The Evolution of Sub-State Nationalism within the European
Integration Process
Master Thesis Political Science – Political Economy
Harald Moosleitner
June 2019
Lecturer: Dr. Mike Medeiros
Second Reader: Dr. Imke Harbers
Abstract
There is an emerging consensus that the European integration process alters the political arena for regional authorities. The multilevel governance structure of the European Union includes regional actors in the decision-making process. In addition, the supranational structure of the EU undermines the sovereignty of central states within Europe. As a result, regions become less dependent on their central states, particularly in economic terms. Regionalist parties are utilizing these developments to campaign for greater autonomy for their respective regions. However, the heterogeneity of the regionalist party family shows that regionalist parties have different attitudes towards the European integration process. Two competing hypotheses are formulated with the intention to test whether regionalist parties oppose or support the European integration process and the opportunities it entails. By evaluating expert survey data in a statistical analysis, this study concludes that regional parties have favourable orientations to the European integration process over time and across issue areas. The in-depth examination of regionalist parties in Catalonia and Bavaria give qualitative support for the quantitative claims. Interviews with party representatives of the major Catalan and Bavarian regionalist parties demonstrate the motivation for their Europhile nature. The quantitative and qualitative findings shed light upon the influence of the European integration process on sub-state nationalism.
Keywords: European integration; regionalist parties; centre-periphery cleavage, Bavaria,
Table of Contents
1. Introduction ... 1
2. Literature Review ... 3
2.1. Regionalist Political Parties in the European Integration Process ... 3
2.2. The European Union: A changing Political Environment ... 9
2.3. Centre-Periphery Cleavage as Theoretical Background... 11
3. Research Design... 19
4. Statistical Analysis... 23
5. In-depth Analysis ... 33
5.1. Regionalist Parties in Catalonia ... 33
5.2. Regionalist Parties in Bavaria ... 40
6. Conclusion ... 45
7. References ... 47
1
1. Introduction
With the Scottish independence referendum of 2014 and the controversial Catalan independence referendum of 2017, European soil has witnessed two regions that struggle for their own statehood just within the last five years. Even though both referendums did not lead to the formation of sovereign states, it seems that sub-state movements are on a general rise in Europe. Even in Bavaria, a region that is mostly excluded by scholars in the debate about emerging sub-state nationalism, the amount of people that want independence climbed to a record share of 30 percent (RT 2017, n. p.). Since these secessionist tendencies come hand in hand with the European integration process, the impression is given that the EU increasingly becomes an opportunity for stateless nations to gain their own statehood. Thus, it would be expected that the representatives of these regions, namely regionalist parties, are promoting their support for the European integration process. Particularly the support for economic integration is self-explaining, as it makes regions more sovereign and active players in economic terms. However, the results of the European elections in 2019 depict that with the Vlaams Belang in Belgium and the Lega Nord in Italy, two regionalist parties with Eurosceptical views could dramatically increase their share of votes in their respective countries (Financial Times 2019, n. p.). These results show that there are regionalist parties that successfully mobilize their electorate by opposing the process of European integration. Thus, it is not entirely clear how the European integration process affects European regions and whether their political representatives oppose or support it. Therefore, the research question “How does the European integration process influence sub-state nationalism?”, was developed to further examine the underlying topic. In general, this study demonstrates the altered opportunity structure for regions that comes with the European integration process. In fact, territorial self-governance becomes a more realistic objective for regions due to nation states transferring competences to the European level. In addition to this, the multilateralism of the EU includes regions in the decision-making process. By testing two competing hypotheses in a statistical analysis, it becomes apparent that regionalist parties support the European integration process. The in-depth analysis of the case studies of Catalonia and Bavaria gives qualitative support for this claim. Furthermore, interviews with party representatives of Catalan and Bavarian regionalist parties show the
2
intention of their Europhile nature. The collected evidence is used to verify the impact of the European integration process on regional nationalism.
The paper proceeds as follows: At first, the emergence of the regionalist party family within the European integration process is presented. Afterwards, the creation of European institutions for regional authorities is displayed to clarify the altered political arena that comes with the integration process. The theoretical background of the centre-periphery cleavage helps to predict whether regionalist parties oppose or support European integration. After formulating two competing hypotheses, a statistical analysis using expert survey data is conducted to test the hypotheses. This is followed by an in-depth analysis of the regional case studies Catalonia and Bavaria to support the quantitative results. Finally, in the conclusion the research question is answered, making this study a contribution to the literature on regionalist parties.
3
2. Literature Review
2.1. Regionalist Political Parties in the European Integration Process
When it comes to examining the current literature on the impacts of European integration on the party family of regionalist parties, it is important to mention the work of Peter Lynch, Lieven de Winter and Seth Jolly. Particularly in the beginning of the European integration process in the early 90s, when the European regions have increasingly been in the focus of academic attention, their contributions to the corresponding research field were essential (Lynch 1996, de Winter & Türsan 1998, Lynch & de Winter 2008, Jolly 2006 & 2007). However, the same cannot be said about regionalist parties, party politics and party alignments, as these topics received only little scholarly attention (de Winter & Gomez-Reino Cachafeiro 2002, 484).
In fact, the European integration process made regionalist parties a permanent feature of the European political landscape, therefore in the next steps the existing literature regarding the emergence of the party family of regionalist parties shall be analysed (Hepburn 2009, Lynch 1996, Elias 2008, Massetti 2009, Urwin 1983, Fagerholm 2016, Jolly 2006 & 2007).
In general, the notion of a regionalist party family was initiated by scholars in the 1980s where parties were grouped together based on their limited territorial appeal (Hepburn 2009, 480). According to Derek Urwin, the common feature of regional parties is that they “[restrict] their activities to only one or a few regions” (Urwin 1983, as cited in Fagerholm 2016, 307). The naming for the regional party family varies across the literature as until recently the terms ‘regional’ and ‘regionalist’ were used interchangeably by scholars. However, current authors tend to separate both terms: regional parties on the one hand are defined as parties whose policies, ideologies and electoral mobilisation are attached to interests of a certain territory or region. Regionalist parties on the other hand are characterized as subtype of regional parties that propagates demands for a certain degree of sovereignty or self-governance for their respective territories (Fagerholm 2016, 307). In short, if a regional party attempts to gain greater autonomy or independence in their region it can be defined as regionalist party. Regionalist parties can also be ethnic parties if their electorate majorly consists of an identifiable ethnic group (idem, 307-308). Furthermore, some authors use the notion ‘minority nationalist parties’ when referring to regionalist parties (Chaney 2014, Elias 2008,
4
Lynch & de Winter 2008) and other scholars combine terms which lead to notions, such as ‘ethnoregionalist parties’ (de Winter & Gomez-Reino Cachafeiro 2002, Lublin 2012, de Winter 2001). Other examples that express the variety of names given to the regionalist party family are the term ‘Stateless nationalist and regionalist parties’ (SNRPs) used by the scholar Eve Hepburn (2009) or ‘autonomist parties’ applied by de Winter et al. (2006). In fact, Hepburn (2009, 482) argues that ‘regionalist’ is the most used label for the party family, consequently this study follows this strategy by using predominantly the term ‘regionalist’ when specifically addressing sub-state parties.
The multiplicity of terms for defining the regionalist party family combined with the problem of using a uniform label when addressing regionalist parties prevented to some extend the conceptual development of the field. Consequently, these controversies regarding nomenclature of the party family resulted in the fact that some cases were simply omitted or not examined altogether, such as regionalist parties in East Germany and Bavaria (Hepburn 2009, 481). The failure of providing a uniform label for the regionalist party family can be explained by the difficulties of grouping these parties together into a single party family in the first place (Fagerholm 2016, Massetti 2009, Hepburn 2009, Mair & Mudde 1998, de Winter & Türsan 1998). The current debate agrees that the regionalist party family shows much more heterogeneity on the left-right ideological spectrum than other party families (Massetti 2009, 501-502). On the one hand, there are regionalist parties with left stances, such as the Esquerra Republicana de Catalunya (ERC) in Catalonia, on the other hand, there are regionalist parties, such as the Bavaria Party (German: Bayernpartei, BP) in Bavaria that campaign more for right positions. In fact, Mair and Mudde (1998, 221-222) go even further by suggesting the dispersion of the regionalist party family among other party families due to its crucial ideological diversity. Thus, the question arises why regionalist parties are grouped into a single party family? In order to accurately address this issue, it is necessary to describe the impacts of the European integration process on regions within Europe.
In general, the European integration process had specific impacts on reinforcing regional interests and identities and offered new possibilities to organize regional actors due to emerging political institutions (Jolly 2007 & 2006, Lynch 1996, de Winter & Gomez-Reino Cachafeiro 2002, de Winter 2001, Lynch 2007, Lynch & de Winter 2008). Therefore, the
5
emergence of European institutions representing regional authorities is illustrated in the following.
In the late 1980s, the idea of a Europe of Regions received increasing attention in the European Union, however, there was a crucial lack of regional actors included in the EU decision-making process. Within the debate of the European Union’s structural fund reform in 1988, the voices of regional governments that demand for institutional representation on European level became louder. These circumstances lead to the creation of the Committee of the Regions as a consultative body within the EU which consisted of members of regional authorities. Furthermore, the Maastricht treaty was signed in 1992, which gave regional governments the constitutional possibility to represent member state interests within the Council of Ministers. Thus, regional governments secured themselves a certain level of institutional representation on European level (Lynch 2007, 1-2).
These developments helped the regional governments to group together in further representative organisations, such as the Congress of Local and Regional Authorities in Europe (Lynch & de Winter 2008, 584). Apart from the mentioned institutions, there exist other network bodies for regional governments that support the mobilisation of regional agendas, namely the Conference of Presidents of Regions with Legislative Power (RegLeg) and the Conference of Presidents of European Legislative Assemblies (CALRE) (Bourne 2003, 349). Thus, it became apparent that regions within Europe were acting beyond their corresponding national state structure. In short, there were shifts towards multi-level governance because nation states were no longer single actors in European policy making (idem, 349-350). These institutional developments created possibilities for regions with legislative powers to address issues, such as the protection of constitutionally established powers of regional authorities. In fact, regional actors wanted to avoid encroachment by the European Union upon exclusive areas of competences of member states and particularly regions (idem, 350). In addition, another issue highlighted by Catalan regional authorities was the attempt to grant special rights to non-state cultural communities, such as Catalonia. However, in order to give legislative regions a special status, these non-state territories need to be put on the same level as nation states (idem, 352). Hence, many regional entities within Europe called for protection of non-state cultural communities from EU policies, thus, the idea of a Europe of Regions gained increasingly popularity among regional players. However, despite the
6
successful regional institutional developments in the EU, the ultimate objective of a Europe of Regions still seemed to be an unrealistic prospect for regional authorities. In general, the idea of a Europe of Regions propagates the unification of Europe as a federation of regions, while artificially constructed national states disappear from the political landscape (Lynch 2007, 2-3). Logically, the prospect of a Europe of Regions is not a popular option for central governments and their national states.
While regional institutions provide access for regional governments to European decision-making level, there exist non-state communities that are not represented in regional governments, consequently, their emancipation and institutional expansion is still required. In fact, there exist many regional governments even though there are no regionalist parties involved (Lynch 2007, 3). Nevertheless, regionalist parties organize themselves as members of the European Free Alliance (EFA), which acts as a transnational organisation that allows collective action of regionalist parties in the European Parliament (Lynch 2007, de Winter 2001, Bourne 2003, Lynch & de Winter 2008).
In the beginning of the European integration process, regionalist parties generally opposed European integration due to the widening gap between subnational (or regional) policies and supranational (or European) decision-making level (de Winter 2001, 6). However, with the expansion of Europeanisation of decision-making and the prospect of a Europe of Regions, regionalist parties hoped that the impact of European institutions reduces the power of nation states, as they must transfer competences to the European level. While state power decreases upwards, regional governments can restrict influences of national states downwards (Lynch 2007, 2-3). In general, the European integration creates a policy space that consists of multiple levels of government, namely subnational, national and supranational (Marks et al. 1996, 341-342).
The European Free Alliance was created in 1981 by six regionalist parties due to direct elections in the European Parliament (Lynch 2007, 3). It was formed to enable regionalist party initiatives at the European level and “to give a role in European politics to parties which, by virtue of their own size or the size of the geographical area they represent, would inevitably be excluded from that arena (EFA 2008 as cited in Gupta 2008, 70).” Hence, it could be argued that European elections drove the creation of the EFA to enhance cooperation between regionalist parties within EU member states (Lynch 2007, 14).
7
While the EFA started with six regionalist parties in 1981, it increased the numbers of membership substantially to 46 party members in 2019 (EFA 2019, n. p.). However, besides its growth in members, the EFA had historically only small numbers of Members of European Parliament (MEPs) being elected. Therefore, the EFA has usually joined forces with non-regionalist players to form a European party, due to parliamentary requirements regarding minimum size for party groups (Gupta 2008, 70-71). Thus, the EFA is commonly regarded as a weak but growing party family as most regionalist parties in Western Europe were successfully organised into the transnational European party (Lynch 2007, 6). Throughout most legislatures following the first European elections in 1979, the EFA grouped together to alliances with the Greens or extreme left parties in order to reach the required minimum size. In fact, currently the EFA is part of the Greens/European Free Alliance (Greens/EFA), that consists of regionalist, green and left-wing nationalist parties representing stateless nations and disadvantaged minorities (Greens-EFA 2019, n. p.). At present, the Greens/EFA supplies 52 MEPs, while only eleven belong to the regionalist party family (European Parliament 2019. n. p.).1 Thus, it becomes evident that the EFA representation in the European Parliament was
never sufficiently strong to form an individual regionalist parliamentary group (de Winter 2001, 6). In fact, only in the period from 1989 to 1994 the EFA could form a homogenous European party family, however, bad election results in 1994 forced the regionalist parties to create alliances again (idem, 6-7). Apart from the small numbers of MEPs elected from regionalist parties, there is another reason that undermines the representative character of the EFA party family. As a matter of fact, certain regionalist parties prefer to join other parliamentary groups making the European Free Alliance a less cohesive and inclusive European party family. For example, the Lega Nord (Movement for a Europe of Nations and Freedom, MENF), the Labour Party in Northern Ireland (Progressive Alliance of Socialists and Democrats, S&D), and also the Christian Social Union in Bavaria (European People's Party, EPP) choose different parliamentary groups and are therefore excluded from the EFA (Gupta 2008, 70-71). Overall, from the 46 party members of the EFA, only eleven MEPs are provided, which in turn results in the fact that numerous regionalist parties (without MEPs) have no representation in the European Parliament. Consequently, these parties lack access to formal
8
institutional activities; thus, they do not have equal opportunities for integration or transnational networking (idem, 71).
Another issue that is highlighted in the literature is the enlargement of the EU and the lack of institutional recognition of non-state cultural communities (Lynch & de Winter 1998, Lynch 2007, Gupta 2008, Bourne 2003). The EU enlargement process entailed full recognition of small states, such as Cyprus or the Baltic states, whereas other stateless regions, such as Catalonia or Scotland struggle for institutional recognition. As a result, the EFA demands better incorporation of sub-state units into the constitutional framework of the EU because otherwise the political survival on European level for non-state nations depends on forming an independent state (Lynch & de Winter 1998, 602). In addition to this, the expansion of the EU meant less seats per member state in the European Parliament. Since newly added regions only brought few member parties for the EFA, the EU enlargement can be seen as a disadvantage for small parties occupying marginal seats (idem, 603). All in all, in comparison to other European party families, the EFA demonstrates weaknesses regarding organisational resources, representativeness of the regionalist party family and forming a single identifiable parliamentary group (Lynch & de Winter 1998, 603). However, for all member parties, regardless of supplying MEPs or not, the European Free Alliance provides possibilities for regionalist transnational organisations and creates political structures and formal access for regionalist parties that consider networking with other regionalist actors (Gupta 2008, 71). Overall, it could be argued that the election for the European Parliament in the course of the European integration process brought the necessity for regionalist parties to group to a party family in order to improve regional mobilization in European regions, organization strategies and transnational networking. Before the creation of the European Free Alliance in 1981, no regionalist party family existed (de Winter & Gomez-Reino Cachafeiro 2002, 485). Furthermore, the importance of European elections for regionalist parties to gain visibility and legitimacy should not be underestimated. In fact, the Lega Nord in northern Italy is a product of an electoral coalition for European elections (idem, 494). Despite crucial differences regarding ideology, autonomy goals, or electoral success, the regionalist parties share the same ‘core business’, namely regional autonomy, which in turn qualifies them to group to a single party family (Lynch 2007, 2). As a result, regionalist parties became
9
subnational actors in national and supranational arenas creating transnational associations (Marks et al. 1996, 346).
Since the emergence of the European Free Alliance as European regionalist party family has been clarified, the next chapter examines the current literature on how effects of the European integration process changes policy goals for regionalist parties and how these parties generally perceive European integration. Furthermore, theories are provided that predict new opportunities for regionalist players in a changing political arena and whether the regionalist party family support or oppose the EU integration process which entails these altered opportunities. Ultimately, by providing this information, the influence of the European integration process on sub-state regions can be determined.
2.2. The European Union: A changing Political Environment
The broad literature agrees that the European integration process had crucial influences on reinforcing and restructuring regionalist interests and providing new opportunities for autonomy claims (Lynch 1996, Hepburn 2009, Jolly 2006 & 2007, de Winter & Gomez-Reino Cachafeiro 2002, Bourne 2003, Aspinwall 2002, Marks & Wilson 2000, Elias 2008, Massetti 2009, Marks et al. 1996).
In general, the development of regional institutions and the inclusion of regionalist authorities suggest that the European integration is a polity creating process in which the decision-making competency is shared across multiple levels of government and not exclusively by the central state. However, there exist alternative opinions that argue that the European integration process does not challenge the autonomy of the nation states. The state-centric theory sees bargaining among national governments of European member states as driving force behind the integration process (Marks et al. 1996, 342). Kahler & Lake (2004, 410) argue for example that even in areas where integration proceeded furthest, such as international finance and monetary affairs, no clear tendency towards supranational or regional governance has emerged. Consequently, the national states remain the dominant actor within the European system for some scholars. On the contrary, there are authors, such as John Loughlin (2008, 477) that clearly identify sub-national players developing distinctive roles within the changing framework of European governance (Marks & Wilson 2000, Jolly 2006, Hepburn 2009). Thus, these scholars support the multi-level governance theory that
10
suggests, besides decision-making competencies being shared across multiple authority levels, that political arenas are interconnected (Marks et al. 1996, 346). Consequently, subnational actors are operating in national and supranational arenas which in turn leads to the emergence of transnational associations (ibid). Nation states, on the other side, are still an integral part of the EU, however, they increasingly share political competences with supra- and sub-national actors (idem, 347). Hence, it could be argued that European integration undermines the sovereignty of traditional nation states but simultaneously creates new opportunities for other concepts of political authorities. In general, the political opportunity theory states that “social movements emerge as a result of expanding political opportunities (Goodwin & Jasper 1999, 30).” In the context of this study, these political opportunities are framed by new multi-level opportunity structures for political influence in the EU, such as the EFA, that promotes transnational networking between regional entities (Bourne 2003, 356). In general, regional issues, such as economic discrepancies or cultural distinctiveness between sub-state territories, drives the creation of political representation in form of regionalist parties (Hearl et al. 1996, de Winter & Türsan 1998, Jolly 2006). The political opportunity for regionalist parties herby also depends on the strategic behaviour of mainstream parties and their accommodation of regional agendas. In short, mainstream parties can undermine electoral success of potential new parties (e.g. regionalist parties) by incorporating their new issue (Jolly 2006, 44). Consequently, it is apparent that the European integration and new issues and opportunities this process entails, is fuelling the conflict between the establishment representing the central power of a state and new regionalist authorities which can be considered as the periphery of a nation state. The current debate agrees that the European integration process is highly substantial for the party family of regionalist parties as it redefines centre and periphery in new ways (de Winter & Gomez-Reino Cachafeiro 2002, Hepburn 2009, Jolly 2006 & 2007, Marks & Wilson 2000, Elias 2008, Massetti 2009, Lipset and Rokkan 1967). Therefore, in the next chapter, the theoretical background of this thesis, namely the centre-periphery cleavage, shall be presented to provide a better understanding of the altered political opportunities for regionalist parties within the European integration process. Moreover, the theory helps to predict whether the regionalist party family generally opposes or supports the European integration process. The attitudes of regionalist parties towards the European integration process is of special interest
11
for this study because it shows whether regionalist parties are favourable to these new opportunities the integration process entails or not.
2.3. Centre-Periphery Cleavage as Theoretical Background
Political parties generally propagate various ideologies, their positioning on the issue of European integration is therefore different. Before explaining the theoretical approaches of the centre-periphery cleavage, it is important to clarify two important elements of the European integration in order to comprehend how political parties respond to this process. Firstly, the economic integration, meaning free movements of goods, labour, services and capital and the monetary union with a barrier free single market has been the core of European integration since its beginning in the 1950s (Marks & Wilson 2000, 436). Secondly, the political integration which refers to the multi-level policy system of the EU where European institutions national and subnational governments increasingly share decision-making competencies (ibid). In fact, the literature suggests that the regionalist party family is more supportive of economic integration than of political integration (Marks et al 2002, 586-587). In order to understand the attitudes of regionalist parties towards the integration process, it needs to be clarified how political parties generally react to the issue of European integration.
According to Lipset and Rokkan (1967) political parties have historically rooted ideological orientations. As a result, it could be argued that political parties evaluate European integration within the scope of their long-standing ideological convictions (Elias 2008, 464). In general, European political party families are structured according to their positioning in the left-right political spectrum. Thus, political parties mostly restrict new European issues to this left-right dimension (Elias 2008, 564). Overall, left-right ideology is considered as a reliable indicator for determining positions on European integration for party families. Previous research suggests that centre parties favour integration, whereas right-wing and left-wing parties oppose it (Aspinwall 2002, Massetti 2009, Marks et al. 2002). Historically, mainstream parties try to undermine the salience of new upcoming issues, while minor or peripheral parties tend to utilize new issues by taking extreme positions that emphasize the importance of these issues (Marks et al. 2002, 588). A way to moderate the salience of the issue of European integration is taking positive instead of sceptical positions. As a result, most
12
major (established) parties are in favour of European integration, whereas minor parties are more sceptical in order to polarize the masses (ibid.). In addition to this, if parties are excluded from the government, they also tend to be less supportive of the European integration (ibid.). When turning to the regionalist party specifically, a slightly different picture arises. In this study, it was already mentioned that the party family of regionalist parties is ideologically diverse (de Winter & Gomez-Reino Cachafeiro 2002, Hepburn 2009, Fagerholm 2016). Hence, positioning on the left-right ideology has been a major problem for the regionalist party family. In fact, the variety of ideologies of regionalist actors reduced their coalition potential and often resulted in party divisions (Hepburn 2009, 489-490). However, when it comes to pursuing socio-economic goals within the European integration process, the location of regionalist parties on the left-right spectrum becomes crucial (Elias 2008, 564). In this context, it is predicted that regionalist parties with leftist ideologies will be more supportive of the European integration process because issues that resonate with leftist stances are increasingly transferred to the European level (ibid.). In other words, ideologically left sided regionalist parties favour supranational controls of markets (Marks & Wilson 2000, 442). Regionalist parties with right ideologies on the other hand, support economic integration, particularly supranational efforts to liberalize the market without any political regulations (Elias 2008, 564.). Overall, it has been demonstrated that the left-right cleavage is a useful predictor for political parties to determine their general response to the issue of the European Union. However, the left–right political spectrum is the traditional cleavage upon which state-wide parties contest (Hepburn 2009, 489). For traditionally non-state-state-wide parties, such as the party family of regionalist parties, a certain degree of territorial self-governance is the core of their party ideologies, thus regionalist parties tend to interpret European issues through the lens of the centre-periphery cleavage (Elias 2008, 563-564). In general, for regionalist parties the European integration process had a significant impact on their policies as the traditional sovereignty of nation states is negatively influenced and centres and peripheries are redefined (de Winter & Gomez-Reino Cachafeiro 2002, 487-488). The centre-periphery cleavage was developed by Lipset and Rokkan (1967) as one of four social cleavages that are responsible for the emergence of all European political parties and their ideologies (von Schoultz 2017, 32-33). “The centre-periphery cleavage is about political control over a peripheral territory inside a state (Alonso et al. 2013, 189).” Historically, it appeared when
13
“[…] groups within the distant, distinct, and dependent peripheries of a state opposed the state- and nation-building processes initiated by central elites (ibid.).” This division between centre and periphery results in political parties defending cultural autonomy for certain regions and conversely, parties propagating a central public authority (von Schoultz 2017, 23-33). Thus, these circumstances can cause responses in form of regional nationalism on the peripheral side and can lead to the emergence of national party systems on the central side (de Winter & Gomez-Reino Cachafeiro 2002, 487). In short, the centre-periphery theory is the cleavage on which regionalist parties contest. Marks and Wilson (2000, 438) criticise that Lipset and Rokkan do not differentiate between territorial minorities that are concentrated in certain regions, such as Catalonia or Bavaria, and those that are dispersed over multiple territories, such as Scandinavian farmers. In fact, these two minority groups need to be treated distinctively because territorially dispersed peripheral minorities oppose all forms of central authority, regardless if on nation state level or European level because decision-making power shifts away from their control (ibid.). Territorially concentrated peripheral minorities, however, can be supportive for the European integration process because it decentralizes power from the nation state to regional authorities (ibid.).
According to Hooghe and Marks (2001, 166), “the single European market reduces the economic penalty imposed by regional political autonomy because regional firms continue to have access to the European market.” The reason for this lies in the supranational character of the EU as regional entities can enter the European market without being regulated by a central government (Jolly 2006, 50). In general, regionalist players represent the ‘periphery’ side of the cleavage that tries to obtain territorial self-government concessions from the central government which in turn stands for the ‘centre’ side (Hepburn 2009, 481). In their quest for a certain degree of self-government, regionalist parties follow two alternative strategies. While some regionalist parties demand autonomy in an indivisible state, some of those parties promote secession form the central state in order to practice the right of self-determination in an independent territory (Sorens 2008, 325-326). Rokkan and Urwin (1983, 111) go even further and describe seven paths towards political self-determination, including peripheral protest, federalism and confederalism. This provided, it is crucial to understand that the supranational opportunity structure that comes with the European integration process, changed the political arena in which regionalist parties operate. Further integration
14
aims at undermining the sovereignty of nation states, the traditional enemy of peripheral regionalism, in essential sectors. These new conditions drive some regionalist parties to overthrow their ambition to create an independent nation state as it was common in the 19th
century (de Winter & Gomez-Reino Cachafeiro 2002, 488). To a certain extent, the European integration process can be interpreted in an ambiguous manner by self-governance seeking regionalist parties. On the one hand, the increasing transfers of political competences from the nation states to a supranational level can be considered as an argument against classical central states and for regional governance. On the other hand, the integration process made the possibilities to create an independent state more feasible and appealing as the economic and military costs for secession within Europe are reduced (ibid.). Consequently, the provided evidence suggests that the European integration process changes the relationship between peripheries and centres.
Before demonstrating the altered connection between peripheries and centres and how these developments influence the attitudes of regionalist parties towards the European integration, it needs to be clarified how the left-right ideology and the centre-periphery cleavage interact when it comes to party positioning of the regionalist party family. In other words, why do regionalist parties show such a great degree of variety in their ideologies? As aforementioned, the party family of the regionalist parties stretches over the entire left-right and centre-periphery spectra due to its heterogenous character (Hepburn 2009, 492). The variety of ideological positions emerges because regions provide different structural characteristics and political environments, such as electoral and party systems (Massetti 2009, 524). If a region is culturally distinct from the central state, including ethnic divisions, such as a different language, this is a strong indicator for the presence of secessionist parties in this respective region (idem, 524-525). However, cultural grievances do not explain all secessionist tendencies of regionalist parties. In fact, in northern Italy, the Lega Nord turned to secessionism in the late 90s in order to protect its economy (idem, 509). Thus, perceived cultural and economic grievances determine whether regionalist parties promote self-determination in form of increased autonomy or secessionism (centre-periphery dimension). Furthermore, the economic status of a region can also help to determine the positioning on the left-right dimension for the most successful regionalist party in a specific sub-state. According to Massetti (2009, 510-512), in wealthier or ‘bourgeois’ regions, there are no major
15
regionalist parties located left of the centre and, vice versa, in working class regions, there is no major regionalist party positioned on the right side. Thus, examining the cultural and economic situation of a region helps to understand party positioning on the left-right and centre-periphery dimension. Apart from that, the electoral or party system of a region is also a strong indicator of regionalist party positioning on the ideological axis. In majoritarian and in two-party systems usually only one regionalist party per region emerges that promotes mainstream ideologies to attract voters from the left and the right side (idem, 514-516). In proportional and multi-party systems, however, two or more regionalist parties arise that position themselves either on the radical left or radical right (ibid.). If there are more regionalist parties for one ethnic minority, intra-ethnic party competition often emerges and drives regionalist parties to pursue more radical strategies (Horowitz 1985, 342-360). Overall, Massetti’s research suggests that the analysis of the relationship between centre-periphery and left-right dimension of regionalist parties results in the following patterns: left-right radicals are more secession prone than left-right mainstream parties and left parties tend to be more secessionist than right parties. Furthermore, he argues that centre-right parties are the least secessionist (Massetti 2009, 525). Ultimately, involvement in office is one more reason mentioned that moderates the positioning of regionalist parties along the centre-periphery dimension because entering a nation-wide government coalition means weakening self-governing demands for regional parties (idem, 515). Overall, it has been demonstrated why the regionalist party family provides such a variety of ideologies. Furthermore, examining the combination of the centre-periphery with the left-right dimension allows to make general assumptions regarding how the ideological character of regionalist parties (left-right dimension) might predict their autonomist (or secessionist) objectives (Centre-periphery dimension) and vice versa. Since these fundamental questions are clarified, this study proceeds with explaining on how the European integration process impacts the strategy and mobilisation of peripheral regions and whether regionalist parties support or oppose the European integration.
Since the European integration process had a decisive impact on the centre-periphery cleavage for the regionalist party family (de Winter & Gomez-Reino Cachafeiro 2002, Hepburn 2009, Jolly 2006, Elias 2008) the political arena they operate in has been altered. As mentioned above, the supranational character of the EU makes regions economically less
16
dependent on the central state, as the monetary union and international trade laws have been introduced. In addition to this, the transnational construct of the European Union increasingly decreases the necessity for nation states to have an individual foreign and defence policy which also makes regions to rely less on their central state. Therefore, it could be argued that the multi-level governance of the European Union increases the viability of regions (Jolly 2006, 4). A general principle of the European law that also benefits the viability of regions is the principle of subsidiarity. In general, subsidiarity is a concept for the division of legislative powers at the lowest possible and efficient level (EUABC 2019, n. p.). The regionalist parties use the European Free Alliance to promote the application of the subsidiarity principle to secure certain degrees of sovereignty on regional level (de Winter & Gomez-Reino Cachafeiro 2002, 497). Initially, the principle was introduced to organize the division of competences between EU and central states, but regional institutions increasingly propagate it to expand their influence (idem, 489). In short, since the sub-state level is a lower level than the nation-state or European level, regional authorities believe that they are the most efficient decision-making level. In this new context, regionalist parties follow a method that de Winter & Gomez-Reino Cachafeiro (2002, 497) labelled ‘pincer strategy’:
Regionalist parties generally favour the transfer from policy competencies of the central state to European level which are beyond the scope of exercising them on regional level, such as foreign and defence policies. However, regional authorities also want to expand their competencies, but only at the cost of the central state and not of the European Union (idem, 497). Overall, given the new opportunities for regional autonomy and for European decision-making participation, it could be argued that the claims of regionalist parties are to some extent Europeanized (Hepburn 2009, 486). Consequently, regionalist parties are not exclusively emphasising local issues anymore.
In the next steps, the literature regarding whether the Europeanized party goals of regionalist parties influenced their activity or electoral results.
In general, the findings in the literature concur that the Europeanized party ideology of regionalist parties, combined with the new multilevel governance structure of the EU, which result in more realistic claims for self-governance, made regionalist parties more attractive for voters. Additionally, these developments boosted the mobilization of the regionalist party family (Jolly 2006, Schakel 2018, Hearl et al. 1996, Hepburn 2009). There exists an increasing
17
body of evidence that demonstrates that the number of regionalist parties contesting in elections is increasing (Schakel 2018, 690). In fact, new opportunity incentives, brought by the integration process, encourage regionalist parties to access political competition (Jolly 2006, 42). Furthermore, deeper European integration is linked to growing electoral support for regionalist parties (idem, 7). Reasons for this lie in the integration process itself, because it makes self-government goals for regions more realistic, therefore, more attractive for voters (idem, 55). In fact, the European integration process, and the decentralization of state structures has strengthened regional electoral arenas and increased the credibility of regionalist party claims for territorial and political autonomy (Hepburn 2009, 488). These developments, combined with the fact that EU supporters generally oppose national party policies and rather vote for regionalist parties, result in growing votes for these parties in national and regional elections (Jolly 2006, 86).
Overall, since it is clarified that the European integration process increased mobilization and electoral results for regionalist parties, it is now crucial to understand whether this is the consequence of supporting or opposing the EU.
In general, when it comes to attitudes towards the European Union for regionalist parties, two contrasting positions emerge: On the one hand, regionalist parties are supportive of the integration process because it changes the centre-periphery cleavage between regions and central states by reducing the sovereignty of nation states and making territorial self-governance a more realistic objective (Jolly 2006 & 2007, Hepburn 2009). In addition, due to the new multilevel governance structure of the EU and the economic integration, regional actors become more active in the decision-making process while being economically less dependent on the central state (Hooghe & Marks 2001, Jolly 2006, Marks & Wilson 2000, Hepburn 2009, Loughlin 2008). In fact, Jolly (2006) implies that regionalist parties are convinced and consistent advocators of the European Union.
On the other hand, there exist regionalist stances that oppose the European integration process by exploiting fears of economic migration and European intervention in regional politics (Jolly 2006 & 2007, van Houten 2003, Marks et al. 1996). There are recent tendencies in the literature that suggest that the EU has been unsuccessful in fulfilling the hopes for greater territorial autonomy for regions (Elias 2008, Hepburn 2009, Massetti 2009). In particular, the failure to create new European polities, such as the Europe of Regions made
18
some regionalist parties abandone their support for the EU (Elias 2008, 559-560). Furthermore, it has been explained that minor parties may be sceptical towards the integration process (Marks et al. 2002). Since many regionalist parties are minor parties that are not participating in governments, this would indicate that they generally oppose the European integration process.
Overall, it could be argued that regionalist parties support the EU as long as it positively influences the territorial reorganization of the state to gain greater autonomy for the region. In short, the redefinition of the centre-periphery cleavage remains the main goal for regionalist parties (idem, 560-561). Thus, the theoretical approach of the centre-periphery cleavage predicts that the regionalist party family supports the European integration because it undermines the monopoly of sovereign states while offering new opportunities for the territorial demands of the regions (Elias 2008, Jolly 2006 & 2007, Marks & Wilson 2000). The provided evidence combined with the theoretical prediction lead to following general assumptions regarding attitudes of the regionalist party family towards the EU:
Hypothesis 1 (H1): “Regionalist parties support the process of European integration.”
However, as aforementioned, there also exist regionalist authorities that exploit fears of European encroachment on regional competencies and economic mass migration in respective territories which in turn threatens the homogenous culture of some regions (Hepburn 2009, Elias 2008). Furthermore, established (central) parties are usually Europhile and the main enemy of regionalist parties which in turn may lead to a more sceptical view towards the EU of the regionalist party family (Marks et al. 2002). The suggested facts induce contrasting expectations that are manifested in a competing hypothesis:
Hypothesis 2 (H2): “Regionalist parties oppose the process of European integration.”
Since there exist qualitative arguments for both hypotheses, a quantitative analysis that includes numerous regionalist parties and their positioning towards the European Union is required. By using regression analyses on expert survey data, the development of EU positioning per party family over the last 20 years can be demonstrated. As a result, general assumptions for the regionalist party family and their support for the EU can be made. In this context it is not enough to simply evaluate whether regionalist parties are pro or contra EU
19
integration, but comparisons with the EU attitudes of other party families lead to more concrete inferences. After determining the general attitudes of the regionalist party family towards the EU, it shall be tested whether regionalist support for the EU varies over time and across issue area. The quantitative study is based on Seth Jolly’s (2006 & 2007) research model which concludes that regionalist parties are consistently Europhile across time and issue area. However, Jolly’s research ends in 2002, thus, this study focuses on the development of the EU positions of regionalist parties since the millennium. After the statistical analysis, the case studies of Bavaria and Catalonia shall be examined to provide qualitative arguments that help verifying the two competing hypotheses. Interviews with party representatives of the major regionalist parties in both regions have been conducted and are used as qualitative data to support the quantitative results. Ultimately, these findings are used to clarify the influence of the European integration process on European regions. In the next steps the research design and the data for the statistical analysis shall be presented.
3. Research Design
In this study, the Chapel Hill Expert Survey Data (CHES data) that estimates party positioning on European integration is used for the statistical investigation. In the surveys, country experts answer questions on general positions of parties on EU integration and several other key topics, including general left/right ideology or government participation. The CHES data depicts averages of these expert evaluations. Since the survey was conducted in subsequent waves from 1999 until 2017, developments of attitudes towards the EU can be examined (CHESDATA 2019, n. p.). In this analysis, the combined CHES data set from 1999 – 2014 (Bakker et al. 2015) is merged together with the CHES data set from 2017 (Polk et al. 2017).2
Together these data sets enable the creation of a time series from 1999-2017 tracking the position of parties on European integration issues. The number of national parties included in the survey increased from 143 in 14 European countries in 1999 to 268 in 31 countries in 2014. The 2017 survey focused on a smaller set again as only 132 national parties in 14 EU member states are part of the data set. The data sets can be merged together because the surveys target similar key questions. Thus, even though there are less observations for the
20
2017 data set, all standard questions on EU placement and ideological positioning are included in the survey. In short, the year 2017 has a slightly lower statistical significance. The party positions on the European integration process that are used as dependent variable in this research3, were initially defined on a seven-point scale ranging from strongly opposed
to European integration to strongly in favour. However, for the sake of clarity, this approach follows the strategy of Marks et al. (2002) and Jolly (2007) by rescaling the variable from zero (strongly opposed) to one (strongly in favour).
The Chapel Hill Expert Survey data is used in this study due to the amount of different national parties and European countries included in the data. Table 1 presents the large amount of different regionalist parties involved in the CHES data set. The data set contains vote percentages received by parties in the national election most prior to the respective survey year, thus, the average vote share for all contested elections has been calculated for each regionalist party. It is striking that almost all regionalist parties can be considered as minor parties due to the low share of votes these parties receive in national elections. Since regionalist parties usually only compete in their respective sub-state or region, even the largest parties have relatively small vote shares on national level.
21
Table 1: Average national vote shares of regionalist parties in Europe (1999-2017)
Country Party National vote
share
Elections contested
Belgium People's Union (VU) 5.60 1
Democratic Front of Francophones (FDF) 2.10 2
New Flemish Alliance (NVA) 11.60 4
Germany Christian Social Union (CSU)4 7.40 5
Spain Convergence and Union (CiU) 3.86 5
Democratic Convergence of Catalonia/Catalan European Democratic Party (CDC/PDeCAT)
2.00 1
Basque Nationalist Party (PNV) 1.37 6
Eusko Alkartasuna (EA) 0.40 6
Euskal Herritarrok (EH) 0.70 1
Aragonese Party (PAR) 0.20 1
Republican Left of Catalonia (ERC) 1.63 5
Valencian Union (UV) 0.40 1
Andalusian Party (PA) 0.90 1
Galician Nationalist Bloc (BNG) 0.92 5
Canarian Coalition (CC) 0.74 6
Chunta Aragonesista (CHA) 0,28 2
Amaiur 1.20 1
Ireland Sinn Féin (SF) 6.49 5
Italy Lega Nord (LN) 5.83 6
Sardinian Action Party (PsDA) 0.70 1
South Tyrolean People’s Party (SVP) 0.68 5
Movement for the Autonomies (MpA) 1.10 1
Aosta Valley (VdA) 0.01 2
UK Scottish National Party (SNP) 1.95 6
Plaid Cymru 0.58 6
Democratic Unionist Party (DUP) 0.90 1
Sinn Féin (SF) 0.70 1
Finland Swedish People's Party of Finland (SKP-RKP) 4.74 5
Estonia Constitution Party (KP) 2.20 1
Latvia Latvian Association of Regions (LRA) 6.70 1
Lithuania
Lithuanian Russian Union (LRS) 1.20 1
Electoral Action of Poles in Lithuania – Christian Families Alliance (LLRA)
4.08 4
Romania Democratic Alliance of Hungarians in Romania (UDMR)
6.08 4
Slovakia Party of the Hungarian Community (SMK-MKP) 7.13 5
Most–Híd 6.50 1
Croatia Istrian Democratic Assembly (IDS) 1.50 1
Croatian Democratic Alliance of Slavonia and Baranja (HDSSB)
3.00 1
4 There is a discrepancy between my coding and the coding of the CHES data. In the data set the CSU is coded
as a Christian democratic party, however, I recoded the CSU as regionalist parties. Justifications for these actions follow in the analysis of the case studies (In-Depth Analysis).
22
Since the survey was conducted in subsequent waves from 1999 – 2017, the development of the party support for the EU can be revealed. The positions are measured at six points: 1999, 2002, 2006, 2010, 2014 and 2017. Furthermore, the data divides the parties into different party families, thus, the general EU support for regionalist parties can be demonstrated. It should be mentioned that the Regional Manifesto Project data set also includes regionalist parties, but it only observes political parties in three countries (Gómez et al. 2019). Another reason that justifies the usage of the CHES data is that the experts party positioning on the EU displays high levels of reliability and validity (Marks et al. 2002, 589).
There are several variables in the data set that can be used as independent variables. Each of these variables might be significant when explaining the support for European integration of parties.
One independent variable of interest is ‘party families’ as it clarifies the differences in EU support between the regionalist parties and other party families. Other explanatory variables that might influence EU party positioning are government participation, electoral support, salience of the European integration issue and left/right extremism. The expert evaluations whether parties favour powers to the European Parliament and whether parties support the creation of an internal market can be used to compare party opinions on different issue areas. In the next step, the development of EU support by party family over the last 20 years shall be graphically presented to clarify differences between the regionalist parties and other party families. As a result, conclusions can be drawn that help verifying the competing hypotheses.
23
4. Statistical Analysis
The level of EU support of regionalist parties can be best expressed by comparing the EU attitudes of different party families. Therefore, a line chart is created with the rescaled dependent variable, party positions on European integration on the y-axis and the observed years on the x-axis (see Figure 1). The graph shows that the regionalist party family is consistently supportive of the European Union. The only party that shows continuously higher levels of EU support is the liberal party family. The figure depicts that the regionalist party family is equally favourable to the EU as other mainstream parties, such as the Christian democrats or the socialists. In fact, most recent developments indicate that regionalist parties have become the second most Europhile European party family.
Figure 1: EU support by party family
The regionalist party family mostly consists of minor parties (see Table 1); therefore, the constant support for EU contradicts the expectation that small parties take a more critical stance towards the EU than mainstream parties (Marks et al. 2002, 588). Besides the regionalist parties, the green parties are the only small party family that shows a similar
24
Europhile character as the mainstream parties. The lines of the extreme right and the extreme left party family clearly indicate a more Eurosceptical attitude. Overall, the results of interpreting the figure show support for Hypothesis 1 because the regionalist party family is consistently supportive of the EU since 1999. Combined with Seth Jolly’s (2006 & 2007) research, who clarified the constant Europhile character of regionalist parties from 1984 – 2002, it can be argued that the regionalist party family is supportive of the EU from the beginning of the European integration process until the present.
Figure 2 shows the average support for the EU from 1999-2017 by party family. The red dots represent the mean position of the respective party family regarding EU support over the given period. The blue line depicts the average of the mean EU positions of the party groups.
Figure 2: Distribution of average EU support positions by party family (1999-2017)
The figure demonstrates that on average, regionalist parties are highly supportive of the European integration process. In fact, the boxplots confirm earlier results from Figure 1 that the regionalist party family has a similar Europhile character as mainstream parties. The whisker, which indicates the standard deviation around the mean, illustrates that the regionalist parties are a more coherent party family when it comes to EU support than the conservative, green or radical left parties. However, the socialist, Christian democratic and
25
liberal parties appear to be tighter party families in support for European integration. The length of the bottom whisker of the regionalist party family shows that there are some regionalist parties that have Eurosceptical views. Regionalist parties with extreme ideologies, such as the Lega Nord in Italy, which is considered as right-wing party, explain the distribution of some parties below the average. Overall, Figure 2 also gives support for Hypothesis 1 due to the Europhile mean positions of the regionalist party family.
To examine the differences within the regionalist party family when it comes to EU positioning, other explanatory variables that might have an impact on EU attitudes must be included in the analysis. Therefore, this study reproduces a multivariate ordinary least squares (OLS) regression analysis of party positions on European integration based on the strategy of Jolly (2007) and Marks et al. (2002). Before performing the regression, the categorical variables used in the analysis are recoded into quantitative dichotomous variables, also called dummy variables. In practice, the regression results are easier to interpret when dummy variables are limited to the values 0 or 1, indicating the absence (0) or the presence (1) of a qualitative attribute. Consequently, the party family, country and year variables were recoded to dummy variables. For the eleven party families, ten dummy variables were created, using the radical right family as the reference. For the years, five dummy variables were coded, taking 1999 as the reference category. Similarly, N-1 dummy variables were created for the N countries available in the data set, using Austria as the reference. The remaining explanatory variables are rather straightforward, left/right extremism for instance, is coded on a ten-point scale, ranging from zero (= extreme left) over five (= centre) to ten (extreme right). The salience of the European integration variable is defined on a similar scale, ranging from zero (= European integration is of no importance) to ten (= European integration is the most important issue). Furthermore, the electoral support variable indicates the share of votes a party received in national elections most prior to the respective survey year. Government participation is a dummy variable which takes the value of one if a political party has been in government for full or part of the observed year and zero if otherwise. According to the literature, government participation has a positive effect on moderating extreme positions of parties, therefore, it is expected that this variable is positively related to EU support (Massetti 2009, 515). A high vote share for parties is also thought to be positively related to support for European integration, while low vote shares might lead to more sceptical stances (Marks et
26
al. 2002, 588). The current debate agrees on the negative effect of left/right radicalism on the support for the integration process, whereas centre parties favour it (Aspinwall 2002, Massetti 2009, Marks et al. 2002). In addition to this, a high salience of the issue of EU integration is expected to have a negative relation to EU support as historically minor parties utilize new issues by taking extreme positions (e.g. Eurosceptical) that emphasize the importance of these issues. Conversely, mainstream parties tend to downplay the importance of new issues by adopting positive stances (Marks et al. 2002, 588). Overall, the level of statistical significance of the explanatory variables indicates to what extent party support for EU is influenced by each one of these variables.
Before regressing the party position on European integration on the explanatory variables, the correlation between some individual independent variables and the dependent variable is examined separately. Table 2 shows a positive highly significant relation between government participation, electoral support and salience of European integration issue with position on European integration. Only the left/right extremism variable depicts a negative, but still significant correlation with support for EU.
Table 2: Pearson correlations between the independent variables and the EU position variable
Independent variables Correlation
Government participation Electoral support
Salience of European integration issue Left/right extremism 0.335** 0.245** 0.202** -0.064* *p < 0.05; **p< 0.01
Overall, when the explanatory variables are tested separately, they all behave as predicted. An exception is the variable salience of European integration issue, which gives a positive correlation with the EU support.
When the regression includes the full set of explanatory variables a slightly different picture emerges, as displayed in Table 3. The adjusted R-squared is 0.61 which means that the regression model explains 61 percent of the variance in party positions on European integration.
The strongest predictors in the model are the dichotomous variables of the parties as indicated by the omega squared (ω²) value (see Table 3). Omega squared is a commonly used
27
estimate of how much variance in the party position variable are accounted for by the explanatory variables. When sample sizes are small, omega squared is considered a less biased estimate than eta-squared (Statistics 2019, n. p.). The omega squared values are acquired from the ANOVA (Analysis of Variance) output and a high ω² value means that the variable has a strong impact on the dependent variable party position. The party family variable has by far the strongest influence (0.43) on party position on European integration. It has already been clarified that social cleavages that are responsible for the emergence of all European political parties, party families and their ideologies (von Schoultz 2017, 32-33). One cleavage that shapes party families included in this analysis is the left-right cleavage. Since the left/right extremism variable directly competes with the party family variable, the omega squared value is slightly higher when the left/right extremism variable is excluded (0.45).
It needs to be mentioned that left/right extremism is one dimension that explains discrepancies among party families, but there are other cleavages that are not included in the data set that also shape party families. The centre-periphery cleavage, for instance, is another dimension which is particularly significant for the regionalist party family when it comes to support for the European integration. The second strongest influence on party position is government participation, explaining 11 percent of the total explained variance. The remaining explanatory variables only have a marginal influence on the party position on EU support. For example, the highly significant variable left/right extremism only explains 1.4 percent of the total explained variance.
Since the party family variable has the strongest influence on party position on European integration, it is not surprising that all the party family dummies except the extreme left are highly significant with positive coefficients (see Table 3). The coefficients show a positive value because the party families are more supportive of the European integration process than the extreme right parties which are used as the reference category. Only the extreme left party dummy is insignificant because the radical left parties only shows slightly more favourable orientations to the EU than the reference category extreme right. In fact, the Eurosceptical character of both extreme parties has already been demonstrated in Figures 1 & 2.
28
Having clarified the statistical significance of the party families, the interest is now turned to the explanatory power of the other independent variables. As mentioned before, the omega squared value of the country variables is quite small (0.02). Thus, it comes as no surprise that apart from the United Kingdom, Ireland and Spain, all country dummies are insignificant. The influence of the year dummies on party positioning is even lower (0.002) than the country dummies. In fact, the year dummies included in the model are all statistically insignificant. Recalling that the reference category for the year dummies is 1999, it is noteworthy that the coefficients of the year dummies depict almost equal values. Combined with the fact that there is no change in the significance for the year dummies, this means that the party positions on European integration are robust over time. In short, over the years there are no significant changes on the party support for the EU. The correlation of the remaining explanatory variables (electoral support, government participation, left/right extremism, issue salience of European integration) with positions on European integration has already been tested individually (see Table 2). In the multivariate analysis, these variables continue to be highly significant even though their causal effect is decreased by the impact of other independent variables. An exception is the electoral support variable that becomes insignificant. Although more explanatory variables have been added in the OLS regression, the salience of European integration issue variable still shows a significant positive correlation with party support for EU. Therefore, the argument raised in the literature that a high issue salience of the EU integration comes with Eurosceptical orientation cannot be confirmed in this analysis (Marks et al. 2002, 588). The positive significance of the government participation variable confirms that governing parties are more supportive of the European integration process. The high significance of the left/right extremism variable and its negative correlation coefficient suggest that parties with radical left or right ideologies are less favourable to European integration. Overall, the key finding of the multivariate regression is the combined explanatory power of the party family dummies, as they explain more than 40 percent of variance in party position on European integration. Applying this finding to the regionalist party family, it could be argued that the centre-periphery cleavage on which regionalist parties contest has an independent effect on EU support.
