This on-line version is the pre-copyedited, preprint version. The published version can be found here:

‘Multi-level governance and multi-level meta-governance’, in I. Bache and M. Flinders, eds, Multi-Level Governance, Oxford: OUP, 49-74, 2004.


This chapter develops a strategic-relational approach to the European Union as a key point of intersection in the transformation of statehood. The argument unfolds in four steps. It first introduces the strategic-relational approach and its implications for the study of government and governance. This approach gives equal attention in principle to the structural and strategic features of the state as a social relation and also provides a basis to periodize state formation. Second, interest turns to competing theoretical accounts of the EU as an emerging state form or political regime. Thus I distinguish state- and governance-centric approaches, consider two variants of each, and offer three criticisms of each main approach. The two statist approaches are liberal intergovernmentalism and supranationalism (a superstate); and the two governance accounts comprise multilevel governance and the ‘network polity’ (or, sometimes, network state). A third section develops a strategic-relational account of the emergence, restructuring, and strategic reorientation of the European Union and the development of new forms of metagovernance such as the open method of coordination. A key element of my critique and reformulation of other analyses is a distinctive, ironic approach to market, state and governance failures and the correlative need for sophisticated forms of reflexive meta-steering of state development.

 A Strategic-Relational Approach to the State and Governance

The strategic-relational approach (hereafter SRA) is a general theoretical framework for addressing structure and strategy at various scales of social life from its micro-foundations to its most general macrostructural dynamics. Applied to the state (and, a fortiori, to its role in governance), it regards the state neither as a unitary political subject nor as a passive, instrumentalizable thing but as a complex social relation (cf. Poulantzas 1979). It treats the state as a relatively unified ensemble of socially embedded, socially regularized, and strategically selective institutions, organizations, social forces and activities organized around (or at least involved in) making collectively binding decisions for an imagined political community. While it certainly accepts that there are material and discursive lines of demarcation between the state qua institutional ensemble and other institutional orders and the lifeworld, the SRA also emphasizes the material interdependence of the state apparatus and state practices with other institutional orders and social practices. In this sense it is socially embedded. Moreover, as Tim Mitchell argues, ‘[t]he state should be addressed as an effect of detailed processes of spatial organization, temporal arrangement, functional specification, and supervision and surveillance, which create the appearance of a world fundamentally divided into state and society. The essence of modern politics is not policies formed on one side of this division being applied to or shaped by the other, but the producing and reproducing of this line of difference’ (1991: 95). These detailed processes also divide the globe fundamentally into different states and societies and thereby create a more or less complex inter-state system. The manner in which these divisions are drawn and reproduced has specific effects on the political process and the effectivity of state power. This is why the SRA emphasizes the heuristic concept of strategic selectivity. This refers to the ways in which the state considered as a social ensemble has a specific, differential impact on the ability of various political forces to pursue particular interests and strategies in specific spatio-temporal contexts through their access to and/or control over given state capacities – capacities that always depend for their effectiveness on links to forces and powers that exist and operate beyond the state’s formal boundaries. This means that analysts must 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 (or civil society). 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.

The SRA’s main methodological conclusion is that state power must be studied not only in terms of the state’s basic structure, institutional architecture, and specific organizational forms but also from the viewpoint of its strategic capacities both within the political system more generally and vis-à-vis the wider nexus of functional systems and the lifeworld. In this sense, the SRA is inconsistent with a purely state-centred approach insofar as the latter assumes that the state can be examined in isolation and treated as an independent variable. In contrast the SRA examines the state in terms of its structural coupling and co-evolution with a wider set of institutions and social practices. Putting the state in its place in this way does not exclude (and, indeed, presupposes) specifically state-engendered and state-mediated processes; but it does require that these be considered in their broader social context and that their effects are related to the strategic choices and conduct of particular actors within and beyond the state (see Poulantzas 1979, Jessop 1990).

It follows that to talk of state managers, let alone of the state itself, exercising power is at best to perpetrate a convenient fiction that masks a far more complex set of social relations that extend far beyond the state apparatus and its distinctive capacities. Interestingly, this is reflected in the practices and discourses of state managers themselves. For, whilst they sometimes proudly claim the credit for having initiated and carried through a general strategic line or a specific policy, at other times they happily seek to offload responsibility for state actions and/or outcomes to other social forces (or to force majeure) at one or more points elsewhere in the ongoing struggle over power. While the constitutionalization and centralization of state power enable responsibility to be formally attributed to named officials and bodies, this should not lead us to fetishize the fixing of formal political responsibility at specific points and/or in specific personages. We should always seek to trace the circulation of power through wider and more complex sets of social relations both within and beyond the state. This is especially important where the growing complexity and mass mediatization of the exercise of state power lead to a search for charismatic leaders who can simplify political realities and promise to resolve them. For charisma actually serves to hide complex, if not chaotic, behind-the-scenes practices which would be hard to explain or defend in public (Grande 2000). Attention to the circulation of power also matters where state power is undergoing major transformations and losing some of its formal constitutional and hierarchical aspects – as is the case with recent developments in the emerging Europolity as part of a more general reorganization of the inter-state system (see below).

Studying the state from a strategic-relational perspective requires attention to at least six interrelated dimensions. Three dimensions primarily concern formal institutional aspects of the state regarded as a social relation. They are modes of political representation and their articulation; the internal articulation of the state apparatus; and modes of intervention and their articulation. Each of these dimensions has its own structurally-inscribed strategic selectivities and, while analytically distinct, all three typically overlap empirically. They can be studied at different levels of abstraction and complexity, ranging from the most basic state forms through to specific regimes in particular conjunctures. The other three dimensions concern the discursive and action-oriented aspects of the state qua social relation and give the first three dimensions their strategic meaning. These aspects give content to the more formal features of the state and it is the contest among social forces over their definition, articulation, and implementation that mediates structural and strategic changes in the state in given conjunctures. These three aspects comprise, first, the political projects articulated by different social forces that are represented within the state system, seek such representation, or contest its current forms, functions and activities; second, the prevailing state project with its raison d’état – or governmental rationality – and statecraft that seeks to impose an always relative unity on the various activities of different branches, departments and scales of the state system and that also defines the boundaries between the state and its environment as a precondition of the ongoing attempts to build such an improbable internal unity; and, third, the hegemonic projects that seek to reconcile the particular and the universal by linking the nature and purposes of the state into a broader – but always selective – political, intellectual and moral vision of the public interest, the good society, the commonweal, or some analogous principle of societalization. These six aspects also provide one basis for periodization of state formation and transformation.

The forms of intervention associated with the state and statecraft are not confined to imperative coordination, i.e., 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’. Studies of governance treat it as a general phenomenon concerned with issues of strategic co-ordination rather than as a state-specific matter. In broad terms, governance refers to mechanisms and strategies of co-ordination adopted in the face of complex reciprocal interdependence among operationally autonomous actors, organizations, and functional systems. Thus governance occurs in all social fields and its students have examined a wide range of such mechanisms and strategies, including markets, clans, networks, alliances, partnerships, cartels, associations, and states. But governance is sometimes identified more narrowly with one specific mode of coordination: reflexive self-organization based on continuing dialogue and resource-sharing among independent actors to develop mutually beneficial joint projects and to manage the contradictions and dilemmas inevitably involved in such situations. In these terms governance can be contrasted with ex post coordination based on the formally rational pursuit of self-interest by individual agents; and with various forms of ex ante imperative coordination concerned with the pursuit of substantive goals established from above. This definition is the primary one adopted below in my discussion of multilevel governance (hereafter MLG). Such heterarchic mechanisms (as contrasted to market anarchy or bureaucratic hierarchy) have long been used in co-ordinating complex organizations and systems. They are especially suited for systems (non-political as well as political) that are resistant to top-down internal management and/or direct external control and that co-evolve with other (complex) sets of social relations with which their various decisions, operations, and aims are reciprocally interdependent.

State-Centric Perspectives

State-centred approaches tend to adopt, albeit implicitly more than explicitly, the ideal-typical late nineteenth-century sovereign national state as their reference point and examine the EU in one of two ways. Some commentators note the emergence of an increasingly important new supranational arena in which sovereign national states attempt to pursue their own national interests. This new arena is a site of intergovernmental (here, international) relations rather than a site to which important sovereign powers have been transferred and so, however important it has become for the joint pursuit of intergovernmental interests, it does not culminate in a new state form. This approach is often termed liberal intergovernmentalism (see especially Hoffman 1995; Moravcsik 1998). Other commentators identify a tendential, emergent, upward re-scaling of the traditional form of the sovereign state from the national to the supranational level. This is expected to culminate sooner or later in a new form of supranational statehood. They suggest that the associated re-allocation of formal decision-making powers is leading to a more or less complex form of multilevel government under the overall authority of a supranational superstate (see Pinder 1991; Taylor 1975; Weiler 1981). Whether the joint decision-making that characterizes this emerging superstate is a transitional feature or will remain once the superstate is consolidated is still uncertain.

For liberal intergovernmentalists, on the one hand, national states are, and will necessarily remain, the key players in the emerging European political space. States abandon little or none of their sovereign authority and retain a comprehensive constitutional mandate in contrast to the limited powers of the European Union. Thus inter-state interactions overwhelmingly take the form of international relations oriented to the pursuit of national interests, involving at best the provisional pooling of sovereignty for the pursuit of joint interests. For some this provides a new means to enhance the power and authority of the national state (e.g., Moravcsik 1998). More generally, rather than the leading to the transcendence of the national state, intergovernmental cooperation is said at most to produce a set of interlocking international arrangements among a self-selected group of national states. While this may eventually lead to a Staatenbund or confederation (e.g., a United Europe of National States), it could be blocked at any stage if one or more national states feel that their respective national interests would be hurt if the process continued.

Supranationalists, on the other hand, must posit a paradoxical transitional process in which national states conspire in their own transcendence (Aufhebung) as they promote supranational state formation. This involves a re-territorialization of political power as the three key features of the modern sovereign state are re-scaled upwards and re-differentiated vertically: Staatsgewalt (organized coercion), Staatsgebiet (a clearly demarcated territorial domain of state authority), and Staatsvolk (state subjects). This is linked to the re-scaling (and, perhaps, re-organization) of mechanisms for constitutionalizing and legitimating state authority in the expanded territory. Two factors distinguish the emergence of the supranational state (or superstate) from the simple territorial expansion of a single national state that absorbs all (or some) of the territories occupied by other relevant national states. First, it emerges from an agreement among independent national states to surrender their sovereignty and transfer it to a higher authority. Second, each of the affected national states becomes a subordinate unit of the new state whilst keeping the same territorial boundaries. Thus the new superstate is a multi-tiered state apparatus.

What do these two approaches imply for the analysis of multilevel political relations? First, in the case of the upward re-scaling (or re-territorialization) of state sovereignty, the development of multilevel government could be seen as a transitional effect of the transition. In other words, it would take the form of relations between an emergent, but still incompletely realised, supranational state and existing, not yet transcended, national states. Moreover, if the emergent, but still incomplete, supranational state were to assume the form of a bi- or multi-tiered federal superstate (Bundesstaat), there would also be scope for analysing the relations between the different tiers of government with the tools previously used for analyzing the dynamics of other federal states. Second, in the case of international relations, multilevel government could be interpreted in terms of distinctive features of the intergovernmental institutional arrangements established by national states and/or the specific governance strategies that they pursue from time to time. In terms of the language introduced by Collinge (1999) to analyse the relativization of scale, while the European Union level becomes an increasingly important nodal scale in the overall exercise of state power, national states continue to form the dominant scale. Given the complexities of state power in such circumstances, it might be more appropriate, then, to call this multilevel governance in the shadow of national government(s).

While liberal intergovernmentalism appears more persuasive than supranationalism, especially for the earlier stages of European economic and political integration, the statist approach as a whole errs on three main grounds: it adopts a restricted account of the state as a sovereign territorial apparatus, employs an anachronistic reference point, and is marred by its very state-centrism. First, although the essence of the state may well consist in the territorialization of political power, politicalpower can nonetheless be territorialized in different ways. Yet analyses of the EU as an emerging supranational state tend to focus on three features of the state apparatus: (a) its monopoly of organized coercion; (b) the constitutionalization of state power through the rule of law and a clear allocation of authority; and (c) control over its own money, taxes, and state budget. This implies that the most significant criteria for assessing whether a European superstate has emerged are the development of a European Kriegs- und Friedensgemeinschaft (a War- and Peace-Community, complete with a European army subject to supranational control, a European police force for internal security, and a European foreign and security policy oriented to the pursuit of distinctively European interests in the wider world system of states), an explicit European constitution (which locates sovereign power at the apex of a multi-tiered political system, defines the relationship between a jointly sovereign European executive, legislature, and judicial system, and determines the division of powers and competencies between the different tiers of government), and a European monetary system, fisco-financial system, and a large, centralized budget. Anti-federalists already claim that the EU has developed these features or, at least, will soon do so. Liberal intergovernmentalists note the absence of all or most of these same features and conclude that the EU is primarily an arena in which traditional national territorial sovereign states compete to influence European policies, politics, and political regimes. Despite these disagreements, however, both sides fetishize formal constitutional and juridical features and ignore de facto state capacities and the modalities of the exercise of state power. They also focus excessively on territoriality at the expense of extra-territorial and non-territorial features.

Second, state-centred theorists overlook the successive historical transformations of the modern territorial state forms from the mid-to-late nineteenth century onwards. This means that they adopt an anachronistic model of the national territorial state as their criterion for judging whether and how far a European superstate has emerged. This claim can be illustrated from Willke’s periodization of the modern state. He distinguishes four stages: the Sicherheitsstaat, which is concerned to defend its territorial integrity at home and abroad; the Rechtsstaat, which provides legal security for its subjects; the Sozialstaat, which establishes and extends welfare rights to its subjects; and the Risikostaat, which protects its citizens from a wide range of unexpected and uncontrollable risks. These stages are associated with the primacy of different state resources, namely, Gewalt (organized coercion), Recht (law grounded in a constitution), Geld (national money and state budgets), and Wissen (organized intelligence) (Willke 1992). Although I do not accept that the ‘risk state’ is the most useful concept for the contemporary state, Willke’s approach does highlight changes in the relative primacy of state resources. This suggests that the absence of a European army-police, constitution, and massive budgets may be less important than the presence of the EU’s ability to mobilize organized intelligence and other forms of soft intervention that shape how national and regional states deploy their respective capacities (cf. Sbragia 2001). Overall, this suggests, first, that the key resources in today’s Staatenwelt (world of states) – at least as far as relations among advanced bourgeois democratic states are concerned – are not so much coercion or money but soft law and intelligence; and, second, that the appropriate model for analyzing EU state building is not an idealized 19th century liberal state but the actually existing late 20th century state – whether this be a competition state, the regulatory state, or the Schumpeterian workfare post-national regime

A related aspect of this second problem is the adoption of anachronistic normative assumptions about European political democracy. We should compare the still emergent EU polity with actually existing national democracies rather than earlier democratic systems – whether nineteenth-century liberal nightwatchman states, interwar interventionist states, or postwar Keynesian welfare national states with catchall governing parties. Contemporary western states tend to towards authoritarian statism, with strong executives, mass-mediatized plebiscitary democracy, and authoritarian mass parties (cf. Poulantzas 1979). Thus, if there is a democratic deficit in the European Union, it may be linked to the contemporary form of statehood more generally, with deficits on different scales reinforcing each other. This in turn suggests that attempts to develop more democratic forms of representation and greater democratic accountability must be oriented to a different understanding of the nature and feasibility of democracy.

The third problem with state-centric analyses is precisely their state-centrism. In particular, they tend to naturalize the state-society distinction. Yet the boundary between state and society is socially constructed, internal to the political system, and liable to change. Thus, adequately to interpret changes in the EU as moments in the reorganization and reorientation of contemporary statehood, we must consider how the wider political system is organized and how changes in its territorial boundaries may contribute to the more general reorganization of state power (see below). The latter must also be related to the changing patterns of strategic selectivity linked to a changing institutional architecture and new forms of political mobilization. This implies in turn that the European Union is not a fixed form of state (apparatus) but an aspect, a path-shaping as well as path-dependent institutional materialization, of a new balance of forces that is expressed, inter alia, in state building.

Governance-Centric Approaches

Simple governance-centric approaches hold that the constitutionalized monopoly of violence and top-down modes of intervention associated with modern states are irrelevant or even harmful in an increasingly complex global social order. Thus they focus on the tendential de-statization of politics (or de-hierarchization of the state) rather than the de-nationalization of statehood; and they emphasize the enhanced role of reflexive self-organization in solving complex coordination problems that involve a wide range of partners or stakeholders beyond as well as within the state. This provides two bases on which to analytically distinguish de-centred forms of governance from the activities of centralized sovereign states. On the one hand, the sovereign state can be seen as the quintessential expression of hierarchy (imperative coordination) because it is, by definition, the political unit that governs but is not itself governed. Hence, beyond the sovereign state, we find the anarchy of interstate relations and/or the heterarchy of a self-organizing international society. And, on the other hand, it is primarily concerned with governing activities in its own territorial domain and defending its territorial integrity against other states. In contrast, governance is based on reflexive self-organization (networks, negotiation, negative coordination, positive concerted action) rather than imperative coordination. And it is concerned in the first instance with managing functional interdependencies, whatever their scope (and perhaps with variable geometries), rather than with activities occurring in a defined and delimited territory.

Adopting this approach leads to the view that the EU is a major emerging site of governance that involves a plurality of state and non-state actors on different levels who attempt to coordinate activities around a series of functional problems. Without reference to non-state as well as state actors and to functional as well as territorial issues, the multilevel governance approach would be hard to distinguish from intergovernmentalism. Thus the key question becomes how state and non-state actors manage, if at all, to organize their common interests across several territorial levels and/or across a range of functional domains. In this respect there are two main approaches: the self-described multilevel governance approach with its primary stress on the vertical dimension of multilevel governance and a parallel body of work that puts more emphasis on its horizontal dimension through the notion of the ‘network polity’ (sometimes referred to, less fortunately, as the ‘network state’).

In the present context, multilevel governance involves the institutionalization of reflexive self-organization among multiple stakeholders across several scales of state territorial organization. This has two implications. First, state actors would cooperate as negotiating partners in a complex network, pooling their sovereign authority and other distinctive capacities to help realize collectively agreed aims and objectives on behalf of the network as a whole. They would operate at best as primus inter pares in a complex and heterogeneous network rather than as immediate holders of sovereign authority in a single hierarchical command structure. Thus the formal sovereignty of states is better seen as one symbolic and/or material resource among others rather than as the dominant resource. Indeed, from a multilevel governance perspective, sovereignty is better interpreted as a series of specific state capacities (e.g., legislative, fiscal, coercive, or other state powers) rather than as one overarching and defining feature of the state. Thus states will supply other resources, too, that are not directly tied to their sovereign control over a national territory with its monopoly of organized coercion, its control over the national money, and its monopoly over taxation (Krätke 1984; Willke 1992). State involvement would therefore become less hierarchical, less centralized, and less directive in character. Other stakeholders in turn contribute other symbolic and/or material resources (e.g., private money, legitimacy, information, expertise, organizational capacities, or power of numbers) to advance collectively agreed aims and objectives. Second, in contrast to the clear hierarchy of territorial powers associated in theory with the sovereign state, multilevel governance typically involves tangled hierarchies and complex interdependence. Thus the EU functions less as a re-scaled, supranational sovereign state apparatus than as a nodal point in an extensive and tangled web of governance operations concerned to orchestrate economic and social policy in and across many different scales of action with the participation of a wide range of official, quasi-official, private economic interests, and representatives of civil society.

The network polity (or state) provides a complementary account of the nature of the European state political system. Three variants can be noted: Castells’ ambiguous claims about the European network state, a Foucauldian view that interprets recent patterns of European governance as a shift to an advanced (neo-)liberal form of governmentality, and governance-theoretical accounts of the network polity. The third variant is the most widespread but I will comment briefly on each.

Castells has gained attention in some circles for his recent extension of the overworked ‘network’ metaphor from the economy and society to the state. But his brief account of European governance is rendered ambiguous by his confused and confusing attachment to the concept of ‘network state’. Castells claims that the EU is organized essentially as a network that pools and shares sovereignty. As such, rather than involving the transfer of authority up to a European state that thereby supplants existing European nation-states, the European Union as a whole tends to operate as a network state. He defines the latter as a state that shares authority (that is, in the last resort, the capacity to impose legitimized violence) along a network. Here Castells retains the conventional Weberian notion of the state as an apparatus possessing a legitimate monopoly of violence over a given territorial area, implying that the authority of the European Union corresponds to a specific territorial domain. But he adds that, by definition, a network has nodes, not a centre, thereby implying that control over this monopoly is more or less dispersed rather than centralized. This is particularly likely because nodes may vary in size and be connected to the network by asymmetrical ties. Indeed, the nodes formed by member states do differ in their respective powers and capacities; and even the three largest member states have different strengths (technological, industrial, financial, military). This dispersion of authority and influence among nodes in the European ‘network state’ is reflected in the complex, variable, and changing geometry of European institutions. Thus Castells notes that the EU combines control of decision-making by national governments (the European Council, its rotating presidency, and regular meetings of the Council of Ministers); management of common European business by a euro-technocracy, directed by the politically-appointed European Commission; and symbolic legitimacy vested in the European Parliament, the Court of Justice and the Court of Auditors. The endless negotiations in this set of institutions, and among the national actors pursing their strategies, may, argues Castells, look cumbersome and inefficient. But it is exactly this indeterminacy and complexity that enable the EU to muddle through, accommodating various interests and changing policies, not only from different countries, but from the different political orientations of parties elected to government (2000: 2, 5). From this brief summary it can be seen that Castells’ account is still strongly imprinted by state-theoretical assumptions and that his major innovation relative to their usual expression in liberal intergovernmentalism and/or supranationalism is to use the idea of ‘network’ to reveal some (but far from all) of the complexity of the linkages between national states and European institutions.

A useful overview of the Foucauldian approach to the changing ‘arts of international government’ (distinguishing between imperialism, developmentalism, and the ‘new regionalism’) is presented in Larner and Wallace (2002). Whilst referring in general terms to the EU as the most advanced form of the ‘new regionalism’ considered as a form of liberal governmentality, they enumerate several key distinguishing features of this new art of international governmentality drawing on developments in North America and the Asian Pacific as well as Europe. Haahr (2002) provides a parallel Foucauldian study focused primarily on the emerging ‘open method of coordination’ (OMC) that was formally announced during the 1999 Portuguese Presidency. The OMC is a ‘soft’ form of governance that differs from traditional top-down ‘positive government’ and the previous trend towards a European ‘regulatory state’ based on a neo-liberal Ordnungspolitik (on which, see Majone 1997). As such, it involves concerted, centralized formulation of objectives, quantification of indicators for measuring progress towards such objectives, decentralized implementation, and systematic monitoring of different member states’ progress. It thereby enables member states to address problems at the European level without ceding new juridical competencies to the European Union (Haahr 2002: 7-8). In this sense, argues Haahr, the OMC reflects advanced liberal forms of government. These embody a notion of structured and conditioned freedom: they are ‘practices of liberty’ that establish and facilitate liberty but also discipline it and constrain its exercise. They govern through the manipulation of techniques and mechanisms rather than more directly through a classical liberal and/or Keynesian welfarist manipulation of processes. Thus, while these new techniques of rule involve contracts, consultation, negotiation, partnerships, empowerment, and activation, they also set norms, standards, benchmarks, performance indicators, quality controls and best practice standards. In short, ‘[a]dvanced liberal rule operates through our freedom, through the way this freedom is structured, shaped, predicted and made calculable’ (Haahr 2002: 9). Re-scaled from national states to the European level, this is seen in the declared role of the OMC to help member states evolve their own policies in new and major areas in line with the constitutional principle of subsidiarity and in consultation with relevant regional and local political authorities as well as with the social partners and civil society. By developing strategies, setting and monitoring targets, and forming partnerships, such ‘advanced liberal’ forms of governmentality can both mobilize and discipline the energies of civil society. In this way, while member states appear as agents capable of devising strategies and achieving objectives without being directly subject to EU diktat, the European Commission can appear as the institution empowered to assess their relative performance in attaining the consensually-determined agreed objectives (Haahr 2002: 10-14).

More conventional governance-theoretical analyses of the emerging European network polity start out from the difficulties of relying on rigid hierarchical coordination in contexts characterized by complex reciprocal interdependence among different fields across different scales (Ladeur 1997; Pitschas 1995). Ansell provides a good overview of this approach and summarizes his (and other) findings as follows:

… the networked polity is a structure of governance in which both state and societal organization is vertically and horizontally disaggregated (as in pluralism) but linked together by cooperative exchange (as in pluralism). Organizational structures in the networked polity are organic rather than mechanistic, which means that both knowledge and initiative are decentralized and widely distributed. Horizontal relationships within and across organizations are at least as important as vertical relationships, and organizational relationships in general follow a pattern of many-to-many (heterarchy) rather than many-to-one (hierarchy). Exchange is diffuse and/or social rather [than] discrete and/or impersonal. The logic of governance emphasizes the bringing together of unique configurations of actors around specific projects oriented toward integrative solutions rather than dedicated programs. These project teams will criss-cross organizational turf and the boundary between public and private. State actors with a high degree of centrality in the web of interorganizational linkages will be in a position to provide facilitative leadership in constructing or steering these project teams’ (Ansell 2000: 311).

Three main criticisms can be levelled at the main governance-centred approaches, excluding for present purposes Castells and the Foucauldians (I will return to Haahr’s analysis in my comments on ‘multilevel metagovernance’). First, reflecting its different disciplinary roots and wide range of applications, work on governance often largely remains at a pre-theoretical stage: it is much clearer what the notion of governance excludes than what it contains. This is reflected in a proliferation of typologies of governance mechanisms constructed for different purposes and a large measure of (often unspoken) disagreement about what is included, what excluded, from the overall concept. Thus many early analyses served to establish that the EU political system cannot easily be assimilated to, and studied in terms of, a traditional conception of government; but it was unclear exactly how multilevel governance operates to produce the European polity, how objects of governance are defined in this context, and how stakeholders are defined. Later work has begun to address these problems but often does so for specific policy areas or policy networks, leaving open the issue of how different multilevel governance regimes are connected, let alone how, if at all, they may acquire a relative unity. Related to this comparative underdevelopment of the governance concept are marked ambiguities in the referents of multilevel governance. For the term is used to capture several trends in the development of the contemporary state – the de-nationalization of statehood, the de-statization of politics, and the re-articulation of territorial and functional powers. The fact that it is used to describe the interaction of three analytically distinct trends (each with its own countertrend) or, at least, to characterize their combined impact, suggests that the concept may obscure as much as it clarifies about recent changes.

Second, governance theories tend to be closely connected to concerns about problem-solving and crisis-management in a wide range of fields. This has led some governance theorists to focus on specific collective decision-making or goal-attainment issues in relation to specific (socially and discursively constituted) problems and to investigate how governance contributes to problem-solving (for a belated self-criticism on this score, see Mayntz 2001). But this can easily lead to a neglect of problems of governance failure, i.e., the tendency for governance to fail to achieve its declared objectives; and, a fortiori, neglect of the various responses of different agents or subjects of governance to such failures as they attempt to engage in different forms of metagovernance (on governance failure and metagovernance, see Jessop 2000a, 2002b). Two aspects of metagovernance are relevant here. On the one hand, because many studies of governance are concerned with specific problem fields or objects of governance, they tend to ignore questions of the relative compatibility or incompatibility of different governance regimes and their implications for the overall unity of the European project and European statehood. And, on the other hand, many empirical studies have overlooked (or, at least failed to theorize) the existence of meta-steering. This complicated process, which Dunsire (1996) has aptly termed ‘collibration’, involves attempts to modify the relative weight and targets of exchange, hierarchy, and networking in the overall coordination of relations of complex interdependence. Yet such meta-steering is central to many of the disputes over European integration and/or state formation and has long been a key issue on the agenda of the European Union itself, especially regarding the different steps in integration. This is reflected in the increasing resort to partnerships, comitology, social dialogue, and the mobilization of non-governmental organizations and social movements as additional elements in the attempts to guide European integration and to steer European Union policy-making and implementation (cf. Scott and Trubek 2002). The ‘Lisbon Strategy’, with its advocacy of the extension of the ‘open method of co-ordination’, and the recent White Paper on Governance are the latest phases in this search for appropriate mechanisms of metagovernance (see below).

Third, work on multilevel governance and the network polity poses fundamental issues about the extent to which a network polity will remain tightly anchored in territorial terms (as opposed to being necessarily territorially embedded) despite its highly pluralistic functional concerns and its equally variable geometries. Schmitter raised just such issues in another context, when he identified four possible, ideal-typical future scenarios for the emerging Europolity. These scenarios were generated in true sociological fashion through the formation of a two-by-two property space based on two dichotomized dimensions of political regime formation: (1) an essentially Westphalian versus ‘neo-medieval’ form of territorial organization; and (2) heterogeneous and flexible versus tightly ordered and highly stable functional representation. The most interesting (and, he suggested, more plausible) scenarios both involved flexibility on the second dimension. They are a condominio (a neo-medieval state system with flexible functional representation and policy-making) and a consortio (a largely intergovernmental Europe des patries with polycentric, incongruent flexible functional representation and policy-making). A Westphalian state re-scaled to the European level with a well-ordered and congruent European system of functional representation (to produce a stato, which could be consideredequivalent to a supranational European superstate) was deemed unlikely; and a confederatio (a neo-medieval territorial arrangement with tightly organized and stable functional representation) was judged even more implausible (Schmitter 1992).

As an open-ended thought experiment only loosely linked to empirical analysis, Schmitter’s typological exercise is not directly relevant to my main objectives in this chapter. But it does serve two purposes in the present context. For it suggests, first, that studies of multilevel governance and/or network forms of political organization should not ignore issues of territorial organization; and, second, once both sets of issues are posed together, issues of multilevel metagovernance become central both in practice and in theory. To these twin conclusions we might add a third, namely, the need to question how far analyses of the political actors in the European Union can be confined territorially to its member states (and, perhaps, candidate states) and functionally to organized interests and movements that are anchored primarily within the political space directly organized and controlled by the European Union and its member states. For the forms, pace, and extent of European integration are also relevant to other states (notably the USA and USSR) and to a wide range of non-state forces with strong roots outside the European Union.

Each of these problems has severe implications for an adequate analysis of multilevel governance. In particular, they have produced a situation described by Weiler and Wessels as comprising ‘too many case studies, ad hoc lessons from limited experiences and organizational description [and] too little theoretical mediation’ (1998: 230n). In part, of course, this ad hoccery reflects the real complexities of the emerging European polity. Indeed, it would be surprising if no such complexities existed. For national states also involve 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. In this sense, perhaps, we may be witnessing a re-scaling of the complexities of government and governance rather than a re-scaling of the sovereign state or the emergence of just one more arena in which national states pursue national interests. It is to these complexities at the national scale that I now turn in order to provide some insights into how we might re-think the emerging EU polity.

Changes in Statehood in Advanced Capitalist Societies

Here I advance three inter-related propositions about recent trends in national states to provide just one possible approach to the ‘theoretical mediation’ called for by Weiler and Wessels. These three trends are derived from theoretically-informed observation of developments in developed capitalist economies in all triad regions rather than Europe alone. In this sense, their generality across these regions (and, hence, their occurrence elsewhere) suggests they are not generated by processes peculiar to the European region. For this reason they can help to contextualize and interpret recent trends in the development of European statehood (cf. Ziltener 1999). But it must also be recognized that these trends have many different causes. They should not be treated as singular causal mechanisms in their own right, which would mean neglecting their essentially descriptive, synthetic, and generalized nature. Nor should they be thought to entail unidirectional movement or multilateral convergence across all national or triadic regimes. Instead, they can, and do, take different empirical forms (for example, on the differing dynamics of cross-border regions in North America and Europe, see Blatter 2001).

First, there is a general trend towards the de-nationalization of territorial statehood. This trend should not be mistaken, pace Shaw (2000), for the rise of a ‘global state’ – at least if the concept of state is to retain its core meaning of the territorialization of a centralized political authority so that a ‘global state’ amounts to a single ‘world state’. Instead it represents a re-articulation of different levels of the territorial organization of power within the global political system. As such it is reflected empirically in the ‘hollowing out’ of the national state apparatus with old and new state capacities being reorganized territorially on subnational, national, supranational, and translocal levels. State powers are moved upwards, downwards, and sideways as state managers on different scales attempt to enhance their respective operational autonomies and strategic capacities. One aspect of this is the gradual loss of the de jure sovereignty of national states in certain respects as rule- and/or decision-making powers are transferred upwards to supranational bodies and the resulting rules and decisions are held to bind national states. Another aspect is the devolution of authority to subordinate levels of territorial organization and the development of transnational but inter-local policy-making. The overall result is the proliferation of institutionalized scales of political decision-making, the increasing complexity of inter-scalar articulation, and a bewildering variety of transnational relations.

Countering this trend is the enhanced role of national states in managing inter-scalar relations. That is, national states seek to control what powers or competencies go up, down, or sideways and to exercise this control so as to enhance their capacities to realize their current state projects; and they also seek, as far as possible, to retain the competence to revoke such transfers of powers and/or to implement them in ways that do least damage to their capacity to secure institutional integration and social cohesion with their corresponding territories. In this sense, even if state powers and competencies are no longer exercised in the framework of the national state qua ‘power container’, the advanced states still retain considerable autonomy in regard to how to organize and re-scale state powers to promote state projects. The key question then becomes how far the movement of competencies or powers away from the national state is irreversible either constitutionally (such that attempted repatriation of powers would be subject to legal sanction or a constitutional use of armed force) or informally (such that attempts to repatriate powers would be regarded as politically illegitimate and/or economically infeasible). This said, it is generally easier for national states to reclaim powers and competencies devolved downwards or sideways than those that are shifted upwards (but see below).

Second, there is a trend towards the de-statization of the political system. This involves a shift from government to governance on various territorial scales and across various functional domains. There is a movement from the central role of the official state apparatus in securing state-sponsored projects and political hegemony towards an emphasis on partnerships between governmental, para-governmental, and non-governmental organizations in which the state apparatus is often only first among equals. Governance involves the complex art of steering multiple agencies, institutions, and systems that are both operationally autonomous from one another and structurally coupled through various forms of reciprocal interdependence. Governments have always relied on other agencies to assist them in realizing state objectives or projecting state power beyond the formally-defined state apparatus. But this familiar reliance on government and governance has been increased and re-ordered. This is reflected in the proliferation of public-private partnerships in various guises to supplement and sometimes to replace more traditional forms of corporatism and concertation. The relative weight of governance has also increased on all levels – including horizontal cross-border transactions as well as coordination across different vertical scales from the local through the triads to the global level. Nonetheless this need not mean the loss of state capacity. Indeed a shift to governance can enhance the capacity to project state power and achieve state objectives by mobilizing knowledge and power resources from influential non-governmental partners or stakeholders.

Countering this shift is government’s increased role in metagovernance. For political authorities (on and across all levels) are becoming more involved in all aspects of metagovernance: they get involved in redesigning markets, in constitutional change and the juridical re-regulation of organizational forms and objectives, in organizing the conditions for self-organization, and, most importantly, in the overall process of collibration. In this last respect, they provide the ground rules for governance and the regulatory order in and through which governance partners can pursue their aims; ensure the compatibility or coherence of different governance mechanisms and regimes; act as the primary organizer of the dialogue among policy communities; deploy a relative monopoly of organizational intelligence and information with which to shape cognitive expectations; serve as a ‘court of appeal’ for disputes arising within and over governance; seek to re-balance power differentials by strengthening weaker forces or systems in the interests of system integration and/or social cohesion; try to modify the self-understanding of identities, strategic capacities, and interests of individual and collective actors in different strategic contexts and hence alter their implications for preferred strategies and tactics; and also assume political responsibility in the event of governance failure. These emerging metagovernance roles mean that different forms of coordination (markets, hierarchies, networks, and solidarities) and the different forms of self-organization characteristic of governance take place ‘in the shadow of hierarchy’ (cf. Scharpf 1994: 40).

Third, there is a complex trend towards the internationalization of policy regimes. The international context of domestic state action has extended to include a widening range of extra-territorial or transnational factors and processes; it has become more significant strategically for domestic policy; the key players in policy regimes have also expanded to include foreign agents and institutions as sources of policy ideas, policy design, and implementation; and there is an increasing number of increasingly influential international regimes across a growing range of policy fields. This trend is reflected in economic and social policies as states become more concerned with ‘international competitiveness’ in the widest sense and with the transnational constraints, consequences, and conditions of state action. It is also reflected in the development of global public policy networks and increasingly ambitious plans for the harmonization (not standardization) of policy regimes across many policy fields. Somewhat ambiguously countering yet reinforcing this trend is a growing ‘interiorization’ of international constraints as the latter become integrated into the policy paradigms and cognitive models of domestic policy-makers. And, more clearly representing a counter-trend, we find that states (on all levels) are now increasingly active in attempting to shape the nature of international regimes, their interrelations, and their local implementation. Currently, it is again the national state that generally has the key role in these struggles to shape international regimes and to shape their local implementation, using whatever arenas are available to this end. This is because the national state is generally the only political organization that is empowered to bind its citizens to international commitments. The European Union is an increasingly important exception to this rule, however, differing significantly from the supranational organization of the other triad regions in this regard.

In short, these three changes do not exclude a continuing and central political role for national states. But it is a role that is necessarily redefined as a result of the more general re-articulation of the local, regional, national, and supranational levels of economic and political organization. Unless or until supranational political organization acquires not only governmental powers but also some measure of popular-democratic legitimacy, the national state will remain a key political factor as the highest instance of bourgeois democratic political accountability. How it plays this role will depend on the changing institutional matrix and shifts in the balance of forces as globalization, triadization, regionalization, and resurgent local governance proceed apace. Possibly the most important role for the national state here is in metagovernance, i.e., coordinating different forms of governance and ensuring a minimal coherence among them. In this sense Shaw (2000) is right to claim that the national state core to governance will not go away. But this will be less governmental and more oriented to issues of metagovernance. But there is no point at which a final metagovernance instance can be established to coordinate the myriad subordinate forms of governance – this would re-introduce the principle of sovereignty or hierarchy that growing social complexity and globalization now rule out.

I have introduced these three trends for one major reason. If the national state can no longer be understood in terms of the received notion of the sovereign national state, then perhaps this notion is also inadequate for studying the evolving EU as a state form. Indeed, we can go further: if the national state is changing in the ways that I have suggested, then the future position and activities of the EU within a re-territorialized, de-statized, and internationalized Staatenwelt must be very carefully reconsidered. What we are witnessing is the re-scaling of the complexities of government and governance rather than the re-scaling of the sovereign state or the emergence of just one more arena in which national states pursue national interests.

Much the same point can be made through changes in the state’s form and functions regarding capital accumulation. These can be studied along four key dimensions. The first concerns the state’s distinctive roles in securing conditions for profitable private business. This is the broad field of economic policy and matters because market forces alone cannot secure these conditions. The second dimension refers to the state’s distinctive roles in reproducing labour power individually and collectively over various timespans from quotidian routines via individual lifecycles to intergenerational reproduction. This is the broad field of social policy and matters because market forces and civil society alone cannot fully secure these conditions in contemporary conditions. The third dimension refers to the main scale, if any, on which economic and social policies are decided – even if underpinned or implemented on other scales. This is important as economic and social policies are politically mediated and the scales of political organization may not coincide with those of economic and social life. The fourth concerns the relative weight of the mechanisms deployed in the effort to maintain private profitability and reproduce labour-power by compensating for market failures and inadequacies. Top-down state intervention is just one of these mechanisms; and, as is well known, states as well as markets can fail. This suggests the need for additional governance mechanisms and, a fortiori, for an active collibrating role for the state (see above).

Referring to these four dimensions, the postwar state in northwestern Europe can be described ideal-typically as a Keynesian welfare national state (or KWNS). First, the state was distinctively Keynesian insofar as it aimed to secure full employment in a relatively closed national economy and did so mainly through demand-side management and national infrastructural provision. Second, social policy had a distinctive welfare orientation insofar as it (a) instituted economic and social rights for all citizens so that they could share in growing prosperity (and contribute to high levels of demand) even if they were not employed in the high-wage, high-growth economic sectors; and (b) promoted forms of collective consumption favourable to the Fordist growth dynamic based on mass production and mass consumption. Third, the KWNS was national insofar as these economic and social policies were pursued within the historically specific (and socially constructed) matrix of a national economy, a national state, and a society composed of national citizens. Within this matrix, the national territorial state was mainly responsible for developing and guiding Keynesian welfare policies. Local and regional states acted mainly as relays for these policies; and, in addition, the leading international regimes established after WW2 were mainly intended to restore stability to national economies and national states. And, fourth, the KWNS was statist insofar as state institutions (on different levels) were the chief supplement and corrective to market forces in a ‘mixed economy’ concerned with economic growth and social integration.

For reasons explored elsewhere (Jessop 2002b), the KWNS experienced a major and multiple crisis in the late 1970s and early 1980s. It has since been tendentially replaced by a new form of state with new functions on these four dimensions. I have termed this the Schumpeterian workfare post-national regime (or SWPR). It can be described in ideal-typical terms as follows. First, it is Schumpeterian insofar as it tries to promote permanent innovation and flexibility in relatively open economies by intervening on the supply-side and to strengthen as far as possible their overall competitiveness. Second, as a workfare regime, the SWPR subordinates social policy to the demands of labour market flexibility and employability and to the demands of economic competition. This includes putting downward pressure on the social wage qua cost of international production but, given the economic and political limits to welfare cuts, it is especially concerned with the re-functionalization of the inherited welfare state to serve economic interests. The state also attempts to create subjects to serve as partners in the innovative, knowledge-driven, entrepreneurial, flexible economy and its accompanying self-reliant, autonomous, empowered workfare regime. Third, the SWPR is ‘postnational‘ insofar as the national territory has become less important as an economic, political, and cultural ‘power container’. This is associated with a transfer of economic and social policy-making functions upwards, downwards, and sideways and with an increasing role for supranational non-governmental bodies and quangos in policy advocacy and policy transfer. And, fourth, given the growing recognition of state as well as market failure, the SWPR is associated, as we have seen, with a dual shift from government to governance and, just as importantly, from government to metagovernance.

This account of the three trends and countertrends in statehood and the distinction between the KWNS and SWPR also applies, of course, to the emerging Europolity. For the latter is an integral moment in the de-nationalization of the state, the de-statization of politics, and the internationalization of regimes – without being the highest level to which national state powers are shifted upwards, at which new forms of partnership are being organized, or on which the internationalization of policy regimes is occurring. Likewise, the changes in the dominant strategies used to build the Europolity, however ineffective, the forms it assumes, however impure, the functions that it exercises, however imperfectly, and the theoretical paradigms used to interpret it, however flawed, are all related to the periodization of the state in the advanced capitalist economies. Thus, to narrate rather breathlessly the phases of Europolity development detailed at length by Ziltener (1999), the initial steps towards Western European integration were initiated through a transatlantic coalition aiming to promote a postwar reconstruction that would integrate Western Europe into the economic and political circuits of Atlantic Fordism (cf. van der Pijl 1984); during the boom years of Atlantic Fordism, the ‘Monnet mode of integration’ was concerned to create a ‘Keynesian-corporatist’ (sic) form of statehood on the European level that could secure the conditions for different national fordist modes of development (cf. Ruigrok and van Tulder 1996) but developed crisis symptoms as the member states pursued divergent strategies in response to the interconnected economic and political crises in/of Fordism; the resulting crisis in European integration prompted the search for a new mode of integration and led to the internal market project – with important conflicts between neo-liberal, neo-corporatist, and neo-statist currents – and the development of new modes of economic and political coordination (cf. van Apeldoorn 2002); after a period of experimentation with new modes of coordination, the provisional outcome of these conflicts can be discerned in a new ‘Schumpeterian workfare’ mode of integration and coordination concerned to promote the structural competitiveness of Europe in a globalizing knowledge-based economy (cf. Telò 2002). Ziltener also suggests that the dominant line of conflict regarding Europolity construction during the Monnet period of Keynesian-corporatism was supranationalism versus intergovernmentalism and that the Delors project could be seen as a failed supranational attempt to rescale the Keynesian welfare state to the European level (1999: 129-30, 18-184). To this we might add that the shift towards a European SWPR is clearly associated with a shift in the dominant line of conflict around appropriate forms of multilevel governance and the ‘network polity’. This story is too brief, of course, to do justice to the complexities of European integration, the complexities of modes of coordination, their intended functions, and their variation with the objects of coordination, and the complexities of competing paradigms and paradigm shifts within and across different disciplines. But it does illustrate the need to periodize the Europolity and, indeed, to locate it in a broader context concerned with the reorganization of statehood on a still more global scale.

The European Union and Multilevel Metagovernance

I now apply the preceding arguments on metagovernance to the European Union as part of the more general change in the forms of statehood. There are at least two levels of failure – the failure of particular attempts at governance using a particular governance mechanism and the more general failure of a mode of governance. Thus, corresponding to the three basic modes of governance (or coordination) distinguished above, we can distinguish three basic modes of metagovernance and one umbrella mode. First, there is the reflexive redesign of individual markets and/or the reflexive reordering of relations among two or more markets by modifying their operation, nesting, articulation, embedding, disembedding, or re-embedding. There are also ‘markets in markets’. This can lead to ‘regime shopping’, competitive ‘races to the bottom’, or, in certain conditions, ‘races to the top’. Moreover, because markets function in the shadow of hierarchy and/or heterarchy, attempts are also made by non-market agents to modify markets, their institutional supports, and their agents to improve their efficiency and/or compensate for market failures and inadequacies. Second, there is the reflexive redesign of organizations, the creation of intermediating organizations, the reordering of inter-organizational relations, and the management of organizational ecologies (i.e., the organization of the conditions of organizational evolution in conditions where many organizations co-exist, compete, cooperate, and co-evolve). This is reflected in the continuing redesign, re-scaling, and adaptation of the state apparatus, sometimes more ruptural, sometimes more continuous, and the manner in which it is embedded within the wider political system. And, third, there is the reflexive organization of the conditions of self-organization through dialogue and deliberation. There are many activities involved here from organizing opportunities for spontaneous sociability through various measures to promote networking and negotiation to the facilitation of ‘institutional thickness’. Fourth, and finally, there is collibration or ‘metagovernance’ proper (see above). This involves managing the complexity, plurality, and tangled hierarchies found in prevailing modes of co-ordination. It is the organization of the conditions for governance and involves the judicious mixing of market, hierarchy, and networks to achieve the best possible outcomes from the viewpoint of those engaged in metagovernance. In this sense it also means the organization of the conditions of governance in terms of their structurally inscribed strategic selectivity, i.e., in terms of their asymmetrical privileging of some outcomes over others. Unfortunately, since every practice is prone to failure, metagovernance and collibration are also likely to fail. This implies that there is no Archimedean point from which governance or collibration can be guaranteed to succeed and that there will always be an element of irony in attempts to engage in collibration in the face of likely failure.

Governments play a major and increasing role in all aspects of metagovernance: they get involved in redesigning markets, in constitutional change and the juridical re-regulation of organizational forms and objectives, in organizing the conditions for self-organization, and, most importantly, in collibration. Thus metagovernance does not eliminate other modes of co-ordination. Markets, hierarchies, and heterarchies still exist; but they operate in a context of ‘negotiated decision-making’. On the one hand, market competition will be balanced by cooperation, the invisible hand will be combined with a visible handshake. On the other hand, the state is no longer the sovereign authority. It becomes just one participant among others in the pluralistic guidance system and contributes its own distinctive resources to the negotiation process. As the range of networks, partnerships, and other models of economic and political governance expand, official apparatuses remain at best first among equals.

It is in this context that we can best interpret the continuities and discontinuities in the development of the European Union as a moment in the structural transformation and strategic reorientation of statehood in a world of states that is not limited to Europe but extends to the global polity (cf. Hettne 1997; Shaw 2000; Sørensen 2001). For the European Union can be seen as a major and, indeed, increasingly important, supranational instance of multilevel metagovernance in relation to a wide range of complex and interrelated problems. While the sources and reach of these problems go well beyond the territorial space occupied by its member states, the EU is an important, if complex, point of intersection (or node) in the emerging, hypercomplex, and chaotic system of global governance (or, better, metagovernance) and is seeking to develop its own long-term ‘Grand Strategy’ for Europe (Telò 2002: 266). But it is still one node among several within this emerging system of global metagovernance and cannot be fully understood without taking account of its complex relations with other nodes located above, below, and transversal to the European Union.

It is clearly premature at a time when the European Union is conducting yet another debate on its future governance to predict the eventual shape of what is bound to be a complex and compromise-based form of multilevel metagovernance in the shadow of a post-national form of statehood. This underlines that the development of multilevel metagovernance is a reflexive process, involving intergovernmental conferences and other modes of meta-constitutional conversation (Walker 2000). But there can be little doubt that the overall movement is towards metagovernance rather than a rescaling of the traditional form of sovereign statehood or a revamped form of intergovernmentalism inherited from earlier rounds of European integration. As an institutionalized form of metagovernance, emphasis falls on efforts at collibration in an unstable equilibrium of compromise rather than on a systematic, consistent resort to one dominant method of coordination of complex interdependence. Apparent inconsistencies may be part of an overall self-organizing, self-adjusting practice of metagovernance within a complex division of government and governance powers. Seen as a form of metagovernance, the emphasis is on a combination of ‘super-vision’ and ‘supervision’, i.e., a relative monopoly of organized intelligence and overall monitoring of adherence to benchmarks. But in this evolving framework, there is also a synergetic division of metagovernance labour between the European Council, the specialized Councils, and the European Commission. The European Council is the political metagovernance network of prime ministers that decides on the overall political dynamic around economic and social objectives, providing a ‘centripetal orientation of subsidiarity’ (Telò 2002: 253), acting by qualified majority, and playing a key intergovernmental and monitoring role. The European Commission plays a key metagovernance role in organizing parallel power networks, providing expertise and recommendations, developing benchmarks, monitoring progress, exchanging best practice, promoting mutual learning, and ensuring continuity and coherence across Presidencies. This is associated with increasing networking across old and new policy fields at the European level as well as with a widening range of economic, political, and social forces that are being drawn into multilevel consultation, policy formulation, and policy implementation.

New methods of multilevel metagovernance are being developed and combined in a complex system of metagovernance (cf. Scott and Trubek 2002) that is ‘being made more precise and applied (with adaptations as for its intensity) to other fundamental policy fields, traditionally under the competence of national and sub-national authorities: education, structural reform and internal market, technological innovation and knowledge-based society, research and social protection’ (Telò 2002: 253). From a strategic-relational perspective, this clearly implies a shift in the strategic selectivities of the modes of governance and metagovernance in the European Union. For, while it builds on past patterns of liberal intergovernmentalism and neo-functionalist spillover, it has its own distinctive momentum and will weaken more hierarchical forms of coordination (whether intergovernmental or supranational). It also entails complementary changes in the strategic selectivities of national states and subordinate levels of government and governance, calling for new forms of strategic coordination and new forms (meta-)governance in and across a wide range of policy fields.

The pattern of multilevel metagovernance in the European Union is still evolving and, given the inherent tendencies towards failure typical of all major forms of governance (market, hierarchy, network, etc.) as well as metagovernance itself (Jessop 2002a, 2002b), continuing experimentation, improvisation, and adaptation is only to be expected. Nonetheless:

‘the perspective would be that of a new system of democratic legitimacy and governance: multilevel (international, national, supranational, transnational), multifaced (territorial, functional, modern and post-modern) and with a multitude of actors (social, economic, political and cultural; institutional and extra-institutional), rather than that of a classical democratic normative model – federal/constitutional or democrat/republican’ (Telò 2002: 266; cf. Schmitter 1992).

Thus the key issue for a research agenda into this new form of statehood becomes the manner and extent to which the multiplying levels, arenas, and regimes of politics, policy-making, and policy-implementation can be endowed with a certain apparatus and operational unity horizontally and vertically; and how this affects the overall operation of politics and the legitimacy of the new political arrangements

Concluding Remarks

I have argued that neither the state- nor the governance-centric perspective is adequate for analyzing the complexities of multilevel governance in Europe. Each approach is flawed theoretically in its own distinctive ways; nor can their respective deficits be overcome by combining them to produce a more coherent account. Each approach is also plagued to different degrees by anachronistic views about the contemporary world – whether about the state or about the objects and subjects of governance. This means that neither approach can capture the novelty of the emerging European polity as a ‘political machine’ for multilevel metagovernance (cf. Barry 2001). The alternative approach offered here draws on the strategic-relational approach and its application to the more general transformation of contemporary political economy as a means to contextualize and ‘theoretically mediate’ recent changes in statehood. This has two sets of implications for future research.

First, regarding the more immediate questions of governance and metagovernance, the SRA emphasises the strategic selectivity of institutional arrangements. Multilevel government, multilevel governance, and multilevel metagovernance arrangements will all have their own distinctive strategic selectivities, i.e., they are never neutral among actors, interests, spatio-temporal horizons, alliances, strategies, tactics, etc. They also have their own distinctive modalities of success, failure, tension, crisis, reflexivity, and crisis-management. These selectivities and modalities depend on specific institutional, organizational, and practical contexts and few generalizations are possible about them (for further discussion, including some possibly hubristic generalizations, see Jessop 2002b). Nonetheless, one generalization that can safely be hazarded is that the belief that multilevel governance can solve old problems without creating new ones is wishful thinking (cf. Mayntz 2001).  In turn this implies the need for an ironic, experimental approach to multilevel metagovernance that is concerned to ensure requisite variety in available modes of coordination as well as appropriate levels of reflexivity, super-vision, and supervision in their combination and implementation.

Second, and perhaps more importantly, it is only by situating the changing political forms of the European Union as part of the ongoing transformation and attempted reregulation of global capitalism as well as part of the more general transformation of statehood in response to major socio-cultural as well as politico-economic changes that one can adequately understand what is at stake in these changes. Among other interesting results, this approach reinforces the importance of examining not only multilevel governance but also multilevel metagovernance. For the development of the European Union can be seen as part of continuing efforts (often at cross-purposes) by key economic and political actors to produce an appropriate balance between different modes of economic and political coordination across functional and territorial divides and to ensure, under the primacy of the political, a measure of apparatus unity and political legitimacy for the European Union. This has taken different forms at different periods in the pursuit of the European project, especially as this has been shaped at different times by shifts in the relative weight of Atlanticist and European economic and political strategies, by shifts in the relative weight of liberal and neo-liberal échangiste (money capital) perspectives and neo-corporatist and neo-mercantilist productivist projects, and by the tendential shift from a KWNS approach concerned to create a single market to realize economies of scale to a SWPR approach concerned to transform the European Union into the most competitive and dynamic knowledge-based economy and to ‘modernize’ the European social model. In addition, of course, this European-wide multilevel metagovernance project is being conducted in conditions of successive rounds of expansion (which have increased the heterogeneity of the growth dynamics and modes of regulation of different regional and national economies as well as the forms and extent of uneven development and inequalities) and in conditions where national economies and national states have been subject to their own individual structural problems and crises as well as the shared crisis-tendencies derived from their integration into the circuits of Atlantic Fordism and into the emerging globalizing knowledge-based economy. And, finally, this multilevel metagovernance project is part of a broader post-Westphalian ‘meta-constitutional conversation’ that is occurring between non-state and state actors (including meta-states such as the EU) as they struggle to develop and institutionalize a new political order (Walker 2000). While it has not been possible within the limits of a short chapter to develop the analysis that this approach demands, I hope enough has been said to show the promise of the strategic-relational approach as one among several alternatives to be explored in future work on the European Union.


The arguments on governance in this chapter have benefitted from discussions at various times with Henrik Bang, Ulrich Beck, Frank Deppe, Edgar Grande, Liesbet Hooghe, Beate Kohler-Koch, Gary Marks, Markus Perkmann, Rod Rhodes, and Gerry Stoker. I also learnt much from the Sheffield conference. Given the idiosyncrasies of its approach to governance as well as many other matters, it is especially important that the usual disclaimers apply.


Ansell, C. (2000) ‘The Networked Polity: Regional Development in Western Europe’, Governance, 13 (2), 303-333.

Barry, A. (2001) Political Machines: Governing a Technological Society, London: Athlone.

Blatter, J. (2001) ‘Debordering the World of States. Towards a Multi-level System in Europe and a Multi-Polity System in North America? Insights from Border Regions’, European Journal of International Relations, 7 (2), 175-210.

Castells, M. (2000) End of Millennium, 2nd edition, Oxford: Blackwell.

Collinge, C. (1999) ‘Self-organization of Society by Scale: a Spatial Reworking of Regulation Theory’, Environment and Planning D: Society and Space, 17 (5), 557-574.

Dunsire, A. (1996) ‘Tipping the Balance: Autopoiesis and Governance’, Administration and Society, 28 (3), 299-334.

Duchacek, I.D., Latouche, D., Stevenson, G., eds (1988) Perforated sovereignties and international relations: trans-sovereign contacts of subnational governments, New York: Greenwood Press.

Gill, S. (2001) ‘Constitutionalising Capital: EMU and Disciplinary Neo-Liberalism’, in A. Bieling and A.D. Morton, eds, Social Forces in the Making of the New Europe, Basingstoke: Macmillan, 47-69.

Gramsci, A. (1971) Selections from the Prison Notebooks, London: Lawrence & Wishart.

Grande, E. (2000) ‘Charisma und Komplexität. Verhandlungsdemokratie, Medien-demokratie und der Funktionswandel politischer Eliten’, Leviathan, 28 (1), 122-41.

Grande, E. (2001) ‘Die neue Unregierbarkeit. Globalisierung und die Grenzen des Regierens jenseits des Nationalstaats’, Munich: Technische Universität München.

Haahr, J.H. (2002) ‘Open Methods of Coordination as Advanced Liberal Government’, Paper prepared for the ECPR 1st Pan-European Conference on European Union Politics, 26-28 September, Bordeaux.

Hettne, B. (1997) ‘A New Europe in the changing global system’, in R. Falk and T. Szentes, eds, A New Europe in the Changing Global System, Tokyo: United Nations University Press, 16-40.

Hoffman, S. (1995) The European Sisyphus. Essays on Europe 1964-1994, Boulder: Westview Press.

Jessop, B. (1990) State Theory: Putting the Capitalist State in its Place, Cambridge: Polity.

Jessop, B. (2002a) ‘Governance and Metagovernance in the Face of Complexity: On the Roles of Requisite Variety, Reflexive Observation, and Romantic Irony in Participatory Governance’, in P. Getimis and G. Kafkalas, eds, Participatory Governance and Multi-Level Governance, Opladen: Leske + Budrich, 33-58.

Jessop, B. (2002b) The Future of the Capitalist State, Cambridge: Polity.

Krätke, M. (1984) Kritik der Staatsfinanzen. Zur politischen Ökonomie des Steuerstaats, Hamburg: VSA Verlag.

Ladeur, K.H. (1997) ‘Towards a Legal Theory of Supranationality — the Viability of the Network Concept’, European Law Journal, 3 (3), 33-54.

Larner, W. and Wallace, W. (2002) ‘The Political Rationality of “New Regionalism”: Toward a Genealogy of the Region’, Theory and Society, 31 (3), 391-432.

Majone, G. (1997) ‘From the Positive to the Regulatory State: Causes and Consequences in the Mode of Governance’, Journal of Public Policy, 17 (2), 139-167.

Majone, G. (2002) ‘The New European Agencies: Regulation by Information?’, Journal of European Public Policy, 4 (2), 262-275.

Mayntz, R. (2001) ‘Zur Selektivität der steuerungstheoretischen Perspektive’, Working Paper 01/2, Köln: Max Planck Institut für Gesellschaftsforschung.

Mitchell, T. (1991) ‘The Limits of the State: Beyond Statist Approaches and their Critics’, American Political Science Review, 85 (1): 77-96.

Moravcsik, A. (1998) The Choice for Europe: Social Purpose and State Power from Rome to Maastricht, Ithaca: Cornell University Press.

Pinder, J. (1991) European Community: the Building of a Union, Oxford: Oxford University Press.

Pitschas, (1995) ‘Europäische Integration als Netzwerkkoordination komplexer Staatsaufgaben’, in T. Ellwein, D. Grimm, J.J. Hesse, and G.F. Schuppert, eds, Jahrbuch zur Staats- und Verwaltungswissehschaft. Band 8, Baden-Baden, Nomos Verlag, 379-416.

Polanyi, K. (1957) The Great Transformation, Boston: Beacon Press.

Poulantzas, N. (1979) State, Power, Socialism, London: Verso.

Ruigrok, W. and van Tulder, R. (1996) ‘The Price of Diversity: Rival Concepts of Control as a Barrier to an EU Industrial Strategy’, in P. Devine, Y. Katsoulacos and R. Sugden (eds), Competitiveness, Subsidiarity, and Industrial Policy, London: Routledge, 79-103.

Sbragia, A. (2000) ‘The European Union as Coxswain: Governance by Steering’, in J. Pierre, ed., Debating Governance: Authority, Steering, and Democracy, Oxford: Oxford University Press, 219-240.

Scharpf, F.W. (1994) ‘Games Real Actors could Play: Positive and Negative Coordination in Embedded Negotiations’, Journal of Theoretical Politics, 6 (1), 27-53.

Scharpf, F.W. (1998) Governing in Europe: effective and democratic? Oxford: Oxford University Press.

Schmitter, P.C. (1992) ‘Representation and the Future Euro-Polity’, Staatswissen-schaften und Staatspraxis, 3 (3), 379-405.

Scott, J. and Trubek, D.M. (2002) ‘Mind the Gap: Law and New Approaches to Governance in the European Union’, European Law Journal, 8 (1), 1-18.

Shaw, M. (2000) Theory of the Global State. Globality as an Unfinished Revolution, Cambridge: Cambridge University Press.

Sørensen, G. (2001) Changes in Statehood. The Transformation of International Relations, Basingstoke: Palgrave.

Taylor, P. (1975) ‘The Politics of the European Communities: the Confederal Phase’, World Politics, 27 (4), 336-360.

Telò, M. (2002) ‘Governance and Government in the European Union: the Open Method of Coordination’, in M.J. Rodrigues, ed., The New Knowledge Economy in Europe, Cheltenham: Edward Elgar, 242-272.

Tömmel, I. (1994) ‘Interessenartikulation und transnationale Politikkooperation im Rahmen der EU’, in Europäische Integration und verbändliche Regulierung, Marburg: Metropolis.

van Apeldoorn, B. (2002) Transnational Capitalism and the Struggle over European Integration, London: Routledge.

van der Pijl, K. (1984) The Making of an Atlantic Ruling Class, London: New Left Books.

Walker, N. (2000) ‘Flexibility within a Metaconstitutional Frame: Reflexions on the Future of Legal Authority’, in G. Búrca and J. Scott, eds, Constitutional Change in the EU: from Uniformity to Flexibility? Oxford: Hart, 9-30.

Weiler, J.H.H. (1981) ‘The Community System: the Dual Character of Supranationalism’, Yearbook of European Law, vol 1., 268-306.

Weiler, J.H.H. and Wessels, A. (1988) ‘EPC and the Challenge of Theory’, in A. Pijpers, E. Regelsberger, and W. Wessels, eds, European Political cooperation in the 1980s: a Common Foreign Policy for Western Europe? Dordrecht: Martinus Nijhoff, 229-258.

Wessels, W. (2000) Die Öffnung des Staates. Modelle und Wirklichkeit grenzüberschreitender Verwaltungspraxis 1960-1995, Opladen: Leske + Budrich.

Willke, H. (1992) Ironie des Staates, Frankfurt: Suhrkamp.

Ziltaner, P. (2001) Strukturwandel der europäischen Integration. Die Europäischer Union und die Veränderung von Staatlichkeit, Münster: Westfälisches Dampfboot.

Zürn, M. (2000) ‘Multilevel governance: on the State and Democracy in Europe’, in A. Mathias, L. Brock, and K.D. Wolf, eds, Civilizing World Politics. Society and Community beyond the State, Lanham, MD: Rowman & Littlefield, 148-168.

Multilevel Governance and Multilevel Metagovernance