THE DYNAMICS OF FEDERALISM: BELGIUM AND SWITZERLAND COMPARED

Transcription

1 Working Paper No. 138 April 2014 THE DYNAMICS OF FEDERALISM: BELGIUM AND SWITZERLAND COMPARED Jan Wouters Sven Van Kerckhoven Maarten Vidal

2 THE DYNAMICS OF FEDERALISM: BELGIUM AND SWITZERLAND COMPARED Jan Wouters Sven Van Kerckhoven Maarten Vidal ABSTRACT In recent years, several countries have moved to a more federal state structure. In particular in Europe, regional decision-making is on the rise. In order to evaluate these dynamics, this paper looks into the cases of Belgium and Switzerland. Both countries share a large number of characteristics, but have witnessed moves in opposite directions regarding the decentralization of decision-making powers. Whereas Belgium has increasingly hollowed out its central decisionmaking, Switzerland s regional powers have transferred more competences to the central level. This paper argues that these differing dynamics are the result of socio-economic variables. In particular, the decision to (de)centralize is the result of a trade-off between heterogeneity of preferences, and economics of scale. The former has been increasing in Belgium, resulting in more decentralization, while the latter has pushed the Swiss cantons to centralize more competences. KEYWORDS Federalism, Belgium, Switzerland, centralization, decentralization AUTHORS Jan Wouters is s Full Professor and Jean Monnet Chair ad personam, Director, Leuven Centre for Global Governance Studies Institute for International Law, KU Leuven and Visiting Professor, College of Europe (Bruges), Sciences Po (Paris) and Luiss (Rome). Sven Van Kerckhoven is a PhD candidate and Research Fellows at the Leuven Centre for Global Governance Studies Institute for International Law, KU Leuven. Maarten Vidal is legal advisor at the Flemish Department of Foreign Affairs. ADDRESS FOR CORRESPONDENCE Jan.Wouters@ggs.kuleuven.be Sven.VanKerckhoven@ggs.kuleuven.be 2014 by Jan Wouters, Sven Van Kerckhoven and Maarten Vidal. All rights reserved. No portion of this paper may be reproduced without permission of the authors. Working papers are research materials circulated by their authors for purposes of information and critical discussion. They have not necessarily undergone formal peer review. 2

3 THE DYNAMICS OF FEDERALISM: BELGIUM AND SWITZERLAND COMPARED Jan Wouters Sven Van Kerckhoven Maarten Vidal TABLE OF CONTENTS 1. Introduction Centralization versus Decentralization: Socio-economic Theory Federalization in Belgium and Switzerland: political and legal realities Belgium s rising regional powers... 6 History Reasons for federalization and how it came about... 6 Structure and competences... 8 Party-political structure...10 Belgium s federalism and the European Union Swiss traditional federalism...11 History Reasons for federalization and how it came about...11 Structure and competences...12 Party political structure...13 Switzerland s federalism and the European Union What do the theory and the reality teach us? Conclusion...15 References

4 THE DYNAMICS OF FEDERALISM: BELGIUM AND SWITZERLAND COMPARED 1. INTRODUCTION Jan Wouters Sven Van Kerckhoven Maarten Vidal In recent decades, several countries have moved towards a more decentralized state structure. This has resulted in an increase of decision-making powers, including legislative competences, at subnational levels. Although the move to decentralization in decision-making has taken place all around the world, the recent rise of decision-making at the sub-state level has first occurred in Europe. This can be explained from an historical perspective. European countries borders have shifted continuously throughout history. Large European empires and states have broken up and smaller countries have been incorporated in larger entities. Recent break-ups of nations (or aspirations to break up an existing state) have come in all colors and varieties. The roots of these separatist movements often reside with a people within a country that stresses the need for a higher level of self-rulemaking. The struggle of people to achieve autonomy in decisionmaking has taken place both violently (for example Northern Ireland) and rather peacefully (for example Catalonia). A potential solution to accommodate these demands is the construction of a federalized state, which is the object of study in this paper. As elaborated below, the decision to centralize or decentralize decision-making powers critically depends on the heterogeneity of preferences and the economics of scale. On the one hand, groups within a nation state can feel that they cherish similar preferences and share some sense of homogeneity with each other but not, or much less, with other inhabitants of the same country. On the other hand, economics of scale and scope might push centralization. In the next section, we discuss the trade-off between centralization and decentralization in more detail. We subsequently look deeper into the cases of Belgium and Switzerland. Both countries are instances of a European state which is weaker than traditional nation states, and where more decision-making powers are exercised at the regional level as a result of both heterogeneous preferences of strong societies and economics of scale and scope. As will be seen, the main difference between them are the opposite starting points and directions of the federal dynamics: in the Swiss case, federalism was the starting point after the aggregation of previously independent cantons, whereas Belgium started as a unitary nation state which gradually saw an ever stronger disaggregation. The paper ends with a short conclusion. 2. CENTRALIZATION VERSUS DECENTRALIZATION: SOCIO-ECONOMIC THEORY Centralization is the transfer of competences from a lower-level entity (in this paper regions ) to a higher-level entity (in this paper the state ). Decentralization is the opposite evolution. 1 For the purposes of this paper, federalization is the creation of a state structure whereby decisionmaking powers, including legislative powers, are shared by the state and a number of regions, encompassing the entire territory of the state. For the definition of the term region we build upon VAN LANGENHOVE (2013) and refer to regions as a governance unit that has some statehood 1 Decentralization can take place in various other forms as well. There can be e.g. a mere territorial deconcentration of services, a decentralization of only executive competences or the granting of autonomy to a limited number of entities while maintaining centralized government for other parts of the nation. 4

5 properties. In this paper, regions are defined more restrictively as sub-national entities, i.e. entities at a governmental level directly underneath the national level. 2 Our analysis builds upon the framework of ALESINA and SPOLAORE (2005) whereby the centralization versus decentralization debate boils down to a trade-off between heterogeneity and economics of scale and scope. Generally, one can think of heterogeneity in two dimensions: geography and ideology. The geographical dimension requires jurisdictions to be geographically compact for the provision of a certain good. Transaction costs render it virtually impossible for a Belgian to set up an educational system with an Australian, even though both might perfectly agree on educational policies. The ideological distance represents how far a policy is from an individual s preferences. The ideological and the geographical dimension generally go hand in hand. People living close to one another tend to have similar preferences, through self-sorting and common historical experiences. Due to this heterogeneity, devolving decision-making powers to the regional level might provide more distant citizens with a higher utility than centralizing them. In particular, people will have more diverse preferences in culturally or linguistically diverse countries and when the different groups are concentrated in regional areas. Moreover, regional governments have better information about local conditions and preferences, and can consequently make better informed decisions, thereby increasing welfare (HAYEK, 1945). This is particularly clear as far as education and cultural issues are concerned. The counterweight to heterogeneity are economies of scale and scope. Public goods can only be provided at a certain cost, which includes some fixed cost. Sharing the burden of this fixed cost among a larger number of taxpayers will render public good provision cheaper. Consequently, centralization will result in cheaper public good provision. Clearly, there are some issues for which economies of scale are so important that providing them at the regional level would not make sense. Military defense and the protection of the internal market through a common contractual environment give rise to large economies of scale and should hence be assigned to the central level. Another reason to centralize is that the central level may be better equipped to deal with certain interregional externalities (OATES, 1972). Coordination might then solve the beggar thy neighbor problem. Economies of scope refers to jurisdictions providing multiple public goods. In general, it will be cheaper to have a certain entity providing multiple public goods. Certain costs, such as establishing an administration, only have to be incurred once in order to allow for the provision of multiple public goods. These costs can then be shared among the different public goods. Additionally, it can be argued that a single policy for the whole federation can result in a feeling of fairness. Redistribution between richer and poorer regions provides some kind of federal insurance, whereby regions can interchangeably support each other. Federalism without centrally-planned redistribution schemes may increase inequality among the regions (although this is not always the case, see QIAN and WEINGAST, 1997). It is also important to note that in the ALESINA and SPOLAORE (2005) framework, it is imperative that decision-making competences and funding for implementing them should go hand in hand. If the expenses are on the regional level, the funding hereof should also be generated at the regional level. 2 We do not pay attention here to supranational regions, cross-border regions or local self-government. 5

6 Last but not least, a federalized structure increases the number of relevant actors in decisionmaking, as subnational entities will also have decision-making powers. Subsequently, within a decentralized state, the decision to centralize competences will become more arduous since more actors will be able to veto a transfer to the central authority. When preferences are heterogeneous at the regional level, and decision-making is initially assigned to this level, it might consequently be the case that the decision-making remains at the regional level, even though centralization would result in significant economies of scale and scope. 3. FEDERALIZATION IN BELGIUM AND SWITZERLAND: POLITICAL AND LEGAL REALITIES From a superficial point of view, Belgium and Switzerland are rather similar. Both are small nations (11 million vs. 8 million people) at the intersection of two great cultural zones in Europe (Latin and Germanic). This is reflected in their respective language set-up. Both countries have multiple official languages: Dutch, French and German are official languages in Belgium, whereas Switzerland has French, German, Italian and Romansh as official languages. 3 Both countries did not achieve unification in pre-napoleonic times and both currently have a federal state structure. Despite their small size, the two countries are economically prosperous 4, even though Belgium s external sovereign debt rate is considerably higher than that of Switzerland. 5 Yet, the federal dynamics in both countries differ on a significant number of points, as will be made clear below. 3.1 BELGIUM S RISING REGIONAL POWERS History Reasons for federalization and how it came about In pre-napoleonic times, the area currently constituting the territory of the Kingdom of Belgium, then known as the Southern Netherlands, had been for centuries a conglomerate of independent feudal entities, most of which were bound together in a personal union under Hapsburg rulers. For a brief period in 1790, these entities managed to depose their feudal overlords and established a confederation which was known as the United States of Belgium. Yet, the Hapsburgs returned, yielding to the French revolutionary armies only a couple of years later. The area was incorporated in the French Empire under Napoleon, and united with the Northern Netherlands in a united kingdom from 1815 until When the country gained independence through secession from the Netherlands in 1830, it was organized as a unitary state, with local self-government for its nine provinces largely based on the territories of the previous feudal entities and many communes. 6 Legislative powers were exercised at the central level only. 163 years later, in 1993, Belgium was declared a federal state, as part of the fourth state reform. 7 The sixth state reform of further regionalized certain competences. This has In both countries, the smallest of these languages, i.e. German in Belgium and Romansh in Switzerland, in both cases spoken by approximately 0.7% of the population, enjoys official status in the area where it is traditionally spoken and in contacts of the federal government with citizens speaking the language, but laws and official acts do not necessarily have to be adopted or promulgated in it. According to the IMF s figures for 2012, Switzerland had a GDP (PPP) per capita of 44,864 USD whereas Belgium s GDP (PPP) per capita amounted to 37,459 USD compared to Article 31 of the original constitution of 1831 (currently article 41 of the Constitution) already provided that the provinces and communes could regulate matters of provincial or communal interest. The Belgian constitution has been amended a number of times, more in particular to gradually introduce universal suffrage, in the period until the 1960s. The six state reforms refer to the gradual process which transformed Belgium into a federal state from that period on. The first state reform of 1970 introduced the 6

7 resulted out of a gradual devolution process, that has gathered pace since the 1960s (FITZMAURICE, 1996). Hence, Belgium s transformation into a federal state did not come about as a result of a merger of various states or other entities, unlike most federations. The current Regions are relatively homogenous units in a heterogeneous state. They are quite different in terms of population, socio-economic development and native languages (SWENDEN, BRANS and DE WINTER, 2006). The first reforms towards a federalist structure took place in This was mainly to ease the linguistic conflict by constitutionally entrenching powersharing between the two main language groups. However, apart from the language communities, it also entrenched the Regions, entities with a clear territorial basis (HOOGHE, 2004). Further reforms deepened the federal character of Belgium, which was, as indicated above, officially recognized in the Belgian Constitution of Decentralization to the Regions was not only meant to attenuate tensions between the two main language groups but also intended to enhance policy-making, facilitate policy reform and improve the effectiveness of public policies. This evolution from a unitary state to a federal structure was caused by several factors. The first one is the language issue. At first, Belgium s cultural diversity was not reflected in its institutional set-up. This was mainly because at that time the Flemish economic elite and middle class were closely involved in French-speaking social and cultural life (BILLIET et al., 2006). The elites were predominantly French-speaking, but this has been reversed in contemporary Belgium (HOOGHE, 2004). The Belgian Constitution of 1831 granted linguistic freedom in the private sphere, whereas the law could establish an official language for the public sphere. In practice French became the only official language. In the decades following the 1830 independence, the Flemish cultural elite increasingly advocated a more important role for Dutch. Nonetheless, it took Belgium a long time before Dutch became an official language, resulting in serious grievances. These continued even after Dutch gained an equal footing with French. 8 As noted by SWENDEN and JANS (2006): Several Flemish activists who had been socialised into the more linguistically averse climate of the inter-war period were not willing to reconsider their loyalty visà-vis the Belgian state despite the latter s accommodation of their initially non-territorial grievances. The increasing linguistic homogeneity in the Flemish and Walloon Region over the last century has further spurred the divergence of both regions. The fact that Dutch-speakers and French-speakers are to a large extent geographically segregated has certainly facilitated the transformation of Belgium in a federal state (HOOGHE, 2004). 9 Currently, the Flemish and 8 9 concept of Cultural Communities (Flemish, French-speaking and German-speaking) with legislative powers for the two main Cultural Communities in a number of fields (in essence cultural competences), as well as some of the current consociational characteristics of the central governmental organs (parity between the two main language groups in the Council of Ministers, alarm procedures and special majority laws in parliament). It also announced the creation of three Regions. The second state reform of 1980 widened the competences of the Cultural Communities beyond the cultural sphere, which was reflected by the fact that from then on they were simply called Communities, and created two Regions, the Flemish Region and the Walloon Region, with socioeconomic legislative competences. The third state reform of transferred almost all educational competences to the Communities and created the bilingual Brussels Capital Region, with power-sharing between Dutch-speakers and French-speakers. The fourth state reform of 1993 transformed Belgium in a federal state. The fifth state reform of devolved further competences to the regions (foreign trade, agriculture, organization of local government). The sixth state reform of reformed the financing of the Communities and the Regions and devolved further competences, for the first time also involving parts of the social security system. See CULLEN, 1990 for an-indepth investigation of the first three reforms and SWENDEN, BRANS and DE WINTER (2006) for more on recent reforms. Although officially this was already achieved with the equalization act of 1898, in reality French remained the main official language until much later. This is in particular the case since the 1963 law dividing Belgium into 4 language areas. A number of Frenchspeaking municipalities lying in Flemish provinces were transferred a the neighboring Walloon province and 7

8 French-speaking media are almost completely living apart, resulting in increasingly diverging cultures. This trend was institutionalized and enhanced by subsequent state reforms. Culture and education were the first competences to be transferred to the Communities, leaving the central authority only little policy instruments to promote a shared Belgian culture or identity. In particular in Flanders, this has resulted in a larger group of people with a predominantly regional identity (although this is often perceived to be complementary to the identification as Belgian). The result is a political attitude in Flanders supporting increased decentralization (BILLIET et al., 2006). Furthermore, the ideological conflict has been increasingly embedded along the linguistic divide, which has resulted in a different party political landscape in the respective regions (see below) (BILLIET, MADDENS and FROGNIER, 2006). Secondly, after the Second World War, the rapid economic growth of Flanders and the difficult conversion from an industrial economy based mainly on steel and coal in Wallonia, put more pressure on the unitary state structure. The result of this cultural and economic divide was an initial focus of Dutch-speaking Belgians on language and cultural issues, and an initial focus of French-speaking Belgians on socioeconomic issues. This partly explains the complex Belgian political patchwork, with the Communities focusing on the former issues and the Regions taking up the latter. As a result, French-speaking Belgians pay more attention to the Regions, while Dutch-speaking Belgians tend to focus more on the Communities (SWENDEN and JANS, 2006). Structure and competences Nowadays Belgium is composed of two types of federated entities, viz. the Communities and the Regions. 10 This odd federal set-up is due to the language pattern and results in a patchwork of overlapping jurisdictions. In Belgium, there are four language areas 11 : the Dutch, the French and German language areas, and the bilingual (Dutch and French) capital. These form the territorial basis for three Communities 12 and an equal number of Regions. 13 The Flemish Region and Community are based on the Dutch language area in the north, while the Walloon Region and the French Community are based in the French-speaking south. Brussels, the bilingual capital is a separate Region, but not a Community. Indeed, the Flemish and French Communities are each responsible for their respective language-based decision-making in Brussels, with complex institutional machinery for aspects that cannot clearly be attributed on a linguistic basis. The German-speaking Community is based in the German-speaking east of Belgium but it was deemed too small to be a region by itself, and therefore forms part of the Walloon Region Even in-between the different Regions and Communities, de jure asymmetries are present (DE WITTE, 2005). In practice, none of them has exactly the same powers. For example, the small German-speaking Community does not possess the same powers as its Flemish and Frenchvice-versa. Furthermore, due to socio-linguistic shifts, French ceased to be the language of the Flemish economic elite. Article 1 of the Belgian Constitution. Article 4 of the Belgian Constitution. Article 2 of the Belgian Constitution provides that there is a Flemish Community, a French Community and a German-speaking Community. Article 3 of the Belgian Constitution provides that there is a Flemish Region, a Walloon Region and a Brussels Capital Region. 8

9 speaking counterparts. 14 The same holds true for the Brussels Capital Region, whose powers are at the same time more limited and more extended than the Flemish and Walloon regions. 15 The asymmetry is further strengthened by the imperfect congruence between regional and community borders (SWENDEN, 2002). The organs of the Flemish Community, i.e. the Flemish Parliament and the Flemish Government, also exercise the competences of the Flemish Region, while such an institutional merger did not happen between the Walloon regional and French community institutions. To the contrary: the French Community has transferred a number of its competences to the Walloon Region and to the French Community Commission in Brussels, an organ made up of the French-speaking members of the Parliament of the Brussels Capital Region. SWENDEN and JANS (2006) argue that this can be attributed to the fact that only 3% of the total Dutch-speaking group lives in Brussels, while 18% of the French-speaking Belgians live in Brussels. Moreover, the Dutch-speaking Belgians share more socio-economic preferences than the French-speaking Belgians living in Brussels and Wallonia. The Regions and Communities are autonomous institutions and have executive legal powers, as granted to them in the Constitution or by virtue of it. The Regions have autonomous powers in mainly economic fields such as environmental protection, regional development, housing, employment, public works, agriculture and transportation, as well as in the field of the organization of local government, while the Communities have autonomous powers in issues related to culture, education and personal social services. The Communities and Regions can cooperate amongst each other, but also with foreign states, including the conclusion of treaties (art. 167 of the Belgian Constitution). The division of powers has been delineated in the Special Act on Institutional Reforms, which was adopted in 1980, and has been adapted numerous times afterwards. All issues not devolved to the Regions and Communities remain federal competences. In a succinct list one can summarize the federal competences as labor law, civil law, trade law, defense, justice, security, social security, fiscal and monetary policy. 16 Competences were drafted to overlap as little as possible and decisions of the federal State, the Regions and the Communities are on equal footing. As expected, there is still considerable overlap in areas such as health and labor market policies, and to a lesser extent in infrastructure and education (SWENDEN and JANS, 2006). A neutral Constitutional Court guards the division of competences and the existing institutional outlook, and also protects the fundamental rights and liberties. Regions have gained considerable spending autonomy, but very little fiscal and taxation autonomy (SWENDEN, 2006), even though this will increase substantially due to the latest state reform. Communities depend fully on federal grants. This is a very important characteristic. As indicated by the theoretical framework in the first section of this paper, the problem is that if there is no correspondence between the collection and the expenditure of taxes, this often results in fiscal imbalances. Due to the federal nature of social security and the regional nature of issues such as employment policy, there exists a mismatch in funding and (the consequences of) decision-making. This could allow a Region to free-ride. However, with full In contrast to the two other Communities, it is not competent for language legislation (except for the use of languages in education). On the other side, due to the fact that the Walloon Region has transferred some of its competences (immovable cultural heritage, archeology, employment, local government, tourism) to the Germanspeaking Community, this Community also exercises regional competences. Whereas the Flemish Region and the Walloon Region adopt decrees, which are legally equivalent to federal laws, the Brussels Capital Region adopts ordinances, which are subject to a higher standard of judicial review. As far as competences are concerned, the sixth state reform devolved a number of cultural competences of regional importance to the Brussels Capital Region, thereby blurring the distinction between the two types of competences. Monetary and fiscal policy are however to a certain extent EU competence now. 9

10 correspondence there will be very little, if any, room for re-distributional transfers, which are relatively large in Belgium. Lastly, Belgium s federal Government has very limited powers to coordinate policies of the Regions and Communities. It only has a power of substitution in the case of the condemnation of the Kingdom of Belgium by the European Court of Justice for the non-compliance of regional or community legislation with European Union law. 17 There are also a number of situations in which a government mandatorily has to consult with other governments. In most cases cooperation amongst them is voluntarily, and there are over 100 cooperation agreements and a number of permanent structures facilitating this cooperation. Party-political structure Political parties have also jumped on the regional bandwagon. Until the 1970s, Belgium was a classic example of a consociational partitocracy, whereby parties connected societal segments with decision-making among elites (DESCHOUWER, 2002, 2006). The secularization has allowed ideological conflicts to ebb away. The political parties, divided along linguistic and territorial lines, began to co-opt nationalist demands and responded by reorganizing into regional wings. Consequently, the Belgian party system has gradually been broken up into a Flemish and a French-speaking party system. The traditional parties (Christian-Democrats, Liberals and Social-Democrats) all broke up along linguistic lines between 1968 and This allowed the formerly unitary parties to pursue more regional strategies (BILLIET, MADDENS and FROGNIER, 2006). A result of this breakup is that party competition is limited to the respective linguistic territory. Bilingual Brussels is the only place where the Flemish and French-speaking parties do not live in different worlds. The linguistically divided political parties in general still form a family along ideological lines with their other-language counterparts, but only the two green parties (which were established in the 1980s and have never constituted a single party) form a single faction in the federal parliament. The other parties are only very loosely linked and do not coordinate extensively (SWENDEN (2002). Yet, parties from the same political family have relatively similar party platforms in both regions. Since there is no federal constituency, parties from the same family often get very different electoral results in both language groups. However, when it boils down to government formation, the parties generally get their counterparts on board. The result is that it has been rather difficult to find Belgian governments not including the Social-Democrats and the Christian-Democrats, as both party families are electorally strong in respectively Wallonia and Flanders. Due to the inclusion of their regional counterparts, the Belgian federal government often consists of oversized coalitions. Initially, the regional Parliaments and Executives were composed of federally elected politicians who acted in two capacities. Direct election of all regional politicians was only implemented in Moreover, until 2003, the federal and regional elections were held on the same day. This logically resulted in relatively similar outcomes on both levels. The latest state reform provides for a return to elections that are held simultaneously. Federalization in Belgium has been a top-down process in which the federal parliament has decided to increasingly transfer competences to the Regions and Communities (SWENDEN and BRANS, 2006). This was a means for the political elites to design the rules in ways allowing them to keep their positions (HOOGHE, 2004). They could easily do so thanks to the strong societal entrenchment of the traditional parties in the consociational system (DESCHOUWER, 2002). Each Article 16, 3, of the Special Act on Institutional Reforms. The Parliaments of the Brussels Capital Region and of the German-speaking Community have been directly elected from the start. 10

11 state reform was the result of a political deal between parties representing a majority in each of the two main linguistic groups in the federal parliament. From an institutional point of view, the Regions and Communities themselves were not involved in this process. Belgium s federalism and the European Union Belgium is a Member State of the European Union, and hence faces the risk of being hollowed out from below (through its Regions) and above (through the European decision-making level). However, it has also been argued that these two evolutions might strengthen the central authority by taking away the less gratifying functions from the state (KEATING, 1995). DE WITTE (2005) argued that the European integration process initially weakened regional decisionmaking, by taking away from it some decision-making powers. The same happened to central governments, but in this case, the latter still had a saying through the Council and European Council. Regions were initially completely left out of European decision-making. Only with the Maastricht Treaty, subnational governmental levels were given a role in the European decisionmaking processes. 19 Internal mechanisms have been set up within Belgium in order to define Belgian positions on European affairs. 20 Since Belgium was a founding member of the European Communities, its membership predates the federalization of the country. Existing EU primary (i.e. treaty-level) and secondary law (i.e. legislation) restrains the legislative autonomy of the Communities and the Regions, and narrows the margins for diverging policy options. EU law has been successfully relied on in disputes between the country s federated entities (WOUTERS and VIDAL, 2013). At the same time, the existence of a European legislative framework which has to be respected can take away fears for the devolution of further policy areas. Interestingly, the process of European integration has not eliminated the pressure for decentralization in Belgium. It has been argued that it even makes a full separation less costly. That explains why nationalists claim to be pro-european, a phenomenon that can also be witnessed in Scotland and Catalonia. 3.2 SWISS TRADITIONAL FEDERALISM History Reasons for federalization and how it came about Switzerland was founded around 1300 as a confederation by three founding cantons in the High Alps (Uri, Schwyz, and Unterwalden) with the Federal Charter (Bundesbrief) of Nowadays, it consists of 26 cantons, and it has been a federal state in its current form since Federalism in Switzerland constituted a response to a powerful security threat (GILPIN, 2001), i.e. a 27-day civil war, the Sonderbundkrieg. The war started when Roman-Catholic cantons formed a separate alliance ( Sonderbund ), in opposition to the centralization of power, which The current article 16.2 of the Treaty on European Union (which was introduced by the Maastricht Treaty) provides that the Council shall consist of a representative of each Member State at ministerial level, who may commit the government of the Member State in question and cast its vote. On the basis of this provision, especially Belgium and Germany, but also other Member States, have mandated regional ministers to represent themselves in the Council. Furthermore, the Maastricht Treaty created the Committee of the Regions. See the Cooperation Agreement of 8 March 1993 between the Federal State, the Communities and the Regions on the representation of the Kingdom of Belgium in the Council of Ministers of the European Union. 11

12 was the result of the electoral victory of the Radicals and the Liberals. The conservative Roman- Catholic cantons objected to proposals for more centralization and progressive reforms, resulting in the short war. The Protestants won, and the outcome was the new Swiss Constitution, drafted in 1848, ending the almost complete independence of cantons, and transforming Switzerland into a federal state (COOLIDGE, 1911; for more information, see CHURCH and DARDANELLI, 2005). A significant difference with the Belgian structure is that the Swiss cantons do not coincide with language communities. The cantons are based on jurisdictions predating the Napoleonic age, when language issues only played a limited role. The speakers of the two main languages of Switzerland, German and French, are divided over a number of cantons. Whereas there are 22 unilingual cantons 21, there are also 3 bilingual cantons 22 and even a trilingual canton. 23 These have remained relatively stable and there is only one example of a secession of part of a canton on a linguistic basis. 24 The media in Switzerland are, like in Belgium, also highly segmented by language (IPPERCIEL, 2007). Yet primary identification remains with the Swiss nation, rather than with the cantons or the language groupings (DARDANELLI, 2008). Structure and competences The original Swiss Constitution was adapted several times, and completely replaced in Nowadays, Switzerland is officially a confederation of 26 cantons. 25 The term confederation has its roots in a history of cantons willing to keep their decision-making powers, but in practice Switzerland is a federal republic. There are three levels of government: the federation, the cantons, and the communes. The cantons powers have over time slowly eroded. However, they still exercise a significant policymaking and revenue raising role (full taxation power), within the limits of the federal constitution. Also, the cantons are still responsible for policy implementation, giving them a certain leverage. Even though there are significant differences in size and political power of the cantons, which can be seen as a form of de facto asymmetric federalism 26, all of them enjoy the same de jure rights, and even the same voting power and representation in the upper house 27, the Council of States (DARDANELLI, 2010). Cantons are formally involved in all phases of federal law-making, although the extent to which they exercise real influence can be highly variable (VATTER, 2004). Federal government s role has expanded in recent decades. The federation is currently in charge of most decisions, but issues such as culture and education still reside under the cantons flag. Indeed, all issues not There are 17 German-speaking cantons, 4 French-speaking cantons and 1 Italian-speaking canton. Vaud (French and German), Valais (French and German) and Berne (German and French). Graubünden (German, Romansh and Italian). The canton of Jura was created in 1979 out of a number of French-speaking districts of the mainly Germanspeaking canton of Berne. Article 1 of the Swiss Constitution The least numerous canton, Appenzell Innerrhoden, has inhabitants, whereas the largest canton, Zürich, has inhabitants. For historical reasons, some cantons are considered to be half-cantons (Obwalden and Nidwalden; Basel- Landschaft and Basel-Stadt; Appenzell-Innerrhoden and Appenzell-Ausserrhoden), and they each only have one representative in the Council of States whereas the other cantons have two. 12

13 explicitly given to the federal government continue to be the cantons responsibilities (CHURCH and DARDANELLI, 2005). Interestingly, any amendment to the constitution, e.g. in order to increase the federation s powers, needs to be approved by a majority of the people and the cantons. This makes the increasing centralization even more striking, as this quite demanding requirement needs to be satisfied every time the federal level s powers are strengthened (CHURCH and DARDANELLI, 2005). Indeed, federalism provides a powerful institutional veto point (IMMERGUT, 1992). This has strongly constrained the maneuvering space of the Swiss federal government and increases policy stability (TSEBELSIS, 1995). When a policy competence is initially not assigned to the federal level, it will be difficult to transfer it to that level afterwards. Social policy for example has always been a competence of the cantons, but over time the federal level became responsible for inter alia health and occupational hazard insurance, family allowances and unemployment insurance. However, as these were originally canton responsibilities, their transfer was contested and time-consuming (OBINGER, 1998). Moreover, the centralization of additional policy-making powers is submitted to an endorsement of a popular referendum. Sometimes these referendums are obligatory (for amending the constitution or joining a supranational organization), and then not only a majority of the population but also a majority of the cantons has to agree. Referendums can also be optional (when signatures are collected in 100 days after a bill has been passed) or can be requested by at least eight cantons. Then a single popular majority is decisive. These differing requirements can also block the federalization of a competence. Party political structure The Swiss parliament, the Federal Assembly, has two chambers, both with equal rights. The National Council s 200 members are elected by nation-wide direct suffrage, and according to proportional representation. The Council of States 46 members are elected on the basis of the cantons and according to majority representation. Switzerland has a rich party landscape, but political parties have not been split up along linguistic lines and are still organized nationally. Each of the 26 cantons is a constituency. For the Council of States each Canton has one or two directly elected representatives. For the National Council, each canton elects between 1 to 34 members. It is clear that the electoral system in the Council of States favors the smaller cantons, while this is less the case for the National Council. Switzerland does not have a system of parity between language groups in its federal executive, the Federal Council. For a long time, its composition was based on the magic formula, in which its seven members were appointed by political parties in approximate relation to their respective strength in the United Federal Assembly, i.e. the joint meeting of both chambers. Since 2008, this formula has been abandoned. Apart from this ideological equilibrium, article of the Swiss Constitution provides for an equitable distribution of seats among the cantons and language groups, but does not set concrete quotas. In practice, five members have traditionally been German-speakers, whereas the two other ones are French-speaking or Italian-speaking. Switzerland s federalism and the European Union At heart, the idea behind the Swiss federal democracy rests on the belief that local is better than distant (CHURCH and DARDANELLI, 2005; DARDANELLI, 2010). This can also explain the traditional aversion to membership of the EU. The accession of Switzerland to the EU would have to be approved by not only a majority of its inhabitants but also a majority of its cantons. Until now, Swiss voters have not only rejected their country s participation in the European 13

14 Economic Area 28 and the start of accession negotiations with the EU 29, but also approved the popular initiative Against mass immigration, which effectively contradicts its obligations concerning free movement under bilateral agreements with the EU. 30 This already gave rise to countermeasures by the EU. It is unlikely that Switzerland will become a EU Member State in the foreseeable future. 4. WHAT DO THE THEORY AND THE REALITY TEACH US? Belgium has been considered as a prime example of federalization by disaggregation and it is currently very close to a confederal structure of constitutionalized units (SWENDEN, BRANS and DE WINTER, 2006). This is the result of the increasing cultural linguistic and socio-economic divide between its northern and southern parts. It has been noted that, in Belgium, federalization has proven to be an efficient way to deter conflicts along language lines but that at the same time it might not be durable (for more on this, see SWENDEN and JANS, 2006). In comparison, Swiss history displays a process of increasing centralization (for some examples in this history, see CHURCH and DARDANELLI, 2005). This trend was reflected in the more than 100 changes to the 1874 constitution in the course of the following century (CHURCH, 2004). However, since Switzerland initially was a very decentralized country, it is still relatively decentralized compared to other federations (MCKAY, 2001). Swiss federalism has therefore been the result of the decision of the cantons to join ranks. Interestingly, these tendencies continue to play out. The theory outlined in the first section helps to shed a light on these developments. In Belgium, the increasing linguistic and economic divide and segregation of the two biggest regions has resulted in more heterogeneous preferences. This posed a big challenge to the initially relatively centralized state. Decision-making in areas with higher heterogeneous preferences were increasingly transferred to the regions, while subject matters with larger economies of scale have remained at the federal level. However, even the latter are coming under pressure of being regionalized. This is, amongst others, a result of the split of political party families among linguistic lines. In order to face up to the rise of more nationalist parties, the traditional parties have become increasingly responsive to regional aspirations. Switzerland, which started out with only very few competences at the national level, is increasingly centralizing its decision-making. Over the last centuries, economies of scale have spurred cantons to transfer more competences to the federal level. At the same time, the Swiss national identity rose, lowering the heterogeneity of preferences. This has taken place even though the media have been divided along linguistic lines. The centralization process has been a time-consuming one, since cantons can block the transfer to the national level. An important difference with Belgium is that, while Belgium is mainly divided in two strong blocks along language lines, Switzerland is divided in 26 cantons, some of whom comprise at least two language blocks. In terms of competences, some subject matters characterized with high heterogeneity continue to be the cantons responsibility. Both countries have a vibrant civil society, which is also reflected in the numerous associations existing in each of them, yet they have a very different attitude towards direct democracy. Switzerland has a strong tradition of direct democracy at the local, regional and national, which, Referendum of 6 December Referendum of 4 March Referendum of 9 February

15 though sometimes exposing regional differences (the so-called Röstigraben), is also part of the Swiss national identity. In contrast, Belgium s only experience with a national referendum on the return of King Leopold III in 1950 is widely seen as a national trauma, due to the very different outcome in the three Regions. Even though non-binding, popular consultations have recently been made possible at the local and regional levels, similar consultations at the national level and binding referendums at either level remain a bridge too far. They would conflict with the tradition of the famous compromis à la belge, reached in nightly conclaves between parties and linguistic groups, and would also render Belgium s participation in the EU more cumbersome. In Belgium, due to its history of foreign rulers, the local level has continued to be an important player, and communities are strongly embedded, mainly at the regional level. Although Switzerland was not to the same extent subject to foreign rule, its mountainous geography has also allowed for a strong local entrenchment. 5. CONCLUSION Both Belgium and Switzerland are instances of a central state which is weaker than traditional nation states, because more decision-making powers are exercised at the regional level. As has been explained throughout this paper, their federal state structures are the result of both heterogeneous socio-economic preferences and economics of scale and scope. The big difference between both is the way federalism was achieved. In the Swiss case, federalism was the starting point after the aggregation of previously independent cantons. The Belgian case took an opposite road to federalism: it was the outcome of a struggle towards more disaggregation. In Switzerland, there is a tendency of transferring more competences to the federal level, while in Belgium, exactly the opposite evolution is playing out. One of the main differences is that Belgian politics are dominated by the interaction between the two main language groups, while in Switzerland language groups are dispersed across 26 cantons, which differ extensively in area, population, religion, socio-economic profile, and economic power. The Belgian traditional party families have also split up along linguistic lines, while such a development did not take place in Switzerland. Finally, and paradoxically, whilst the EU can be said to have been a factor that has in a sense facilitated the decentralization of competences in Belgium, Swiss federalism (and tradition of direct democracy) is a factor which makes it more difficult for the country to join the EU. At the same time, the complex bilateral relationship between the EU and Switzerland diminishes the leeway of the cantons to adopt legislation in fields covered by the bilateral agreements, and favors further centralization of powers within Switzerland. 15

16 REFERENCES Alesina, A. and Spolaore, E. (2005). The Size of Nations. The MIT Press. Billiet, J., Maddens, B. and Frognier, A.-P. (2006). Does Belgium (Still) Exist? Differences in Political Culture between Flemings and Walloons, West European Politics, 29(5), pp Church, C. (2004). The Politics and Government of Switzerland. Palgrave Basingstoke Church, C. and Dardanelli, P. (2005). The Dynamics of Confederalism and Federalism: Comparing Switzerland and the EU, Regional and Federal Studies, 15(2), pp Coolidge, W.A. B. (1911). Switzerland/History/Constitution, in Chisholm, H.(ed.) Encyclopædia Britannica, 26 (11th ed.). Cambridge University Press. p. 259 Cullen, R. (1990). Adaptive Federalism in Belgium, UNSW Law Journal, 13(2), pp Dardanelli, P. (2008). Communicative nation and Multi-Nationalism, Swiss Political Science Review, 14(3), pp Dardanelli, P. (2010). Federal Democracy in Switzerland, in Burgess, M. and Gagnon, A.G. (eds.) Federal Democracies. Routledge Series in Federal Studies. Routledge, London, pp De Witte, B. (2005). Regional autonomy, cultural diversity and European integration: the experience of Spain and Belgium, in Ortino, S., Zagar, M. and Mastny, V. (eds.) The Changing Faces of Federalism: Institutional reconfiguration in Europe from East to West, Manchester University Press. Deschouwer, K. (2002). Falling apart together. The changing nature of Belgian consociationalism, , in J. Steiner & T. Ertman (ed), Consociationalism and corporatism in Western Europe. Still the Politics of Accommodation?, Special issue of Acta Politica, vol 37, p Deschouwer, K. (2006). And the peace goes on? Consociational democracy and Belgian politics in the twenty-first century, West European Politics, 29(5), pp Fitzmaurice, J. (1996). The Politics of Belgium: A Unique Federalism. Hurst and Company, London. Gilpin, R. (2001). Global Political Economy: Understanding the International Economic Order. Princeton University Press. Hayek, F.A. (1945). The Use of Knowledge in Society, American Economic Review, 35, pp Hooghe, L. (2004). Belgium: Hollowing the Center, in Amoretti, U and Bermeo, N. (eds.) Federalism, Unitarism and Territorial Cleavages, John Hopkins Press. Immergut, E.M. (1992). Health Politics and Institutions in Western Europe, Cambridge University Press. 16