East and West: Testing the Five Theories in Times of Crisis

Transcription

1 East and West: Testing the Five Theories in Times of Crisis Simona Guerra, University of Leicester Fabio Serricchio, Università del Molise Abstract. Fifteen years after Matthew Gabel s seminal work on support for European integration (1998a), this paper empirically tests again the five theories to explain attitudes towards the EU. In Gabel s article, results show that citizen s support could be explained by both cognitive mobilization and political values (affective dimension), in particular across original member states, while policy effects on welfare, political parties and government support (utilitarian dimension) were determinant in later member states. This paper replicates Matthew Gabel s test on five theories, using the Eurobarometer data. First, it briefly presents the five theories and the most recent findings in the literature; second, it tests the five theories in two case studies: Italy, as one of the founding member states that has undergone economic recession for four years in a row; and Poland, as one of the recent new member states that represented half of the 2004 enlargement and is the eight biggest economy in the EU. Finally, the theories will be tested in comparative perspective. This analysis expects that elites have greater opportunity to influence public opinion in Central and Eastern Europe. However, high levels of distrust for political elites and very low levels of turnout can indicate that the egocentric and sociotropic utilitarian indicators and the affective dimension can be determinant also in post-communist countries; and recent analyses on the current financial crisis show that national identity and political institutions are increasingly significant in explaining support for the EU. While it is worth noting that party-based euroscepticism does not necessarily translate into public euroscepticism, and there may not be a relationship between eurosceptic or euroenthusiast voting into eurosceptic and euroenthusiast political parties at national and European Parliament elections, this paper stresses the differences and possible convergences across Western and Eastern member states. Paper prepared for presentation to the Panel East-West Comparisons Electoral Behaviour in EP Elections and Beyond, 7 th ECPR General Conference, Bordeaux, 4-7 September First draft.

2 European citizens attitudes increasingly influence and constrain the process of EU integration; candidate countries at accession referendums, with some countries amending their constitution prior to the referendum, fearing low turnout and a possible rejection, as Hungary, Poland and Croatia, or through referendums, with Ireland voting twice for the Treaty of Nice and the Treaty of Lisbon, and record low turnout at European Parliament (EP) elections in 2009 (43%) since the members of the EP (MEPs) have become directly elected in 1979 (61.99%) (EP data). Hence, public opinion is crucial to the relationship between the EU and member states; on one hand domestic politics is affected by this relationship, developing since the first stages of the candidacy of countries; on the other political parties can influence the domestic and European political debates, in a dialectic relationship where domestic politics affects and is affected by the EU and the nature of the relationship of the EU with each member state (Guerra 2013). Research on attitudes towards the EU have developed with a focus on attentiveness to politics and political psychology factors, as with Ronald Inglehart s cognitive mobilization (1970); the use of shortcuts or domestic proxies, the evaluation of the performance of the government, national institutions and domestic economic and political factors (Anderson 1998; Jasiewicz 2004); trust towards domestic institutions and the relationship between their performance and trust towards EU institutions (Sánchez-Cuenca 2000); national and EU factors (Kritzinger 2003); rational utilitarianism (Gabel & Whitten 1997) and an affective dimension (Gabel 1998a); identity (McLaren 2006), explored in its exclusive and inclusive qualities (Bellucci et al. 2012; Hooghe & Marks 2004); mass-parties linkage (Marks et al. 2002; Ray 2003a, 2003b; Steenbergen et al. 2007) and, recently, on the role of the media (de Vreese & Boomgaarden 2006; de Vreese et al. 2006; Peter 2007).

3 Studies on Central and Eastern European countries emerged in the run-up to the fifth enlargement of the EU (Klingemann et al. 2006; Rohrschneider & Whitefield 2006), mainly applying the same frameworks used on the Western member states (Christin 2005; Cichowski 2000; Tucker et al. 2002), comparing attitudes in the post-communist region with Western member states (Marks et al. 2006); examining the accession referendum (Caplanova et al. 2004; Tverdova & Anderson 2004) and determinants and behaviour in the first EP elections (van der Eijk & van der Brug 2007; Koepke & Ringe 2006; Marsh et al. 2007; Schmitt 2005). In 1998, Matthew Gabel addressed some methodological questions on how we can compare across different studies, with analyses on different data sets, using different dependent variables, and the problem with bivariate analyses that may conceal intervening or spurious relationships (1998a: 334). The aim of this paper is to test again Matthew Gabel s five theories in times of crisis. Using the Intune data set (Cotta et al. 2009), already used for broad comparative analyses, and using independent and dependent variables, which are common to other data sets (as the Piredeu project and European Election Studies), and regression analysis, this study compares attitudes in two member states in comparative perspective. The hypotheses and theoretical frameworks are first tested at the regional level (east and west member states); second the analysis tests the comparative generalization to two case studies. This analysis examines Italy, as one of the founding member states that has undergone years of recession, an unstable political situation and views decreasing EU support; and the case of Poland, as the largest of the candidate countries that joined in Poland has enjoyed a positive economic situation with a turning point in 2013 due to the lack of important reforms, and suffers a high rate of unemployment (up to 10.6% in the first four months of 2013, OECD data); and a lower and lower GDP per capita, down to 1.9% in 2012 (it was 6.2% in 2006, 6.8% in 2007 and 4.5% in 2011, OECD data), with a small drop to support for EU integration (CBOS 2013).

4 The case study and the comparative framework enable to underline common patterns and differences between Western and Central and Eastern European countries. Compared to Gabel s theories, this study examines together variables related to political cues and class partisanship ; also it does not test support for government as a single theory, but investigates perceptions on the performance of national institutions and it tests whether recent findings on the increasing salience of national identity and political institutions (Serricchio et al. 2013) are still valid and to what extent. First, this paper presents an overview of the literature, in particular on the expectations of the impact of the current crisis in the two regions; second, it presents the data set and the relationship between the variables included in the different frameworks and the Western and Central and Eastern member states. Third, the regression analysis underlines common characteristics and differences in the two regions; the utilitarian model is most explanatory in the Eastern case, while identity and political cues explain attitudes towards the EU across the Western member states. Poland results quite in line with the general Central and Eastern patterns and by contrast Italy is far from them in the case of the Western member states. EU identity (achieved, by choice) is the common determinant factor and can affect attitudes towards the EU positively and negatively; on one side the EU four free movements, and working and studying abroad, can reinforce EU identity; on the other identity can also refer to the performance of the EU institutions and the impact of the financial crisis may further affect attitudes towards the EU, and turnout and results in the next European Parliament elections. Framework of analysis and hypotheses As in Matthew Gabel s seminal article (1998a), this analysis selects and examines the five main theoretical frameworks he adopted, collapsing political cues and class partisanship

5 together, with a focus on national institutions, with socio-demographic factors as control variables. Cognitive mobilization In the late 1960s and 1970s, the impact of education on citizens perceptions of politics and the performance of political institutions increasingly became a focus of studies on political behaviour. Literate citizens could follow political developments through the media and improve their political skills. Ronald Inglehart found that increasing cognitive mobilization seem(ed) to favour European integration (1970: 785). Higher levels of education could enable citizens to understand and positively perceive the process of European-level politics and policies, and develop their sophisticated understanding of international organizations. Findings showed a close relationship between high levels of education and positive attitudes towards the EU that still held true after citizens became more responsible towards the process of EU integration in the 1990s (McLaren 2005). In the comparative study carried out by Matthew Gabel (1998a) cognitive mobilization was only explanatory only in the six founding member states. In this study it will be tested through variables as political sophistication, influence on politics, education, non electoral participation, and media consumption and it is expected to have a higher impact, as in Gabel s study, in the founding, and west, member states. H1: Cognitive mobilization is likely to be more explanatory in the west member states than the east region. Political cues and Class Partisanship Based on Ronald Inglehart s silent revolution, proposed by Matthew Gabel s political values section, where the analysis focused on how political attitudes are shaped by the

6 socioeconomic conditions surrounding their formative, or preadult, years (1998a: 336), this study identifies and tests possible factors affecting both west and east member states. The distinction between materialist values, fulfilling physical and security priorities, and postmaterialist values, fulfilling intellectual priorities, seem to be less applicable to the case of the post-communist region. Communism had implications for the cultural and social development of Central and Eastern European countries and brings enduring legacies on the process of democratization. Leslie Holmes addressed the possible subsequent rejection of grand theories and an ideological vacuum, where any political and cultural proposal, pointing to a set of norms or rules, could be rejected; in particular if proposed by the elites or any form of hierarchy that could be linked to the Communist past and the us vs. them mentality (1997: 18). This can have an impact also on levels of trust towards political parties making their cues less valid for post-communist countries where trust towards parties and domestic institutions is lower compared to western countries (Klingemann et al. 2006). Also, existing research was found inconclusive by Matthew Gabel with possible impact of further variables, as education, that could influence the impact of political values on support for EU integration. Hence, this analysis includes variables that investigate political cues, but also class partisanship. This study comprises those variables that were found determinant in the literature and tests factors as national institutions, EU institutions, left and right, and political party identification. H2: Political cues are likely to be more explanatory in the west member states than the east region. Utilitarian Appraisals of Integration Policy This study will also test the utilitarian model; studied, and compared, by Matthew Gabel in the run-up of the ratification of the Treaty of Maastricht (1998b) with the affective model,

7 and previously developed by Matthew Gabel and Harvey Palmer (1995). Their findings showed that levels of education and occupational skills had a positive relationship with support for EU integration. More prosperous citizens would profit from market liberalization, while citizens with a low income were expected to oppose it as more dependent upon social welfare spending (Gabel 1998a: 337). Further, living close to the border could provide economic advantages, because of trade exchanges and have a subsequent positive impact on attitudes towards EU integration. That is further examined by Matthew Gabel (1998b) and Simona Guerra (2013); in her analysis, Polish citizens living at the border with Germany viewed the advantages of EU integration in economic terms, while citizens in the eastern regions referred to culture and the return to Europe. Overall, results based on perceptions and comparisons between the national and personal levels show that citizens may refer to the subjective evaluation of their personal economic situation (Gabel & Whitten 1997; Guerra 2013). Also, comparing attitudes before and after joining the EU in Central and Eastern Europe, it was mainly the economic variable that changed, where benefits at the personal level could be more visible after accession (Guerra 2103). This analysis measures the impact of the economic perceptions, at the national and individual levels. H3: The utilitarian model is likely to be more explanatory in the east member states than the west region. National Identity and Trust towards other Europeans (sense of community) Liesbet Hooghe and Gary Marks (2004) underline the exclusiveness of national identity, where people holding multiple territorial identities are more likely to support the EU, while in-group identity can help to explain the strengthening of self-esteem and opposing attitudes towards the EU (Turner 1985). The conceptualization of national identity and European identity as a political and social identification has been the subject of analysis in Paolo

8 Bellucci, David Sanders and Fabio Serricchio s work on European identity (2012). The concept has developed from analysis of general attitudes towards EU integration and has been synonymous of support until the Treaty of Maastricht (1991) and the contestation of the concept in the public debate. In their analysis, they distinguish the concept following a social theory approach, and underline two dimensions, the first relying on the feeling of collective identity and belonging and the second on its emotional attachment and its salience (Bellucci et al. 2012: 62-63). The identification towards a group and the protectiveness of in-groups feeling have been identified (Tajfel 1970) and explained with difficulty by the use of protecting identity to reinforce self-esteem or to simplify and smoothly engage with the world (Turner 1985). As stressed, Liesbet Hooghe and Gary Marks (2004) addressed the exclusiveness quality of national identity, as examined in the Eurobarometer questions. This analysis examines EU identity and also adds feelings about the EU, on what the EU ascribes (inherited identity) and achieves (acquired by choice) and compares them to the same variables at the national level, and to trust towards the EU. As already underlined, Ignacio Sánchez-Cuenca (2000) found that the perception that national institutions were performing poorly correlated with higher levels of trust towards the EU, as joining the EU could positively affect the work of domestic institution and influence the governance at the national level - as with the case of candidate countries with political conditionality, when this is not perceived a threat towards national sovereignty. The expectation is that these variables will be more significant across the Central and Eastern European region, although EU identity is likely to correlate with older member states. H4: The national identity and trust models are likely to be more explanatory in the east member states than the Western region.

9 Socio-demographic factors Analyses on support for EU integration identify some specific socio-demographic characteristics; generally a young highly educated man or a high income businessman living in an urban centre would be the most supportive of EU integration. Businessmen benefit from an integrated EU market, just as young citizens can benefit from travelling and studying abroad. Also, as seen in Gabel s section on Utilitarian Appraisals of Integrative Policy, low income citizens are generally hurt by market liberalization as fixed salary people do not benefit from a competitive market, where instead skilled employee and higher education can provide usable transferable skills all across the EU. Further, religion is likely to impact on level of support, although in varying ways. Committed Catholic citizens in the founding member states are more supportive of the EU integration process (Nelsen et al. 2011), while religion was found more significant at the macro level compared to the micro level (Boomgarden & Freire 2009). This analysis would not expect a significant difference across the two regions, although the religion factor could view differences, as found in previous studies (Nelsen et al. 2011), and are here analysed as control variables. H5: Socio-demographic factors are likely to have a similar impact in the east and west member states. Aggregate (contextual) variables In our analysis we add a minimum set of aggregate variables. These are: Corruption Perception Index, a measure of Level of corruption, GDP growth, as percentage differences with previous year, Unemployment rate and Annual inflation rate. Although the role of context is relevant (Marks & Hooghe 2005), these variables are used as control, without particular expectation on impact and sign.

10 Data set and analysis This analysis uses data generated by the IntUne project coordinated by the University of Siena and funded by the EU Sixth Framework Programme (Cotta et al. 2009). Data covers 16 EU member states and Serbia i. Although in Gabel s study, the analysis was undertaken on the Eurobarometer data, seeking to be independent from time and nation (Gabel 1998a: 340), the question we address is the comparability across the Western and Central and Eastern European regions and identifying differences and commonalities, while controlling for internal relationships across variables (and multicollinearity). The dependent variable is given by a scale (0-10) and the questions on whether the EU is a good thing and the EU should strengthen. First the analysis provides an overview of the correlates between support and socio-demographic characteristics. Correlates are then measured with all the factors comprised in the theories ii and across west and east member states and the case studies of interest, included in the IntUne project, while finally the regression analysis examines determinants of support across the two regions, answering the main questions, on how attitudes vary across west and east member states, and to what extent the theories tested by Matthew Gabel, on data from 1978 and 1992, are still explanatory at this stage of the EU integration process, after the ratification of the Lisbon Treaty and in a challenging financial situation. Correlates are presented across all values, first examining gender, age groups, education, occupation, religion and Church attendance, with the standard deviation for each value and the eta-square. (Table 1 about here)

11 As expected men have a higher correlate compared to women, and young people are generally more positively linked to levels of support for the EU. The highest correlate is with the value for higher education and being a student, confirming all the expectations drawn from the literature and previous empirical research. When the same analysis is carried out with the values of those factors included in the theories, as close as possible to Gabel s study, the relationships are confirmed with some important observations. (Table 2 about here) In comparative perspective the only factor that does not correlate with EU support is EU identity is ascribed (inherited). As noted by Paolo Bellucci et al. (2012: 85), European attachment is fostered by national traditions and those citizens who rely more on their ascribed factors with reference to their national identity would be the ones who would perceive a weaker attachment to Europe, contrary to the possible expectation that national identity may correlate with the perceived threat towards national sovereignty. All the factors included in the theories correlate with EU support in the 16 west and east countries, generally with the expected directions. The only variable that does not have any correlate with support is the question on whether the EU is ascribed, while the differences between east and west regions are explained by four factors, national identity, national identity is ascribed, quality of governance and (annual) inflation rate. In the first three cases the variable correlates with west member states, but with different relationships, while the inflation rate correlates only in the east region. National identity correlates in the west member states, while it does not in the post-communist region; when we read this result with national identity is ascribed (inherited), which is negatively correlated to west, it is possible to suggest the first difference across the two regions. The process of EU integration could be

12 framed in the Eurosceptic public discourse as a threat to sovereignty, but does not have a significant relationship with support. When national identity is viewed as achieved (by choice) it becomes positively related in both cases, with national identity reinforcing support for EU integration. When we look at the variable national identity is ascribed (as inherited), the analysis shows that it holds a negative relationship with the west member states, making the identity dimension of how citizens perceive themselves at the national level a factor that is likely to decrease levels of public support and worth to further investigate in the regression analysis, as it was expected that identity could be a more contested variable across the east region, following the debates on EU accession (Taggart & Szczerbiak 2004). In the case of the expectation that Central and Eastern member states are likely to be more supportive when looking at levels of trust towards national institutions (and here looking at the quality of governance ), following Sánchez-Cuenca s findings across west member states (2000), the hypothesis holds true, when we examining correlates; while in the case of the Western region, the relationship is negative. Similarly, the two regions show one further difference in the case of the unemployment rate, in a negative relationship with east member states. Focus groups with Polish young citizens show that they generally see the EU as a provider of chances (Guerra 2013). This can explain the impact of this factor in Poland and the lack of an impact in the Western member states. Further, the legacy of the Communist past is likely to have an impact also on the left-right scale, viewing a positive correlate with west member states and a negative relationship with the post-communist region. As Holmes (1997) stressed, the them and us mentality remained, with mistrust of political institutions and cynicism towards politics and political participation, and low turnout. Mistrust was further fuelled by cases of bribery and scandals during the process of democratization, iii in particular towards parties on the left wing of the

13 spectrum. The legacy also worked out as a magnetic field for the axis of ideological competition, on the left-tan and right-gal iv dimensions (Vachudova & Hooghe 2009), but this analysis suggests that the process of integration and the negotiation process also blurred differences between the left and the right, as both had to support the implementation of the reforms, required by conditionality. In an atomized civil society (Holmes 1997), this could further maintain the distance between political parties and citizens, and could slow down the emergence of cleavages. Therefore, this analysis expected a difference between east and west at the political level and the different results on the left-right spectrum likely explain the political culture that has developed in the post-communist region. When we look at the two case studies these deviate from the usual pattern in the region. Poland shows a lower impact of those factors comprised in the cognitive mobilization framework, as expected, while Italy does not show any relationship with positive retrospective economic evaluations, national institutional confidence and party identification. In order to look more in-depth at the relationship of these variables in the two case studies and the regions, the analysis runs the regression analyses, testing these same theories. Discussion Eurobarometer data on general patterns of support for EU integration in the two regions show that east and west member states are decreasingly supporting the EU, with positive attitudes dipping in 2004, when the EU enlarged, and in 2009, when the financial crisis hit all countries across the EU. (Figure 1 about here)

14 Poland is the deviant case, when looking at trends across EU member states. While attitudes towards the EU become more negative, Poland views increasing levels of public support, as also national survey reports, stabilizing between 80% and 89% ( , CBOS data), with increasing levels of opposition only since April 2010 (it was 9%, increased to 15% in December 2012 and 19% in April 2013, CBOS). Poles generally had very low expectations from EU membership; the long waiting for EU membership brought dissatisfaction, the costs of accession became visible and provoked mixed feelings; further, although Polish citizens thought membership could benefit the country, they did not assume any personal benefit (Guerra 2013). When fears were neutralized by benefits which materialized both at the country and personal level, after joining the EU, public support stabilized, as seen, at very high levels (CBOS data). (Table 3 about here) Cognitive mobilization H1: Cognitive mobilization is likely to be more explanatory in the west member states than the east region. The first hypothesis expected a higher impact of the cognitive mobilization model on west member states due to the legacy of the Communist past and findings from previous research. Also Gabel (1998a) found that, before the Treaty of Maastricht, cognitive mobilization was only holding true only in the six founding member states. As expected the theory is not very explanatory, but contrary to the hypothesis it explains more in the case of the east member states compared to west member states. The only factor that has an impact across the west member states is education, but in a negative relationship; on the other hand, it does not have any relationship with the Central and Eastern European region and the case studies. In the

15 post-communist region, influence on politics and media consumption are determinant factors; as in the case studies, media consumption remains an important variable for Poland. Previous research (de Vreese et al. 2006) had found that before the 2004 European Parliament elections citizens in Central and Eastern Europe had better access to political debates from MEPs more compared to citizens in Western member states. Further, the large amount of information that citizens in Central and Eastern Europe have accessed prior to joining the EU and once they joined the EU, may have had an impact on the role of media consumption. Also, research, using focus groups, shows that, although citizens do not perceive the impact of information, debates on the Constitutional Treaty referendum, later Lisbon Treaty, were discussed using the daily newspapers language (Guerra 2013). Overall, as in the case of Gabel, fifteen years ago, the cognitive mobilization framework is not very explanatory and the H1 is rejected, as it holds true more across the post-communist region, while it is not explanatory in the Italian case. Political Cues and Class Partisanship H2: Political cues are likely to be more explanatory in the west member states than the east region. As in the previous hypothesis political cues and class partisanship were expected to be more explanatory in the case of western member states. Communism influenced the cultural and political developments and low levels of trust towards political institutions and politicians may have enduring legacies in the region. The hypothesis holds true and shows that it is explanatory in the west case, where all the variables, but confidence in national institutions, impact on support for the EU, with a negative relationship in the case of the right-wing scale, and a positive impact with EU institutions and political party identification. Again, Italy does not show any significant relationship with these variables and is a deviant case across patterns

16 in its region, while Poland is completely in line with the post-communist region, where unexpectedly trust towards EU institutions is linked to levels of public support. The hypothesis is accepted and is very explanatory across the Western member states. Utilitarian Appraisals of Integration Policy H3: The utilitarian model is likely to be more explanatory in the east member states than the west region. In the case of the utilitarian model, we contend that Central and Eastern European member states may perceive more advantages from EU membership, although factors as education, living close to the national border and unemployment may affect the results also across the Western member states. As expected, all the three variables included in the data set, positive retrospective economic evaluation, EU benefits nation, EU benefits for people like me are positive and significant across the post-communist region. EU membership represented the return to Europe, but it was also benefits, in particular when these materialized after joining the EU. Students, businessmen, but also farmers, were the three categories which showed the highest levels of support after joining the EU in Poland (Guerra 2013), and if Poland was successful in receiving more funds in the agricultural sector, we can also expect clear benefits for students and businessmen across the entire post-communist region. It is worth to note that, as before accession, benefits at the personal level are not likely to be significant in the case of Poland; while it is the complete opposite in the case of Italy, where personal benefits are the only significant factor in the utilitarian framework. Overall, this hypothesis is accepted, and the case studies, as in the case of patterns across time, show completely opposite results. Italy is becoming more Eurosceptic, while Poland is still showing very high levels of Euroenthusiasm; and Italy views a very personal utilitarian

17 attitude towards support for the EU, that is based on positive retrospective evaluations, while benefits for the country are a salient factor in the case of Poland. National Identity and Trust towards other Europeans (sense of community) H4: National identity and trust models are likely to be more explanatory in the east member states than the west region. This analysis has adopted, and tested, the same conceptualization of national and European identity, as a political and social identification, as in Paolo Bellucci, David Sanders and Fabio Serricchio s analysis (2012). The hypothesis is rejected, as comprising seven variables, six of them are significant in the case of Western member states and only five are determinant across the Central and Eastern European region. In particular EU identity (ascribed) and national identity have a negative relationship with support for the EU; while only trust towards other Europeans is a positive determinant in the case of the post-communist region, where identity seems to represent a more contested variable compared to Western member states. Also, in the two case studies, here analysed, Italy shows three significant variables, while Poland only shows two. Trust towards other Europeans is always an important positive factor, in the regions and in the case studies, but they differ as European identity is important, in Italy, as a founding member state, but is lacking in the case of Poland. EU identity, as achieved (by choice) and ascribed, is important in the case of Poland, showing and supporting the mobilizing strength of the EU, termed before joining the EU as unconditional support (Guerra 2013). Without knowing how life could be in the EU, membership represented the final goal of the democratization process and also a clear benefit for the young generations and for the country, as also the utilitarian framework shows. Overall, this hypothesis is not supported, and the framework is not fully explanatory.

18 Socio-demographic factors H5: Socio-demographic factors are likely to have a similar impact in the east and west member states. Results on the socio-demographic factors are quite similar, do not show many significant factors, and the hypothesis can be accepted. The only main difference regards age, as this is determinant in Western case, but not in the post-communist region. That is surprising as young people are generally very positive towards EU integration, although analysis on Croatia, that joined the EU in July 2013, also stresses that young people are generally more Eurosceptic and/or apathetic towards the EU (Covic 2012). Gender is significant in the case of Italy, while occupation results determinant in the Western case, and partly also in the case studies. Religion does not seem to be determinant, possibly following latest findings and expectations of a less salient impact of religiosity in Western member states (Nelsen et al. 2011). Conclusion This analysis shows that the utilitarian model is most explanatory in the Eastern case, with Poland quite in line with patterns in the region. In the west region, identity and political cues are the ones that explain attitudes towards the EU, while Italy shows far from general patterns across Western member states. (Table 4 about here) The main difference across the two regions can be explained by what the EU is still representing; the EU represented the civilizational choice, a historical return to the geographical, political space in which Central and Eastern European states naturally

19 belonged; the final goal of the democratization process, the demonstration that countries that were considered far from EU standards could also achieve membership. After joining the EU, benefits materialized, citizens perceived that the country was moving in the right direction and benefits were perceived both at the personal and country level (Guerra 2013). This can also become a source of Euroscepticism, as general trends in Central and Eastern Europe show (see figure 1). When the financial crisis provoked growing levels of unemployment and lowering levels of GDP, even in countries that were considered the most successful at the economic level, as Poland (see OECD data), those unexpected benefits dissolved and general trends show a slight decrease in support for the EU. Previous research (Serricchio et al. 2013) found that attitudes towards the EU are increasingly becoming embedded at the national level. This emerges in this analysis in the case of Western member states, while in the regions, east and west, EU identity achieved (by choice) is the common positive determinant factor, with increasing importance of the civic dimension of EU identity. These findings can be further corroborated by a recent study on the Erasmus experience which showed students with this experience were more interested in other European countries (90%) and felt more European, albeit that students attracted to studying abroad could already possess a more pro- EU attitude beforehand. (Mitchell 2012) If this is the common factor across the two regions and the two case studies, it can also suggest that despite the financial crisis, the enlargement of the EU and the free movements can reinforce and strengthen EU identity and support for the EU. Identity, however, may refer to the performance and efficacy of EU institutions, where the slowdown of the integration process and the impact of the financial crisis may further affect attitudes towards the EU, turnout and results in the next European Parliament elections.

20 Tables and figures Table 1. Socio-demographic correlates of European Support (0-10 scale) E-Identity St. Dev. (mean) Overall Gender Female Male Age cohort Education elementary secondary intermediate university Occupation self-employed non-manual employee civil servant manual employee student retired unemployed other inactive Religion None Catholic Orthodox Protestant Muslim Other Religion attendance Never Seldom Often Source: Intune 2009 Note: Eta-sq is association with EU support Eta-sq

21 Table 2. Correlates of European Support (0-10 scale) Economic ALL (16 countries) WEST EAST ITALY POLAND Positive Retrospect econ. eval..091(**).054(**).169(**) (**) EU benefits nation.416(**).440(**).371(**).302(**).384(**) EU benefits for people like me.337(**).373(**).262(**).291(**).298(**) Identitarian European Identity.347(**).391(**).244(**).302(**).296(**) EU Identity is ascribed Eu identity is achieved.252(**).302(**).155(**).185(**).286(**) National identity.023(**).034(**) National identity is ascribed -.073(**) -.089(**) (**) -.079(*) National identity is achieved.037(**).031(**).050(**) Trust other Europeans.270(**).275(**).256(**).274(**).258(**) Political cues National institutonal confidence.201(**).193(**).216(**) (**) EU institutional trust.426(**).442(**).396(**).243(**).435(**) Left-Right Scale -.019(*) -.045(**).040(**) -.142(**) Party identification.087(**).081(**).097(**) (*) Cognitive mobilization Political sophistication.103(**).102(**).105(**).161(**).090(**) Influence on politics.122(**).132(**).085(**).110(**) Education.099(**).096(**).102(**).152(**).160(**) Non electoral participation.040(**).028(**).051(**).087(**) Media consumption.092(**).096(**).079(**) (*) Aggregate (contextual) Quality of Governance (CPI index) -.084(**) -.142(**) Gdp growth.067(**).055(**).092(**) - - Inflation rate (annual) -.030(**) (*) - - Unemployment rate.091(**).148(**) -.053(**) - - ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). N

22 Figure1. EU support ( ) Italy Poland Mean value EU-27 Mean value Western Country Mean value Eestern Country Source: Eurobarometer (various) Note: % citizens answering EU membership is a good thing (missing excluded)

23 Table 3. OLS regressions: West and East Italy and Poland OVERALL WEST EAST ITALY POLAND (16 countries) Economic Positive Retrospect economic ** * evaluation (0.0393) (0.0366) (0.0430) (0.150) (0.103) EU benefits nation 1.012*** 0.986*** 1.374*** *** (0.199) (0.209) (0.247) (0.375) (0.392) EU benefits for people like me 0.510*** 0.514*** 0.229** 0.554** (0.0773) (0.0903) (0.0869) (0.265) (0.254) R-squared Identitarian European Identity 0.173*** 0.187*** * (0.0241) (0.0232) (0.0353) (0.0771) (0.0816) EU Identity is ascribed ** ** ** * (0.0207) (0.0220) (0.0193) (0.0677) (0.0615) EU Identity is achieved 0.215*** 0.209*** 0.225*** 0.342*** 0.309*** (0.0196) (0.0228) (0.0545) (0.0946) (0.0837) National identity *** *** ** (0.0173) (0.0197) (0.0173) (0.0614) (0.0717) National identity is ascribed * * (0.0398) (0.0396) (0.0287) (0.0867) (0.0942) National identity is achieved ** (0.0418) (0.0448) (0.0151) (0.0609) (0.0935) Trust other Europeans 0.122*** 0.127*** *** 0.252*** (0.0166) (0.0167) (0.0126) (0.0689) (0.0739) R-squared Political cues National institut. conf (0.0371) (0.0409) (0.0459) (0.0664) (0.0681) EU institutional trust 0.256*** 0.243*** 0.295*** *** (0.0425) (0.0459) (0.0315) (0.0746) (0.0882) Left-Right Scale *** *** (0.0398) (0.0449) (0.169) (0.128) (0.168) Left-Right Scale squared *** *** ( ) ( ) (0.0152) (0.0122) (0.0151) Party identification * * (0.0449) (0.0409) (0.175) (0.238) (0.239) R-squared Cognitive mobilization Political sophistication ** (0.0183) (0.0206) (0.0466) (0.0600) (0.0591) Influence on politics ** (0.0154) (0.0165) (0.0200) (0.0397) (0.0437) Education *** ** (0.0155) (0.0183) (0.0366) (0.144) (0.138) Non electoral participation (0.0169) (0.0160) (0.0161) (0.0470) (0.0575) Media consumption * *** * (0.0158) (0.0172) (0.0182) (0.0527) (0.0530)

24 R-squared Socio-dem. Gender (0=female; 1=male) ** (0.107) (0.128) (0.0850) (0.241) (0.230) Age * ( ) ( ) (0.0103) (0.0353) (0.0366) Age squared * * 8.17e e (8.72e-05) (9.87e-05) ( ) ( ) ( ) Occupation (not/manual) ** ** *** (0.1000) (0.103) (0.0804) (0.482) (0.351) Occupation (not/unemployed) * (0.121) (0.0969) (0.360) (0.500) (0.502) Religiosity * (0.0266) (0.0302) (0.0177) (0.0545) (0.0670) R-squared Aggregate (contextual) Quality of Governance (CPI *** *** 0.402*** index) (0.0284) (0.0371) (0.0832) Gdp growth * 0.173*** (0.0531) (0.0501) (0.0150) Inflation rate (annual) ** (0.0294) (0.0985) (0.0336) Unemployment rate * 0.123*** (0.0306) (0.0250) (0.0500) R-squared Constant 3.277*** 2.948*** ** 3.178** (0.489) (0.768) (1.121) (1.261) (1.385) Observations 9,156 6,335 2, R-squared Dependent variables: European support Model 1, 2 and 3 are OLS regression with standard error corrected for nation (cluster) Robust standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1 Cases weighted Intune 2009

25 Table 4. Hypotheses and findings Theory H West East Italy Poland Cognitive Mobilization Rejected 1/5 2/5 X 1/5 Political Cues and Class Partisanship Accepted 4/5 1/5 X 1/5 Utilitarian Model Accepted 2/3 3/3 1/3 2/3 National Identity Rejected 6/7 5/7 3/7 2/7 Sociodemographic factors Accepted 3/6 1/6 1/6 1/6

26 Table 5. Description of the variables Dependent variable Description Source European Support Independent variables Positive Retrospect econ. Evaluation EU benefits nation EU benefits for people like me European Identity EU Identity is ascribed EU identity is achieved National identity National identity is ascribed National identity is achieved 0-10 scale as result of combination of two questions: 1) Membership (Do you thing that our country s membership in EU is a good thing, a bad thing, neither good nor bad? 2) Strenghten EU How do you think the general economic situation in (OUR COUNTRY) has changed over the last 12 months? Five point scale from Got a lot worse to Got a lot better Taking everything into consideration, would you say that (OUR COUNTRY) has on balance benefited or not from being a member of the European Union? Has not benefited or has benefited And what about of people like you? Have people like you on balance benefited or not from (OUR COUNTRY)'s EU membership? Has not benefited or has benefited 0-10 scale ad results of combination of four variables: How much does being a European have to do with how you feel about yourself in your day to day life? People feel different degrees of attachment to their town or village, to their region, to their country and to Europe. What about you? Are you very attached, somewhat attached, not very attached or not at all attached to the following? [EUROPE] Do you see yourself as? 1) (NATIONALITY) only; 2) (NATIONALITY) and European; 3) European and nat 4) European only; 5) None of the above How far do you feel that what happens to Europe in general has important consequences for people like you? 0-10 scale. And for being European, how important do you think each of the following is? (To be a Christian) (To be born in Europe) (To have European parents) 0-10 scale. And for being European, how important do you think each of the following is? (To respect European Union s laws and institutions) (To master any European language) (To exercise citizens' rights, like being active in politics of the European Union) 0-10 scale. People feel different degrees of attachment to their town or village, to their region, to their country and to Europe. What about you? Are you very attached, somewhat attached, not very attached or not at all attached to the following? [Your town/village] [Your region] [Our country] 0-10 scale. People differ in what they think it means to be (NATIONALITY). In your view, how important is each of the following to be (NATIONALITY): To be a Christian To be born in (OUR COUNTRY) To have (NATIONALITY) parents 0-10 scale. People differ in what they think it means to be (NATIONALITY). In your view, how important is each of the following to be (NATIONALITY)? Intune wave II, 2009

27 Trust other Europeans National institutional confidence EU institutional trust Left-Right Scale Party identification Political sophistication Influence on politics Education Non electoral participation To respect (NATIONALITY) laws and institutions To master (COUNTRY LANGUAGE) to master one of the official languages of (OUR COUNTRY) To exercise citizens' rights, like being active in the politics of (OUR COUNTRY) Please tell me on a scale of 0 to 10, how much you personally trust each of the following groups of people. '0' means that "you do not trust the group at all" and '10' means "you have complete trust" People in other European countries 0-10 scale results from combination of several variables: Please tell me on a scale of 0 to 10, how much you personally trust each of the following institutions to usually take the right decisions. '0' means that "you do not trust an institution at all" and '10' means "you have complete trust" The (NATIONALITY) Parliament The regional or local Government The (NATIONALITY) Government On the whole, how satisfied are you with the way democracy works in (OUR COUNTRY)? Are you? I am going to read a few statements on politics in (OUR COUNTRY) and in Europe. Could you please tell me whether you strongly agree, somewhat agree, somewhat disagree or strongly disagree with each of them? Those who make decisions in (OUR COUNTRY) do not care much what people like me think Those who make decisions in (OUR COUNTRY) are competent people who know what they are doing Please tell me on a scale of 0 to 10, how much you personally trust each of the following institutions to usually take the right decisions. '0' means that "you do not trust an institution at all" and '10' means "you have complete trust" The European Parliament The European Commission On the whole, how satisfied are you with the way democracy works in the European Union? Are you? Those who make decisions in European Union are competent people who know what they are doing In politics people sometimes talk of "left" and "right". Where would you place yourself on a scale from 0 to 10 where '0' means "the left" and '10' means "the right", and '5' means "neither left nor right"? ( 0 left - 10 right) Do you usually think of yourself as being close to any particular political party? (1 no, 2 yes) 0-10 scale as result of combination of: -Interest in politics -Effective knowledge of Europe On a scale from 0 to 10, where '10' means "a great deal of influence" and '0' means "no influence", how much influence do YOU have on politics? (1-11 scale) (recoded as 0 No influence 10 a great deal of influence) Level of education: 1) elementary 2) secondary 3) a-level 4) university Now a few questions about how active you are in politics and community affairs. In the last 12 months...? Have you worked for a party or a political candidate

28 Media consumption Aggregate variables Level of corruption.have you been active in a voluntary organization, such as a community association, a charity group, a sport club, etc. Have you been active in a voluntary organization, such as a community association, a charity group, a sport club, etc. Have you boycotted certain products or services for political or ethical reasons Combination of two questions: Normally, how many days a week do you watch the news on television? Normally, how many days a week do you read about politics in the newspaper? Corruption perception index (0-high corruption 10-low corruption) Transparency International Gdp growth Differences in % with previous year Eurostat Unemployment rate Annual unemployment rate Eurostat Annual inflation rate Annual inflation rate Eurostat Please note that great part of the recoding and computing new variables/indexes or scale was done by Intune Mass Group. Fabio Serricchio was directly involved in that group but the work done was a collective effort.