From constitutional rule to loosely coupled spheres of liquid authority: a reflexive approach

Abstract

This article develops a reason-based social foundation of new forms of authority, which often are liquid and sectorally limited. The recognition of authority hinges, in this view, on reflexive actors who are aware of their own limits of rationality regarding the lack of either information or a perspective that allows for the pursuit of common goods. In such a reflexive concept of authority, authority takers tend to monitor the authorities closely, and the internalization of the subordinate role is not a necessary part of it. Reflexive authority is embedded in the acceptance of a knowledge order that reproduces the authority relationship. In spite of a tendency toward institutionalization, reflexive authority often comes in a liquid state of aggregation, and almost always with a restricted functional scope. Moreover, this new set-up of authority creates social dynamics that add to liquidity. First, the encompassing constitutionalized rule with majoritarian decision making as major source of legitimacy is increasingly undermined by loosely coupled spheres of specialized authorities, which are most often justified on the basis of expertise. We can observe both the rise of international authorities in the absence of coordination between them, and the rise of similar authorities within the nation state that escape control of the democratic core institutions. As a result, authority gets fragmented and different authorities need to adjust to each other. The second implication of the argument is that democratic legitimation narratives become rare, leading to an ongoing legitimatory contestation of authorities. Both these processes make authority even more liquid.

Liquid forms of authority are ‘characterized by a lower degree of consolidation and a significant dynamism in the configuration and reconfiguration of authority structures’. These forms of authority defy the notion of authority as ‘the right to command’, and thus question the mental map of ‘solid authority’ with far-reaching implications for our understanding of political institutions and their legitimacy (see Krisch this issue). I want to take up this characterization of current forms of public authority and carry it on by theorizing their social foundations. For this purpose, I will develop a reflexive conceptualization of authority and legitimacy that emphasizes their epistemic foundations in a globalized society. In such a reason-based account, authority hinges on subordinates’ perception of limitations of their own rationality. At the same time, reflexive authorities are under permanent scrutiny. They do not operate through demands but through requests, and they need to leave choices for the addressees of authorities. Nevertheless, they can induce deference and even get perpetuated via objectivization and institutionalization. Reflexive authority therefore comes often, but not always in a liquid state of aggregation. In addition, it creates social dynamics that even increase liquidity over time. This contribution first proposes that the encompassing constitutionalized rule as archetype of solid authority is increasingly undermined by loosely coupled spheres of specialized authorities. Loosely coupled spheres of authorities may collide and interact with each other, but they are not centrally coordinated by meta-authorities. The rise of loosely coupled spheres of authority cuts across different levels of governance: we see the rise of international authorities in the absence of coordination between them, as well as the rise of authorities within the nation state that escape control of the democratic core institutions, such as parliaments and governments. As a result – and this is the second proposition – the legitimation of reflexive authority becomes contested. Both of these processes add to the liquidity of authority.

The meaning of authority in global governance

The foundational treatment of authority in modern social sciences is still the one by Max Weber in his sociology of domination. He defines authority and domination as the chance

that certain specific commands (or all commands) will be obeyed by a given group of persons. It thus does not include every mode of exercising ‘power’ or ‘influence’ over other persons. Domination (‘authority’) in this sense may be based on the most diverse motives of compliance: all the way from simple habituation to the most purely rational calculation of advantage. Hence every genuine form of domination implies a minimum of voluntary compliance, that is, an interest (based on ulterior motives or genuine acceptance) in obedience (Weber 1978 [1925], 212).

Authority thus describes an asymmetrical relationship between actors that is recognized by both sides. The core of such a relationship is deference of the subordinate party to the authority holder describing a specific form of power. ¹ This holds for different strands of thinking about authority. Horkheimer (1987 [1936], 24) describes authority as ‘approved dependence’, according to Arendt (1957, 117) authority ‘always demands obedience’ and for Simmel (2009 [1908], 130–1) authority ‘presupposes in large measure […] a freedom of those subject to the authority; it is, even where it seems to ‘crush’ them, not based on coercion and pure resignation alone’.

An authority relationship plays out in specific, but recurrent situations and thus contains a certain stability over time. To the extent that authority describes an ongoing relationship, it expresses a form of hierarchy, and therefore it is necessary to legitimate it. Authority thus can be abstractly grasped as a continuing relationship that is reproduced by subordinates deferring to the authority holders and by authority holders justifying it in terms of common interests and goods.

This notion of authority consisting of the components of deference, perpetuation, and legitimation captures the paradox that is built into such relationships: voluntary subordination. Such an understanding is open for different sources of authority. According to Weber, authority can be based on simple or dull habituation [dumpfe Gewöhnung] as well as on instrumental rationality. He maintains that both the disciplining of subordinates through manipulation and internalization, as well as well-reasoned decisions by autonomous actors, can be the source of authority relationships (see also Sending this issue). He brought in the socialization-based perspective not by disavowing, but by critically questioning the sources of authority. He pointed to the reason-based perspective not with the goal to equate authority and legitimacy, but to allow for the possibility that authority relationships may be rational in a functionally differentiated society.

In the next three subsections, I will develop a concept of reflexive authority which – not least due to the specific context of global governance – starts out as reason-based (or actor oriented), but does not exclude the possibility of internalization and habituation. I then continue to show how liquid authority becomes institutionalized and point to the legitimatory predicaments of reflexive and liquid authority in global governance.

Reflexive authority

Reflexive authority is reason-based and differs from traditional authority. Traditional authority points to a social relationship and a form of power that is, according to Arendt (1970, 45), characterized by ‘unquestioning recognition by those who are asked to obey; neither coercion nor persuasion is needed’. In contrast, the recognition of reflexive authority does not equate with blind and mindless submission.

The reflexivity in reflexive authority relationships is twofold. On the one hand, it speaks to an element of ongoing reflection about the quality of the authority holders. While reflexive authority involves dispensing with an exact examination of the specific judgment or decision – because such an examination would involve enormous efforts and recognition on the basis of trust appears reasonable – it, at the same, involves a permanent monitoring and consideration of the standards that make an authority appealing and trustworthy. For instance, most people would believe in the climate models of a Nobel laureate without checking all the parameters and equations – which would take quite a long time. At the same time, the credentials and the reputation of the models are checked permanently.

On the other hand, these advantages do not exclusively derive from a rationalist conception based on pre-defined interests. It is especially in situations without pre-defined interests that reflexive authority plays out. By providing information and new perspectives, authority may even shape actors’ preferences. A listener may re-evaluate his wealth preferences after listening to the climatologist. As Furedi (2013, 52) states, ‘[t]he need for a concept of authority only emerges when communities are forced to contend with uncertainty about questions who to believe, trust, follow or obey’. In the case of reflexive authority, the recognition of external authorities is based on the knowledge about the limitations of one’s own rationality, and thus the dependence on epistemic constructions that identify these limits. This form of self-reflection on the side of authority holders points to the second component of reflexive authority.

Reflexive authority thus has advantages for both the authority holders and authority receivers. In this sense, I follow the ‘service conception’ of authority as developed by Raz (2006). At the same time, it points to a special role of information and epistemic processes.

Reflexive authority seems especially apt to capture the relationship between international institutions and governments of states. ² The conceptualization of international institutions – which is based on a private law paradigm, according to which states produce consensually passive treaties in order to serve their interests – is increasingly questioned. Both International Law and International Relations have developed approaches to capture the new quality of international institutions. ³ In this view, the core decision makers of international institutions consisting of the leadership of the secretariat and the executive representatives of the more powerful states may exercise authority over other states. According to the notion of reflexive authority, states recognize the authority of international institutions when they provide valuable information and when they request from the perspective of impartiality to do something in order to achieve common goods or goals. States, however, do not accept unquestionable commands to do things they do not like to do at all. States are actors that have various established mechanisms for reflecting about obligations in the international realm – both within the government and outside in the form of expertise in the parliament, universities, and think tanks. It therefore seems dubious that states do subordinate themselves to international institutions without further questioning as most conceptions of authority want it. The deference toward the International Criminal Court or the willingness to accept the authority of the International Monetary Fund (IMF) regularly contains an element of reflection. A ruling of the World Trade Organization’s (WTO) Dispute Settlement Body can therefore be considered and taken seriously without complying with it. In the case of labeling of genetically modified organisms (GMOs), the European Union (EU), for instance, accepts to pay fines for their GMO policy, thus affirming the WTO’s authority, without following the prescription to change their policy.

Reflexive authority does not usually work with ‘commands to do x’, but rather with ‘requests to consider y’ (see also Krisch this issue). Reflexive authorities do not usually give reasons that underlie a command. ⁴ No international organization (IO) can ask a state to ‘ignore’ the so-called national interest. IOs do not ‘command’ states. They can ‘request’ to take into account information and reasons as a means to pursue international or global goals. The authority of transnational and international institutions is often based on something that is labeled ‘secondary reasons’ in pragmatist philosophy. Especially in cases where one is uncertain about the primary reasons for making a decision (for instance, ‘Is the food good in this restaurant?’), secondary reasons (such as ‘If they have a certificate, it must be good’) become decisive. ⁵ This deference to the certification scheme, however, excludes neither an own judgment afterwards (‘The food was, in spite of the recommendation, mediocre’), nor the possibility to shift to another certification scheme after negative experiences have multiplied. Different reflexive authorities may compete with each other, thus allowing for forum shopping.

Moreover, according to the logic of secondary reasons, it is not the perceived ‘duty to follow’ – emphasized by many scholars of traditional authority – toward the authority that is decisive. One does not usually follow a restaurant guide because one feels a duty to the restaurant guide. Secondary reasons do not create a duty to follow. In fact, states usually follow IO prescriptions because they feel that the authority is doing a good service. Even if states feel an obligation to follow, this obligation is hardly derived from a ‘right to rule’ – a formula often emphasized by theorists of traditional authority. States usually do not follow IOs, because they recognize their right to rule. If states feel an obligation to follow, they feel a duty toward their fellow states to contribute to the public good or toward future governments not to destroy the reputation of a country. It is thus not IOs’ ‘right to rule’ that is at stake, but the willingness of states to take into account and to consider requests and interpretations from IOs.

This concept of reflexive authority differs from other important contributions to authority in International Relations that build on such notions like commands, right to rule, and obligation to follow, and are often associated with the notion of solid authority. First, socialization-based accounts of international authority that prevail in post-structuralist and critical accounts aim at debunking the knowledge order (episteme) that underlie authority relationships and lead to unquestioned and internalized subordination (Ashley 1988; Anghie 2007; Adler and Pouliot 2011). Yet, deference to transnational and international institutions is often not internalized, but is permanently under observation. This does not exclude internalization per se, which is however not a necessary part of the story in global governance. Many of the new authorities are young, and they target corporate actors, which often have significant analytical capacities. In this situation, dull obedience based on an internalization of the authority relationship can be expected to be rare. The concept of reflexive authority does also not depend upon the internalization of the norms postulated by the authority, as Hurd (2008) suggests. Moreover, even when deference to the authority gets internalized over time, this should not be equated with an internalization of the norms postulated by the authority. The payment of taxes to states, for instance, is rather ‘grudgingly’ accepted, without being transformed into an internalized desire to pay, even if the state, as authority, is recognized or even internalized (see also Haas 1983).

Second, and in contrast to other reason-based accounts of international authority (see, e.g. Hawkins et al. 2006), reflexive authority emphasizes the role of episteme – or knowledge orders – as constitutive background of authority relationships and deference (see also Sending this issue; Venzke 2013). Lake (2009, 2010), who significantly contributed to a reason-based understanding of international authority, has analyzed the bilateral relationships between the United States and its allies. According to his analysis, the smaller partners recognize US leadership in return for protection and the provision of collective goods. Therefore, authority expresses a right to command, which needs to be acquired and thus results from a social contract based on interests existing independently of the epistemic foundation that constitutes authority. In this conceptualization, however, deference evaporates. If the recognition of an authority is completely fluid and can be explained completely with a reference to pre-defined interests, then it is the outcome of negotiations and not the authority relationship as such that is accepted. Such a perspective misses the special epistemic role of the authority holder. All the classical figures of authority – the theological father, Hegel’s master, or the judge – have a cognitive component that can roughly be described as ‘knowing better’. Authority relationships are thus embedded in and reproduce belief systems and interests, and must be conceived in a broader sense than just a right to command.

The epistemological grounding of authority relationships points to the relevance of the social process in which superior knowledge is established (see also Sending this issue). The notion of reflexive authority does not depend on the assumption that those who are considered as authorities are indeed closer to the truth. It rather points to social processes of knowledge production and the role of science therein. The whole notion of ‘scientification’ of social processes, on the one hand, indicates the importance of science in this process; on the other, those parts of the society who question these authorities often question science as such. Those who reject genetically modified food, for instance, tend not to challenge the results of studies which point to the non-hazardousness of these products. They rather doubt whether science can assign ethical considerations an appropriate place, and whether science is able to take the ‘unknown unknowns’ into account. Most strategies that aim at de-legitimating authority therefore involve epistemic challenges to the dominant knowledge order. For instance, many transnational movements are challenging international institutions question neo-liberal thinking at the same time.

To sum up, reflexive authority is not only in form, but also in substance different from solid authority. In such relationships, the authority holder sends requests instead of demands, and authority takers tend to monitor the authority closely. The internalization of the subordinate role is not a necessary part of it. Whereas reflexive authority can, under special circumstances, transform into dull habituation, especially on the side of those actors that lack informational and monitoring capacities, this is not a constitutive part of it. Reflexive authority is based on the acceptance of a knowledge order that reproduces the authority relationship. At the same time, the underlying knowledge order is dynamic. Moreover, reflexive authority is always under scrutiny by the addressees of authorities, whether it delivers and is in line with the expectations. Both of these aspects lead to permanent contestation and adjustment of the reflexive authorities. This makes reflexive authority ceteris paribus more fluid than traditional authority (see Krisch this issue).

Perpetuation of reflexive authority

Without commands and a right to rule, can international institutions as reflexive authorities be consequential after all? The answer is yes, and the reason is processes of perpetuation. If we describe authority as the chance to achieve deference in specific recurrent or in all situations within a social relation, then recurrence is built into that definition. As authority contains more than brute force and is not based on case-specific persuasion, it describes an ongoing relationship between two social actors that contains a specific epistemic element of recognizing the authority as ‘knowing better’. While such a relationship may, in some cases, emerge spontaneously – for instance, in crises situations – the future of this relationship will not forget the spontaneous moment. ⁶ Any authority relationship has therefore a tendency not to be a one-shot instance, but to be durable for at least some time. Talking about authority thus points to the inscribed tendency of transforming an asymmetrical relationship into hierarchy. Such a transformation can come in form of either ‘objectivation’ or of ‘institutionalization’.

An authority relationship is ‘objectivized’ if the knowledge order that underlies the relationship becomes a dominant worldview or ideology that reaches beyond the immediately involved actors. It is then a shared understanding reaching beyond the direct authority relationship that the authority holder represents the ‘objective’, ‘non-particularistic’ perspective. This role assignment can be more or less formalized. Scientists with public appeal are cases of this role assignment. To the extent that this role assignment gets broadly accepted, objectivation points also to social pressure and coercion. If an actor rejects insights that are seen as ‘objective’ and unquestionable, the danger of social exclusion and being seen as weird arises. In this case, the voluntariness of subordination is reduced. Similarly, if a political regime is worldwide broadly perceived as democratic and human rights observing, it can more easily depict a violent opposition as terrorism that needs to be combatted than an authoritarian dictatorship. It is then the external recognition of the (democratic) authority holders that allows them to be coercive to their opponents.

Moreover, in some cases, reflexive authority can transform into indirect coercion when third parties are dependent on the targets of authority. For instance, many lenders take the judgment of rating agencies as guidance, without checking the judgments in each individual case and without feeling an inscribed duty to follow these requests. Rating agencies thus exercise reflexive authority over lenders. In this way, rating agencies also exert pressure on states to follow the policies that give them a good rating; otherwise borrowing becomes impossible or at least more expensive. It is in this indirect way in which reflexive authority can become coercive. Objectivation, therefore, is one mechanism through which even secondary reasons and liquid authority can turn out to be very consequential. Liquid authority does not mean inconsequential authority (see Krisch this issue). As the saying goes, constant dropping wears away a stone.

In global governance, the ‘institutionalization’ of reflexive authority is the most important mechanism through which authority relationships become consequential. An authority relationship is institutionalized when the decisions and interpretations are taken by an institutionalized body. Especially reason-based accounts point to institutionalization, to law and assignment of institutional competences as mechanisms of perpetuation. Governments represent an ideal type for this form of authority perpetuation.

What types of reflexive public authority can be distinguished in global governance? Are there good measures to show that such liquid authority has gained in importance over time? In order to show the relevance of reflexive authority in the international and transnational realm, two basic types of public authority can be distinguished: the authority to make decisions, and the authority to provide interpretations. The authority to make decisions can be labeled ‘political authority’, the authority to make interpretations can be called ‘epistemic authority’. ⁷

In the case of ‘political authority’, stipulations, rules, and norms are viewed as ‘binding’ for a certain collective. Political authority rests on the acknowledgment that there needs to be an institution that is authorized to make collectively binding decisions in order to promote the common good and to prevent chaos. Political authorities thus have the right to make decisions that violate the particular interests of members in the short term. Political authority rests on a cognitive framework that permits a common interpretation of the common good. A political authority may have the right to enforce rules, but its influence does not rest on force alone.

Newly institutionalized political authorities have emerged in the last decade, especially on the transnational and international level. The principle that a given territory is governed exclusively by the national government does not hold any longer. International institutions circumvent the consent principle by taking decisions through forms of majority voting or through the informal dominance of hegemonic powers. Majority decisions and the exercise of dominance by strong countries enhance the capacity of international institutions to act by avoiding vetoes by single states and overcoming blockades. Today, roughly two thirds of all IOs – in which at least one major power participates – have the possibility of taking decisions by a majority vote (Breitmeier et al. 2006; Blake and Payton 2008). Moreover, states are increasingly delegating power directly to IOs. The authority index developed by Liesbet Hooghe and Gary Marks show a marked increase in the last decades. Their standardized delegation index for 78 IOs rose from 23.5 in 1975 to 33.2 in 2010 (Lenz et al. 2015).

To the extent that the new international institutions exercise political authority, they set rules that reduce the room of maneuver of national states and govern formerly domestic affairs either directly or indirectly. Especially powerful states aim to use such authorities to exercise influence outside of their territory; at the same time, they often try to limit the authorities’ influence on their own affairs. On the one hand, the most important IOs with majority voting also come with weighted votes or even vetoes – the IMF, the World Bank, and the United Nations Security Council are the best known examples. On the other hand, powerful states keep in the possibility of evaluation and of shifting loyalty, and they even create circumstance in order to extend these options. The notions of ‘counter-multilateralism’ (Morse and Keohane 2014) and ‘deliberate fragmentation’ (Benvenisti and Downs 2009) point to this strategy: powerful states recognize ‘multiple’ authorities with overlapping competences in order to put pressure on authorities to act not against their interests, and in order to keep the option of shifting loyalty. Less powerful states often do not have this option, and therefore are less able to maintain choice in such a relation. Non-state actors and individuals have, of course, even less possibilities to circumscribe political authorities. In this way, reflexive authority goes hand in hand with institutionalized inequality (Viola et al. 2015).

In general, it seems odd to conceive the practice of international and transnational political authorities in line with the mental map of solid authority as ‘commands’ issued by one actor that are expected to be obeyed by a second, and regularly backed up by force (Lake 2009, 18). Political decisions by reflexive authorities are, in principle, permanently monitored by those who are subjects of political authority, and they do not necessarily provide exclusionary reasons for actions. They sometimes come with the right to enforce decisions – as in the case of the Security Council – but more often come without it. Even the EU had no sanctioning capacity for a long time. While the subjects of reflexive political authority accept an inducement to defer, they keep alternatives and the exit option open, and thus keep it liquid. This capacity, however, varies across different targets of authority.

The other basic type of authority can be labeled ‘epistemic authority’. Epistemic authorities provide interpretations that structure the behavior of others. Epistemic authority is based on expert knowledge and moral integrity. The views and positions of an authority are adopted because they appear to be both knowledgeable and nonpartisan at the same time. Epistemic authority is based on the assumption that knowledge and expertise are unequally distributed, but that there is a common epistemological framework that makes it possible to ascertain knowledge inequality. An epistemic authority needs not, in all cases, convince people factually and in detail. It is, therefore, not the quality of the specific argumentation, but rather the general reputation of an institution or a person that is decisive. What is involved is governance by reputation (Schuppert 2010, 94) and by indicators (Davis et al. 2012).

The significance of institutions with purely epistemic authority has unquestionably increased on the transnational level. This particularly includes influential and ‘credible’ non-governmental organizations (NGOs) such as Greenpeace in the environmental area, or Amnesty International in the area of human rights, which are active in setting as well as monitoring norms. The growth of transnationally active NGOs has been enormous – today there are roughly eight times as many of them as there were in 1975 (United Nations Department of Economic and Social Affairs 2009). The access to IOs has as well increased dramatically, by a factor of three since 1975 (see Tallberg et al. 2013, 68).

A special version of this type of public authority arises when an epistemic authority is assigned to that status by political institutions. Then we may speak of ‘politically assigned epistemic authorities’ (PAEAs). It is especially this type of authority that has gained enormously in importance and has changed the constellation in the exercise of public power. At the transnational as well as the international level, such politically assigned epistemic authorities have moved into the focus of attention particularly during the past two decades. The rating agencies, the Intergovernmental Board for Accounting Standards (Büthe and Mattli 2011), the International Panel for Climate Change (Beck 2012) and the international evaluation institutions in the area of environmental policy in general (Mitchell et al. 2006), the Organisation for Economic Co-operation and Development (OECD) in the area of educational policy (Martens and Jakobi 2010), and international courts (cf. Alter 2011; Bogdandy and Venzke 2014) are all political institutions that have become more influential over the past two decades. In general, the authority of IOs has risen steeply regarding the policy functions of agenda setting, knowledge generation, and norm interpretation (Zürn et al. 2015), pointing to the growing relevance of PAEAs. While a comprehensive systematic survey of this type of authority – including both international and transnational institutions – is still lacking, it appears that, particularly in recent decades, it has increased enormously in significance.

While the rise of PAEAs is most visible in global governance, we can see related developments on the national level. Ethics commissions, rankings, and benchmarking have gained importance in the domestic context as well. This rise of PAEAs goes hand in hand with non-majoritarian institutions, which act exactly in the field between decisions and interpretations. One example is Constitutional Courts. Hirschl (2004, 1) points out that ‘[o]ver the past few years the world has witnessed an astonishingly rapid transition to what may be called juristocracy. Around the globe, in more than eighty countries and in several transnational entities constitutional reform has transferred an unprecedented amount of power from representative institutions to judiciaries’. A very similar development can be discerned with respect to central banks. They have been gradually introduced in many countries, and their independence has been strengthened. Between 1990 and 2008, no fewer than 84 countries passed legislation to enhance the formal autonomy of central banks (Rapaport et al. 2009; see also Cukierman 2008). At the same time, the importance of monetary policy tools in the general economic control toolbox has principally increased with the spread of monetarism. Central banks thus became more autonomous and more important. Ultimately, however, they are only the manifestation of a broader development: the increase in so-called ‘independent agencies’ (Shapiro 1997). According to a quantitative study, ‘autonomous regulatory agencies’ play a role in 73% of all policy areas in the countries under investigation (Jordana et al. 2011). The study shows steady quantitative growth between 1966 and 2007, becoming almost exponential in the 1990s, and declining slightly only after the turn of the century.

Reflexive authorities can be influential in spite of being liquid instead of being solid. No one doubts that the WTO, the OECD, rating agencies, international courts, constitutional courts, and central banks – to name but a few – are consequential. Their commonality is that they exercise reflexive authority, which is either objectivized or institutionalized, but cannot be backed with force by these authorities. As opposed to traditional authority, reflexive authority is always dynamic and contested. It contains, however, a tendency toward institutionalization and therefore can gradually move away from the pole of liquidity – although it is unlikely to be internalized moving fully to the solidity pole.

Legitimation of reflexive authorities

When reflexive authority is perpetuated, the voluntariness of subordination, as well as the liquidity of authority, is reduced. Perpetuation thus shifts the balance of voluntary subordination away from the voluntary side of the equation. It is for this reason that the need for legitimation is a constitutive component of the conceptualization of reflexive authority. To the extent that authority gets perpetuated via objectivization and institutionalization, there is a need for legitimating the relationship. Therefore, the rise of authority ‘is normally accompanied by the permanent attempt to arouse and nurse the beliefs in legitimacy’ (Weber 2013b, 450; trans. Michael Zürn). This attempt may or may not be successful – that is, it may or may not lead to legitimate authority (see also Black this issue). ⁸

An appeal to the public interest, or to the common good of the collective that is being regulated, represents the foundation of any legitimation and legitimate authority (see Macdonald and Macdonald this issue). Even traditional and religious justifications are embedded in narratives that emphasize the common good, and may work successfully in specific circumstances even today. Yet, faith in legitimacy in any case dissipates immediately if the rulers are perceived as being exclusively self-serving or selfish.

The mere reference to the common good, however, is insufficient as a legitimation strategy in modernity. The public interest claim needs to be supplemented with a reference to sources and procedures by means of which the public authority could indeed give the appearance of serving the common good. On the basis of a set of different sources through which the exercise of public authority can be bound to the common good, we may distinguish four dominant narratives of legitimation for public authority. ⁹

Authority was, for a long time, a concept emphasized by conservatives (see Eschenburg 1976; Furedi 2013), one of the reasons of which is that tradition served as the most important legitimation of authority – especially pre-modern authority. The basic justification is that long-standing practices could not be long-standing if they were false. The ‘traditionalist legitimation narrative’ can come in different embodiments as religion or the traditional right of certain families to rule. It is closely associated to habituation and the dull obedience, based on the acceptance of a certain worldview with disciplinary effects as emphasized by Foucault (1980) and others. The traditionalist legitimation narrative is often used to justify solid authorities, but is hardly used by transnational and international institutions.

The ‘technocratic legitimation narrative’ is mainly based on results or output (Scharpf 1999). The process of decision making is only secondary in this account. Good policy results are based on non-prejudiced ‘expertise and knowledge of the facts’. This expertise is derived from the concept of science as an independent search for knowledge, with no regard for particular interests, based on a systematic methodology. Connected with this is the hope for successful goal-oriented policies that especially promote the welfare of the community and generally support solutions to problems. This technocratic legitimation is often accompanied by the idea of ‘accountability’. However, accountability should not be confused with democracy in this context. Accountability does not require the participation of those affected by regulations in specific decisions; it merely makes the decision makers responsible ex post for the decisions they have made – legally, electorally, or in some other way. ¹⁰ At issue here is control, not participation.

The ‘liberal or legal narrative of legitimation’ is based on the ‘protection of basic rights and the promotion of legal equality’. Especially in liberal thinking, individual rights and legal equality constrain the exercise of authority (Tamanaha 2004). At the same time, the reference to decisions that protect and promote individual rights can have a legitimacy-generating effect. National and international courts play a key role in this legitimation narrative, which moves it closer to the technocratic narrative.

Finally, the ‘participatory legitimation narrative’ is based on the ‘equal opportunity of participation’ of all those affected by regulation, or at least of all those to whom the regulations are addressed (Dahl 1989). Participation may be carried out directly or via representation. Representation can be organized functionally or territorially. Accountability is absolutely central to the concept of political representation. In institutional terms, participation and representation mechanisms may be represented in multiple and strongly differing ways. In addition, the theory of deliberative democracy refers to a further procedural source of democratic legitimacy: public discourse and contestation. It is based on the normative conviction that the aggregation of interests must, in the course of participation or representation, be accompanied by an open dispute regarding the common good (Habermas 1994).

Implications

The concept of reflexive authority provides a reason-based account of liquid authority (see Krisch this issue). It rests on the knowledge about the limitations of one’s own rationality, and thus the dependence on epistemic constructions that identify these limits. At the same time, it is always under close scrutiny by the receivers of authority. Reflexive authority therefore has a built-in tendency toward liquidity. This may be countered, to some extent, by the tendency toward institutionalization. Due to this institutionalization, the voluntariness of subordination however gets qualified, which, in turn, makes legitimation necessary (see also Black this issue).

In the remainder, I want to show that this re-conceptualization is consequential by pointing to implications that are different from the expectations that can be derived from the concept of solid authority. Both of these implications point to processes via which authority becomes even more liquid.

From constitutional rule to loosely coupled spheres of authority

Rule/domination and authority as forms of power are both based on recognition and deference. Yet, the two terms can be – contrary to Weber ¹¹ – distinguished according to their scope. Authority, then, is a functionally differentiated form of a right to do something; it is specialized in the sense that it is limited to certain tasks and functions. ¹² The authority of the driving instructor is an example. In contrast, rule refers to a social relationship as a whole – as in the master–slave relationship. We therefore speak of political rule if a political system that can potentially take action regarding all issues has developed, and thus involves a meta-authority to decide on issues about which different spheres of authority are in conflict. System-wide rule typically also involves the backing-up through institutionalized force. Whereas rule, in this use of the term, is encompassing or systemic, and often involves the recognition that force can be used in order to enforce rules, authority is more limited to specific issues and functions, and it does not necessarily involve the acceptance of rule enforcement. While courts hold authority, the modern state with its monopoly of the legitimate use of force and a general responsibility for the development of a given society exercises rule. To the extent that the state is depersonalized and that the rulers are bound by rules as well, it may be considered as a constitutionalized rule. Constitutionalism thus points to a system of rule that allows a community to govern itself (foundational aspect), but imposes limits on the freedom of choice for the rulers (limitational aspect).

This notion of constitutionalism is most often associated with a democratic political system in which parliaments – in connection with the government – play the decisive role. As both parliaments and governments in democratic political systems consist of elected representatives – who, in turn, are organized in political parties – they are, at the same time, the most typical cases of ‘majoritarian institutions’; that is, those institutions that are legitimated on the basis of a participatory narrative. Non-majoritarian institutions, such as courts and central banks, do play an important role in democratic political systems as well. Non-majoritarian institutions can be defined as governance entities ‘that (a) possess and exercise some grant of specialized public authority, separate from that of other institutions, but (b) are neither directly elected by the people, nor directly managed by elected officials’ (Thatcher and Sweet 2002, 2). They are most often based on a technocratic or legal legitimation narrative. In democratic theory, their major task is to control and limit the public powers, and to implement the norms set by the legislative. ¹³ In this conception of democratic rule, parliaments are the norm setters and, together with the executive, are considered the most important majoritarian institutions that stand for the foundational component of constitutionalism. Non-majoritarian institutions have a mainly limitational function.

It is this constitutionalized system of rule based on majoritarian institutions that has lost in relevance and importance in the last decades. Instead, mostly specialized authorities – somewhat independent of the constitutionalized rule and most often not backed by force – have gained major relevance and importance. Without a question, such functional and non-majoritarian authorities have always been there and were part of the model of constitutional democracy. Yet, in the last decades – so the first thesis goes – specialized authorities have risen in quantity and quality so that the integrated institutional system of democratic constitutionalism is now being challenged by a system of loosely coupled spheres of authority ¹⁴ that not only play a limitational, but also a foundational role – that is, a norm setting one.

While the fully developed political rule in the form of the territorial state is fraying at the edges, specialized public authorities of different types have, in recent decades, increased in significance – both on national and international, as well as transnational levels. They have significantly undermined the notion of democratic rule based on national parliaments. These extra-parliamentary institutions control the executive, and give outsiders a voice – they create, in the words of Keane (2009), a new historical form of democracy; the ‘monitory democracy’. This is good news. But this development contains bad news as well. These authorities do not only implement and control policies – as foreseen by the notion of democratic constitutionalism – they also are strongly involved in setting norms and rules, directly or indirectly. Moreover, these new public authorities have gained independence from the majoritarian institutions of the constitutional state.

Especially when it comes to transnational and international institutions, we face a set of plural authorities that can collide with each other sectorally, as well as with constitutionalized states. Leaving the era of neatly separated territorial states does not lead us to a world state, or to a moving of the constitutional state to a higher level. There is no constitutionalized place for the final decision. It may be too far-fetched to talk about full-scale institutional fragmentation, yet it is an institutional architecture that lacks centralized coordination ¹⁵ – which is why we move from encompassing constitutional rule to plural and only loosely coupled spheres of authority. A system of loosely coupled spheres of authority is however full of interface conflicts, forum shopping, and mutual adjustment of partially competing authorities. This adds to the liquidity of reflexive authorities.

The paradox of the democratic legitimation narrative

The second implication that needs sketching refers to changed legitimation processes that are associated with the rise of loosely coupled spheres of reflexive authorities. As mentioned, it is mainly constitutional rule that builds primarily on the participatory legitimation narrative. Parties, parliaments, and governments refer to participation and public opinion, to votes and voices when they justify their exercise of authority. They are often backed by the tradition of the nation and the symbols of the state and the nation. On the contrary, central banks, regulatory agencies, and also transnational and international institutions display primarily a technocratic legitimation narrative when justifying their authority. In general, their legitimation narrative points to effectiveness, wealth, and superior knowledge. Politically assigned epistemic authorities emphasize expertise even more than international political authorities that often also emphasize accountability to member states. National and international courts, in addition, build on the liberal narrative of legitimation; that is, the protection of individual rights and the rule of law. In general, the legitimation of spheres of authority mainly depends on results and their ability to adapt to new circumstances (Sabel and Zeitlin 2012; see also Macdonald and Macdonald this issue) (Figure 1).

Figure 1 From democratic rule to loosely coupled spheres of authority.

At first sight, this looks like a logical and smooth development, as it is in line with reasoning about different legitimacy standards for different types of authority (Majone 1994; Moravcsik 2006). A fully constitutionalized political system of rule is most strongly dependent on democratic legitimacy. By contrast, a purely epistemic authority requires hardly any additional legitimation, for its authority stems exclusively from the credibility of its assessments. Politically assigned epistemic authority requires rather greater efforts for legitimation. Especially the possible suspicion of partisanship must be eliminated. For example, panels of experts need not only demonstrate the expertise of their members, but also a certain degree of plurality of representation. In any case, limited public authorities seem to be less demanding in terms of legitimation than the constitutionalized state.

However, a second look makes clear that the simplified graphical presentation should not hide all the disruptions and nonlinearities involved in this development. It has its ups and downs, and it is often challenged – by political parties on the right and left that want to empower the people, as well as by states defending their national sovereignty.

The movement from encompassing constitutional rule to plural and only loosely coupled spheres of authority will, therefore, not necessarily go on indefinitely. One of the reasons for this is that it involves an inconsistence in the currently dominant belief system about the legitimation of authority and rule. On the one hand, democracy is, according to survey data, accepted worldwide as a desirable political order. In all countries covered by the World Values Survey, the majority of respondents are in favor of democracy. With only two exceptions (Colombia and Russia), this view is taken by more than 80% of the population. On the other hand, political institutions that do not draw on genuinely democratic sources of legitimation – central banks, constitutional courts, independent agencies, the EU, and international institutions – are gaining in importance throughout the world. As a result of this development, the effective power of political institutions that justify themselves on participatory and majoritarian grounds is reduced.

This contradiction fully plays out and becomes an inconsistency in the dominant belief system when we look at survey data that show widespread backing for this development in society. People in parliamentary democracies seem to especially like epistemic authorities, whereas they dislike mostly the political institutions that are inherently connected to democratic rule in a constitutionalized state: parties and parliaments. Among the multitude of public institutions, political parties in the United States and the EU consistently scored worst on confidence, worse even than big business and the media. Parliaments, too, score badly, generally ranking in the last four of 12 public institutions (Norris 2011).

Compared with parties and parliaments, constitutional courts and central banks have a very high reputation in most countries. In many countries, especially in Germany, the constitutional court consistently outscores all other national political institutions. More generally, in all 22 countries covered by the 2008 European Social Survey, people had greater confidence in the legal system than in parliament and parties. International institutions are also grounded in a remarkable degree of recognition. In the parliamentary democracies of the West, the United Nations enjoys greater political trust than national parties and parliaments (United Nations Department of Economic and Social Affairs 2009, 34). The succeeding waves of the World Values Survey reports even show that a majority of people in the world has confidence in the United Nations (Furia 2005).

There has thus been a remarkable internal shift in key legitimation sources in western democracies. Established majoritarian procedures in parliamentary democracies have lost strongly, not only with regard to relevance in political decision-making processes, but also with respect to public support. As far as attitudes are concerned, dissatisfaction with the core institutions of parliamentary democracy – such as parliaments and parties – contrasts with the considerable esteem enjoyed by epistemic authorities. Therefore, the core institutions for majority formation – which define the realm of politics in the narrower sense – are viewed with an extremely critical eye, whereas institutions empowered to make decisions affecting society as a whole while escaping from the political process, such as central banks, constitutional courts, and international institutions, have a much better standing in public opinion than the core democratic institutions. One may label this as ‘democratic paradox’: while participatory democracy requires a constitutionalized system of rule in which especially the coordination of governance and the delegation of authorities is democratically controlled, the people who, in principle, are in favor of democracy trust especially non-majoritarian institutions that are beyond the reach of majoritarian institutions.

This democratic paradox can be expected to create precarious dynamics in the processes of political legitimation. As a result, the process of legitimation becomes reflexive as well, and conflicts over which justification is appropriate for which form of political authority may arise more and more. ¹⁶ In most of the 20th century, the question about legitimacy was one in which the standards of democratic decision making in national political systems were more or less consented and uncontested. The legitimacy question was whether and to which extent these standards were fulfilled. Today, those very standards are contested. In such reflexive legitimation conflicts, not only the question ‘Was the decision compatible with democratic constitutionalism?’, but also the standards themselves get questioned; not only the ‘what’ (What counts as effective basis for legitimation?) gets disputed, but also the ‘who’ (Who decides about this?) and the ‘how’ (How and under which conditions can we answer these questions and institutionalize the answers?) have begun to be debated. ¹⁷ It, therefore, can be expected that reflexive authority will lead to reflexivity in legitimation conflicts. These legitimation conflicts will affect the authorities, adding further to their liquidity.

Conclusion

Authority relationships in global governance are reflexive. Reflexive authority’s state of aggregation is ceteris paribus more liquid than traditional authority. To the extent that the built-in tendency toward institutionalization takes effect, reflexive authority moves away from liquidity toward the solidity end of the spectrum. This is especially the case when the institutionalization is the outcome of delegation by traditional authorities. However, this counterforce can be expected to have a limited effect. Over time, reflexive authority produces social dynamics that even increase liquidity over time. Namely, the loose coupling of reflexive authorities and their specific legitimation problems produce fragmentation and contestation, which, in turn, lead to a permanent competition and mutual adjustments of reflexive authorities. The concept of reflexive authority thus provides a reason-based account of increased liquidity and dynamism of public authority. Whereas the concept of liquid authority mainly sheds light on the form (or the aggregate state) of transnational and international institutions, reflexive authority elaborates its social foundation. While simple authority most often appears in solid form, reflexive authority in general tends to be more liquid.