The de-institutionalisation of power beyond the state

Abstract

The making of modern authority centred on efforts to formalise and de-personalise power, and transnational orders such as the European Union have often been viewed as an extension of that project. As this article argues, recent developments tell a different story. More than a decade of crisis politics has seen institutions subordinated to and reshaped by individuals and the networks they form. Locating these tendencies in a wider historical context, the article argues that greater attention to informality in transnational governance needs to be paired with greater recognition of the normative questions it raises. Just as a separation between rulers and the offices of rule was central to the making of modern legal and political structures, the weakening of that separation creates legitimacy problems for contemporary authorities both national and supranational. Rather than acclaimed as flexible problem-solving, the step back from institutions should be viewed as a challenge to accountable rule.

International organisations are conventionally portrayed as the sphere of law, institutions, and anonymous bureaucracy. ¹ Power, goes the thought, has been abstracted from individuals and embedded in impersonal frameworks, leading to an approximation of the rule of rules. In an age of crisis decision-making, can such a perspective still be maintained?

Today’s era of crisis politics forces us to reckon with the de-institutionalisation of power beyond the state. Fast-moving, challenging circumstances encourage agents to sidestep chains of command, pointing to exceptional circumstances as reason to depart from accepted script. While deviations of a constitutional kind are a well-recognised feature of emergency governing, for example in the form of executive dominance, less discussed are those at the level of institutions, as arrangements based on informality arise. Such dynamics can be observed in the recent experiences of the European Union (EU), the most consolidated of transnational orders, as its leaders have grappled with a succession of challenges centred on finance, migration and health.

This article develops an account of authority beyond the state that captures these tendencies and why they matter. Against the assumptions of institutionalist perspectives in International Relations (IR) and EU studies, it shows how informalisation and personalisation prosper in this context and come to the fore in hard times. A precarious achievement at the best of times, institutions and their procedures tend to soften quickly when the politics of emergency takes hold. Against the grain of a specialist literature that acknowledges and engages with these patterns, the article urges more attention to the normative costs. Scholars of informal governance stress its functional benefits, be it in solving coordination problems or managing power shifts (Abbott and Faude, 2020; Vabulas and Snidal, 2013). The flexibility of informality is said to make it ‘especially valuable during crises when leaders must move quickly’ (Vabulas and Snidal, 2013: 211). As this article argues, more attention should be paid to what these arrangements imply for the accountability, contestability and legitimacy of authority (cf. Roger, 2020: 213). Circumventing institutions can aid actor speed and discretion, but for the same reason it can shade into arbitrary rule.

The article’s point of departure is a contrast: the wider development of modern political authority as a project of institutionalising power. As the first section shows, by revisiting its history we can retrieve some of the normative issues at stake. Drawing on the work of Gianfranco Poggi (1990), the text shows how the making of the state involved the pursuit of a separation between the persons who ruled and the offices of rule. It entailed abstracting from the particularities of the former – their motivations, opinions and interpersonal ties – so as to establish a framework that could shape and constrain them. Institutions had to be made, and their construction supported a project of rationalising the exercise of rule.

Transnational authority structures, the second section observes, tend to be viewed as an extension of this. In the political science of transnational orders, the centrality of institutional dynamics has tended to be assumed (e.g. Garrett and Tsebelis, 2001). Yet there are good reasons to see things differently. Structural features of the transnational realm create distinct possibilities for the de-institutionalisation of power, ones likely to be actualised in crisis moments. The section outlines the basis of an alternative perspective, combining recent IR scholarship on informal arrangements (Mérand, 2021; Roger, 2020; Westerwinter et al., 2021) with work on transnational emergency politics that highlights exceptional methods and the legitimacy questions they raise (Auer and Scicluna, 2021; Heupel et al., 2021; Kreuder-Sonnen, 2019; Kreuder-Sonnen and White, 2021; White, 2019).

The article’s third, empirical section examines what the de-institutionalisation of power means in practice. In the contemporary EU, a period of crisis politics has seen institutions subordinated to and remade by individuals and the networks they form. In addition to the use of informal forums of decision-making, one can observe changes in the established sites of supranational authority. Looking at the Commission and European Central Bank (ECB), the section highlights the concentration of power in individuals, such that key decisions are taken by the few; the collaboration of leaders across institutional boundaries, such that ties of trust override the formal definition of roles; and a reliance on personalised authority, such that emphasis falls on personal judgement, expertise and discretion. As the article argues, these tendencies create a template liable to recur, in the EU and transnational governance more widely.

The final section conveys why this matters. While the informalisation of power tends to be acclaimed as improving the effectiveness of rule, and personalisation as aiding its legibility, things look more challenging from the perspective of political thought. By weakening the separation between rulers and offices that was central to modern legitimacy, these tendencies weaken power’s accountability and the facility to contest it at the level of ideas. They ease the passage of controversial measures in the guise of crisis response. They also make it harder for the polity to retain public acceptance despite the failings of particular officials. Institutions that can withstand the repercussions of emergencies are a neglected precondition of any desirable transnational consolidation, and to reflect on their foundations is ultimately to revisit some basic questions of sovereignty.

Modern authority and the institutionalisation of power

At least since Max Weber, the making of modern authority has been grasped by state theorists as the rise of rules, institutions and procedures. IR’s story of state-formation has tended to focus on units in interaction – their relative power, claims to non-interference, and the properties of the Westphalian system (e.g. Waltz, 1979). The history of domestic change underpinning these developments has always received less attention (for critical discussion: Osiander, 2007, ch. 1). But at the core of the state-building process in Europe was a project of separating political power from the persons able to wield it at a given moment – what Weber (1968 (1922)) called the ‘general tendency to impersonality, the obligation to conform to abstract norms’ (p. 294). ² Whereas the feudal orders of late-medieval Europe were based mainly on informal and personal ties, later political forms sought to abstract from these.

It is important to begin by tracing these changes – first, because they highlight that the institutionalisation of power was historically tied to state-formation, hence not necessarily to be expected in non-state contexts, and second, because they indicate some of the normative issues at stake, ones relevant to the assessment of contemporary trends. Engaging with the scholarship on these changes also yields categories of analysis that can be transposed to the supra-state setting.

The transformation of power in early-modern Europe was on the one hand a phenomenon of changing political language. In a classic study, Ernst Kantorowicz famously documented the conceptualisation of authority as it evolved in the medieval period. From the late 1100s, he observed, discourse in England and France increasingly came to distinguish the king, as an individual, from the ‘idea of an impersonal crown’ (Kantorowicz, 1957: 347). The latter was something that would outlive the king, and to which the king had certain responsibilities. He would start to be viewed as its guardian, expressed by the early 1400s in an oath sworn at the point of coronation (Kantorowicz, 1957: 358). Such distinctions were inspired in part by the ecclesiastical model, where it was already customary to adopt the corporative view, abstracting from tangible individuals to evoke a larger, enduring collective (Kantorowicz, 1957: 194, 209). ³ In the image of the Church, monarchy started to be experienced as an institution.

Such distinctions anticipated the later language of statehood. In the medieval period, such terms as status, stato and état still carried personal connotations. They referred to the attributes of individuals in their specificity, resulting from their lineage and their prowess in battle – something closer to the ‘status’ of a ruler. In Harvey Mansfield’s words, in this period ‘the Latin status does not stand alone, but requires some accompanying word or phrase to specify whose status’ (Mansfield, 1983: 851). One could write of status regalis – something like ‘the king’s majesty’ – but not of status as an independent arrangement. Only from the 1400s did the term acquire more abstract connotations – for example, status regni, the ‘state of the realm’, evoking a territorial space – and with it the potential to be detached from the figure of the prince (Mager, 1968: 83; cf. Shennan, 1974: 25). ⁴ A depersonalised conception of political authority, and with it the capacity to speak of such abstractions as ‘the state’ without adjectives, and of citizens rather than subjects, arguably found its clearest expression in Renaissance republican thought (Skinner, 2009). It would be refined in key texts of the late 16th and early 17th centuries, notably Bodin’s Six Books of the Commonwealth and Hobbes’ Leviathan.

The transformation of power was equally a matter of changing practices. Medieval monarchies were for the most part maintained on relations of interpersonal trust and loyalty. The king surrounded himself by favoured associates and relied on the networks of his household to rule. Coordination between agents in the feudal system was achieved by oaths of allegiance and related rituals; a vassal paid ‘homage’ to his lord, indicating that he was his lord’s man (homme). These were first and foremost ties between determinate individuals (Poggi, 1990: 38). By contrast, what started to emerge in the early-modern period were less personalised modes of rule. Impersonal ‘oaths of office’ started to replace oaths of allegiance to a particular lord (Kantorowicz, 1957: 354). Early bureaucracies began to take shape, differentiated in their purposes, regulated by procedures such as auditing, and staffed by those chosen mainly for their ability (Shennan, 1974: 70). Such enduring entities could be tasked indefinitely with managing the affairs of the ‘state’, outliving the individuals that composed them.

Both then in the language and symbolism of political authority, and in the way decisions were taken, one can observe moves towards impersonal forms. The making of the modern state was the making of an abstraction: a political order detachable in some measure from the identities and inclinations of individuals. IR scholarship, insofar as it engages with the history of state-building, has tended to emphasise centre-formation and technological change (see notably Spruyt, 1994). But it was the abstraction of power from determinate individuals that was arguably the decisive novelty – the basis of efforts to systematise rule, and also to grasp it as a system. It was, on one level, the making of the distinction between offices and persons. Officials could represent the state, but did not embody it. Offices and the roles attached could be transferred from one individual to another, even left vacant for a while. On another level, it was the making of an order expected to transcend idiosyncratic or sectoral preferences: to be impersonal in the sense of ruling for non-personal ends, responsive to generalisable priorities rather than arbitrary whim (Shennan, 1974: 28ff; cf. Mansfield, 1983: 852; for a rare engagement from an IR perspective: Osiander, 2007: 8–10).

These tendencies were a combination of practical innovations, imperfectly realised, with normative ideals that reached beyond them. Political and constitutional thought has developed and embroidered them in countless ways, from defences of the Rechtsstaat to accounts of the substitution of the ‘rule of men’ with the ‘rule of law’ (Bobbio, 1987, ch. 7; cf. Eulau, 1942). The rise of modern legal systems, centred on norms of general application, has been a natural focus, but analogous processes have been discerned in contexts beyond the legal field. The embedding of leadership in political parties, technocratic agencies and other forms of organisation are all variants of the same wider movement towards abstraction.

In his landmark account of the making of the modern state, Gianfranco Poggi summarises the transformation of early-modern political power as a process of institutionalisation (Poggi, 1990: 18ff., 33ff., 75ff.; cf. Popitz, 1986). It encompassed, he suggests, three parallel tendencies. One was the ‘depersonalisation’ of power, whereby individual identities and interpersonal ties became less significant on the side of both rulers and ruled. The second was the ‘formalisation’ of power, whereby its exercise came to be standardised through laws, rules and procedures. The third was the ‘integration’ of power into a greater social whole, whether by reference to ideas of nationhood or other principles of representation. The state, he suggested, embodied these tendencies ‘to a high degree’ (Poggi, 1990: 33), but they expressed a transformation of authority more generally. As we shall see, these categories hold considerable potential for the analysis of politics beyond the state.

The institutionalisation of power was something distinct from its democratisation. The reordering of early-modern rule proceeded for some time without widespread demands that it accommodate popular will. Only with the emergence of the legislature as the primary source of law, and with parties as collective agents seeking to control it, did institutionalised rule start to intersect with the democratic (Bobbio, 1987: 143). But even if institutionalisation was distinct in this sense, it was always a process with normative weight. Abstracting power from persons and informal ties was a necessary condition of making it more accountable. Only then could it be scrutinised from without and contested at the level of general principle. Through driven by a variety of factors, the making of institutions was first and foremost about reining in arbitrary rule.

Transnational authority as extending the institutionalisation process?

How do transnational authority structures relate to the above-mentioned dimensions of institutionalisation? The third of Poggi’s categories – the integration of power into a social whole – has generally been considered weak. There is no transnational demos, it is said – no cross-border community to which supranational authority can meaningfully appeal. Authority in this context is technocratic authority, at least in its primary features. Given these points are well-rehearsed, we shall postpone further discussion of integration to the final section. But the other two elements of institutionalisation in this scheme – the depersonalisation and formalisation of power – are widely assumed to be very present in transnational contexts. As I want to argue here with reference to the EU, the making of transnational authority has generally been seen as consolidating and extending the institutionalisation of power – a reading which in key respects is misleading.

Sometimes fondly, sometimes disparagingly, the EU has been seen since its inception as the archetype of power detached from individuals and their confidants. By its admirers, it has been celebrated since the days of Walter Hallstein as a community of law – as a rule-bound order, even if lightly democratic. The appeal of European integration in the post-War period lay in its promise to constrain the wayward exercise of executive power (Müller, 2011). The lesson taken by many from the experience of fascism was that the regularisation of power was a precarious achievement, liable to be reversed by the appearance of charismatic leaders pursuing authoritarian rule. Precisely because the project of formalising and depersonalising power was incomplete and threatened, the role of transnational structures was to consolidate it.

Equally, by their detractors, transnational structures have been criticised as depersonalising and formalising power to an excessive degree. They have been caricatured as ‘faceless’ bureaucracies, overly devoted to rules and procedures, and with an exaggerated fear of political agency. Such critiques have become prominent in the characterisation of the EU in recent decades, in the age of Brexit and public euro-scepticism (Lacey, 2017: 2). They portray the EU as a kind of reductio ad absurdum – the culmination of the modern project of institutionalising power, taken to its logical extreme.

Less normatively, accounts of how transnational structures function have tended to internalise the idea that they are a continuation of the formalisation and depersonalisation of power. When scholars take aim at IR-realism’s image of an anarchic international order, it is generally by casting the transnational sphere as a world of institutions and procedures. This is true of scholarship on international organisations generally (e.g. Barnett and Finnemore, 2004), and of that on the EU specifically. Varieties of ‘institutionalism’ are the backbone of research in this field, based on the view that the EU is ‘without question the most densely institutionalised international organisation in the world’ (Pollack, 2018: 108). Adopting an ontology of politics centred on institutions both national and supranational, these accounts embrace the view that the Union is best understood in terms of the interplay of corporate forms, relegating to a subordinate role such factors as the leadership of individuals, the networks surrounding them, and the material and ideological forces to which these are exposed (for critical overviews: Jenson and Mérand, 2010; Schmidt, 2018). ⁵ Treating the EU as an extension of the institutionalisation process, they are predisposed to see its actions in these terms.

Without discounting the value of institutionalist perspectives, one may note some important limitations. To begin with, such accounts take as their empirical starting-point what was always an ideal, only ever approximated. Just how far modern political power was ever truly depersonalised and formalised in any context, including the state, is debatable. Weber tends to be read as treating charisma as a pre-democratic logic, supplanted by bureaucratic rationality, but there can be little doubt that it lingered in modern politics. Marxists are among those who have argued that concepts such as the state can be misleading in their abstraction from social power, given powerful interests and ideologies tend to skew the exercise of rule in favour of some (Mansfield, 1983). In the most sceptical line of reasoning, institutions are little more than a cover for the exercise of power by individuals and the classes they represent. One does not need to accept this strong version of the thesis to observe that the distinction between rulers and the offices of rule is a precarious one, and that there may be settings where it is especially vulnerable.

One should note moreover that several global changes in contemporary societies run counter to the institutionalisation of power. Poggi writes of ‘a tendency for contemporary state structures to regress toward political arrangements similar to those preceding the emergence or the maturity of the state’ (Poggi, 1990: 184), and this general observation applies well to the features in question. Partly this stems from changes in the media environment, as emphasised in an extensive contemporary literature on personalisation (e.g. Garzia, 2014; McAllister, 2007). Important also to observe is that the depersonalisation and formalisation of modern politics, to the extent it was achieved, corresponded to the demilitarisation of societies. The concentration of power at the top of the feudal pyramid had always in part been an artefact of preparations for war, while the personal authority associated with prowess (courage, valour) tended to be won on the battlefield (Poggi, 1990: 35). These were modes of rule oriented to crisis settings. Only when politics was reoriented to a predominantly peacetime footing, with the Westphalian order and successive arrangements, could other logics decisively emerge.

While open military engagements remain rare in today’s Europe, an orientation to security is not. The emergence of Covid-19 has amplified this, but there is a longer trend in play. Recent decades have seen the reappearance in peacetime of emergency modes of governing and the securitising logic that underpins them (Calhoun, 2004). Emergency rule is a paradigm with a variety of sources today: the material pressures associated with finance capitalism and social acceleration, the influence of powerful interests aiming to consolidate preferred policy regimes under conditions that challenge them, and changes within executive power itself, notably leaders searching for new ways to secure their authority with mass publics (Scheuerman, 2004; White, 2019). Important for our purposes is that this is a mode of rule that fosters irregular forms of decision-making and individual discretion.

If these are general tendencies in contemporary politics, one can assume they will make themselves felt in the transnational setting. Recent research suggests emergency modes of government are now a feature of both supranational and intergovernmental authority, offering much opportunity for informality and personalisation (for a ‘grammar’ of exceptionalism in international organisations (IOs): Kreuder-Sonnen, 2019: 32ff; cf. Kreuder-Sonnen and White, 2021; White, 2015a, 2015b). What is more, when such tendencies unfold, their impact will be more problematic. Transnational orders tend to be advocated as communities of law and procedure, meaning that practices at odds with this weaken their rationale. Moreover, the link is weaker in the transnational sphere between authority and electoral accountability. In a conventional presidential system, emergency situations encourage the personalisation and centralisation power, but for some this is acceptable insofar as the president must present themselves to an electorate (Posner and Vermeule, 2011). At the transnational level, when politics takes place outside institutions constrained by the law, it tends to be by actors over whom voters have little control. The decisions they take can also be especially hard to undo, given the weakness of judicial review and the many veto players involved. The normalisation of emergency measures becomes a likely prospect, such that the effects of exceptionalism are long-lasting (Kreuder-Sonnen, 2019: 43ff).

Finally and most importantly, the transnational sphere offers its own additional possibilities for the de-institutionalisation of power. To the extent that tendencies towards depersonalisation and formalisation were successfully fostered in the modern state, they were supported by the development of binding constitutional arrangements. The functioning of institutions and the relations between them had to be codified, and this backed by meaningful sanctions. The ideal of sovereignty and its attendant hierarchies encouraged this process of codification (Eulau, 1942). While some argued this process was unavoidably incomplete (Schmitt, 2005), these constitutional structures nonetheless put obstacles in the way of unchecked personal discretion. So too did the emergence of corporate forms of agency such as the political party, defining themselves by collective goals and tying their leaders to an organisation (Bobbio, 1987: 51ff.). These features have always been less developed transnationally. Although the workings of international organisations have been circumscribed by conventions of practice and statements of principle, these constraints have been weakly undergirded by sanctions for their breach and typically made conditional on their capacity to serve particular policy commitments (Isiksel, 2016). When emergency powers are pursued in international organisations, it is generally in a manner that has not been codified in advance, and by actors who lack sovereign authority (Kreuder-Sonnen, 2019: 38, 2021; White, 2015b). They must therefore cajole others into supporting their actions, with all the reliance on informal networks this is likely to entail.

These features present a challenge to the tendencies towards de-personalisation and formalisation highlighted by Poggi. One can expect them to be actualised in various forms. One will be informal IOs that lack much in the way of operating procedures to begin with, as IR scholars have started to observe (Vabulas and Snidal, 2013). Entities such as the G20 arguably exist to give national leaders a context in which to make policy without the procedural constraints of domestic politics (Roger, 2020). Equally, one can expect irregular practices within and around ostensibly more formal supranational structures (Kleine, 2013; Stone, 2013; Westerwinter et al., 2021). To get the measure of what these tendencies to informalisation and personalisation entail, and as the basis for their critical appraisal, it is worth examining them more closely in the context arguably least hospitable to them, yet dense with decision-making power: the EU.

De-institutionalisation in practice: emergency rule in the EU

There is a long-standing tradition in EU scholarship of narrating EU politics in terms of the changing fortunes of its leading institutions. This focus has been evident as authors take stock of the Union’s transformations since the early 2010s, charting the relative gains of supranational and intergovernmental institutions (e.g. Bauer and Becker, 2014; Bickerton et al., 2015). One reason such accounts are valuable is that they connect current affairs to historical processes and wider explanatory theories. A focus on institutional dynamics helps avoid casting episodes of high drama as worlds unto themselves. Yet as I want to suggest in this section, much that has happened in the EU’s recent history is difficult to grasp in these terms. As the accounts of participants, journalists and some scholars suggest, crucial aspects of the handling of the Eurozone crisis, migration politics and Covid-19 do not map easily onto an institutional grid. One sees actions that cut across institutional boundaries and blur them, and which divide institutions from within, bringing individual discretion to the fore. What these actions amount to, I suggest, are de-institutionalisation. ⁶

Keeping to Poggi’s categories, we are interested in patterns at odds with the formalisation and de-personalisation of power. Starting with the former, first to note is the reliance on informal forums of policy-making in contemporary EU decision-making. The ‘Eurogroup’ is one example – an unofficial entity rather than an institution as such, bringing together national finance ministers and Commission representatives. It was central to efforts to manage the Euro crisis in the 2010s and was also relied on to initiate the EU’s response to Covid-19 in spring 2020. ⁷ Its existence indicates the importance of informal ties and has been celebrated by participants exactly for its absence of codified procedures. ⁸ With no rules by which weaker members might seek to hold their ground, for example, by threatening a veto, pressure is more readily applied (Transparency International EU, 2019: 5). As a recent study puts it, European economic governance relies on informal arrangements ‘as it remains in the interest of powerful countries to exert power behind the curtain’ (Abels, 2019: 532). While the Eurogroup has been the standout informal forum of the EU sphere in recent years, until recently the European Council played a similar role. Summits between heads of state were a way to pursue crisis decision-making unencumbered by the constraints of the Community method (White, 2019: 56–60), a practice that retained its informal character until brought into the EU framework with the Lisbon Treaty.

What is notable in the EU context is how informal structures of this kind are paired with informal practices in and around the established sites of supranational power. A pattern recorded in both the ECB and the Commission in this period has been the ad hoc concentration of power in individuals, as bureaucratic hierarchies are circumvented in the name of crisis response. At the ECB, decision-making at key moments has involved a powerful individual (Draghi, Trichet) surrounded by a small circle of trusted aides, governing by informality and discretion (Verdun, 2017). A persistent observation of those close to Draghi was that he operated in a highly centralised fashion, running affairs through a ‘kitchen cabinet’. Members of the ECB’s Governing Council have spoken of informal ‘meetings before meetings’ and ‘Draghi’s practice of preparing big decisions with a close circle of advisers and involving the ECB Governing Council only late in the process’. ⁹ Likewise at the Commission during Covid-19, it has been said by close observers that ‘decisions are made too fast and in a narrow circle’. ¹⁰ In the context of criticisms of the Commission’s vaccine procurement, President Von der Leyen was said to employ a ‘silo’ style of management that excludes commissioners and senior officials in favour of a ‘tightly knit group’. ¹¹ These criticisms echo those made in an earlier phase of the pandemic, ¹² and are also consistent with research on the Commission during the economic upheavals of the 2010s (Mérand, 2021: 16). While systematic studies of power-concentration in this context are few, one may note also the small body of EU scholarship that identifies tendencies towards ‘presidentialisation’ in the supranational sphere, accelerated in an emergency context (Becker et al., 2016).

This concentration of power has tended to be coupled in the EU with a different kind of informality – the collaboration of leaders across institutional boundaries. In the name of solving immediate problems, those formally tied to different institutions have collaborated in ad hoc ways. The activities of the eurozone Troika, which brought together figures associated with the ECB, Commission and International Monetary Fund (IMF) to oversee the conditions for emergency lending, are the clearest example. Internal evaluations of the Troika suggest a pattern of individuals and groups substituting institutional procedure with improvised collaboration ¹³ (cf. Beach and Smeets, 2020a, 2020b; Henning, 2017). Likewise, an analysis of EU migration policy after 2015 observes that ‘collaboration was informal and flexible, enabling actors to sidestep bureaucratic hierarchies and to cross the intergovernmental / Community pillars’ (Beach and Smeets, 2020a: 12, 2020b: 143). Informal networks, taskforces and other extra- and trans-institutional arrangements have become recurrent features of EU governance. Individuals acquire the role of ‘fix-it’ people, redeployed from one area to another to reinforce the EU’s basic policies against the challenges of the moment. Such formations do not map well onto institutional units and are defined less by issue-specific competence than the capacity of individuals to form good workings ties (Beach and Smeets, 2020a: 14–15, 2020b: 147). Importantly, they are networks founded on interpersonal trust, and their centrality has corresponded to the distrust of established institutions and chains of command, notably around the Community Method. They are primarily connections based on persons in their specificity rather than codified relations between offices.

The centrality of trust-based networks itself amounts to a kind of personalisation of power, ¹⁴ even when largely behind closed doors. But it also fosters changes in the public image of power, involving a reliance on personalised authority. In media commentaries, but also in the self-presentation of leaders, one sees an emphasis on key persons and their attributes – understandably, given their tendency to free themselves from bureaucratic routines. Draghi is again one example, lauded for instance in 2012 by the Financial Times as ‘person of the year’, ¹⁵ and in many reports over the years as a hero for his crisis management, often as ‘Super Mario’. ¹⁶ Clearly, in part such discourse simply reflects the values of the media, keen to enliven its stories with figures of praise or blame. (One is reminded of such constructions as ‘Merkozy’ and ‘Merkron’ used to describe Franco-German cooperation in this period, serving to personalise one of the key dynamics of power in the Union (Schoeller, 2018).) But an emphasis on individuals and their qualities is also fostered by leaders themselves. Draghi’s peers hailed him on his departure from the Bank as the man who ‘saved the eurozone’, ¹⁷ deserving a ‘place in the pantheon’. ¹⁸ Moreover, his self-presentation has been analysed as an instance of charismatic authority (Lokdam, 2020; Pansardi and Tortola, 2019), invoked to legitimise exceptional measures beyond the terms of his office. Perhaps the most striking feature of his ‘whatever it takes’ speech was the reliance on personal assurance over technical detail – ‘believe me, it will be enough’. ¹⁹ Forms of authority based on personal capacity – technocratic, charismatic, or some combination – have also been claimed by figures at the Commission (Pansardi and Battegazzorre, 2018; Vauchez, 2016; White, 2021). Jean-Claude Juncker’s concept of a distinctively ‘political Commission’, and of himself as a distinctively ‘political president’, ²⁰ entailed his self-presentation as a powerful leader able to remake the meaning of his role. Rather than a case of the individual taking his authority from the office, here was an individual asserting his personal authority to redefine and reshape what the office entailed.

Recent EU decision-making thus displays four features relevant to our concerns: the use of informal forums, the concentration of power within institutions, the close cooperation of leaders across institutional boundaries, and a reliance on personalised authority. ²¹ While each has only been sketched, the basic contours seem clear. One sees methods that bypass formal procedures or take advantage of their absence, and that rely on, and in some measure celebrate, the unscripted capacity of individuals and their networks. Such patterns of informalisation and personalisation challenge an account centred on institutional agents. Certainly, individuals continue to occupy official roles and draw on the material and symbolic resources associated with these – de-institutionalisation describes a tendency rather than an end-state – but the exercise of power becomes more elastic. In the case of national leaders, this entails taking advantage of the latitude traditionally afforded to executive representatives in international affairs, expressed with recourse to the informal forums increasingly common in global politics (Roger, 2020). In the case of supranational leaders, it entails policy-making irreducible to the competences formally conferred. As argued by the Bundesverfassungsgericht in May 2020, ²² one sees actions that exceed what the terms of office foresee – at least until the office is later redefined, or a new one is created for the purposes of accommodation. The appearance of ‘de novo’ institutions (Bickerton et al., 2015) in the EU sphere can be regarded as the retrospective institutionalisation of practices that, at the moment of their initiation and greatest impact, were conducted outside existing structures.

One reason these patterns of rule tend to be insufficiently discussed is that they are inherently hard to record. They depart from the expected template. The authorities in question tend to publicise the hoped-for outcome of these decision-making methods (crisis resolution etc.) rather than the methods themselves. Informal, personalised practices are therefore difficult to discern (though ethnography can bring them to light; see Mérand, 2021). But a second reason they tend to be overlooked is conceptual. Because they are at odds with how we have come to expect modern political authority to look, they tend to be discounted or treated as transitional. Those who acknowledge these tendencies tend to naturalise them as logical developments – as examples of adaptation by governing authorities to the pressures posed by events (for critical discussion: Kreuder-Sonnen, 2016). They are seen as episodes of experimentalism and collaborative problem-solving, helping sustain the EU before more regular forms are restored. The temptation, in other words, is to read them as signalling the incompleteness of institutionalisation rather than a departure from it, to be addressed once conditions permit.

To see these as merely transitional phenomena would be optimistic though. While the de-institutionalisation of power is most pronounced in periods of heightened exceptionalism, and some degree of regularisation can be expected thereafter, these practices leave traces that outlive any one episode. Networks of interpersonal trust built in one governing context are liable to be redeployed in others (see e.g. Beach and Smeets, 2020b). Transgressing institutional norms, or exceeding their reach, is likely to be undertaken more freely once a precedent is established. One of the findings of existing research on informal arrangements is that they tend to be durable: the actors involved aim to preserve them (Vabulas and Snidal, 2021: 9). Even if institutional formats ultimately emerge from these interactions, what is crucial is that – at least for a period – these are beyond the reach of existing frameworks. ²³ Especially striking in the EU context is how the informality of bodies like the Eurogroup overlaps and interacts with informality in the established sites of supranational authority (cf. Roger, 2020: 213). One may speculate that the two rub off on each other, as actors familiar with the constraints of formal structures acquire a taste for informality that leads them to prefer such arrangements (the Eurogroup, and before it the European Council), while the use of these encourages informal methods in supposedly more formal settings.

Likewise, personalisation tendencies can be expected to endure. Public reputations built in a period of crisis management become a resource that can later be reactivated. Like the prowess once displayed on the battlefield, modern leaders can earn and enhance their personal authority by convincing others of their aptitude in crisis resolution. They can develop a name for themselves as the possessors of fire-fighting expertise, transferable from one institutional setting to another and irreducible to the authority of a particular office (consider Draghi’s re-emergence in February 2021 as ‘Super Mario’ on the domestic Italian scene). For trusted individuals with a record built in adversity, there will be opportunities long after the crisis.

Implications

Although widely viewed as an extension of the modern project of institutionalisation, transnational integration as it has been pursued in recent years leads elsewhere. It is more accurately seen as enabling new opportunities for rule that operates outside formal procedures and that relies on, and even celebrates, the unscripted capacity of individuals and inter-individual networks. In this section, I offer an evaluation of these patterns and what they imply for transnational authority.

Such an assessment seems especially due given that, while our focus has been on the EU sphere, the patterns described are not unique to it. The IMF is another ostensibly rule-bound setting in which informalisation and personalisation can be studied (Chwieroth, 2013). Its involvement in the Troika led to the same dynamics of power-concentration, cross-institutional collaboration, and the elevation of personal authority. An internal IMF report of the time criticised a tendency for decisions to be taken informally by a small group, to the exclusion of other board members. ²⁴ Efforts to coordinate with the Commission and ECB in a fast-moving situation encouraged shortcutting the official procedures of the institution (Henning, 2017: 34). Leading officials – Strauss-Kahn, Blanchard, Lagarde – emphasised the importance of exercising personal judgement and discretion in their policy-making (Clift, 2018: 8; 107–109). While these practices and their evolution demand closer study, there is a clear prima facie case for charting the de-institutionalisation of power beyond the EU. ²⁵

In assessing the significance of the patterns in question, our focus is on the normative aspects. Evidently these patterns have implications for empirical study – fields, networks and power elites are among the concepts one may adopt when the hold of institutions is in question – but the first goal is to clarify what is at stake. At odds as they may be with wider norms of modern authority, are the patterns described necessarily unwelcome? From some perspectives – well represented in the scholarship on IOs that picks up on these trends – a politics centred on informal and personalised decision-making has positive things to offer. It is a good thing, either intrinsically or as a corrective to existing failings. As I will argue, such readings are unsustainable – the de-institutionalisation of power weakens the accountability and contestability of authority. But before reaching this conclusion, let us consider some of the arguments made.

In one line of reasoning, such forms of governing are to be endorsed because they promise greater effectiveness of rule. A long-standing argument for granting significant power to key figures of the executive, or for allowing them to claim it in crisis moments, is that individuals are better able than corporate bodies to adapt to fast-changing circumstances. Informal rule is more agile than that based on rules, procedures and circumscribed offices (classically: Machiavelli, 2008 (1531), I.34-38). Where effectiveness is the main yardstick of evaluation, constraints on discretion may be viewed as obstacles. The use of informal gatherings and trust-based networks might be welcomed, cautiously, if it facilitates the capacity to make decisions and act – output legitimacy thereby substituting for the legitimacy of procedures (Scharpf, 1999).

Such a position is to be found among scholars of recent EU history who engage with the informal and improvised character of decision-making. Some applaud what they see as the emergence of stronger leadership and a willingness to exercise discretion (van Middelaar, 2019). These readings tend to accept the claims of those involved to be acting without alternative under the pressure of necessity, seeing their actions therefore as more or less successful forms of functional adaptation. This type of account is also well established in the IR literature on informal governance, the advantages of which are said to lie in its functionality and capacity to manage power asymmetries. ‘States’, it is said, ‘opt for less formality by using informal intergovernmental organisations (IIGOs) when the advantages of lower sovereignty and negotiation costs, flexibility and speed outweigh the need for enforcement, commitment, consensus, and the bureaucratic centralization’ (Vabulas and Snidal, 2013: 219; cf. Westerwinter et al., 2021: 11). Informality and personal networks are said to be attractive ‘low-cost’ alternatives to more codified forms of rule (Abbott and Faude, 2020; Alter and Raustiala, 2018). To be sure, these are explanatory rather than justificatory accounts, but the emphasis is on the rationality of the arrangements described.

The informalisation of power presents some evident problems however. It is not just that such arrangements are not always particularly functional (on this see Roger, 2020: 205–206). Rather, the more rule is enacted in informal forums or at the margins of formal institutions, the harder to scrutinise and contest it becomes. Established chains of delegation become unreliable guides to where power lies and how decisions are taken. A mismatch emerges between the order ‘on paper’ and how it is enacted in practice. In his study of authoritarian rule, Ernst Fraenkel detailed the difficulties arising when formal rules co-exist with informal methods. The former present an image of procedural order belied by recourse to the latter. A ‘normative state’, shaped around laws and procedures and the basis of the order’s legitimacy, comes to be undermined by a ‘prerogative state’ whose agents are willing to override them. His term to describe this was the ‘dual state’ (Fraenkel, 2017 (1941)), and it was objectionable both due to its susceptibility to arbitrary rule and because the outward appearance of order concealed this.

To endorse the informalisation of power as an effective way to govern can also be to depoliticise the choices that get made. If it is believed that decisions are taken under the determining pressure of ‘events’, perhaps it matters little in principle who takes them and how. ²⁶ All reasonable agents could be expected to pursue largely the same goals, and they would be distinguished from each other only in their competence in delivery. But if one accepts that politics, even in emergency moments, is never simply about problem-solving but entails choices between competing problem-definitions, values and interests – that it is politics, in other words, and not just crisis ‘management’ – then it matters greatly how decisions are arrived at. To act in an emergency is to make decisions about which parts of the status quo are to be protected or discarded. As work on transnational exceptionalism has highlighted, there are always competing goals to be evaluated even in crisis decision-making, ones hard to revisit after the fact (Kreuder-Sonnen, 2016, 2019: esp. 43ff; see also Heupel et al., 2021). The de-institutionalisation of rule obscures the basis on which these decisions are taken. ²⁷ Moreover, as episodes of emergency politics recur, ever more spheres of decision-making acquire this character.

One should resist the thought that some situations demand informal methods – that they can be governed in no other way. Defenders of exceptionalism may see a necessary trade-off in hard times between procedures and effective rule. Emergencies demand speed and discretion, they will say, and the costs to transparency must be accepted. But as Stephen Holmes has argued, quite the reverse may be true (Holmes, 2009). It is in extreme circumstances that rules and procedures may be most useful. Formal channels offer scripts of action and coordination when thinking-time is short. They keep objectives in view that might otherwise be lost in the scramble. They are enabling as much as constraining. To the extent that there are cases where this does not hold, because the rules are dysfunctional or lacking, this should be seen as a design failure rather than an inevitable fact. Informality is ultimately the expression of a choice.

In their analysis of various forms of informal governance, Abbott and Faude (2020: 25–26) acknowledge in passing that ‘these avenues raise thorny normative issues: they can reduce transparency and accountability, favour powerful actors, and produce relatively weak collective action. In positive terms, however, they may well enhance or protect cooperation’. Treating the benefits of cooperation as potentially balancing the normative considerations seems misleading however. One of the reasons transparency and accountability matter is to allow a wider public to shape which forms of cooperation are considered worth having. Without the circumscription of rule with procedures, such matters are left to the discretion of the few, who in turn are more vulnerable to the pressure of private interests (recall the early-modern arguments for impersonality). Even where they do not exploit the situation for self-serving ends, their criteria of the public good remain hard to identify and contest.

What is also at stake is perceived legitimacy (Beetham, 1991). The basic principle of separating rulers from the offices of rule was crucial to the modern state’s capacity to achieve a measure of public acceptance. It became possible to accept power while opposing those who wielded it: one could oppose the incumbents without opposing the system. For rulers themselves this presented a threat. If the King could be challenged in the name of the Crown (e.g. for ruling irresponsibly and depleting its reserves), he could never be certain of his hold on power (van Apeldoorn, 2020: 51). But for the emerging political order it promised the possibility of radical reform within an association strong enough to sustain it. Conversely, when power is exercised informally, the distinction between person and office becomes harder to sustain, and it becomes harder to abstract from the failings of individuals. When populations have little reason to be confident in the structuring capacity of institutions, to oppose those in charge is necessarily to oppose the system as a whole. Just as in the late medieval period, such authority is brittle.

In a second line reasoning, one has to distinguish further between the dynamics in play. While a tendency towards informalisation is rightly criticised, it may be said a tendency towards personalisation has positive features and is perhaps ultimately something to endorse. Such a view relaxes the connection between the two identified in the state-building accounts we began with. If power is prone to be concentrated in the hands of the few, it may be said that the appropriate goal is to allocate this formally, realigning power with institutional boundaries. The elevation of key figures of the transnational sphere – including, in the EU, the presidents of the Commission, Council and ECB – makes sharper the delineation of power: what it needs is better codification.

Especially in a context where ideas of nationhood and common identity cannot easily be invoked, a measure of personalisation may seem welcome. Arguably it has positive implications for that third dimension of institutionalisation highlighted in accounts of state-formation – the integration of power into a larger social whole. When governing authority is centred on key individuals, citizens have something to point at, to endorse or to criticise, enabling a better connection between rulers and ruled. Whereas bureaucratic institutions tend to anonymise power, personalised arrangements foster the intelligibility of decision-making and create objects of scrutiny and allegiance. Such arguments have a long pedigree in presidential systems, being advanced in the early US (Federalist Papers 70: 348-9). They might seem all the more pertinent beyond the state, given the complexity of governing arrangements and the difficulty of attributing responsibility. Arguments to this effect have been made in relation to the EU Commission President and Council President (Crum, 2009; Schmitt et al., 2015) and could in principle extend to the ECB.

While stronger codification would surely be welcome, consolidating this personalised model seems a dubious goal. Worth emphasising is the importance not just of clarity on where power lies, but of its exposure to contestation. More than just lines of responsibility, high offices should be expected to offer targets of influence, such that the goals informing policy can be disputed and shaped. The more power retains a personalised basis, the more difficult this is likely to be. Electoral mechanisms for selecting supranational leaders are undoubtedly part of the answer, as a way to further separate persons and offices, but are unlikely to suffice on their own. As the experience of contemporary presidential systems nationally suggests, personalised power tends to be judged mainly on questions of character, charisma and competence rather than the ends to which it is put (McAllister, 2007). A democratic rationale for depersonalisation is that it allows abstraction from the qualities of individual rulers to the ideas promoted by wider collectives. When ideologies were developed by movements and parties of the left in the late 19th century, not only did this allow the integration of power into a social whole, but in such a way as to encourage principled contestation. Embedding the offices of state in party politics meant embedding them in intermediary organisations that were themselves instances of depersonalised agency, and that contributed as such to making power more democratic.

Fostering something similar transnationally has been notoriously difficult, but there are steps that might support it. Parliamentarisation stands as the most logical route in the EU. Further linking the composition of the Commission to the outcome of European-Parliamentary elections, for example, by requiring that Commissioners be MEPs, would be one plausible initiative (Lacey, 2017: 221–223). Rather than individuals selected on personal criteria, incumbents would then be the representatives of Europe-wide parties. The same argument is applicable to central banking, where it would support moves to increase European-Parliamentary involvement in how ECB officials interpret their mandate (de Boer and van ’t Klooster, 2020: 26). Depersonalising EU power in such ways would aid contestation of the ends it serves.

Ultimately, however, the far-reaching question for transnational politics is how far the institutionalisation of power can be detached from the making of sovereignty. The formalisation, depersonalisation and integration of power in the modern state, though always incomplete, was facilitated by hierarchies of authority that could be invoked when agents departed from script, and that allowed expectations to form about where power should lie. The absence of such hierarchies transnationally both enables and encourages irregular forms of rule, especially in times of stress. A functional constitution that embeds certain policy commitments is no substitute for a legal constitution that establishes constraints on action and identifies who has the authority to impose them. This may suggest the desirability of restoring decision-making to the sovereign nation-state, as the most plausible context of efforts to renew the institutionalisation of power. But it could also be the rationale for a transformative approach, aimed at the construction of supranational sovereignty. Demanding as such a constitutional project may be, the current of model of power looks hard to defend.

Conclusion

The making of modern authority was a project of institutionalising power. Transnational political orders have often been seen as an extension of that project, even an effort to preserve it against threats. The EU in particular has been cast as thoroughly modern in its attachment to rules and procedures, sometimes to the point of excess. As this article has sought to suggest, the EU of recent years is better approached as an arrangement that challenges the formalisation and de-personalisation of power rather than one that marks their consolidation. Here, as in other respects (Zielonka, 2006), there is a ‘neo-medieval’ quality to the contemporary EU.

One of the insights from the tradition of studying emergency politics is that one learns from extreme situations about possibilities that are latent and long-standing. The features we have examined – the use of informal forums, the concentration of power in individuals, the collaboration of leaders across institutional boundaries, and a reliance on personalised authority – are most pronounced in crisis conditions. But rather than the shortcomings of a moment, they reveal vulnerabilities that are structural. The extreme situation reveals what is possible and sets precedents that can be reactivated. Contra what tends to be written on informal governance, this should be seen as a disturbing prospect, since the looser the constraints on the exercise of power, the less confident one can be that it is put to good use.

To chart the de-institutionalisation of power transnationally is not to suggest domestic contexts are spared it. Increasingly the distinction between the two spheres is hard to draw. Alongside figures of the supranational world, the EU’s crisis politics has involved national leaders using transnational coordination to evade domestic constraints. All are subject to influences that invite discretionary forms of decision-making, from the challenges of interdependence to the material pressures of economic power, from the weakening of representative ties to failings of institutional capacity. But even if the working methods of transnational authorities are ultimately no worse than elsewhere, hopes that they might offer a positive contrast currently seem misguided.

One may further note that many of the movements challenging supranational authority in recent years have been keen to accentuate these features. Some Eurosceptics highlight the informality of rule as the basis for conspiracy theories. The suggestion that leaders are surrounded by informal networks becomes the basis for denunciations centred on exposing a sham, and sometimes for efforts to discredit the principle of political representation more generally. The de-institutionalisation of power at the supranational level thus presents an inviting target for critics at the national level. Even should EU leaders seek to change course, stepping back from the methods of recent years, they will be confronted by those intent on personalising authority as a way to discredit it.

Developing transnational authority structures in a different direction is no easy thing therefore. First though is to recognise the dynamics in play and why they matter. De-institutionalised power need not be arbitrary, but little prevents it becoming so. Such problems can be anticipated especially in situations governed as emergencies, which offer opportunities and licence for irregular methods. As emergencies come to preoccupy authorities at all levels, one should look to the transnational setting expecting not so much the further institutionalisation of power as steps that disrupt or reverse it.