Attitudes on (future) EU integration and cooperation among youth: Regional divide?

2.1 European identity

European citizenship to become meaningful for its citizens has to be based on a sense of shared identity and common belonging (i.e. European identity) [6]. This common identity is crucial for at least three reasons. When identifying with the EU, citizens will a) accept decisions that are made in Brussels [6, 7]; b) be more prone to overcome divisions between and stereotypes towards different countries within Europe; c) be more prone to develop cosmopolitan dispositions, linked to greater trust and tolerance of immigrants, and ethnic minorities [8, 9].

European identity has been studied extensively over the past decades, however, the multifaceted nature of this construct makes it difficult to define European identity unambiguously [10]. For example, it can be distinguished between civic and cultural components of European identity [12]. A European civic identity refers to the perception to be part of the EU political system that defines rules, laws, and rights with relevance to one’s own life. Its identifiers are related to the experience of open borders, mobility of citizens, common civic area, and economic prosperity. A European cultural identity refers to shared culture, values, and history. Its identifiers are related to peace, harmony, the fading of historical divisions, and cooperation between similar people and cultures. Similarly, it can be differentiated between two forms of European identity –synonymic and synecdochical ones [12]. The synonymic concept presents the EU and EU membership as the only possible way the nation and society can partake in the European identity. The synecdochical concept understands the European identity as something that can be developed and constructed in wider Europe (not necessarily within the EU.

Based on its different conceptualizations of European identity, it can in general be described as “a feeling of being European as an integral part of one’s own social identity” [13]. According to this understanding, it is important to point out that there is a wide scholarly agreement that adolescence is a decisive period for political identity formation [14] and that a sense of European identity starts to emerge in the middle/late childhood [15]. International Civic and Citizenship Education Study (ICCS) 2016 framework, including a sample of students in Grade 8 (also used datasets for our secondary analyses), therefore, presents a very appropriate time frame for its assessment.

Numerous studies highlight the different elements that contribute to the construct of European identity [10]. At least three types of factors influencing adolescents’ European (as well as national) identity can be distinguished [8]: a) individual (psychological) factors; b) social factors formed by characteristics of adolescents’ families and schools, and c) macro factors formed by country-level characteristics. Regarding individual socio-demographic variables, the previous research shows that girls and students with immigrant backgrounds are less likely to identify with Europe. The evidence is less clear for SES, with some studies finding that SES is positively associated with European identity, while others found no significant association (for review see [8]).

Diversity of attitudes to the EU and the knowledge of the EU, leads to diversity in target visions of common life in Europe [16]. The (future) visions of Europe among youth can therefore be supported in educational practices by both, cognitive (knowledge) and affective (identity) formation.

2.2 Regional (east-west) divide

The discussions about the East-west divide in the EU became prominent with the process of the EU enlargement in 2004 when 10 new member states (Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, and Slovenia) entered the EU. These discussions are mainly related to the rule-of-law consolidation, economic performance, and functional collaboration between the states [17]. To some extent, they relate to the attitudes and values [18] and European identity as well [8]. Jugert et al. [8] reported that late accession to the EU (2004–2007) and therefore communist past is significantly positively associated with European identity. Cabada [12], however, recognized that after the successful Europeanization in these states in the first years of entering the EU, we can observe the strengthening of anti–EU and anti–European attitudes in the years that followed. These can be mainly attributed to the economic crisis in 2008 and the migration crisis in 2015. According to Cabada [12], such a multi-crisis situation strengthens the so-called Core-Periphery divide in the EU and mental gaps between so-called old and new EU member states. However, it should be noted that country-level differences inside the two “blocks” of countries should be carefully taken into consideration when studying certain phenomena. This was evident also when investigating youth future civic participation in Europe - based on ICCS 2009 data - the results suggested differentiated patterns of future civic participation between the new and established democracies, but they are not that clear, because participation seem to be related not only with the countries’ experience with democracy, but also with their cultural similarities, common history, as well as often associated with family background of students [19]. From this perspective, the article tries to reveal, whether some patterns between two “blocks” of the EU member states exist as regards the young future vision of the EU, and how this differentiates between member states.

3.1 Data, research question

The 2016 cycle of the International Civic and Citizenship Education study (ICCS 2016), internationally coordinated by the International Association for the Evaluation of Educational Achievement (IEA), and nationally by the national study centres, assessed 8th grade students (in case of Norway 9th grade students) from 24 educational systems. This article uses data only from European countries. The total number of participating students from the European region was 52,788. The basic descriptive statistics and the later secondary analyses use the ICCS 2016 European regional questionnaire data, and data on student background characteristics are from the ICCS 2016 International Student Questionnaire. Our research questions are:

R1: To what extent different background characteristics are related with students’ attitudes, opinions and expectations about sense of European identity, future of Europe (and EU), and student’s endorsement of European cooperation.

R2: How strong is the association between student positive expectations towards EU and other attitudes related with Europe and EU 2 , and is there clear divide between eastern European countries and the rest?

All of the data analyses were performed through the R Analyzer for Large-Scale Assessments (RALSA) [20], an R package for analysing data with complex sampling and assessment designs, as ICCS and other studies.

3.3 Analyses and results

We first started our secondary analyses with descriptive statistics, the second part presents analyses on the extent different background characteristics are related with their attitudes, opinions and expectations. The third part tests the association between students’ positive attitudes towards EU with different other attitudes and expectations towards EU and Europe.

The results from the descriptive analyses are described below. Table 1 to Table 5 present the averages of the following complex scales derived from ICCS 2016 European Module data:

Students’ sense of European identity (see myself as European, proud to live in Europe, feel part of Europe, see myself first as a citizen of Europe and then as a citizen of the world);

Note that the tables in the article are sorted, so that the post-communist countries (new democracies) appear on top and the rest of the countries appear at the bottom. Also, within each of these groups the tables are sorted in ascending order of the averages.

Results in the Table 1 presents the averages for the “Students’ sense of European identity” scale. The lowest average is in Latvia (48.33) and the highest is in Finland. In general, the averages tend to be a bit lower in the post-communist countries compared to the established democracies. However, in Croatia the average is higher to the one in Norway (55.13) and quite close to this in Finland. Slovenian students have slightly lower average for the sense of European identity (54.80) than the Croatian ones but still quite high. That is, the sense of European identity of Croatian and Slovenia students is higher than the rest of the post-communist countries and resembles more the averages of the established democracies.

Results in the Table 2 presents the averages of the “Students’ endorsement of European cooperation” scale. The lowest averages are for Latvia and The Netherlands (47.91 and 47.48 respectively) and highest for Croatia (54.47) again. The variation among the two groups of countries (new and established democracies) is nearly the same, with the only exception being Croatia which has the highest average of endorsement of European cooperation.

Results in the Table 3 presents the averages for the “Students’ negative expectations for European future” scale. The variation of averages within the two groups is again nearly the same. The lowest negative expectations are found in Slovenia (47.09) and Italy (47.22) and the highest in Lithuania (51.62) and Denmark (51.54). It is interesting that the students from Lithuania show relatively high averages for the previous two scales (European identity and endorsement of European cooperation), but at the same time have negative expectations. Denmark, on the other hand, has low averages for the previous two scales. Also, just like the Danish students, students from The Netherlands reflect their opinion for the previous two scales –they have low sense of European identity, low endorsement of European cooperation and highly negative expectations for the European future. Also, Croatia has the second highest average for negative expectations for the European future, although they were scoring the highest in the post-communist countries group for the previous two scales. Sweden has average position compared to other established democracies on the students’ sense of European identity and endorsement of European cooperation, but positions among the countries with the highest negative expectations on the European future. Norway is another interesting case, positioning among the countries with highest sense of European identity, among countries with lower endorsement of European cooperation and lower negative expectations for European future.

Results in the Table 4 presents the results on the positive expectations for European future scale. The expectations are lowest in Slovenia (47.75) and Germany, North-Rhine Westphalia (47.85) and highest in Lithuania (49.99) and Malta (52.66). The variation in the averages among the new and established democracies is rather different. Within the group of new democracies, the difference between the highest and the lowest averages is just above two points on the scale and all of the countries have averages below the center point of the scale of 50 points. That is, students in post-communist countries tend to be less optimistic about the European future. On the other hand, the variation of the averages in the group of older democracies is higher and all, except for Germany (North-Rhine Westphalia) and Belgium (Flemish), are slightly above the center point of the scale, i.e. their students tend to have more positive expectations about the European future. Surprisingly, Slovenia was the country with the lowest negative expectations (see Table 3), but is also the country with the lowest positive expectations. The results for Denmark and Lithuania are the opposite –these are the countries with the highest negative expectations, but also the countries with the highest or among the countries with the highest averages for the positive expectations for European future. Also, the averages for students from Norway on negative (see Table 5) and positive expectations (Table 4) are about the same as average values. For Finland both the average values and positions in the two tables are the same. Malta is the only country with low negative and high positive expectations compared to all other countries.

Results in the Table 5 presents the averages on the “Students’ positive attitudes toward European Union” scale. Note that this is different than the previous scale that reflects students’ opinion for the European future. The lowest averages are found in Latvia (47.32) and Denmark (47.28) and the highest in Lithuania (52.14) and Malta (54.28). The variation of the averages within the two groups of countries is different, being slightly higher in the group of established democracies. In general, in both groups the positive attitudes towards EU are higher than the positive expectations for European future (see Table 4). However, the positions of the countries are quite different. Latvia, Bulgaria, Estonia and Croatia (new democracies) now position lower. Lithuania remains as the country with the highest position. Slovenia remains at about the same position. There are big shifts in the group of established democracies. While Malta remains with the strongest positive attitudes towards EU, Denmark, being the second highest in positive expectations for European future is the one with the lowest positive attitudes towards EU. Belgium (Flemish), Norway, Italy, Sweden and The Netherlands who had lower positive expectations for the European future (see Table 4) have much higher positive attitudes towards EU. Finland and Sweden rank about the same for their positive attitudes towards the EU as for positive expectations for the European future.

An important part of the differences between the old and new democracies is to what extent different background characteristics are related with students’ attitudes, opinions and expectations. Two regression models were fitted with the “Students’ sense of European identity” and “Students’ endorsement of European cooperation” as dependent variables. For the students’ sense of European identity, the following independent variables were used:

Student gender;

For the student endorsement of European cooperation model, the independent variables are the same, except for the last one (student endorsement of European cooperation) which was replaced with the students’ sense of European identity. Table 6 and Table 7 present the results for these two regression models.

The first model uses the “Students’ sense of European identity” as dependent variable. Student civic knowledge shows coefficients are close to zero in all European countries participating in ICCS 2016. In Estonia, Latvia, Belgium (Flemish), Denmark and Norway these coefficients are strictly zero. In these countries (and Germany [North-Rhine Westphalia]) the p values also show non-significant coefficients (p > 0.05). In all other countries the coefficients are negative (i.e. the higher the student civic knowledge is, the lower their sense of European identity tends to be) and statistically significant (p < 0.05). The coefficients, however, are very close to zero, ranging from –0.01 to –0.02. The students’ expected further education has slightly higher coefficients, but still close to zero and none of them is statistically significant, except in Denmark (b = 0.58, p < 0.05). The student immigration status is significantly related with students’ sense of European identity (p < 0.001) in all countries except in Bulgaria and Croatia. The coefficients in countries where the relationship is stronger are stronger, ranging from 2.07 (Sweden) to 5.88 (Estonia). That is, the less migration status the students have, the higher their sense of European identity tends to be. In general, the relationship is stronger in the new democracies. Students’ family SES is significantly related (p < 0.05) in just three countries, all of them being old democracies (Germany [North-Rhine Westphalia], Malta and Norway). The coefficients are positive (i.e. the higher the students’ family SES is, the stronger their sense of European identity tends to be), although they are very close to zero. Student gender is significantly related (p < 0.05) with student sense of European identity in just one of the new democracies (Slovenia), but in five of the old democracies –Belgium (Flemish), Germany (North-Rhine Westphalia), Italy, Malta and Norway. The coefficients in all countries are negative, meaning that female students have a lower sense of European identity. Finally, the student endorsement of European cooperation is significantly related with the students’ sense of identity in all countries in this analysis. The coefficients are positive, i.e. the higher the endorsement is, the stronger the sense of European identity tends to be. The explained variance of this model ranges from 9% (Norway) to 17% (Malta). In general, the variability in the explained variance is the same in both set of countries (old and new democracies).

Table 6. Results from the linear regression model with the “Students’ sense of European identity” scale as dependent variable.

The results from the regression model with the “Students’ endorsement of European cooperation” scale as dependent variable are presented in Table 7. As with the previous model, the coefficients for civic knowledge are close to zero, although this time they are slightly stronger and all of them are positive (i.e. the higher the student civic knowledge, the more they tend to endorse European cooperation) and statistically significant in all countries. In general, the coefficients show that civic knowledge is slightly stronger predictor in the new democracies. The expected further education is a significant predictor in Estonia, Latvia, Denmark, Finland, Italy and Malta (p < 0.05). In general, the expected further education is stronger positive predictor of endorsement of European cooperation in the established democracies (four countries) than in new democracies (two countries). It worth noting, though, that in the two new democracies (Estonia and Latvia) the coefficients tend to be higher than in the old democracies. The situation is similar with the student migration status. It is a significant predictor of student endorsement of European cooperation in two of the new democracies (Bulgaria and Slovenia) and in six of the old democracies (Belgium [Flemish], Finland, Germany [North-Rhine Westphalia], The Netherlands, Norway and Sweden). The strongest coefficient is in Bulgaria (b = 2.82, p < 0.05) and this is the only country where the coefficient is positive; that is, in Bulgaria the less migrant status the students have, the more they tend to endorse European cooperation. On the contrary, in all other countries, the lower the student migrant status is, the less they tend to endorse European cooperation. The student SES is significant predictor of student endorsement of European cooperation only in Croatia and The Netherlands. The coefficients in both countries, however, are quite close to zero. In Croatia the coefficient is negative (higher SES students tend to endorse European cooperation less) while in The Netherlands the coefficients are positive. Student gender is significant predictor of student endorsement in four new democracies (Estonia, Latvia, Lithuania and Slovenia) and in five old democracies (Denmark, Italy, The Netherlands, Norway and Sweden). The proportion of new democracies where this predictor is related significantly with the outcome variable is higher than in older democracies. The coefficients in all countries where the relationship is significant for student gender and endorsement are negative, i.e. female students tend to endorse European cooperation less than male students. Finally, the students’ sense of European identity is a strong and significant (p < 0.001) predictor of student endorsement of European cooperation in all countries in this analysis. The coefficients are positive, i.e. the higher the students’ sense of European identity is, the more they tend to endorse European cooperation. The explained variance from the model is the lowest in Norway (9%) and highest in Malta (26%). On the average, the explained variance is higher in the new democracies.

In this part bellow secondary analyses for our second research question were conducted. And the results are as follows. Table 8 shows results on coefficients for the correlations between scales “Students’ positive attitudes toward European Union” and “Students’ sense of European identity", “Students’ endorsement of European cooperation", “Students’ negative expectations for European future” and “Students’ positive expectations for European future".

The “Students’ sense of European identity” and “Students’ endorsement of European cooperation” scales show moderate to relatively strong association with the “Students’ positive attitudes toward European Union” scale. The weakest association between the positive attitudes towards EU scale and the European identity scale is found in Belgium [Flemish] (0.23), Norway (0.28) and Germany [North-Rhine Westphalia] (0.29), and the strongest in Latvia (0.40), Croatia (0.41), Malta (0.41) and Lithuania (0.42). The correlation coefficients between student sense of European identity and endorsement of European cooperation range in a similar fashion. However, the ranking of the countries differs. The weakest association is found in Italy (0.28), Slovenia (0.29) and Belgium [Flemish] (0.30), and the strongest in Finland (0.39), Lithuania (0.40) and Malta (0.41). All of the correlation coefficients between students’ positive attitudes towards EU and these two scales (sense of European identity and endorsement of European cooperation) are statistically significant (p < 0.001).

It would be expected that students’ negative expectations for the European future will correlate negatively with the students’ positive attitudes towards EU. However, this holds in just three countries - Malta (–0.08), Norway (–0.05) and the Netherlands (–0.01). The coefficient in Netherlands, however, is very close to zero and is statistically insignificant, but in Malta and Norway the coefficients are slightly stronger and statistically significant (p < 0.001 and p < 0.01, respectively). In all other countries the coefficients are positive. In Bulgaria, Sweden, Belgium [Flemish], Estonia and Germany [North-Rhine Westphalia] the coefficients are positive but rather weak and insignificant. In Denmark, Lithuania, Italy, Slovenia and Finland the coefficients are statistically significant, but rather weak. Croatia and Latvia show the strongest correlation coefficients, also statistically significant, although below 0.20. This is somewhat surprising, however, because these two countries are among the ones with the strongest correlation coefficients between students’ positive attitudes towards EU and students’ sense of European identity.

Nevertheless, the correlation coefficients between students’ positive attitudes towards EU and students’ positive expectations for European future are fairly strong and statistically significant (p < 0.001) in all countries in this analysis. The weakest correlation coefficients are found in Norway (0.39), Denmark (0.40), Sweden (0.41) and Malta (0.41), and the strongest in the Baltic countries –Lithuania (0.48), Estonia (0.49) and Latvia (0.49).

In general, some countries in South and East Europe, as well as Scandinavian countries tend to group together in part of the analyses.