Measuring the Europeanization of Everyday Life: Three New Indices and an Empirical Application

For decades, comparative sociology has analyzed European societies pluralistically by focusing on similarities and differences between countries. Only recently has there been growing interest in a European society in the singular sense – the idea that Europe may form an integrated entity not only in political terms (the European Union [EU]), but also in social terms. Two large schools of thought have emerged. The first school is inspired by Stein Rokkan's comparative research into nation-building (Rokkan 1999), and investigates the process of Europe's political and administrative center formation (Bartolini 2005; Ferrera 2005). The second school emphasizes the network character of European society, inspired by Karl Deutsch's transactional integration theory and its emphasis on cross-border exchange and communication (Deutsch 1966). Recently, the transactional paradigm has been rediscovered by sociologists. Michael Mann (1998) has defined society in terms of overlapping social interaction networks, asking the provocative question: ‘Is there a society called Euro?’ Others have taken the transactional approach to the micro-level of ‘social transnationalism’ and ‘horizontal Europeanization’, focusing on the transnational practices and attitudes of ordinary citizens (Fligstein 2008; Mau 2010; Mau and Verwiebe 2010; Favell and Guiraudon 2011; Favell et al. 2011; Recchi 2012).

Despite the growing popularity of this approach, an accurate empirical assessment of the Europeanization of everyday life is still largely missing. Scholars have mainly concentrated on absolute levels of activity in Europe (e.g., Diez Medrano 2008; Mau and Mewes 2012; Risse 2010) or on the strength of these European activity levels relative to national ones (e.g., Bruter 2005). What is missing in most accounts, however, is an assessment of the salience of the European reference frame vis-à-vis the global reference frame. Building on Mann (1998) and others (e.g., Gerhards and Rössel 1999), we argue that the latter is needed in order to arrive at a more elaborate understanding of Europeanization. A twofold relative perspective is essential because we cannot investigate Europeanization in isolation from globalization (cf. Delanty and Rumford 2005) – rather, we have to investigate it relative to the world and the nation at the same time.

In this article we sketch out a refined relative approach to European society and illustrate it empirically. Conceptually, our main contribution is to systematically integrate ‘external closure’ into the concept of Europeanization. We further suggest three new indices that capture relative Europeanization and its major components. Using large-scale survey data from EU citizens (mainly from the Eurobarometer [EB]) we empirically show which life domains are most Europeanized, and the difference the new relative approach makes, compared to traditional accounts of Europeanization.

The article is organized as follows: Section 2 introduces the transactional paradigm and its contributions to the sociology of Europe. Section 3 explains our concept of relative Europeanization. The data and methods are described in Section 4, while Section 5 presents the empirical evidence. Limitations and implications of our findings for further research are discussed in Section 6.

The theory of transactional integration originated in international relations in the 1950s and 1960s with Karl W. Deutsch being its most prominent proponent. To Deutsch, the road to peace was to intensify cross-border communication. He deemed economic, cultural, and social exchanges between nations to have an educational effect, leading to a growing sense of community among the elites and the masses, and to the creation of common institutions. In an upward spiral, formerly independent nation-states merge into larger political entities. Deutsch's theory has proven itself fruitful for the sociology of Europe in three major ways.

First, the theory offers a clear idea under which conditions Europe becomes an important object of identification, viz. when citizens make ‘European’ experiences. A number of recent studies have investigated this causal link at the micro-level (Fligstein 2008; Kuhn 2011; Mau 2010) or at the macro-level (Delhey 2007).

Second, the theory has proposed that there is not one single, closed society (the nation-state society), but rather a complex constellation of multiple transaction networks which differ in geographical scope and density. In a similar vein, Michael Mann has argued that:

Third, transactional theory opens up the possibility to map the European society. This mapping may involve transactions on the one hand, from cross-border trade and investment, migration flows, student exchange, and travel to information exchange, news coverage, family ties, and friendships. On the other hand, it can also include various expressions of sense of community, such as territorial identification, attachment, or solidarity. For the purpose of this article, the second and third sociological applications of Deutsch's framework are highly relevant.

Cross-border transactions and people's sense of community are not only central to the sociology of Europe, but also for research on transnationalization (Mau 2010), globalization (Held et al. 1999), and cosmopolitanism (Beck and Grande 2007). Whereas all these processes have a lot in common, it is important to differentiate between them. Transnationalization can be viewed as the general process of reaching out beyond national borders, whereas Europeanization and globalization specify the territorial reference frames (cf. De Swaan 1995; Gerhards and Rössel 1999). In line with Michael Mann, we argue that so far Europeanization research has not done enough to differentiate between various ‘networks of interaction’.

To systematically elaborate a more refined concept of Europeanization, we first revisit the traditional approach via national openness (Section 3.1), then discuss the (largely) missing dimension of external closure (Section 3.2), and finally develop a new concept of relative Europeanization via three new indices (Section 3.3).

In order to map the European society from a transactional perspective, contemporary scholars take Deutsch chiefly as an inspiration for what to look at: transactions and sense of community. This is often done with an eye on absolute numbers (e.g., Mau and Mewes 2012). Yet for Deutsch, such absolute numbers (‘volumes’) were just the first step before transforming them into relative figures (‘ratios’) that compare external and internal exchanges: ‘[t]he proportions of such transactions which cross the boundaries of the country, as against those that remain entirely within it’ (Deutsch 1960: 148).

To get an idea about how relevant Europe is as a reference frame for transactions and sense of community, one needs a standard for comparison, and the obvious benchmark is the nation-state (cf. Immerfall 2000; Vobruba 2008). European society gains strength through the opening of the national ‘container’ (Beck and Grande 2007) toward Europe. The stronger the European reference frame becomes, the less important, in relative terms, the national one gets. This key idea of Europeanization as de-nationalization has been applied in several studies, e.g., on identity (Bruter 2005). With good reason, ‘European versus national’ has been described as the axial principle out of which the entire sociology of Europe has developed (Vobruba 2008).

Karl Deutsch's chief concern was with internationalization, for which tracking national and international transactions was sufficient. This is different for Europeanization (as for any other continent-wide integration process) with its position ‘sandwiched’ between the nation-state and world society. Europeanization does not take place in isolation, but is isochronic with globalization (Delanty and Rumford 2005). Hence, the question is whether the societies in Europe are indeed becoming Europeanized, or rather globalized. This is precisely what Mann (1998) referred to as the issue of ‘external closure’ – a closure in a mere statistical sense, not as a strategic decision to build a ‘fortress Europe’. From a transactional perspective, the very idea of a European society in the singular requires Europeans to create more European than extra-European experiences, and to be cognitively and emotionally more attached to Europe than to the world in general. This reintroduces Deutsch's internal/external schema, but this time elevated to another level; now Europe is ‘internal’ and the world outside is ‘external’. As a consequence, a second axial principle, European versus global, is required to capture the degree of Europeanization.

The literature has to a large extent ignored the need to investigate EC, with some notable exceptions. For Germany, Gerhards and Rössel (1999) investigated three societal subsystems, namely science, art, and the economy, and their data point to Europeanization rather than globalization. Subsequent studies have drawn a similar conclusion for the economy (e.g., Fligstein and Merand 2002; Kim and Shin 2002) and science (Heilbron 2013; Smeby and Trondal 2003). Popular culture might be a different story, since a European popular culture appears to develop only via the detour of Americanization (Therborn 1995). Systems theorists have tended to dismiss the European closure of Europe, arguing that any modern communication is potentially global (Stichweh 2000). Similarly, it has been speculated that ‘Europe as a focus of attachment is increasingly pressured by what lies behind: by a global or universal human identity’ (Kohli 2000: 123).

To our knowledge, however, none of the key publications on social transnationalism (Favell 2008; Fligstein 2008; Mau 2010; Recchi and Favell 2009) has addressed the issue of European closure systematically. Often, the data used do not address the geographic scope of Europeans' cross-border transactions, making it unclear to what extent it actually sheds light on Europeanization as opposed to globalization. One recent qualitative study (Andreotti and Le Galès 2011) has touched upon this issue more successfully, pointing to main differences between practices and identities: whereas the mobility trajectories and social networks of Europeans featured a distinct European focus, they were nevertheless found to also feel part of world society on the attitudinal level.

To summarize, we have brought forward the argument that the Europeanization of everyday life should be defined in relative terms: how relevant the European reference frame for Europeans is as compared to (1) the national reference frame (indicating the degree of national openness) and (2) the global reference frame (indicating Europe's external closure). This twofold relative concept results in (3) a novel definition of Europeanization as the concurrence of openness and closure, i.e., relative Europeanization (see Figure 1). It is possible to apply this approach to both survey-based data collected at the micro-level and process-generated, macro-level data (in the tradition of Mann and Deutsch).

In what follows we apply our novel approach empirically to EU citizens’ everyday life and corroborate it vis-à-vis traditional approaches. Specifically, we seek to explore the following research questions:

1. Which practices and attitudes of the EU citizens are most Europeanized?

2. Does external closure provide additional information independent of the degree of national openness?

3. Does relative Europeanization provide additional information independent of the degree of absolute Europeanization?

From these surveys, we selected eight items that cover both practices and attitudes of EU citizens and that shed light on the three reference frames most relevant to the study of Europeanization: the nation, Europe, and the world (see Table 1). The original items were recoded and aggregated to denote macro-level agreement rates, which for practices represent the proportion of the sample having done a particular activity and for attitudes the percentage acquiescing to the items (cf. Appendix for further details). The EU-25 means were computed giving equal weight to each country. Computing the population-weighted EU-25 average did not change our results. While the item selection has been determined by data availability, the eight domains covered lie at the heart of social transnationalism research (cf. Hanquinet and Savage 2011).

Nevertheless, there are several potential drawbacks inherent to the item selection, which are addressed in the following.

The first issue is data availability. For two of the items, travel and family, data on the national reference frame is lacking in the original surveys. This was resolved as follows: for travel, information on domestic trips is taken from Eurostat (2008). Information is provided for 2006, the same year in which the respective EB (66.1) has been conducted.¹ For family, it was not possible to unearth an alternative data source. So as to not lose this important dimension, we settled on a guesstimate which assumes that national citizens, but not first-generation immigrants, have a family member living in the country of residence. The low regional immobility of Europeans (Favell and Recchi 2011) strongly suggests that most citizens’ family members indeed do live in the same country. Following this approach yields guesstimates between 74% (Luxembourg) and 99% (Hungary) of the population having a family member on the national level. The EU-25 mean is 94%. Wherever this guesstimate is involved, results are marked with an asterisk (*) to remind us of this special circumstance in the data. Fortunately, the surveys provide full information on at least the European and global reference frames across all items. This allows for an accurate measurement of the EC, which constitutes the major innovation of this article after all.

One of the surveys employed provides no information on the citizens of Bulgaria and Romania. In order to homogenize the findings and facilitate comparisons across the different items under consideration, we focus exclusively on the populations of the EU-25 member states. We checked whether including these two countries with those items where data are available would substantially change results and found this not to be the case.

A third potential shortcoming is the fact that the data originate from different sources and years. This is less problematic than it may seem at first glance, since the surveys have been conducted within the rather narrow range of three years (2005–2008) and it appears unlikely that outcomes at the aggregate level would fluctuate wildly from one year to another. It is therefore plausible to assume that any patterns we discover stem for the utmost part from substantive differences between the various practices and attitudes under scrutiny.

A fourth data limitation concerns the inconsistency of labels used for the three reference frames in the original survey items (see Appendix for details). Most importantly, whereas some question wordings refer to Europe, others refer to the EU. This introduces a potential margin of error, since some of the non-EU European experiences (e.g., those involving Switzerland) might be classified as non-European (and hence global) experiences when the question inquires after the EU (as opposed to Europe). A possible consequence is a slight underestimation of Europe's salience, should respondents have mobility, purchasing, and travel experiences solely with non-EU European countries. For reasons outlined along two exemplary domains, we believe that this potential underestimation should turn out rather small. As for the mobility item, in 2008 two million EU citizens immigrated to an EU-27 country, compared to just 159,000 immigrating to a non-EU European country (Eurostat 2011). In the case of travel, in 2006 for all EU countries except Lithuania and Slovenia, the three top outbound destinations were other EU countries (Eurostat 2008). Thus, the intra-EU mobility captures the lion's share of total European mobility, and the same holds true for travel.

Finally, our indices are designed to meter Europeanization for aggregates of people. No inferences can be made about individuals and their personal life practices and attitudes. Consequently, similar index values might have been generated by very different social patterns.² This, however, is a property many aggregate indicators share (e.g., average income).

Bearing these potential shortcomings in mind, we proceed to apply the novel conceptual approach to Europeanization to the selected practices and attitudes of EU citizens.

We first analyze levels of national openness, external closure, absolute Europeanization, and relative Europeanization across the eight domains (Section 5.1) and then compare them across the EU-25 countries for two exemplary domains (mobility and identity, Section 5.2), in order to answer the three research questions outlined above.

Following the order of the conceptual part (Section 3.3), we first describe the national openness and external closure patterns across the eight domains under study and then turn to relative Europeanization and its relation to absolute Europeanization. Figure 2A shows the levels of NO (displayed on the x-axis) and EC (displayed on the y-axis) across domains. It reveals substantial differences between domains concerning both dimensions. NO is especially low for mobility (.06), purchase (.18), and family* (.24), and high for travel (.61). For the three attitudes, identity (.44), solidarity (.45), and cultural interest (.47), Europe is almost as salient as the nation.

EC, on the other hand, is particularly high for purchase (.73), travel (.69), and mobility intention (.67). For these domains, the European reference frame is more relevant than the global one. This is different for attitudes. For three of the attitudinal domains, EC scores are around the point of equality of .5; as many EU citizens identify as world citizens as with Europe (.50), little preference is given to European over global culture (.51), and solidarity with all humans around the world is slightly more widespread than solidarity with fellow Europeans (.46).

Taking national openness and external closure together leads to further insights. Whereas in the case of NO all domains except for travel lie below the point of equality, for EC all domains except for solidarity are situated above it. Travel is exceptional; it is the only domain where the European reference frame is stronger than both the national and the global ones. The other domains form two distinct clusters: one comprises the three attitudinal domains, cultural interest, identity, and solidarity, which are relatively de-nationalized, but lack a clear cognitive and/or emotional boundary toward the world. The other cluster covers the domains mobility, family*,³mobility intention, and purchasing, which take place mostly nationally, but for which Europeans prefer Europe to the world.

Figure 2B illustrates the degrees of relative and absolute Europeanization in the EU-25 aggregate for each domain. In terms of RE, travel is far ahead; almost half (.48) of all travels take place on the European level. The four attitudinal domains follow neck and neck, with RE scores around .3. Then come family* (.21), purchasing (.17), and lastly mobility (.06), which are hardly Europeanized at all. When contrasted with the maximum score of 1, RE appears to still be low across most domains studied.

As Figure 2B shows, the AE scores are very similar to the RE scores for four of the domains (mobility, purchase, mobility intention, and travel), while they deviate across the other four (where RE scores are lower than AE scores). The difference is especially large for identity, where we find an RE score of .30 and an AE score of .68. In general, empirical divergences between the two indices occur in those domains where the global reference frame has substantial weight. Where this is the case, AE scores tend to overestimate the salience of Europe. The RE index proposed here can therefore be deemed the more accurate measure of Europeanization.

To further corroborate the added value of our more laborious approach to Europeanization and to address research questions (2) and (3), we assess the relations between NO and EC, and between AE and RE across the EU-25 member states with respect to one exemplary practice (mobility) and one exemplary attitude (identity).

We first concentrate on the advantages of adding the concept of external closure to national openness (Figure 3A). Looking at mobility (depicted by triangles), it is immediately apparent that EC adds extra information on the Europeanization of this domain. If we were to only rely on NO, we would see that the populations of the EU-25 member states are not very Europeanized, with NO scores scattered closely between .01 (Estonia) and .21 (Ireland). However, we would miss out on the evidence of the EC dimension, which reveals that in nearly all member states considered, citizens prefer Europe over the rest of the world when moving abroad (scores above the .5 point of equality for all countries but Lithuania, Estonia, Latvia, France, and Malta). In other words, the transnational fraction of mobility is substantially Europeanized. The graph also shows that the EU populations are heavily scattered along the EC Index, ranging from the least externally closed population of Lithuania (EC score of .35) to the most externally closed population in Slovakia (.89).

The identity domain (depicted by circles in Figure 3A) displays a similar pattern; whereas the European countries are close together on the NO-axis (from .35 in the UK to .52 in Luxemburg), the EC-axis reveals a wide dispersion between countries (from .43 in the UK and Cyprus to .70 in the Czech Republic). These different degrees of closure in the identity domain underline that the EC Index does provide insights into the Europeanization of everyday life independent from the degree of national openness. Figure 3A also reveals that most countries find themselves in the upper left square of the graph, indicating that a relatively low degree of NO and a high degree of EC prevail for both domains.

Figure 3B shows the country scores on the RE Index as contrasted with the AE values for the two example domains. This comparison discloses at least three empirical differences.

First, AE scores need not always be higher (or lower) than the respective RE scores. Whereas for mobility (depicted as triangles), AE scores are lower than the respective RE scores in all 25 countries, for identity (depicted as circles), the opposite is the case; all AE scores are higher than the respective RE scores. This observation is important, as the domain comparison (Section 5.1) seemed to show that AE scores often tend to be higher than RE scores.

Second, the dispersion of AE and RE scores can differ, meaning that the range of the RE scores can be both larger and smaller than that of the AE scores; whereas for mobility, the range of the RE values (.19–.01 = .18) is bigger than that of the AE values (.15–.01 = .14), the reverse is true for identity, where the range of the RE values (.38–.24 = .14) is much smaller than that of the AE values (.90–.49 = .41). In other words, with an absolute approach we sometimes overestimate country differences in the degree of Europeanization and sometimes we underestimate them, depending on the domain in question.

Third, the rank order of EU-25 member states with respect to Europeanization can change both across and within domains. Consider, for instance, Sweden's different rank positions shown in Figure 3B; its AE ranks are 3rd for identity and 4th for mobility, while the respective RE ranks are just 10th for identity and 8th for mobility.

These three discrepancies show that absolute and relative Europeanization are different concepts of Europeanization. In fact, even where AE scores are high, Europe can still be trailing behind in terms of salience relative to the nation and the world taken together (as is the case, e.g., with the identity domain). High AE scores can go together with low RE scores and vice versa.

In this paper, we have argued that the principal difference between national and European – the main currency of Europeanization research so far – needs to be complemented by a second principal difference, between Europe and the world. Incorporating external closure into the definition of Europeanization allows assessing more accurately the extent to which the everyday life of Europeans is Europeanized. Based on this conceptualization, we have proposed three concrete indices for metering Europeanization: the NO Index, the EC Index, and the RE Index. We have tested these novel indices on large-scale survey data, which effectively allowed us to map the degree of Europeanization of eight practices and attitudes across the EU-25 aggregate. Moreover, we contrasted the findings of our relative measures with results of the traditional absolute approach on the country level to highlight the added value of our contribution. Altogether, this generated two central new insights:

1. Life domains are Europeanized to different degrees: for most of the practices investigated, the national reference frame remains dominant compared to the European one (low NO), while the global reference frame is particularly weak (high EC). In contrast, across most of the attitudes studied, the nation, Europe, and the world proved more or less equally salient. Thus, practices and attitudes formed two distinct clusters. Travel featured exceptionally high levels of NO and EC.

2. EC constitutes a distinct component of RE, adding to our understanding of Europeanization beyond the contribution of NO. Further, our RE measure leads to different outcome patterns than AE, thereby providing valuable additional information as well.

Both insights warrant some further discussion; concerning (1), the divergence between practices and attitudes, the differing role of resources and opportunities may be important. What humans do depends to a large extent on available resources and opportunity structures. With higher income, for instance, practices beyond national borders increase (Mau and Mewes 2012). Geographic distance can be expected to be a central determinant of many activities, most obviously for traveling. Also, the EU's active facilitation of engagement in European practices via cultural exchange initiatives or its internal market may possibly be a driving factor behind the pronounced external closure of the investigated practices. Attitudes by contrast are hardly dependent on resources, infrastructure, and opportunities – there is no cost involved in identifying as a world citizen. The RE patterns of attitudes can consequently be interpreted as adhering to what cosmopolitans have described as the ‘both/and principle’ (Beck and Grande 2007: 14); people do not have to decide between Europe and the world, but can go for both. As mentioned in Section 3.2, similar differences between practices and attitudes have been reported in the qualitative study by Andreotti and Le Galès (2011).

Concerning (2), the added value of our approach, we would like to mention the modification potential inherent in the proposed indices in future research. While our results have shown the benefits of incorporating EC into the RE concept, one could argue that pitting Europe against the entire world has set the bar too high. Instead, it might be sufficient to test the salience of the European reference frame relative to single other world regions. For example, do Europeans travel more within Europe than between Europe and North America? Are Europeans more interested in European than in African culture? Deliberately, the indices introduced are flexible and can be adapted to any reference frame comparison of interest. If desired or needed, additional geographic reference frames such as the local region of residence, the Organisation for Economic Co-operation and Development (Gerhards and Rössel 1999), or ‘the North’ (Mann 1998: 187) could then be added as well. We further want to stress that our framework can be applied to other world regions, too, in order to measure processes of ‘Africanization’, ‘Asianization’, and so on. Finally, monitoring the development of RE scores over time in future research would add the much-needed time dimension.

To conclude, we think that the twofold relative approach proposed here allows for a more adequate measurement of Europeanization than standard absolute and relative (nation versus Europe only) conceptualizations. By positioning Europe between the nation-state and the world, it manages to incorporate effects of both de-nationalization and globalization. As everyday life today takes place at all conceivable social and geographic distances, approaching Europeanization from these two perspectives appears pivotal.

We are very grateful to Jan Lorenz for his advice on the mathematical properties of Europeanization indices and to two anonymous reviewers for their helpful comments. All remaining errors are our own.

This research is part of the project ‘Cross-border Interactions and Transnational Identities’, which is supported by the German Science Foundation (DFG) within the framework of the DFG research unit FOR-1539 ‘Horizontal Europeanization’. For more information, see http://www.horizontal-europeanization.eu/en.

Jan Delhey is Professor of Sociology at Jacobs University Bremen, School of Humanities and Social Sciences (SHSS). He holds a diploma in Sociology from Otto-Friedrich-University Bamberg and received his doctoral degree from Free University Berlin. His research interests include cross-national differences in subjective well-being, trust and social cohesion, and sociological aspects of European integration. He is currently leading the project ‘Cross-border Interactions and Transnational Identities’, which is funded by the German Science Foundation (DFG) and part of the DFG research unit ‘Horizontal Europeanization’.

Emanuel Deutschmann is a Research Associate at Jacobs University Bremen and an Affiliated PhD Fellow at the Bremen International Graduate School of Social Sciences. He holds a BSc in Social Sciences from the University of Cologne and an MSc in Sociology from Nuffield College, University of Oxford. His research interests include social networks and transnational social interaction, with a regional focus on Africa, Europe, and Latin America.

Timo Graf is a regular PhD Fellow at the Bremen International Graduate School of Social Sciences. He holds a BA in Integrated Social Sciences from Jacobs University Bremen and an MA in Politics from the University of Cambridge. His research interests include civilizational identity and intergroup conflict; state failure; and China-EU relations.

Katharina Richter is a Research Associate at Jacobs University Bremen. She holds a BA in Integrated Social Sciences from Jacobs University Bremen and an M.Soc.Sci in Ethnic Relations, Cultural Diversity, and Integration/Sociology from the University of Helsinki. Her research interests include social transnationalism, transnational sense of community, and denationalization, especially with respect to the European integration process.