Multilevel Governance in the EU

The emergence of the ‘third level’ within the EU, in particular since the second half of the 1980s (due to the combined impact of decentralisation processes at national level and the progressive opening up of the Union to input from the sub-national authorities), led scholars in the late 1980s and in the early 1990s to rethink the structure of the European decision-making in view of the role played by the sub-national level. A crucial development occurred in 1988, when a major reform of structural funding took place.¹ Under the new regime, the funds had to be administered through partnerships established within the Member States, consisting of representatives of national and regional and/or local authorities and the Commission. It was indeed from the study of the renewed European structural policy and particularly of the partnership principle that in the early 1990s Gary Marks developed and launched the notion of ‘multilevel governance’.²

The multilevel governance analysis framework was later extended to the whole spectrum of EU action, and in this way multilevel governance became a general theory on the functioning of the EU (albeit, more precisely, the multilevel governance paradigm is usually applied only to the former ‘First Pillar’,³ due to the mainly intergovernmental nature of the ‘Second Pillar’ and ‘Third Pillar’⁴).⁵ Over the last 20 years, also thanks to research into multilevel governance, it became increasingly clear that sub-national authorities perform a role in EU decision-making (for example, through consultation with the Commission or through national participation channels) and are involved in the implementation of EU law and policy (in some Member States, they are responsible for the transposition of directives and for the implementation of the structural or cohesion policy⁶). According to multilevel governance theory, in the atypical European ‘multilevel polity’, decision-making in a number of policy areas in the former First Pillar is not dominated by national governments, as put by liberal intergovernmentalism,⁷ nor is it entirely in the hands of supranational institutions. In those policy areas and to a varying degree across the different Member States, European decision-making rather reflects a ‘triangulation of political relationships’ among national, supranational and sub-national players,⁸ i.e., a system of multilevel ‘co-governance’ in accordance with the constitutional structure of the Member States (federal, regionalised or otherwise decentralised).

Over the years, the multilevel governance paradigm has been fundamentally accepted by the majority of legal and political scholars and became a key pattern for understanding the functioning of the EU and its territorial dynamics.⁹ The summa divisio in the field of multilevel governance is between two ideal types. The first (Type I multilevel governance), whose prototype is federalism, includes general-purpose territorial jurisdictions (states, regional and local authorities). These ‘bundle together multiple functions, including a range of policy responsibilities, and in many instances, a court system and representative institutions’. In this context, ‘every citizen is located in a Russian Doll set of nested jurisdictions, where there is one and only one relevant jurisdiction at any particular scale’. The second type of multilevel governance (Type II) embraces ‘task-specific jurisdictions’, such as specialised agencies and task-specific organisations that provide a certain local service, select a product standard, monitor water quality in a particular river and so forth. This study is concerned with Type I multilevel governance only, as it focuses on the role and mutual interaction of territorial authorities that are expression of regional and local communities (general-purpose territorial jurisdictions).¹⁰

Concerns relating to Type I multilevel governance are at the forefront in the Commission’s White Paper on European Governance of 2001 and in the Committee of the Regions’ White Paper on Multilevel Governance of 2009.¹¹ In the White Paper on European Governance, the Commission places substantial emphasis on the interaction and dialogue between the EU and the sub-national authorities and lays down recommendations (for example, ‘There needs to be a stronger interaction with regional and local governments’, ‘Establish a more systematic dialogue with representatives of regional and local governments through national and European associations at an early stage in shaping policy’, ‘Bring greater flexibility into how Community legislation can be implemented in a way which takes account of regional and local conditions’¹²). ‘Multilevel governance’ is expressly mentioned at p. 34 of the document in relation to how competence is shared, not separated, between the EU and the Member States:

The Union needs clear principles identifying how competence is shared between the Union and its Member States. In the first place this is to respond to the public’s frequent question ‘who does what in Europe?’ A common vision is needed to answer this question. The White Paper has highlighted […] a Union based on multilevel governance in which each actor contributes in line with his or her capabilities or knowledge to the success of the overall exercise. In a multi-level system the real challenge is establishing clear rules for how competence is shared – not separated; only that non-exclusive vision can secure the best interests of all the Member States and all the Union’s citizens.¹³

This idea of interwoven and overlapping competences between the different tiers of government is strongly rooted in cooperative federalism theory¹⁴ and constitutes the foundation of multilevel governance. The key challenge for multilevel governance is bringing together and coordinating the action of different territorial levels. The specific instruments identified by the Commission to this purpose are the ‘structured dialogue’ and the ‘tripartite agreements and contracts’.¹⁵ The ‘structured dialogue’ consists of regular meetings between the Commission and the European and national associations of sub-national authorities. These meetings may concern issues of broad interest, such as the annual work programme of the Commission (general dialogue), or a specific policy area (thematic dialogue).¹⁶

The ‘tripartite agreements and contracts’ are stipulations between the Commission, a Member State and one or more sub-national authorities concerning the implementation of binding law (contracts) or soft law (agreements).¹⁷ Up to now, this instrument has found only limited application, despite the submission of some projects concerning environmental protection in Birmingham (UK), Lille (France) and Pescara (Italy). The first and to date only tripartite agreement was signed on 15 October 2004 by the Commission, the Italian Government and the Region Lombardy in relation to sustainable mobility.¹⁸ Although tripartite stipulations have not found extensive application, probably due to the lack of planned European funding,¹⁹ their existence is emblematic, as it shows that the EU ascribes to the sub-national authorities the status of contractual partners on an equal footing with other tiers of government and the ability to negotiate contracts and agreements with the Commission. Only the Member State is responsible vis-à-vis the EU in case of non-compliance with the contract or agreement. However, as shown previously (cf. supra Chap. 2), in some Member States with federal or regional structure, there are mechanisms ensuring that a sub-national authority that causes a non-compliance with EU obligations is held financially accountable and is involved by the Member State in the pre-judicial and judicial phases of the infraction proceedings. This corroborates the submission that in the EU multilevel system the status of the sub-national authorities derives from the combination, ‘amalgamation’, of rules emanating from the European and the national levels.

Eight years after the Commission’s White Paper, the Committee of the Regions published the White Paper on Multilevel Governance, which defines multilevel governance as follows:

coordinated action by the European Union, the Member States and local and regional authorities, based on partnership and aimed at drawing up and implementing EU policies.²⁰

The document further specifies:

Multilevel governance is not simply a question of translating European or national objectives into local or regional action, but must also be understood as a process for integrating the objectives of local and regional authorities within the strategies of the European Union. Moreover, multilevel governance should reinforce and shape the responsibilities of local and regional authorities at national level and encourage their participation in the coordination of European policy, in this way helping to design and implement Community policies.²¹

In the light of this definition, multilevel governance appears to be a dynamic concept (‘coordinated action’, ‘translating into local or regional action’, ‘process for integrating the objectives’, etc.) consisting of two major elements:

The ‘dynamic’ notion of multilevel governance emerging from the Committee of the Regions’ White Paper is fundamentally ‘procedural’, as it lays down certain guidelines in relation to how decisions are to be made.²² Multilevel governance emerges as a ‘method’ or an ‘approach’,²³ the appropriate method or approach according to the Committee, for bringing together and coordinating the action of the different layers of government in the EU multilevel system.

So far, the notion of multilevel governance has remained mainly the province of students of politics and of official statements or policy documents of Union bodies and institutions. However, whilst political studies deserve great credit for defining the concept and for explaining its importance for the functioning of the EU, the notion has remained mainly descriptive.²⁴ From a legal perspective, we need to move from a descriptive to a prescriptive notion of multilevel governance. Law is not only concerned with ‘what is’ but principally with ‘what ought to be’. If, as this study shall demonstrate, multilevel governance is a legal concept, it will determine how the EU has to handle the territorial pluralism within it. The Committee of the Regions’ White Paper, due to its objective to shape future European integration, is an essential starting point for formulating a hypothesis in relation to the legal content of multilevel governance.²⁵

In light of the definition of multilevel governance in the White Paper (‘coordinated action by the European Union, the Member States and local and regional authorities, based on partnership and aimed at drawing up and implementing EU policies’²⁶), the working hypothesis is that, as a minimum, multilevel governance should include the following:

The foundation of multilevel governance lies in the first place in the Treaties (constitutional charter of the EU):

The constitutive elements of the notion of multilevel governance are mirrored in the EU primary law. However, despite the fact that the roots of multilevel governance are in EU primary law, and that the EU lays down a framework for regional and local participation, it cannot oblige the Member States to create participation channels for the sub-national authorities or to use those prompted at Union level. For example, participation in the Council is dependent not only on the EU but especially on if and to what extent each Member State allows the sub-state entities to be involved in the Council activity (cf. infra Sect. C.2). In accordance with the ‘united in diversity’ motto, the EU cannot impose uniform patterns to all the Member States. The EU can open doors and windows, but it is largely up to the Member States and their regions to seize the opportunity. Therefore, the legal source of multilevel governance cannot be the EU primary law only. There are two parallel dimensions of multilevel governance in the EU, one supranational (the EU) and one national (the Member States).²⁹ Furthermore, the insights offered by the EU primary law are rather minimal. Whilst the EU opens up its processes to forms of regional and local participation inspired to multilevel governance (e.g., Council, Committee of the Regions, involvement of regional parliaments in the ‘early warning system’), there is little prescriptive indication of how these forms of participation should work in practice. For example, the decision on the composition of each (sub-)national delegation to the Committee of the Regions is left to the Member States, so it is up to each Member State to strike a balance in relation to the representation of the different tiers of government (regional and/or local).³⁰ Alone, the opportunities prompted by the EU primary law do not offer the whole picture of regional and local participation in the EU. Multilevel governance, far from being a monolithic notion finding application everywhere in the same way, is a largely asymmetrical concept receiving differentiated application in each Member State. Accordingly, there is no single pathway to enforcing multilevel governance but potentially 28 partly different pathways.

The concept of multilevel governance is no longer confined to political discourse or theory. As well as being rooted in EU primary law as a concept embracing potentially all fields of Union action, it is also reflected in EU secondary law specifically concerning energy policy, as well as economic, social and territorial cohesion.

Regulation (EU) No 1233/2010 concerning EU financial assistance to projects in the field of energy³¹ contains a strong reference to multilevel governance. Point 3 of the Preamble states that ‘Cooperation among the various tiers of government (multi-level governance) is essential in th[e] context [of the development of further renewable energy sources and the promotion of energy efficiency]’. Regulation No 1233 also creates a dedicated financial facility in support of investment projects related to energy efficiency and renewable energy by local, regional and national public authorities (cf. Point 4 of the Preamble and Annex II). The notion of multilevel governance as ‘cooperation’ incorporates the idea of partnership-inspired ‘coordinated action’, synergy for the achievement of Union objectives (β) and also the respect for the role of the sub-national authorities in conformity with subsidiarity (α). For example, Regulation No 1233 establishes that the Union financial support shall facilitate the financing of investments in energy saving, energy efficiency and renewable energy projects by local, regional and, ‘in duly justified cases’, national public authorities (cf. Annex II). An interesting aspect of Regulation No 1233 is that it highlights a ‘positive dimension’ of multilevel governance in the EU. Through its financial intervention, the EU enhances the role of regional/local authorities in achieving the Union’s objectives.

The strongest reference to multilevel governance can be found in Regulation (EU) No 1303/2013 on EU funding promoting economic, social and territorial cohesion.³² This act indicates multilevel governance as a ‘principle’ that, along with subsidiarity and proportionality, must be ‘respected’ by the Member States when creating partnerships with the sub-national authorities and other economic and social actors for the implementation of the EU economic, social and territorial cohesion policy.

Like in the Committee of the Regions’ White Paper, also in Regulation No 1303 multilevel governance emerges as a ‘procedural’ concept, i.e., as a ‘governance method’ that requires the involvement of the sub-national authorities (‘In accordance with the multi-level governance approach, the partners (…) shall be involved by the Member States in the preparation of Partnership Agreements and progress reports and throughout the preparation and implementation of programmes’³³) and that requires ‘coordinated action’ between the different levels of government (‘In order to respect [the] principles [of partnership and multi-level governance] coordinated action is required, in particular between different levels of governance’³⁴).

Regulation No 1303 also contains indications concerning the raison d’être of multilevel governance, where it says that respect for the principles of partnership and multilevel governance is required ‘in order to facilitate achieving social, economic and territorial cohesion and delivery of the Union’s priorities of smart, sustainable and inclusive growth’.³⁵ Accordingly, in the view of the EU legislator, the partnership with the sub-national authorities finds its justification in the need to enhance the effectiveness of the EU social and cohesion policy. Similarly, Regulation No 1303 also highlights that the ultimate purpose of a partnership in the context of the EU regional policy is ‘to ensure the ownership of planned interventions by stakeholders and build on the experience and the know-how of relevant actors’ (Point 11 of the Preamble). The phrase ‘ownership by stakeholders’ suggests that the participation of sub-national authorities in the EU regional policy could contribute to strengthen acceptance of these policies, i.e., to their legitimacy (‘ownership’) as well as to their effectiveness (‘build on the experience and the know-how of relevant actors’).

The inclusion of multilevel governance as a key principle of the EU regional policy sets aside any remaining doubt in relation to the nature of multilevel governance as a legal concept. Multilevel governance emerges as a principle of procedural nature (i.e., it requires a certain ‘method’ or ‘approach’), which must be respected by the Member States in the field of economic, social and territorial cohesion. In the context of this policy field, multilevel governance is a concept, more specifically defined as a ‘principle’, established by Union law, and as such its interpretation may be referred to the ECJ in preliminary rulings stemming from legal disputes in domestic courts. However, given that constitutional arrangements vary asymmetrically across the Union, the ECJ is likely to stick to a minimal notion of multilevel governance whilst applying it to a specific State context, rather than dictating prescriptive solutions for the Member States. The Union notion of ‘multilevel governance’ must necessarily be minimal and procedural. It must be minimal because the Union can require the Member States to respect a certain ‘method’ or ‘approach’ when implementing Union policies; however, it cannot impose a specific multilevel structure on them. The multilevel governance ‘method’ shall include an appropriate involvement of the sub-national authorities and of other relevant players in the preparation and execution of projects in the EU economic, social and cohesion policy. In this way, far from laying down a set of prescriptive rules, multilevel governance in this policy area emerges as a principle and, more specifically, as a ‘procedural principle’ indicating in general how decisions are to be made and policies implemented. The expected results and added value of that method are the effectiveness and legitimacy of EU policies.

All Member States have a constitutional framework that includes regional and/or local self-government. Given that the Member States are also part of the EU, and that almost all EU policies interfere with the competences of the regional or local level, we are faced with mutually contradicting propositions. On the one hand, the Member States’ constitutions value regional/local autonomy, but, on the other, a considerable degree of that autonomy has been handed over through the acceptance of the EU membership. These are opposite poles pulling into different directions, regional and/or local autonomy vs. creation of a supranational entity dragging a substantial part of that autonomy away from the sub-national level. Can a constitution command one thing and its opposite at the same time? Is it possible to reconcile such conflicting propositions? Whilst contradictions and inconsistencies may exist in any human artefact, including a constitution, the duty of lawyers is to identify and to resolve them.³⁶

A first hypothetical but simplistic way to reconcile local/regional autonomy and the growing role of the EU could be the principle of supremacy of EU law. According to that principle, as construed by the Court of Justice in Internationale Handelsgesellschaft,³⁷ EU law shall prevail over any rule of domestic law, including constitutional law. By analogy, one may argue that the centripetal push towards European integration could legitimately forfeit the autonomy of the sub-national authorities. However, it is well known that the national courts have set limits to the prevalence of EU law in order to protect the Member States’ constitutional identity.³⁸ After the Treaty of Lisbon, the constitutional identity of the Member States is protected also by EU primary law. Article 4(2) TEU includes regional and local self-governments among the ‘fundamental constitutional structures’, which are part of the ‘national identity’ of the Member States that the Union is obliged to respect.³⁹ The constitutional laws of the Member States underpin and add new vigour to this conclusion. Some examples will illustrate the point more effectively.

The Swedish Instrument of Government stipulates that ‘Swedish democracy is founded on the free formation of opinion and on universal and equal suffrage. It shall be realised through a representative and parliamentary polity and through local self-government’ (Chapter 1, Section 1, para. 2, Instrument of Government). It emerges clearly from this constitutional provision that the existence of local self-government is a qualifying feature of Swedish democracy and of the Swedish State. Similarly, in the Czech Republic, the right to self-government is regarded as one of the fundamental and inalienable rights of local communities (cf. Art. 100 Const.), whilst in Italy local autonomy and decentralisation are fundamental principles of the Constitution (cf. Art. 5 Const.).⁴⁰

These examples underpin the opinion that self-government and local autonomy are constitutional cornerstones. Parliaments cannot sic et simpliciter abolish or restrict them. In some Member States, this would be limited or impossible even through an ordinary constitutional revision. For example, in Austria, the federal principle is a core element of the constitutional setting. This implies that the suppression of the Länder or of the Municipalities, or a major restriction to their power, would require a total revision of the Federal Constitution pursuant to Article 44(3) of the Federal Constitutional Law. In addition to the passage of a constitutional amendment by the National Council and the Federal Council by two-thirds majority, a total revision would also require a referendum.⁴¹

Similarly, in Germany, Article 79(3) of the Basic Law rules out any constitutional amendment affecting the division of the country into Länder, their participation on principle in the legislative process and the federal character of the State. These limits to constitutional revision do not apply only internally within Germany but also in relation to transfers of powers to the EU (cf. Art. 23(1) Basic Law). Furthermore, the German Federal Constitutional Court held that the Länder are endowed with a ‘hard core’ of responsibilities that cannot be taken away from them.⁴²

Due to the fluidity of its unwritten constitution, a very interesting pattern in this context is the UK. According to the orthodox doctrine of parliamentary sovereignty,⁴³ the UK Parliament can modify any earlier statute, i.e., in principle it would have the power to give any responsibility to the local authorities, but it could also take everything away from them. However, over the last few decades, the doctrine of parliamentary sovereignty, in its traditional form, has been repeatedly challenged. Notably, in the case Thoburn v Sunderland, Laws LJ held that the doctrine of implied repeal did not apply to the conflict between an earlier ‘constitutional statute’ and a later one.⁴⁴ According to Laws LJ, the category of constitutional statutes includes any statute that ‘(a) conditions the legal relationship between citizen and State in some general, overarching manner, or (b) enlarges or diminishes the scope of what we would now regard as fundamental constitutional rights’. In Laws LJ’s view, any statute not fitting this description can be classed as an ‘ordinary statute’. Important examples of constitutional statutes are the Magna Charta of 1215, the Bill of Rights 1689 and the Act of Union 1707 and also more recent statutes such as the Human Rights Act 1998, the Scotland Act 1998, the Government of Wales Act 1998 and the European Communities Act 1972.⁴⁵ The Localism Act 2011 contains an interesting development in relation to local government in England.⁴⁶ Section 2(4) stipulates that the powers of local government are not subject to implied repeal. By contrast, the acts that Thoburn considered did not include explicit provisions on implied repeal. Laws LJ actually stated that the constitutional acts were such ‘by force of the common law’. The further question of whether (expressly, of course) it is legally possible for Parliament to remove local government or devolution goes completely to the heart of the constitution in the UK. If Parliament ever tried to remove local government or devolution or to render them irrelevant, most certainly there would be court cases that would test the courts’ adherence to the principle of parliamentary sovereignty to the maximum extent. To date, the only case to have suggested a possible willingness of the courts to find certain legislation passed by Parliament to be unlawful is R (Jackson) v Attorney General. The House of Lords found, in an obiter dictum, that the courts could find a statute to be in breach of the principle of legality.⁴⁷ However, whilst, politically, the abolition of local self-government or of devolution is currently unthinkable, it would appear uncertain, to say the least, whether Parliament would be legally authorised to dismantle them. Such uncertainty demonstrates that, even in a flexible constitutional system like the UK, the regional and local self-governments are so deeply rooted in the constitutional tradition that they could even limit the principle of parliamentary sovereignty.

The above examples corroborate the conclusion that local/regional autonomy is an essential element of the constitutional identity of the Member States. The key importance of local/regional autonomy for the constitutional identity of the Member States can be seen also in the European Charter of Local Self-Government of the Council of Europe, which has been ratified by all the EU Member States. Multilevel governance, by commanding the involvement of the sub-national authorities in the EU decision-making process and in the implementation of EU law/policy, enables the sub-national authorities to maintain and possibly expand their constitutional role whilst being involved in supranational integration. This (and not supremacy) is the correct way for reconciling regional and local autonomy with European integration. Multilevel governance has a key constitutional function to play in the EU context. It is constitutionally required by both the Member States and the Union primary law for the protection and the development of an essential part of the Member States’ constitutional identity.

The justiciability of a principle is not essential in order to conclude that it is a juridical and not a philosophical or political notion. For example, in the next chapter, it will be argued that the justiciability of the principle of subsidiarity is limited and yet that subsidiarity remains a legal principle and one of the constitutional cornerstones of the EU.⁴⁸ Like subsidiarity, multilevel governance too lives on the edge between law and politics. The judicial enforcement of multilevel governance in Union courts is not the primary route for the application of that principle on the EU level. The principal sanction in case of non-compliance with the requirements of multilevel governance by the EU or the Member States is the loss of legitimacy of Union action or of the Member States’ participation in the EU. Still, Union courts have an important role to play in enforcing and shaping multilevel governance. There are fundamentally three ways in which Union courts can play a role: 1) in relation to the Treaty articles embodying the idea of multilevel governance indicated supra at Sect. B.3, 2) in relation to the concept of ‘national identity’ of Article 4(2) TEU and 3) in relation to acts of secondary law that contain reference to the concept (principle) of multilevel governance.

In the context of this study, the term ‘legitimacy’ indicates the acceptability, albeit not necessarily the inner/psychological acceptance, of authoritative decision-making by the subjects to power and by all the players concerned. A distinction can be made between two types of ‘legitimacy’: one being ‘output legitimacy’, the second being ‘input legitimacy’. ‘Output legitimacy’ is linked to the content of a decision. The fundamental output legitimacy alternative is between ‘approval’ and ‘disapproval’ of a decision by the subject (or, which is ultimately the same, between ‘like’ and ‘dislike’, ‘good’ and ‘bad’). Such ‘approval/disapproval’, ‘like/dislike’, ‘good/bad’, or dichotomy is the result of a subjective evaluation, which is ‘political’ in nature. Being a political concept, ‘output legitimacy’ does not legitimise law and policy objectively, only subjectively. Different groups of subjects, political parties or even individuals may well have different opinions as to what amounts to ‘good’, ‘not so good’ or ‘bad’ governance. This study is only concerned with ‘input legitimacy’.⁵⁵

‘Input legitimacy’ conveys ‘acceptance’ of authoritative decisions not for their ‘content’ but for the way in which a decision is taken. In this way, it leads to the objective legitimacy of a decision, i.e., to the legitimacy of a decision vis-à-vis everybody, objectively, i.e., irrespective of whether a group or an individual agrees or disagrees with it. Legitimacy, here, derives from respect for certain objective procedural criteria; i.e., it results from compliance with a ‘procedure’. Like for democracy, observance of a democratic procedure generates objective legitimacy, that is, objective acceptance. Every member of the community will (have to) accept a democratically taken decision, or a decision made in accordance with a previously agreed procedure, and could not challenge its (objective) legitimacy, even if he or she disagrees with the merit. A decision could be ‘bad’ or ‘inexpedient’ from an individual or group point of view but still be fully ‘legitimate’.⁵⁶

Legitimacy should not be mistaken for mere ‘legality’. A decision can be formally adherent to ‘procedural law’ but still be in conflict with fundamental constitutional values, such as human dignity or, for what concerns this book, regional and local autonomy. Accordingly, a further distinction needs to be made between ‘formal’ and ‘substantial’ legitimacy. ‘Formal legitimacy’ means simple adherence to a procedure. ‘Substantial legitimacy’ includes respect for substantial constitutional requirements. Even a democratic regime should respect certain fundamental values to maintain its substantial legitimacy along with its formal legitimacy. At the same time, it needs to be taken into account that procedures are never ‘neutral’. They facilitate the achievement of certain, previously identified, substantial objectives. For example, the right of the citizen to a hearing in an administrative procedure contributes to the ‘fairness’ of the decision-making activity. Other types of procedures serve different values; for example, judicial proceedings serve the principle of fairness and equality between the parties in the superior interest of justice. The question is not only whether the procedures in force in the single countries and in the EU for regional/local participation have been complied with but also if these procedures are coherent with, and fully satisfactory for, regional and local autonomy, which is a constitutive part of ‘national identity’ (cf. supra Sect. B.5; Article 4(2) TEU). Accordingly, legitimacy in the context of this study has a ‘constitutional’ foundation.⁵⁷