It is fitting to reflect on UK devolution in the context of the current EU-wide trend towards greater policy autonomy being given to regions even within its more 'centralist' member states. Although the structure of regional governance in the EU falls significantly short of that which proponents of a 'Europe of the regions' might advocate, devolution in the UK nonetheless is consistent with a general shift in the direction of an emerging EU system of multi-level governance comprising supra-national, national, and sub-national (or 'third level') elements.60 This raises the question to what extent is devolution a practical example of the principle of subsidiarity - a principle that has, since the ratification of the Treaty on European Union (TEU), informed many contributions to the 'Europe of the regions' literature? What, if any, lessons may the advocates of a greater measure of subsidiarity in EU governance draw from the process of devolution as it has been applied in the UK? In this section we argue that the UK's devolution experience in fact demonstrates the limitation of subsidiarity as a device for transcending the policy and legitimacy problems inherent to a process of trans-national governance generally, and European Union governance in particular.
The role of the sub-national (or 'third') level of governance within the evolving structure of overall EU governance has, in recent years, become something of a preoccupation on the part of EU scholars. For many, a shift towards an multi-level governance system for the EU, in which regions play a larger role in the overall policy process, is considered to be both possible and desirable.61 Enhancing the authority of the regions not only permits EU economic and social policies to be better adapted to dissimilar conditions within individual member states, it also represents a barrier to an over-concentration of power at the highest - EU - governance level. Instead, a shift to genuine multi-level governance in the EU will ensure that governance remains (will be brought) 'close(r)' to the citizens and societies it serves and thus will retain (acquire) popular legitimacy as a result.62 However, much of this discussion of regionalism in EU governance is conducted in informal language. Multi-level governance is presented as decision making system in which regions within EU member states become involved as independent participants in the arrangements of EU policy-making and have competence for common policies which they are best placed to formulate and implement. The difficulty is - as the experience of UK devolution shows - multi-level governance inevitably will be characterized by a plurality of sovereignties involving multiple agencies with shared and overlapping constitutional authorities. In the context of the EU debate, subsidiarity as a 'rule of policy assignment' has been advanced as a device to ensure coherence in an EU multi-level governance system that includes sub-national authorities. The general subsidiarity 'rule' stipulates that the competence for a policy should be assigned to that level of governance which can discharge it most efficiently, and that wherever possible this should be at that level which is 'closest' to the citizen. Given that policy overlap and policy contagion are unavoidable features of multi-level governance, this will yield an EU governance arrangement in which each of the EU institutions, the member state and the sub-national governments singly and severally exercise policy competencies.
An important element in the growing preoccupation with multi-level governance has been the decline in the importance of the sovereign nation state within the EU decisional architecture. Indeed, some have questioned whether the nation state remains at all relevant to the current phase in the evolution of the EU. An example of this type of thinking is to be found in Neil MacCormick's collection of essays, Questioning Sovereignty.63 There, he demonstrates how membership of the European Union has changed fundamentally the nature of national sovereignty;64
"...it is clear that absolute or unitary sovereignty is entirely absent from the legal and political setting of the European Community. Neither politically nor legally is any member state in possession of ultimate power over its own internal affairs. Politically, the Community affects vital interests, and exercises political power on some matters over member states. Legally, Community legislation binds member states and overrides internal state-law within the respective criteria of validity. So the states are no longer fully sovereign states externally, not can any of their internal organs be considered to enjoy present internal sovereignty under law; nor have they unimpaired political sovereignty."65
He concludes: "Western Europe's successful transcendence of the sovereign state and of state sovereignty is greatly to be welcomed".66 Later he suggests that subsidiarity can be a basis for recognizing "...further levels of system differentiation...",67 where this points to arrangements of governance 'beneath' the nation state as well as beyond it. In short, subsidiarity is seen as offering a mechanism for reconciling the demands of sub-national groups for a greater measure of legislative autonomy in a way that does not undermine the coherent and unitary economic and political framework of the EU. It provides a non-divisive method for promoting self-determination and cultural diversity. But is this a correct reading of subsidiarity?
It is certainly true that European integration has changed the nature of national sovereignty available to the EU member states. Similarly, subsidiarity provides a conceptual framework for assigning competencies between national and EU levels of authority. However, we take issue with two propositions implicit or explicit in the aspirations for EU multi-level governance as suggested by MacCormick. First, is it true to say that European integration has transcended "the sovereign state [and] state sovereignty"? Second, is subsidiarity capable of providing an intellectual framework for determining policy assignment beneath the level of the nation-state as well as beyond that level? The conclusion we derive from our analysis of the devolved governance arrangements in the UK instead is that subsidiarity, as defined in the TEU, has a meaning only within the unique setting of the European Union as a union of sovereign states. And, insofar as the EU remains solely a union comprising of nation states, while subsidiarity may continue to 'inspire' thinking about the structure of multi-level governance in the EU, it cannot be applied instrumentally to achieving that end.
Elsewhere we have argued, as have others, that the 'principle' of subsidiarity within the EU debate, and as reflected in European Treaty reform, in fact comprises two elements.68 On the one hand it is presented as a tenet of democratic government which stipulates that decisions should be taken as closely to the citizens as possible.69 As MacCormick notes;
"The doctrine of subsidiarity requires decision-making to be distributed to the most appropriate level. In that context, the best democracy - and the best interpretation of popular sovereignty - is one that insists on levels of democracy appropriate to levels of decision-making."70
On the other hand, subsidiarity is advanced as a (potentially justiciable) procedural rule for assigning powers between different levels of government according to specific 'efficiency' criteria, within the constraints imposed by compliance with democratic principles.71 When presented as a doctrine or canon of 'good' governance, the appeal of subsidiarity is self-evident. Further, it is a doctrine that plausibly can be applied to conceptualise a range of alternative arrangements for the governance of a system in which public policies exist, and in which decisions about the nature of these policies have to be taken.
But persuasive as it may be as a doctrine, subsidiarity suffers from two principal defects as a constitutional tool. In the first place the complexities arising from policy externalities already described (i.e. overlap and contagion) have to be effectively managed if a binary rule of policy assignment is to be avoided and, instead, subsidiarity applied meaningfully.72 This is because subsidiarity is a process for managing competing claims with respect to the application of concurrent powers. To achieve this, subsidiarity has to have a capacity to direct a legislative inquiry to determine the consequences should one level of governance refrain from exercising its legitimate authority over a measure in deference to considerations raised by another level of governance with whom it shares competence over that measure.73 If it is to command legitimacy, such an inquiry must be transparent and engage the legislatures involved (not the executives). It must evaluate both elements (legitimacy and efficiency) that comprise the 'principle' of subsidiarity, and the process should be justiciable.74 Second, the corollary to a consistent application of subsidiarity is the requirement that the multi-level governance structure is sufficiently fluid to permit policy competencies to be re-assigned between the different levels as circumstances dictate. It is unlikely that a particular configuration of policy assignment will remain optimal indefinitely. As the barriers that separate the internal economy and society (i.e. the policy jurisdictions) of the EU progressively fall, the interdependencies between the regions will intensify. The logic of this, as the history of the EU has demonstrated, is for more and more authority automatically to gravitate to the 'higher' level of governance.75 This can be avoided only by correctly specifying the conditions necessary to facilitate that application of subsidiarity as a process - that is, as an on-going and ever present legislative inquiry. If we are to define a constitutional architecture for EU governance, first we have to stipulate the objectives that we wish it to meet, and the criteria by which we might appraise or revise it. This is not a novel argument. Any account of the origins of the prevailing constitutional arrangements of EU governance would focus on the primary objective of designing a governance system that revolved around (and buttressed) the nation state but which, at the same time, eroded or contained particular elements of national sovereignty through a partial shift to supranational governance.76 Further, a prominent role would be accorded to the judicial process within that arrangement. In a similar way, only a comprehensive analysis of the role that subsidiarity can play in legitimating the EU governance system will enable us to design the constitutional arrangements necessary for its realization.77
To be sure, the governance system of the European Union has, partially and imperfectly, accommodated both defects noted above. The formulation of subsidiarity set out in Article 5 (ex 3b) of the TEU was an attempt to manage the problem of concurrent powers by setting an effectiveness 'test' that all proposed EU policy actions must meet before they could become EU legislative proposals, and authority consequentially gravitate to the EU level. But that test has neither been conducted transparently, nor has it been subject to judicial review.78 Subsidiarity as practiced in the EU therefore fails the test of a meaningful legislative inquiry and, consequently, is unlikely to enhance confidence within the polity. Similarly, the EU governance system has proven itself to be fluid within its constitutional limitations. Competencies have been transferred from national to the EU governance level, both by formal Treaty revisions, and informally by use of the provisions of Article 308 (ex 235).79 In most, if not all, cases this transfer of competence has been predicated on efficiency considerations. Frequently, the competencies newly assigned to the EU level touched on policies hitherto the responsibility of sub-national governments of member states, and it is those governments who have been the integration 'losers'.80 Therefore while the extension of EU competencies may be regarded as extending the scope of supranational governance in the aggregate policy process at the expense of national governance, at the same time it has served to empower national governments at the expense of sub-national governments. This tendency for integration simultaneously to extend the competencies of EU governance and to strengthen the national institutions of member state governments is a well known feature of the integration process.
Despite its partial and imperfect application thus far, the principle of subsidiarity as a process mediating policy assignment between the EU and the member state remains conceptually and constitutionally plausible. But this is the extent of its plausibility. When we turn to national and sub-national systems of governance, however, subsidiarity is not plausible. As implicit in MacCormick's arguments, the matter turns on competing interpretations of sovereignty and, in particular, on his claim that "...the [member] states are no longer fully sovereign states externally". Krasner lists four usages of the term 'sovereignty'81 - domestic sovereignty, which is the organization of authority within the nation state; interdependence sovereignty which refers to a state's ability effectively to control external problems independently of other states; international legal sovereignty, defined as a residing with a state whose claim to territoriality is recognized by other states and which voluntarily may enter into, or leave, agreements with other states; and Westphalian sovereignty, describing the non-interference by external actors in domestic authority arrangements. The key point is that while EU membership has eroded significantly both interdependence sovereignty and (consequently) Westphalian sovereignty - and has thereby restricted the scope of domestic sovereignty (although formally only in the sphere of laws82) - it is not clear that it has undermined the international legal sovereignty of the member states. The litmus test of the international legal sovereignty is whether or not a member state has the competence to withdraw from the EU. If it can do so, international legal sovereignty must be intact; if it cannot, then it has surrendered that sovereignty.83
It is tempting to regard this question solely as a - probably hypothetical - matter of EU or international law.84 But the implications extend beyond this legal technicality and impact directly on the application of subsidiarity in the manner proposed by advocates of multi-level governance. This is because it is only when a level of governance has international legal sovereignty that it is equipped to participate directly in a trans-national legislative inquiry concerning the management of concurrent powers.85 And because levels of member state governance other than the national level are not sovereign in that sense, and accordingly have no independent constitutional status in the EU Treaties, they are unable to participate as partners in a process of policy assignment involving such an inquiry. Sub-national governments have no competence under EU law independently to participate in, or to conclude, agreements at the EU level. They may only do so in the restrictive sense of being delegates of the member state, as under Article 203 (ex 146). Were they to have such independent competence, the international legal sovereignty of member states would, by definition, have ended. The conclusion is clear. Subsidiarity in the sense applied in EU discussions and in the TEU provides little effective purchase in effecting directly a transition to a genuine system of multi-level governance in the EU. It does, however, indicate the constitutional conditions required for such a transition to begin - namely that regions acquire a position of constitutional authority within the arrangements of EU governance allowing them to become a participant in a policy assignment process mediated by the principle of subsidiarity. Our review of subsidiarity indicates that the EU system of governance not only buttresses the role of the nation state rather than jeopardizing it, at the same time it creates an incentive structure for sub-national government to campaign for independent statehood within the European Union. This is the only route currently available for sub-national governments to participate in the subsdiarity process and, consequently, to react to demands for a greater measure of self-government and local autonomy as citizens increasingly question the legitimacy of a governance system in which regions have no constitutional 'voice'.
These questions of governance, legitimacy and subsidiarity are not confined the European Union but have broader relevance as reflected in research being conducted by international relations theorists and international legal scholars.86 A pervasive feature of the organization of the contemporary global economy is the emergence of international institutions which incorporate binding and enforceable rules and which are replacing the organizations of (commercial) diplomacy hitherto based on voluntary codes, behavioral norms, and policy discretion.87 Although there is considerable dispute in the literature over the nature of the autonomy which these new institutions command - autonomy as delegated power versus autonomy as independent authority - few take issue as to the nature of the transformation underway. This poses a clear and distinctive set of challenges for national governance systems everywhere, challenges that are foreshadowed in the subsidiarity debate in the EU. Similar questions are bound to arise as the competence of global governance extends into new areas; that is, as policy spillover proceeds.88 What powers will sub-national authorities be able to retain as global governance moves into new policy areas and nation states are required to acquire greater authority over domestic policies in order that they may successfully bargain within these international governance institutions? To turn the argument around - in a world where an ever increasing part of our economic and social interactions are governed by international institutions according to rules, regulations and conventions, what scope remains for meaningful regionalism that does not undermine the international legal sovereignty of the state?
60 For a critical and informed review of the role of the 'third level' in EU governance, see C Jeffery "Sub-National Mobilization and European Integration", Journal of Common Market Studies, Vol 38, No 1 at 1-24
61 For a comprehensive review of this debate, see G Marks, L Hooghe and K Blank, "European Integration from the 1980s", Journal of Common Market Studies, Vol 34, No. 3, at 341-378
62 Among the objectives listed in a recent strategy document, the Commission includes new forms of European governance which include "new forms of partnership between different levels of governance in Europe", and strengthening "civil society's voice in the process of policy shaping and implementation to ensure a proper representation of Europe's social and economic diversity at European Union level". EC, Shaping the New Europe; Strategic Objectives 2000-2005, COM(154), February, 2000 p5-7
63 ibid, p135
64 N MacCormick, Questioning Sovereignty: Law, State and Nation in the European Commonwealth (1999) at 123-136
65 This argument is worked through in Chapters 8 ('On Sovereignty and Post-Sovereignty') and 9 ('Democracy and Subsidiarity in the European Commonwealth') of the volume.
66 ibid, p133, emphasis added
67 ibid, p133
68 See A. Scott, J. Peterson and D. Millar, "Subsidiarity: A Europe of the Regions vs The British Constitution?", Journal of Common Market Studies, Vol. 32, No. 1, p47-67
69 This depiction of subsidiarity appeared in Article A, TEU (Article 1, Treaty of Amsterdam).
70 Neil MacCormick, Questioning Sovereignty: Law, State and Nation in the European Commonwealth at 135
71 Article 5 (ex 3b), Treaty of Amsterdam). Under this rule, the EU will act in areas of concurrent competence "...only if and insofar as the objectives of the proposed action cannot be sufficiently achieved by the Member States". However, as has been argued, this is a very imprecise rule and alternative formulations exist which better lend themselves to objective operationalisation.
72 A binary rule simply states that if one level of governance has competence over the policy, no other level of governance has authority over how that competence is exercised. While a binary rule is possible, it raises serious problems of policy spillover and policy contagion unless a strict separation of the policy jurisdictions can be enforced.
73 The suggestion of a legislative inquiry originates in George A Bermann, 'Taking Subsidiarity Seriously: Federalism in the European Community and the United States", Columbia Law Review, Vol.94, No.2 at 335.
74 Legitimacy requires accountability, which in turn necessitates transparency. Bermann describes this as the 'confidence building' function of subsidiarity, ibid at 367
75 In effect this describes the historical dynamic of 'state-building' - see Samuel H Beer, To Make a Nation: The Rediscovery of Amerian Federalism, (1993) at 11
76 See, for instance, Alan S Milward, The European Rescue of the Nation-State (1992)
77 Multi-level governance differs from multi-level administration, a difference which is not adequately addressed in the 'regionalist' literature. Governance involves participation in each of the four stages of policy-making (formulation, negotiation, implementation, enforcement); administration in only the latter two of these.
justiciability of subsidiarity is a subject of considerable debate within EU
legal discourse. See the discussion in Gráinne de Búrca,
"Reappraising Subsidiarity's Significance after Amsterdam", Jean Monnet Working
Papers No.7/99, Harvard Law School.
79 In revisions to the Treaty of Rome there has only been one instance where a binary rule was applied - this was competence over monetary policy post-EMU. For obvious reasons competence for EU monetary policy could not be a shared between EU and national organs
80 This explains why the TEU provided for the creation of the Committee of the Regions and introduced Article 203 (ex 146).
81 Stephen D Krasner, Sovereignty; Organized Hypocrisy (1999) at 9-25
82 Krasner's definition of domestic sovereignty refers to 'authority' to act and not 'autonomy' to act. Only in the legal sphere does EU membership impinge on domestic 'authority' structures.
83 The comparison is usually made with the US Constitution and the nationalist versus compact theorist debate which surrounded that Union in the run-up to the Civil War. The triumph of the nationalist rather than compact interpretation led to the conclusion that a state could not legally withdraw from the Union. In the EU any similar suggestion almost certainly would be considered preposterous.
84 As is well known, the problem stems from the silence of the EU Treaties on the issue of secession.
85 Although most federal systems have constitutional rules governing policy assignment between levels of government within the nation state, and means of changing this assignment, clearly these rules are not transferable to trans-national governance structures such as the EU.
86 For an excellent review of this literature see Anne-Marie Slaughter, Andrew S. Tulumello and Stephen Wood, "International Law and International Relations: A New Generation of Interdisciplinary Scholarship", American Journal of International Law, Vol. 92, No. 3, July, 1998, p367-397. However, the genesis of this work is much older. It is foreshadowed by J H H Weiler in his 1982 exhortation that political scientists take more account of the role of the European Court of Justice in analysing the political dynamics of European integration (J H H Weiler, "Community Member States and European Integration", Journal of Common Market Studies, Vol.XXI, No.1&2 (1982) p39-65. For the seminal analysis of the interaction between the EU legal order and the integration process generally, see Joseph H H Weiler, "The Transformation of Europe", Yale Law Journal 100 (1991)
87 Slaughter et al note that "...much [international] institutionalized cooperation has taken an increasingly 'legalized', ' judicialized' or constitutional form", op cit at 370
88 For example, spillover is evident by the emergence of international debates over common rules with respect to a raft of trade-related national measures including competition policy, environmental issues and labour standards.