YOUNG EUROPEAN CITIZENS 1
This is an Accepted Manuscript of an article published by Taylor & Francis in
‘European Journal of Developmental Psychology’ on 28 August 2017, available online:
http://www.tandfonline.com/10.1080/17405629.2017.1366308.
Young European citizens: An individual by context perspective on adolescent European
citizenship
Jan Šerek a
Masaryk University
&
Philipp Jugert b
University of Leipzig
a
E-mail: serek@fss.muni.cz
b
E-mail: philipp.jugert@uni-leipzig.de
YOUNG EUROPEAN CITIZENS 2
Abstract
This study examined the effects of individual, school-level and country-level variables and
their interactions on two components of adolescents’ active European citizenship: trust in
European institutions and participation at the European level. For comparison, country-related
institutional trust and participation were also predicted. Using multilevel regression models,
we re-analysed a subsample of survey data from the International Civic and Citizenship
Education Study (ICCS), collected from 14-year-old students (n=72,466) in 22 European
countries in 2009. Results showed that higher cognitive engagement with politics (e.g.,
political interest), more opportunities for learning about Europe at school, and country wealth
and social equality were positively associated with both aspects of adolescents’ active
European citizenship. In contrast to country-related participation, the participatory dimension
of active European citizenship was also positively related to a higher socioeconomic status of
adolescent’s classroom and family, an association that was more pronounced in less wealthy
and post-communist countries.
Keywords: active European citizenship, adolescence, institutional trust, participation,
socioeconomic status
YOUNG EUROPEAN CITIZENS 3
Young European citizens: An individual by context perspective on adolescent European
citizenship
Active citizenship of youth is believed to be a crucial prerequisite for the functioning
of democracy. Thus, it is not surprising that past research paid considerable attention to the
question of what individual and social factors shape young people’s active citizenship within
their countries (e.g., Barrett & Zani, 2015; Zukin, Keeter, Andolina, Jenkins, & Delli Carpini,
2006). However, active citizenship can be practiced not only in local or national contexts but
also internationally. One example is the European Union, a project aimed at creating a
political community that would, at least partially, function beyond the scope of nation states.
Yet, can we simply assume that the same factors that predict general active citizenship at the
national level will also be predictive of active citizenship at the European level or is there
something distinct about European active citizenship?
The concept of active citizenship
The concept of active citizenship has no single agreed upon definition (cf. Hoskins &
Mascherini, 2008). In this study, we build on a heuristic (rather than normative)
understanding of active citizenship that encompasses two interrelated dimensions: while
psychological citizenship refers to one’s cognitive and affective ties to some political
community, participatory practices refer to one’s active involvement within this community
(see editorial of this special issue). Psychological citizenship encompasses a large number of
perceptions and attitudes, such as civic identity, sense of empowerment, or perceived citizens’
rights and duties. For the purpose of this study, we focus on one selected aspect of
psychological citizenship – institutional trust, which not only affects citizens’ compliance
with public policies (Hetherington, 2007) but also their tendencies toward noninstitutionalized
political behaviour (Kaase, 1999). Moreover, lack of institutional trust might
relate to adolescents’ involvement in illegal political activism (Dahl & Stattin, 2016).
YOUNG EUROPEAN CITIZENS 4
Regarding participatory practices, we take a broad understanding of participation as an
activity involving regular meeting, communicating and cooperating with people from the
wider community (cf. Schulz, Ainley, & Fraillon, 2011). Both institutional trust and
participation might be related not only to the national, but also to the European level (i.e. trust
in institutions of the EU and participation involving people from different European
countries).
Although active European citizenship is widely perceived as a mere higher-level
extension of national citizenship, there are some characteristics of European politics making
this analogy questionable. For many adolescents, European institutions and their structures
might be less familiar and comprehensible compared to country-related institutions. In the
same vein, European-level participation might pose more barriers and demands, such as the
need to gather more information, spend more resources, or establish cross-border
collaborations. Consequently, the relative importance of particular predictors of national
versus European citizenship might be different, too.
Predictors of active citizenship at the local or national level
According to ecological systems theory (Bronfenbrenner, 1979) individual
development is embedded into multiple systems that interact with each other and influence
one’s development. Some systems of influence are proximal, that is formed of people and
institutions with whom a person directly interacts (e.g., school or family), while some of them
are shared by the whole society (e.g., a country’s history or economic situation;
Bronfenbrenner, 1979; Wilkenfeld, Lauckhardt, & Torney-Purta, 2010). Based on these
considerations, at least three types of factors influencing adolescents’ active citizenship can be
distinguished: (1) individual psychological factors supporting or hindering trusting attitudes
and participatory tendencies, (2) social factors formed by the characteristics of adolescents’
YOUNG EUROPEAN CITIZENS 5
schools or families, and (3) macro factors formed by the country-level characteristics (cf.
Barrett, 2015).
In the present study, we consider the effects of two broad individual factors. One of
them is cognitive engagement with politics, understood as paying attention to and being
interested in public affairs (Zukin et al., 2006). Using both adolescent and adult samples,
previous studies have shown positive links from political interest or following political news
in the media to trust in institutions (Catterberg & Moreno, 2006; Claes, Hooghe, & Marien,
2012; Šerek & Macek, 2014) or civic participation (Pasek, Kenski, Romer, & Jamieson, 2006;
Verba, Schlozman, & Brady, 1995). On the border between social and individual factors,
positive associations have been also shown between political discussions with parents or peers
and civic participation (McIntosh, Hart, & Youniss, 2007; Zukin et al., 2006). The second
individual factor is post-materialist value orientation, emphasizing autonomy and selfexpression,
and manifesting itself, for instance, through the support for equal rights of
different social groups. People with such values are expected to be more involved in noninstitutionalized
civic participation (Theocharis, 2011) but to have rather critical and negative
views of public institutions, hence a rather negative effect on trust can be expected
(Catterberg & Moreno, 2006).
Among social factors, we focus particularly on those related to school. First, there is
a well-known link between open classroom climate, characterized by the possibility to engage
in respectful debates and to freely express opinions, and students’ higher political trust (Claes
et al., 2012) or participatory tendencies (Campbell, 2008; Manganelli, Lucidi, & Alivernini,
2015; Quintelier & Hooghe, 2013). Second, we expect that students’ trust and participation,
particularly at the European level, will be enhanced by opportunities for learning new
information about European issues (cf. Claes, et al., 2012, Torney-Purta, Barber, Richardson,
2004). From the macro level perspective, we assume that citizens’ institutional trust is eroded
YOUNG EUROPEAN CITIZENS 6
by worse living conditions in the country (cf. Torney-Purta, et al., 2004) or great social
inequalities (Wilkinson & Pickett, 2009). In a similar manner, poor economic conditions and
social inequalities might pose a barrier to adolescents’ participation (Solt, 2008). Another
country-level factor in the European context is a communist past of the country. Even though
current adolescents do not have direct experiences with authoritarian communist regimes,
certain tendencies such as distrust in public institutions might still be present in and passed on
through political cultures of the post-communist countries (e.g., through political socialization
by parents and teachers).
Finally, an important source of differences in adolescents’ active citizenship is their
status in the society, defined by their gender, socioeconomic situation, or migration
background (cf. Verba et al., 1995). For instance, willingness to participate is lower among
young people coming from low status families (Gaby, 2016) or first-generation immigrant
youth (Lopez & Marcelo, 2008). At the same time, social and macro factors might moderate
the consequences of social status on active citizenship. As noted above, adolescents’
development is embedded in multiple contexts (cf. Bronfenbrenner, 1979) that interact with
person’s individual characteristics such as social status. Hence, due to person-by-context
interactions, the impact of adolescents’ individual status might have different forms and
extents across contexts. For instance, although young people from immigrant or ethnic
minority groups have generally lower rates of civic participation than native and majority
youth, data from different European countries show that this pattern might be absent or even
reversed for some groups in specific countries (Barrett & Zani, 2015). Similarly, certain
contextual characteristics of the classroom (e.g. open climate) are expected to mitigate the
gaps in active citizenship stemming from students’ socioeconomic background, although
these moderation effects have not been convincingly demonstrated yet (Castillo, Miranda,
Bonhomme, Cox, & Bascopé, 2015).
YOUNG EUROPEAN CITIZENS 7
The present study
Due to the lacking knowledge on the predictors of adolescents’ active citizenship at
European level, this study takes an exploratory approach employing the predictors that were
previously described for the national level. The main goal is to assess individual,
school/classroom and country-level predictors of European-level institutional trust and
participation and to differentiate European from general active citizenship. At the same time,
we aim to test whether potential effects of person’s social status are moderated by higherlevel
variables (i.e. person-by-context interaction). Overall, we assume that active European
citizenship is predicted by similar variables as national active citizenship although the effects
in the former case might be weaker due to supposed adolescents’ less frequent and less direct
experience with the European versus the national level. However, it may be that at least some
factors are more predictive of national than of European level citizenship and vice versa.
Thus, it may be that migration background is negatively associated with national and
positively with European citizenship, as previous studies have shown that ethnic minority
members are more likely to identify with the EU, but are less likely to identify with their
nation (Hadler, Tsutsui, & Chin, 2012). This is presumably because the EU pressures member
states to give more rights to ethnic minorities. Socioeconomic status may also relate more
strongly to European than to national-level citizenship because social and economic elites
gain the most from new opportunities arising from European integration (e.g., traveling freely,
studying abroad; Hadler et al., 2012). Post-materialist tendencies such as granting equal rights
to different groups (e.g., to immigrants) should also be more strongly related to European
level citizenship as the EU advocates and enforces policies meant to guarantee equal rights to
different groups much more strongly than its member states. Lastly, opportunities for learning
about Europe should most strongly be associated with European rather than with national
citizenship.
YOUNG EUROPEAN CITIZENS 8
To distinguish whether the effects established by our study are specific for European
active citizenship, or whether they concern active citizenship in general, we will include both
European and country-related trust and participation as outcome variables. In addition, we
will control for county-level trust when predicting European level trust and for countryrelated
participation when predicting European level participation.
Method
Participants and procedure
We reanalysed survey data from the International Civic and Citizenship Education
Study (ICCS) conducted in 2009 in 38 countries across the world. Questionnaires were
administered in schools to eighth-graders (50% females) who were on average 14 years old.
Only one classroom per school was sampled (Schulz et al., 2011). For our analyses, we used a
subset of 22 countries that were members of the European Union and where a module on
issues relevant to Europe was administered. Data had a three-level structure: a total number of
72,466 students (level 1) were nested in 3,632 classrooms (level 2; mean classroom size was
19.62) that were nested in 22 countries (level 3). Due to occasional missing data, the final
numbers of analysed participants differed slightly from analysis to analysis1
.
Measures
If not stated otherwise, we used scales developed and provided by the authors of the
ICCS. Using item response modelling, they computed the total scores of these scales from the
items based on the weighted likelihood estimates (logits) of the latent dimensions. The scales
were transformed to have an international average of 50 and a standard deviation of 10 (these
values were slightly different in our study because we used a subsample of 22 countries). For
more details see Schulz et al. (2011).
1
Data used in this study are publicly available from the IEA Study Data Repository (http://rms.iea-dpc.org).
Description of the steps through which we created the working dataset, used in our paper, can be found in the
Masaryk University Repository (https://is.muni.cz/repo/1387306/en).
YOUNG EUROPEAN CITIZENS 9
Trust in country-related institutions. Participants indicated how much they trusted
in six institutions (a four-point response scale from “not at all” to “completely”): national
government, national parliament, local government, courts, police, and political parties
(country-level Cronbach alphas ranged from .76 to .89; M=49.57, SD=9.57).
Trust in European institutions. Using the same response scale, respondents
indicated how much they trusted in European Commission and European Parliament. We
computed the total score by ourselves by averaging the items (inter-item correlation was >.70
in all countries). For the sake of consistency with other outcome variables, mean was
transformed to 50 and SD to 10.
Participation in the wider community. Participants were asked whether they had
participated (“never”, “more than a year ago”, “within the last 12 months”) in seven different
organizations, clubs or groups in the wider community (e.g., environmental organization or a
voluntary group doing something to help the community; alphas from .60 to .80; M=48.18;
SD=9.25).
Participation at the European level. With the same response scale, participants
indicated their participation in eight activities involving another European country (e.g.,
activities in local area that involve meeting people from other European countries or school
trips to another European country; alphas from .63 to .83; M=49.78; SD=9.79).
Migration background. A categorization of participants based on the country of
birth was provided by the ICCS. We transformed this categorization into two categories: a
person and at least one parent born in the country (=0; 91%), a person and/or both parents not
born in the country (=1; 9%).
Socioeconomic background. The ICCS provided the index of socioeconomic
background derived from parental highest occupational level, parental highest educational
YOUNG EUROPEAN CITIZENS 10
level, and the approximate number of books at home (the score was standardized to have
national means of 0 and SDs of 1).
Political interest. Using five items, participants assessed their interest (a four-point
scale from “not interested at all” to “very interested”) in local, national, foreign and
international political issues and national social issues (alphas from .82 to .92; M=48.58,
SD=10.17).
Discussing political issues. Using four items, respondents assessed the frequency of
their participation (a four-point scale from “never or hardly ever” to “daily or almost daily”)
in discussion with parents and friends about political or social issues and international events
(alphas from .63 to .81; M=48.95, SD=10.07).
Watching news on TV. One item measured the frequency of watching television to
inform yourself about European news (a four-point response scale was from “never or hardly
ever” to “weekly”; M=2.80, SD=1.10).
Post-materialism. Post-materialist value orientation was indicated by two scales
(four-point response scales from “strongly disagree” to “strongly agree”). Support for equal
rights for immigrants was measured by five items (e.g., immigrants should have the
opportunity to continue speaking their own language; alphas from .74 to .89; M=.49.00,
SD=10.08), and support for gender equality was measured by six items (e.g., men and women
should get equal pay when they are doing the same jobs; alphas from .73 to .87; M=51.47,
SD=10.16).
Open classroom climate. Students’ perceptions of openness in classroom
discussions (a four-point response scale from “rarely to “often”) was measured by six items
(e.g., teachers encourage students to discuss the issues with people having different opinions;
alphas from .66 to .81; M=50.55, SD=9.68).
YOUNG EUROPEAN CITIZENS 11
Opportunities for learning about Europe at school. Using nine items, students
indicated (a four-point response scale from “strongly disagree” to “strongly agree”) how
much opportunity their school gives them for activities such as finding out what is happening
in other European countries or meeting young people from other European countries (alphas
from .78 to .85; M=50.11, SD=9.78).
Country-level variables. Country’s wealth was represented by its gross domestic
product per capita (GDP) standardized according to the European Union average (Eurostat,
2009a; M=101.27, SD=41.56). Income inequality was indicated by the GINI index (Eurostat,
2009b; M=0.29, SD=0.04) and gender inequality by the Gender Inequality Index (GII; UNDP,
2016; M=0.14, SD=0.06). Communist (=1; eight countries) or non-communist past (=0; 14
countries) was indicated by a dichotomous variable.
Data analysis
Data was analysed using multilevel linear regression models in Mplus 7.4 (Muthén &
Muthén, 1998-2015). Sampling weights at levels 1 and 2 recommended by Brese, Jung,
Mirazchiyski, Schulz, & Zuehlke (2011) were used.
For each outcome, several three-level models were estimated using the maximum
likelihood estimator with robust standard errors (MLR). First, an intercept-only model without
predictors was estimated to decompose individual-level, classroom-level and country-level
variance (M0). Second, we added three individual sociodemographic status predictors
(gender, immigration and socioeconomic background; M1), and, after that, we added the
remaining seven individual predictors (political interest, discussing political issues, watching
news on TV, support for equal rights for immigrants, support for gender equality, subjective
perceptions of open classroom climate, and subjective perceptions of opportunities for
learning about Europe at school; M2). All non-dichotomous individual-level predictors were
grand mean centred. Third, three classroom-level predictors were added: classroom average
YOUNG EUROPEAN CITIZENS 12
socioeconomic background, classroom average perceptions of open climate, and classroom
average perceptions of opportunities for learning about Europe (M3). Fourth, four countrylevel
predictors were considered. Because of the small sample size at level 3 (n=22), countrylevel
predictors were tested individually in separate models (M4a-d). Finally, country-level
trust or participation were added as control variables into the models predicting Europeanlevel
trust and participation (M5a-d).
For statistically significant level 1 and level 2 sociodemographic predictors with nonnegligible
effect sizes (β>.10), cross-level interactions were tested. First, we estimated a
random-slope model for the predictor to test whether its effect significantly varied at higher
levels (this model included all level 1 and level 2 predictors and the most important level 3
predictor). If the variance was significant, we tested whether it could be explained by higherlevel
predictors (i.e. cross-level interactions). For the sake of easier interpretation of
significant interactions, simple slopes were computed for higher-level units with lower (25th
percentile), median (50th
percentile), and higher (75th
percentile) levels of moderator variable
(Aiken & West, 1991). For specific values of the moderator, see Table A3 in Online
Appendix.
Results
Preliminary analyses
Intercept-only models showed that the variance in outcome variables was attributable
mostly to differences between individuals (see Notes in Tables 1-4). Differences between
classrooms were the most pronounced for European-level participation (10%), while
differences between countries were the strongest for trust in country-related institutions (7%).
There was a strong correlation between trust in country-related and European institutions
(r=.60) and a moderate correlation between community-level and European-level
YOUNG EUROPEAN CITIZENS 13
participation (r=.26; see Online Appendix A for the full correlation table and country-level
descriptives).
Predictors of trust
Trust in country-related institutions. On the individual level, trust was not
associated with gender, migration or socioeconomic background (M1 in Table 1). After other
individual predictors were added, significant effects of seven predictors appeared (M2).
Standardized coefficients revealed that most of these effects were negligible except for the
effects of political interest (β=.26) and opportunities for learning about Europe (β=.16). On
the classroom level (M3), students in classrooms with a higher average socioeconomic
background expressed a lower institutional trust (β=-.19). Finally, on the country level (M4ad),
institutional trust was higher in countries with a higher GDP (β=.56), smaller economic
(β=-.55) and gender inequalities (β=-.45), and without communist past (β=-.55).
[Table 1]
A random-slope model showed that the effect of classroom socioeconomic
background significantly varied across countries (σ2
=2.44, p<.01) and interacted with GDP
(unstandardized B=0.02, p<.05), income inequality (B=-15.89, p<.05) and gender inequality
(B=-17.33, p<.01), and communist past (B=-2.23, p<.01). Specifically, the negative effect was
present only in countries with a lower GDP (B25=-1.31, p<.01; B50=-0.56, p<.05; B75=-0.12,
p=.74), greater income inequality (B25=-0.08, p=.83; B50=-0.51, p=.08; B75=-1.07, p<.01),
gender inequality (B25=0.27, p=.18; B50=-0.54, p<.05; B75=-1.39, p<.01), and with communist
past (Bnon-postcommunist=0.23, p=.32; Bpostcommunist=-2.00, p<.01).
Trust in European institutions. On the individual level, trust was positively but
only marginally predicted by male gender and socioeconomic background (M1 in Table 2).
After other individual predictors were added, all of them had significant effects (M2).
However, most of them were negligible except for the positive effects of political interest
YOUNG EUROPEAN CITIZENS 14
(β=.26) and opportunities for learning about Europe (β=.12). On the classroom level (M3),
there was a significant but rather negligible (β=-.09) effect of opportunities for learning about
Europe at school. Finally, trust was higher in countries with a higher GDP (β=.29) and smaller
income (β=.36) or gender inequalities (β=-.46; M4a-d). Random slopes of sociodemographic
predictors were not tested because their main effects were negligible.
[Table 2]
When trust in country-related institutions was controlled for (M5a-d), it had a strong
effect on trust in European institutions (β=.57). All other previous effects became negligible
except for the effect of political interest that became considerably smaller (β=.12). At the
same time, two new effects appeared: on the country level, trust in European institutions was
higher in post-communist countries (β=.48), and on the classroom level, trust in European
institutions was higher for students from classrooms with a higher average socioeconomic
background (β=.29). An additional random-slope model showed that the effect of average
socioeconomic background significantly varied across countries (σ2
=0.57, p<.01) and
interacted with GDP (B=-0.01, p<.01). Specifically, the effect was stronger in countries with a
lower GDP (B25=0.84, p<.01; B50=0.58, p<.01; B75=0.42, p<.01).
Differences between effect sizes. The effects of migration background (Δβ=.03),
individual socioeconomic background (Δβ=.04), support for immigrants’ rights (Δβ=.00),
support for gender equality (Δβ=.05) or opportunities for learning about Europe (individual
Δβ=.05, classroom Δβ=.04) in M4 were similar when predicting trust in country-related
versus European institutions. However, the effects were different for classroom
socioeconomic background that predicted negatively country-related but not European-level
trust (Δβ=.25).
Predictors of participation
YOUNG EUROPEAN CITIZENS 15
Participation in the wider community. On the individual level, there were positive
effects of female gender and socioeconomic background, but both these effects were
negligible (M1 in Table 3). After the addition of other individual variables, seven predictors
were significant (M2). According to standardized coefficients, most of these effects were
negligible except for the effects of political discussions (β=.18) and political interest (β=.12).
On the classroom level (M3), students from classrooms with a higher average socioeconomic
background participated less in the wider community (β=-.14). On the country level (M4a-d),
participation was higher in countries with greater gender inequalities (β=.46).
[Table 3]
According to a random-slope model, the effect of classroom socioeconomic
background significantly varied across countries (σ2
=0.85, p<.05) and interacted with GDP
(B=0.01, p<.01) and communist past (B=-1.07, p<.01). The negative effect was stronger in
countries with a lower GDP (B25=-1.04, p<.01; B50=-0.69, p<.01; B75=-0.48, p<.01) and with
communist past (Bnon-postcommunist=-0.30, p=.16; Bpostcommunist=-1.37, p<.01).
Participation at the European level. On the individual level, participation was
positively predicted by socioeconomic background (β=.19) and marginally positively by
immigration background (β=.06; M1 in Table 4). After other individual predictors were
added, we found positive effects of seven predictors (M2). Most of these effects were
negligible, except for the effects of opportunities for learning about Europe (β=.22) and
political discussions (β=.11), and still a considerable effect of socioeconomic background
(β=.17). On the classroom level (M3), we found positive effects of average socioeconomic
background (β=.32) and opportunities for learning about Europe (β=.19) and a negative effect
of open classroom climate (β=-.15). Finally, participation was higher in countries with a
higher GDP (β=.65) and smaller income inequalities (β=-.53; M4a-d).
[Table 4]
YOUNG EUROPEAN CITIZENS 16
A random-slope model for individual socioeconomic background suggested that the
effect significantly varied between countries (σ2
=0.13, p<.01) but not between classrooms
(σ2
=0.01, p=.91). Significant cross-level interactions further suggested that the effect of
socioeconomic background interacted with GDP (B=-0.004, p<.05) and communist past
(B=0.36, p<.05). Specifically, the positive effect of individual socioeconomic background was
stronger in countries with a lower GDP (B25 B=1.54, p<.01; B50=1.40, p<.01; B75=1.31,
p<0.01) and with communist past (Bnon-postcommunist=1.27, p<.01; Bpostcommunist=1.63, p<.01). A
model with a random slope for classroom socioeconomic background showed significant
variance across countries (σ2
=1.08, p<.01), but no interactions with country-level variables
were found.
Finally, when participation in the wider community was added as a predictor (M5ad),
it had a moderate effect on European participation (β=.24), while all other effects
remained similar to the previous model (the most notable changes were a slight increase of the
country-level effect of gender inequality that became significant at the .05 level and a slight
decrease of the individual-level effect of political discussions that became negligible).
Differences between effect sizes. The effects of socioeconomic background
(individual-level Δβ=.12, classroom-level Δβ=.46) and opportunities for learning about
Europe (individual-level Δβ=.13, classroom-level Δβ=.20) in M4 were stronger for Europeanlevel
participation than participation in the wider community. The effects of migration
background (Δβ=.06), support for immigrants’ rights (Δβ=.01), and support for gender
equality (Δβ=.01) were similar for both types of participation.
Discussion
The aim of this study was to find out whether active European citizenship can be
separated from general active citizenship by examining differences in predictors of
institutional trust and civic participation at the national versus the European level. Our
YOUNG EUROPEAN CITIZENS 17
findings point out that it is possible to distinguish between general active citizenship and
European citizenship on the participatory dimension but not on the psychological dimension
of trust. More specifically, results indicate that the participatory dimension of EU citizenship
depends to a large degree on socio-economic resources at the individual, school, and country
level. In addition, findings point to the important role of adolescents’ cognitive political
engagement, and school opportunities for learning about Europe as predictors of active EU
citizenship.
Distinguishing between national and European active citizenship
With regard to institutional trust, a comparison of the models predicting national
versus European level trust revealed, with two exceptions, few differences in predictors. By
far the biggest influence on trust in European institutions was trust in national institutions,
which accounted for a large share of the variance. This suggests that once we can explain
institutional trust, we can also explain trust in European institutions. Such a finding is insofar
surprising because one could assume that in countries with a negative discourse towards the
European Union (e.g., Greece) trust towards authorities of the nation state (e.g., government,
courts, police) need not necessarily be related to trust in institutions at the European level. Of
course, we need to take into account that 14-year-old adolescents may not be politically savvy
enough yet to make clear distinctions between institutions at the national and the European
level. Thus, future research should examine whether the distinction between national and
European institutional trust is equally small among adults.
One exception to this pattern was that students in classrooms with higher than
average levels of socio-economic background had lower trust in national institutions but,
controlling for country-related institutional trust, higher trust in EU institutions. Notably, this
tendency was present only in less wealthy, more unequal and post-communist countries. A
possible explanation is that national institutions are generally less trustworthy in these
YOUNG EUROPEAN CITIZENS 18
countries and students from more educated backgrounds and privileged schools (e.g., higher
track schools) are more aware of that fact. Consistent with this explanation, the distrust of
these students was primarily oriented towards national but not European institutions, which
were perceived relatively positively. The second exception was a finding that students in
countries with a communist past expressed lower trust in national institutions but, controlling
for their country-related trust, they expressed higher trust in European institutions. Again, it
seems that adolescents living in post-communist countries trust in institutions less than their
peers living in countries without a communist past but their distrust pertains to national rather
than European institutions. A relatively more positive image of the European Union among
young people from new (i.e. mostly post-communist) versus old member states has been
already found by other studies (Eurobarometer, 2007) and it is probably associated with
positive perceptions of the European Union enlargement in 2004 (Nancy, 2016).
When comparing the models predicting participation at the national versus European
level, results revealed more notable differences in predictors than in the models predicting
trust. Thus, both socio-economic background and opportunities to learn about Europe were
more strongly related to participation at the European level than at the national level, both at
the individual and the classroom level. Importantly, controlling for participation at the
national level did not lead to substantial changes in the relationships to other variables. This
indicates that participation at the European level may indeed be substantially different from
participation at the local or national level, and that European level participation cannot be
sufficiently explained by knowing what drives local or national participation. Thus, as the
relatively weak association between participation at the wider community and participation at
the European level suggests, different from institutional trust, both levels need not go hand in
hand. This maybe because actually participating at the European level poses greater
challenges to 14-year-olds than trusting institutions at the European level does. It likely
YOUNG EUROPEAN CITIZENS 19
requires prior knowledge about how one can participate in activities at the European level and
where to find them. This may also explain why socio-economic resources at the individual,
school- and country-level are so strongly related to participation at the European level.
The contributions of predictors at different levels of analysis
For all dimensions and levels of citizenship, at least five sixths of their variance was
attributable to the individual level. It means that the differences between students in both
national and European active citizenship were primarily given by individual-level factors.
School- and country-level factors were not negligible, but their roles were less pronounced.
Results on individual psychological predictors showed that adolescents’ active
European citizenship was primarily associated with their greater cognitive engagement with
politics. Institutional trust was positively related particularly to political interest, suggesting
that adolescents high in political interest might have more information on the positive aspects
of European institutions and thus a more positive view of them than those low in political
interest. At the same time, European-level participation was associated with political
discussions with parents and peers that are likely to stimulate adolescents’ further engagement
in the public realm (McIntosh et al., 2007; Zukin et al., 2006). On the other hand, postmaterialist
values did not considerably contribute to any component of active European
citizenship. It is possible that the effects of post-materialism on trust and participation are
rather complex and cannot be described by simple linear trends (e.g., European institutions
might be positively perceived as advancing some post-materialist values such as gender
equality, but distrusted due to their alleged bureaucracy). The overall patterns of
psychological predictors were similar to country-related trust and participation, suggesting
that similar psychological processes operate in the expression of active citizenship at both the
national and European level.
YOUNG EUROPEAN CITIZENS 20
In the school context, active European citizenship was positively related to
opportunities for learning about Europe. This effect was found for adolescents’ subjective
perceptions of these opportunities and, in the case of participation, also for average
perceptions within classrooms. These results suggest that school civic education targeting
European issues might contribute to the development of trust and participatory tendencies at
the European level. On the other hand, no support was found for the positive effect of open
classroom discussions, which were only negligibly related to trust and had even a negative
association with participation. Although some previous studies have found links between
open classroom and young people’s active citizenship, these studies often found relationships
of open classroom with cognitive variables (e.g., civic knowledge) or participatory intentions
(Torney-Purta, 2009; Torney-Purta, Lehman, Oswald, & Schulz, 2011) rather than with
participation itself (Jugert, Eckstein, & Noack, 2016). Hence, it is possible that an open
classroom climate, particularly at the relatively low age of 14, affects cognitive
predispositions of active citizenship, but that the direct effect on its participatory dimension is
less visible because young adolescents have limited opportunities to act upon their knowledge
or intentions.
Next, lower levels of both components of active European citizenship were found in
less wealthy countries and in countries with greater income or (in the case of trust) gender
inequality. This is in agreement with the expectation that the contexts characterized by
economic problems and social inequalities undermine young people’s institutional trust and
pose a barrier to their participation in cross-border activities (cf. Solt, 2008; Wilkinson &
Pickett, 2009). Consistent with this explanation, the same country-level factors were found to
undermine country-related institutional trust, and, as suggested above, the negative
perceptions of European institutions in less wealthy and more unequal countries were an
expression of a general distrust in institutions rather than a distrust in European institutions as
YOUNG EUROPEAN CITIZENS 21
such. On the other hand, country-related participation in the wider community was only
poorly explained by the examined country-level predictors, which suggests that a different set
of factors plays a role here. Based on the indication that country-related participation was
slightly higher in more unequal countries, one of these factors may be collective grievances
(Walker & Smith, 2002).
Regarding individual social status, no substantial differences in trust in European
institutions were found between males and females, people with and without immigration
background, or adolescents from different socioeconomic conditions. Also, no gender or
immigration-based differences were present for European-level participation.
However, we found that adolescents coming from families with higher
socioeconomic status, compared to those from low status families, tended to participate more
in European activities. This effect interacted with country-level context and was specifically
pronounced in less wealthy and post-communist countries. Moreover, the effect was present
also at the school level, suggesting that European-level participation was limited in schools
with higher proportions of socioeconomically disadvantaged students. Hence, poor
socioeconomic conditions in multiple contexts seem to add up to limit adolescents’
opportunities for meeting and cooperating with people from different European countries.
Consequently, participation in European-level activities is probably, to a large extent,
facilitated by privileged socioeconomic conditions, in which an adolescent lives and studies.
On the contrary, the negative effects of individual and contextual socioeconomic
conditions were not present for national active citizenship. Moreover, in less wealthy, postcommunist,
and (in the case of trust) less equal countries, the effect of classroom average
socioeconomic background on country-related trust and participation was even negative.
Together with the abovementioned finding that country-related participation was slightly
higher in less equal countries, the tendency of students from classrooms with lower average
YOUNG EUROPEAN CITIZENS 22
socioeconomic background to participate more in their communities might indicate that
country-related participation in the wider community often addresses the grievances and
everyday problems in one’s surroundings. Such motivations are probably not common for the
participation at the European level that has rather positive associations with socioeconomic
conditions.
Limitations
Three limitations of this study should be mentioned. First, our analysis intentionally
focuses only on two components of active citizenship. The inclusion of other components,
such as national/European identity or participatory intentions, could bring further details on
studied predictors. Second, the data is cross-sectional, hence causal interpretations should be
made only with caution. Finally, following Bronfenbrenner’s (1979) model, some
developmental contexts are considered only marginally in the data, specifically adolescents’
parents and their attitudes.
Conclusion
Overall, our findings suggest that active European citizenship is different from
general active citizenship on the participatory dimension but not on the psychological
dimension of trust. Moreover, adolescents’ active European citizenship can be boosted if
school curricula involve ample opportunities for students’ learning about Europe. This also
implies that schools, particularly those with large proportions of students from disadvantaged
socioeconomic backgrounds, need to be sufficiently resourced to offer opportunities for
student exchange with other European countries. Finally, it seems that the gaps in
adolescents’ active European citizenship can be reduced by decreasing income and other (e.g.,
educational) inequalities at the country level.
YOUNG EUROPEAN CITIZENS 23
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Thousand Oaks: SAGE.
Barrett, M. (2015). An integrative model of political and civic participation: Linking the
macro, social, and psychological levels of explanation. In M. Barrett & B. Zani (Eds.),
Political and civic engagement: Multidisciplinary perspectives (pp. 162-187). Hove:
Routledge.
Barrett, M., & Zani, B. (Eds.). (2015). Political and civic engagement: Multidisciplinary
perspectives. Hove: Routledge.
Brese, F., Jung, M., Mirazchiyski, P., Schulz, W., & Zuehlke, O. (2011). ICCS 2009 User
Guide for the International Database. Amsterdam: IEA.
Bronfenbrenner, U. (1979). The ecology of human development. Experiments by nature and
design. Cambridge: Harvard University Press.
Campbell, D. E. (2008). Voice in the classroom: How an open classroom climate fosters
political engagement among adolescents. Political Behavior, 30, 437–454.
doi:10.1007/s11109-008-9063-z
Castillo, J. C., Miranda, D., Bonhomme, M., Cox, C., & Bascopé, M. (2015). Mitigating the
political participation gap from the school: The roles of civic knowledge and classroom
climate. Journal of Youth Studies, 18, 16-35. doi:10.1080/13676261.2014.933199
Catterberg, G., & Moreno, A. (2006). The individual bases of political trust: Trends in new
and established democracies. International Journal of Public Opinion Research, 18, 31-48.
doi:10.1093/ijpor/edh081
Claes, E., Hooghe, M., & Marien, S. (2012). A two-year panel study among Belgian late
adolescents on the impact of school environment characteristics on political trust.
International Journal of Public Opinion Research, 24, 208-224. doi:10.1093/ijpor/edr031
YOUNG EUROPEAN CITIZENS 24
Dahl, V., & Stattin, H. (2016). Beyond the limits: Involvement in illegal political activities.
European Political Science Review, 8, 125-145. doi:10.1017/S1755773914000435
Eurobarometer (2007). Young Europeans: A survey among young people aged between 15-30
in the European Union. Retrieved from:
http://ec.europa.eu/commfrontoffice/publicopinion/flash/fl_202_en.pdf
Eurostat (2009a). GDP per capita in PPS. Luxembourg: Eurostat. Available from:
http://ec.europa.eu/eurostat/tgm/table.do?tab=table&init=1&language=en&pcode=tec0011
4&plugin=1
Eurostat (2009b). Gini coefficient of equivalised disposable income – EU-SILC survey.
Luxembourg: Eurostat. Available from:
http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=ilc_di12&lang=en
Gaby, S. (2016). The civic engagement gap(s): Youth participation and inequality from 1976
to 2009. Youth & Society. Advance online publication. doi:10.1177/0044118X16678155
Hadler, M., Tsutsui, K., & Chin, L. G. (2012). Conflicting and reinforcing identities in
expanding Europe: Individual- and country-level factors shaping national and European
identities, 1995-2003. Sociological Forum, 27, 392-418. doi: 10.1111/j.1573-
7861.2012.01323.x
Hetherington, M. J. (2007). Why trust matters: Declining political trust and the demise of
American liberalism. Princeton: Princeton University Press.
Hoskins, B. L., & Mascherini, M. (2009). Measuring active citizenship through the
development of a composite indicator. Social Indicators Research, 90, 459-488.
doi:10.1007/s11205-008-9271-2
Jugert, P., Eckstein, K., & Noack, P. (2016). Differential effects of school experiences on
active citizenship among German and Turkish‐origin students. International Journal of
Psychology. Advance online publication. doi:10.1002/ijop.12409
YOUNG EUROPEAN CITIZENS 25
Kaase, M. (1999). Interpersonal trust, political trust and non-institutionalised political
participation in Western Europe. West European Politics, 22, 1-21.
doi:10.1080/01402389908425313
Lopez, M. H., & Marcelo, K. B. (2008). The civic engagement of immigrant youth: New
evidence from the 2006 Civic and Political Health of the Nation Survey. Applied
Development Science, 12, 66-73. doi:10.1080/10888690801997051
Manganelli, S., Lucidi, F., & Alivernini, F. (2015). Italian adolescents' civic engagement and
open classroom climate: The mediating role of self-efficacy. Journal of Applied
Developmental Psychology, 41, 8-18. doi:10.1016/j.appdev.2015.07.001
McIntosh, H., Hart, D, & Youniss, J. (2007). The influence of family political discussion on
youth civic development: Which parent qualities matter? Political Science and Politics, 40,
495-499. doi:10.1017/S1049096507070758
Nancy, J. (2016). Major changes in European public opinion regarding the European Union.
European Parliament Research Service. Retrieved from:
http://www.europarl.europa.eu/pdf/eurobarometre/2016/major/desk_research_historique_2
016_en.pdf
Pasek, J., Kenski, K., Romer, D., & Jamieson, K. H. (2006). America’s youth and community
engagement: How use of mass media is related to civic activity and political awareness in
14-to 22-year-olds. Communication Research, 33, 115-135.
doi:10.1177/0093650206287073
Quintelier, E., & Hooghe, M. (2013). The relationship between political participation
intentions of adolescents and a participatory democratic climate at school in 35 countries.
Oxford Review of Education, 39, 567–589. doi:10.1080/03054985.2013.830097
Schulz, W., Ainley, J., & Fraillon, J. (2011). ICCS 2009 Technical Report. Amsterdam: IEA.
YOUNG EUROPEAN CITIZENS 26
Šerek, J., & Macek, P. (2014). Antecedents of political trust in adolescence: Cognitive
abilities and perceptions of parents. Journal of Applied Developmental Psychology, 35,
284–293. doi:10.1016/j.appdev.2014.05.001
Solt, F. (2008). Economic inequality and democratic political engagement. American Journal
of Political Science, 52, 48-60. doi:10.1111/j.1540-5907.2007.00298.x
Theocharis, Y. (2011). Young people, political participation and online postmaterialism in
Greece. New Media & Society, 13, 203-223. doi:10.1177/1461444810370733
Torney-Purta, J. (2009). International psychological research that matters for policy and
practice. American Psychologist, 64, 825-837. doi:10.1037/0003-066X.64.8.825
Torney-Purta, J., Barber, C. H., & Richardson, W. K. (2004). Trust in government-related
institutions and political engagement among adolescents in six countries. Acta Politica, 39,
380-406.
Torney-Purta, J., Lehmann, R., Oswald, H., & Schulz, W. (2001). Citizenship and education
in twenty-eight countries. Civic knowledge and engagement at the age fourteen.
Amsterdam: IEA.
UNDP (2016). Human development report 2016: Human development for everyone. New
York: United Nations Development Programme. Available from:
http://hdr.undp.org/sites/default/files/2016_human_development_report.pdf
Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality. Cambridge: Harvard
University Press.
Walker, I., & Smith, H. J. (2002). Relative deprivation: Specification, development, and
integration. Cambridge: Cambridge University Press.
Wilkenfeld, B., Lauckhardt, J., & Torney-Purta, J. (2010). The relation between
developmental theory and measures of civic engagement in research on adolescents. In L.
YOUNG EUROPEAN CITIZENS 27
R. Sherrod, J. Torney-Purta, & C. A. Flanagan (Eds.), Handbook of research on civic
engagement in youth (pp. 193-219). Hoboken: Wiley.
Wilkinson, R., & Pickett, K. (2010). The spirit level: Why equality is better for everyone.
London: Penguin Books.
Zukin, C., Keeter, S., Andolina, M., Jenkins, K., & Delli Carpini, M. X. (2006). A new
engagement? Political participation, civic life, and the changing American citizen. Oxford:
Oxford University Press.
YOUNG EUROPEAN CITIZENS 28
Acknowledgements
This study was produced as a part of the CATCH-EyoU project that has received funding
from the European Union’s Horizon 2020 research and innovation programme under G.A. n.
649538. The study reflects only the authors’ view. The Research Executive Agency of the
European Commission is not responsible for any use that may be made of the information it
contains.
YOUNG EUROPEAN CITIZENS 29
Tables
Table 1. Multilevel linear regression model predicting trust in country-related institutions (n=67,035).
M1 M2 M3 M4a-d
B (SE) β B (SE) β B (SE) β B (SE) β
Individual level
Gender (Female) 0.26 (0.20) 0.01 0.02 (0.23) 0.00 0.01 (0.23) 0.00 0.01 (0.23) 0.00
Migration background -0.64 (0.36) -0.02 -1.82 (0.41)* -0.06 -1.86 (0.42)* -0.06 -1.86 (0.42)* -0.06
Socioeconomic background 0.08 (0.15) 0.01 -0.21 (0.11) -0.02 -0.11 (0.09) -0.01 -0.11 (0.09) -0.01
Political interest 0.23 (0.01)* 0.26 0.23 (0.01)* 0.26 0.23 (0.01)* 0.26
Discussing political issues -0.05 (0.01)* -0.05 -0.05 (0.01)* -0.05 -0.05 (0.01)* -0.05
Watching news on TV 0.25 (0.07)* 0.03 0.24 (0.07)* 0.03 0.25 (0.07)* 0.03
Equal rights for immigrants 0.08 (0.01)* 0.08 0.08 (0.01)* 0.08 0.08 (0.01)* 0.08
Gender equality -0.01 (0.02) -0.01 -0.01 (0.02) -0.01 -0.01 (0.02) -0.01
Open classroom climate 0.06 (0.01)* 0.07 0.06 (0.01)* 0.06 0.06 (0.01)* 0.06
Opportunities for learning about Europe 0.14 (0.01)* 0.16 0.14 (0.01)* 0.16 0.14 (0.01)* 0.16
Classroom level
Classroom socioeconomic background -0.67 (0.34)† -0.19 -0.67 (0.34)† -0.19
Open classroom climate 0.02 (0.02) 0.05 0.02 (0.02) 0.05
Opportunities for learning about Europe 0.02 (0.02) 0.05 0.02 (0.02) 0.05
Country level
Gross domestic product per capita 0.04 (0.01)* 0.56
Income inequality -39.52 (11.78)* -0.55
Gender inequality -22.05 (8.58)* -0.45
Communist past -3.14 (0.87)* -0.55
Variance explained
Individual level .00 .13 .13 .13
School level .04 .04
Country level (M4a) .31
Country level (M4b) .30
Country level (M4c) .21
Country level (M4d) .31
Note. Variance components in the intercept-only model (M0): individual level – 81.53 (88%), classroom level – 4.95 (5%), school level – 6.40
(7%). Unstandardized regression coefficients, standard errors, and standardized coefficients are presented. Country level predictors were tested
one at a time in separate models (M4a-d). Since regression coefficients and standard errors of individual and classroom-level predictors were
almost identical, only values from M4a are presented. * p<.01. † p<.05.
YOUNG EUROPEAN CITIZENS 30
Table 2. Multilevel linear regression model predicting trust in European institutions (n=66,749).
M1 M2 M3 M4a-d M5a-d
B (SE) β B (SE) β B (SE) β B (SE) β B (SE) β
Individual level
Gender (Female) -0.94 (0.18)* -0.05 -1.52 (0.16)* -0.08 -1.52 (0.16)* -0.08 -1.52 (0.16)* -0.08 -1.52 (0.10)* -0.08
Migration background 0.42 (0.28) 0.01 -0.84 (0.27)* -0.03 -0.83 (0.27)* -0.03 -0.84 (0.27)* -0.03 0.30 (0.20) 0.01
Socioeconomic background 0.76 (0.08)* 0.08 0.35 (0.06)* 0.04 0.32 (0.05)* 0.03 0.32 (0.05)* 0.03 0.38 (0.06)* 0.04
Political interest 0.26 (0.01)* 0.26 0.25 (0.01)* 0.26 0.25 (0.01)* 0.26 0.12 (0.01)* 0.12
Discussing political issues -0.02 (0.01)* -0.02 -0.02 (0.01)* -0.02 -0.02 (0.01)* -0.02 0.01 (0.00) 0.01
Watching news on TV 0.35 (0.06)* 0.04 0.35 (0.06)* 0.04 0.35 (0.06)* 0.04 0.21 (0.05)* 0.02
Equal rights for immigrants 0.08 (0.01)* 0.08 0.08 (0.01)* 0.08 0.08 (0.01)* 0.08 0.04 (0.01)* 0.04
Gender equality 0.04 (0.01)* 0.04 0.04 (0.01)* 0.04 0.04 (0.01)* 0.04 0.05 (0.01)* 0.05
Open classroom climate 0.06 (0.01)* 0.06 0.06 (0.01)* 0.06 0.06 (0.01)* 0.06 0.02 (0.01)* 0.02
Opportunities for learning about Europe 0.12 (0.01)* 0.12 0.11 (0.01)* 0.11 0.11 (0.01)* 0.11 0.03 (0.00)* 0.03
Trust in country-related institutions 0.58 (0.01)* 0.57
Classroom level
Classroom socioeconomic background 0.18 (0.21) 0.06 0.17 (0.21) 0.06 0.57 (0.15)* 0.29
Open classroom climate 0.00 (0.02) 0.00 0.00 (0.02) 0.00 -0.01 (0.02) -0.05
Opportunities for learning about Europe 0.03 (0.01)† 0.09 0.03 (0.01)† 0.09 0.02 (0.01) 0.08
Country level
Gross domestic product per capita 0.01 (0.01)† 0.29 -0.01 (0.01) -0.37
Income inequality -15.74 (7.72)† -0.36 7.63 (6.10) 0.26
Gender inequality -13.66 (5.06)* -0.46 -0.55 (4.66) -0.03
Communist past -0.74 (0.66) -0.21 1.12 (0.41)* 0.48
Variance explained
Individual level .01 .14 .14 .14 .42
School level .01 .01 .09
Country level (M4a) .09 .13
Country level (M4b) .13 .07
Country level (M4c) .21 .00
Country level (M4d) .05 .23
Note. Variance components in the intercept-only model (M0): individual level – 92.86 (93%), classroom level – 4.30 (4%), school level – 2.84
(3%). Unstandardized regression coefficients, standard errors, and standardized coefficients are presented. Country level predictors were tested
one at a time in separate models (M4a-d; M5a-d). Since regression coefficients and standard errors of individual and classroom-level predictors
were almost identical, only values from M4a and M5a are presented. * p<.01. † p<.05.
YOUNG EUROPEAN CITIZENS 31
Table 3. Multilevel linear regression model predicting participation in the wider community (n=67,109).
M1 M2 M3 M4a-d
B (SE) β B (SE) β B (SE) β B (SE) β
Individual level
Gender (Female) 1.47 (0.21)* 0.09 1.63 (0.19)* 0.09 1.63 (0.19)* 0.09 1.63 (0.19)* 0.09
Migration background 0.62 (0.51) 0.02 -0.12 (0.47) 0.00 -0.15 (0.47) -0.01 -0.15 (0.47) -0.01
Socioeconomic background 0.50 (0.08)* 0.06 0.19 (0.07)* 0.02 0.27 (0.06)* 0.03 0.27 (0.06)* 0.03
Political interest 0.10 (0.01)* 0.12 0.10 (0.01)* 0.12 0.10 (0.01)* 0.12
Discussing political issues 0.16 (0.01)* 0.18 0.16 (0.01)* 0.18 0.16 (0.01)* 0.18
Watching news on TV 0.10 (0.05) 0.01 0.10 (0.05) 0.01 0.10 (0.05) 0.01
Equal rights for immigrants -0.01 (0.01) -0.01 -0.01 (0.01) -0.01 -0.01 (0.01) -0.01
Gender equality -0.04 (0.01)* -0.04 -0.04 (0.01)* -0.04 -0.04 (0.01)* -0.04
Open classroom climate 0.02 (0.01)* 0.03 0.02 (0.01)* 0.03 0.02 (0.01)* 0.03
Opportunities for learning about Europe 0.07 (0.01)* 0.08 0.07 (0.01)* 0.08 0.07 (0.01)* 0.08
Classroom level
Classroom socioeconomic background -0.56 (0.20)* -0.14 -0.56 (0.20)* -0.14
Open classroom climate 0.03 (0.02) 0.05 0.03 (0.02) 0.05
Opportunities for learning about Europe -0.01 (0.02) -0.01 -0.01 (0.02) -0.01
Country level
Gross domestic product per capita 0.00 (0.01) -0.05
Income inequality 17.31 (9.89) 0.34
Gender inequality 16.14 (6.57)† 0.46
Communist past 0.78 (0.82) 0.19
Variance explained
Individual level .01 .10 .10 .10
School level .02 .02
Country level (M4a) .00
Country level (M4b) .11
Country level (M4c) .21
Country level (M4d) .04
Note. Variance components in the intercept-only model (M0): individual level – 75.14 (88%), classroom level – 5.72 (7%), school level – 4.90
(6%). Unstandardized regression coefficients, standard errors, and standardized coefficients are presented. Country level predictors were tested
one at a time in separate models (M4a-d). Since regression coefficients and standard errors of individual and classroom-level predictors were
almost identical, only values from M4a are presented. * p<.01. † p<.05.
YOUNG EUROPEAN CITIZENS 32
Table 4. Multilevel linear regression model predicting participation at the European level (n=67,162).
M1 M2 M3 M4a-d M5a-d
B (SE) β B (SE) β B (SE) β B (SE) β B (SE) β
Individual level
Gender (Female) -0.44 (0.17) -0.02 -0.14 (0.14) -0.01 -0.13 (0.14) -0.01 -0.13 (0.14) -0.01 -0.52 (0.15)* -0.03
Migration background 1.97 (0.22)* 0.06 1.36 (0.23)* 0.04 1.42 (0.22)* 0.05 1.42 (0.22)* 0.05 1.43 (0.25)* 0.05
Socioeconomic background 1.77 (0.06)* 0.19 1.59 (0.08)* 0.17 1.40 (0.07)* 0.15 1.40 (0.07)* 0.15 1.33 (0.07)* 0.14
Political interest 0.07 (0.01)* 0.08 0.07 (0.01)* 0.08 0.07 (0.01)* 0.08 0.05 (0.01)* 0.05
Discussing political issues 0.11 (0.01)* 0.11 0.10 (0.01)* 0.11 0.10 (0.01)* 0.11 0.07 (0.01)* 0.07
Watching news on TV 0.00 (0.07) 0.00 0.01 (0.07) 0.00 0.01 (0.07) 0.00 -0.02 (0.07) 0.00
Equal rights for immigrants -0.02 (0.01)* -0.02 -0.02 (0.01)* -0.02 -0.02 (0.01)* -0.02 -0.02 (0.01) -0.02
Gender equality -0.04 (0.01)* -0.04 -0.04 (0.01)* -0.05 -0.04 (0.01)* -0.05 -0.03 (0.01)* -0.04
Open classroom climate -0.01 (0.01) -0.01 0.00 (0.01) 0.00 0.00 (0.01) 0.00 -0.01 (0.01) -0.01
Opportunities for learning about Europe 0.21 (0.01)* 0.22 0.20 (0.01)* 0.21 0.20 (0.01)* 0.21 0.18 (0.01)* 0.19
Participation in the wider community 0.24 (0.01)* 0.24
Classroom level
Classroom socioeconomic background 1.48 (0.20)* 0.32 1.48 (0.20)* 0.32 1.62 (0.21)* 0.35
Open classroom climate -0.09 (0.03)* -0.15 -0.09 (0.03)* -0.15 -0.09 (0.03)* -0.16
Opportunities for learning about Europe 0.12 (0.05)* 0.19 0.12 (0.05)* 0.19 0.12 (0.04)* 0.20
Country level
Gross domestic product per capita 0.04 (0.01)* 0.65 0.04 (0.01)* 0.65
Income inequality -35.40 (9.20)* -0.53 -39.49 (9.02)* -0.58
Gender inequality -18.32 (9.37) -0.41 -22.10 (9.39)† -0.48
Communist past -1.54 (1.13) -0.29 -1.73 (1.16) -0.32
Variance explained
Individual level .04 .13 .12 .12 .17
School level .15 .14 .17
Country level (M4a) .43 .42
Country level (M4b) .28 .34
Country level (M4c) .16 .23
Country level (M4d) .09 .10
Note. Variance components in the intercept-only model (M0): individual level – 84.87 (85%), classroom level – 9.50 (10%), school level – 5.45
(5%). Unstandardized regression coefficients, standard errors, and standardized coefficients are presented. Country level predictors were tested
one at a time in separate models (M4a-d; M5a-d). Since regression coefficients and standard errors of individual and classroom-level predictors
were almost identical, only values from M4a and M5a are presented. * p<.01. † p<.05.