Multi-level governance in a ‘Europe with the regions’

The concept of multilevel governance was introduced to reveal the impact of regions on European Union (EU) policy making in Brussels. In this contribution, I show that multilevel governance also exposes regional involvement in EU affairs within the member states. In contrast to perspectives that focus on formal decision-making by central governments, multilevel governance uncovers significant sharing of authority between governmental actors within and beyond national states even in cases where the formal right to make a decision lies with national governments or the EU legislator. I argue that the concept of multilevel governance is key for understanding developments within a three-tiered EU polity because it directs scholarly attention to the incentives for regions to be involved in EU affairs and for national governments and EU institutions to share their authority with regions.


Introduction
The concept of multilevel governance (MLG) can be traced back to the seminal work of Liesbet Hooghe (1995Hooghe ( , 1996 and Gary Marks (1992Marks ( , 1993 of the 1990s. During this time, regions started to become prominently present and active in Brussels because of a revision of Cohesion Policy in 1988 which required national governments to co-develop plans to implement the funds with subnational governments. Furthermore, the Maastricht Treaty of 1992 granted regions access to the Council of Ministers and established the Committee of the Regions (CoR) which has to be consulted by the European Union (EU) legislator on key areas of regional concern. Liesbet Hooghe and Gary Marks' (2001) landmark study Multi-Level Governance and European Integration clearly revealed that a new conceptual tool had to be developed to account for the presence of regions in Brussels and the impact they made on EU policy. Theories of European integration which were in fashion at the time -that is, neofunctionalism and intergovernmentalism -could no longer capture and explain what was going on in EU governance. MLG was introduced as an original concept to understand this new mode of EU governance which involved a third regional tier alongside member states and EU institutions (Jeffery and Peterson, this issue). The concept of multi-level governance is still widely used in the literature that deals with regions in the EU. 1 Whereas Hooghe and Marks' research was pivotal in revealing the involvement of regions in EU affairs at the EU level, later research exposed that regions have undermined the capacity of central governments to maintain a monopoly over EU policy in the intra-state arena (Jeffery, 2000). A process of regionalization has shifted political authority from the national level down to subnational levels of government and has increased the involvement of regions in EU affairs beyond but also within the member states. One of the main analytical leverages of MLG is its implied shift in focus from formal decision-making by governments to sharing of authority between levels of government during the entire process from policy making to implementation (Hooghe and Marks, this issue;Tortola, 2017). As I will show in this contribution, MLG reveals that regional governments and actors are increasingly involved in EU-affairs beyond and within member-states even in cases where the formal right to make a decision lies with national governments or the EU legislator.
In the next two sections, I briefly discuss the research that reveals the involvement of regions in EU affairs at the EU level and within the member states. In both sections, I discuss the formal legislative role of regions in EU-decision-making processes as well as the involvement of regions in EU policy beyond their formal legislative role. In the fourth section, I argue that MLG is key for understanding the developments in a threetier EU polity because it steers scholarly focus towards the incentives for regions to be involved in EU affairs and for national governments and EU institutions to share their authority with regions. Marks (1996, 2001) have been at the forefront of exploring the involvement of regions in EU decision-making, and they identify three main channels: regional representation the CoR, regional access to the Council of Ministers, and regional lobby offices. 2 The CoR gives subnational authorities a formal role in the EU decision-making process since the CoR was established by the Maastricht Treaty in 1992. 3 The CoR consists of 350 representatives from regional and local governments who should hold a regional or local authority electoral mandate or who are politically accountable to an elected assembly. 4 Since the Maastricht Treaty, the involvement of the CoR in the EU legislative process is mandatory for policy related to 'social and economic cohesion, public health, trans-European networks in the fields of energy, transport and telecommunications, education and youth, culture'. 5 The powers for the CoR have gradually increased over time and with the Treaty of Lisbon mandatory consultation was extended to include employment, social affairs, sport, environment, energy, and climate change. Furthermore, the CoR gained the right of issuing an opinion on its own initiative and to protect its prerogatives before the European Court of Justice. 6 Within the EU legislative process, the role of the CoR may be conceived to be limited because it does not have veto-powers or co-decision rights and can only provide advice. In addition, in almost all member states, national governments decide on the list of candidates for the CoR, although regional governments and/or local and regional government associations can propose candidates (Committee of the Regions, 2009b). 7 However, the CoR is taken seriously by the Commission which reports to the CoR how its opinions have been taken into account even when there is no treaty obligation on the part of the Commission to do so. 8 A closer look at the profiles of the CoR regional government representatives reveals that, in December 2019, 71% are the head or member of a regional executive, whereas 29% is member of a regional assembly. The corresponding percentages for the representatives of local government are 50% and 50%. 9 Nominating highprofile candidates may signal that regional governments take their representation in the CoR very seriously and manage to get their preferred candidates approved by their national governments.

Regional involvement in EU affairs at the EU level
Apart from the CoR, the Maastricht Treaty introduced another legislative role for regions which is the possibility for regional ministers to represent the member state in the Council of Ministers. 10 Numerous regions have sought and gained access to heart of EU decision making. Regional ministers from the federal states of Austria, Belgium and Germany (may) replace national ministers and exercise the full rights associated with the position for policies that are exclusively assigned to the regions. Regional representatives from the regionalized states of Italy, Spain, and the United Kingdom may be authorised to speak during council meetings (Högenauer, 2008;Tilindyte, 2016).
Although access to the Council of Ministers grants regions a formal and direct role in the EU legislative process, a focus on formal decision-making provides for a sceptical view on the possibility of regions to impact EU policy. Central governments act as 'gatekeepers' and regional minister access to the Council of Ministers depends in many member states on the agreement of central government (Högenauer, 2008;Noferini, 2012;Tatham, 2008). However, those regions which have gained the right to replace their national minister are actively involved in the preparation phase of EU decision-making through their participation in member state delegations and through their membership in the working groups of the Council of Ministers.
A major development in the EU multi-level polity has been the growth in the number of regional lobby offices in Brussels. There were 15 of such offices in 1988, and this number increased to more than 200 in 2013 (Callanan and Tatham, 2014;Donas and Beyers, 2013;Greenwood, 2011;Hooghe and Marks, 1996;Marks et al., 2002;Tatham, 2010). This growth in the number of regional lobby offices can arguably be linked to the 1988 reform of the structural funds which instituted 'partnership' among the Commission, national governments, and local and regional authorities in designing, implementing, and monitoring economic programmes (Hooghe and Marks, 2001: 83-86). The partnership principle induced the Commission to involve regional representatives in EU policy-making and incentivized regions to establish representations in Brussels.
Regional lobby offices in Brussels do not have a formal role in the EU decision-making process. However, research reveals that regions perform a dual role as top-down and bottom-up information broker. The most important role regional lobby offices see for themselves is to gain information about funding opportunities and EU legislation relevant to a region and to explain a region's position on issues to EU decision makers as to influence the EU decision-making in favour of the region. Regional lobby offices are welcomed by the European Commission which relies on regional representatives for information because the European bureaucracy lacks the expertise and resources to gather insightful regional-level data. (Greenwood, 2011;Marks et al., 2002;Tatham, 2017). In addition, most regions intensively collaborate with central state authorities rather than bypassing them (Beyers et al., 2015;Callanan and Tatham, 2014) but contextual factors such as different parties in government at the regional and national levels, (fiscal) resources of the office, and length of exposure the integration process may affect the incidence of bypassing versus cooperation (Tatham, 2010).

Regional involvement in EU affairs within the member states
The literature identifies three ways in which regions are involved in EU affairs within the member states: regional involvement in Cohesion Policy, intergovernmental meetings to coordinate EU affairs, and subsidiarity monitoring by regional parliaments.
Perhaps the most visible regional involvement in EU affairs within the member states concerns EU Cohesion Policy. When (candidate) member states, in the opinion of the Commission, did not have adequate regional institutions capable of managing cohesion funds than they were expected to create them (Piattoni, 2009). This expectation arose after the 1988 Cohesion Policy reform which enshrined the partnership principle. This principle stipulates that (implementing) decisions about operational programmes -that is, the plans to implement and spent the funds -should be taken in partnership among the Commission, the member state governments, and regional governments (Allen, 2010). Piattoni (2010: 128) notes that it is one thing to expect cohesion policy to empower the regions of Europe, in the sense of making them more capable of contributing to the policy-making process (policy empowerment), while it is an entirely different issue to expect cohesion policy to bring about a redefinition of the institutional or even constitutional setup of the member states (institutional empowerment).
Indeed, Cohesion Policy seems not to have significantly strengthened regional government. During the 2014-2020 programme period, national governments are responsible for administering and implementing Cohesion Policy in 19 out of 28 member states. Nevertheless, regions in six member states are fully in charge of Cohesion Policy -that is, Austria, Belgium, France, Germany, Italy, and Poland -and regions in three member states -that is, Finland, Sweden, and the United Kingdom -are in full control of administering and implementing Cohesion Policy (Balsiger, 2016;Charles, 2016;Kovács, 2016;Muravska et al., 2016;Polverari, 2016). Although national governments seem to remain in control of Cohesion Policy, regions have become increasingly involved in the governance of Cohesion Policy (Baun and Marek, 2008). In all EU member states regional representatives advise and inform managing authorities and, in some cases, they also help managing authorities with implementing Cohesion Policy. In 14 member states, the aforementioned tasks are integrated into general-purpose regional government, 9 member states have set up specific regional development councils, and only in 5 member states regional development is an exclusive competence of the central government (Hooghe et al., 2016;Schakel, 2018). 11 The possibility for regional ministers to replace their national ministers in the Council of Ministers at the EU level led to the establishment of formalised intra-state coordination mechanisms between central and regional governments. Eight member states established formalised intergovernmental meetings specifically dedicated to EU affairs soon after the Treaty of Maastricht introduced the right for regions to participate in the Council of Ministers. In Austria, Belgium, Germany, and Spain regions can exercise a veto on member states positions to be taken up in the Council of Ministers. Regions in Finland (Åland), Italy, Portugal (Azores and Madeira), and the United Kingdom (Northern Ireland, Scotland, and Wales) do not have a veto, but they can voice their concerns to their national governments before the latter expresses the member state opinion in Brussels (Högenauer, 2008;Hooghe et al., 2016;Lanceiro, 2015).
Although these regions have a formal role in determining a member state position to be taken up in the Council of Ministers this may not translate into a strong impact of regions on EU decision making. Each member state representative should be able to commit the government of their member state because each member state has only one vote. 12 This requirement implies that there has to be one common position among 9 Austrian, 5 Belgian, 16 German, 21 Italian, and 17 Spanish regions. This often results in a fairly low common denominator (Jeffery, 2007). Nevertheless, research reveals that regions which have the opportunity to meet with their central governments to discuss EU affairs prefer to mobilise their rights through rather than beyond the established structures of the member state (Börzel, 1999;Jeffery, 2000;Keating, 2008).
The Treaty of Lisbon introduced the early warning system for the principle of subsidiarity. 13 This is a legal instrument for national parliaments to review EU legislation before it is discussed and enacted at the EU level. Parliaments have a chance to raise objections and prevent EU action when they have the opinion that the EU legislative proposal violates the subsidiarity principle (Abels, 2015). Each member state has two votes which are both assigned to unicameral national parliaments or divided between the lower and upper chamber in the case of bicameral national parliaments. 14 Subsidiarity concerns may also involve regional autonomy and 'it is for each national parliament to consult, where appropriate, regional parliaments with legislative powers'. 15 National parliaments are required to consult regional parliaments on those matters which are encompassed by a region's legislative powers and thereby regional legislatures can raise objections against breaches of the subsidiarity principle for those powers that are within their competences. 16 Seen from a perspective that focuses on the EU legislative process, the early warning system does not provide regions much leeway to impact on EU affairs even for regions in the member states which consist fully or in large part of regional representatives appointed by regional executives or regional legislatures. 17 In addition to the need to find agreement among up to one or two dozen regions, agreement also need to be found with the lower and upper chambers of national parliament and this needs to be achieved within the challenging time frame of 8 weeks. In addition, the role of the parliaments of the regions with legislative powers 18 is restricted to consultation. Despite these reservations, most regional parliaments have set up EU affairs committees when they did not have one previously, many regional parliaments have changed their (internal) regulations or rules of procedure, and some regional parliaments negotiated inter-institutional agreements on information rights and scrutiny power regarding EU affairs with their executives (Abels and Eppler, 2015;Borońska-Hryniewiecka, 2017; Committee of the Regions, 2013). These reforms may eventually lead to an increased influence of regional parliaments on EU affairs through an enlarged ability to hold regional and national executives accountable.

Multilevel governance in the three-tiered EU polity
The above overview on regional involvement in EU decision-making processes at the EU level and within the member states reveals that regions have become an unextractable and important tier of government within the EU multi-level polity. Regions have not gained much traction on EU affairs through an increased legislative role within the EU decisionmaking process in Brussels. However, such a perspective does not recognise the involvement of regions in the EU policy making process beyond their formal legislative role and within the member states. Hooghe and Marks (2001) published their landmark study (2001) during the heydays of the idea of a 'Europe of the Regions' -that is, a vision that foresaw development of a EU polity in which the EU and the regional levels would gradually gain ever more competences at the expense of the member-states which might even eventually disappear (Loughlin, 1996;Madrino, 2008). The EU clearly did not develop into a federal polity consisting of a federal government at the European level and state government at the regional level. Therefore, several scholars prefer to describe the EU as 'Europe with the Regions' rather than 'Europe of the Regions' (Hepburn, 2008;Hooghe and Marks, 1996;Moore, 2008). The main contribution of the concept of multilevel governance is that it helps to describe as well as to explain the development of a 'Europe with the Regions' (Piattoni, 2010).
Liesbet Hooghe and Gary Marks attribute increasing multi-level governance in Europe to two developments: European integration which has shifted authority from national states to European-level institutions and regionalization which has shifted political authority from the national level down to subnational levels of government (Hooghe and Marks, 2001: xi). Continuing processes of European integration and regionalization during the 2000s and 2010s have further increased the interdependency between regional and national governments and the EU institutions. Recent estimates reveal that 60% of the decisions taken by local and regional authorities are influenced by European legislation and nearly 70% of EU legislation is implemented by local and regional authorities (Committee of the Regions, 2009a;Council of European Municipalities and Regions, 2016). The Commission has much interest in making sure that EU policy is appropriately implemented in the regions and it strongly induces member states to share authority with their regions when they implement EU policy. Regions have much reason to influence EU legislation from the initial stages onwards and especially regions with a large policy portfolio -that is, regions which more often are required to implement EU law -have a higher inclination to monitor EU affairs and are keener to make the EU institutions aware of regional preferences regarding EU policy (Donas and Beyers, 2013;Hooghe and Marks, 1996;Noferini, 2012). Regions which seek to protect their competences from EU interference have come to learn that their interests are best served by collaborating with their central governments through participating in intergovernmental meetings on EU affairs (Börzel, 1999). Regions with access to central state decision making are less prone to bypassing their member states and less often directly seek contact with EU institutions. But these regions will resort to bypassing strategies and directly contact EU officials in cases when they are dissatisfied or disagree with the national position (Högenauer, 2014;Jeffery, 2007;Tatham, 2010Tatham, , 2014Tatham, , 2017. A multilevel governance perspective reveals that a major challenge within the threetiered EU polity is to coordinate governments across member states and regions as well as between regions within member states (see also Hooghe and Marks, this issue). Deepening European integration and continuing regionalization have significantly increased the overlap of competences between governments and an action or decision at one level is increasingly more likely to have an impact at another level. Coordination across levels of government is necessary to improve policy outcomes and effective coordination depends on the willingness of governments to share authority (Allain-Dupré, this issue; Charbit, this issue). Multilevel governance directs scholarly attention to the means through which governments try to achieve coordination. Typical examples of research questions raised by multilevel scholars are when and why regions are present and active in Brussels, when and why EU institutions provide a listening ear to them, when and why national governments are willing to co-determine member state positions with their regions, and when and why regions prefer to work through rather than beyond central state institutions. In other words, knowing the incentives for regions to be involved in EU affairs and for national governments and EU institutions to share their authority with regions is key for understanding the EU multilevel governance system.

Author's note
A previous version of this commentary was presented during the MLG 'breakthrough' political science symposium held at the European University Institute in Fiesole, Italy on 23 May 2019. I am grateful to Liesbet Hooghe, Charlie Jeffery, Gary Marks, and John Peterson for organising the symposium.
16. Full detail on the early warning system for the principle of subsidiarity can be found in Kiiver (2012). 17. The upper chambers in five out of eight member states with legislative regions -that is, Austria, Belgium, Germany, Italy, and Spain-consist fully or in large part of regional representatives appointed by regional executives or regional legislatures. 18. There are 74 regional parliaments in eight member states which possess legislative powers: nine Austrian Landtage, five regional and communal parliaments in Belgium, the parliament of Åland in Finland, 16 Germany Länder parliaments, 19 parliaments of ordinary and special regions plus two parliaments of two special provinces in Italy, the parliaments of Azores and Madeira in Portugal, the 17 parliaments of the comunidades autónomas in Spain, and the parliaments of Northern Ireland, Scotland, and Wales in the United Kingdom (Committee of the Regions, 2013).