Domestic transnationalism and the formation of pro-European sentiments European Union Politics 0(0) 1–25 ! The Author(s) 2016 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1465116516633536 eup.sagepub.com Article Domestic transnationalism and the formation of pro-European sentiments Juan J Fernández Departamento de Ciencias Sociales, Universidad Carlos III de Madrid, Spain Monika Eigmüller Seminar für Soziologie, Europa-Universität Flensburg, Germany Stefanie Börner Institut für Soziologie, Otto-von-Guericke-Universität, Magdeburg, Germany Abstract Increasingly, research on attitudes towards the European project focuses on trans- national practices. This article furthers the transnational approach by offering the first systematic analysis of how domestic transnationalism – i.e. transnational practices conducted in the home country – influences the formation of pro-European sentiments. We argue that domestic transnational activities foster recognition of common, trans- national interests and identities that support the European integration project. Using a 2013 Eurobarometer, we show the distinct need to pay attention to domestic trans- nationalism. Individuals engaging in more domestic transnational activities display more pro-European sentiments in four of our five dependent variables. Moreover, the effect of domestic transnationalism is particularly intense among less-educated citizens. Keywords Attitudes, domestic transnationalism, Europeanization, education Transnational interactions are taking center stage in the analysis of attitudes towards the European project. This stems from a reconsideration of Karl Corresponding author: Juan J Fernández, Departamento de Ciencias Sociales, Universidad Carlos III de Madrid, c/de Madrid 126, Getafe 28093, Spain. Email: jjfgonza@clio.uc3m.es at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Deutsch’s ‘transactional theory’ (Deutsch, 1952; Deutsch et al., 1957) and his insight that international engagement helps dispel prejudices and fosters percep- tions of common interests that affect individual identities and support for regional integration. Several recent studies confirm these principles, showing that a general international orientation is associated with an increased identification with Europe (Fligstein, 2008; Kuhn, 2011, 2012; Sigalas, 2010). While this new perspective is commendable because of its focus on attitude formation as the focal point (Gerhards and Hans, 2014), current research in this area has overlooked the role of highly common, transnational activities (TAs) conducted in the home country (see also Recchi, 2014). This article advances the ‘transactional approach’ by spe- cifically targeting the relationship between attitudes towards the European project and transnational experiences gained domestically without physical border cross- ing. Our research addresses two main questions: Can TAs conducted in the home country affect individual attitudes towards Europe? Is the role of domestically based, transnational experiences more salient among less-educated than highly educated individuals? To answer this, we focus on domestic TAs. While the ‘transactional approach’ sees foreign-based activities (e.g. having lived or vacationed abroad) as de facto, ideal forms of transnationalism (Mau and Mewes, 2012; Roudometof, 2005), we argue that domestic transnationalism equally shapes individual attitudes and iden- tities. Domestic transnationalism is defined as activities conducted in one’s home country that, nevertheless, transcend national, linguistic and cultural borders. Such activities might encompass Internet communications; socializing in one’s home country with people from other EU countries; reading newspapers and magazines or, where possible, watching television in a language not one’s own – each a domestic means of virtually crossing borders to participate in transnational experiences. Occurring either daily or weekly, these practices frequently recur, thereby constantly reinforcing individual awareness of the possibilities and advan- tages of regional integration. Such activities also shape pro-European sentiments – defined as a personal, affective identification with Europe, a positive evaluation of EU institutions and/or a cognitive familiarity with the European project. We further contend that domestic transnationalism is particularly consequential among less-educated citizens. As education levels decline, TAs tilt more towards domestic-based activities, making them appreciably more influential in forming pro-European attitudes and beliefs. This pattern is of utmost importance because – given the focus of previous research on elitist transnationalism – our study notes that European integration affects the attitudes and beliefs of broader groups than previously expected. The article’s empirical base consists of a multivariate analysis of pro-European sentiments in 28 member states. To assess the role of domestic transnationalism, we estimate interactions between various domestic-based activities and experiences of border crossing. Five dependent variables measure pro-European sentiments: attachment to Europe, self-identification as European, having a good image of the EU, objective knowledge of EU facts and familiarity with EU rights. 2 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Our main data source is the Eurobarometer 80.1 conducted in October 2013. The logit models include 21 individual-level control variables and country fixed-effects to address unobserved heterogeneity. Our research reveals that individuals who have recently conducted more TAs in their home country are significantly more likely to have pro-European sentiments in four of the five dependent variables considered. This supports our main argument that domestic transnationalism influences pro-European senti- ments. It also discloses that education levels significantly moderate the effect of domestic transnationalism in four out of the five indicators of pro-European atti- tudes. These results are in line with our second expectation that the influence of domestic transnationalism declines as the level of education increases. Dimensions of pro-European sentiments Attitudes and preferences towards the European project have become well-estab- lished research topics in recent decades in the expectation that public opinion could have institutional consequences. The latter could determine multiple aspects of the European project such as the speed of European integration, the types of EU policies, even the platforms of domestic political parties (Fligstein, 2008; McLaren, 2006). Existing research commonly considers either support for the EU (e.g. Lubbers and Scheepers, 2007) or individual identification with Europe (e.g. Gabel, 1998). The literature has been criticized, however, on the grounds that it usually considers only one aspect of the European project – despite European integration being fundamentally multidimensional. Recent work appropriately calls for a more holistic approach, drawing on the theoretical literature about Europeanization to elaborate a more encompassing map of sentiments towards Europe (e.g. Boomgaarden et al., 2011). Scholars have devised several typologies to address individually relevant atti- tudes towards Europeanization. Some use a twofold differentiation between spe- cific and utilitarian attitudes; others choose diffuse and affective oriented attitudes (Boomgaarden et al., 2011; Scharpf, 1999). Still others broaden this approach by integrating a third dimension and distinguishing between a cognitive-evaluative, an emotive and a projective-connative dimension of Europeanization (Best et al., 2012; Scheuer, 2005) – i.e. between feeling, thinking and doing Europe (Immerfall et al., 2010). This approximates the social-psychological, three components model of attitudes consisting of an affective, a behavioral and a cognitive dimension. Since the nexus between actions and pro-European attitudes lies at the heart of the ‘transactional approach’ and attitudes and behavior have a rather complex rela- tionship, we modify these components by discarding behavior as an attitudinal dimension, and substituting evaluations in its place (see Augoustinos and Walker, 1995). Partially building on the aforementioned approaches and without seeking to establish a definitive typology of individual dimensions of pro-European sentiments, we, thus, believe it helpful to distinguish between affective, cognitive and evaluative dimensions of beliefs about Europe. Fernández et al. 3 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ The cognitive dimension refers to the objective and self-perceived familiarity with central elements of the European tradition and EU institutions. Knowledge of EU institutions does not necessarily lead to a positive assessment of the EU (Bruter, 2005), but an awareness of individual rights may facilitate support for European integration. The affective dimension captures one’s personal identi- fication with Europe, i.e. to feel ‘not only as part of, but also with the collective’ (Kuhn, 2015: 41). It includes self-categorization as a European citizen as a form of collective identity that can overlap (or replace) national, regional or local identities with an overall attachment to the political, territorial and institutional body of the European Union (Kuhn, 2012). Like the affective dimension, the evaluative dimension is also subjective in character and refers to personal assessments made by citizens regarding the performance of European institutions. The assessment could be positive or negative depending on the capability of a given institution, as well as the diverse, normative frames of individual actors. An evaluative assessment could further derive from abstract constructs (e.g. ‘European values’) or very con- crete organizations (e.g. European Commission). Most work to date focuses on the latter. We consider the affective, cognitive and evaluative dimensions as distinct – yet related – aspects of a general, pro-European system of integrated, EU beliefs that includes attitudes, forms of awareness and identities. Domestic transnationalism TAs have gained increasing scrutiny in analyzing the three dimensions of Europeanness. Based on Karl Deutsch and associates’ seminal work, this approach argues that ‘individual’ and ‘horizontal’ transnationalism fosters favorable atti- tudes towards European integration (Fligstein, 2008; Kuhn, 2011, 2012). Our the- oretical framework combines principles of Deutsch’s transactional theory with recent advances in political psychology summarized by John Zaller (1992). Deutsch (1952; Deutsch et al., 1957) argues that institutional projects of transna- tional integration (like the European project) trigger the type of practices and beliefs that underpin new supranational identities and support further institutional integration. Removal of borders, in this view, facilitates the expansion of cross- border, personal interactions and networks that help dispel prejudices about for- eigners; help people realize common interests in further cooperation; and ultimately generate a form of transnational, collective identity. Having developed novel, transnational interests and identities, a ‘new sense of community’ can arise, with citizens thus becoming increasingly averse to potential, inter-state, military con- frontations and more supportive of regional, institutional integration. Acts of individual transnationalism, in Deutsch’s theory, lead to the revaluation of broader, social communities (Nelsen and Guth, 2015: 13), thereby increasing personal identification with Europe and support for the European project in gen- eral. Based thereon, recent key empirical contributions have analyzed the attitudi- nal impact regarding ‘the extent to which individuals are involved in cross-border interaction and mobility’ (Kuhn, 2011: 814) with generally positive results. 4 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Fligstein (2008) shows that the most educated- and upper-classes are the most likely to travel to other EU countries and more likely to have pro-European attitudes. A range of studies also show that intra-EU migrants are likely to have a European identity (Rother and Nebe, 2009) and that proximity to a country’s borders fosters Europeanism (Gabel, 1998). Kuhn (2011, 2012) points out that citizens engaging in a broad array of TAs (e.g. travel abroad, contacts with other EU-nationals) tend to support the EU. While recent contributions underscore the general role of a transnational orienta- tion, they leave unexplored the main types of TAs important to forming European attitudes (Recchi, 2014).1 These, as well as previous studies on transnationalism (e.g. Favell and Recchi, 2009; Mau and Mewes, 2012; Roudometof, 2005), consider international migration and travel as the implicit, ideal, transnational actions. Less demanding forms (e.g. socializing with other EU nationals in one’s home country) are, instead, conceptualized as imperfect substitutes. Only recently have scholars recognized domestic transnational practices as proper cross-national activities; the main difference being that border crossing is virtual. For example, Favell et al. (2011) differentiate between physical and non-physical types of mobility; the latter category includes activities like consuming international media, or having contact with people from other countries. Since most people are ‘stayers’ (Favell and Recchi, 2009: 4), we suggest to take these domestic TAs into account when studying the attitudinal effect of ‘virtual’ border-crossing practices. We draw specifically on Deutsch’s work to differentiate between types of TAs. Deutsch sets similarities across personal, cultural toolkits at the center of his expla- nation of national identity and support for international ‘security communities’. For him, broad-based identifications do not emerge mechanically from common material interests that lead only to superficial interactions, but interactions cemented in ‘com- municative efficiency’ (Deutsch 1952: 96). This ‘efficiency’ occurs when actors from very different milieus – e.g. different territories or social classes – have cultural structures of symbols, values, habits and/or common interpretations with what Parsons (1951: 11) calls sufficient ‘pattern-consistency’. Complementarity in cultural resources across diverse social groups is needed to achieve the communicative fluidity and meaningful informational exchanges required to develop a perception of com- monness. The communicative structures carried by individuals are in constant trans- formation for Deutsch and, most important, can be learned collectively, setting the conditions for evolving, collective identities. Nothing therein assumes that preconditions for information exchanges can be achieved only through physical mobility. Deutsch (1952) stresses, in fact, the insufficiency of a ‘mere interchange of goods and services’ for this purpose (91). The key mechanism is the existence of ‘unbroken links of social communication’ (Deutsch et al., 1957: 51) out of which a common language and set of predis- positions can evolve. Border crossing, therefore, is not a precondition for forming these links. Individuals can also develop communicative resources comple- mentary to those of distant groups through domestic-based experiences of transnationalism. Fernández et al. 5 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ We turn now to an as yet marginalized but important area of research: domestic- based transnational activities (DBTAs). DBTAs refer to encountering the people, cultures, markets or discourses of other countries without physically crossing bor- ders. Socializing with EU nationals from other countries; watching TV or reading in another language; purchasing products from other countries via the Internet – all without border crossing – are forms of DBTAs. DBTAs could enhance levels of individual commitment to the European project and reshape educational bias in pro-European attitudes. We rely on European identity literature as well as Zaller’s (1992) classical con- tribution to sociopolitical attitude formation, The Nature and Origins of Mass Opinion, to formulate our arguments. Zaller condenses an extensive literature on political psychology in his book, contending that most people do not have a set of modular, coherent beliefs readily available to formulate an explicit preference when pressed to do so. Their decisions are instead based on ‘considerations’ – or mental points of reference (e.g. factual information, expected effects or moral implications) – brought to mind by the questionnaire. To Zaller (1992), only when individuals have sufficiently complex, integrated sets of ‘considerations’ can they connect a survey question to their own values and interests. Otherwise, their response will hinge on recent, random experiences or cues included in the survey question. We use Zaller’s framework because it has direct implications for linking transnation- alism and European attitudes. It also calls our attention to recurrent DBTAs and the general, attitudinal stability of citizens with higher levels of education. In Deutsch’s model, DBTAs like foreign media use, cross-border Internet pur- chases and domestic interactions with foreigners from other EU countries are per- sonal experiences having the necessary properties to generate pro-European considerations. Historically, news media use fosters a sense of identity because news media are cultural products representing the society to which they refer (Anderson, 2006). Similarly, domestic use of foreign media helps people recognize shared values, challenges and mutual dependencies with other countries, which could stimulate cross-national empathy. It also bolsters communicative skills contributing to fluid, cultural exchanges that lead to cross-border, emotional identification. In short, foreign media use creates transnational dispositions and provides tools to interact effectively with other cultures (Norris and Inglehart, 2009). Tariff-free, inter- national purchases via the Internet also offer immediate, practical instances of the benefits intrinsic to regional, economic integration. Even though DBTAs are virtual, they indeed increase awareness of advantages to be gained by more international, cultural and economic exchanges among European countries. Such awareness under- pins the identification with and general support essential to the European integration project. DBTAs help forge conditions favorable to pro-European sentiments. Likewise, Deutsch’s theory has affinities with intergroup contact theory, a body of research in social psychology that considers the effects of cross-group interactions. Intergroup contact theory (Allport, 1979) argues that long-term contact between individuals of different groups, involving a common goal, fosters mutual knowledge and helps reduce prejudices. Recent comparative research supports this approach, 6 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ especially regarding attitudes towards foreigners. In Europe, having regular, mean- ingful, personal interactions with foreign friends or foreign coworkers reduces indi- vidual perceptions of external ethnic threats (McLaren, 2006) and anti-immigrant prejudices in general (Quillian, 1995; Schneider, 2008). We can infer from such research that these interactions also facilitate intercultural bonds and a sense of commonality. Another form of DBTAs, direct domestic interactions with other EU-nationals, could constitute practical, easily interpretable, micro-demonstrations of successful communicative exchanges that activate pro-European considerations. Another of Zaller’s axioms regarding public opinion formation suggests that the recency of DBTAs could also contribute to favorable attitudes. The ‘accessibility axiom’ of his theory states that the more recently a consideration [or message] has been called to mind or thought about, the less time it takes to retrieve that consideration or related considerations from memory and bring them to the top of the head for use. (Zaller, 1992: 42) Recent experiences, therefore, are more likely to be the considerations that are used in response to a question. Political campaign research, as example, shows that nonpartisan phone calls made a week before election to encourage participation actually increase voter turnout (Nickerson, 2007: 276). The survey analyzed in this study asks questions only about the recurrence of DBTAs. We strongly believe that recency is co-related with the recurrence. The more recurrent an activity is, the higher the chance that it was recently conducted. Further research (based on new data) could explore the relationship between recurrence and recency in TAs, but relying on a proxy of recency is the best option we have at this point. How recurrent are DBTA, however? Table 1 includes an indicator of recurrence for actors engaged in different types of TAs. Based on the 2013 Eurobarometer 80.1, it provides descriptive evidence of whether, during the last 12 months, each of five activities were conducted ‘once or twice’ or ‘on several occasions’. It excludes individuals who did not conduct a DBTA in the last year. Percentages of clear recurrence (‘on several occasions’) are much higher among DBTAs than cross- border activities because border crossing has higher emotional and economic costs than DBTAs. This means that DBTAs have a high potential for being recalled, at the moment of the survey interview, as a recent personal experience with attitude-formation content driving the ultimate response. H1: Individuals who conduct domestic TAs are more likely to hold pro-European sentiments than those who do not. DBTAs, the education-bias and pro-European sentiments We contend that DBTAs could gradually create the foundations for persistent, pro- European attitudes, and play an important role in moderating the education bias in Fernández et al. 7 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ pro-European sentiments. Education levels are a substantial focal point in the empirical literature on attitudes towards the European project. Scholars from diverse, theoretical approaches concur that elites (Checkel and Katzenstein, 2008; Dı́ez Medrano, 2008) and citizens with higher education are more likely to hold pro-European attitudes. These approaches postulate numerous reasons why better-educated citizens are more pro-European. They are better able to accumu- late transnational social capital and face the challenges of transnationalization to their social networks. They have superior bargaining positions in the labor market and can leverage these better positions to migrate abroad (Gabel, 1998). They are more exposed to the pro-European, doxic discourse of educated elites. They are also less worried about an immigrant, cultural threat or the undermining of national cultural symbols (McLaren, 2006; Schneider, 2008). Using an array of surveys, samples and countries, studies show that, consistent with these arguments, education is positively related to a range of pro-European attitudes and identities (Fligstein, 2008; Gabel, 1998; Kuhn, 2011; Lubbers and Scheepers, 2007; Marks and Hooghe, 2003; Rother and Nebe, 2009). Hakhverdian et al. (2013) provide indications, in fact, that educational bias has grown since at least 1992. Kuhn (2012) further argues that education level is relevant because it moderates the attitude-formation power of TAs. Education produces a ‘ceiling effect’, in this view, such that TAs have less formative power for those highly rather than less-educated. The descriptive evidence regarding the five dimensions of Europeanization used in this Table 1. Recurrence levels of transnational activities among citizens that conducted them in the last 12 months, 2013. On several occasions Once or twice Total Percent conducted this activity and visited other country Half least years educ. Half most years educ. Visited other EU country 43.0 57.0 100 – – Read in other language in her home country 56.5 43.5 100 31.7 53.9 Socialized in her home country with a foreigner from other EU country 61.4 38.6 100 74.9 82.5 Watched TV in other language in her home country 67.4 32.6 100 46.6 65.5 Made an Internet purchase from another EU country from her country of residence 45.7 54.3 100 30.2 46.2 EU: European Union. 8 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ study, however, is not consistent with the ceiling effect argument (Figure 1). Although education is positively associated with pro-European attitudes, between 58.6 and 24.6% – depending on the indicator – of very highly educated individuals (i.e. two standard deviations above the mean) do not display the five types of pro-European sentiments. Hence, additional TAs could still, theoretically, increase pro-European sentiments among highly educated individuals. We argue as an alternative that the role of TAs, in particular DBTAs, increases as the level of education declines due to (a) differences in attitude stability and (b) transnational omnivorousness.2 Regarding attitude stability, political psychol- ogy research concurs that the attitudes of citizens with lower levels of political sophistication and education tend to be less stable and are more affected by recent, personal experiences (Zaller, 1992). A higher education level enhances cog- nitive skills and the level of political information, thereby increasing the number of considerations to evaluate and judge any given issue. This ensemble of multiple considerations provides better skills to connect an external input of information (like a questionnaire item) with the person’s values and experiences. Individuals, therefore, with higher education levels tend to have a more cohesive, all-encom- passing corpus of fairly stable beliefs, dispositions and attitudes. Since having more considerations facilitates the connection with personal values, it accordingly leads to highly consistent interpretations of the world, less likely influenced by each new input of information. Schuman and Presser (1996: 150) show that for a wide range 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 8 12 16 20 24 28 32 36 Education years 8 12 16 20 24 28 32 36 Education years 8 12 16 20 24 28 32 36 Education years 8 12 16 20 24 28 32 36 Education years 8 12 16 20 24 28 32 36 Education years Attachment to EU Feels European Good image of EU Knows the EU rights Objective knowledge P ro -E ur op ea n se nt im en t Education years Figure 1. Predicted level of pro-European sentiments by education level in the 28 EU member states, 2013. EU: European Union. Fernández et al. 9 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ of issues ‘less-educated persons are more apt to give ‘‘DK’’ [don’t know] or ‘‘no opinion’’ answers’, thereby supporting the claim that the more educated have more considerations available for any given survey question. Political non-sophisticates and the less educated, by contrast, tend to mobilize fewer considerations. They have inconsistent perceptions and ideas, without a well- integrated set of hierarchical principles of appreciation and perception. Facing new statements, citizens with lower levels of education often lack the informational tools needed to assess them in light of their values, making recent experiences especially influential in their stated attitudes. Collingwood’s (2012) survey experiment demon- strates how education fosters the stability of attitudes towards direct democracy. The results show that evaluations of direct democracy among non-college educated citi- zens were more sensitive to the inclusion of prior questions that could discourage support for direct democracy, than those of college educated citizens. Although the level of education should shape the influence of TAs in general, it should for two reasons be especially consequential for the effect of DBTAs: (a) individuals differ in the structure of their TAs and (b) the level of education affects this structure of activities. If domestic TAs are not counter-balanced by direct, foreign experiences, they must prove especially influential in attitude formation. Moreover, DBTAs should be especially relevant for citizens with lower education levels, because they particularly dominate the structure of TAs in this group. A higher level of education provides more contexts and tools to reap the social benefits of a truly diverse, cultural disposition (Bourdieu, 1984), thus incentivizing omnivorous cultural attitudes. Better educated individuals have, in fact, more omnivorous tastes in music, theater and the visual arts (Bryson, 1996; Warde and Gayo-Cal, 2009) – a gap which could have grown over time (Garcı́a-Álvarez et al., 2007). This variable structure and its relationship to an individual’s educa- tion have clear implications for the impact of TAs in general. Consequently, the imprint of domestic activities among those with less educa- tion cannot be balanced out by personal experiences of border-crossing transna- tionalism, common to the highly educated. A given unit of DBTAs, therefore, does not have the same meaning and influence across the education spectrum. Since DBTAs are practically the sole forms of activity, their influence on attitude or belief formation should be higher at lower levels of education. The Eurobarometer analyzed in this project provides strong indications, in fact, of a positive relationship between education and transnational omnivorousness in TAs. When considering only individuals who indulged at least one form of TA (e.g. reading in another language, or socializing with an EU national from another state), the proportion of citizens that also traveled abroad increases substantially with the level of education (Figure 2). Moreover, citizens with lower levels of education are less likely to have engaged in each DBTA considered and also travelled abroad (Table 1). For these reasons, we hypothesize that: H2: The level of education moderates the effect of domestic-based TAs on pro- European sentiments. 10 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Data and methods To test the hypotheses, we rely on the Eurobarometer 80.1. We chose the 80.1 Eurobarometer because, at the time of writing, it was the latest one available that includes (a) five different proxies of pro-European preferences necessary to con- struct the dependent variables, and (b) detailed information about the types of TAs undertaken by the respondents during the last year. The fieldwork of the 80.1 Eurobarometer was conducted in November 2013. The following analysis includes all 28 EU member states at the time of the survey. Five EU candidate members are excluded from the analysis due to lack of information for the dependent variables. Dependent variables To create a comprehensive perspective on the determinants of beliefs and attitudes about the European project, we construct five, dichotomous, dependent variables to capture the affective, evaluative and cognitive dimensions of these sentiments. Two indicators seize the affective element: attachment to the EU, distinguishing respondents who feel ‘very’ or ‘fairly attached’ to the EU (1) from those who do not feel this attachment (0); and self-identified European, identifying respondents with some form of exclusive European identity (1) or a combination of European and national identity (1) from respondents with exclusive national identification (0). Having a positive image of EU addresses evaluative Europeanization, which .3 .4 .5 .6 .7 .8 -2SD -1SD Mean +1SD +2SD Figure 2. Proportion of individuals with cross-border experience among those with transnational experiences by level of education. Fernández et al. 11 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ differentiates individuals with a ‘very’ or ‘fairly positive image’ of the EU (1) from those with a ‘neutral’ or ‘negative image’ of the EU (0). Two indicators capture the cognitive dimensions: knowledge of the EU rights, distinguishing persons who answer ‘definitely’ and ‘to some extent’ to the statement ‘you know your rights’ as EU citizens (1) from the rest (0); and objective knowledge about the EU, a less commonly used but very informative indicator. The latter has a value of (1) for individuals correctly answering (a) the number of EU member states at the time of the survey (correct answer: 28); (b) whether European Parliament members are elected directly or not (correct answer: directly); and (c) Switzerland’s status in the EU (correct answer: non-member); and (0) for respon- dents who did not answer or provide correct answers. Independent variables and analytical strategy Based on the theoretical discussions in ‘Domestic transnationalism’ and ‘DBTAs, the education-bias and pro-European sentiments’ sections, the main independent variables address the level of education and the type of TAs. We do not dichot- omize the years of education into three or more groups because that would require arbitrary decisions about thresholds (e.g. high, medium, low) and the expected age of transition between education levels, which varies by domestic educational system. We also have no theoretical reasons to believe that education should have a non-linear effect. The variable education years refers, therefore, to the age at which the person completed his or her full-time education. Following Lubbers and Scheepers (2007), we take the respondent’s age at the time of the interview to assign a value for those still studying and not yet finished with their full-time education. We define less-educated individuals as those with less than one standard deviation of the mean, which represents 10.9% of the European population. The 80.1 Eurobarometer poses a major challenge in examining the attitudinal effect of TAs because the questionnaire does not ask about the location where some TAs took place. One group of questions inquires about experiences abroad (e.g. having traveled abroad recently). Another asks about having watched TV or read in another language, or having socialized with an EU national from another EU member state without specifying whether these activities occurred in the respon- dent’s home country or abroad. It is impossible to capture DBTAs, therefore, simply by constructing an index of activities explicitly bound to a given country and assessing its effect. It is necessary instead to estimate an interaction term between having conducted TAs and having foreign experiences, which then allows us to discern the effect of DBTAs when the respondent clearly has no foreign experience. To this end, we construct two indexes. The first – foreign experience – is a cumulative. It assigns a value of one to respondents who ‘ever lived’, ‘ever worked’ or ‘ever studied’ abroad – adding to a maximum value of three; and a value of one to those having visited another EU country ‘in the last 12 months’, ‘once or twice’ and a value of two to those who travelled abroad ‘on several 12 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ occasions’. Foreign experience ranges from zero to five. A value of zero means no background of having lived abroad and no recent border crossing. The second index – TAs – summarizes four different activities: having ‘read a book, newspaper or magazine in a language other than your mother tongue’; ‘socialized with people from another EU country’; ‘watched TV programs in a language other than your mother tongue’; and ‘used the Internet in order to pur- chase a product or service from another EU country’. For each, a value of one indicates having conducted it ‘once or twice’ in the last 12 months, and a value of two if ‘on several occasions’ in the last 12 months. The index has a range from zero – for a person not conducting any of these activities in the last year – to eight, for a person conducting all of them several times in the last year.3 This is a valid indi- cator of our ‘domestic transnationalism’ concept because it captures at least three of five hypothetical DBTAs: face to face interactions with foreigners; interactions with foreigners via other means of communication; use of foreign media; foreign language use; and use of foreign-produced, inanimate objects. Given that foreign experience¼ 0 identifies domestically bound individuals – i.e. never lived abroad and did not travel abroad recently – including these two indexes and their interaction term foreign experienceTAs allows us to interpret the coefficient of TAs as the effect of transnationalism among specifically immobile or domestically bound individuals. Mokken scale tests indicate the construct validity of foreign experi- ence and TAs. The scales have overall HjK coefficients of .491 and .569, respectively, that are above the standard threshold (.30) for a satisfactory scale (van Schuur, 2003). All following models include a range of control variables to address alternative explanations. According to the utilitarian approach, individuals with more market- able skills and higher incomes have more to gain from European integration (Fligstein, 2008; Gabel, 1998). In response, the models control for nine social classes and working status (managers, employed professionals, white-collar workers, supervisors, manual workers, house person, unemployed, retired and student) and the subjective, ego-tropic, financial situation. Household financial situation identifies respondents with a self-described ‘very good’ or ‘rather good’ financial situation in their household (1) from the rest (0). The national attachment approach believes collective loyalties and identities display zero-sum relationships (Carey, 2002). If this is the case, individuals ‘very’ or ‘fairly attached’ to their home country (national attachment) should display lower commitments to Europe (see Nelsen and Guth, 2015; for the opposite view, see Dı́ez Medrano and Gutiérrez, 2001). According to the institutional trust approach, citizens rely on domestic cues to form their atti- tudes towards the EU. In this view, trust in the national government and the EU should be linearly related (Armingeon and Ceka, 2014). We therefore control for trust in national government that differentiates individuals in this situation (1) from the rest (0). The models also control for female, age and town size, which have been found relevant in previous empirical studies (see Mau, 2005; Nelsen and Guth, 2000). Eurobarometer 80.1 offers no questions regarding a respondent’s political party preferences or position in a continuous, left–right index (Marks and Hooghe, Fernández et al. 13 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ 2003). They therefore could not be included in the analysis. We have no reason to believe, however, that these two variables should be strongly related to the level of education or engaging in TAs. Given the presence of interaction terms, all contin- uous variables (education years, age and age2) have been centered. Finally, the dimensions of pro-European sentiments could be causally related. Based on Inglehart (1970), objective knowledge should favor a positive image of the EU; based on Dı́ez Medrano and Gutiérrez (2001), a positive image of the EU should favor self-identification as European and attachment to Europe. Regarding statistical models, one possible strategy would involve ordinal logit models. Yet, these models build on the commonly unmet, parallel lines assumption (Long and Freese, 2006). Brant tests (reported in the online appendix) indicate the violation of this assumption for the five dependent variables. We therefore test our hypotheses through logit models. Since the two hypotheses refer to the individual level, our main results do not draw on multilevel models. Instead, we control for unobserved heterogeneity at the country-level through country fixed-effects and clustered robust standard errors. Results Table 2 indicates that the respondent’s gender, collective identity, subjective finan- cial situation, town of residence size and trust in national government all have robust impacts on pro-European sentiments. Women are significantly less likely to self-identify as a European; think they know their EU-related rights better than men; and have objective knowledge of the EU. Interestingly, strong national attach- ment is related to feeling attached to the EU, having a positive image of the EU and a better objective knowledge of Europe. Further work could explore if this effect is due to a syndrome of multi-level communitarianism across groups of Europeans. Supporting the utilitarian approach to pro-European attitudes (Gabel, 1998) – household financial situation is positive and significant in all models. Consistent with the expectation of Inglehart (1970), objective knowledge about the EU increases the probability of having a positive image of the EU. Regarding the role of education, the evidence reveals a strong, positive effect in all five dependent variables. Ceteris paribus, a standard deviation increase in the years of education (4.52 years) is associated with a 11.0%, 17.0%, 10.1%, 14.9% and 22.8% increase in the probabilities of feeling attached to the EU; identifying as European; having a positive image of the EU; objective knowledge about the EU; and knowing the EU rights, respectively. Given that the Models control for social class and, indirectly, for purchasing power, education does not capture only the structural position of the individual. Socio-demographic conditions, however, do not exhaust the determinants of these attitudes and beliefs. Experiences of transnationalism also have an impact on pro- European sentiments. Because of the interaction terms included in Models 1–5, there- fore, the indicators of foreign experience and TAs must be interpreted in conjunction. Foreign experience indicates that, among individuals without any TA in the last year 14 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ T a b le 2 . L o gi t m o d e ls p re d ic ti n g fiv e in d ic at o rs o f E u ro p e an iz at io n in 2 8 E U -m e m b e r st at e s, 2 0 1 3 . M o d e l 1 M o d e l 2 M o d e l 3 M o d e l 4 M o d e l 5 M o d e l 6 M o d e l 7 M o d e l 8 M o d e l 9 M o d e l 1 0 Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts Fo re ig n e x p e ri e n ce .2 4 5 ** * .2 2 7 ** * .1 9 5 ** * .1 6 2 ** .1 6 0 ** * .2 1 3 ** * .2 1 2 ** * .1 9 7 ** * .1 4 9 ** .1 3 4 ** * (. 0 5 7 ) (. 0 3 9 ) (. 0 3 5 ) (. 0 5 1 ) (. 0 3 2 ) (. 0 4 8 ) (. 0 3 5 ) (. 0 3 5 ) (. 0 4 8 ) (. 0 3 0 ) T ra n sn at io n al ac ti vi ti e s .0 9 2 ** * .1 3 8 ** * .1 0 6 ** * .0 1 4 .1 9 0 ** * .0 8 7 ** * .1 3 9 ** * .1 1 0 ** * .0 2 1 .1 8 6 ** * (. 0 1 2 ) (. 0 2 4 ) (. 0 1 7 ) (. 0 1 6 ) (. 0 1 0 ) (. 0 1 1 ) (. 0 2 5 ) (. 0 1 9 ) (. 0 1 7 ) (. 0 1 1 ) T ra n s. ac ti vi ti e s Fo re ig n e x p . .0 1 8 .0 1 2 .0 1 4 .0 2 1 ** .0 1 2 .0 0 8 .0 1 0 .0 1 6 .0 2 2 ** .0 0 5 (. 0 1 1 ) (. 0 1 0 ) (. 0 0 9 ) (. 0 0 7 ) (. 0 0 9 ) (. 0 1 0 ) (. 0 1 1 ) (. 0 1 0 ) (. 0 0 8 ) (. 0 0 9 ) E d u ca ti o n ye ar s .0 2 5 ** * .0 5 9 ** * .0 2 3 ** .0 4 7 ** * .0 5 4 ** * .0 6 2 ** * .0 8 5 ** * .0 2 0 .0 7 5 ** * .1 0 1 ** * (. 0 0 4 ) (. 0 0 9 ) (. 0 0 9 ) (. 0 0 6 ) (. 0 0 7 ) (. 0 1 2 ) (. 0 1 6 ) (. 0 1 6 ) (. 0 1 2 ) (. 0 1 1 ) Fo r. e x p e r. E d u c. ye ar s .0 1 2 .0 0 4 .0 0 7 .0 1 0 .0 3 0 ** * (. 0 0 7 ) (. 0 0 8 ) (. 0 1 2 ) (. 0 0 6 ) (. 0 0 8 ) T ra n s. ac t. E d u c. ye ar s .0 1 2 ** .0 1 4 * .0 0 1 .0 1 3 ** * .0 1 5 ** * (. 0 0 4 ) (. 0 0 6 ) (. 0 0 3 ) (. 0 0 3 ) (. 0 0 3 ) Fo r. e x p e r. T ra n s. ac t. E d u c. ye ar s .0 0 1 .0 0 3 .0 0 0 .0 0 4 ** * .0 0 5 ** * (. 0 0 1 ) (. 0 0 2 ) (. 0 0 2 ) (. 0 0 1 ) (. 0 0 1 ) Fe m al e .0 7 3 .1 6 6 * .1 1 0 .4 8 0 ** * .2 3 3 ** * .0 6 9 .1 6 4 * .1 1 1 .4 7 8 ** * .2 3 0 ** * (. 0 4 5 ) (. 0 6 5 ) (. 0 6 3 ) (. 0 9 3 ) (. 0 6 0 ) (. 0 4 3 ) (. 0 6 4 ) (. 0 6 3 ) (. 0 9 2 ) (. 0 6 0 ) A ge .0 1 4 .0 2 1 * .0 3 3 ** * .0 5 4 ** * .0 2 1 ** .0 1 4 .0 2 2 * .0 3 3 ** * .0 5 4 ** * .0 2 1 * (. 0 0 9 ) (. 0 1 0 ) (. 0 0 9 ) (. 0 0 9 ) (. 0 0 8 ) (. 0 0 9 ) (. 0 1 0 ) (. 0 0 9 ) (. 0 0 9 ) (. 0 0 9 ) A ge 2 .0 0 0 .0 0 0 * .0 0 0 ** .0 0 0 ** * .0 0 0 ** * .0 0 0 .0 0 0 * .0 0 0 ** .0 0 0 ** * .0 0 0 ** (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) (. 0 0 0 ) W o rk st at u s (r e f. se lf- e m p lo ye d ) M an ag e rs .1 9 5 .2 6 9 .0 3 7 .0 9 8 .1 7 3 * .2 2 2 .2 8 7 .0 3 5 .0 8 0 .1 9 9 * (c o n ti n u e d ) Fernández et al. 15 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ T a b le 2 . C o n ti n u e d M o d e l 1 M o d e l 2 M o d e l 3 M o d e l 4 M o d e l 5 M o d e l 6 M o d e l 7 M o d e l 8 M o d e l 9 M o d e l 1 0 Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts (. 1 4 4 ) (. 1 6 8 ) (. 1 3 3 ) (. 0 5 1 ) (. 0 8 6 ) (. 1 4 0 ) (. 1 7 2 ) (. 1 3 4 ) (. 0 5 1 ) (. 0 8 3 ) E m p lo ye d p ro fe ss io n al .1 5 2 .4 3 0 .3 0 9 * .0 0 2 .0 5 9 .1 7 5 .4 5 1 .3 0 6 * .0 1 8 .0 3 9 (. 2 0 0 ) (. 2 2 1 ) (. 1 2 7 ) (. 1 6 2 ) (. 0 9 3 ) (. 2 1 2 ) (. 2 3 3 ) (. 1 2 8 ) (. 1 6 4 ) (. 0 9 3 ) W h it e co lla r w o rk e r .0 6 2 .0 1 7 .1 6 4 .1 3 4 * .1 9 7 * .0 5 4 .0 1 8 .1 6 2 .1 3 6 * .1 9 0 * (. 1 0 0 ) (. 0 8 0 ) (. 0 8 8 ) (. 0 5 9 ) (. 0 9 0 ) (. 1 0 1 ) (. 0 7 9 ) (. 0 8 9 ) (. 0 5 9 ) (. 0 9 0 ) Su p e rv is o r .1 4 2 .3 3 4 .1 3 2 .0 9 3 .0 2 5 .1 2 3 .3 2 4 .1 3 0 .0 8 5 .0 0 5 (. 2 0 5 ) (. 3 0 2 ) (. 1 9 4 ) (. 1 9 4 ) (. 1 7 0 ) (. 2 0 4 ) (. 2 9 8 ) (. 1 9 6 ) (. 1 9 1 ) (. 1 7 8 ) M an u al w o rk e rs .0 6 2 .0 0 7 .2 1 8 * .3 6 7 ** * .0 0 9 .0 6 5 .0 1 5 .2 1 4 * .3 5 9 ** * .0 1 0 (. 1 0 0 ) (. 0 6 9 ) (. 1 0 4 ) (. 0 8 3 ) (. 0 5 7 ) (. 1 0 0 ) (. 0 6 7 ) (. 1 0 4 ) (. 0 8 4 ) (. 0 5 8 ) H o u se p e rs o n .0 5 9 .0 3 1 .1 5 7 .2 9 3 ** * .1 2 0 .0 6 9 .0 2 0 .1 5 5 .2 7 8 ** * .1 3 5 (. 1 0 1 ) (. 1 0 2 ) (. 1 9 9 ) (. 0 8 3 ) (. 1 1 8 ) (. 1 0 5 ) (. 1 0 4 ) (. 2 0 1 ) (. 0 8 2 ) (. 1 1 9 ) U n e m p lo ye d .1 0 0 .1 2 3 .0 1 0 .2 4 0 * .0 7 4 .0 9 8 .1 2 0 .0 1 3 .2 3 8 * .0 7 2 (. 1 5 4 ) (. 0 9 5 ) (. 0 6 4 ) (. 0 9 7 ) (. 1 1 3 ) (. 1 5 8 ) (. 0 9 6 ) (. 0 6 5 ) (. 0 9 4 ) (. 1 1 7 ) R e ti re d .2 3 0 .0 5 8 .0 1 0 .1 5 1 * .2 6 6 .2 2 2 .0 5 5 .0 0 7 .1 5 0 * .2 5 9 (. 1 4 3 ) (. 1 1 6 ) (. 0 6 4 ) (. 0 7 2 ) (. 1 4 6 ) (. 1 4 3 ) (. 1 2 1 ) (. 0 6 5 ) (. 0 7 1 ) (. 1 4 7 ) St u d e n t .1 4 1 .1 3 4 .0 6 3 .1 5 0 .1 6 0 .1 5 8 .1 4 4 .0 6 5 .1 3 1 .1 6 9 (. 1 9 0 ) (. 1 8 5 ) (. 1 0 5 ) (. 0 9 8 ) (. 1 3 4 ) (. 1 8 2 ) (. 1 8 2 ) (. 1 0 6 ) (. 0 9 6 ) (. 1 2 9 ) Fo re ig n e r .1 6 4 .3 5 5 * .0 0 6 .1 5 3 .0 3 1 .1 2 1 .3 3 5 * .0 0 3 .1 7 9 .0 7 2 (. 1 0 6 ) (. 1 6 1 ) (. 1 0 4 ) (. 1 3 6 ) (. 1 2 9 ) (. 1 1 1 ) (. 1 6 4 ) (. 1 0 5 ) (. 1 4 5 ) (. 1 2 6 ) N at io n al at ta ch m e n t 1 .2 4 8 ** * .2 1 8 .2 5 0 ** .2 6 1 ** * .0 7 0 1 .2 6 5 ** * .2 1 4 .2 5 1 ** .2 7 1 ** * .0 8 1 (. 1 1 1 ) (. 1 9 4 ) (. 0 9 4 ) (. 0 5 7 ) (. 0 9 6 ) (. 1 1 4 ) (. 1 9 4 ) (. 0 9 4 ) (. 0 5 8 ) (. 0 9 7 ) H o u se h o ld fin an ci al si tu at io n .4 1 2 ** * .2 7 5 ** * .5 0 0 ** * .3 1 3 ** * .4 3 3 ** * .3 9 8 ** * .2 6 7 ** * .5 0 1 ** * .3 0 3 ** * .4 2 1 ** * (. 0 3 9 ) (. 0 7 8 ) (. 0 5 6 ) (. 0 8 0 ) (. 0 7 5 ) (. 0 3 9 ) (. 0 8 0 ) (. 0 5 6 ) (. 0 8 2 ) (. 0 7 4 ) (c o n ti n u e d ) 16 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ T a b le 2 . C o n ti n u e d M o d e l 1 M o d e l 2 M o d e l 3 M o d e l 4 M o d e l 5 M o d e l 6 M o d e l 7 M o d e l 8 M o d e l 9 M o d e l 1 0 Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts Fe e ls at ta ch e d to th e E U Se lf- id e n ti fie d E u ro p e an P o si ti ve im ag e o f th e E U O b je ct iv e k n o w le d ge K n o w th e E U ri gh ts To w n si ze (r e f. ca t. sm al l) L ar ge .2 3 4 ** * .2 0 2 * .0 6 2 .2 2 6 ** * .1 2 4 ** .2 3 2 ** * .2 0 0 * .0 6 3 .2 2 1 ** * .1 2 2 ** (. 0 6 0 ) (. 0 8 5 ) (. 0 8 9 ) (. 0 6 2 ) (. 0 4 0 ) (. 0 6 0 ) (. 0 8 5 ) (. 0 8 9 ) (. 0 6 2 ) (. 0 3 9 ) M id d le .1 7 8 ** .1 2 1 * .0 7 1 .1 8 6 ** .1 2 5 ** .1 8 0 ** .1 2 4 * .0 7 1 .1 8 8 ** .1 3 2 ** (. 0 6 4 ) (. 0 6 1 ) (. 0 8 7 ) (. 0 7 1 ) (. 0 4 7 ) (. 0 6 3 ) (. 0 6 1 ) (. 0 8 7 ) (. 0 7 3 ) (. 0 4 9 ) T ru st in n at io n al go ve rn m e n t .5 3 6 ** * .3 4 0 ** * 1 .2 0 8 ** * .2 1 7 ** .4 9 6 ** * .5 3 7 ** * .3 3 8 ** * 1 .2 0 8 ** * .2 1 4 ** .4 9 6 ** * (. 1 0 4 ) (. 0 8 3 ) (. 1 3 1 ) (. 0 7 6 ) (. 0 6 1 ) (. 1 0 2 ) (. 0 8 4 ) (. 1 3 3 ) (. 0 7 7 ) (. 0 6 1 ) P o si ti ve im ag e o f th e E U 1 .3 9 0 ** * 1 .0 5 1 ** * 1 .3 9 6 ** * 1 .0 5 3 ** * (. 0 5 2 ) (. 0 6 0 ) (. 0 5 4 ) (. 0 6 0 ) O b je ct iv e k n o w le d ge .2 9 2 ** * .2 9 1 ** * (. 0 5 5 ) (. 0 5 5 ) C o n st an t 2 .4 4 3 ** * .1 6 3 1 .8 7 4 ** * 1 .2 1 7 ** * 1 .9 5 9 ** * 2 .4 1 3 ** * .1 3 8 1 .8 8 1 ** * 1 .2 0 3 ** * 1 .9 1 9 ** * (. 1 5 8 ) (. 1 5 7 ) (. 1 4 2 ) (. 1 1 4 ) (. 1 0 1 ) (. 1 5 9 ) (. 1 5 7 ) (. 1 4 3 ) (. 1 1 9 ) (. 1 0 1 ) R 2 .1 7 3 .1 5 4 .1 1 9 .0 6 6 .1 2 8 .1 7 6 .1 5 6 .1 1 9 .0 6 8 .1 3 1 O b se rv at io n s 2 3 ,4 5 3 2 3 ,2 6 5 2 3 ,7 3 3 2 4 ,0 1 5 2 3 ,8 3 7 2 3 ,4 5 3 2 3 ,2 6 5 2 3 ,7 3 3 2 4 ,0 1 5 2 3 ,8 3 7 C o u n tr y co e ff ic ie n ts ar e n o t sh o w n . Tw o -t ai le d te st , *p < 0 .0 5 , ** p < 0 .0 1 , ** *p < 0 .0 0 1 . St an d ar d e rr o rs ar e gi ve n in p ar en th e se s. E U : E u ro p e an U n io n . Fernández et al. 17 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ (¼0), having more experiences of border crossing does have positive effects in all five dependent variables. For those citizens without TAs, a unit increase in foreign experi- ence is linked to a 25.2%, 14.5%, 19.4%, 11.3% and 14.5% increase in the probability of feeling attached to the EU; self-identified as a European; having a positive image of the EU; having objective knowledge and knowing the EU rights, respectively. This offers strong evidence that experiences of traveling abroad do have formative influ- ences on attitudes and beliefs about Europe. Most important for our purposes, TAs is positive and significant in four of the five models.4 Again, due to the interaction term TAs foreign experience, the coefficients of TAs reveal that, among domestically circumscribed citizens (foreign experi- ence¼ 0), each additional unit of these activities has a positive effect in four of the dependent variables. For these immobile citizens, a unit increase in this variable is linked to a 8.9%, 8.8%, 10.2% and 17.5% increase in the probability of feeling attached to the EU; self-identify as European; having a positive image of the EU; and knowing the EU rights, respectively. This is also a strong evidence in support of H1, i.e. domestic transnationalism is conducive to pro-European sentiments. -.2 -.1 0 .1 .2 .3 -.2 -.1 0 .1 .2 .3 0 1 2 3 4 5 Foreign experience 0 1 2 3 4 5 Foreign experience 0 1 2 3 4 5 Foreign experience 0 1 2 3 4 5 Foreign experience 0 1 2 3 4 5 Foreign experience Attachment to Europe Self-identified European Good image of the EU Knows her/his EU rights Objective knowledge on EU E ff ec t o f tr an sn at io na l a ct iv it ie s E ff ec t o f tr an sn at io na l a ct iv it ie s Figure 3. Estimated effects (and 95% confidence intervals) of transnational activities in pro-European sentiments at different levels of foreign experience. 18 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ The interaction terms in Models 1–5 indicate that foreign experience does not significantly increase the impact of TAs. In fact, given the negative signs of the coefficients in all models – and in light of Figure 3 – the effect of TAs peaks among immobile citizens and declines with the level of foreign experience. In the case of attachment to Europe, TAs is only positive and significant among those who have a value of two or less in foreign experience.5 We hypothesized in ‘DBTAs, the education-bias and pro-European sentiments’ section that the effect of domestically circumscribed, TAs should be particularly strong as the level of education declines. To test this hypothesis, we estimate a three-way interaction between foreign experience, TAs and education years (Models 6–10 in Table 2). These complex interaction terms allow us to determine the effect of TAs at different levels of education, specifically among individuals with no border- crossing experience (foreign experience¼ 0). Given that we now focus on the moder- ating role of education for TAs when foreign experience¼ 0, the two key terms are TAs and TAs education year. Since education years is centered, TAs represents the -.4 -.2 0 .2 .4 .6 .8 -.4 -.2 0 .2 .4 .6 .8 -2 .0 S D -1 .0 S D M ea n + 1. 0 SD + 2. 0 SD Years of education -2 .0 S D -1 .0 S D M ea n + 1. 0 SD + 2. 0 SD Years of education -2 .0 S D -1 .0 S D M ea n + 1. 0 SD + 2. 0 SD Years of education -2 .0 S D -1 .0 S D M ea n + 1. 0 S D + 2. 0 S D Years of education -2 .0 S D -1 .0 S D M ea n + 1. 0 S D + 2. 0 S D Years of education Attachment to Europe Self-identified European Positive image of the EU Knows her/his EU rights Objective knowledge on EU E ff ec t o f tr an sn at io na l a ct iv it ie s E ff ec t o f tr an sn at io na l a ct iv it ie s Figure 4. Estimated effects (and 95% confidence intervals) of transnational activities in pro-European sentiments at different levels of years of education and foreign experiences ¼ 0. Fernández et al. 19 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ effect at the mean value of education. For this level, these activities have a positive and significant effect in four of the dependent variables. More importantly, TAs education years is negative and significant in four models. This is already strong evidence that education moderates the influence of these activities. To provide a clearer depiction, Figure 4 displays the coefficient and confidence intervals of TAs at different, standardized levels of education years and for immo- bile individuals (foreign experience¼ 0). This figure shows that the education level has a strong moderating influence on the impact of TAs on four dependent vari- ables: attachment to Europe; self-identified as European; knows her EU rights and objective knowledge on the EU. The coefficient of TAs peaks, in all cases, at the lowest levels of education and then declines linearly with education. The second hypothesis is therefore confirmed. The evidence supports the expectation that activities of domestic transnationalism are particularly influential in opinion for- mation among citizens with lower levels of education. Robustness checks We conduct a series of robustness checks to examine the stability of the results. First, given that TAs includes second language use in two components, and several European societies have two or more, co-official languages, second language may refer to another domestic language. We therefore replicate the models excluding countries with co-official languages.6 Second, although the indexes TAs and foreign experience were constructed summing unequally common individual activities, these indexes can also be constructed giving the same weight to all individual activities. These robustness checks produce equivalent results. Dropping officially multilin- gual countries and using principal component factors to construct the transnational indexes, the variable TAs remains mostly significant. TAs is positive and significant in eight of the 10 additional models replicating Table 2 – i.e. all but two regarding objective knowledge, which was already not significant in Table 2. Moreover, TAs education years has a significant moderating effect of TAs in eight of the 10 models – i.e. all but two regarding a positive image (for details, see the online appendix). Discussion This article contributes to the transactionalist approach to individual, European- related sentiments by disaggregating the notion of TAs and emphasizing the role of domestic-based transnational actions – i.e. domestic transnationalism. Based on Deutsch’s work, a growing literature argues that international transactions foster interest-based and affective bonds across European countries (Fligstein, 2008; Kuhn, 2011, 2012, 2015; Mau and Mewes, 2012) – usually positing international travel as the ideal form of transnationalism. Our main argument, however, is that experiences of transnationalism – e.g. interactions with migrant EU-citizens, or foreign media consumption – do not 20 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ require physical border crossings to be influential. This is because TAs that occur in the familiar settings of one’s home country can also bolster the communicative complementarities with distant groups that undergird transnational identification. Domestic transnationalism may be, in fact, particularly influential in forming beliefs because these activities benefit from more recurrence than do border crossings. We further argue that domestic-based activities of transnationalism are particu- larly influential as the level of education declines. The less educated have, in gen- eral, less coherent evaluative structures, which makes recent activities more relevant in their opinion-formation processes. Additionally, among lower levels of education, DBTAs cannot be balanced out by foreign experiences, which means for such citizens that domestic experiences are more cognitively relevant than border-crossings. The analysis of Eurobarometer 80.1 confirms these expectations. Engagement with TAs, occurring in the home country, is positively and significantly associated with four of the five, considered indicators of the affective, cognitive and evaluative dimensions. The probability of having pro-European sentiments is therefore higher among citizens that engage in domestic transnationalism. This evidence is consis- tent with our theoretical argument that DBTAs facilitate the endorsement of the European project. While domestic transnationalism is related to pro-European beliefs, the level of education also affects it. In particular, the impact of the domestic transnationalism index peaks at low levels of education and, in fact, becomes non-significant among the most educated citizens. Such evidence provides strong support for our argu- ment that domestic transnationalism can have a particularly formative influence among the less educated. This is extremely significant for a comprehensive under- standing of the dynamics driving the European integration process – a process often criticized as being only for European elites and ignored by less-educated Europeans. Our results clearly show that the development of pro-European atti- tudes is not exclusively linked to high individual investments in transnational mobility, reserved only for a small European elite who possess the necessary social and economic capital to participate, but that domestic TAs foster pro- European sentiments – especially among citizens with lower levels of education. One caveat bears mentioning. The cross-sectional nature of Eurobarometers does not allow us to eliminate the risk of endogeneity. Like all previous, survey- based research regarding the link between transnational practices and attitudes, we cannot discount the possibility that pro-European attitudes could also partially determine TAs. In our view, however, the endogeneity risk is rather small, espe- cially concerning the moderation effect of education on the influence of DBTAs. With strong reasons to believe that education moderates the impact of domestic transnationalism, we have no grounds to believe – following the inverse logic – that these attitudes moderate the effect of education in the number of TAs. That said, research on European attitudes and identity faces a major challenge in specifying the causal relationships between transnational practices and public sentiment Fernández et al. 21 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ towards the European project. Following decades of research, based mainly on cross- sectional data that cannot fully separate explanans from explanandums, our understand- ing of popular views towards Europe would be particularly improved through panel data studies and clever experimental designs that elucidate the net influence of certain actions on attitudes formation (e.g. Sigalas, 2010; Cram et al., 2011). Our results also have noteworthy policy implications. The autonomous role of domestic transnationalism, theorized above, illuminates the less spectacular bene- fits of European citizenship – namely, facilitated cross-national communications and trade. While perhaps less dramatic than freedom of mobility, they can still benefit a larger pool of citizens who, indeed, do not usually stay abroad. Since the European Commission has historically concentrated most of its resources in pro- moting physical mobility, like the Erasmus program, with rather limited effects (Kuhn, 2012; Sigalas, 2010), it may be time to consider other strategies. Should further research confirm that domestic transnationalism stimulates a sense of European commonness, its practices could be promoted (e.g. through a more integrated European media market and inexpensive, cross-national communica- tions) as a viable and maybe cost-saving option for the European Union. Acknowledgements Juan Dı́ez-Medrano, Antonio Jaime-Castillo, Theresa Kuhn and Klarissa Lueg gave exten- sive comments on earlier drafts. Richard Frederick edited the manuscript. Funding The author(s) disclosed receipt of the following financial support for the research, author- ship, and/or publication of this article: The authors thank the German Volkswagen Foundation (Grant No: 88-260 2) for their generous support. Notes 1. This also translates to the otherwise very informative works on the determinants of transnational engagement itself (Gerhards and Hans, 2014; Mau and Mewes, 2012), which do not discuss variations in the relative importance of practices like living abroad or having foreign language skills. 2. With transnational omnivorousness, we adapt the concept of cultural omnivorousness defined as ‘simultaneous appreciation of both ‘‘highbrow’’ and ‘‘lowbrow’’ genres’ (Prieur and Savage, 2013: 255) and refer to persons who engage in a wide range of transnational activities that can be described as both transnational and domestic. 3. The correlation between education and transnational activities is positive but not very high (r¼ .329, p < .05). 4. In an alternative index of objective knowledge that exclude the most temporary element of the total number of current member states, transnational activities is positive and signifi- cant (online appendix). 5. We also estimated additional models with an index of pro-European sentiments that adds all five dependent variables and has a 0–5 range. Using both ordered logit and a regres- sion model, this evidence is consistent with Table 2. Transnational activities remains 22 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ positive and significant in both cases and transnational activities education years is negative and significant in both cases, too (for detains see the online appendix). 6. We refer to countries with co-official languages spoken by at least 5% of the total population: Austria, Belgium, Finland, Ireland, Luxembourg, Malta and Spain (Central Intelligence Agency, 2014). References Anderson B (2006) Imagined Communities: Reflections on the Origin and Spread of Nationalism. Revised Version. London: Verso. Allport GW (1979) The Nature of Prejudice. Reading: Addison-Wesley. Armingeon K and Ceka B (2014) The loss of trust in the European Union during the great recession since 2007: The role of heuristics from the national political system. European Union Politics 15: 82–107. Augoustinos M andWalker I (1995) Social Cognition: An Integrated Introduction. London: Sage. Best H, Lengyel G and Verzichelli L (eds) (2012) The Europe of Elites: A Study into the Europeanness of Europe’s Political and Economic Elites. Oxford: Oxford University Press. Boomgaarden HG, Schuck ART, Elenbaas M, et al. (2011) Mapping EU attitudes: Conceptual and empirical dimensions of euroscepticism and EU support. European Union Politics 12(2): 241–266. Bourdieu P (1984) Distinction. Harvard: Harvard University Press. Bruter M (2005) Citizens of Europe? The Emergence of a Mass European Identity. Basingstoke: Palgrave Macmillan. Bryson B (1996) ‘‘Anything but heavy metal’’: Symbolic exclusion and musical dislikes. American Sociological Review 61: 884–899. Carey S (2002) Undivided loyalties: Is national identity an obstacle to European integration? European Union Politics 3(4): 387–413. Central Intelligence Agency (CIA) (2014) The World Factbook. Washington: CIA. Checkel JT and Katzenstein PJ (2008) European Identity. Cambridge: Cambridge University Press. Collingwood L (2012) Levels of education and support for direct democracy. American Politics Research 40(4): 571–602. Cram L, Patrikios S and Mitchell J (2011) Interacting identities: A comparative experimen- tal study of the impact of existing and emerging identity(ies) in the EU on attitudes to integration. In: ECPR conference, Reykjavik, August 2011. Deutsch KW, Burrell SA, Kann RA, et al. (1957) Political Community and the North Atlantic Area: International Organization in the Light of Historical Experience. Westport: Greenwood Press. Deutsch KW (1952) Nationalism and Social Communication: An Inquiry into the Foundations of Nationality. Cambridge: M.I.T. Press. Dı́ez Medrano J (2008) The public sphere and the European union’s political identity. In: Checkel JT and Katzenstein PZ (eds) European Identity. Cambridge: Cambridge University Press, pp. 81–111. Dı́ez Medrano J and Gutiérrez P (2001) Nested identities: National and European identity in Spain. Ethnic and Racial Studies 24: 753–778. Favell A and Recchi E (2009) Pioneers of European integration: An introduction. In: Recchi E and Favell A (eds) Pioneers of European Integration. Citizenship and Mobility in the EU. Cheltenham: Elgar, pp. 1–25. Fernández et al. 23 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Favell A, Recchi E, Kuhn T, et al. (2011) The Europeanisation of Everyday Life: Cross- Border Practices and Transnational Identifications Among EU and Third-Country Citizens – State of the Art Report. EUCROSS Working Paper 1. Fligstein N (2008) Euroclash. The EU, European Identity and the Future of Europe. Oxford: Oxford University Press. Gabel M (1998) Interests and Integration: Market Liberalization, Public Opinion, and European Union. Ann Arbor: University of Chicago Press. Garcı́a-Álvarez E, Katz-Gerro T and López-Sintas J (2007) Deconstructing cultural omni- vorousness 1982–2002: Heterology in Americans’ musical preferences. Social Forces 86(2): 417–443. Gerhards J and Hans S (2014) Explaining citizens’ participation in a transnational European public sphere. Comparative Sociology 13(6): 667–691. Hakhverdian A, van Elsas E, van der Brug W, et al. (2013) Euroscepticism and education: A longitudinal study of 12 EU member states, 1973–2010. European Union Politics 14(4): 522–541. Immerfall S, Boehnke K and Baier D (2010) Identity. In: Immerfall S and Therborn G (eds) Handbook of European Societies: Social Transformation in the 21st Century. New York: Springer, pp. 325–353. Inglehart R (1970) Public opinion and regional integration. International Organization 24(4): 764–795. Kuhn T (2011) Individual transnationalism, globalisation and euroscepticism: An empirical test of Deutsch’s transactionalist theory. European Journal of Political Research 50: 811–837. Kuhn T (2012) Why education exchange programmes miss their mark: Cross-border mobi- lity, education and European identity. Journal of Common Market Studies 50(6): 994–1010. Kuhn T (2015) Experiencing European Integration: Transnational Lives and European Identity. Oxford: Oxford University Press. Long SJ and Freese J (2006) Regression Models for Categorical Dependent Variables Using Stata. College Station: Stata Press. Lubbers M and Scheepers P (2007) Explanations of political Euro-scepticism at the indivi- dual, regional and national levels. European Societies 9(4): 643–669. Marks G and Hooghe L (2003) National identity and support for the European Union. WZB Discussion Paper, SP IV 2003-202. Mau S (2005) Europe from the bottom: Assessing personal gains and losses and its effects on EU support. Journal of Public Policy 25(3): 289–311. Mau S and Mewes J (2012) Horizontal Europeanisation in contextual perspective: What drives cross-border activities within the European Union? European Societies 14(1): 7–34. McLaren L (2006) Identity, Interests and Attitudes to European Integration. Basingstoke: Palgrave. Nelsen BF and Guth JL (2000) Exploring the gender gap women, men and public attitudes toward European Integration. European Union Politics 1: 267–291. Nelsen B and Guth JL (2015) Religion and the Struggle for European Union: Confessional Culture and the Limits of Integration. Washington, DC: Georgetown University Press. Nickerson DW (2007) Quality is job done: Professional and volunteer voter mobilization calls. American Journal of Political Science 51(2): 269–282. Norris P and Inglehart R (2009) Cosmopolitan Communications: Cultural Diversity in a Globalized World. Cambridge: Cambridge University Press. 24 European Union Politics 0(0) at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/ Parsons T (1951) The Social System. London: Routledge. Prieur A and Savage M (2013) Emerging forms of cultural capital. European Societies 15: 246–267. Quillian L (1995) Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review 60: 586–611. Recchi E (2014) Pathways to European identity formation: A tale of two models. Innovation: The European Journal of Social Science Research 27: 119–133. Rother N and Nebe T (2009) More mobile, more European? Free movement and EU identity. In: Recchi E and Favell A (eds) Pioneers of European Integration: Citizenship and Mobility in the EU. Cheltenham: Elgar, pp. 120–155. Roudometof V (2005) Transnationalism, cosmopolitanism and glocalization. Current Sociology 53: 113–135. Scharpf F (1999) Governing in Europe: Effective and Democratic? Oxford: Oxford University Press. Scheuer A (2005) How Europeans see Europe. Amsterdam: Amsterdam University Press. Schneider SL (2008) Anti-immigrant attitudes in Europe: Outgroup size and perceived ethnic threat. European Sociological Review 24(1): 53–67. Schuman H and Presser S (1996) Questions & Answers in Attitude Surveys. Thousand Oaks: Sage. Sigalas E (2010) Cross-border mobility and European identity: The effectiveness of inter- group contact during the ERASMUS year abroad. European Union Politics 11(2): 241–265. van Schuur WH (2003) Mokken scale analysis: Between the Guttman scale and parametric item response theory. Political Analysis 11: 139–163. Warde A and Gayo-Cal M (2009) The anatomy of cultural omnivorousness: The case of the United Kingdom. Poetics 37: 119–145. Zaller J (1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press. Fernández et al. 25 at FU BerlinLateinamerikanisches on April 21, 2016eup.sagepub.comDownloaded from http://eup.sagepub.com/
-
File type
- application/pdf
-
Referenced at
-