Bringing back Max Weber into Network Governance Research

ABSTRACT

The current debate on the consequences of governance networks (GNs) for the future of democracy is excessively normativized at the expense of its empirical accuracy and critical force. This paper therefore aims to bring back Weber’s approach to legitimacy as an empirical problem. This approach allows studying GNs as improbable achievements that require constant legitimacy work as part of the art of networking. Weber offers a theory of ethical power, a power which nature and rationality can be grasped by studying rituals that establish and validate authority with extrarational means. Studying these microphysics of power opens up analysis to institutional complexity and promises a more accurate taxation of the possibility of democracy in contemporary governance. This paper additionally shows why this Weberian approach constitutes a distinctive voice within critical governance and elaborates the methodological tools it has to offer empirical researchers.

KEYWORDS:

• Legitimacy
• network governance
• democracy
• Max Weber
• authority

Although Network Governance Research (NGR) is still a relatively young field, there have been consistent calls for a ‘second generation’ of research that focuses less on the problem-solving effectiveness of governance networks (GNs) and more on their legitimacy and relations with representative democracy (Torfing 2005; Sørensen and Torfing, 2007; Klijn 2008). GNs are claimed to represent: 1) a rising discrepancy between traditional government and the actual arenas of politics and policy, often explained as a function of a general crisis of governability (Hajer 2003; Pierre and Peters 2005); 2) a coordinative logic of ‘heterarchy’ and negotiated action inherently different from the hierarchical and rule-following logic of bureaucracies or the competitive and utility-maximizing logic of markets (Torfing 2005; Rhodes 2007); and 3) polycentric (Ostrom 2010) fields of political power that challenge and redefine the traditional organization of legitimate state power (Pierre and Peters 2000; Hajer 2003; Rhodes 2007; Jessop 2016). As the modern state has been historically the most prominent and enduring institutional anchorage of democracy, it becomes clear that the study of GNs is also the study of the future and feasibility of democracy.

Pessimists argue that GNs are not so much hollowing out the state but hollowing out democracy. Under the ‘realistic’ reign of Alternativlos politics, democratic legitimacy is exchanged for problem-solving effectiveness. GNs signal a separation of politics and policy from democratic accountability and will-formation, a decoupling of ‘frontstage’ democratic politics coordinated by the electoral logic of party and vote, and its ‘backstage’, the politics of policy-making and implementation coordinated by the logic of effectiveness and strategic bargaining (Papadopoulos 2013). GNs indicate the ‘privatization’ of power (Jessop 1998) and the transformation of democratic arenas into ‘policy-takers’ rather than ‘policy-makers’ (Papadopoulos 2013). Moreover, GNs tend to be technocratic, opaque, nonresponsive, exclusive, and biased toward organized interests, which cannot be offset by the cryptonormative fig leaf of output-legitimacy. Although GNs might be effective strategies for ‘getting things done’ in the face of ‘wicked problems’ (Koppenjan and Klijn 2004; Provan and Kenis 2007), the dispersion of politics and policy undermines the already strained legitimacy of institutions of representative democracy, explaining political distrust, cynicism, and a vicious spiral of increasing ‘irrational’ and symbolic electoral politics decoupled from policy.

Optimists claim that the crisis of representative democracy is not the result of GNs, but conversely, GNs are the result of a crisis of representative institutions. GNs are functional responses to governability issues, which include the problem of legitimacy and authority (Hajer 2009). Legitimacy has become a scarce resource in complex society, and the legitimacy of representative democracy no longer automatically spills over into ‘policy-specific legitimacy’ (Warren 2009). The rise of GNs should therefore not be understood as a trade-off between effectiveness and legitimacy (Sørensen 2005), but rather as a search for new legitimation practices. Although this does not mean that GNs are necessarily democratic, this positive account at least opens up analysis to the possibility that democracy is transforming, the possibility of a different institutional anchorage where the old institutions seem to fail, providing a more open outlook on the future of democracy and the ground for a second generation of critical research at the ‘frontier’ of democratic theory and innovation (Warren 2009).

The main problem with this debate, this paper argues, is that legitimacy is exclusively formulated as a normative research problem at the expense of its empirical accuracy and critical force. This paper, therefore, aims to bring back Weber’s sociological approach to legitimacy as an empirical problem. This might sound surprising as the Weberian paradigm is perceived as the exact opposite of NGR, the antipode of its self-identity. Weber’s sociology concerns the politics of hierarchy, of command and dutiful obedience, and the rule of the ‘bureaucratic machine’ in state and party democracy. It is a sociology of modernity, its institutions and rationality, at odds with the complexity of contemporary late-modern society. Yet, looking beyond this stylized clash of hierarchies and networks, Weber struggled with similar problems that confront NGR today. He tried to analyze the historical rise of bureaucracy and its consequence for the organization of state power and democratic politics, providing the sociological basis for generations of critical-democratic theory. The analogy with the contemporary rise of networks seems apparent and is what this paper tries to pursue.

Bringing back Weber is to bring back power to the center of analysis, which corresponds with the general aim of the recent rise of ‘critical governance’ (Newman 2014), sometimes labeled as the ‘third generation’ of NGR (Skelcher, Sullivan, and Jeffares 2013). However, as this paper will argue, Weber’s approach to power is profoundly different, especially from the Gramsican and Foucauldian approaches in this literature.

Bringing back Weber is to study GNs as improbable achievements, as normative spaces that require constant legitimacy work as part of the art of networking. Weber offers a theory of ethical power, in which nature and rationality can be grasped by studying rituals of power through which authority is established and validated with extrarational means. Studying these microphysics of power, moreover, opens up analysis to institutional diversity, ambiguity, and conflict, which helps us assess more accurately the future of democracy, or in Weber’s words, whether democracy is ‘at all possible’ (1978, 1403).

The aim of this paper, then, is to bring back Weber’s approach to legitimacy and power and argue for its continued relevance for contemporary governance. In the final section, it will argue that it is possible to recognize a Weberian spirit in contemporary governance studies as it cautiously shifts away from power as discourse toward power as performance. By reviewing some of these studies, this paper will draw out the methodological tools a Weberian approach has to offer empirical scholars.

A normativized debate

The second generation of NGR calls for ‘radical discussion about the future of democracy’ (Sørensen and Torfing 2007, 233), which means both ‘the need for a critical theory’ assessing current changes in politics and policy (Warren 2009, 11) and a prescriptive theory on how to address practical problems encountered. Central within this discussion is the question of legitimacy. How legitimacy is organized within GNs and how these networks consequently ‘couple’ or ‘challenge’ traditional democratic institutions are the key questions nourishing the different scenarios of future democracy (Klijn and Skelcher 2007). However, the literature tends to confuse the normative relevance of institutional transformations of state and democracy, by considering legitimacy primarily as a normative research problem. This leads to an excessively normativized debate undermining its own goal of a critical-democratic research program.

First, the debate is overly preoccupied with the question of how to assess the democratic ‘quality’ or ‘performance’ of GNs. To assess GNs, it is argued, we need a ‘post-liberal’ theory freeing democracy from its traditional connection with liberal institutions (Sørensen 2002, 2005; Sørensen and Torfing 2007). Such a post-liberal democratic theory, however, is little more than a deinstitutionalized list of democratic values such as transparency, participation, representation, deliberation, responsiveness, and learning (Bogason and Musso 2006; Esmark 2007; Dryzek 2007; Warren 2009), which can be used to evaluate any institution upon its democratic quality. But what does it mean if networks ‘score’ upon some value and not upon others? This ‘assessment literature’ is foremost an attempt to show that networks are not necessarily democratic threats and traditional democracy should not be idealized either. But as a consequence, it shows an astonishing lack of institutional concreteness, ignoring ‘the pluralism of our moral universe’ and ‘the indeterminacy of our normative principles’ (Bader and Engelen 2003).

Difficult trade-offs between democratic (or other politically relevant) values are hardly considered, and the idea that all these values neatly accord under a single universal moral premise of ‘democracy’ is questionable. Congruency, if thematized at all, seems to be based upon the unproblematized acceptance of the deliberative tradition (Dryzek 2007; Esmark 2007). However, this normative ideal tends to subordinate legitimate politics to universal morality with little to say about the reality of value-pluralism, thick-ethical group identities, or democracy as interest articulation (Netelenbos 2016). Post-liberalism is therefore also silent on the need and desirability of institutional differentiation. Difficult trade-offs between majority rule, public interest, and expert knowledge have historically been addressed by some form of institutional differentiation, giving rise to an institutional complexity of different and often conflicting value-logics. Post-liberal theory, however, holds not only the questionable idea that all political institutions or ‘any situation, structure or process’ must be evaluated and coordinated by the same universal list of democratic values (Dryzek 2007), but also that an increase in the democratic quality of networks therefore automatically ‘supplements’ representative democracy (Sørensen and Torfing 2007; Warren 2009). Post-liberalism completely ignores the institutional complexity of democracy.

Second, even studies that address the historical transformation of democratic governance are drawn into the normative realm at the expense of empirical accuracy, as the empirical problem of political legitimacy is reduced to normative discussions about democratic accountability (Pierre and Peters 2005; Bogason and Musso 2006; Sørensen and Torfing 2007; Esmark 2007; Papadopoulos 2013). For sure, most studies conceive that governance institutions need to be accepted as legitimate by the actors involved (Skelcher, Sullivan, and Jeffares 2013; Klijn and Koppenjan 2016) and reasonably argue that practices of accountability have been historical key sources of legitimacy (Olsen 2015). It follows that the rise of governance constitutes a legitimacy crisis to the extent that traditional accountability practices are weakened. The problem, however, is not just that these studies tend to use reified ‘textbook models’ of democratic accountability (Papadopoulos 2013, 3) but also that it allows them to discuss empirical legitimacy problems in terms of alternative models of accountability.

It gives birth to a literature in which the crisis of democracy is not so much an empirical-institutional problem but rather a theoretical problem (Sørensen 2002; Dryzek 2007; Durose, Justice, and Skelcher 2015). Weale, for example, argues that if we conceptually deconstruct the nature of democratic accountability into its different ‘principles’ and ‘rationales’ and use these as a the proper ‘standard of evaluation’ instead of historically given democratic institutions, legitimacy problems might be shown to be exaggerated (Weale 2011). Others propose alternative types of accountability that are assumed to be functional equivalents of traditional practices and which relax the problem of network accountability (Erkkillä 2007; Esmark 2007; Klijn and Koppenjan 2016). This kind of reasoning, however, takes us a long way from the original empirical problem that actors need to accept institutions as legitimate. It threatens to become research by assumption only as ‘empirical’ legitimacy crises are merely reflections of normative theories and conceptual models used.

This normative slippage, moreover, confusingly clashes with the central governance narrative in which legitimacy problems, as part of a general governability crisis, are considered drivers of institutional change and the search for innovative practices (Pierre and Peters 2005). Legitimacy problems that explain governance, as it now turns out, can miraculously be solved by changing our normative concepts and ‘theoretical lenses’ (Weale 2011). Legitimacy as an empirical resource, however, is not some form of ‘window dressing’ but a crucial aspect of concrete governance practices (March and Olsen 1984). The legitimation of political power and institutional rules is a difficulty that requires constant attention in different practices and institutions, targeting different publics. Political legitimacy cannot be unproblematically presumed, nor can we ignore the problem that political power can be legitimated in nondemocratic practices. To understand the empirical transformation of democracy and its institutional anchorage, we cannot replace empirical problems with normative theory.

Finally, this lack of empirical anchorage and curiosity also affects the practical goals formulated. It is difficult to see how democratic theory that is not firmly grounded in empirical and sociological accounts of its feasibility can get beyond rather gratuitous calls that we need to assure inclusiveness and equality. The rise of GNs is not the result of a democratization process but a response to structural pressures and problems of governability that drive the transformation of state. Merely hoping that dilemmas shaped by political struggles, power, and interests can be fixed by ‘strategic leadership’ of democratically inspired ‘brokers’ at the boundaries of network and hierarchies (Bogason and Musso 2006; Sørensen and Torfing 2009), without a critical account of structural pressures and limitations, does not seem to get us anywhere. These empirical problems cannot be understood or overcome by a post-liberal recipe book listing the normative ingredients of democracy. Democracy is an institution that gets its historical shape and specific nature within a complex field of powers, legitimations, interests, institutional logics, and social and political struggles. Democracy is not about free-floating values, but how these values sediment and are shaped within this sociological and historical complexity. If the practical goal is to enhance the democratic quality of politics, we cannot ignore these issues; we cannot ignore politics, power, and conflict, and how these are legitimated in practice.

As Shapiro once put it, ‘speculation about what ought to be is likely to be more useful when informed by relevant knowledge of what is feasible’ (2003, 2). To critically debate ‘the future of democracy’, we need to redress this overly normative, deinstitutionalized, ahistorical debate, without hiding politics, power, and conflict behind a ‘harmonizing rhetoric’ of consensus and voluntary cooperation at the same time (Offe 2009).

Bringing back Weber

What separates Weber’s sociology from current NGR, therefore, is less his concern with bureaucracy and more his approach of political legitimacy as an empirical problem. What that exactly means, however, is often subject to confusion.

In general, a sociological approach to legitimacy concerns the study of normative institutions that structure political life (Marsh and Olsen 1984). It implies an interest in describing and explaining how normative rules, norms, and expectations influence and structure political action, in what Ostrom calls ‘rules-in-use’, those formal or informal rules that ‘prescribe’ which actions are ‘required, prohibited or permitted’ (1986, 5). The sociological approach, then, concerns empirical research of a normative phenomenon, clearly distinguishable from the moral-philosophical tradition, the ‘dogmatic sciences’ (Weber 1978, 33) that try to prescribe what ought to be accepted as legitimate. More difficult however, this divide between the sociological and philosophical traditions does not neatly parallel the difference between empirical and normative studies. Much empirical work on legitimacy is cryptonormative as empirical practice is evaluated by normative standards. This includes all democratic studies that assess political practice on transparency, output effectiveness, responsiveness, inclusiveness, or deliberation. Although empirical in orientation, these studies remain inherently normative.

If sociological studies of legitimacy concern the study of institutionalized rules-in-use, Weber’s approach, more specifically, is based upon the distinction between: (1) ordinary institutional life in which ‘objective’ norms of appropriateness and claims to legitimacy are routinely accepted as socially valid and therefore structure everyday action and (2) extraordinary practices of legitimation that cultivate subjective beliefs and obligations and therefore help sustain these ordinary expectations of legitimacy (1978, 213). What characterizes Weber’s approach is this dual distinction between ordinary institutional routines and extraordinary practices of legitimation, on the one hand, and objective or social meanings and subjective interpretations and beliefs, on the other (Netelenbos 2016). Ultimately, Weber’s project might be summed up in the key question: how are subjective beliefs, duties, and obligations that sustain institutional norms and routine practices mobilized and cultivated in extraordinary practices of legitimation?

This question clearly shows the fundamental distinction between Weber’s sociology and contemporary NGR. What sets Weber’s theory apart is not his emphasis of government over governance, hierarchy over heterarchy, or bureaucratic rule-following over horizontal negotiated action, but his curiosity in how legitimacy is done in practice. Why do actors accept certain rules, norms, and expectations as appropriate, even against their own preferences? How is that improbable achievement accomplished, especially when routines are uprooted and dislocated, when norms are contested, ambiguous, or indeterminate? Indeed, when legitimacy beliefs themselves lead to highly emotional conflicts and resistance?

Bringing back power

Weber studied legitimacy to grasp political power, its probability, its specific nature and rationality, and its institutional complexity and historical configurations. This provided him the analytical tools to make sense of the historical rise of the bureaucratic state and its consequences for democracy. This section aims to explain Weber’s critical theory of power and its continued relevance for NGR.

The probability of power

A common misunderstanding equates the study of legitimacy with the study of political stability. Legitimacy is perceived as an essential condition for political rule as feelings of duty are boundaries on the centrifugal forces of group conflict and self-interest, while coercion is excluded as inefficient or normatively undesirable. As moral reasoning is feeble or inherently lacks motivation, the ‘sociology of belief’ studies moral sentiments and obligations as ‘artifice’ of authority, as Hume called it (1992, 51), without which the body politics must collapse and dissolve into conflict.

Many also interpret Weber’s study of legitimacy as a study of stability. Indeed, he claims that beliefs in the right to rule are the most ‘reliable basis for a given domination’ (1978, 213). If people’s obligations to obey political rule are ‘internally guaranteed’, domination is the most stable and efficient. However, Weber was well aware that political order can be explained by its ‘objective’ side only, by ‘external guarantees’ limiting choice alternatives. To explain political order, as Weber admits, actors do not have to be oriented to the belief in its legitimacy, ‘even at all’ (1978, 214). From a social action perspective, the question is not whether people believe in the legitimacy of institutional rules-in-use, but whether these rules are capable of effective social coordination. Like pluralist analyses of political bargaining games, it does not matter whether actors believe in the legitimacy of the rules of the game. What matters is the social validity of these rules, their facticity.

This begs the question why normative beliefs and sentiments should be studied at all. Some therefore conclude that Weberian legitimacy is irrelevant for the social sciences (Marquez 2016). Similarly, some network scholars start from the assumption that ‘legitimacy is not an issue for network governance’ (Börzel and Panke 2007, 160). They assert that networks are game-like interactions between strategic actors constrained by mutual dependency, institutionalized rules of the game, and a shared interest in ‘getting things done’ − so why study legitimacy?

What is at stake, however, is not the stability of political order grounded in moral duties and beliefs, but the effectiveness of legitimate power as a coordinating force: the problem of the ‘reliability’ and the ‘probability’ of power, in Weberian terms (1978, 31). In everyday bureaucratic life, for example, the rules-in-use are routinely expected to be socially valid and therefore coordinate bureaucratic action independently from underlying motivations. However, Weber was well aware of the vulnerability of these normal expectations, which thus require constant symbolic work in legitimation practices, especially in times of crisis when expectations are uncertain, contested, or violated. Expectations need to be established, reconfirmed, and reinvigorated in extraordinary practices that prove and validate these norms. The importance of legitimacy, then, is not to explain everyday rule-following behavior or motives of obedience, but the coordinative force of ‘rules-in-use’, the improbability of which requires a specific kind of legitimacy work.

Bringing back Weber means to study governance practices as vulnerable and normative spaces. Indeed, the vulnerability of social coordination is a fundamental characteristic of networks for at least three reasons. First, networks are not about routine actions, as norms, rules, and expectations must be negotiated in relatively deinstitutionalized settings in between different social systems with their own value-logics (Torfing 2005). The complex mixture and layering of different forms and styles of governance (Rhodes 2007) ensure that expectations of what is proper and right, of obligations and responsibilities, are indeterminate, unstable, or incomplete. Second, despite the ‘rhetoric’ of consensus, trust, and cooperation (Offe 2009), it is quite clear that networks are littered with conflicts, tensions, and disruptions, including role conflicts between politicians and network managers (Pollitt 2003), boundary conflicts determining who is in or out (van Stokkum 2006), or conflicts of interests between participating stakeholders (Klijn 2008). Finally, despite its ‘shadow’, governance takes place beyond the direct reach of state coercion and control, on the one hand, and cannot be reduced to market competition on the other. So, even if the rules of the game are clear and common interests identified, their coordinative force is inherently vulnerable, as it is not guaranteed by either coercion or competition, but by normative sanction only (Powell 1990). The risks of noncompliance, free-riding, and contestation must be normatively absorbed in networks by cultivating value commitments. Offe therefore concludes that if we cannot explain the persuasive and normative force of governance, the whole concept may be pointless (2009).

If anything, the study of legitimacy as an empirical problem helps us analyze the effectiveness of network coordination as an improbable achievement that requires continuous legitimacy work. Larsson similarly argues that the power that ‘resides in networks’ is a ‘sovereign power’ that defines it as a ‘political space’, a ‘constitutive’ power ‘detached from the coercive and violent instruments’ (2013, p. 101). In Weberian terms, this means that the study of networks is the study of legitimacy as an empirical problem.

The nature of political power

Weber observes that the need to legitimate goals, actions, and decisions is an inherent human condition, especially when power and inequality are involved (1978, 491). These legitimations and the normative values they communicate are not political facades but have an autonomous force and logic of their own. Political power constitutes a normative field, not just a constellation of interests, and it matters whether it is legitimated in terms of expertise, tradition, religion, law, or democracy (1978, 953). Weber’s main reason for studying legitimacy, therefore, is to understand the specific nature of a political order, its specific mode of power and the logic, rationality, or ‘character’ of its political organization (1978, 213). The rise of bureaucracy, Weber claimed, cannot be properly understood without explaining the rise of its specific ‘ethos’ without which ‘the whole apparatus would fall to pieces’ (Weber 1958a, 95). What is to be explained is not so much the formal organization of political power in offices, rules, and responsibilities, but the corresponding change in the nature of this power. The study of legitimacy as an empirical problem – the study of bureaucrat’s particular ‘moral discipline’ and ascetic ‘self-denial’ – is Weber’s method to unlock that nature. Weber therefore searches for ultimate sources of legitimacy that allow him to identify ‘pure’ types of legitimate domination and their ‘corresponding forms of action’ (1978, p. 20) that allow him to describe the nature of bureaucracy as the legal-rational organization of ‘domination through knowledge’ (1978, 225).

What does this mean for NGR? Without a doubt the whole field starts from the assumption that ‘organisational form matters’ (Powell 1990). Weber in addition assumes that the organizational form political power can take depends upon its specific nature and type of legitimation, and not the other way around. It hardly seems sufficient to understand networks solely in abstract terms such as the ‘logic of heterarchy’ (Jessop 1998) or the logic of strategic games under conditions of dependency and relative autonomy (Torfing 2005). To grasp its nature, we must ask what kind of legitimations makes networks possible? And what does that tell us about the nature of power in networks? Power has not suddenly and mysteriously disappeared from governance. Indeed, many scholars acknowledge that networks can be elitist, technocratic, exclusive, or dominated by powerful actors or special interests and therefore tend to be anti-political, provide poor channels for interest articulation, and marginalize the unorganized. Rhodes (2007) even claims that the study of governance is the study of how networks exclude, limit participation, privilege interests, or mobilize bias. The study of network legitimacy, the study of networks as normative fields of power with a specific nature and logic, tries to understand this bias not as some unfortunate contingency, but as part of its specific constellation.

Bringing back Weber into NGR is therefore to bring back power into the center of analysis, but on different grounds than other traditions of critical governance. At least three traditions might be discerned here. First of all, some scholars call for more attention to power understood as resource asymmetry and conflicts of interests (Marsh 2011; Kjaer 2011). Although important, this critique remains safely within the view of GNs as strategic spaces of negotiated action. A more profound critique, therefore, concerns the Gramscian approach to power as a dialectic between consent and coercion. Davies (2011), for example, argues that governance is part of a neoliberal hegemonic project, which not only explains why the reality of governance never lives up to its ‘self-governing’ ideals, but especially why the ‘hard power’ of repressive and coercive state institutions remains pervasive. Indeed, governance seems to strengthen rather than undermine state hierarchy (Jessop 2016),

Governance, then, is a ‘neo-liberal ideology’ (Eagleton-Pierce 2014). In Marxist tradition, the concept of ideology discloses the ‘ruling ideas’ as the ‘ideas of the ruling class’ (Marx and Engels 2004, 64). This ‘mental production’ of public consent points to both the power of naturalization and legitimation. The former is to present domination as without alternative (Cox 1999). Naturalization is based upon the force of factual claims about how the world, society, or human nature functions, about its essence, and which therefore excludes alternatives as unrealistic or impossible. This ‘politics of knowledge’ (Davies 2011) explains how facticity organizes and structures action alternatives. The power of legitimation, in contrast, is to present domination as right and just. It points to the power of counterfactual claims about how the world ought to be. Its force, therefore, lies not in an appeal to reality as it is but to what it has to be.

The third, Foucauldian tradition within critical governance, can be said to advance the power of naturalization, even if Foucault forcefully distanced himself from Marxist Ideologiekritik. The Foucauldian approach, however, does not deny the power of naturalization, but rather criticizes its instrumentalist character in Marxist analysis, i.e. the idea that real power somehow lies unchanged and untouched behind this ideological cloak and that it is the task of the critical scholar to reveal its true workings (Triantafillou 2012, 4). In Foucauldian studies, different ‘regimes of truth’ make possible different kinds of power. Triantafillou (2012), for example, argues that governance must be understood as a ‘scientific body of knowledge’, a ‘new epistemological terrain’ that ‘problematizes’ traditional bureaucratic government. This not only creates new objects for political intervention, but also leads to new technologies ‘through which power is exercised’, such as new public management or participatory policy-making. These techniques reproduce the ‘unquestioned’ character of these problematizations, but also produce ‘neoliberal subjects’ (Johnson 2007, 107) as they internalize an ‘entrepreneurial ethos’ through their exercise of freedom (Triantafillou 2012, 71).

Governance, then, is a body of knowledge, a set of techniques, and a specific practice of freedom. The task of the critical scholar is still to reveal power in the taken-for-granted world – in discourses, policy frames, or background habitus – but not to reveal the ‘real’ machinations of power behind it, but to show that reality can be different. The critical force of this tradition is to challenge the power of facticity and break open reality to alternatives.

Weber shares with Foucault his critique on Marxism – as legitimacy does not mask power, but changes its nature, rationality, and organization – and also his interest in the relation between power and truth. However, Weber develops the other strand of ideology and studies legitimacy as an ethical power that claims to be right and just. He studies power as counterfactual truth-claims that are valid even despite reality and in opposition to everyday facts. Such power, therefore, is not based upon the authority of cognitive knowledge but upon the authority of normative convictions. Normativity means that even if the facts tell otherwise, it ought to be different (Luhmann 1985).

Although discourses ‘legitimate’ policies, prescribe what is normal and appropriate behavior, and allow a ‘normative leap’ from ‘is to ought’ (Stone 2012), the sociological force of discourse is nevertheless based upon its appeal to factual truths. Discourse is a cognitive, epistemic, and factual power – an expert power that essentializes, categorizes, naturalizes, and problematizes reality ‘as it is’.

The sociological force of legitimacy, Weber argues, cannot be understood by studying the force of cognitive facts, bodies of expert knowledge, or philosophical or scientific reasoning. Power, for Weber, is not an epistemic problem of this kind. Legitimacy validates its counterfactual truth-claims in non-cognitive truth-experiences, in rituals of proof that mobilize extrarational passions, sympathies, and convictions (Netelenbos 2016), and ‘obligate’ those present to ‘recognise’ power as valid (Weber 1978, 242). In other words, if scientific methods obligate the observer to recognize the validity of factual claims, legitimation practices obligate the witness to accept the validity of counterfactual claims. The first appeals to the factual, cognitive, and rational, and the latter to the counterfactual, non-cognitive, and extrarational. For Weber, therefore, power is foremost an ontological problem as it must be experienced as right and proper.

Weber provides a theory of ethical power, of authority that claims rights others have the duty to recognize as just. A power, first of all, that establishes what is normatively appropriate within specific relationships, institutions, or social configurations (Tyler 2006). Second, this ethical power is not hidden in the taken-for-granted world waiting to be revealed by critical scholars, but is a visible power. It explains the authority of leaders, rules, and norms, of which actors are in principle conscious. However, and third, legitimacy is not necessarily repressive even if it justifies violence. Although it increases the probability of power – enlarging the freedom of ‘superiors’ (Weber 1978, 212) – it also binds power – increasing the ‘degrees of freedom’ of the ruled (Luhmann 1975, 45ff.). What matters is not the false choice between domination and non-domination, but what kind of authority legitimacy makes possible. Finally, legitimacy is an extrarational power, an enchanting power relatively immune to facts and rational argument. Its persuasive force explains the specific ‘pathos’ of power and the emotive side of politics, not as some irregularity but as central to its functioning.

A critical project

Weber asks us not to study governance in terms of asymmetric negotiating power, as an ideology backed up by coercion, or as epistemic fields allowing specific governmentalities, but as a normative field of power, as a moral space that constitutes a specific kind of authority. The critical force of this project, however, might be less obvious as Weber was foremost interested in describing and explaining different types of authority, while ardently trying to uphold a self-restrained ‘politics-free’ science (Weber 1958b, 145–6). Nevertheless, there are at least two reasons why a Weberian approach to power contributes to critical governance.

First, Weber’s analysis of power opens it up to the political understood as ‘the warring of the gods’, as conflict between ultimate values (Weber 1958b, 152). In his discussion of scientific practice, for example, he argues that when reason interrogates reason, at a certain point we hit the hard layers of the axiomatic and the normative (Weber 1958b, 143ff.). At that point where reason falters and where we are left without answers, the words of Luther might echo: ‘Here I stand, I can do no other’ (Weber 1958a, 127). Revealing these extrarational foundations of social order is an inherently democratic project to the extent that it challenges the subjugation of politics by science, philosophy, law, or any other system of truth. There are no single, univocal answers that can replace political judgment.

Second, a Weberian approach takes the future of democracy more seriously. Pointing out the ‘political nature’ of networks and the problem of power asymmetries might have launched an interest in democracy (Larsson 2013), but it fails to escape the normativized debate discussed above. Gramscian or Marxist approaches, by contrast, have a strained relationship with democracy. Representative democracy is primarily a functional means to legitimate state power and absorb the contradictions of capitalism (Offe 1984). Democracy in a capitalist system is almost per definition pathological. In governance studies, this translates into the position that GNs are nothing new, but ‘governance as usual’ (Davies 2011, 3). Foucault’s interrogation of power, finally, might invoke an ‘an ethos of democracy’, but he hardly has anything to say about democracy as a set of institutions (Connolly 1993). Weber, conversely, approaches democracy as a historical, dynamic, and composite institution. Indeed, despite his bleak view of a bureaucratized and disenchanted future, his hesitation whether ‘new formations of authority’ are possible at all (1978, 989), Weber was committed to study the possibilities of democracy (1978, 1403). To do so, we need to take institutions seriously.

Institutional complexity

On a general level, democracy can be understood as a specific method of legitimating state power. Yet Weber was wary of ‘reifying’ a uniform state concept or legalistic notions of state sovereignty (1978, 14ff). To grasp the modern state ‘as a complex of social interactions’ is to grasp how authority establishes and validates itself differently in various institutional settings and relationships. This not only means there exist multiple kinds of authority within a democratic state, but especially that we cannot mystify legitimacy as some kind of societal value consensus. Contrary to post-war pluralism or functionalism, there is no singular, harmonious societal normative order, some kind of consensus ‘out there’ that can be unproblematically assumed. Authority necessitates legitimacy work in diverse practices addressing a plurality of publics.

Letting go the myth of a uniform, monolithical political system underpinned by societal consensus is to take seriously the differentiated polity and the pluralist state. Democracy consists of an assemblage of institutions, relationships, and authorities. While Weber is renowned for his ideal-typical approach, his more historical-institutional analysis shows how different and often conflicting modes of legitimacy exist within and between institutions. The study of legitimacy is not to explain political stability, status quo, or consensus, but institutional tensions, dynamics, and conflicts that propel history.

Weber analyzes the political dynamics of ‘modernity’ from inherent conflicts between different kinds of authority claims and legitimation practices (Netelenbos 2016). The future of democracy depends on how these institutional value-conflicts play out. Weber was especially concerned with value-conflicts between politics and administration or leadership and party. If he thought that mass democracy is hardly imaginable without bureaucracy, he showed how the ‘public’ nature of democratic authority clashes with the ‘secrecy’ of bureaucratic authority (1978, 991). Therefore, democracy must inevitably come ‘into conflict with the bureaucratic tendencies’ (1978, 985). He also studied value-conflicts of democratic leadership, torn between a politics of conviction and responsibility (Weber 1958a, 126), and the normative ambiguity of bureaucracies based upon legality and expertise, management and professionals, or the ‘moving spirit’ of politics and the ‘ethos of office’ (1978, 1404).

The point here is not to put forward these value-conflicts as universal laws of democratic politics. Rather, modern government (or any complex organization) cannot be properly understood without being aware of its different normative orders and how these historically play out. The empirical study of legitimacy shows the complexity of ‘the’ political system, its conflictive, indeterminate, and ambiguous nature, which cannot be reduced to a play of interests and bargaining power only, let alone to (post-)liberal theories of justice that disregard institutional complexity and value-plurality. If anything, disclosing the different and often conflictive normative fields within the modern democratic state provides Weber the tools for grasping the historical consequences of the rise of bureaucratic organizations for representative democracy and the forces that determine its future.

This sensitivity for institutional complexity is much needed within NGR. Indeed, one of the key questions in discussing the ‘future of democracy’ concerns the institutional linkage between public policy negotiated in multi-actor networks and traditional arenas of democratic decision-making. How networks ‘couple’ with hierarchies is a recurring theme in the literature (Pollitt 2003; Klijn and Skelcher 2007; Offe 2009; Papadopoulos 2013), but without much empirical research. Not only do we need to take the institutional pluriformity of ‘hierarchy’ seriously, we also need to study networks as normative spaces with their own repertoires of legitimacy. What is the nature of authority that emerges in networks? How do value-conflicts at the interface between networks and hierarchies propel ‘shifting patterns of governance’ (Rhodes 2007)? And how does network authority challenge and transform the traditional bureaucratic and political repertoires of legitimacy?

Equally important is the institutionally anchored or loosely coupled relation between networks and the general public. GNs address public problems, make collective decisions, and are therefore part of the authoritative allocation of values. The problem of their external legitimacy, the problem of governance ‘persuasiveness’ (Offe 2009) or ‘authority’ (Hajer 2009), however, has not received much attention. How innovative governance practices are able to operate in highly mediatized environments and how they are able to mobilize political trust and support beyond the circle of direct participants are nevertheless crucial. Despite the general narrative in which governance is a strategic response to legitimation deficits, it remains unclear how networks mobilize their own legitimacy or necessarily live off the legitimacy of traditional hierarchies and legal systems. It matters, however, for assessing the empirical validity of the different scenarios describing the relationbetween state, networks, and democracy.

In short, ‘the future of democracy’ depends on how networks are institutionally embedded – how new repertoires of network legitimacy conflict, supplement, or modify traditional repertoires of authority. The study of legitimacy not only comprises the study of networks as normative fields of power, but also opens up analysis to institutional differentiation and value-conflicts, providing more leverage on organizational complexity and institutional dynamics.

From hierarchy to networks

Bringing back Weber promises to address the overly normativized debate on the future of democracy. Weber asks us to study GNs as 1) vulnerable spaces, requiring legitimacy work; 2) normative fields of power, making possible specific kinds of authority, rationality, and freedom; and 3) embedded in a wider normative-institutional context, allowing us to study the tensions, dynamics, and value-conflicts that propel state transformation. Studying legitimacy as an empirical problem therefore provides the necessary sociological building blocks for a sound critical-democratic theory of governance. The more difficult question, however, is how such Weberian research of networks must look like.

It is one thing to pursue the analogy between ‘the bureaucratic machine’ at the turn of the twentieth century and contemporary GNs; it is also clear that Weber’s specific and historically situated theory of legitimacy is a theory of hierarchy and bureaucracy, hardly describing governance processes of networked and negotiated action. Trying to make sense of the ‘historically unique configurations’ of politics in his own day and age (Weber 2011, 93ff.), Weber shows how appropriate beliefs sustaining the ‘right to rule’ are cultivated in charismatic revelations of truth that ‘move the soul’, in sacred rituals in which actors experience the sanctity of ‘unalterable’ traditions ‘overshadowing’ the actor, or through ‘ascetic discipline’ or vocational ‘calling’ found in a rationalized order or bureaucracy (Weber 1978). Weber, in short, provides a quasi-religious theory that explains the ‘internally sanctioned’ duty to recognize and accept legitimate domination in terms of existential meaning, in one way or the other (Netelenbos 2016). These rituals of power make possible a kind of authority that is stern, unapproachable. and unquestionable – an ‘authoritarian power of command’ and unconditional obedience (Weber 1978, 946), establishing relationships of deference between ruler and ruled, master and servant, leader and followers, or patriarch and household.

This particular historical interpretation of the question of authority provides little leeway for the analysis of more horizontally organized governance practices in which decision-making powers appear to be dispersed and lack clear authorizing moments and the distribution of roles and responsibilities caught up in ongoing negotiations and deliberations. This does not mean that GNs should therefore be understood as ‘non-normative’ fields of strategic negotiations or as ‘domination-free’ spaces of democratic deliberation. Rather, we must study networks as normative spaces that make possible different kinds of authority.

It might be instructive, in this regard, to take a closer look at the literature that focuses specifically on participatory or deliberative forms of governance understood as experimental and innovative spaces of doing policy and politics. Although this literature is often cryptonormative, indeed, part of what Davies (2011) would call the ideology of innovative government not living up to its promises, and often overzealously contributing to what Triantafillou (2012) calls the ‘epistemic field’ problematizing traditional government, it is nevertheless telling that this literature entails a cautious shift from discursive power – the power of language, knowledge, frames, storylines, etc. – toward power as a performance – the enactment of authority in symbolic practices.

This shift is instigated by the recognition that success or failure of deliberative practices cannot be explained by discourses itself or by studying the exchange of arguments. The muddy reality and the vulnerability of deliberative practices rather show the importance of ‘participatory rituals’ (Forester 1999) in which emotions are vented, relationships established, and commitments forged. As such, scholars point out the importance of entertainment in deliberative settings (Akkerman 2001), of field trips and listening sessions (Hendriks 2009), ritualized storytelling (Wagenaar 2002), and joint-fact finding practices (Laws and Forester 2010). These rituals are less about explaining the epistemic rationality promised by deliberative theory or post-positivism, but more about an ontological rationality through which a collective ‘we’ is experienced and recognized. Other studies are explicitly Weberian when they highlight the importance of ‘magical moments’ created by expressions of art (Kersten et al. 2001) or charismatic leaders that inspire and enthuse (van Stokkum 2006).

What all these scholars have discovered is that deliberative settings are inherently vulnerable to distrust, disillusionment, and anger, and therefore require ‘emotion work’ that establishes normative commitments to collective projects, as well as ‘vertical’ commitments to formal facilitators, experts, administrators, community leaders, or politicians, as to the rules and design of deliberative experiments in all their empirical variety (van Stokkum 2006, 67ff). They show the importance of the extrarational and emotive side of collective action. Moreover, these rituals are not some kind of window dressing, but constitutive of a persuasive authority that institutes normative relationships and prescribes what is right and appropriate. Or, in Weberian terms, how deliberative authority ‘establishes and cultivates’ itself in participatory rituals explains the ‘rationality’ and organizational form this ethical power can take.

This Weberian spirit might be the most visible and convincing in Hajer’s study of ‘governance authority’ (2009). Indeed, despite his well-known epistemic and discursive view of policy processes in which politics follows the facts, Hajer gradually discovered the limitations of this view (p. 6–7). The real problem seems to be how political authority is possible at all when the facts are contested, indeterminate, or distrusted. To understand the possibility of authority in contemporary governance, Hajer proposes to study new repertoires of authority. Although he too easily dismisses ‘traditional’ Weberian repertoires as irrelevant (p. 179), his approach remains quite Weberian. First of all, he focuses on extraordinary moments in which ‘things get unhinged’ and authority needs to establish and validate itself (p. 5). Moreover, these moments of crisis in which routine expectations are challenged not only show the vulnerability of power, but also provide a methodological lens through which to study power. It is at these moments that Hajer is able to study the necessary legitimacy work (which Hajer calls ‘dramaturgical work’) and the symbolic performances, the ‘ceremonies and rituals’, through which authority tries to establish and validate itself (p. 49). Finally, as before, the importance of these rituals is to explain the persuasive, enchanting, and counterfactual force of authority (p. 3–4).

More concretely, in one of his case studies, Hajer analyzes repertoires of legitimacy that the British Food Standards Agency as a regulative and expert authority develops in the aftermath of the BSE ('mad cow disease') crisis in the 1990s (2009, 125ff). In a conscious effort to restore authority, this expert agency tried to ‘stage’ and ‘enact’ a deliberative authority in contrast with ‘old style’ expertise. But to understand what this really means, Hajer studies its micro-practices and rituals to disclose its specific nature and ‘particular genre’ (p. 127).

Hajer shows, for example, how the specific staging of authority in recurrent board meetings aims to establish an authority ‘committed to openness and transparency’ (2009, 131), even if the actual opportunities for participation are quite limited, ‘tightly controlled’, and tend to ‘depoliticise’ issues. He shows an authority staging itself as trustworthy, as an independent agent of consumer interests, in confessions rituals – where board members declare their interest in specific cases (p. 142); in ‘disruptive’ rituals of accountability – in the form of lay members in scientific committees (p. 152); and even in eating rituals – where committee members and the public have lunch together as equals (p. 152). In addition, the agency continuously stages itself as an ‘honest’ authority committed to science, in rituals that allow ‘the possibility of discursive challenge’ (p. 131), and by constantly displaying the uncertainty of scientific knowledge as well as the necessity to make decisions nevertheless (p. 138), obliging the public to be ‘reasonable’ as well (p. 148). Finally, Hajer points out how the agency’s commitment to its new role and professional self-identity is validated in an original ‘charismatic’ moment, a highly emotional encounter with families of BSE victims, and in rehearsing and reliving this ‘foundational myth’ in rituals of storytelling (p. 146).

The nature of this deliberative authority as it surfaces in its various rituals and enactments hardly compares with Weber’s stern and unapproachable authority. Yet, the food agency still makes binding collective decisions. It still claims the right to do so. However, as Hajer stresses, even if part of a conscious strategy, these rituals are not just ‘pretence’ (2009, 138). Legitimacy work has real consequences for the organization and rationality of the food agency, which becomes apparent in value-conflicts with other institutions – e.g. when scientists need to change their traditional repertoires of expert authority – or when specific regulative issues lay bare institutional contradictions – e.g. when discussing safety risks and trying to avoid social panic it is torn between being a ‘transparent’ and ‘paternal’ authority. These frictions explain the constant struggle of the FSA to establish and legitimate its authority, the constant need for legitimacy work, as well as the continuing institutional struggles that propel its institutional ‘history’ (Hajer 2009, 165).

What Hajer and these other scholars show is that in deliberative practices, authority takes on different forms (Warren 1996). By studying the rituals of power, the practices of legitimacy that are part of the art of participatory governance, it becomes possible to understand the kinds of authority these normative spaces make possible. These empirical studies therefore show the methodological tools a Weberian approach has to offer NGR:

1. A focus on rituals of power. GNs are improbable achievements that require legitimacy work. Such legitimacy work becomes especially visible in extraordinary moments when routine expectations and repertoires are contested, unclear, or doubted.

2. A focus on relationships of power. Authority is more than Weber’s ‘right to command’. Analysis should focus on the kinds of normative relationships these rituals seek to establish, and how rights and duties are distributed.

3. A focus on the specific pathos of power. A Weberian approach, above all, distinguishes itself by putting the emotive force of legitimacy work at center stage to explain the persuasive, inspiring, and ‘magical’ character of counterfactual power. This might also be the most elusive part of Weber’s methodology as it remains difficult to rationalize the extrarational. Yet, this should not be exaggerated. If we are comfortable in describing political crises in terms of moral outrage, anger, distrust, or alienation, it must be equally possible to understand how authority is supported by loyalty, pride, deference, duty, trust, solidarity, respect, conviction, or enthusiasm. Moreover, as I tried to show elsewhere, there exists a long history of political sociology that analyze this ‘emotion work’ not just in quasi-religious terms, but also in terms of trust, dramaturgy and reasonability (Netelenbos 2016).

The difficulty of studying network legitimacy and authority concerns its ‘contextual and entrepreneurial’ character, caught up in ongoing negotiation processes and new forms of politics these processes attract (Pierre and Peters 2000). This should not stop us, however, from trying to make sense of how legitimacy is done as part of the art of networking. Replacing empirical questions with normative theory, as is the general inclination of contemporary NGR, hardly solves anything.

The call for a second generation of governance research, putting the problem of network legitimacy and the future of democracy at center stage, must be embraced. But a genuine critical-democratic approach must take history and institutional reality seriously and cannot replace empirical questions with normative theory. This paper tried to show why bringing back Weber’s sociology of legitimacy proves valuable for a critical theory of networks, and how we might go about studying network legitimacy as an empirical problem. Due to the multiplicity and complex nature of networks, this will not be an easy assignment. But the question of democracy is just too important to be left to the exchange of normative platitudes.

Acknowledgments

An early version of this article was presented in the workshop Democratic Network Governance? ECPR-Joint Sessions Nottingham 2017. The author would like to thank the workshop participants, but especially Eva Sørensen and Jacob Torfing, for their helpful comments. He would also like to extend his appreciation to the anonymous reviewers of Critical Policy Studies for their sustained efforts and insightful commentaries.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Benno Netelenbos

Benno Netelenbos is a lecturer at the political science department of the University of Amsterdam. His ongoing project investigates ‘how legitimacy is done’ in innovative political spaces of governance and how this affects traditional institutions of democratic government. His latest book Political Legitimacy beyond Max Weber (2016) is published by Palgrave-Macmillan.