Stalled Political Euroscepticism Despite Erosion of Trust in the EU: The Case of Immigration Policy

Transcription

1 Stalled Political Euroscepticism Despite Erosion of Trust in the EU: The Case of Immigration Policy Domanov Aleksey, LCSR HSE, MGIMO-University, Moscow Abstract The study deals with predictors of the reluctance of EU citizens to communitarize a policy, i.e. to transfer powers to the EU from national level (political Euroscepticism). The subject is worth examining, as the unwillingness appears to determine the extent to which the EU can afford to deepen integration in order to resist the current economic crisis. Furthermore, Euroscepticism has been considerably analysed from supply side of parties (party-based Euroscepticism), but demand side (public Euroscepticism) remains underanalysed or analysed on national level without EU-wide generalizations. The study is based on theories of basic trust and trust in institutions, moreover the author adopts a securitization perspective. The unwillingness to communitarize a policy is analysed with the use of aggregated Eurobarometer, Eurostat and European Commission data on the example of the attitude towards common immigration policy in At first stage the author defines main predictors of public Euroscepticism with the help of correlation analysis. In order to assess the force of the relationships, we carry out regression analysis. Preliminary factor analysis lets us avoid endogeneity and multicollinearity and makes the regression analysis more valid.

2 Introduction As in response to the economic crisis and political unrest the EU seeks to deepen integration in some policy areas, it seems important to forecast what new policies soon may be within the EU s competence. Therefore, it is interesting to see the factors influencing the decision of national governments to confer powers to the EU. Decisions of democratic national governments are supposed to depend on the will of peoples, so we will look for these factors in decrease in unwillingness of citizens to communitarize policies (i.e. to make them supranational, common policies). The power of this factor was demonstrated when the EU leaders agreed upon a breakthrough in the policy area (the European Pact on Immigration and Asylum of 2008 establishing foundations of the common policy) in 2008 as soon aspercentage of the EU citizens unwilling to let bodies of the EU to take part in decision-making in immigration policy area decreased from 49% in June 2000 to 33% in October Thus, in pursuit of predictors of this unwillingness, we address the following research question: what does the unwillingness to transfer powers in immigration policy to the supranational level depend on? So far the public attitude towards communitarization remains relatively underanalysed in comparison with parties attitude. In relation to parties the explored phenomenon was clearly defined as political Euroscepticism. Lubbers and Scheepers reconsidered the studies like Gabel (1998) and De Winter and Swyngedouw (1999), who examined if people were ready to allocate decision-making responsibilities to the EU in some policy areas, and determined political Euroscepticism as public refusals to provide more legitimate power to supranational institutions to deal with policy issues (Lubbers and Scheepers 2005: 224). It should be noted that political Euroscepticism defined that way was different from usual meanings of Euroscepticism: - from what Lubbers and Scheepers called instrumental Euroscepticism (assumption that membership of a country in the EU costed more than gave back benefits) and - from the meaning in De Vries and Edwards (2009) (desired speed of European integration). Nevertheless, public attitude towards deepening European integration seems to be at least of the same importance as party-based Euroscepticism. For instance, constructivist approach to international relations (Wendt 1999) presupposes that we have to consider society and citizens to be crucial international relations actors, as they shape identity of states. The study seeks for predictors of (public) political Euroscepticism. Usually, facing a problem, people have two possible solutions: either to suppose that the problem will be managed by national authorities and that the EU shouldn t interfere with the policy area, or to call for 1

3 supranational intervention and to vest the EU with powers in the policy area. We look for the conditions under which people prefer not to make supranational bodies responsible for the policy. We will explore people s reluctance to confer additional powers to the EU on the example of political Euroscepticism about common immigration policy, as problems related to migrants are difficult to solve unilaterally. Thus, according to national difficulty hypothesis of Dalton and Eichenberg (1998), supranational governance of such difficult issues is more desirable for citizens, than issues that seem to cross borders to a lesser extent for example, sociocultural policies on education, health care and culture (Lubbers and Scheepers 2005: 226). A decline in public political Euroscepticism about common immigration policy in may have allowed politicians to communitarize the policy area by adopting the EU Immigration Pact in 2008, therefore this form of Euroscepticism is worth being explored. Nevertheless, the case-study of one policy area (immigration policy) will outline the principle of political Euroscepticism with respect to all the policy areas. Indeed, Gabel and Anderson (2002), Lubbers and Scheepers (2005) confirm, that, if one is sceptical about extension of powers of the EU concerning an issue, one is more likely to doubt the Union s interference in any other issue. The work doesn t just cover the topical issue for the EU, but also attempts to fill a gap in researches of Euroscepticism. As we mentioned above, the phenomenon has been considerably analysed from supply side of parties (party-based Euroscepticism), but demand side (public Euroscepticism) remains underanalysed or analysed on national level with few EU-wide generalizations (for example, analysed only in the Netherlands: Lubbers and Jaspers 2011). The famous study by De Vries and Edwards (De Vries and Edwards 2009) also does not cover the given issue: although the research is EU-wide and explores public Euroscepticism, the choice of dependent variable is arguable. The word speed in the Eurodynamometer question What is the current speed of building Europe? And which corresponds best to the speed you would like? may be understood differently by different respondents, and there are no negative answers corresponding with desire to turn the integration around and to hinder it. Moreover, we use newer data and broader range of independent variables than De Vries and Edwards. Theoretical framework At first sight, the decrease in political Euroscepticism may be explained in terms of basic trust and trust in institution, explored by Luhmann, Giddens, Sztompka, Fukuyama, Coleman, Quéré. Trust is a precondition of any cooperation (Coleman, 1990; Deutsch, 1958; Gambetta, 1988), i.e. also of vesting powers in an institution. It is worth distinguishing between cognitive and active components of trust: for example, in our case people may consider the EU trustworthy (cognitive component), but not dare to provide it with additional powers (active component). 2

4 Indeed, simple trustworthiness of an object is not enough for entrusting powers: as Sztompka emphasized, trust is an active attitude. Quéré (2010) also distinguished between trustworthiness and entrustment (i.e. action of trust, real cooperation with an object). It should be mentioned that trust in the EU seems to be a safer independent variable than trust in national authorities, which turned out to have insignificant effect on a similar indicator of Euroscepticism used by De Vries and Edwards (2009). In our case, we argue that active component of trust in the EU (unwillingness or desire to vest powers in the EU, i.e. political Euroscepticism) depends on its cognitive component (usually measured trust in an institution). However, judging by Eurobarometers of , trust in the EU is poorly correlated with the political Euroscepticism in immigration policy (about ; see the indicator of political Euroscepticism in Variables and Hypotheses section below). Therefore, we have to look for less obvious predictors. Conferring powers to the EU may also seem suspicious for the people who are very much attached to their country. Indeed, political Euroscepticism may be perceived as a manifestation of nationalism, as the EU governance is often opposed to sovereignty of nation-states (De Winter and Swyngedouw 1999: 49) and nationalism is said to be common denominator of many Eurosceptic positions (Halikiopoulou et al., 2010). Adopting a securitization perspective, we observe that it is growing threat that can make people seek strong governance of the policy field and make them call for supranational intervention. The securitization theory of Buzan and Wæver explains the role of the actors that operate inside nation-states (e.g., civil society actors and individuals) in foreign and security policy of states. According to the theory, it is public discourse that shapes foreign and security policy: a threat becomes so unbearable that citizens urge the authorities to combat the problem using extraordinary instruments, then under that pressure state pays attention to the threat and includes this issue in its security policy. It seems like we may continue these ideas and apply the approach to communitarization of immigration policy: people perceive a big inflow of migrants as a threat, urge their governments to fight the problem (for example, using the resources of the EU and other member states). People wanting to vest powers should perceive a threat to the policy object big enough to call for supranational intervention. A partial confirmation of this assumption is found in Lubbers and Jaspers (2011), who noticed stable importance of perceived ethnic threat for Eurosceptic attitude. It would be logical to account for the number of immigrants coming in a country. However, some scholars assume it to be insignificant factor of perceived immigrant threat, in contrast to perceived size of immigrant population (Ceobanu and Escandell, 2008; Semyonov et al., 2008; Strabac and Listhaug, 2008). 3

5 The perception of immigrant threat may be influenced by other threats experienced by a respondent (his or her «general threatened condition ). For instance, the condition provoked by economic scarcity. Indeed, several scholars have shown that support for or opposition to European integration varied with macroeconomic indicators (Anderson and Kaltenthaler, 1996; Eichenberg and Dalton, 1993). From previous studies we derive the following possible macroeconomic factors of political Euroscepticism: aggregated EU GDP (Hix 2005: 595), GDP per capita (Eichenberg and Dalton 1993), inflation measured as Consumer Price Index (De Vries and Edwards 2009), unemployment rate (De Vries and Edwards 2009; Olzak 1992; Koopmans et al. 2005) and institutional variables like type of welfare state and variety of capitalism (Brinegar, Jolly and Kitschelt 2004). However, our study of public opinion should be concerned with the public view of economic conditions: with perceived economic growth, perceived inflation etc. In case of the EU, we focus on the assessment of the situation in European economy. Besides we are interested, to what extent does a respondent feel safe economically in such conditions. De Vries and Edwards (2009) suggest measuring that characteristic with an economic anxiety index composed of evaluations of household financial and personal job prospects (or expectations, as measured in the analysed Eurobarometers). However, as our study is done on aggregate level, it would be more convenient and precise to replace job prospects by the above-mentioned unemployment rate that does not depend on accuracy of a poll sample. We also suggest using satisfaction with life as a comprehensive indicator of the general threatened condition reflecting all the threats to a respondent, which were not taken into account by the above-mentioned variables. In order to entrust anything, people have to be sure that there is no risk, that their interests will not be taken into account or betrayed by the entrusted actor (Baier, 1986; Hardin, 1998). It seems reasonable to divide the determinants of risk assessment in two groups: "present dimension" and "future dimension". Present dimension includes estimations of the current state of affairs: if the EU plays a positive role in immigration policy, if it commands existing powers beneficially for the country (i.e., in the words of Lubbers and Scheepers, instrumental Euroscepticism). An objective measure of benefits from membership of the EU (for instance, the most obvious: net transfers from the EU budget to the country) should also be regarded as a possible predictor. The interdependence of political and instrumental Euroscepticism was confirmed by Lubbers and Scheepers (2005). Moreover, De Winter and Swyngedouw (1999) showed how strongly the latter affected the former. Likewise, assessment of membership of the EU as a bad thing and overall image of the EU may strongly influence the level of political Euroscepticism. 4

6 In addition, people who are unaware of the EU bodies may be more unlikely to entrust additional powers to the Union. Indeed, Luhmann (1979), Giddens (1988), Hardin (1993) argue that one can reduce risk of misplaced trust by improving the amount and quality of information about the object. The theory of bounded rationality also suggests that human s decisions may vary from acceptable to optimal with increasing the amount of disposable information. Moreover, awareness was proven to be important for Euroscepticism: a focus-group study conducted by Göncz in 2009 (see Lengyel 2010) showed that respondents became more Eurosceptic after they had got new information during group deliberation. People will consider risk of vesting powers to be low if they are informed about the EU. On the contrary, people will not trust the EU to manage immigration policy if they do not know anything about the EU bodies, which would pursue a common policy. Thus, it is important to measure the awareness of the European Commission namely, as it is responsible for execution of decisions in any common policy. As to "future dimension", it includes perceived ability to influence the EU s policy in the future (i.e., satisfaction with democracy in the EU). Norris (1999) showed that dissatisfaction with supranational EU institutions and the lack of democratic procedures was a sign of Euroscepticism. Democratic deficit has an effect on the desire to confer power to the EU (Hooghe, 2003) and the perceived lack of accountability erodes legitimacy of the EU (Scharpf, 1970). Furthermore, it seems important to include crime rate in our model, as citizens may blame the EU for failure in prevention of crime within the framework of cooperation in home affairs, and that assumption may alienate them from the EU. Besides, De Vries and Edwards (2009) suggest testing occupation as a dummy variable for unskilled manual workers. In our case we suggest testing a relationship to the rate of industrial workers in a country. Variables, data Dependent variable is the typical measure of political Euroscepticism (Lubbers and Scheepers, 2005): answers Should be decided by the (national) Government to the question «For each of the following areas, do you think that decisions should be made by the (nationality) Government, or made jointly within the European Union? Immigration» of Eurobarometer. Timeframe for our longitudinal study is limited by the number of Eurobarometer surveys when the dependent variable was measured. The question was asked 14 times almost twice a year from November 2003 to November In sum, there are 378 observations for 27 EU countries in our dataset. Consequently, for this period we aggregated answers to Eurobarometer questions measuring the dependent variable and independent variables (see the table of questions measuring independent variables below) by using Eurobarometer Interactive Search System. We also calculated crime rate, 5

7 unemployment rate, rate of industrial workers in total number of employees and migrant influx rate for every country in every of these 14 timepoints relying on Eurostat data and net transfers to country from the EU budget in relation to the country s GNI by using the website of the European Commission (see links to sources in the end). Question Options Please tell me how attached you feel to (our country)? Very attached, fairly attached, not very attached, not at all attached. How would you judge the current situation in each of the following? 1 5 The situation of the European economy What are your expectations for the next twelve months: will the next 1 4 twelve months be better, worse or the same, when it comes to...? The financial situation of your household On the whole, are you very satisfied, fairly satisfied, not very Very satisfied, fairly satisfied or not at all satisfied with the life you lead? satisfied, not very satisfied, not at all satisfied. Tend to trust, tend not to trust. For each of the following institutions, please tell me if you tend to trust it or tend not to trust it? The European Union. How about the way democracy works in the European Union? Very satisfied, fairly satisfied, not very satisfied, not at all satisfied. I feel I am safer because (our country) is a member of the European Union. Generally speaking, do you think that (your country''s) membership of the European Community is...? Taking everything into consideration, would you say that (your country) has on balance benefited or not from being a member of the European Community? In general, does the European Union conjure up for you a very positive, fairly positive, neutral, fairly negative or very negative image? Have you recently seen or heard, in the papers, on the radio, or on television, anything about the European Commission in Brussels, that is the Commission of the European Community? For each of the following issues in (our country), do you think that the European Union plays a positive role, a negative role or neither positive nor negative role? Immigration Table 1. Questions measuring independent variables. Tend to agree, tend to disagree A good thing, a bad thing, neither good nor bad Benefitted, not benefitted Fairly positive, very positive, fairly negative, very negative, neutral Yes, no Positive role, negative role, neither positive nor negative role The sample of Eurobarometer is 1000 men per country (except 500 in Luxembourg, 2000 in Germany and relatively more in the United Kingdom: 1000 in Great Britain and 300 in Northern Ireland). 6

8 Hypothesis and method I test the following hypothesis: H1: The level of political Euroscepticism about immigration policy is positively related to: - expectations of respondents for the year to come about financial situation in their households; - satisfaction with life of a respondent; - assessment of situation in the European economy; - assessment of membership of the EU as a bad thing; - attachment to nation-state; - crime rate; and negatively related to: - awareness of the European Commission; - trust in the EU; - satisfaction with democracy in the EU; - perceived benefits for respondent s country from its membership of the EU; - net transfers from the EU budget to their country; - overall image of the EU; - the assessment of the role of th EU in immigration policy; - unemployment rate; - the rate of industrial workers in a country. This hypothesis is tested in two stages. Firstly, using correlation matrix we sort out the significant predictors of the dependent variable. We look for significant correlations between the dependent and the independent variables on aggregated country level. Secondly, seeking to assess the force of the relationships, we carry out linear regression analysis. But before that, as a preparation for the regression analysis, we group independent variables by conducting factor analysis (maximum likelyhood method, rotation method promax). The factor analysis is necessary, as it removes endogeneity problems and multicollinearity between the grouped independent variables. Then we test several models, in which the dependent variable is regressed upon the factor values or the factor values supplemented with the independent values having no correlation with them. The tests let us choose the best model (and, consequently, the best factor) by comparing R-squares of the model, as utility of a model or a factor is characterized by percentage of variance in the dependent variable explained by it. Afterwards the factor is checked with the help of two-level confirmatory factor analysis. We need this step to ascertain that the factor is valid for every EU-27 country, i.e. that there is no 7

9 considerable variation in factor loadings across the studied countries. That would be shown by a good model fit of the confirmatory factor analysis. In the end regression coefficients of the best model including this factor are presented. Results Table 2 is the correlation matrix. Surprisingly, correlation coefficients of the dependent variable with immigrant inflow rate, bad assessment of situation in European economy and attachment to nation-state appeared to be insignificant. These variables were omitted during factor analysis. During the factor analysis the independent variables with significant correlation coefficients were grouped in different combinations. As a result, we discovered several factors having strong factor loadings and explaining more than 60 percent of variance of their composing variables. We regressed the dependent variable on every factor in order to choose the best one (explaining the biggest percent of variance, i.e. having the biggest R-square in linear regression; see table 3). All regression coefficients in the models are significant (p<.05). Variable Independent variable Correlation number Socioeconomic coefficient 1 - Unemployment rate -.236** 2 - Immigrant inflow rate to population, per year Crime rate.579** 4 - Rate of industrial workers -.267** 5 - Net transfers to the EU budget, percent of GNI -.296** Public opinion 6 - Bad situation in European economy Attached to nation-state Expectations about household financial situation for next year: worse -.355** 9 - Not very or not at all satisfied with life -.485** 10 - Distrust in the EU.519** 11 - Dissatisfaction with democracy in the EU.524** 12 - Do not feel safer because a country is a member of the EU.392** 13 - Assessment of membership of the EU as a bad thing.476** 14 - Country didn t benefit from membership to the EU (instrumental.345** Euroscepticism) 15 - Negative overall image of the EU.582** 16 - Heard about the European commission -.314** 17 - Negative role of the EU in immigration policy (only before 2008).427** Significance: ** p<0.01 (two-fold) ; * p<0.05 (two-fold) Table 2. Correlation matrix. 8

10 Model Independent variables composing a factor R- square 1 Dissatisfaction with democracy in the EU, distrust in the EU, negative overall.411 image of the EU, crime rate 2 Dissatisfaction with democracy in the EU, distrust in the EU, negative overall.368 image of the EU, membership to the EU bad and/or unbeneficial 3 Distrust in the EU, negative overall image of the EU, membership to the EU.328 bad and/or unbeneficial 4 Expectations about financial situation, not very or not at all satisfied with life,.22 net transfers to the EU budget 5 Dissatisfaction with democracy in the EU, net transfers from the EU budget, do.194 not feel safer because a country is a member of the EU 6 Membership to the EU bad and/or unbeneficial, do not feel safer because a.169 country is a member of the EU Table 3. Regression of the dependent variable on factors. The factor included in the first model explains the biggest part of variance and has the biggest predictive value. However, perhaps other models will appear to be more valuable. These are the models that include not only a factor, but also an uncorrelated independent variable (having no significant correlation with the dependent variable, but not with the factor included in the same model). Following the same logic, we have to test regression models with separate uncorrelated independent variables. Table 4 shows R-squares for multivariate linear regressions of these models. Thus R-squares of models 7 and 8 exceed the one of model 1. Model 7 seems to be more logical, as the factor in the model groups more similar independent variables (all related to the EU), while in the factor of model 8 crime rate stands apart. As model 7 appeared to be the best model, we should name it (for example, caution in regard to the EU ) and present its parameters in detail. Factor loadings are presented in table 5 (sum of squares of loadings explain 77,4% of variance of respective variables). 9

11 Model Independent variables composing a factor Independent variable R- square 7 Membership to the EU bad and/or unbeneficial, Heard about the EC.434 distrust in the EU, negative overall image of the EU 8 Dissatisfaction with democracy in the EU, Unemployment rate.433 distrust in the EU, negative overall image of the EU, crime rate 9 Dissatisfaction with democracy in the EU, Unemployment rate.398 membership to the EU bad and/or unbeneficial, distrust in the EU, negative overall image of the EU 10 Dissatisfaction with democracy in the EU, net transfers from the EU budget, do not feel safer because of the EU Unemployment rate Membership to the EU bad and/or unbeneficial, do not feel safer because of the EU Unemployment rate Negative role of the EU in.24 im.migration policy, unemployment rate 13 - Negative role of the EU in im.migration policy, heard about the EC.313 Table 4. Regression of the dependent variable on factors and separate uncorrelated independent variables. Loadings Commonalities (extracted) Membership to the EU unbeneficial Membership to the EU bad Distrust in the EU Negative overall image of the EU Table 5. Factor loadings of the factor in model 7. Then I proceeded to ensuring the validity of this factor with the help of two-level confirmatory factor analysis. On this stage I assumed that there is variance of factor loadings not only among respondents within every EU member state, but also between various countries. The task was to show the robustness of the factor across all the countries by computing factor loadings on both individual and country levels. As individual level values were aggregated by countries on previous steps, while we have to examine factor loadings on individual level, a survey containing data specific to every respondent was needed. In this case, the most recent Eurobarometer containing questions both about the 10

12 dependent variable and the questions included in our factor is Eurobarometer 73.4 of June The results of the two-level confirmatory factor analysis are given below. We see that the robustness of the factor is almost good: just RMSEA is slightly bigger than.06 and TLI is slightly less than.9. Moreover, if we take into account only the ten countries of EU enlargement of 2004, robustness of the factor becomes perfect, as RMSEA and TLI assume totally acceptable values. P-value.079 for the residual variance of good assessment of membership does not create any trouble, as that is a two-tailed p-value, while the variance can never be negative, so I need a one-tailed p-value. In other words, I may rightfully divide the p-value by two and get the p- value of.038 an acceptable one. Unfortunately, I did not manage to find the scalar invariance of these results across all the countries, as the tests for scalar invariance of the factor appeared to have considerably worse model fits: on the sample of the EU-27 and the enlargement countries). Nevertheless, in the absence of scalar invariance I have at least configural invariance of the factor. As my research question and theoretical framework did not presuppose the presence of scalar invariance, the given list of individual variables composing this factor remained the same. Finally, after the confirmatory factor analysis verified the validity of the factor for all the EU-27 countries, I have a right to assume that the factor may indeed serve as a latent independent variable in model 7. Hence, regressing the dependent variable on it will provide meaningful results. The regression parameters of model 7 are as follows: R=.659, intercept=49,958, b factor =.575, b heard about the EC=-.326 (coefficients are standardized and significant on.001 level). 11

13 Conclusion All in all, by analysing models 7 and 8 I found out that 43,4% of variance in public political Euroscepticism about immigration policy were explained by awareness of the EC and caution in regard to the EU (a latent factor composed of membership to the EU perceived as bad and unbeneficial, of distrust in the EU and of negative overall image of the EU). Besides, first relationship was negative, while second was positive and stronger than first. The discovered latent factor appeared to be valid enough: it had configural invariance across the EU-27 and the EU-10 countries in June 2010, i.e. components of the factor were identical in all the countries. The findings show that the factor may be used for other studies of public attitude towards the EU during different timeframes. Moreover, the dependent variable had links to other predictors sorted out during correlation analysis. According to its results, the hypothesis about the direction of links was confirmed. Indeed, the level of political Euroscepticism about common immigration policy in a country is higher: - the better financial situation citizens expect for their household in the year to come; - the more citizens are satisfied with life; - the less they trust in the EU; - the less they are satisfied with democracy in the EU; - the less they think that the membership to the EU gives them safety; - the worse and the less beneficial they assess the membership of their country to the EU and the less transfers from EU budget their country gets; - the more negative overall image of the EU they have; - the less they are aware of the European Commission; - the more negatively they assess current role of the EU in immigration policy; - the higher crime rate in their country is; - the higher unemployment rate is; - the lower the rate of industrial workers is. The negative relationship of the dependent variable to unemployment rate and the positive relationship to expectations about financial situation in households, to satisfaction with life and to assessment of situation in the European economy seem to be the most curious. I would suggest that citizens do not need any supranational intervention in their lives, if situation is good without the involvement of the EU (e.g. concerning unemployment): why entristing governance to some new organs and assigning them taxpayers money? This causal chain upholds securitization theory, which predicted such relationships. Unfortunately, the analysis did not give us certainty about directions of links between the dependent variable and immigrant inflow rate, bad assessment of situation in European economy 12

14 and attachment to nation-state, as correlation coefficients appeared to be insignificant. It is probable that missing data for some independent variables in some years and generally short period of observations did not let us to achieve the significance of results. Or maybe the situation in national economy is more important than situation in the European one. However, this is hardly possible, as such conclusion would contradict the study of McLaren (2007), who found that attitudes towards European integration were driven mostly by EU-level variables than by national-level assumptions. More possible, the perception of economic condition simply does not absorb all the explored socioeconomic variables mentioned in the section Theoretical framework or we need more observations for significant correlation coefficient. 13

15 Data sources Budget section of the European Commission website ( ) Eurobarometer Interactive Search System ( ) Eurostat ( ) Bibliography Anderson, C.J. & Kaltenthaler, K.C. (1996). The dynamics of public opinion toward European integration, European Journal of International Relations 2(2): Baier, A.C. (1986) Trust and Antitrust. Ethics 96: Brinegar, A., Jolly, S., Kitschelt, H. (2004) Varieties of Capitalism and Political Divides Over European Integration. In: G. Marks and M. Steenbergen (eds.) European Integration and Political Conflict. New York: Cambridge University Press. pp Buzan, B. and Waever, O. (2003) Regions and Powers: The Structure of International Security. Cambridge: Cambridge University Press. Coleman, J.S. (1990) Foundations of Social Theory. Cambridge, MA: Harvard University Press. Dalton, R.J. and Eichenberg, R. (1998) Citizen Support for Policy Integration. In W. Sandholtz and A. Stone Sweet (eds.) European Integration and Supranational Governance. Oxford: Oxford University Press: Deutsch, M. (1958) Trust and suspicion. Journal of Conflict Resolution 2: De Vries, C.E. and Edwards E.E. (2009) Taking Europe to Its Extremes: Extremist Parties and Public Euroskepticism. Party Politics 15 (1): De Winter, L. and Swyngedouw, M. (eds.) (1999) The scope of EU government. Oxford: Oxford University Press. Eichenberg, R.C. and Dalton, R.J. (1993) Europeans and the European Union: the dynamics of public support for European integration. International Organization 47: Gabel, M.J. (1998) Interests and Integration: Market Liberalization, Public Opinion, and European Union. University of Michigan Press. Gabel, M.J. and Anderson, C.J. (2002) The Structure of Citizen Attitudes and the European Political Space. Comparative Political Studies 35 (8): Gambetta, D. (ed.) (1988) Trust: making and braking cooperative relations Oxford, N.Y.: Basil Blackwell. 14

16 György, L. (2011) Supranational Attachment of European Elites and Citizens. Europe-Asia Studies 63 (6): Halikiopoulou D. (2010) The Paradox of Nationalism: The Common Denominator of Radical Left and Radical Right Euroscepticism. London: LSE. Hardin, R. (1998) The Dear Self and Others. Annual Review of Law and Ethics. 6: Hardin, R. (1993) Altruism and Mutual Advantage. Social Service Review. 67: Hix, S. (2005) The Political System of the European Union. 2 nd ed. Basingstoke: Palgrave Macmillan. Hooghe, L. and Marks, G. (2003) Unraveling the Central State, but How? Types of Multi- Level Governance. The American Political Science Review. 97 (2): Hooghe, L., Marks, G. and Wilson C.J. (2002) Does left/right structure party positions on European integration? Comparative Political Studies. 35 (8): Koopmans, R. et al. (2005) Contested Citizenship: Immigration and Cultural Diversity in Europe. University of Minnesota Press. Kopeckỳ, P. and Mudde, C. (2002) The Two Sides of Euroscepticism: Party Positions on European Integration in East Central Europe European Union Politics 3 (3): Lubbers, M. and Jaspers, E. (2011) A longitudinal study of euroscepticism in the Netherlands: 2008 versus European Union Politics 12 (1): Lubbers, M. and Scheepers, P. (2005) Political versus Instrumental Euro-scepticism: Mapping Scepticism in EuropeanCountries and Regions. European Union Politics 6 (2): Luhmann, N. (1979) Trust and Power. Toronto: Wiley. Marks, G. and Wilson, C.J. (2000) The Past in the Present: A Cleavage Theory of Party Positions on European Integration. British Journal of Political Science. 30: McLaren, L. (2007) Explaining Mass-Level Euroscepticism: Identity, Interests, and Institutional Distrust. Acta Politica 42: Norris, P. (1999) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press. Olzak, S. (1992) The Dynamics of Ethnic Competition and Conflict. Stanford: Stanford University Press. Quéré, L. (2005) Les «dispositifs de confiance» dans l'espace public. Réseaux. 4 (132): Scharpf, F. (1970) Die politischen Kosten des Rechtsstaats: Eine vergleichende Studie der deutschen und amerikanischen Verwaltungskontrollen. Tübingen: J. C. B. Mohr, Taggart, P. (1998) A Touchstone of Dissent: Euroscepticism in Contemporary West European Party Systems. European Journal of Political Research 33 (3):