highest international civil servant initiates policies, brokers deals, and leads a variety of global actors based on unparalleled moral authority. Second, how this critical individual is selected is politically significant. Historically, “the person tasked with these extraordinary responsibilities [was] chosen through a process geared to select only the least objectionable candidates.”Footnote 59 Yet under the new process, the P5's preference for “a pliant Secretary-General” has become much more difficult to express, which is likely to transform the position's profile and visibility. According to observers, Guterres prevailed in 2016 because he “made the strongest impression” in the newly devised procedure.Footnote 60
Third, the case matters because of its spillover effect in global governance. Given the prominence of the UN SG position, other IOs have already drawn inspiration from recent changes. For example, in 2018 the UN PGA election included for the first time informal interactive dialogues with candidates, together with vision statements by applicants. In 2017 UNESCO collected similar documents from prospect directors, later to conduct public interviews in a plenary meeting of the executive board. In November 2016, the World Health Organization (WHO) plenary held a live forum with candidates, followed by an informal debate a few months later, departing from its old practice of holding executive board meetings behind closed doors. All in all, with a prominent precedent set at the UN SG level, “the new process set a higher standard for future office holders.”Footnote 61
Fourth and finally, the renewal has already transformed the political calculations that member states, including the P5, put in the selection process. According to Samantha Power, the American ambassador to the UN at the time, “these conversations mattered—there is no question that the General Assembly and other dialogues shaped perceptions, informing the Council and broader UN membership thinking from the outset.”Footnote 62 For example, the publicity about the new process “has made it much more difficult for the Council to recommend a weak candidate or to veto an outstanding candidate.”Footnote 63 The new process has also made it virtually impossible for the P5 to pull a “dark horse candidate” at the last minute, as the fate of latecomer Kristalina Georgieva demonstrated.Footnote 64 Intergovernmental dynamics, while remaining central to the selection process, have been reshaped by the grafting of more transparent practices.
How did the selection process unfold prior to the latest episode? The UN Charter does not provide much detail on how exactly the SG should be chosen. Article 97 states that “the Secretary-General shall be appointed by the General Assembly upon the recommendation of the Security Council.” Given the indeterminacy, in January 1946 the GA adopted a resolution setting parameters that remain valid to this day.Footnote 65 First, it established the length of the mandate to five years, renewable once. Second, it explicitly mentioned the need for an affirmative vote of all SC permanent members for the nomination to proceed. Third, the resolution recommended that the SC put forward only one candidate to the GA for voting. Fourth, the resolution provided for secret voting ballots, both at the SC and the GA, when deciding over the appointment of the SG.
Soon thereafter, the council adopted rules that enshrined secrecy as the cardinal principle of the nomination process. Rule 48 of its Provisional Rules of Procedure states that “any recommendation to the General Assembly regarding the appointment of the Secretary-General shall be discussed and decided at a private meeting.”Footnote 66 In ensuing decades, the procedure evolved incrementally and at the margins. Major innovation came in the early 1980s with the invention of “straw polls,” which I return to later. These incremental changes, which are largely uncodified, essentially solidified the council's predominance in the SG's selection.
Compared with earlier episodes, the process that led to the appointment of Antonio Guterres in 2016 saw significant changes, including the upstream formalization of the procedure via a joint letter signed by the presidents of the GA and of the SC, the publication of the candidates’ CV and vision statements, and public hearings and town hall meetings with the participation of all interested states as well as civil society. As the British ambassador put it in the midst of the process: “The days of smoke-filled rooms, of rumours and speculation on the runners and riders for the job, are over. Through consensus today we have brought overdue transparency to an archaic and opaque practice.”Footnote 67 In total, a dozen candidates were nominated for the position—the longest and most gender-balanced list in UN history. How did such an institutional change occur?
Timing: Practices Affording Relational Crossover
A set of practices that were part of the UN web in 2015–2016 enabled the generation of new social ties between challengers and some incumbents, allowing the pressure for institutional change to pick up. This relational crossover attenuated the weight of structural constraints, especially the P5 veto, helping challengers to open the proverbial window of opportunity. By showing how these very practices were either absent, or at least much more marginal in the 1996 and 2006 episodes, I account for the timing of the selection procedure's renewal.
It is worth noting that in the previous two appointment episodes, the GA had already passed resolutions demanding change. In the 1990s, after the American veto over Boutros Boutros-Ghali's reappointment, the assembly demanded for the first time that “the process of selection of the Secretary-General shall be made more transparent.”Footnote 68 Nothing happened at the time and the council even inscribed its prerogative over nomination through the Wisnumurti guidelines (discussed later). The next window of opportunity came a decade later when the search for Kofi Annan's successor began. In October 2006, the GA adopted a resolution reiterating the principle of transparency and raising “the need for the process of selection of the Secretary-General to be inclusive of all Member States.”Footnote 69 The resolution also encouraged the “formal presentation of candidatures … in a manner that allows sufficient time for interaction with Member States,” further “request[ing] candidates to present their views to all States members.”Footnote 70 Despite some attempts, these calls for action were not implemented though, and the SC forwarded the name of Ban Ki-moon to the GA for its endorsement.Footnote 71
Given this long-standing yet ineffective call for change, how can we explain the 2015–2016 layering? Three new practices in the UN web afforded unprecedented relational crossover among challengers, as well as with incumbents. First, a joint letter by the PGA and the president of the security council (PSC) made it possible for the PGA to take the lead in effecting change. Second, the practice of forming a group of friends fostered mobilization among (and beyond) a critical group of challengers. Third, hybrid panels expanded these connections between activist member states, the 1 for 7 billion NGO coalition, the Elders, and other interested parties, including from the P5. Neither of these practices were prevalent enough in earlier episodes to generate these decisive relational effects.
Joint letter
Let us start with the joint letter between the PGA and the PSC, an (almost) unprecedented move at the UN that proved to be a game changer.Footnote 72 By allowing challengers to break into the council's monopoly over the process, the new practice created the social conditions for “the key role played by the proactive president.”Footnote 73 There is no doubt that Mogens Lykketoft, a former social-democratic finance minister hailing from Denmark, was central to the renewal. Unless we study how the web of practices afforded his critical agency, though, it is difficult to understand how his preferences were turned into action. This is all the more significant that, by his own account, Lykketoft did not harbor strong views on the matter prior to his arrival in New York. His own election as PGA, largely thanks to the support of leftist governments in Paris and Berlin and to the practice of regional rotation, had little to do with this specific agenda.Footnote 74 In other words, the PGA's effective leadership was more the result of the affordances of the UN web of practices than of intense preferences or inherent institutional power.
Conditions were set in September 2015 when the GA adopted a resolution laying the groundwork for key changes to come. While most of the text was a repeat of similar past resolutions, it also “request[ed] the Presidents of the General Assembly and the Security Council to start the [selection] process … through a joint letter addressed to all Member States, containing a description of the entire process.”Footnote 75 This was not the first attempt to put the PGA on an equal footing with the PSC, but in 2006 a Canadian proposal had eventually failed to take off despite some support.Footnote 76 This time, the PGA office rapidly took it upon itself to draft the letter.Footnote 77 For its part, the SC endorsed the UK ambassador, Matthew Rycroft, as “penholder” on this issue—another relatively recent practice that proved critical because Rycroft was the most enthusiastic of his P5 colleagues.Footnote 78
Equally crucial was another coincidence owing to the practice of monthly rotating presidencies at the SC. From October through December, the three countries that held the chair (Spain, the UK, and the US) were more open to making changes than many others. For instance, in October the Spanish presidency invited the PGA to an open debate on working methods, during which he was able to lay out his vision publicly. The following month, the British presidency circulated a draft joint letter to all SC members. Had a less enthused country such as Russia occupied the presidency during the PGA-PSC negotiations, the drafting process would likely have occurred differently. Indeed, Russia opposed the draft letter during the fall, until a tripartite meeting with Rycroft at the PGA's office, during which the Russian ambassador finally agreed to let Lykketoft have his way, thinking that “no one would be interested in the procedure anyway.”Footnote 79 The American ambassador then successfully put the draft letter under silent procedure in early December.
ACT group lobbying
The second practice affording unprecedented relational crossover in 2015–2016 involved forming a “group of friends.” Under this mode of operation, which has significantly grown at the UN in the post-Cold War era, an ad hoc coalition of countries forms around a given issue to facilitate coordination.Footnote 80 United in their desire to infuse more transparency in UN working methods, notably at the SC, a group called Accountability, Coherence and Transparency (ACT), made up of twenty-five small and mid-sized countries from all regions of the world, formed in 2013 to succeed the defunct Small Five group (S5). The creation of this group enabled three forms of relational crossover. First, it allowed a variety of dissatisfied member states to come together early in the GA framework. For example, the Non-Aligned Movement (NAM) had long been calling for more transparency in the appointment process. In April 2015, ACT teamed up with them and used a GA working group to discuss the issue of SG selection. During the meeting, more than thirty delegations spoke out in favor of the changes that ACT was promoting.Footnote 81 Then in June, ACT proposed that a letter cosigned by the PGA and PSC outline in advance the selection process's key steps.Footnote 82 The group also advocated the formal presentation of the candidates at the General Assembly, council hearings of the candidates in Arria-formula format (a relatively new practice by which council members informally may meet with non-state actors), as well as a preset timeline for the entire process.
Soon thereafter, the co-chairs of the GA working group, Croatia and Namibia, began the process of drafting a resolution in line with ACT's proposal.Footnote 83 A contentious negotiation eventually delivered an agreement in September 2015 in the form of Resolution 69/321. The PGA also established close ties with ACT, sometimes out of coincidences that proved quite important. For instance, Lykketoft's point person on the selection process, Meena Syeed, was previously working for the Norwegian delegation—an ACT member. Similarly, the leader of Antonio Guterres's campaign, a Portuguese diplomat also previously involved in ACT meetings in New York, later proved receptive to the suggested changes, adding momentum to the push for change.
Finally, the ACT group was able to get a foot in the council's door by leveraging the timely election of some of its members, especially Chile, Jordan, New Zealand, and Uruguay. The group also worked closely with other elected members interested in the matter, such as Egypt, Japan, Malaysia, and Spain.Footnote 84 In terms of timing, these ties were critical in putting the issue on the council's agenda thanks to rotating presidencies. In June 2015, PSC Malaysia took advantage of the traditional end-of-month “wrap-up session” to discuss the nomination process. New Zealand repeated the exercise the following month, this time under “any other business.” Then, in October, under Spain's presidency, the PGA was invited to address SC members directly as part of the biannual debate on working methods, during which seven countries sitting on the SC voiced support for the proposed changes. In the spring of 2016, Egypt and Spain produced a “nonpaper” to resolve some outstanding issues within the council, including the modalities for meetings with candidates.Footnote 85 Overall, by engaging the SC through deft diplomacy, ACT was able to influence its dynamics to an unprecedented degree.
Hybrid panels
The last practice instated by a coalition of challengers in 2015–2016 was hybrid panels, which describe unofficial gatherings hosted in New York by think tanks or private foundations involving diplomats, international civil servants, experts, and civil society representatives. The practice emerged during the 1990s and has grown steadily since, to the point that “it is now commonplace for not only NGOs but other members of the wider UN diplomatic community (e.g., think tanks, foundations), to convene meetings outside the formal UN rules and structures, their purpose being to attempt to influence UN resolutions, UN actions, and even the Secretary-General's statements and speeches.”Footnote 86 Thanks to hybrid panels, limited but real overlap between institutional outsiders and a few insiders was generated in 2015–2016. Compared to the previous two practices, hybrid panels were critical in enlarging the circle of challengers beyond member states, allowing for close coordination with two groups hailing from civil society: 1 for 7 Billion and The Elders.
In November 2014, a coalition of NGOs named “1 for 7 Billion” distributed an open letter to all member states calling for a more transparent selection process. By 2016, 750 NGOs had signed up to the campaign, including the likes of Amnesty International and the World Federation of UN Associations. The campaign engaged in “intense lobbying” in New York to provide “diplomats with specific information and ideas prior to debates and during negotiations.”Footnote 87 For instance, it produced tables to publicize the different positions held by member states on a number of issues relevant to the appointment procedure. It also disseminated information on “comparative appointment practices of executive heads in UN Headquarters, the International Labour Organization, the World Trade Organization, and the World Health Organization.”Footnote 88
The 1 for 7 Billion campaign was aligned with a parallel civil society push initiated by The Elders, a self-appointed group of retired diplomats and political figures involved in matters of peace and human rights. In February 2015, The Elders took advantage of the Munich Security Conference—itself an increasingly significant practice allowing for significant relational crossover—to issue a statement called “Strengthening the United Nations” that contained proposals regarding the selection process of the SG.Footnote 89 They also published an op-ed in the New York Times cosigned by Kofi Annan and Gro Harlem Brundtland drawing much attention to their cause. Overall, the alignment of views between not only 1 for 7 Billion and The Elders, but also with ACT member states, for instance in advocating the nomination of female candidates, paved the way for an unprecedented coalition of challengers.
The three groups teamed up in a series of public events starting in June 2015, when ACT organized a panel discussion on the renewal of the SG selection process that featured representatives from The Elders and 1 for 7 Billion, as well as the UK ambassador. Then in September 2015, ACT and The Elders held a second panel on the issue, gathering high-level figures such as the president of Costa Rica, Estonia's minister of foreign affairs, Finland's permanent representative to the UN, as well as two Elders, former Norwegian Prime Minister Bruntland and former Mexican President Ernesto Zedillo.Footnote 90 Several supportive countries took the floor on this occasion, including some SC members, sending a strong signal of support among UN member states. Meanwhile, NGOs continued to lobby like-minded member states, such as Algeria for the NAM and Estonia for ACT.Footnote 91 In this endeavor, they received help from the PGA who had already started to establish ties with “a range of UN Ambassadors from those countries most interested in reforming the UN,” as well as with prominent Elders such as Brundtland, Martti Ahtisaari, and Lakhdar Brahimi.Footnote 92
In sum, the three critical practices that I highlighted go a long way in explaining the timing of the renewal. By comparison, in 2006 a small set of thirteen NGOs gathered around the World Federalist Movement had “signed an open letter to all heads of state, foreign ministers and diplomats in New York.”Footnote 93 Meanwhile, Canada had produced a nonpaper to propose some changes to the procedure.Footnote 94 Both of these isolated initiatives failed to take off, despite the fact that the 2006 Assembly Resolution contained a call for change similar to the 2015 one. Now mobilized and connected thanks to the PGA's brokerage position, ACT group lobbying, and the networking of civil society advocates, challengers achieved in 2015–2016 what they had long tried to do before: impose the issue on the council's agenda, despite the resistance of some of its members.
Form: Practices Affording Competence Transfers
Challengers mobilized a narrative of “best practices” to transfer standards of competence applied in other parts of the UN to the SG selection process. They also built on heterodox practices of global governance, such as public hearings and worldwide surveys, to shape the specific changes that occurred. Such practice transfers made it possible to problematize older ways of doing things and push for new modes of operation at the UN. Crucially, these three practices—the narrative of best practices, public hearings, and worldwide surveys—had become much more readily available within the UN since the last appointment in 2006.
Best practices narrative
As I argued earlier, to graft new ways of doing things onto an existing institution, challengers must be able to problematize how constituents assess competent practice. Historically, the UN SG nomination process was structured by standards associated with traditional great power politics. According to this view, the nominated candidate should be a compromise figure between the conflicting interests of the SC heavyweights. The practice of straw polls exemplifies this point particularly well: as an informal, secretive, and socially stratifying practice, straw polls belong to a form of exclusive and shadowy diplomacy that has long structured political relationships at the top of the system, including in selecting the SG.Footnote 95
Elsewhere in the UN, though, things started to change in the 1990s, when a counter-discourse of best practices made its way into various parts of the organization. For example, in 1997 Kofi Annan's UN reform plans were cast in terms of best management practices. The UN warmed to the narrative, which became a commonplace in its administrative work.Footnote 96 In a 2009 report, the Joint Inspection Unit (JIU) endorsed best practices for senior nominations within the UN, underlining the need for such processes to become more transparent and inclusive and pointing to “the necessity of ensuring high standards of professional performance.”Footnote 97 While noting that the selection process of the SG is “unique,” the report stressed that “prevailing practice in other system organizations is also suitable for the election of the Secretary-General.”Footnote 98 New narrative resources thus became available for challengers.
The very structure of the best practices discourse implies a transfer from one social domain to another because it “rests on the mobilization of practitioners toward transforming their practices” based on “the dual recognition of a lack of know-how and of a gap between available resources and the more or less explicit demands for transformation and adaptation.”Footnote 99 Proponents of best practices give direction to change by advocating a new set of standards of competence against which to judge performance within the organization.Footnote 100 In the case at hand, the narrative entailed that nominating a new SG no longer belonged to power politics and diplomatic horse trading, as traditionally assumed, but rather to the domain of good governance practices in which transparency is the cornerstone of competence. By implication, orthodox practices such as straw polls would have to give way to heterodox practices like public hearings.
There is ample evidence of such narrative work in the discourse put forward by proponents of change. Critics were quick to point out that “best practice for high-level appointments had significantly evolved in many countries and in some other international organizations.”Footnote 101 Comparisons with other IOs were explicitly drawn by the 1 for 7 Billion campaign. In their 2016 “Call for ACTion,” ACT members stated that “the identification and appointment of the best candidate for the post should be in line with the best practices in high-level appointments of executive heads of the United Nations and other international organizations.”Footnote 102 Most strikingly, the 2015 GA resolution made a direct reference to the 2009 JIU report and mentioned the need to build on “best practices” in selecting the next SG. Reflecting the rise of this narrative, one insider opined that “there really is an expectation now that the business of governance and policymaking should be carried out publicly and transparently.”Footnote 103 Instead of closed-door meetings and shadow deals, public hearings and open consultations were framed as the new, competent way of nominating the next SG.
Public hearings
The most remarkable innovation set in motion in 2015—2016 is the organization of public hearings with candidates. Never before had the selection process of the UN SG contained such a public dimension. Crucially, the precise format of the hearings was not provided for by Resolution 69/231: the GA simply decided “to conduct informal dialogues or meetings with candidates.”Footnote 104 Key in determining how hearings would occur was the leadership of the PGA, who confessed that he was “‘shocked’ by the anachronistic ‘closed-door’ means of selecting the UN's leader when he took up his post.”Footnote 105 Leveraging his brokerage position afforded by the joint letter with the PSC and his close connections with ACT members, the PGA also benefited from the historical proximity between his home country and the P3, comprised of France, the UK, and the US.
Once the Russians had agreed to the joint letter, Lykketoft followed through with a series of letters to the assembly outlining the format of the hearings, emphasizing how they would follow “established practice for informal meetings of the General Assembly (e.g., no pre-established list of speakers, Member States will pose questions from their seats on a first-come, first-served basis, GA seating protocol).”Footnote 106 In combining a reference to tradition with innovations afforded by the narrative of best practices, such as public debates, the PGA was able to imprint on the new practice of hearings with the candidates, starting in mid-April and followed with further rounds in June, July, and October 2016. The hearings, held in the Trusteeship Council room in UN headquarters, were webcast with translation into the UN's six official languages.
On top of this unprecedented publicity, Lykketoft and his team went one step further by allowing representatives from civil society to address questions to the candidates—an opening neither part of the GA resolution nor of the joint letter with the PSC. Building on momentum, the PGA continued to overstep his formal mandate and organized two high-level thematic debates (in collaboration with partners including the New America think tank and The Guardian), later followed by a “global townhall meeting” in the GA hall itself (with a live worldwide broadcast by Al-Jazeera). By Lykketoft's own account, the Russian ambassador “resented the live broadcast of the town hall debate. He questioned our right to conduct a live debate.”Footnote 107 His office pushed forward anyway, resisting the démarches from various ambassadors and compensating his lack of budget through partnerships, including with Al-Jazeera.
SDG-inspired survey
Worldwide consultations are the third practice that shaped the form of institutional change. Deciding to involve civil society in the process—a contentious move at the UN—the PGA's team took it upon itself to build a dedicated web page containing all the information about the selection process, with profiles of the candidates, verbatims and videos of meetings, and links to civil society consultations. At the PGA's request, the UN's Nongovernmental Liaison Service (NGLS) also set up a website through which anyone in the world could submit a question for the candidates using email, Twitter, or Instagram. To impose this initiative, Lykketoft turned to a particularly heterodox practice that had just been tried in the previous year as part of the making of the Sustainable Development Goals (SDGs): direct consultation with individual citizens around the world.
On the face of it, this innovation is puzzling, not only because it results from an initiative on Lykketoft's part (going beyond the GA's mandate), but also because it seems particularly foreign to traditional ways of doing things in global governance—all the more so in selecting the UN SG. The notion that individual citizens from around the world should have a say in selecting the world's top diplomat would surely have shocked the UN Charter's drafters. Yet because running global consultations had recently emerged as a competent way of governing the globe, the PGA was able to pitch the idea, explicitly acknowledging the connection: “That was really the first bottom-up process in the UN, with something like 8 million people around the globe taking part in the discussions and deliberations … That was an inspiration for me.”Footnote 108 The SDG precedent made it much easier to introduce this innovative way of doing things, while also facilitating its normative justification as part of the UN web of practices. In the spirit of bricolage, the practice was not invented from scratch but transferred from another initiative.
Magnitude: Practices Affording Pushback
The 2015–2016 push to renew the UN SG selection process did not go unopposed—far from it. Throughout, some SC members were uneasy with opening up the procedure and diluting their historical prerogatives.Footnote 109 Recall that Article 97 of the Charter grants recommendation powers to the SC, making it a de facto veto player in the nomination. Also key to the council's pushback was the adaptive flexibility of its procedure. As a former diplomat put it, “the Council's practices remain ad hoc.”Footnote 110 While heterodox practices were taking place elsewhere in the UN, at the horseshoe table SC members resorted to three orthodox modes of operation in order to limit the extent of change: private meetings with candidates, straw polls, and nonpapers. These affordances help understand why, despite some momentum, newly acquired transparency took the back seat in the closing stages of the selection process.
Closed meetings
The first pushback practice was closed meetings. As public hearings were unfolding at the GA, SC members decided to organize parallel, private meetings with candidates in the lead-up to the first straw poll in late July 2016. As the leading council voice on the issue, the UK had originally proposed to hold “Arria-formula meetings,” a practice invented in the 1990s to allow for unofficial but open meetings between member states and relevant nonstate actors. Faced with resistance inside and outside, the council finally decided to hold fully private meetings, following tradition.Footnote 111 These informal and closed-door hearings severed the possibility for non-insiders to participate in the critical nomination stage. Behind closed doors, the one-hour meetings “were held at the permanent mission of the Council president of the month, with Council members allowed only two representatives each.”Footnote 112 No records were made, countering in effect the transparency of the newly developed public hearings at the assembly.
Straw polls
The biggest brake on institutional change came from so-called straw polls, an informal practice developed over several decades by which SC members may anonymously “encourage” or “discourage” particular candidacies, short of a constraining vote.Footnote 113 Because straw polls take place in a private meeting, there are no records held, nor is there a communiqué issued afterward. Invented in 1981, the practice has since been entrenched to the point of structuring the nomination process. The practice was further refined in 1991, when color-coded ballots were introduced to allow for informal vetoes (permanent members cast their vote on a red ballot, whereas elected members use white pieces of paper). Finally, in 2006, a “no opinion” ballot was introduced, adding further flexibility to the council procedure.Footnote 114 Mostly unwritten, the practice has come to determine the recommendation that the SC makes to the GA.
Despite its secrecy and exclusivity, which go directly against the other changes made to the procedure in 2015–2016, there was reportedly no resistance in the council to maintaining the practice as it had evolved so far.Footnote 115 Held in the consultations room in the absence of Secretariat officials, straw polls gather only permanent representatives (with one aide) from the fifteen members. The first straw poll was taken in late July 2016, followed by several more rounds. Finally, on 5 October, the SC ran its sixth straw poll, using color-coded ballots for the first time. With thirteen positive votes, two abstentions, and no negative votes, Antonio Guterres was recommended by the SC the following day,Footnote 116 and the GA formally appointed the new SG by acclamation.Footnote 117 Despite the changes made in earlier stages, in the end the recommendation and appointment occurred in a way similar to past episodes, with the SC firmly in control and operating without much transparency.
Nonpapers
Finally, the SC resorted to a third practice to resist the push for change: issuing nonpapers and other soft forms of codification. This move allowed the council to counter the PGA's efforts at establishing new procedures. Historically, the SC has always been wary of codifying its practice, hoping instead to preserve as much flexibility as possible. The UN SG nomination is no exception to this pattern. One limited attempt was made in November 1996 when a set of guidelines for running straw polls, dubbed the “Wisnumurti Guidelines,” were agreed upon at an informal luncheon of SC ambassadors.Footnote 118 According to the document's opening statement, the Italian ambassador “prepared and circulated 100 copies of the ‘Wisnumurti Guidelines,’ and sent to, among others, the President of the General Assembly, the Secretary-General elect, the Secretariat of the Security Council, and the library of the United Nations.”Footnote 119 Importantly, the guidelines were never published as an official UN document, although a fact sheet was later prepared by the Secretariat.
In 2015–2016, facing mounting outside pressure, the SC witnessed additional attempts at inscribing its prerogatives. In May 2016 Russia distributed a nonpaper to formalize the procedure of straw polls.Footnote 120 France circulated another nonpaper with its own modalities in June 2016, leading to a council-wide agreement on the use of color-coded ballots during the straw polls to come. In typical SC practice, none of these documents were published or officially adopted.Footnote 121 The SC even innovated on its own practice by introducing separate ballots for each candidate, “in order to make sure that the pattern of voting did not reveal the Council member who had cast that particular ballot.”Footnote 122 In a clear reaction, then, there was even more secrecy in 2016 than before.
While effective in securing a nomination, the council's pushback quickly backfired. Three practices mobilized by challengers suggest that the political struggle is far from over. First, the PGA strongly objected to the limited communication of results made by rotating PSCs. As Security Council Report (SCR) explains: “following each straw poll, the [council's] president orally informed the president the General Assembly that the vote had taken place but did not communicate its results. The Council president also orally informed the president of the General Assembly of the planned date(s) of future straw polls.”Footnote 123 When the first straw poll was taken in late July 2016, the Japanese PSC sent a letter to the PGA to inform him that the process had started, though without communicating any results, in line with established practice. Lykketoft reacted angrily: “In my view, limiting the communication to the fact that the informal straw poll has taken place without any further detail adds little value and does not live up to the expectations of the membership and the new standard of openness and transparency.”Footnote 124
Second, new information technologies posed a significant challenge to the secrecy of straw polls: “in a matter of minutes in every round, the exact details were quickly known and broadcast around the world.”Footnote 125 It appears that one or more “pro-transparency members” of the SC leaked the information on Twitter.Footnote 126 Faced with growing public pressure, most SC members came to favor less secrecy in divulging straw poll results, though in the end Russia stood firm.Footnote 127 According to one observer, the leaks exposed “the self-defeating and ridicule-inviting secrecy of the process,”Footnote 128 especially after public hearings and consultations with civil society. In a rare admission, the Japanese delegation later recognized how the straw poll practice had lost touch with the era: “There is no denying that the Security Council was unable to respond effectively to calls for transparency, which were brought to a higher level than ever before by the informal dialogues of the General Assembly.”Footnote 129
Third, activist member states moved quickly to codify the changes made in 2015–2016 and establish precedents for subsequent selection processes. Before the appointment was done, the GA adopted a resolution to inscribe the changes set in motion so far.Footnote 130 A Japanese wrap-up letter sent to the SG in February 2017 summarized the council's process. Then in October 2017, ACT member Estonia sent a letter to both the PGA and the PSC, lauding recent changes but calling for more (e.g., abandoning color-coded straw polls, communicating results openly, and setting a deadline for nominations).Footnote 131 The codification efforts have continued since, yet the “lessons learned” remain contested. Without a doubt, the stage is set for a continued political struggle when the time comes to select Antonio Guterres's successor.
Conclusion
Making sense of incremental institutional development is a major task for students of world politics. Layering is highly prevalent on the world stage because IOs are particularly prone to small-steps institutional reform.Footnote 132 Analytically, such institutional change presents daunting challenges, especially at the explanatory level (because the process explains the outcome) and at the methodological level (since informal dynamics are particularly hard to study empirically). In this endeavor, combining insights derived from historical institutionalism and practice theory holds the promise of developing efficient analytical tools.
The pulling theory of agency offers a necessary complement to the intentionalist accounts of institutional change that prevail in IR, including in HI. Practice pulls the train of agency just as much as interest pushes it.Footnote 133 In the case at hand, the preferences harbored by challengers and incumbents do explain the position they took in the process. For instance, the PGA's leadership and civil society mobilization played a key role. Yet in and of themselves, these factors cannot fully account for the timing, form, and magnitude of institutional change. For instance, Lykketoft's oversized influence owes to his connections with other challengers, which were themselves afforded by hybrid panels and group-of-friends diplomacy, in addition to contingent openings at the SC, such as rotating presidencies, a recent joint letter precedent, and British pen holding. For its part, civil society entered the process through public hearings made possible by unprecedented SDG consultations, and the growth of the best practices narrative within the UN. Finally, institutional change was partly limited by the council's practices of closed meetings and straw polls. Overall, a full account of the SG selection process's renewal is impossible without understanding how the UN web of practices afforded some innovations and not others.
The pulling account also matters empirically because it helps us to understand the case's significance. The specific practices documented here will likely continue to transform in the future, though incrementally and based on the UN web's affordances. While certain proposals failed to take off, most strikingly the push for a female SG, some precedents have been set. One observer speaks of “a game-changing process that will be difficult to reverse.”Footnote 134 It seems unlikely that the next SG will be selected without public hearing and consultations with civil society. The SC will probably continue to push back to maintain its institutional prerogatives, although the playing field has already shifted. This is not even to mention the difficulty of maintaining secret straw polls after public hearings and in the shadow of leaks. Precisely because this case of institutional change takes place via contending practices, most of which lack codification, it will continue to be structured by the broader UN web of modes of action, which facilitates some forms of agency but not others.
This is a critical argument for bringing HI and practice theory together in explaining institutional development in world politics. By nature, the study of incremental transformation concerns itself with changes in—as opposed to changes of—a given institution. Here, practices form both the instruments and the increments of layering. It is obviously too early to tell whether the UN as a whole will transform down the line. Yet it is already clear that the changes set in motion in 2016 are diffusing to other UN agencies and beyond, gradually shifting political calculations and expectations on the part of member states, and possibly granting more authority and visibility to the world's top civil servant. Ultimately, the future of the institution will be shaped by how agents struggle with one another under the pull of the UN web of practices.
Acknowledgments
For excellent research assistance, I am grateful to Nina Jaffe-Geffner, Zaheed Kara, and Scott Patterson. For useful comments on earlier versions, I thank the IO editors and reviewers; Steven Bernstein, Tom Biersteker, Christian Bueger, Alena Drieschova, Tine Hanrieder, Ted Hopf, Andrew Hurrell, Erik Kuhonta, Dan Nexon, Lou Pingeot, Meena Syed, Lora Ann Viola, and Michael Zürn; as well as seminar participants at Cardiff University (October 2016), McGill University (February 2017), WZB-Berlin (April 2017), HEI-Geneva (September 2017), and Oxford University (November 2017).
