POLITICAL PARTIES AND THE POLITICS OF DIVERSITY IN THE EUROPEAN PUBLIC SPHERE Robert sata Abstract This article compares how diversity views affect political parties' willingness to engage in trans-European deliberation to create trans-European publics. Relying on data collected within the Eurosphere, we investigate the extent to which European diversity frames the issue of integration in the public discourse of political parties in 16 European countries - 14 members of the EU plus Norway and Turkey as non-members. We identify the homogeneity vs. heterogeneity of political party discourses and the consensus or contestation among these discourses. As a result, we find that parties with more inclusive views of diversity are more likely to be active participants in European arenas irrespective of the parties' government role or ideological background (though limited to mainstream parties). More importantly, the nature of the national public spheres and domestic political competition and cleavages determine whether national publics are willing and able to be more open to transnationalisation efforts. Robert Sata is an Academic Coordinator and Research Fellow at the Department of Political Science, Central European University; e-mail Satar@ceu.hu. LO si rN i—i o rN o~> £ £ 3 a 0 £ * 1 * (R 0 C s LO Introduction The EU subsidiary principle of devolved power demands the difficult balancing of a multiplicity of identities, while immigration flows bring new diversity to member states that challenge the creation and maintenance of national identities. Political discourse on integration policy seems torn between international commitments to accept immigrants and refugees and public opinion unwilling to grant welfare benefits or rights to them. Frames of reference employed by political actors in their discourse can bolster support for and be the most powerful break in convergence and imitation within the European public sphere, yet trans-European deliberation might be the most appropriate framework for achieving integration because this framework allows for an open-ended process of redefining the principles of inclusion and exclusion (Dryzek 1990). To see the potential for a European public sphere, we compare how diversity views and attitudes on migration might affect political parties' willingness to engage in trans-European deliberation and to create trans-European publics. Relying on qualitative data collected within Eurosphere, we investigate the extent European diversity frames the issue of integration in public discourse in 16 European countries - 14 members of the EU plus Norway and Turkey as non-members1 - to identify the homogeneity vs. heterogeneity of European discourses and the consensus or contestation among these. Based on systematic comparative analysis, we analyse the openness or closure of various kinds of public discourses to the idea of societal diversity and immigrant integration on the one hand and European public spheres on the other to answer how diversity and immigration issues affect the potential of democratic deliberation on the European level. Let us begin with a brief overview of the theoretical assumptions behind the concepts we employ in our analysis. To capture immigration issues, we examine citizenship and free-movement policies. We examine the explanatory power of diversity attitudes and integration policy in creating a European public sphere. ^ We introduce theoretical propositions for the Europeanisation of political parties to formulate alternative hypotheses of how the parties' ideological background, geographic location, or government role might affect how they articulate the public sphere. We also test whether diversity interacts with any of these theoretical propositions for party Europeanisation. Europeanisation of Public Spheres and Public Discourses Most commonly cited definitions of Europeanisation conceive it as some process of diffusion/penetration of European rules, norms, policies, etc., into domestic structures, policies, and discourses. In turn, domestic change in response to Europeanisation presupposes that national actors reconstruct their discourse and actively participate in public debate using European references. The overwhelming m s~\f nrmrim ic Thi i t~f~\noanicaf ir\t~% c+n iHiqc fr-\m icnn /~vr\ fV»/-» oi-f or^f o /~»f T7T T m il/no ot-xH A c > majority of previous Europeanisation studies focused on the effects of EU rules and regulations on domestic institutions, emphasising a top-down approach, inquiring how member states respond to European pressure (Olsen 2002). Other studies concentrated on how domestic politics shape attitudes toward Europeanisation and how national structures influence the creation of supranational structures K (Hooghe 1995). Notwithstanding the debate in the relevant literature on what Europeanisation consists of, this paper uses the term to describe a process by which the topics and f^ salience of European themes, issues, and actors become dominant in public dis- ^ courses, identities, and policies. We believe Europeanisation is best conceptualised as an interactive encounter of the domestic with the European. Domestic actors internalise EU norms because of a socialisation process, facilitated by transnational networks. As a result, identity is a critical factor for Europeanisation, and understanding how and when identity is mobilised in relation to Europe is imperative (Hooghe and Marks 2008). Diversity and especially cultural diversity are important because cultures determine group and individual behaviour, and by portraying values and norms, cultures create identities. We believe cultural identity is neither primordial nor instrumental but contingent and contextual. This means identities are historically constructed, and they are always relational and multiple. Contingency refers to a particular type of group self-identification along multiple axes of identification that are salient in greatly diverse group interactions. The particular expression of collective identity is a function of a conjunction and constellation of factors, meaning collective identity is contextual (Bush and Keyman 1997). During the past decade, migration research has seen an increased focus on trans-national communities that have strengthened using modern communications technology. The impact of transnational communities on the sending and receiving countries is undoubtedly extensive, but while processes of transnational networking weaken the role and power of national governments, governments continue to dictate the migration, settlement, and integration conditions. Thus, domestic conditions are essential elements influencing transnational politics even within the EU, where free movement of people has become the principle, but national citizenship still serves as a control device for governments (Baubock 2005). Habermas conceived the public sphere as an arena not only for the perception but also for the treatment of different problems affecting society as a whole (Habermas 1989). We argue that the public sphere is needed if only to provide information on which citizens can form their opinion and base their choices of policy. Forming transnational public sphere(s) that are inclusive and legitimate will enable citizens "to learn to mutually recognise one another as members of a common political existence beyond national borders" (Habermas 2001, 99). We believe this does not have to translate into a demand for a European identity as Habermas suggests, but mediated processes of communication are indispensible for reaching some commonality on the European level. The European public sphere (EPS) should emerge out of the interconnectedness of and mutual exchanges between various national public spheres. Europeanisation of public communication does not need to increase consensus or convergence across countries (see similarly Ladrech 2002), but if we want to make sense of the future of the EU, we need to examine how one can create new transnational public powers that are accountable to new, transnational publics (Fraser 2007, 23). Europeanisation magnifies tensions between transnational and national perspectives because the development of the EU polity - recent EU enlargement or international migration - has increased diversity within the EU. This provides new social and political conditions for very diverse social groups to participate and belong that underscore the need to analyse how different public policy regimes intersect with multiculturalism and diversity. We claim the EPS is a democratic model that can give voice and influence to very diverse social groups, as this sphere is a neutral, shared space for all. Inclusion can be done "from below" along the lines of the "politics of difference" or "from above," fusing the "politics of presence." Many claim Europe is facing a democratic deficit because European citizens have very little information about the EU and its politics and institutions. Others argue the biggest problem is the lack of a common European culture or identity - often identified as a cultural deficit of the EU (Majone 1998; Benz 2006). Most critics agree political parties should bring Europe closer to citizens as the fundamental flow of Europeanisation is this lack of open competition, of public debate, of clear articulation of debate positions (by political parties) that results in a lack of voter salience on European issues. Creating a new public arena at the European level would provide new opportunities for all types of actors for debate and political mobilisation. Although in the last few decades society has experienced a shift from government to governance, a move toward a practice of problem-solving involving multiple actors, political parties across Europe continue to play an important role in articulating responses to diversity because parties play important roles in fostering and maintaining multiple political loyalties in multi-level polities (McKay 2004). Citizens form their views about which policy options they prefer through deliberation and party contestation processes that are essential elements of all democracies. Parties can support or oppose the EU because of spillover effects from other ideological positions they might hold, and if a party is opposed to globalisation, immigration, economic competition and openness, cosmopolitanism in general, the party likely will be opposed to the EU. Maintaining multiple contextual identities is crucial for successful Europeanisation of political parties; therefore, an exclusive (national) identity ascribed to parties will make them more likely to be critical of Europeanisation and EU policies, and the more inclusive parties are regarding y diversity, the more likely they will participate in trans-European communication and collaboration networks and support the idea of a European public sphere. Europeanisation of Political Parties Earlier studies have shown that no electoral forum focuses on European issues (Marks, Wilson and Ray 2002). The key debate regarding political parties and Europe is over the relationship of Europeanisation vis-à-vis the traditional political cleavages, and whether, and to what extent, this constitutes a new basis for party competition. Some see party contestation over Europe having few "spillover" effects and absorbed within pre-existing cleavages mainly along the traditional left/right axis (Mair 2000; van der Eijk and Franklin 2004). Hooghe and Marks (2004) claim the two dimensions of more/less integration and the left/right divide are not nec rtfnnvilfr A n ^ ■rr»/-! t^TK-J-iiif ■ r* nii-1 4-W Atv» t~> 1 rs-\f rtn c > essarily independent of each other and parties instead position themselves on the new politics cleavage" on the green-alternative-libertarian (GAL) versus a traditional-authoritarian-nationalist (TAN) axis. Others claim Europeanisation causes the emergence of a new cleavage in the Rokkanian sense, restructuring political space along the lines of a conflict between losers and winners of the denationalisa-^ tion of politics, economics, and culture (Kriesi 2005). Another set of explanations for the Europeanisation of political parties claims that parties' strategic positioning relative to each other is the key determinant of LH their attitude toward the EU. Thus, Hix (1999) subscribes to the idea of a "politics ^ of opposition" by marginal parties, claiming that mainstream parties will maintain the "status quo" by incorporating European issues into their program, while marginal parties will exploit Euroscepticism since they have to fight the built-in advantage of the mainstream parties over domestic issues (Enyedi 2005). Thus, party positions on Europe cross-cut left/right divisions, and mainstream parties tend to be pro-integrationist, with Euroscepticism confined to the margins of the political spectrum, resulting in the inverted "U" pattern confirmed in empirical studies (Hooghe and Marks 2004; Bielasiak 2005). As our selected cases include Eastern and Central European (ECE) countries, we must note that some claim in ECE countries Europeanisation shows direct effects unlike in the case of Western Europe (Lewis 2005). Thus, ECE countries are considered by most scholars "downloaders" of European norms and values without any input into them (Schimmelfennig and Sedelmeier 2004; Ladrech 2009).We should also warn the reader that this paper uses data from interviews with political party elites, and they tend to be more pro-integrationist than citizens; thus, our findings cannot be easily generalised. Identifying Party Positions Since our primary aim is to examine the role of parties in articulating the public sphere, we analyse party positions, not individual respondents. Whenever possible, we concentrate our analysis on multiple dimensions of the same phenomena. These dimensions might often be counterintuitive and occasionally even contradictive, but we believe responses often contain negative and positive attitudes toward different aspects of the same subject of inquiry (Sicakkan 2009). We conduct factor analysis - a non-deterministic procedure that uncovers multiple dimensions with an often unpredictable item combination - with all variables of the questionnaire by running a variance maximisation rotation model to estimate common factors for variables. The factor analysis results are available from the author. Our analysis reveals several dimensions of attitudes toward diversity, citizenship, free movement, and asylum policy common across Europe. Table 1 shows our findings, identifying relevant factors and the composite indexes we created for further analysis as well as the calculations behind these indexes. We identify the most salient dimensions of political party attitudes toward diversity across Europe as being the degree of inclusiveness of the definition of diversity parties have, the scope of minority groups the parties identify, how much parties view diversity as a normative goal, and whether they identify advantages and disadvantages associated with diversity. We see that parties prefer adaptation to diversity either through separate institutions for minorities or within existing institutions, while the last important dimension of diversity attitudes is the degree of adaptation parties require from immigrants and minority groups. When it comes to European party attitudes toward migration, the most important dimensions that explain party views are opinions about citizenship policy, free movement regimes, and asylum regulations. More specifically, parties voice their opinion about the inclusiveness of citizenship policy, as well as their support Table 1: Attitude Indicators U S I 0 £ * 1 * (ß 0 c .S. LO Dimension Indicators Composite index Explanation Attitudes toward diversity - fluid definitions of diversity - bounded/traditional definitions of diversity - cultural/linguistic definitions Degree of inclusiveness of the definition of diversity Made up of the averaged scores -no minority groups mentioned -all groups equal -some groups mentioned Scope of minority groups Averaging a scale: no groups mentioned = lowest score, all groups equal = medium, groups mentioned = highest Diversity as a: - fact of life - condition for society - normative goal Diversity as normative goal Averaging a scale: diversity as fact of life = lowest score, diversity as condition = medium, diversity as a normative goal = highest - dynamic and globalised identity - freedom, justice, and rule of law - heterogeneous society and individual autonomy Advantages of diversity Made up of the averaged scores - broken solidarity and rigid identity - endangered national identity - unequal society with cultural tensions - closed and unjust society Disadvantages of diversity Made up of the averaged scores Minority regulation: - special status for groups - parallel systems - minority political institutions Adaptation to diversity through separate institutions Made up of the averaged scores Minority regulation: - state neutrality toward groups - multicultural institutions Adaptation to diversity within existing institutions Made up of the averaged scores - type of adaptation required Degree of adaptation required Made up of the averaged scores Attitudes toward citizenship - the case of children - specific conditions for citizenship Criteria for citizenship The degree of inclusiveness of citizenship policy Made up of the averaged scores Preferences in citizenship policy for: - co-ethnics and united family - for culturally similar immigrants - for EU rules on citizenship and immigrants who are accustomed to the host country Degree of state discretion in citizenship policy (made up of the averaged scores of indicators) - number of criteria for citizenship - support for dual citizenship Support for dual citizenship Recoding the two negatively correlated items, made up of the averaged scores - support for supranational EU citizenship Support for supranational citizenship Made up of the averaged scores Attitudes toward free movement - specific restrictions - same rules for all residents - discriminating rules Restrictions on free movement Recoding the negatively correlated items, made up of the averaged scores - support for rights for non-citizens Political rights for non-citizens Made up of the averaged scores - political rights granted to non-citizens Accept migrants: - out of compassion and acceptance of inclusive diversity - out of interest and for human rights Welcomed groups Made up of the averaged scores - free movement policies Preferential policies Made up of the averaged scores - restriction on asylum Limits on asylum Made up of the averaged scores Attitudes toward the EPS - public spaces European communication sub-spaces Made up of the averaged scores - types of exclusion Exclusion from European communication Made up of the averaged scores - support for more collaboration Support for more collaboration Made up of the averaged scores - possible European partners of collaboration - possible non-European partners of collaboration Addressing European institutions/ addressing civil society Made up of the averaged scores of the two sets of indicators for dual and supranational EU citizenship. Additionally, the restrictions on free movement, political rights for non-citizens, welcomed groups of immigrants, and preferential policies on migration are other important themes shared across Europe, ^ as well as the limits on asylum seekers parties would support. Last, political parties have clear views on whether they are willing to become participants of trans-European collaboration and communication networks, and they identify different European communication sub-spaces present, as well as the degree of exclusion from these public arenas. Other important common dimensions are the support parties have for more collaboration, and their possible partners for collaboration, with either European institutions or civic actors. Testing the Traditional Hypotheses on Party Positions We have mapped out party positions on diversity and migration, as well as the EPS using distinctions between government vs. opposition parties, ideological groups of parties, and East vs. West parties using discriminant analysis elsewhere.2 Our most important finding is that party ideology and the left/right divide matter most in determining party positions toward Europe. Counter to our expectations, our analysis of party positions has shown that parties assume positions on issues of diversity and migration irrespective of whether they are part of the government or not, while the East/West separation is significant in what parties think about adaptation of immigrants, the normative value of diversity, and group claims. In fact, the government vs. opposition distinction proves significant only concerning positioning toward the EPS and even here is a relatively weak predictor, meaning government status affects little parties' willingness to participate in European affairs - although one would expect government parties to be more entrenched in these issues since governments interact primarily with European institutions. The East/West distinction between the parties is also weak for most of our analyses, except when it comes to positions about the EPS, where the East/West distinction is a much stronger predictor than the government/opposition divide and the sharper distinction between East and West parties is seemingly due to a much larger interest in the West in interacting with European and civic actors. The clearest distinction among positions on issues of diversity and immigration as well as the EPS is present among ideological party families, where most often the political space is defined by right parties and left parties as predicted by scholars arguing Europeanisation is absorbed into traditional left-right cleavages. Yet there are a few exceptions; regional minority parties have proved to represent very distinct positions when it comes to an exclusive understanding of diversity (confirming one's expectations for these to be most inclusive). There are issues where Conservatives distinguish their preferences mostly against left parties - and migration policy is such, yet we also observed that Social Democrats and Conservatives have very close positions on most of the other issues, confirming our expectations that mainstream parties assume similar positions close to the political centre. This is especially true for the EPS - where party respondents claim some sub-spaces of European communication exist, they would not think that these are exclusive, yet have no interest in addressing European actors (the most important component) but rather civil society; although they claim they would welcome more trans-European collaboration and communication - where the two mainstream party families take up stances in the centre that are hardly distinguishable from each other, while the remaining party groups scatter around their position. Predicting Party Positions Having established the positions parties assume on the different dimensions of diversity, immigration, and the EPS, we run multiple regression analysis to examine our underlying assumption that parties more inclusive regarding diversity would be more open to trans-European deliberation. We have argued that immigration flows are important challenges for countries all over Europe; therefore, we also test if opinions about immigration as reflected by views on citizenship policy, migration, free movement, and asylum policy might affect parties' attitudes toward further European communication and collaboration. All of the regression tables are provided; meaningful adjusted R square values and significant scores are in bold (* p<0.1; ** p<0.05). Table 2: The Effect of Diversity Views on the EPS (R 0 C « 00 Ln European public spheres Extent of European communication space Degree of exclusion from the EPS Interest in addressing European institutions Interest in addressing civil society More possibilities for collaboration and communication Diversity issues Adjusted R Square 0.495** 0.147** -.018 -.067 0.314** Degree of inclusive-ness of the definition of diversity 0.808** 0.481** .072 .077 .193 Advantages of diversity .078 .015 -.017 .117 .184 Disadvantages of diversity -0.262** -0.348** .004 .035 -.020 Adaptation to diversity through separate institutions .026 .034 .180 .141 0.299* Adaptation to diversity within existing institutions .004 .130 .145 -.078 .150 Degree of adaptation required .030 .146 .043 -.083 .033 Diversity as normative goal .122 .144 -.088 .100 0.269** The scope of minority claims -.120 -.137 .094 .091 -.057 Party views on issues of diversity are the most important predictors of support of European communication spaces (Table 2), as the parties' views on diversity explain about half (49.5 percent) of the variation in the extent to which parties believe in the presence of European communication spaces. Looking for individual factors, our data shows the degree of inclusiveness of the definition of diversity has an outstanding effect that is four times larger than the effect of beliefs that object to diversity as having disadvantages, our second most important predictor. Diversity views have a much weaker predictive power (14.7 percent) on how much exclusion from the European public sphere parties claim is there. What is more important is C^ that the same individual predictors can be observed as in the previous model: the ^ inclusiveness of the definition of diversity and parties' rejection of the disadvantages of diversity. Nevertheless, in the second model the two individual predictors are on an equal footing in their importance, unlike in the previous case. Notwithstanding the important role views of diversity play in predicting support for European deliberation and how exclusive European public spaces might be, diversity positions do not affect parties' interest in addressing either European actors or civil society actors. However, a relatively large 31.4 percent of the variation of whether parties welcome further European communication and collaboration opportunities is once again explained by the parties' diversity views, confirming our expectation that parties with more inclusive views of diversity would welcome more European debate. What is interesting is that our previous individual predictors of diversity attitudes lose significance in this model, and diversity perceived as a normative goal and the belief in the adaptation of publics through separate institutions for minorities will predict the extent to which parties welcome further opportunities for trans-European communication and collaboration. Nevertheless, several aspects of diversity seem to have no or little effect on our subjects of inquiry. For example, adaptation to diversity within existing institutions does not score on any dimension, and it seems relatively unimportant the degree of adaptation required from immigrants or the scope of minority claims accepted by the parties - which contradicts expectations of multicultural scholars who focus on the importance of adaptation and minority claim-making. However, in addition to the inclusive definition of diversity as the most prominent predictor, thinking of diversity as a normative goal for society and accepting that adaptation to diversity can happen through separate institutions for minority groups are aspects that turn out to be significant predictors and confirm communitarian scholars' expectations. Turning to the question of how party opinions about immigration might affect support for the EPS and willingness to participate in European communication and collaboration (Table 3), we show that support for dual citizenship is a single predictor for 20.3 percent of the variation in how political parties judge the exten-siveness of European public spaces. Nevertheless, party views on citizenship cannot predict whether the parties judge these deliberative spaces exclusive or not, and explain only 16.2 percent of parties' interest in addressing European-level actors. As the model shows, parties that support granting supranational EU citizenship and at the same time reject inclusive citizenship policy are the ones most likely to address European actors. In contrast, our model is not significant for predicting parties' interest in addressing civil society actors. Our most interesting finding is that citizenship issues predict a relatively high 32.8 percent of the support for more opportunities for trans-European communication and collaboration, and parties that support supranational EU citizenship and dual citizenship will be the ones interested in further trans-European networking, which yet again confirms that inclusive parties are the ones open to more trans-European deliberation. In a similar manner, attitudes related to the free movement of people also affect parties' willingness to participate in more trans-European communication Table 3: The Effect of Immigration Views on the Prospects of the EPS European public spheres Extent of European communication space Degree of exclusion from the EPS Interest in addressing European institutions Interest in addressing civil society More possibilities for collaboration and communication Citizenship Adjusted R Square 0.203** .038 0.162** .113 0.328** Inclusiveness of citizenship policy -.089 -.053 -0.313** -.349 .140 Support for dual citizenship .497 .306 .123 .189 0.508** Support for supranational EU citizenship -.079 .002 0.286** .121 0.295** Migration Adjusted R Square 0.182* 0.205** 0.283** 0.365** 0.188** Degree of restrictions on free movement .021 .244 .083 -.150 0.342** Scope of political rights for non-citizens 0.364** 0.391** .126 .109 0.495** Scope of welcomed groups 0.224* 0.226* 0.469** 0.495** .169 Preferential policies on free movement .159 .167 .097 0.219* -.031 Asylum restrictions -.023 -.064 .154 .179 -.180 and collaboration. Support for political rights for non-citizens and the number of groups that parties welcome predict 18.2 percent of the variation in parties' belief in the existence of European spaces of communication. The same two individual predictors with almost identical weights of importance account for 20.5 percent variance in party attitudes on how exclusive these communication spaces are, meaning that party preferences on free movement policy explain simultaneously y about a fifth of what parties say about the presence and the exclusive nature of European public spaces. Thus, what matters most is how many immigrant group parties would be welcomed, as party choice on the scope of welcomed immigrant groups is the single predictor for 28.3 percent party interest in addressing European-level actors. More than a third (36.5 percent) of the interest in addressing civil society actors is explained by the same views, though a second predictor, with half the weight of importance, can also be identified as party support for preferential policies on free movement for specific groups of immigrants. The only aspect of the EPS, which is unaffected by what choices parties have regarding immigrant groups, is whether political parties would welcome more possibilities for trans-European collaboration and communication - a somewhat unexpected finding as we would expect that parties welcoming immigrants would be more cosmopolitan and therefore welcome more trans-European deliberation. Instead, party choices on restrictions on free movement and political rights granted for non-citizens predict 18.8 percent of the variation in answers to this question. Each measure of party positions on free movement, migration, and asylum (jq policy turns out to be a good predictor of party positions on the EPS, which means c > immigration issues are important for determining trans-European collaboration and communication. Support for dual and supranational citizenship predicts a third of the variation in parties' willingness to embrace further opportunities for trans-European cooperation and collaboration, which suggest that more inclusive parties are the ones we can expect to see in European-level debate. Thus, the parties' diversity attitudes and citizenship preferences each determine a third of the variation in party support of the EPS. At the same time, parties' interest in addressing European-level actors is affected (though to various degrees) by the parties' views on citizenship and free movement policy. Furthermore, party positions on free movement are responsible for about a third of the variation in parties' interest in engaging civil society actors, which might suggest that we not only witness a bottom-up version of Europeanisation but also see European issues descending to domestic-level actors as well. Interactive Models of Party Positions Having examined our hypotheses that diversity views and attitudes on migration affect what parties think of the EPS, we also inquire how established theoretical propositions on party Europeanisation interact with our predictive models. We control for party ideology, government role, and geographic location to examine how these interact with our independent variables in determining party position regarding the EPS, and whether they weaken or strengthen the explanatory power of our predictive models. Due to space limitations, we control ideology by testing the two main groups of parties in our sample: Social Democrats and Conservatives. We employ multiple regression analysis to test interaction effects; the results are available from the author. Examining how diversity views interact with our control variables, we find that interactive models have a higher explanatory power as reflected by the relevant R square scores with the exception of controlling for Social Democrat parties. Notwithstanding this, we observe little interaction as some of our original predictors are significant alone even in the interaction models, and some of the control variables themselves turn out to be single predictors emphasising the important role they play in determining party positions. Our most important finding is that being Eastern affects most significantly what parties have to say about the EPS. The interactive model of the diversity views of Eastern parties explains more than half of the identified extent of the EPS, about 40 percent of exclusion from the EPS, and more than four-fifth of party interest in addressing European and 46.4 percent interest in civil society actors. No other model is significant for all dimensions of the EPS, and none provides these high scores. Furthermore, Eastern membership is a significant single predictor for the models explaining exclusion from EPS, interest in EU institutions, and interest in civil society, while government membership and being Conservative single-handedly affect only the extent of the EPS identified, and being a Social Democrat never proves a single predictor of the relationship between diversity views and the EPS. In turn, government membership interacts with inclusive definitions of diversity and the scope of minority claims in determining the extent of the EPS, and government membership's interaction with definitions of diversity matters for exclusion from the EPS. Being a Social Democrat interacts only with definitions of diversity and the scope of minority claims in explaining the extent of EPS identified - our weakest interaction effect. On the other hand, being a Conservative interacts with the same dimensions for determining the extent and exclusion from the EPS, and in addition, diversity as a normative goal matters for the extent of EPS, while disadvantages of diversity for the exclusion from EPS. Further interaction shows that Conservatives who do not require adaptation from immigrants and do not think of diversity as a normative goal are likely to support further cooperation and collaboration. Eastern membership interacts with four and five original components in explaining exclusion from the EPS and interest in addressing European institutions, respectively. Further interactions show that Eastern parties that do not require adaptation identify more EPS, those opposed to minority claims are likely to address civil society, and those that do not see diversity as a normative goal will welcome more trans-national networking possibilities. Turning to the interactions of our control variables with citizenship preferences, we note that the interactive models once again tend to have a better explanatory power with the exception of the model for addressing European actors. What matters most is again Eastern membership, as combined with citizenship preference, not only makes all our models significant but also explains about half of the party interest in addressing European and civil actors - dimensions that were unpredictable in our earlier model. Our control variables being Eastern or Conservative turn out to be single components for explaining interest in European and civil actors, while government membership matters only for explaining addressing civil society, and being a Social Democrat proves to be insignificant as a single component and produces no interaction effects with citizenship preferences. Being in government and supporting dual citizenship make parties ready for more communication, while Eastern membership interacts significantly with support for inclusive citizenship policy and supranational EU citizenship when predicting exclusion from the EPS and addressing European and civil actors. Eastern parties supporting dual citizenship also address civil society, while those y that oppose EU citizenship welcome more communication and collaboration. While being Social Democrat proved insignificant, being Conservative negatively affects interest in European and civil society actors, but Conservatives in favour of inclusive citizenship policy positively affect these interests (see similarly Van Os, Wester and Jankowski 2007). Last, examining migration-related preferences together with the control variables we cannot observe a straightforward strengthening of the explanatory power of our models. Instead, we see that accounting for government membership in migration preferences significantly raises the explained variance of the extent of and the exclusion from the EPS. Similarly, Eastern party preferences for migration much better explain interest in addressing European and civil actors; while being a Social Democrat or a Conservative does not affect predictive power significantly. Government parties that oppose immigrant groups and preferential policies of free movement identify a wider EPS, while no other interaction with our control variables can be identified. At the same time, government membership alone and in c > interaction with no welcomed groups, no preferential policies, and asylum restrictions explain exclusion from the EPS. Eastern membership alone is a significant (^q component - but is also in negative interaction with restrictions on free movement and the scope of welcomed groups - in predicting the degree of exclusion from the EPS. The Eastern parties opposed to restrictions on free movement and asy- fO lum are most likely to address European and civil actors, while those favouring ^ political rights for non-citizens are the most supportive of more collaboration and communication. When we control for party ideology, Social Democrats turn out to be keen to address European actors irrespective of the missing interaction effect with any aspect of migration. At the same time, Social Democrats opposed to immigrants see exclusion from the EPS, while those opposed to immigrants and favouring asylum restriction welcome more trans-European debate. Similarly, Conservative ideology interacts little with aspects of migration in predicting views about the EPS. Yet, Conservatives who welcome immigrants and political rights for them, and are opposed to preferential policies or asylum restrictions, are likely to identify exclusion from the EPS. Those supporting restrictions in policy and political rights for non-citizens and opposing preferential policies are also the ones open to more collaboration. To summarise, although one would expect that government status, party ideology, or geographic location would interact with our models of diversity views and immigration attitudes explaining party positioning on the EPS, we find very little evidence that these established theoretical propositions on party Europeanisation matter with the exception of Eastern membership. The diversity views of Eastern parties explain to a much higher degree what parties think of the EPS than any other variable from our analysis. We also see most interactions taking place between Eastern membership and our independent variables not only for diversity views but also citizenship preferences and views on migration, which might suggest that it might make sense to revise the models accordingly. In contrast, ideological differences for the two main groups of parties, as well as government role, do not produce much interaction; therefore, we do not need to refine our original hypotheses. Findings that contradict theoretical propositions, despite the limited role of ideological differences, might be because both groups of parties are composed by mainstream parties that often position themselves in the centre of the political scene. The few interaction effects that prove significant might help better predict the positioning of given groups of parties, yet these effects do not show a significant pattern for the overall analysis. Furthermore, although Eastern location proved important, it fails to explain the most important aspect of the EPS, namely, how parties see the extent of public spheres. The same is true for controlling for Social Democrats while in a government role and being a Conservative: they interact to various degrees in explaining the extent of the EPS. The other important aspect of the EPS largely unaffected by interaction effects is whether parties would welcome more communication and collaboration possibilities. Discussion and Conclusions The systematic comparison of responses from political parties in the 16 countries help us identify the most common understanding of what parties think of diversity, what they dispute regarding immigration, or how they conceptualise the EPS. We should underline that the issues we have identified proved significant for all parties across Europe, and thus, these issues constitute significant common European themes contested over the entire continent. Our most important finding is that maintaining multiple contextual identities is crucial for political parties in dealing with diversity, migration, or European public spheres as party views on diversity and immigration affect significantly what parties say about trans-European collaboration. This means that national contexts are important determinants for European integration, and if a party is opposed to an inclusive understanding of diversity, the party will most likely oppose European deliberation and vice versa. Political parties play an important role in articulating the EPS since they aggregate domestic preferences and attitudes on diversity and choice in immigration policy, which then are important predictors of how the European arena is conceived. However, Europeanisation seems to have penetrated domestic political systems since issues such as political rights for non-citizens, adaptation requirements for immigrants, and the inclusiveness of citizenship criteria figure prominently in domestic political discourses across the continent. Diversity views and claims on immigration determine what political parties incorporate in their public discourses not only on the domestic but also on the European level. In turn, common European issues and debates affect what parties think of diversity or migration. In this sense, parties convey the domestic to the European level on the one hand, but on the other, they transmit important European issues and policies back to the domestic civil society actors given the interconnectedness and mutual exchange among various national public spheres. Traditional cleavages matter in how parties position themselves along the different dimensions of diversity, migration, or the EPS; however, these cleavages are weak in explaining how the views of diversity and migration affect what parties say about trans-European communication and collaboration. The only exception is Eastern location, which seems to substantiate the claim that Europeanisation should be understood differently for Central and Eastern Europe. Ideological differences between mainstream parties or government role play a very limited role in explaining the link between diversity and the EPS. Thus, we argue that parties with more inclusive views of diversity are more likely to be active participants in European arenas irrespective of their government role or ideological background (though limited to mainstream parties). The nature of the national public spheres and domestic political competition and traditional cleavages determine whether national collective identities and loyalties prevail or whether national publics are willing and able to be more open to transnationalisa- Q tion efforts. We need to consider the multi-dimensional conditions and processes that affect diversity in contemporary European society as a European-level discussion of common issues could enable national political actors to carry European ideas into their national public sphere, which might prove a new potential for reaching common attitudes and preferences across the different member states. Hi Notes: 2 1. For each country, at least three political parties were selected: the two most important parties (government and opposition) plus a maverick party. Semi-structured interviews were conducted with approximately 37 high-ranking party members. A > org/publications/workpackage-reports/> 2. Sata, Robert. 2010. Eurosphere Task Group Report WP 5.2: Does Europe Matter? The Europeanization of Political Parties across Europe and the Fragmentation of European Public Spheres. Benz, Arthur. 2006. Policy-Making and Accountability in EU Multilevel Governance. In A. Benz and Y. Papadopoulos (eds.), Governance and Democracy: Comparing National, European and Transnational Experiences, 99-114. London: Routledge. Bielasiak, Jack. 2005. Party Competition in Emerging Democracies: Representation and Effectiveness in Post-Communism and beyond. Democratisation 12, 3, 331-356. Bush, Kenneth D. and Fuat E. Keyman. 1997. Identity-Based Conflict: Rethinking Security in a Post-Cold War World. Global Governance 3, 3, 311-328. Dryzek, John. 1990. Discursive Democracy: Politics, Policy and Political Science. Cambridge: Cambridge University Press. Enyedi, Zsolt. 2005. Europeanization of Central Eastern European Party Systems Paper prepared for the POLIS Plenary Conference 2005, Workshop 8: Party Systems and European Integration. Fraser, Nancy. 2007. Transnationalizing the Public Sphere. Theory, Culture, and Society 24, 4, 7-30. Habermas, Jürgen. 1989. The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society. Cambridge: MIT Press. Habermas, Jürgen. 2001. The Postnational Constellation. Political Essays. Cambridge: MIT Press. Hix, Simon. 1999. Dimensions and Alignments in European Union Politics: Cognitive Constraints and Partisan Responses. European Journal of Political Research 35, 1, 69-106. Hooghe, Liesbet. 1995. Sub-national Mobilisation in the European Union. West European Politics 18, 3, 175-198. Hooghe, Liesbet and Gary Marks. 2004. European Integration and Democratic Competition. Europäische Politik 3, 1, 1-8. Hooghe, Liesbet and Gary Marks. 2008. A Postfunctionalist Theory of European Integration: From Permissive Consensus to Constraining Dissensus. British Journal of Political Science 39, 1, 1-23. Kriesi, Hanspeter. 2005. The Mobilisation of the Political Potentials Linked to European Integration by National Parties. Paper presented at the Conference on Euroscepticism, Amsterdam, July 1-2, 2005. Ladrech, Robert. 2002. Europeanization and Political Parties. Towards a Framework for Analysis. Party Politics 8, 4, 389-403. Ladrech, Robert. 2009. Europeanization and Political Parties. Living Reviews in European Governance 4, 1. Lewis, Paul G. 2005. Consolidation or Collapse?: Impacts of EU Involvement on Party Systems in Central Europe. Paper prepared for European Union Studies Association (EUSA) Biennial Conference 2005, March 31-April 2, Austin, Texas. Mair, Peter. 2000. The Limited Impact of Europe on National Political Systems. Western European Politics 23, 4, 27-51. Majone, Giandomenico. 1998. Europe's 'Democratic Deficit': The Question of Standards. European Law Journal 4, 1, 5-28. Marks, Gary, Carole Wilson and Leonard Ray. 2002. National Political Parties and European Integration. American Journal of Political Science 46, 3, 585-594. McKay, David. 2004. The EU as a Self-sustaining Federation: Specifying the Constitutional Conditions. In L. Dobson and A. Follesdal (eds.), Political Theory and the European Constitution, 23-39 London: Routledge. Olsen, Johan. 2002. The Many Faces of Europeanisation, ARENA Working Paper WP01/2. Schimmelfennig, Frank and Ulrich Sedelmeier. 2004. Governance by Conditionality: EU Rule Transfer to the Candidate Countries of Central and Eastern Europe. Journal of European Public Policy 11, 4, 661-679. Sicakkan, Hakan G. 2009. Eurosphere Working Papers No. 19: Measuring Attitudes to Immigrants with Diversity Methodology. van der Eijk, Cees and Mark N. Franklin. 2004. Potential for Contestation on European Matters at National Elections in Europe. In G. Marks and M. R. Steenbergen (eds.), European Integration and Political Conflict, 32-50. Cambridge: Cambridge University Press. Van Os, Renee, Fred Wester and Nicholas W. Jankowski. 2007. Presentation of Europe on Party Websites during the 2004 EP Election Campaign. Javnost-The Public 14, 2, 63-82. s I 0 £ * 1 * (R G C