SYMBOLIC AND PRAGMATIC ASPECTS OF EUROPEAN IDENTITY 1

Transcription

1 György Lengyel Borbála Göncz Corvinus University of Budapest Institute of Sociology and Social Policy Original scientific paper UDK: (4)(439) Received: SYMBOLIC AND PRAGMATIC ASPECTS OF EUROPEAN IDENTITY 1 Simbolički i pragmatički aspekti evropskog identiteta APSTRAKT Čini se da je na mestu tvrdnja da je osnaživanje evropskog identiteta jedan od dugoročnih preduslova evropskih integracija. U suprotnom bi moglo doći do sve većeg jaza izmedju elita i stanovništva kada je u pitanju integracija. Dok se u zapadnoevropskim zemljama koncepti Evrope i EU često podudaraju, u istočnoevropskim zemljama Evropa ima prevashodno kulturno-istorijske konotacije a EU predstavlja ekonomski razvoj i blagostanje. Poredjenjem na međunarodnom nivou, evropski identitet pokazao se kao snažan u zemljama novim članicama, ali je takođe u nekima od njih (Mađarska i Estonija) nacionalni identitet bio među najsnažnijima. Ova studija zasnovana je na ispitivanju javnog mnenja predstavnika mađarskog stanovništva koje je obavljeno 2003, to jest pre ulaska Mađarske u EU. Pretpostavili smo da klasne pozicije, raspoloživost materijalnih, kulturnih i socijalnih resursa vrše jak uticaj na evropski identitet. Ispitivali smo dva aspekta identiteta: simbolički i pragmatički. Simbolički identitet meren je pitanjima koja se odnose na nacionalnu nasuprot nad- i pod-nacionalnoj pripadnosti, dok je pragmatički identitet ispitivan pitanjem u vezi sa pravilnom preraspodelom sredstava od poreza na različitim nivoima. Na taj način mogli smo da upoređujemo ove dimenzije i istražimo moguće razloge za nedoslednosti. KLJUČNE REČI Evropa, Mađarska, simbolički i pragmatički identitet ABSTRACT It seems realistic that one of the long-term preconditions of European integration is the strengthening of European identity. Otherwise, it might happen that a growing split occurs between the elites and the population in the question of integration. In the Western European countries the concepts of Europe and the EU frequently coincide, while in the Eastern European countries Europe has primarily cultural-historical connotations and the EU embodies economic development and welfare. In an international comparison, European identity was stronger in the newly joining countries, but in some of them (i.e. in Hungary and Estonia) the national identity was among the strongest as well. 1 Paper presented at the International Conference Contemporary Structural and Value Changes in CEE Societies, Belgrade, November 2005.

2 2 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 The current study is based on a Hungarian representative survey carried out in 2003 that is before Hungary joined the European Union. We supposed that class positions, the availability of material, cultural and social resources strongly influence European identity. We examined two aspects of identity, a symbolic and a pragmatic one. The symbolic identity was measured by questions addressing national vs. supra- and sub-national belonging, while pragmatic identity was approached by a question addressing the fair redistribution of taxes among the different levels. We could compare these dimensions and investigate the possible reasons for inconsistencies. KEY WORDS Europe, Hungary, symbolic and pragmatic identity Introduction Hungary joined the European Union on the 1 st May 2004 together with other 9 Eastern and Southern European countries. The referenda on the membership were held in the course of , followed by the European Parliament elections right after their accession on th of June 2004, when the citizens of these countries had the opportunity to employ one of the new rights acquired by being citizens of the EU. Looking at the turnover data, it seems that the citizens of the new member states did not make use of this right as much as their counterparts in the old member states did. In a national context, the legitimacy of a state or nation is often apprehended by its citizenship and the existence of a certain national identity. In Europe's case, the notion of European citizenship was introduced by the Maastricht Treaty (1993) in the European integration process together with the appearance of the questions of democratic deficit and the problem of legitimacy of the EU in the public and scientific discourses. The legitimacy question recently became even more important as, with the expanding number of member states, the EU and its institutions might become even more distant from the people than before. This phenomenon put forward the question of European identity. In the paper we are going to address the question of European identity in Hungary by first placing the concept in a theoretical context and drawing the attention to its problematic characteristics and then examining it through the results of a research done in Hungary on this subject. The data for this research was collected in April 2003 (before Hungary s accession to the EU) 3. 2 Malta ( ), Slovenia ( ), Hungary ( ), Lithuania ( ), Slovakia ( ), Poland ( ), Czech Republic ( ), Estonia ( ), Latvia( ). No referendum was held in Cyprus. 3 The sample size (n) is 1032, representative of the Hungarian population of 18 years old or more. Survey data collected by the TÁRKI Institute.

3 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects... 3 In our paper we assume that European identity is highly influenced by one s overall position in society, one s material, cultural and social resources. In order to better understand how people relate to the EU, we are examining two aspects of the identity, a symbolic and a pragmatic one. By the symbolic aspect we measure European identity as it is done in various previous researches (comparing different levels of belonging) whilst by the pragmatic aspect we try to measure how the tax should be redistributed by the different levels. Relying on this we try to measure what people would do for the belonging, what they would sacrifice. In this respect the symbolic attachment is rather a passive one whilst the pragmatic attachment is rather an active one. The problem: different aspects of European identity and their social context We approach the concept of identity as an attempt of individual or collective actors for self-definition. In this respect identity is a social construct, in which the genus proximum, the sense of belonging generated by the interaction between the individual and the social environment plays an important role. The identity can appear at different levels and can have different forms which emerge in different ways depending on the environment and the situation (Rusciano, 1998) - in this respect it is contextual and relational. The identity is highly influenced by the national self-perception but also by the perception of the nation by others what we think about how we are perceived by others. In this interpretation, to be able to talk about identity a third part, a relevant other is needed, against which the identity can be expressed (Rusciano, 1998; Horolets, 2003). Many scholars affirm that in the case of the EU this can be the USA or Japan (Optem, ). The question is to what extent are the different levels of the identity competing with each other or strengthening each other, especially in the case of the European and the national identities. It is possible that the national identity has a positive effect on the European one as both have their origin in the same factors (Inglehart, 1970; Duchesne-Frognier, 1995). Inglehart stated that the expansion of mass media and mass education had an overall integrative role which had its effects on the establishment of the identity both at European and national level. According to Inglehart, there are two main underlying factors when addressing European identity which are (1) the existence of cognitive mobilization capacities (capacity to internalise abstract information, which enables one to understand the complex 4 This was also confirmed by a qualitative research ordered by the European Comission and carried out with the coordination of Optem Institute in 2001 in the 15 member states and 8 candidate countries (4 focus group discussions per countries).

4 4 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 European institutional system and which helps to be more familiar with it) and (2) the role of certain values. The concept of European identity can be approached from several directions: in a stricter way it can be defined as to what extent the citizens of the EU perceive themselves as part of a single political community (Duchesne-Frognier, 1995), whereas in a wider context it can be defined as to what extent they perceive themselves as members of a symbolic cultural community. Regarding the EU, several scholars have emphasised the lack of symbolic, emotional links (Horolets, 2003); nevertheless the EU can be still perceived as a good, positive thing without any emotional attachment to it (Duchesne-Frognier, 1995). According to the results of a qualitative study (Optem, 2001), it is easier for one to identify with Europe than with the EU, since the common historical and cultural roots were more mentioned with regards to Europe, whilst the EU was more perceived as a bureaucratic economic entity. At the same time, in Western European countries Europe and the EU often turned up as synonyms, whereas in former candidate countries Europe had a wider content compared to the EU which was rather associated with the economic prosperity. It seems to be likely that one of the preconditions of the European integration in the long-term is an established and widely expanded European identity. Without this kind of common attachment a very volatile and unstable situation might occur. The European project, being mainly perceived as an elite-driven project so far, might become more and more distant for average people who could become rather indifferent to it. There might appear a clash between the opinion of elite and of the general mass public on the integration (Haller 2004). In general, members of the elites are more likely to understand the EU due to its complexity (cognitive mobilization capacities), whilst the general public might perceive it as a distant, bureaucratic entity. Here we can refer to the low participation level at the EP elections 5, but the recent referenda on the EU constitution are also a good example of the split between the elite and the public. When mentioning the low participation level at the EU referenda and the EP elections in the new member states vs. the old ones, we have to take into account that there are some wider differences in terms of European identity too. It is interesting to note that the European identity was stronger in former candidate countries than in the older member states according to a study carried out by the Gallup Institute in However, Hungary together with Estonia and Great Britain was one of the three countries where the proportion of people stating to have only national identity was the highest too. It appeared that the candidate countries were more committed to the idea of a unified Europe, but this was mainly 5 In the new member states the participation level was even lower but we are not addressing this issue in this paper.

5 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects... 5 due to the expectations of economic development that were associated with it (Optem, 2001). In the case of Hungary the low participation level at both the EU referendum and EP elections should be mentioned as a sign of passivity. Bozóki András and Karácsony Gergely state in their article addressing the question of identity (Bozóki- Karácsony, 2003) that one explaining factor of the low participation can be the fact that the referendum was about the membership and not about belonging. A free rider behaviour can also be mentioned here as people may have thought that the question was already decided and thus have not voted at the referendum. An other proposed explanation of this phenomenon is the lack of trust of the mass public towards the elite that they perceived the EU accession as a new trick from the part of the elite just as in the case of the democratic transformation. The peaceful, elite-led transition did not leave much chance to an active civic participation, thus the lack of trust in institutions and the elite persisted, whilst a wide range of social problems (poverty, elite change, political participation, public moral, etc.) have still not been resolved. All these can be reasons for the phenomenon of low participation and the parallel high proportion of yes votes, but also for the fact that the question of a supra-national membership became mainly a function of internal political issues and debates. Hypotheses In our paper we assume that European identity is highly influenced by the position in society, material, cultural and social resources. We are studying two aspects of the identity, a symbolic and a pragmatic one. Our main hypothesis is that both aspects of identity are strongly influenced by one s material, cultural and social resources, and that this influence is positive saying that the more material and cultural-social resources one has, the more likely one is to have European identity in both measured aspects. We also propose that these two aspects follow the same patterns and are consistent. However, in case there is an inconsistency between the two aspects, we assume that the symbolic attachment is rather defined by cultural and social factors, whilst the pragmatic aspect rather by material resources. We propose this because we hypothesise that the respondents will be consistent in the logic they adopt if we are asking them on tax-related issues, they will decide more upon their financial situation. Thus, in this paper we are examining the following 3 hypotheses: European identity is positively influenced by one s material, cultural and social resources.

6 6 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 Symbolic and pragmatic aspects of identity are consistent. If they are not consistent, this is due to the fact that symbolic attachment is rather influenced by cultural-social resources, whilst pragmatic attachment by material ones. Symbolic and pragmatic identity As mentioned before, the reason to separate two aspects of the European identity, which we call symbolic and pragmatic ones, is to better understand how people relate to the EU. By examining these two kinds of attachment, besides measuring belonging as a passive concept, we also try to measure what people would do for this belonging. The symbolic aspect of European identity is measured by a question addressing the different levels of belongings (I am European, I am Hungarian, I am citizen of my settlement). In Table 1 one can see that being Hungarian was mentioned first by three-quarter and even being citizen of one s settlement was mentioned more (one-fifth) than being European. Roughly about 29% of respondents mentioned European identity at the first or second place. Table 1: Symbolic attachment to different levels of belonging 6 Citizen of my European Hungarian N=1032 settlement First choice 5.9% 74.7% 18.9% Second choice 22.8% 22.4% 53.2% No choice/dk 71.3% 2.9% 27.9% 100.0% 100.0% 100.0% Pragmatic identity is measured by a question on how much of 100 HUF tax would be assigned to EU/ national/ local level for redistribution. One can see in Table 2 that indicating national and local levels for redistribution is quite similar on average 39-40% of tax would be assigned to both national and local level. Assignment for redistribution at EU level seems to be different, on average it would be only 15%. 6 Question: Which one is the most important for you? Being European/ Hungarian/ Citizen of your settlement? And which one is the second most important?

7 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects... 7 Table 2: Amounts of taxes assigned to different levels of authority 7 HUF EU level Hungarian level Local level Mean Std. Deviation The difference among the logics behind the decision on how much tax to assign to the different levels of authority is very clear when one looks at the distribution of the amounts (Figure 1). Whilst taxes at national and local level follow a nearly normal distribution, taxes at EU level show a different pattern, mostly lower amounts were mentioned. Figure 1: Amounts of taxes assigned to different levels of authority Std. Dev = 14,17 Std. Dev = 24,62 Std. Dev = 25,99 Mean = 14,7 Mean = 38,7 Mean = 39,8 0 N = 1016,92 0 N = 1016,92 0 N = 1016,92 0,0 20,0 40,0 60,0 80,0 100,0 0,0 20,0 40,0 60,0 80,0 100,0 0,0 20,0 40,0 60,0 80,0 100,0 10,0 30,0 50,0 70,0 90,0 10,0 30,0 50,0 70,0 90,0 10,0 30,0 50,0 70,0 90,0 EU level Hungarian level Local level Following this, we are going to consider as having symbolic European identity all respondents who have mentioned being European at the first or second place at the symbolic identity question and having pragmatic European identity those who would pay the largest amount/ second largest amount of tax for the EU at the pragmatic identity question. Although the latter is originally a continuous numeric variable, this kind of coding allows us to compare the consistency of the two kinds of identity; at the same time we are aware of the fact that in this way we lose part of the original information. Coded as mentioned, we can see that about a half of the respondents have pragmatic attachment to the EU and only 29% have a symbolic one (Table 3). 7 Question: If Hungary decides to join the EU, what would be fair according to you: utilization of how much of 100 HUF tax should be decided at EU/National/ Local level?

8 8 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 Table 3: Two aspects of European Identity Symbolic identity Pragmatic identity European 29.2% 50.6% Sub-European 70.8% 49.4% N= Nevertheless, when one studies European identity it is worth to state precisely if one refers to the identification with the EU or the emotional attachment to the beautiful old lady (Europe). The questions used were formulated in such a way that two dimensions were present: the symbolic vs. the pragmatic dimension on the one hand, and a European vs. EU-related identity on the other. The question destined for measuring the symbolic aspect was referring to Europe/ being European, whilst the question about the taxes was referring to the EU. As the research was done before Hungary s accession to the EU, the question related to the symbolic identity could not be phrased as I am a citizen of the EU. In an ideal case we would be able to measure both dimensions on their own and one within the other to be able to compare the attitude of members and non-members. In some of the similar researches a global level in addition to local/ national/ European levels is also included. Explanatory variables and methodology The variables we have chosen to answer our questions are addressing material, cultural and social resources. We are measuring the material resources with (1) the household income/capita (grouped in quintiles), (2) the occupation (being blue or white collar), (3) being shareholder or not and (4) being active or inactive on the labour market. Cultural resources are apprehended by (5) the education (elementary, professional, grammar school, higher education), (6) the internet usage (yes/ no), (7) how much time one spends reading newspaper (recorded as more or less than 30 minutes per day - the average time spent on reading). We have also included a variable destined to measure cognitive mobilization capacity as a measure of cultural-social resources. For this purpose we have used a question proposed by Inglehart, (8) the frequency of discussing politics with friends (often/sometimes/ never). Participation in the previous national elections (9) was included too, which is a sign of social mobilization capacity together with (10) political preferences (left/right) which we have left in as a control variable.

9 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects... 9 Beside the above mentioned we have included (11) age (18-24 years old, years old, 55 years old or more), (12) gender and (13) type of locality (village, town, city, and capital) as demographic and geographic control variables. Next, we will first examine the results of cross-tabulation analysis with Chi 2 statistics, afterwards completed with the results of a logistic regression model. We have decided to use logistic regression because both of our dependent variables (symbolic and pragmatic identity) were coded as binary (European/ sub-european) ones. This methodology enables us to obtain the odds of the probability of occurrence of European identity as a result of complex interaction of the explanatory factors. We will also be able to compare the effects of one variable onto the two aspects of identity by controlling the effect of the other variables (see Appendix 1 for more details on methodology). Social characteristics of symbolic and pragmatic aspects of European identity We have examined our first hypothesis about the influence of material, cultural and social resources on the different aspects of identity by cross-tabulation and Chi 2 statistics. Further in the text, statistically significant interrelations will be commented (see tables in Appendix 2). Symbolic identity In terms of material resources, being shareholder, being a white collar instead of blue, being active instead of inactive and having higher revenue indicate a higher probability to have symbolic European identity. In terms of cultural/ social resources we can also say that it has a positive influence: the more educated one is, the more likely one is to have a symbolic attachment to Europe, just as it is the case with Internet-user(s) and those spending more time reading papers than average. Cognitive mobilization capacity also has a positive influence: the more frequently one discusses politics with friends, the more likely one is to have symbolic identity. We can also say that people living in Budapest are more likely to consider themselves as being European, just as men and younger than 55 years of age do. Pragmatic identity Regarding pragmatic identity results are not so evident. Within material resources, occupation and being shareholder have no significant effect. As opposed

10 10 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 to the symbolic aspect, this time the inactive population is more likely to have pragmatic identity, just as people having lower revenue. In terms of cultural/ social factors education has also a significant impact, however, just in the opposite way as in the case of symbolic attachment: less educated people are more likely to assign their taxes to the EU level. Those who are less active in terms of democratic participation (have not voted at the previous elections) would also delegate more power of redistribution to the EU. There is also an interesting skew towards people who cannot place themselves in the domestic political sphere those who do not know whether they are rather left-wing or rightwing would also delegate more of their taxes to the EU level. However, reading more papers is still positively influencing this kind of European attachment. Looking at the results, it emerges that the pragmatic identity is rather characterizing lower strata in the society, less advantaged groups or people with less opportunity in life. The fact that they are rather female is also indicative of this. It appears that the two aspects are quite different in the dimensions studied here. Beside the real differences, this might perhaps also reflect methodological problems: to what extent is tax a hierarchy-related concept, how the coding was done, etc., however, in the following parts of the text we are concentrating on the social meanings of this phenomenon. Regarding our first hypothesis where we would have expected the different factors to have a strong and positive effect on the two aspects of the European identity, we have to conclude that in the case of symbolic identity this appeared to be true several material and cultural factors had positive impact on identity whereas in the case of pragmatic identity the results are not that evident. Material resources had no, or had negative effect on the identity. In terms of cultural resources the trends are just the opposite of the symbolic one; the less cultural/social resources one had at his or her disposa,l the higher was the proportion of European pragmatic identity. Only reading newspapers had positive impact on it. Here we could refer to an earlier work of Inglehart's (Inglehart, 1970), where he has found that amongst less educated people reading papers has a much stronger impact on supporting European integration process than amongst highly educated people this is rather about cognitive mobilization capacity. Thus, all in all, in the case of pragmatic aspect of European identity, the first hypothesis is rather falsified. Consistency of the two aspects of European identity We have already seen in Table 3 that a potential pragmatic attachment to the EU is wider than the symbolic European identity, and we have also seen that the

11 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects studied cultural/ social and material elements have a quite different influence on the two aspects. According to our second hypothesis we would expect all those who have symbolic European identity to have also a pragmatic attachment however this is not the case (see Table 4). Only 17% of the population have consistent European identity and another 37% have a consistent sub-european identity. This means that about half of the respondents had inconsistent answers when talking about symbolic and pragmatic kinds of attachments to Europe, to the EU. Table 4: Consistency of the two aspects of European identity Pragmatic Identity Total European Sub-European European Symbolic Identity 16.6% 12.6% 29.2% Sub-European % 36.9% 70.8% Total % 49.5% 100.0% Here we could consider an explanation based on the EU vs. Europe problematic already mentioned in the introduction part. According to these results people would be more likely to identify themselves with the EU than with Europe. However, this reasoning is in contradiction with other research results (Optem, 2001) stating that in the ex-candidate countries Europe generated wider identification. People already felt as part of Europe before the accession and this was one of the strongest reasoning why these countries should be part of the EU. Thus in this case we would suggest that there are other explanations behind these trends. In order to gain a more complete picture we have run logistic regression analysis for both aspects of the identity (see Table 5). Looking at the results of the logistic regression models one can see that they are very much in line with the trends of the cross-tabulation already mentioned, but this time the interrelations of the explaining variables are controlled and filtered. This gives us a more precise picture and allows us to compare the impacts of different factors on the identity. Although both are significant models, they are explaining only 6-8% of the total variance of the dependent variables we need to note that there should be other factors behind the identity that have more explaining power than the ones included in the analysis.

12 12 SOCIOLOGIJA, Vol. XLVIII (2006), N 1

13 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects Regarding cultural and social factors one can see that education has a significant impact in both cases, but in the opposite sense. Whilst a higher level education indicates a higher probability of a symbolic attachment, pragmatic identity is more likely to appear amongst people with lower level of education. A person with higher education is one and half times more likely to have a symbolic attachment than someone with elementary education, whilst the same person is twice less likely to have pragmatic attachment. Internet usage has a significant positive impact on symbolic identity, whilst reading papers is positively influencing pragmatic identity. Although its impact is not significant, cognitive mobilization is obviously influencing in an opposite way the two kinds of attachment; whilst discussing more often politics with friends increases the probability of symbolic attachment, it decreases the probability of the pragmatic one. Those who cannot position themselves in the domestic political sphere cannot identify themselves with either of the existing pole of power and are more likely to assign more tax to EU level. Amongst demographic control variables gender and locality have a significant effect on identity. Whilst those who live in Budapest are three times more likely to have a symbolic attachment to Europe than those living in a smaller town, females are more likely to have a pragmatic attachment. Regarding our third hypothesis we can say that symbolic identity is in effect explained by cultural resources; this part of the hypothesis is true. On the other hand pragmatic identity is also explained by cultural factors instead of being explained by material ones as we would have expected, therefore the hypothesis is falsified. Conclusion Symbolic identity performed as we expected upon the findings of other researches and theories of other scholars. It is mostly influenced by cultural factors, e.g. education, Internet usage, cognitive mobilization capacities according to Inglehart. However, the two studied aspects of European identity appeared to be quite inconsistent and, furthermore, they contained contradictory elements. Pragmatic identity performed differently than we have supposed. It was wider than the symbolic identity which had rather an elite characteristic and was characterizing rather the lower social strata with less education who had no real political opinion. There can be several explanations for this phenomenon. One of them is that we have found a different kind of attachment to the EU. Scholars are talking about the clash between elites and public s opinion towards the EU, that the elites are more supportive to the integration. However, we have mentioned that one can

14 14 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 consider the EU as a good, beneficiary thing (which is even more important in the case of countries with a lower level of economic development) without having an emotional attachment. If it is so, the fear of a clash between the elites and the public has less basis the wider public also has an attachment, even if it differs from the more alert minority s feeling of belonging. Liesbet Hooghe had a paper concerning the politics that should be delegated to EU level or kept on national authority she has found that the opinion of the elite and the public was different; both were supporting the idea of European integration but with a different content (Hooghe, 2003). However, it also can happen that, in the way we have measured the pragmatic identity, we are measuring a different phenomenon, other than the attachment to the EU. It may be that this is an indicator of the lack of trust in national institutions, domestic political system. As there are a lot of expectations towards the EU in terms of economic development and its representing the wealth and high living standard, people might think that by assigning more tax to that level more efficiency will be reached in the redistribution process. Scholars have found (Czichowski, 2000) that the support for the EU is the highest amongst the higher social strata but also amongst the lowest ones the logic behind is that once people do not trust anymore in the state to resolve their situation, they will turn to a higher level. Furthermore, these people have not much to loose. This explanation can also be applied to this case: people who have not very much tax to pay anyway are not so alert about whom to pay it to. To be able to decide which explanation is the right one, the question should be the subject of further examination. In this paper we have raised a question and drawn the attention to an interesting phenomenon which may point beyond European identity towards underlying and unsolved social problems in Hungary. References Bozóki András Karácsony Gergely (2003): Membership without Belonging? Hungary into the European Union: A Historic Step Passively Approved Central European Political Science Review, 13. Cichowski, Rachel (2000) Western Dreams, Eastern Realities: Support for the European Union in Central and Eastern Europe. Comparative Political Studies, 33 Duchesne, S. - A-P. Frognier (1995): Is there a European Identity? Public Opinion and International Governance. Ed. O. Niedermayer-R. Sinnott. Oxford University Press Haller, Max (2004),The Elections to the European Parliament 2004 An Example for the Increasing Split between Political Elites and Citizens, in György Lengyel (ed.) Cohesion

15 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects and Division of Economic Elites in Central and Eastern Europe. Budapest, Corvinus University Hooghe, Liesbet (2003): Europe Divided? Elites vs. Public Opinion on European Integration. [Paper presented at CEU, Budapest, 03/06/2003] Horolets, Anna (2003): Conceptualising Europe Through Metaphors: a Way to Identity Formation? Polish Sociological Review, I(141) Inglehart, Ronald (1970): Cognitive Mobilization and European Identity Comparative Politics 3. Janssen, J.H. (1991): Postmaterialism, Cognitive Mobilization and Support for European Integration British Journal of Political Science 21 Rusciano, F.L. (1998): World Opinion and the Emerging International Order. Praeger- Westport, Connecticut Optem S.A.R.L. (2001): Perceptions of the European Union - General Report on a Qualitative Study for the European Commission

16 16 SOCIOLOGIJA, Vol. XLVIII (2006), N 1 Appendix 1 Methodological notes An advantage of logistic regression models compared to standard regression and the analysis of discriminance is that (1) it does not suppose the explaining variables to form a meaningful structure, (2) does not expect homoscedasticity and (3) is not influenced by the distribution of the variables. However, the interpretation of the results is more complicated and the redundance amongst the variables included in the model cannot be measured as in the case of a linear regression. In the case of logistic regression models the output to be interpreted are odds which are not simply showing the direct impact of an explaining variable on the explained one as in the case of a linear regression. Instead of explaining the value of the explained variable, here the odds of the probability of the occurrence vs. the probability of non-occurrence is shown (Exp(B) or odds ratio ). As opposed to linear regression, logistic regression is not estimating the regression coefficients with the lowest square method but using maximum likelihood estimation which is maximizing the log-likelihood function (LL) to obtain the best fitting model. In the case of a linear regression a simple R 2 is enough to measure the goodness of the model whilst in a logistic regression model this measure is not enough, thus the following measures are used in this paper 8 : R 2 LA = (((-2LL 0 ) (-2LL M )) 2k) / (-2LL 0 ), where 2LL 0 is a transformation of the log-likelihood function with only positive values following approximately a Chi 2 distribution. It shows the error of the model without the inclusion of any explaining variables but only the constant. The -2LL M shows the error of the model after the inclusion of the explaining variables where k is the number of these variables. Thus we can see the proportional decline of the initial error due to the inclusion of the variables taking into account their number. Lambda measure calculated for the comparison of the observed classification (0 or 1) and the predicted classification of the model (0 or 1). It shows whether the model has significantly increased the right classification compared to a simple classification based on the modus. If the model put all cases in one group (one column of the classification table is empty) this measure cannot be calculated. Adjusted R 2 this measure shows the explaining power of the logistic regression model. It is calculated based on a linear regression where the explaining variable is the original variable and explained one is the predicted variable. 8 Székelyi Mária Barna Ildikó (2003): Túlélőkészlet az SPSS-hez, Typotex Kiadó It should be noted that the SPSS statistical software automatically calculates R 2 measures, but these are only approximative measures which not necessarily fall between 0 and 1, thus their interpretation is not evident these measures are usually called pseudo R 2.

17 György Lengyel, Borbála Göncz: Symbolic and Pragmatic Aspects Appendix 2 The two aspects of European Identities by Explaining Variables Symbolic Identity Pragmatic Identity Sub- European European Total Chi 2 Sub- European European Total Chi 2 Total 70.8% 29.2% 100% 49.4% 50.6% 100% Age % 31.2% 100% * 49.3% 50.7% 100% % 31.8% 100% 51.7% 48.3% 100% % 24.5% 100% 45.8% 54.2% 100% Gender Male 67.7% 32.3% 100% ** 53.2% 46.8% 100% ** Female 73.5% 26.5% 100% 46.1% 53.9% 100% Location Village 78.6% 21.4% 100% ** 44.5% 55.5% 100% - Town 74.5% 25.5% 100% 52.3% 47.7% 100% City 67.8% 32.2% 100% 53.7% 46.3% 100% Capital 51.5% 48.5% 100% 50.8% 49.2% 100% Occupation Blue collar 73.2% 26.8% 100% ** 48.8% 51.2% 100% - White collar 66.4% 33.6% 100% 50.4% 49.6% 100% Presence on the labour Inactive 73.3% 26.7% 100% ** 45.2% 54.8% 100% ** market Active 68.1% 31.9% 100% 54.0% 46.0% 100% Shareholder No 71.4% 28.6% 100% ** 49.2% 50.8% 100% - Yes 52.0% 48.0% 100% 56.8% 43.2% 100% Household income/ 1st 79.7% 20.3% 100% ** 50.4% 49.6% 100% ** capita (quintiles) 2nd 69.5% 30.5% 100% 42.1% 57.9% 100% 3rd 71.8% 28.2% 100% 50.0% 50.0% 100% 4th 65.0% 35.0% 100% 51.0% 49.0% 100% 5th 68.4% 31.6% 100% 58.2% 41.8% 100% Not reported 69.4% 30.6% 100% 44.1% 55.9% 100% Education Elementary 81.3% 18.7% 100% ** 41.8% 58.2% 100% ** Professional education 66.9% 33.1% 100% 47.9% 52.1% 100% Grammar school 65.7% 34.3% 100% 55.3% 44.7% 100% Higher education 62.6% 37.4% 100% 60.2% 39.8% 100% Internet usage Yes 57.0% 43.0% 100% ** 52.0% 48.0% 100% - No 73.7% 26.3% 100% 48.9% 51.1% 100% Time spent a day on 30 min or less 72.4% 27.6% 100% ** 51.6% 48.4% 100% ** reading papers More than 30 min 65.7% 34.3% 100% 42.5% 57.5% 100% Political discussion Never 74.7% 25.3% 100% ** 43.7% 56.3% 100% * with friends Sometimes 72.3% 27.7% 100% 51.2% 48.8% 100% Often 58.8% 41.2% 100% 53.4% 46.6% 100% Political ideology Left 67.3% 32.7% 100% % 48.2% 100% ** Centre 69.6% 30.4% 100% 52.0% 48.0% 100% Right 70.2% 29.8% 100% 55.7% 44.3% 100% DK/NA 77.2% 22.8% 100% 38.3% 61.7% 100% Participation to last No/ don't remember 74.5% 25.5% 100% % 61.2% 100% ** elections Yes 70.1% 29.9% 100% 51.5% 48.5% 100% Chi2 Significance: ** < 0.05, * < 0.1