Metagovernance and the State: How Government and Governance Coexist

Metagovernance and the State: How Government and Governance Coexist Joint RC 31/RC 32 Panel on Governance, Metagovernance and the State International Political Science Association Conference Madrid 8-12 July 2012 Paul Fawcett Department of Government and International Relations The University of Sydney paul.fawcett@sydney.edu.au First Draft

Metagovernance and the State: How Government and Governance Coexist The term governance has been used across a wide variety of academic fields, including public administration and policy, international relations, European Union studies and law (Chhotray and Stoker 2009). In political science, governance theory has had a significant impact by challenging conventional Westphalian understandings of the nation state. Governance has subsequently become a popular shorthand term for describing the displacement of the nation state to the supranational, regional and local level as well as the existence of important mechanisms of social coordination that exist beyond the nation state s direct control (Arts, Lagendijk and Houtum 2009; Colebatch 2009; Lemke 2007). These trends were often linked at the macro level to the growth of networks leading Kahler (2009) to conclude that governance and networks have become the dominant leitmotif of our times. The novelty and pervasiveness of the changes associated with the early governance literature continues to be disputed. However, the fundamental challenge that it has presented to many conventional understandings of public policy and administration is clearly reflected in the bourgeoning literature on partnership working, multi-level governance, participatory governance, policy networks and global governance (Levi- Faur 2012). The debate that the early governance literature generated about the continued relevance of the state as an independent object of study was important in this respect but the popularity of governance theory also had the effect of marginalizing state theory from much of the discussion on public policy and administration. This was reinforced by the new institutionalist turn that was taking place within the broader discipline at around the same time (Sørensen and Torfing 2007, chapter two). Indeed, Sørenson and Torfing (2009, 45, emphasis added) have gone as far as to argue that: In the early 1990s, public administration theory and state theory combined in a shared effort to develop a theory of the emergence, role and impact of self-governing networks within and beyond the state in processes of public governance. 1 This maybe so but the argument of this paper is that public administration theory largely trumped state theory as part of this effort. This is not to deny the important meso-level dimension that governance theory has added to our understanding of public policy and administration but simply to note that, to some extent, governance theory replaced state theory and a more meso level perspective replaced a more macro level perspective as the principal level of analysis in discussions about governance (Bang 2008; Davies 2011, 5). This is reflected in the way in which the contemporary transformation of the state has come to be interpreted and understood largely through the concepts and heuristic devices provided by governance theory (Lemke 2007; Marinetto 2003; Marsh 2011; Davies 2011, chapter two). It is within this context of contemporary developments within the discipline that it is interesting to observe the growing interest that has been shown more macro level perspectives in recent iterations of the governance literature. This is represented by the greater role that has been accorded to the state in networked arrangements, the acceptance that governance may well rely on hard as well as soft forms of power and the growing interest that has been shown in the relationship that exists between processes of governance on the one hand and the large scale societal developments

within which they take place on the other (Esmark and Triantafillou 2009; Stoker 2011). This article contributes to these debates based on the assumption that a sophisticated understanding of the trends and transformations associated with governance requires an engagement with macro as well as meso and micro level approaches (Daugbjerg and Marsh 1998). The questions raised by macro level approaches are important not only because the literature on governance appears to be talking about the state more than it did in the past but also because many of the selfacknowledged limitations of this literature can only be addressed through an engagement with the frameworks that have been developed by more macro level perspectives, including, most notably, questions about power, interests and conflict (Torfing et al. 2012; Kjaer 2011; McGuire and Agranoff 2011; Kahler 2009). Macro level approaches are unlikely to present straightforward answers to any of these questions. However, their overarching concern with examining the broader political structures and processes within which governance networks are located, including the ongoing and changing relationship between the state and civil society, means that they can provide governance theory with a framework for posing a consistent set of questions at the meso-level. The remainder of this article is split into four sections. It begins by briefly outlining the governance narrative before it then discusses three different approaches to the state within governance theory: the networked approach based on the state as a network manager; the neo-weberian approach based on the embedded state ; and the neo-marxist approach based on the political economy of the state. 2 The concept of metagovernance has been used to describe the state s role in all three approaches, although each approach attaches a different meaning to this term. A Networked State? The state has always been a key reference point in political science, which has conventionally focused on the role played by formal institutions such as the government, political parties and parliaments. The early governance literature was important in problematizing this view by arguing that the primacy of the state could no longer be assumed. This opened up a discussion about the state s continued role in an era in which it was argued that its very core was being hollowing out upwards to international organisations, downwards to regional and local government and sidewards to arm s length bodies and the private sector (Rhodes 1997). These arguments have been discussed at length elsewhere so there is no need to rehearse them. Two points will suffice. First, the governance narrative clearly presented a direct challenge to the state (Newman 2003). This is demonstrated in statements such as: Governance means there is no one centre but multiple centres; there is no sovereign authority because networks have considerable autonomy (Rhodes 1997, 109; Ansell 2000; Hajer 2003). The centralised power of the state could no longer be assumed and the boundaries between the state and civil society were becoming increasingly blurred. As Jonas and While (2007, 72) have argued: It is not that the word state has disappeared from academic and political discourse altogether (although one could be forgiven for thinking that social scientists have abandoned any attempts to develop a theory of the state).

Rather, the current preoccupation with the language of governance signifies the extent to which modes of social, economic and cultural recognition in modern democracies have changed almost to the point where the term state has become archaic or so it seems. Governance and networks were therefore viewed as synonymous and the growth of networked arrangements was taken as representative of a new era of governing. These networks were conceptualized as self-organising and generally resistant towards any attempt at being steered and the state was viewed as one actor amongst many in a series of game like interactions (Rhodes 1997). Second, these arguments can be located within a broader set of macro level approaches. These approaches, primarily developed within sociology, were arguing that we were seeing the emergence of a network society based on a putative shift towards an era of late modernity in which many of the key features of industrial society such as class, Fordist production methods, collective action, the family, the nation and the state were under strain (Castells 2000; Beck 1992). 3 These changes could also be understood in the context of a broader deterritorialisation of the state. For example, Castells (2000: 14) has argued that the state is undergoing significant transformations explained, in part, by the ascendancy of networks and the advent of an Information Age: Thus, overall the new state is not any longer a nation-state. The state in the information age is a network state, a state made out of a complex web of power-sharing, and negotiated decision-making between international, multinational, national, regional, local, and non-governmental, political institutions. These sociological approaches are therefore important in identifying a series of economic, social, cultural and technological changes that could be used to help explain the emerging complexity identified by the early governance literature and the challenges that it arguably presented for state governance. However, more often than not the link between the macro level changes identified by sociologists and the transformations in the state identified by the early governance theorists were discussed implicitly within the literature on public policy and administration. The arguments contained in the early governance literature logically called into question the very body which state theory had accorded analytical priority the sovereign national state and it is within this context that the state and state theory were largely marginalized from discussion (Kelly 2000, 288). This fed into a more general trend towards meso-level studies of institutional arrangements and their effect on governance outcomes, including case study based research on the collaborative interaction between state and non-state actors in networked arrangements. This makes governance theory s more recent interest in the state an interesting development. The remainder of this paper examines three different ways in which the state is currently discussed in governance theory starting with the state as a network manager.

Networked Approaches: The State as a Network Manager The state s role has been recast as a metagovernor within more networked approaches. This is an approach that is broadly sympathetic with the network governance narrative and argues that networks have superceded markets and hierarchy as the dominant mode of governing (Klijn and Edelenbos, 2007). Here, I focus on Torfing et al. s (2012) contribution to this debate but draw on other work where appropriate. Torfing et al. (2012, 122) begin their chapter on metagovernance by arguing that: Metagovernance can be defined as the governance of governance as it involves deliberate attempts to facilitate, manage, and direct more or less selfregulating processes of interactive governance without reverting to traditional statist styles of government in terms of bureaucratic rule making and imperative command. This definition is typical of the dominant way in which more attention has been directed towards the state and its role in governance within more network-based approaches. As such, the concept of metagovernance is introduced to refer to the different ways in which the state can govern otherwise self-governing networks, including how: traditional forms of government and new and emerging forms of interactive governance are connected (Torfing et al. 2012, 132). This discussion around metagovernance clearly reasserts a more central role for the state in networked arrangements in comparison with the earlier governance literature (cf. Milward, Provan, and Else 1993). For example, Torfing et al. (2012, 142) have identified public authorities as naturally born metagovernors and metagovernance as a critical instrument in delivering effective and legitimate governance outcomes (cf. Sørensen and Torfing 2009). 4 The state is no longer viewed as one actor amongst many in networked configurations but an actor who may actually exercise at least some influence over the policymaking process: While agreeing that interactive forms of governance prevent states at various levels from exercising the kind of control that they are assumed to have had in the past, we should be careful not to paint a picture of a totally incapacitated State that is rendered powerless by the advent of new forms of governance In short, governments are capable of influencing policymaking in the decentred world of interactive governance through the exercise of metagovernance (Torfing et al. 2012, 131-2). As such, the state s role is reinstated, but in a limited way, as it must still work through networked configurations which circumscribe its role, place limits on the extent to which it can intervene and prevent it from reverting to approaches that rely either wholly or, in part, on command and control type mechanisms. This partial reinvigoration of the state s role is matched by a strongly prescriptive element to some of the more networked based approaches to metagovernance. This is particularly evident in the way in which the delivery of effective and legitimate governance outcomes has been related to the deployment of a series of

metagovernance tools (Sørensen 2006; Sørensen and Torfing, 2007; Kooiman and Jentoft, 2009). For example, Torfing (2012, 107) has argued that: Governance networks should not be left to drift and possibly fail, since reflexive and strategic meta-governance can help to facilitate, manage, and give direction to networked policy processes without reverting to traditional forms of hierarchical command and control. The state is therefore plays an important role in ensuring that suboptimal outcomes are avoided by deploying a series of metagovernance tools during the lifecycle of a network. This can involve both direct hands-on forms of metagovernance (facilitation of and participation in networks) as well as more indirect hands-off forms of metagovernance (network framing) (Sørensen, 2007). Here, the parallels between the literature on metagovernance and the Dutch network management school are clear as both approaches not only share the view that networks are now the primary way through which policy is developed and delivered but also that the state has an important role in structuring and managing networks in order to make sure that they deliver the required results (Klijn 2008; Meulemann 2008). Three points are worth making about this networked approach to bringing the state back into governance theory: first, its inherently problem solving nature; second, its pluralist and politicist tendencies; and third, the uneasy tension that exists between outright hierarchical power relations and metagoverning otherwise self-governing networks. I briefly deal with each of these points below. The first point goes back to Cox s classic distinction between problem-solving theory on the one hand and critical theory on the other. Cox (1981, 129) argued that: critical theory, unlike problem solving theory, does not take institutions and social and power relations for granted but calls them into question by concerning itself with their origins and how and whether they might be in the process of changing. In short, critical theory focuses on social change in a way that problem solving theory does not; and it becomes clear from this brief review that there is a strong tendency for the networked approach to engage in problem solving theory. For example, Sørenson and Torfing (2009, 47-8) note that: It is difficult to measure the degree to which these policy recommendations [directed towards potential metagovernors] have had impact on reforms and praxis in the public sector but the ambition to guide practitioners is definitely there, and the current focus on networks as a tool kit for governance suggest that some impact has been made. The problem solving focus of this literature has meant that it has tended to remain bound to the meso level. There has been little attempt to engage with critical theory in the way in which Cox has defined it despite the effect that such factors may have on the effectiveness and legitimacy of network governance. Thus, the origins of power relations and how and whether they might be in the process of changing remain relatively unexplored.

This meso level focus has also meant that the metagovernance approach has not really developed a coherent macro level perspective and the state has remained a relatively under-theorised concept outside the context of immediate and localized interactions within any one particular network. A great deal of this literature is subsequently underpinned by a largely pluralist and politicist account of governance in which there is a diversity of relatively loose and open networks in which the state reflects the preferences of the strongest groups in society and where there is no discernable or consistent pattern of inclusion or exclusion across different networked arrangements (Daugbjerg and Marsh 1998, 58). The introduction of the state into governance theory within more networked based approaches has therefore occurred without much change to any of the underlying ideas about the nature of the state or state power. As Levi-Faur (2012, 10-11) has argued about the early governance literature: Still, there is a more important and illuminating point here; this perspective is strongly connected with pluralist and neo-pluralist theories of the state, which tend to see the state as a broker or even a weathervane. The autonomy of the state is constrained and it reflects the preferences of most of the strongest groups in society In short, this governance approach is a society-centred analysis despite Rhodes s effort to draw lines between his perspective on governance and pluralism, they belong to the same intellectual and scholarly family. In short, the networked approach to bringing the state back into governance theory lacks any broader consideration of the social, economic and cultural context within which governance takes place (see Bang (2011) for an exception to this mainstream view). Finally, there appears to be an ongoing tension in the way in which the state has been reconstituted in networked approaches. For example, Sørenson and Torfing (2009, 45) continue to ascribe to a definition of a network as: a relatively stable self-governing group of interdependent but operationally autonomous actors who work together in order to reach negotiated goals that contribute to what is at a given point in time defined as public purpose. It is subsequently argued that the autonomy of networks means that their metagovernance is something qualitatively different from a style of governing based on command and control: Metagovernance cannot revert to traditional forms of hierarchical steering as metagovernors must respect the capacity for self-regulation of the interactive governance arenas in order to preserve the commitment of the public and private actors. As such, the concept of metagovernance does not, as it has been suggested, endeavor to bring the State back in by insisting on its omnipotence and, consequently, reduce governance to the tools, strategies and relationships used by governments to help govern (Bell and Hindmoor, 2009: 191). Rather, the notion of metagovernance offers a way of balancing state-centred and society-centred views on how society and the economy are governed (Torfing et al. 2012, 132).

In other words, metagovernance is about achieving a balance: Too tight a metagovernance undermines the motivation of network actors to put resources into solving public policy problems, and too weak a metagovernance reduces the ability of networks to produce negotiated outcomes that go beyond the lowest common denominator and undermine the coordination of public policy (Sørenson and Torfing 2009, 47). In my view, this creates room for analytical confusion between metagoverning otherwise self-governing networks on the one hand and conventional hierarchical relations on the other thereby displacing the principal-agent relationships that are otherwise inherent within metagovernance as a process and the metagovernance tools described above. If it is accepted that metagovernance does introduce a more or less direct form of hierarchy into networks then metagovernance in the networked approach becomes less an argument about a new mode of governing per se and more an argument about how much hierarchy can be imposed before a network no longer functions as a network and we must be called something else. More broadly, I would also argue that the networked based approach to metagovernance significantly downplays the extent to which the state reverts back to hierarchy and the old way of doing things or shifts from what Bang (2011) has called policy-politics to politics-policy. This occurs when the state undermines the very transformative capacities that it requires in order to achieve its success by leaning more and more heavily on the principal-agent model as a way of governing the institutions and networks that are in charge of delivering services (cf. Crozier 2011). I would argue that these issues extend, in part, from the tendency to use a narrow definition of governance within the networked-based approach that reflects a very particular type of governing form. I would argue that, at the very least, engaging with a concept such as metagovernance suggests that there is a need to use a broader notion of governance. Network governance should only be viewed as one sub-type of governance and whereas most accounts within the network tradition have only focused on the network form metagovernance can only really make sense if one recognizes the other sub-types of governance and the variation that exists between them. Neo-Weberian Approaches: The Embedded State Neo-Weberian approaches have been more critical towards the network governance narrative. The overall argument is that whilst important transformations have taken place in the organization of the state the extent of these changes and their effect on the state s capacity to govern has been over-exaggerated (Pierre and Peters 2000; Peters 2004). The state therefore continues to retain significant control over the way in which society is governed and hierarchy continues to be a principal component in the way in which this is achieved. Levi-Faur (2012, 12-13) argues that the argument that governance is marginal and the state needs to be brought back into theories of policy is probably the most popular approach within governance theory:

Taken to the extreme, this view would suggest that polities worldwide are and should be structured around states; governance is either a marginal or temporary solution to state failures. Scholars need to bring the state back in order to tune their theories of politics and policy to the realities out there. Much of the literature of governance, probably most, would be easily classified as belonging to this perspective. Various authors have defended and developed this approach to bringing the state back into governance theory. For example, Dahlstrom et al. (2011), Peters (2010) and others have used a broadly neo-weberian approach to help explain recent trends in public sector reform, particularly the tendency towards strengthened bureaucratic and political control within western governments. Peters (2010) therefore argues that governments are faced with two choices to continue with the New Public Management (NPM) reforms and further decentre their bureaucracies or, more likely, they will attempt to deal with some of the unintended consequences of the NPM reforms, including lack of coordination, accountability and decision-making. This will lead them to strengthen their bureaucracies, particularly the role played by central agencies, using tools such as performance management, strategic management, budget and personnel controls, soft law and trust and values. Peters (2010, 48) also refers to this as a process of metagovernance in which: The choice of meta-governance strategies, then, is an attempt to reassert some balance of power within the policy-making systems of the public sector, and to continue to involve nonstate actors in the process while recognizing the primacy of politics. These arguments are important but public sector reform is representative of only one example of the way in which the neo-weberian approach has been developed. I will focus in the remainder of this section of the paper on Bell and Hindmoor s (2009) state centric relational approach, which is also taken as broadly representative of the neo-weberian approach to bringing the state back into governance theory. Bell and Hindmoor (2009) develop a state centric-relational approach based on the argument that governance is best understood as: the strategies used by governments to help them govern. Two developments are particularly important in this context. First, they argue that the state now has a much broader range of policy instruments at its disposal than it did in the past. This includes a range of hard instruments, including hierarchy and regulation and markets and contracts, but also a range of soft instruments, including persuasion, community engagement and network associations. This mix of old and new policy instruments presents the state with not only a much broader range of options and strategies than in the past but also the ability to experiment with a range of different approaches over time. The second development is the stronger and closer links that the state has been able to establish with non-state actors. These relations should not be seen as a zero sum game but one in which the state actually gains power by being able to reach out and extend its influence into previously uncharted territory, including a wide range of businesses, voluntary groups and charities. This is particularly the case if the state has developed the capacity to delegate and depoliticize certain functions whilst at the

same time recruiting and then regulating those non-state actors who are increasingly being called upon to perform public tasks (Braithwaite 2007; Braithwaite et al. 2008; Flinders 2009). Thus, governance takes place in the shadow of hierarchy in which the state retains the capacity to set standards, monitor whether or not those standards are being achieved and intervene where necessary (Hertier and Rhodes 2011). The state is therefore seen to retain a central role in all governance arrangements. Bell and Hindmoor (2009, chapter 3) also refer to this in terms of the state s role in metagovernance but they clearly use the concept in a very different way than in the networked approach described above. Here, metagovernance is used to refer to the different reasons why the state continues to retain control in all governance arrangements. Bell and Hindmoor identify six reasons, including the role that the state continues to play in steering, ensuring effective policy outcomes and resourcing, as well as its accountability functions, its continued recognition as the highest form of democracy and its legitimacy with the population at large. There are a number of parallels between this argument and neo-weberian state theory. In particular, this argument relates to a classic Weberian theme in which the state s strength comes from being both deeply embedded and partially separated from civil society (Kelly 2000, 221). Similar themes are also found in Mann s distinction between despotic power, which is distributive, and infrastructural power, which is collective. Despotic power therefore refers to the range of actions which the [state] elite is empowered to undertake without routine, institutionalized negotiation with civil society groups whilst infrastructural power refers to the capacity of the state to enter civil society, and to implement logistically political decisions throughout the realm (Mann cited in Kelly 2000, 220). Mann s argument is then extended to demonstrate how the state socially cages its occupants and their various political, economic, ideological and military powers (Kelly 2000, 221). The neo-weberian approach to bringing the state into governance theory is therefore underpinned by an alternative view as one in which the state has actively adapted to the changes identified by governance theory. This can be seen in the way in which the state has developed a much wider range of hard and soft policy instruments at the same time as it has forged a much stronger set of relationships with a broader range of non-state actors. The neo-weberian approach also argues that it has been the state and actors within the state that have been the ones to initiate the changes identified by governance theory in what Offe (2009, 555) has otherwise described as the stateorganized unburdening of the state (Weiss 1998). Neo-Marxist Approaches: The Political Economy of the State The neo-marxist approach also takes a critical approach to network governance theory but primacy is given to the political economy of the state. According to Gamble (1995, 517), who paraphrases Caporaso and Levine, a view of the state based on political economy is grounded in two distinguishing characteristics: firstly [the view] that political and economic processes and institutions are interlinked and should be studied as a complex and interrelated whole rather than as separate spheres; and secondly that a proper understanding of the

politics [sic.] requires giving special explanatory weight to economic structures and processes. Kjaer (2011) has taken this one step further and argued that governance theory would benefit from a closer and more consistent engagement with different approaches within political economy. She argues that this would help governance theory to address some of its key shortcomings, particularly the way in which it addresses questions of interest, power and conflict: We do not know enough about the tensions and dilemmas that arise when new and old forms of governance co-exist and interact, and more research on these topics is needed. In particular, we need to know more about how context affects the nature of tensions and dilemmas. To that end, comparative analysis may help to draw out a more general picture of types of dilemmas and tensions, of how they affect policy outcomes, and on how they are affected by different socio-economic and political contexts Context should be analysed with regard to social, economic and political parameters, including the nature of central and local state bureaucracy, strength of civil society, the organization of interests, and the traditions of state-society interaction (Kjaer 2011, 111). Davies (2011, 144) has made a similar argument: A key challenge for future studies of governance networks will therefore be to develop historically grounded but fine-grained understandings of how they service, or fail to serve, the ends of the connectionist project (or is successor) and to chart how the ideology of networks rises and falls in concrete settings. This means that the study of networks should be placed squarely in the political economy tradition, examining how they reproduce and embed power asymmetries or generate conflict and resistance. 5 Of course, this only takes us so far and there are several different approaches within political economy itself. The key strand discussed here comes out of the Marxist tradition. Marxist theory has changed significantly over the past thirty years. These changes have been informed by the attempt to reject reductionist and determinist standpoints on the one hand but retain the basic assumption of political economy that the state cannot enjoy absolute autonomy on the other (Gamble 1995, 524; Hay 2005; Barrow 1993). The integral role played by the state within these discussions has meant that various neo-marxist approaches have engaged with governance theory over recent years. This means that there is no one neo-marxist position but a diversity of different approaches (Davies 2011). Here, as in the previous sections, I therefore limit my discussion to two different approaches, the Gramscian perspective developed by Davies (2011), and the Strategic Relational Approach (SRA) developed by Jessop (1990; 2007). Davies notes the general unpopularity of Marxist approaches within public policy and administration but nevertheless argues that a Gramsican approach can help to understand important aspects of network governance theory that have been largely

neglected up until now. The connectionist ethos associated with network governance, and its pervasiveness within contemporary public administration, is subsequently understood as one part of a much wider neoliberal hegemonic project. This neoliberal project has involved not only the expansion of markets but also the state s active cultivation and enrollment of new and ever expanding parts of civil society, which has been transformed in the process (Davies 2011, 151). Yet, this ethos and the broader neoliberal ideology of which it is a part continue to confront intractable barriers as a result of the trend towards political and economic centralization characteristic of the crisis-prone nature of capitalist modernity (Davies 2011, 5). This sets up a fundamental incompatibility between the connectionist ethos on the one hand and neoliberal political economy on the other. This contradiction is expressed in a number of ways but Davies (2011, chapter three) highlights four. First, Davies argues that the connectionist ethos has misread both the past and the present and overexaggerated the extent to which broader economic, social and political changes have undermined the state, capitalism and class (cf. Marsh 2011). As such, he argues that: The target of [my] critique is the proposition that network-like institutions and practices are proliferating, that they are based on novel forms of sociability and that they transcend structures of power and domination. Networks can be a powerful organising tool, but whether cooperative or insurgent, they have no special potential (Davies 2011, 152). Once we accept that networking is as routine and unexceptional as command then Davies argues that the most interesting change that we have witnessed over recent times is actually the rise of network governance ideology and its embodiment in everyday practice, rather than any particularly fundamental change in the underlying nature of capitalist relations. As such, Davies argues that: what is different about the past 40 years or so is not so much de-traditionalisation, dispersion and the risk of the network society, as the celebration of networks spearheading neoliberalism. Based on this argument, Davies goes onto use Gramsci s concept of the integral state to challenge the argument that we are moving from a logic of structures to a logic of flows in which network governance trumps modernist tendencies and networked power relations trump hierarchical power relations (cf. Joseph 2010). Second, Davies argues that governance networks are prone to resolve into hierarchy and incremental closure. This leads to a creeping managerialism and the governmentalisation of networks: As long as hegemony is partial and precarious then hierarchy can never completely retreat to the shadows. This dialectic plays out in the day-to-day politics of governance networks through the clash between connectionist ideology and roll-forward hierarchy or governmentalisation (Davies 2011, 5). Third, Davies argues that governance networks actually reproduce inequalities, rather than promote higher levels of democratic inclusion and equality. The argument that traditional structures and conflicts are being superseded by a reflexively modern condition is therefore a false one (Davies 2011, 131). Finally, and following on from

the earlier points, distrust is therefore a common feature in governance networks (Davies 2011, chapter 3). Thus, whilst the connectionist ethos may try to downplay underlying value conflicts, there is no reason to believe that these would be any less likely to be present now than at any other time before. In short, Davies (2011, 84) argues that his theory of governance as hegemony: explains how the ideologies of neoliberalism confront the structural limits of neoliberal capitalism, limits which manifest in the day-to-day politics of governance networks and make authentic network governance utopian. Aspects of these arguments can also be found within the SRA. The SRA is important in this context because it provides a second broadly neo-marxist response to the governance narrative (Jessop 1990, 2007). Perhaps the key concept in the SRA is the notion of strategic selectivity. This is used to explain how all structures state forms, accumulation regimes and hegemonic projects are the outcome of past strategic struggles (Jessop 1990). 6 The capitalist state, understood as a contingent institutional ensemble, achieves a partial unity through a relative convergence between these structures. This process occurs within a strategically selective terrain, which subsequently acts as the context within which current and future struggles occur: Strategic selectivity refers to the structurally mediated bias which means that particular forms of state privilege some strategies over others, some time horizons over others, some coalitions possibilities over others. A given type of state, a given state form, a given form of regime, will be more accessible to some forces than others according to the strategies they adopt to gain state power. And it will be more suited to the pursuit of some types of economic and political strategy than others because of the mode of intervention and resources which characterise that system (Jessop 1990, 209). Elsewhere, Jessop (2004, 50) has argued that: This means that analysts must always look beyond the state to examine its embedding within a wider political system, its relationship to other institutional orders and functional systems, and to the lifeworld. In turn the attempted exercise of state power (or, better, state powers in the plural) will reflect the prevailing balance of forces as this is institutionally mediated through the state apparatus with its structurally inscribed strategic selectivity. It follows that the strategic selectivity of the state is not given in advance but is the result of the interplay between state priorities on the one hand and the ongoing sociopolitical contestation that takes place both within and beyond state institutions on the other (Jessop 1990; Ioris 2011, 127). The strategically selective context within which the state operates therefore helps to explain why certain interests and hegemonic projects can be privileged over others at the same time as the outcome of any struggle can never be determined or read-off from the strategically selective context within which it occurs. Jessop (2007, 225-33) helps to explain this apparent contradiction by developing the notion of contingent necessity as part of his broader support for a critical realist philosophy of the social sciences (Jessop 2005). As Kelly (2000, 219 emphasis in original) explains:

He [Jessop] argues that the structures that make up the social world are selfreproducing, but the causal mechanisms necessary for such reproduction come together in a contingent, non-necessary manner. An ontology of contingent necessity is, then, theoretically indeterminate events cannot be explained within the confines of a single theoretical system but determinate in analysis of the real world, where all events are caused. Jessop (2007, 229) expands on this point elsewhere by arguing that: Contingent necessity implies the infinite complexity of the real world, and the infinite complexity of the real world implies contingent necessity. This requires us to pursue complexity reduction (i.e. to engage in methodological simplificationism) as well as to adopt methodological relationalism. Methodologically, this means that social phenomenon can only be explained as part of an ongoing dual movement from the abstract to the concrete on the one hand and from the simple to the complex on the other (Jessop 2007, 231; Sayer 2000). This is what Jessop has referred to as the method of articulation. This set of philosophical arguments also helps to explain why there can only ever be: a fragile and temporary unity around certain state projects, which define the character and content of state action and non-action in specific historical periods (Heigl 2011, 131). As Ioris (2011, 127) argues: Against both the assumption that the state is a neutral organization promoting the common good (as implicit in the theory of environmental governance) and the limited discussion of the class-tendencies of state action (as still implicit in the work of many critical scholars), Jessop (2007) has argued that state power combines centralised and diffuse forms of authority in conformity with the fundamental features of political economy and profoundly embedded in social relations. Ioris makes an important point here by highlighting the key difference that also exists between the SRA as a neo-marxist approach and the more networked and neo- Weberian approaches discussed above. The state is thus conceptualized as a social relation that offers unequal opportunities to different social groups and: has a specific, differential impact on the ability of various political forces to pursue particular interests and strategies (Jessop 2005, 40). As such, processes of governance will always be characterized by social conflict as agents pursue their own political and accumulation strategies as part of their overall aim to articulate their interests in the institutional terrain and generalize their strategies (Heigl 2011, 131). The inscribed asymmetries of state action, the differences in capacity that exist between different social forces and the extent to which the interests and ideas held by different social forces are aligned with the state s selective functioning will lead to: heterogeneous patterns of government and/or governance, with patterns varying with the objects of state intervention, the nature of policy fields, the changing balance of forces in and beyond the state, and so on (Jessop, 2004, 63; Ioris 2011, 127). It also means that the state, its institutions and the agents that work for it are strategic (not passive) actors who operate within a strategically selective context that facilitates and constrains how they respond to the trends and counter-trends associated with governance.

This brief discussion of some of the SRA is important because it sets the ontological context within which Jessop and others have developed a series of responses to the governance narrative. As such, Jessop argues that the trends identified by governance theory are actually best understood as a series of trends and counter-trends that exist at different scales, including the global, regional and local. The exact way in which the changing relationship within and between these trends and counter-trends develops over time and between different scales depends upon the particularities of the strategic terrain within which they occur. The concept of metagovernance is deployed to help explain this process but in a different way to either of the two ways in which it has been used above. Here, metagovernance is used to refer to a counter-trend to the de-statisation of politics in which: Political authorities [at national and other levels] are involved in organising the self-organisation of partnerships, networks and governance regimes. They provide the ground rules for governance; ensure the compatibility of different governance mechanisms and regimes; deploy a relative monopoly of organisational intelligence and information with which to shape cognitive expectations; act as a court of appeal for disputes arising within and over governance; seek to rebalance power differentials (Jessop, 1997: 575; 2007, 210). 7 Metagovernance therefore refers to the way in which economic and political coordination is achieved through the judicious blending of hierarchy, markets and networks in particular governance arrangements: The forms of intervention associated with the state and statecraft are not confined to imperative coordination, that is, centralized planning or top-down intervention. Paraphrasing Gramsci (1971), who analysed the state apparatus in its inclusive sense as political society + civil society and saw state power as involving hegemony armoured by coercion, we could also describe the state apparatus as based on government + governance and as exercising governance in the shadow of hierarchy (Jessop, 2004, 52). Jessop further argues that this is an iterative process due to the ongoing tendency towards governance failure at both the individual level of each mode of governing and in the process of metagovernance itself. Indeed, this tendency towards failure increases the likelihood of hybrid governance forms as the limits of any one individual form of governance to secure continued patterns of organizational stability are reached. Metagovernance subsequently take place in the context of negotiated decision-making (Jessop 2001, 15; 1997, 575; 1999: 17). This means that the increased importance and relevance of heterarchic modes of governing, a wider and more complex society and increasingly multilevel systems of governance actually increases the state s overall importance simply because it still remains responsible for securing the conditions necessary for the institutional integration and social cohesion of society. Neo-Marxist approaches share a common concern with exploring the underlying conditions that lead to conflict within governance arrangements and base their

understanding of power and interests on a broader political economy of the state. Davies has argued that these issues are best explored from a Gramscian perspective in which the connectionist ideology and its pervasiveness within contemporary public administration is understood as one part of a much wider neoliberal hegemonic project. Jessop addresses similar concerns but appears to be more open to the argument that the broader trends identified by governance theory are having more of an impact on the state than Davies would accept. This is reflected in Jessop s development of the concept of metagovernance and his discussions around the scalar rearticulation of the state, although here it must also be acknowledged that there is a difference between Davies and Jessop in that Jessop s work has been mainly focused around the actual political practices, techniques and punitive actions of the state, than on the broader socio-political and class forces that drive such actions (Whitehead 2003, 8). Whitehead (2003, 9) subsequently concludes that metagovernance is probably best viewed as a useful corollary to strategic relational state theory in that it reflects a somewhat different emphasis from Jessop s earlier concerns. 8 Conclusion This paper started out by noting the growing popularity of governance theory and the declining popularity of state theory. Perhaps the declining interest around state theory is to be expected given that governance theory questioned the very entity to which state theory had accorded analytical priority. It was not so much that the state disappeared from discussions altogether, it was more that it was accorded an altogether secondary role when compared with the language of governance and networks. This makes governance theory s more recent discussion of the state an interesting development. In all three of the approaches discussed above there appears to be a common agreement that government and governance coexist but very different views as to why this is the case. The network management approach argues that the state s involvement is necessary to achieve effective and legitimate policy outcomes. Its prescriptions remain largely at the meso level but it is still underpinned by the assumption that networks consist of members who share similar interests and a common purpose (Kjaer 2011, 107). The state is therefore assigned a role as the key metagovenor but there is little critical engagement with the concept of the state itself. It is largely assumed that the state is a neutral player and, to some extent, it is argued that the achievement of effective and legitimate policy outcomes through the metagovernance of governance networks is said to depend on the state limiting itself to this role. The neo-weberian and neo-marxist approaches both question whether it is possible for the state to limit itself to this role. This can be partly explained by the more sceptical stance that the neo-weberian and neo-marxist approaches have taken towards the early governance literature but also because both approaches have adopted a more inclusive view of governance than the more networked approach discussed above. For example, neo-weberian approaches have argued that it is the state that has initiated the trends identified by governance theory, has implemented strategies to adapt to those trends and has continued to retain control over them by imposing a shadow of hierarchy. It has not weakened itself by forming alliances with non-state actors but has actually extended its reach into areas that were previously

deemed to be off limits. Neo-Marxist approaches generally agree that the state remains an important object of study but focus more on its political economy and the conflict, interests and power dynamics that this engenders. It does so by trying to position the political and economic changes associated with governance within the context of changing patterns of state power, strategy and intervention. In short, there has been much discussion about how governance arrangements are more likely to be characterized by hybrid arrangements going into the future (Börzel 2011). However, we really need to engage with macro as well as meso and micro level approaches in order to be able to adequately address the questions that such approaches raise. This is where the discipline of public policy and administration, as well as governance theory more narrowly defined, could benefit from a more direct engagement with the questions and issues raised by state theory. Bibliography Ansell, C. (2000), The Networked Polity: Regional Development in Western Europe, West European Politics, 13 (3), 279-91. Arts, B., Lagendijk, A. and Houtum, H. van (eds.) (2009), The Disoriented State: Shifts in Governmentality, Territoriality and Governance (Springer). Baker, K. and Stoker, G. (2012), Metagovernance and nuclear power in Europe, Journal of European Public Policy, 19 (7), 1026-51. Bang, H.P. (2008), Political Science in a Swing: Between democracy and good governance, Paper presented at the 58 th Political Studies Association Annual Conference, Swansea University, 1-3 April 2008. Bang, H. P. (2011), The politics of threats: late-modern politics in the shadow of neoliberalism, Critical Policy Studies, 5 (4), 434-48. Barrow, C. W. (1993), Critical Theories of the State: Marxist, Neo-Marxist, Post- Marxist (Madiscon: University of Wisconsin Press). Beck, U. (1992), Risk Society: Towards a New Modernity (Sage: London). Bell, S. and Hindmoor, A. (2009), Rethinking Governance (Cambridge: Cambridge University Press). Bevir, M. and Rhodes, R. A. W. (2010), The State as Cultural Practice (Oxford: Oxford University Press). Börzel, Tanja A. (2011), Networks: Reified Metaphor or Governance Panacea?, Public Administration, 89 (1), 49-63. Braithwaite, J. (2008), Regulatory Capitalism: How it works, Ideas for Making it Work (Cheltenham: Edward Elgar). Braithwaite, J., Coglianese, C. and Levi-Faur, D. (2007), Can regulation and governance make a difference?, Regulation and Governance, 1, 1-7. Castells, Manuel (2000), Materials for an exploratory theory of the network society, British Journal of Sociology, 51 (1), 5-24. Chhotray, V. and Stoker, G. (2009), Governance Theory and Practice: A Cross- Disciplinary Approach (Houndmills: Palgrave).