Confessional Cultures and European Identity: Religion, Ideology, and Economics

Abstract

Proponents of greater European political unity through the development of the European Union (EU) have long sought to foster a “European” identity among citizens as a way of advancing their cause. And there is now a substantial body of scholarship devoted to understanding the social, economic, and demographic factors contributing to the development of such an identity. Recently, there has been a growing interest in cultural influences, but the possible impact of religion has been largely ignored. Using Eurobarometer 65.2 (2006), we show that religious groups differ systematically in their propensity to take “European” identity markers. Using multivariate analysis, we demonstrate that Catholics are most likely to be cognitive “Europeans”, while Protestants and other Christians are less likely to take such perspectives. Religiosity tends to reinforce the dominant propensity of each tradition. These religious differences persist even under statistical controls for other demonstrated influences, although their effects are strongest in the old Western European “core” of the EU. The long-term decline of Catholic religiosity thus has important implications for the future of “European” identities: their growth will depend increasingly on less “diffuse” and more variable influences, such as successful economic management by EU and national governments.

1. Introduction

Proponents of European economic and political integration have always known that ordinary citizens would be crucial to the achievement of their ultimate objective: a new supranational identity, undergirding a united continent. National leaders might develop a deep mutual sense of community through constant interaction and meaningful exchange in European institutions, but if ordinary citizens did not feel that same sense of community, real European unity would not be possible. Economic, social, and political integration would only proceed as far as common identity permitted. Thus, the key to successful European integration has been the creation of a strong “European” identity.1

Over the history of the European Union and its predecessors, political leaders have sought to foster such a “European” identity. And beginning at the turn of this century, social scientists began to judge the success of that project and specify the economic, ideological, and demographic factors underpinning identity formation. The search for determinants of European identity has ranged very widely, but scholars have neglected one important factor: the influence of religion. Although historians have delineated the deep religious roots of the integration project, the burgeoning contemporary social science literature on European identity has either ignored entirely or minimized the contribution of religion. For example, none of the numerous, extensive and sophisticated analyses in Kaina et al. (2016) considers any religious variables. In partial contrast, Keulman and Koós’ (2014) comprehensive assessment does occasionally find traces of religion. They note in passing that religiosity seems to produce more “European” orientations (p. 156) and also observe that “Protestant countries are less likely to harbor European identities than Catholic and Orthodox countries are” (p. 281). But their basic assumption is that “religious affiliations … have mainly become irrelevant in Europe” (p. 166), obviating the need for systematic analysis. We think this is a major omission.

In the pages to follow we show the relevance of religion to the development of European identity. The first section reviews briefly the efforts by EU leaders to create a stronger sense of European identity among its publics. The next section considers the evidence for an evolving European identity, based largely on surveys of the early members of the EU. Although these studies usually ignore religion, they do identify other important factors, factors that will serve as control variables in our analysis. We then suggest that both the early history of the EU and research on public support for its institutions demonstrate the relevance of religion to European identity formation. With this background established, we move to the central task of testing the influence of religious variables, using Eurobarometer 65.2 (Spring 2006). We show that religious factors are not only evident at the bivariate level, but survive extensive controls for other influences more often emphasized in the literature. Finally, we discuss the implications of our findings for the ongoing European integration project.

2. Elite Construction of a European Identity

Although some analysts have argued that the EU does not require a “people” with a conscious “European” identity to produce democratic legitimacy (Weiler 1995), most academic experts and EU political leaders have recognized the need for a common identity strong enough to bind together peoples divided by language, nationality, social class, religion, gender, ethnicity, and other social markers. Indeed, many scholars argue that the EU must have the support of a “people” if it is to make difficult policy choices—exactly the kind of decisions it now faces (see Delanty 1995; Schmitter 2000; Siedentop 2001; Schmidt 2006; Maas 2007). And from the start of European integration in the 1950s, its proponents have sought to construct a new national identity, assuming that eventual union required the emergence of a “we feeling” among national publics (Deutsch et al. 1957).

Continental elites naturally led the European identity construction efforts. The British and the Danes (and later the Finns and Swedes) eventually signed on but did not consider the projects central to integration. For many continental politicians and bureaucrats, however, the building of a “People’s Europe” was indeed their main purpose. Without mass identification with the European Union, they could not push integration forward (Prisacariu 2007; Fornäs 2012). In December 1973 the Copenhagen European Summit (which included new members Britain, Denmark, and Ireland) issued a “Declaration on European Identity” (Declaration on European Identity, Copenhagen 1973). This remarkable document committed the Community to developing a common identity for projection to the world: “In their external relations, the Nine propose progressively to undertake the definition of their identity in relation to other countries or groups of countries”. This would reinforce the “construction of a United Europe thus making easier the proposed transformation of the whole complex of their relations into a European Union”. The Nine recognized that supranational institutions and common policies constituted “an essential part of the European Identity”, but the “fundamental elements” were representative democracy, the rule of law, social justice, and respect for human rights.

These values were hardly the sole property of Europe. What gave “the European Identity originality and dynamism” were cultural features: “a common European civilization, the attachment to common values and principles, the increasing convergence of attitudes to life, the awareness of having specific interests in common and the determination to take part in the construction of a United Europe” (all quotes are from the Declaration). This document did not commit the Nine to a specific strategy, but assumed that “the European Identity” was emerging from actions already taken. Further integration would solidify this identity, grounded in common values.

The 1980s “relaunch” of Europe, emphasizing a “People’s Europe”, focused attention once again on creating identity (Bruter 2005). The new European parliamentary elections in 1979 had already created a Community-wide political event and civic symbol. The European Council had also accepted in principle the idea of a common European passport. And even more was in the works. In June 1984, the Council established an ad hoc committee to assist in “adopting measures to strengthen and promote its identity and its image both for its citizens and for the rest of the world” (Adonnino 1985). Chaired by Pietro Adonnino, the committee proposed to strengthen Community citizens’ rights and ensure ease of movement across borders. It also sought to broaden and deepen cross-border exchanges through commercial transactions, audiovisual productions, tourism, sports, the pairing of cities, and, most important, education. The committee urged an increased emphasis on teaching languages, cross-border student and apprentice exchanges, and mutual recognition of university diplomas, all to encourage the mingling of Europeans and broadening of their concept of “us”.

Coupled with this indirect approach to European identity construction, the Adonnino committee also proposed symbols of nationhood to bond citizens emotionally to the Community and to each other, strengthening “the Community’s image and identity”. The report recommended adopting the Council of Europe’s flag as the Community’s, Beethoven’s “Ode to Joy” as its anthem, a series of European postal stamps, attractive uniform border signs, a new European Academy of Science, Technology, and Art, and even a “Euro-lottery”. In addition, the committee recommended that Europeans should receive more information on the history and benefits of the “construction of Europe”. Europe’s leaders did not take these proposals lightly. Starting in the mid-1980s, officials inside and outside European institutions tried to socialize citizens through education and the creation of new symbols.

Over the next two decades, integration leaders developed a broad range of programs and symbols designed to create more “Europeans”. Educational exchange programs, expanded language training, and “Europeanized” curricula across the EU sought to enhance a sense of common citizenship. Common official symbols made European unity visible to people across the EU: the flag, the EU anthem, a motto (“united in diversity”), a new currency (the “euro”), and Europe Day (May 9, commemorating the Schuman Declaration). Indeed, the “founders”—Schuman, Adenauer, Monnet, and de Gasperi—were raised to an almost mythical status in an effort to shape a common historical narrative for the newly united continent. Common citizenship (in the EU) and a common passport reminded Europeans of their shared identity. And these are only the most visible of the identity-creating efforts (Fornäs 2012).

3. European Identity: Empirical Perspectives

One putative result of all these elite efforts has been the addition of “European” to the multiple identities held by EU citizens. Whatever metaphors might be used to describe simultaneous identification with several communities, the emergence of a European identity was undeniable (Risse 2005, 2010; Hobolt 2004). What is less clear is the strength of this new identity and the factors contributing to its intensity, especially compared to national identity. As Thomas Risse has shown, some national identities are Europeanized by incorporating a European identity into the very content of national identity (Risse 2010). For instance, “a good German equals a good European, supporting a united Europe” (Risse 2005, p. 68; Banchoff 1999). But such melding is not apparent everywhere. Although German, French, Italian, and Luxembourgish identities may sit easily next to a European one, others (such as those in Britain and the Nordic region) seem to resist a new overlay. Some national identities may be weaker than others—Luxembourgers, for example, may not be as emotionally tied to their identity as the English—but some strong national identities, such as the French, seem quite compatible with a European one while others, such as the English, are not (Fuchs 2011). In short, European identity seems to be unevenly distributed across the EU.

A number of early studies captured the rise of a European identity, assessing the extent to which the efforts of EU elites had succeeded. The first wave drew on surveys from the 1990s and early 2000s, and focused on the “EU Nine” or the “EU 12”. One early analyst asked an elementary question: “How many Europeans are there? The short answer seems to be not many” (Green 2007, p. 52). But the response depended on what “European” meant and on how that identity was measured. Opinion surveys have tapped European identity using a confusing array of questions, alternative definitions of “European”, and many options. Some have defined “Europe” as a geographic unit; others have prompted respondents to think about their national (or other) identity and then ask whether they have the same feelings toward Europe (see Green 2007 for a review).

Another approach has been to ask citizens to compare their future attachments with their country and Europe (European Union 2004). Few queries specifically refer to the European Union, but most researchers assumed that “Europe” means the EU. Michael Bruter, however, identified two separate meanings: a civic meaning attached to the EU, and a cultural meaning attached to Europe as a civilization (Bruter 2005, pp. 110–11). When confronted with a general query about “Europe”, the French and Dutch often equated “Europe” with the “European Union” (civic identity), but the British were inclined to identify “Europe” culturally, not civically (Bruter 2005, p. 114).

Thus, the questions used produced somewhat different results. All seem to divide the public into what Green called core, secondary, and non-European identifiers (2007, p. 52). Surveys forcing respondents to make specific choices of identities (European or national) have found that core-“Europeans” (first choice) made up approximately five percent, with an additional twelve percent choosing Europe, but listing it second (Green 2007, p. 56). Thus, about eighty-three percent of Europeans fell into the non-identifier category. Surveys allowing greater flexibility, however, yielded higher proportions of core and secondary identifiers. Fuchs and his colleagues, for instance, used a Eurobarometer 62 (Autumn 2004) item: “People feel different degrees of attachment to their town or village, to their region, to their country or to Europe. Please tell me how attached you feel to (a) your city/town/village, (b) your region, (c) your country, and (d) Europe?” About three percent identified exclusively with Europe (core identifiers), and thirty-one percent identified exclusively with their countries or refused to identify with either their country or Europe (non-identifiers). Thus, sixty-six percent identified with Europe to some extent (secondary identifiers), a much higher proportion than the twelve percent cited above (Fuchs et al. 2009).

After reviewing all the early data, Green offered his “best guess” summary: “It would seem reasonable to conclude that something on the order of fifteen percent of Europeans (roughly one of every six or seven) strongly identifies with Europe, while perhaps another thirty-five percent (roughly every third person) also identifies with Europe, but to a lesser extent. This leaves about half of the population for whom a European identity is either irrelevant or actively rejected” (Green 2007, pp. 54–55). The good news for EU elites, then, was that many people considered themselves “European”. But half did not, and the European identity that existed seemed quite shallow: in the first decade of this century, very few saw themselves exclusively in European terms. In Risse’s telling characterization, it was “European identity lite” (Risse 2010, p. 60).

The news might be better for EU elites if “Europeans” were multiplying, perhaps responding to their continued identity construction efforts. But on the whole, the story since the early 1970s has been “stasis, not movement” (Green 2007, p. 66). Indeed, additional recent surveys show only modest changes. Kuhn’s review of Eurobarometer data through 2019 found very little increase in the number of exclusively “European” identifiers or of “European first” citizens, but did note a modest decrease in “national only” respondents and more citizens who are “nationals” first, but also “European”—at least in the EU 12 (Kuhn 2019, p. 1222).

Despite these persistent patterns, there may still be hope for growing European identity. Bruter hoped that over time exposure to European institutions and symbols should bolster European identities and Risse argued that creation of a “common communicative space”, where Europeans debate common issues, would help Europeanize them (Bruter 2005, pp. 123–29, 140–41; Risse 2010, p. 5). Ceremonies and symbols may have only touched the emotional surface of most Europeans: deep identification requires more than seeing a flag or holding a passport, although familiarity with symbols may at least produce what Foret and Trino (2021) call “banal Europeanism”. If Bruter and Risse are correct, the emergence of a truly European public sphere may be a better indication of growing European identity, but such phenomena are hard to measure. Even if European identity is widening and deepening, it is still fragile and vulnerable to economic and political shocks.

4. Religious Influences on the Creation of European Identity

Scholars have identified a variety of individual traits that help explain the locus of European identity: it is most common among younger, highly educated, politically active, wealthy, and cosmopolitan citizens who live in large urban centers and hold “modern” social and political values (Bruter 2005; Green 2007; Risse 2010; Keulman and Koós 2014). However, it is also evident that “Europeans” are unevenly distributed across the EU, with some countries showing a higher proportion than others. Scholars have offered a variety of historical, political, and cultural explanations for this finding, sometimes incorporated in empirical analyses using a measure of “national tradition”, basically just a placeholder for unspecified or unmeasured influences (Eichenberg and Dalton 1993).

We think that one of those usually unmeasured factors—religion—can add to our understanding of identity construction. Elites can construct national identities, but not out of thin air. They need shared cultural materials that carry emotional weight across social cleavages to bind together a new nation—to create “solidarity among strangers” (Habermas 2006). Religion can be such a cultural “binder”, although that possibility has been ignored or downplayed by most analysts. Nevertheless, there are plenty of reasons to consider religion. The postwar European integration efforts were largely the project of devout Catholic Christian Democratic political leaders and strongly supported by the Vatican and Catholic hierarchy. Protestant political leaders—and Protestant churches—were much less enthusiastic, fearing integration’s threats to national sovereignty (Kaiser 2007; Minkenberg 2009; Leustean 2014; Mudrov 2015; Kratochvíl and Doležal 2015; Nelsen and Guth 2015). Indeed, the Christian Democratic founders saw themselves as “national identity constructors” and sought common symbols for the budding European polity, often grounding them in Christian or Carolingian iconography. The passing of this Christian Democratic generation and dramatic secularization in Europe did not end such construction efforts, but did shift the emphasis away from overt religious symbolism. Still, religious elements remained (Foret 2015).

Public opinion polls reveal that the confessional split among leaders over Catholic and Protestant visions of Europe also divided the Continent’s citizens, especially in the early stages of integration. Before 1970, surveys of public attitudes toward integration were irregular and confined to individual countries, but the limited evidence suggests broad support for the dominant Christian Democratic program of reconciliation and integration. Public backing for integration was strong and rising: Between 1950 and 1962, European support rose from 50 to 70 percent, creating a “permissive consensus” that facilitated the elite construction of a new Europe (Lindberg and Scheingold 1970, p. 39).

Systematic polling of public attitudes began when the European Commission conducted surveys of the Six in 1970 and 1971. Biannual polls of an enlarged Community began in late 1973, with the spring 1974 iteration tagged the Eurobarometer. Although the Eurobarometer series produced a wealth of data on attitudes toward integration, it seldom included questions about religion. The few surveys that did, however, revealed a clear picture of religion’s impact on public attitudes. Historical experience suggests that Catholic transnationalism should produce support for integration, while the nationalism of Protestant churches would work in the other direction. And the most observant members of each should exhibit those confessional tendencies more strongly than less faithful believers.

Eurobarometer data consistently confirmed those hypotheses: Confessional culture (as measured by self-identified religious tradition) had a significant effect, even under rigorous controls for other factors often identified as producing support for integration (e.g., nationality, party identification, ideology, political engagement, postmaterial values, economic situation, gender, class, and education). From 1973 to 2006, Catholics in the Nine (the Six, plus Britain, Denmark, and Ireland) were significantly more favorable to the European Community/Union than those of other faiths or no faith. Furthermore, devout Catholics—regular Mass attenders—showed the strongest support. Protestants showed much less enthusiasm for integration, although in later years greater devotion actually increased support (at least with everything else in the equation), with nominal or “cultural” Protestants as the most negative, suggesting that exposure to the growing ecumenism in Protestant state churches has encouraged internationalism among the more observant. Eurobarometer measures are not usually precise enough to identify the remnant “free church” or sectarian Protestants, but where identified this group was fiercely anti-integration (Hagevi 2002; Madeley 2012; Nelsen and Guth 2017). Although these relationships have changed over the past forty years, religious factors still help explain support for the EU (for reviews see Boomgaarden and Freire 2009; Nelsen and Guth 2015; Scherer 2020).

The crucial point: we expect religion to influence “European” identity in very much the same way as it shapes support for integration. Unfortunately, testing this proposition is not easy: surveys asking about European identity seldom have good religious measures. Nevertheless, analysis of aggregate data suggests that religion does shape “European” identity. For example, Eurobarometer 61 (Spring 2004) gave respondents four options, ranging from exclusive national identity to exclusive European identity. If we correlate the percentage of respondents in each country who answer “nationality only” with the proportion of Protestants, we find a strong relationship (r = 0.419); the more Protestants, the more “national only” identifiers, a finding that survives statistical controls for other influences (cf. Keulman and Koós 2014, p. 281). Thus, religion might well matter at the individual level, even when we incorporate other variables shown to influence identity.

5. Influences on European Identity: Religion, Ideology, and Demography

Scholars have identified many factors that shape European identity. To examine religious variables in this full context, we propose the following hypotheses about the way religion influences the creation of “Europeans” and also outline expectations about the contribution of “control” variables:

Hypothesis 1.

Catholic and Orthodox2 believers will offer more “European” responses than Protestants and “Other Christians”, while members of minority religions as well as atheists and agnostics will be more European (Risse 2010).

Hypothesis 2a.

Religious commitment will reinforce the “normative” position in each major tradition: pro-European among Catholics and Orthodox; anti-European among Protestants and Other Christians.

Hypothesis 2b.

Religious commitment will produce a pro-European effect across all major religious traditions.

To test the independent effects of religion on European identity, we also include influences stressed by previous scholarship. First, ideology should influence identity, with neo-liberals, postmaterialists (Inglehart 1977), and the left as more “European” (Risse 2010; Cirlanaru 2016). Strong national identity and trust in national political institutions will enhance European identity, contrary to some scholars’ expectation but consistent with others (Bruter 2005; Guglielmi 2016). “Cognitive mobilization”, as measured by knowledge about the EU, political engagement, and higher education, will produce a stronger European identity (Inglehart 1970). Economic “winners”, such as business people and professionals, men, the internationally mobile (Weber 2016), young people, and urbanites—as well as those anticipating future prosperity—will be more “European” (Gabel 1998; McLaren 2006; Green 2007; Risse 2010, p. 48; Bruter 2005, pp. 32–33; Cirlanaru 2016). Finally, socialization should have an impact. European identity will be lower in countries where the overall EU approval is low (Eichenberg and Dalton 1993; Bruter 2005, 136ff), in nations participating in fewer EU “formations” (Nelsen and Guth 2015), and in countries only recently admitted (Risse 2010, p. 41).

6. Data and Methods

Unfortunately, given the lack of Eurobarometer surveys tapping both European identity and religion since 2006, we have no comprehensive data for the entire EU over that period.3 However, Eurobarometer 65.2 (Papacostas 2006), coming after several decades of “identity-construction” by EU elites and sampling the entire contemporary EU (including now-departed Great Britain), provides a good bit of insight into the questions of interest. Unlike earlier analyses focusing on the “EU 12” or “EU15”,4 we are able to determine whether the religious and other factors produce different effects in successive accession groups. The survey may be instructive for other reasons: it was administered while the EU faced a series of economic, political, and foreign policy problems that constituted what German politician Wolfgang Schäuble called “its worst crisis” (quoted in Karolewski et al. 2016, p. 1). Thus, the survey should reveal insights about the nature of European identity in a critical conjunction of long and short term forces.

Most important for present purposes, EB 65.2 included critical religious questions and identity items. Two get at Bruter’s “civic” and “cultural” identities. Half the sample was asked “how attached” they felt to “the European Union”, a “civic” identity; the other half was asked about their attachment to “Europe”, a “cultural” identity. A third measure draws from three items on EU threats to national and cultural identities; the fourth is calculated from a battery on the EU flag, which Risse calls an “identity marker” for Europeans (Risse 2010, p. 58; see also Cram and Patrikios 2016). All four measures have solid face validity: citizens who feel attached to the EU (or to Europe), who do not fear that the EU threatens existing national and cultural identities, and who value the EU flag (and want it flown next to their national banner) would seem to be the very “Europeans” that EU elites and integration scholars have long sought (see the Appendix A for full details on variable construction).

Using alternate identity measures also avoids reliance on one or two questions that suffer from a limited range and get at only part of the concept. Multiple measures allow us to assess different emotive dimensions of “Europeanness”. That each taps slightly different facets of a single phenomenon is confirmed by the fact that they are closely related empirically and when combined, form a composite measure which relates in expected ways to a wide range of independent variables.

7. Correlates of European Identity: A First Look

What do we discover? As a first cut,

Table 1. Correlates of Alternate Measures of European Identity (Pearson’s r).

	Attached to EU (“Civic”)	Attached to Europe (“Cultural”)	Do Not Fear Identity Loss	Identify with EU Flag	European Identity Score
Religious Variables
Majority Catholics	0.134 ***	0.064 ***	0.127 ***	0.093 ***	0.135 ***
X commitment	0.133 ***	0.095 ***	0.108 ***	0.079 ***	0.133 ***
Majority Protestants	−0.081 ***	−0.036 ***	−0.137 ***	−0.105 ***	−0.122 ***
X commitment	−0.067 ***	−0.012	−0.113 ***	−0.082 ***	−0.092 ***
Other Christians	−0.076 ***	−0.074 ***	−0.090 ***	−0.069 ***	−0.105 ***
X commitment	−0.049 ***	−0.067 ***	−0.068 ***	−0.057 ***	−0.083 ***
Orthodox	0.025 **	0.022 *	0.050 ***	−0.045 ***	−0.006
X commitment	0.023 **	0.019 *	0.051 ***	−0.050 ***	−0.009
Other religions	−0.009	−0.009	0.017 **	−0.020 ***	−0.007
Atheist	−0.032 ***	−0.030 ***	0.010	0.009	−0.007
Ideological Factors
Neo-liberalism	0.276 ***	0.253 ***	0.119 ***	0.271 ***	0.314 ***
Postmaterialism	0.044 ***	0.048 ***	0.021 ***	0.096 ***	0.081 ***
Right	0.011	0.016	−0.053 ***	0.034 ***	0.007
Left	−0.005	0.035 ***	0.033 ***	0.031 ***	0.035 ***
National Orientations
Attached to Country	0.237 ***	0.271 ***	−0.052 ***	0.105 ***	0.162 ***
Trust National Government	0.199 ***	0.203 ***	0.107 ***	0.181 ***	0.232 ***
Cognitive Mobilization
Knowledge of EU	0.270 ***	0.247 ***	0.066 ***	0.296 ***	0.309 ***
Discuss Politics	0.108 ***	0.135 ***	−0.007	0.172 ***	0.148 ***
Education	0.103 ***	0.126 ***	0.070 ***	0.190 ***	0.183 ***
Demographics
Higher Occupation	0.070 ***	0.051 ***	0.045 ***	0.093 ***	0.094 ***
Male	0.070 ***	0.020 *	0.024 ***	0.107 ***	0.084 ***
Born Other Country	0.064 ***	0.046 ***	0.022 ***	0.031 ***	0.053 ***
Age	−0.136 ***	−0.020 *	−0.067 ***	−0.136 ***	−0.104 ***
Rural	−0.078 ***	−0.028 **	−0.013 *	−0.078 ***	−0.063 ***
Economic Expectations
Positive National	0.246 ***	0.239 ***	0.139 ***	0.195 ***	0.271 ***
Positive Personal	0.182 ***	0.145 ***	0.111 ***	0.204 ***	0.231 ***
Contextual Socialization
National tradition	0.162 ***	0.145 ***	0.257 ***	0.091 ***	0.206 ***
Number formations	0.056 ***	0.062 ***	0.177 ***	0.150 ***	0.171 ***
Late accession	−0.025 **	0.042 ***	0.079 ***	0.003	0.039 ***
N	12,646	12,942	25,589	25,589	25,589

Source: Eurobarometer 65.2 (2006). *** p < 0.001; ** p < 0.01; * p < 0.05.

reports bivariate relationships (Pearson’s r) between indicators for each hypothesis and the four measures tapping European identity, along with the composite overall score. As we anticipated, Catholics in Catholic countries are systematically more attached to both the EU and “Europe” (but especially the former), have less fear of cultural and national identity loss, and respond more warmly to the EU flag.5 And the solid correlation of Catholic affiliation with the composite measure summarizes the overall impact. In addition, greater Catholic commitment (measured by religious attendance) is uniformly correlated with “European” responses, especially on the composite measure.

Protestants also conform to expectations. Those in Protestant-heavy countries are less attached to the EU and Europe, more prone to fear identity loss, and don’t like the EU flag. Their consistency is revealed by the strength of the negative coefficient on the composite measure. And, again, as predicted, the most committed Protestants are less “European”. In both respects “Other Christians” look like Protestants. And that is because they probably are: although the EB’s coding does not allow us to be sure, their geographic locations suggest that they are “evangelical” or “free church” Protestants, whose “less European” identity is fortified by regular church attendance.

The EU’s twenty-first century expansion added many Eastern Orthodox believers. The Orthodox are mildly “European” on most indicators, a tendency enhanced slightly by observance—but are less likely to identify with the flag. Because of the small numbers of Muslims, Hindus, Buddhists, Jews, and other non-Christians in the study, we combine them in an “Other Religions” category. Although we join the scholars suggesting that minority groups should be less identified with their nation of residence and more with “Europe”, others have argued that their non-European origins and anti-immigrant sentiment among national populations might lead religious minorities toward anti-European sentiments (Pastorelli 2012). In fact, they do not differ much from the rest of the public, although, similar to the Orthodox, they are slightly less prone to identify with the flag. Surprisingly, atheists are not very distinctive, even though some scholars have envisioned them leading the way to attachments “beyond the nation state” (Risse 2010, pp. 6, 10; Laitin 2002).

Ideology also influences identity. Neo-liberal ideas are especially powerful: citizens who believe in free markets, business competition, and limited government are more “European” on all four individual indicators and, especially, on the summary measure. Inglehart’s (1977) “postmaterial values” have a very small but consistent positive association, but pale in comparison with neo-liberalism. Apparently those valuing “quality of life” and “modern” social mores are not much more “European” than their “materialist” counterparts concerned with economic well-being. Those on the “Right” or “Left” of the political ideology scale are not much different from the rest of the population, although the Left may be a tad more “European”.

As some scholars have shown, national affections do not necessarily work against European ones (Guglielmi 2016). In fact, attachment to one’s country has a fairly solid positive correlation with attachment to the EU and support for the EU flag, although linked to a slight tendency to fear identity loss as a result of the EUs operation and expansion. Trust in national government institutions, however, is consistently associated with positive attitudes toward Europe. This suggests, perhaps, a tendency for generalized attachments: citizens feeling close to locality, region, and nation also tend to identify with their “supranational” home, the EU, and perhaps, trust its institutions (cf. Bruter 2005, pp. 115–17).

Cognitive mobilization is also tied to European affinities. Self-professed knowledge about the EU and understanding of its institutions has by far the largest connection: those who know most feel most attached to the EU and to Europe, fear identity loss the least, and love the flag. Higher education and frequency of political discussion, have more modest, but consistently positive associations with European attachments, as predicted. On the other hand, the demographic “winners and losers” perspective stressed by many analysts (e.g., Gabel 1998) seems less useful, although the expected patterns appear: higher occupational status produces more “Europeans”, but detailed examination shows the effects are concentrated at the ends of the continuum: business and professional people are more “European” and working class folks less so (data not shown). As scholars have found earlier, males are more attached to Europe, as are younger people6 and national residents born in other countries. Finally, rural residents are slightly less likely to exhibit “European” markers than people in larger urban settings. Although occupational and residence indicators stressed by the “winners and losers” genre are relatively weak predictors, individual economic expectations play a bigger role: those who anticipate a better national economy in the next five years feel distinctly more “European”, while high expectations for one’s personal fortunes have the same effect, if not quite as strong.

Finally, contextual socialization indicators have mixed effects. “National tradition” is measured simply by the mean scores for all a nation’s citizens on support for the EU. This recognizes that a host of historic influences shape responses to the EU in each country; such forces presumably influence all citizens in a nation and are reinforced by social interaction (Eichenberg and Dalton 1993). Although the national mean does correlate with the dependent variables, the coefficients are fairly modest, suggesting a good bit of variation among citizens in each country. A second measure of contextual socialization is each country’s membership in various EU “formations”, or cooperative arrangements among groups of member states. Presumably, citizens in countries participating in more transnational formations should feel more “European”. That proves to be the case: although the bivariate effect is not especially striking on attachment to the EU or Europe, it is a good bit stronger on the other two measures and the summary score. On the other hand, the length of time that one’s country has been part of the European Community/Union has no strong or consistent direct influence on European sentiments; indeed, on two indicators later entrants are more “European”. If time in the EU increases chances for socialization, that process has not had much of an impact.

8. Putting It All Together: Multivariate Analyses

Of course, many of the independent variables are interrelated. To determine the direct influence of each while controlling for that of others, we incorporated the individual-level measures in Ordinary Least Squares (OLS) regressions for each “European” orientation.

Table 2. European Identity and Individual Traits (OLS regression betas).

	Attached to EU (“Civic”)	Attached to Europe (“Cultural”)	Do Not Fear Identity Loss	Identify with EU Flag	European Identity Score
Religious Variables
Majority Catholics	0.108 ***	−0.017	0.152 ***	0.098 ***	0.103 ***
X commitment	0.053 ***	0.097 ***	0.008	−0.019	0.043 ***
Majority Protestants	−0.039 *	−0.091 ***	−0.131 ***	−0.127 ***	−0.141 ***
X commitment	−0.025	0.045 **	0.035 **	0.012	0.028 *
Other Christians	−0.082 ***	−0.036 *	−0.091 ***	−0.056 ***	−0.096 ***
X commitment	0.056 ***	−0.022	0.031 *	0.003	0.024 *
Orthodox	−0.067 *	0.018	−0.020	0.024	−0.011
X commitment	0.042	−0.090 ***	0.013	0.003	−0.011
Other religions	0.024 **	0.013	0.037 ***	−0.012 *	0.015 **
Atheist/agnostic	0.007	−0.010	0.028 *	−0.002	0.005
Ideological Factors
Neo-liberalism	0.141 ***	0.118 ***	0.082 ***	0.160 ***	0.176 ***
Postmaterialism	0.005	0.000	0.013	0.051 ***	0.032 ***
Right	−0.070 ***	−0.017	−0.056 ***	−0.018 **	−0.049 ***
Left	−0.017	0.042 ***	0.026 ***	0.010	0.020 ***
National Orientations
Attached to Country	0.182 ***	0.234 ***	−0.082 ***	0.091 ***	0.120 ***
Trust National Government	0.095 ***	0.106 ***	0.079 ***	0.092 ***	0.123 ***
Cognitive Mobilization
Knowledge of EU	0.201 ***	0.174 ***	0.017 *	0.171 ***	0.186 ***
Discuss Politics	0.041 ***	0.066 ***	−0.008	0.072 ***	0.059 ***
Education	0.012	0.044 ***	0.040 ***	0.043 ***	0.055 ***
Demographics
Higher Occupation	0.031 ***	−0.006	0.033 ***	0.010	0.021 ***
Male	0.020 *	−0.027 ***	0.011	0.038 ***	0.019 **
Born in Other Country	0.058 ***	0.065 ***	0.022 ***	0.025 ***	0.051 ***
Age	−0.009	−0.022 *	−0.004	−0.073 ***	−0.045 ***
Rural	−0.044 ***	−0.012	−0.015 *	−0.040 ***	−0.037 ***
Economic Expectations
Positive National	0.118 ***	0.124 ***	0.081 ***	0.063 ***	0.124 ***
Positive Personal	0.026 *	0.000	0.032 ***	0.041 ***	0.035 ***
Adj. R squared	0.245	0.230	0.098	0.209	0.286
N	12,646	12,942	25,589	25,589	25,589

Source: Eurobarometer 65.2 (2006). *** p < 0.001; ** p < 0.01; * p < 0.05.

reports standardized coefficients (betas) for each regression, as well as the amount of variance explained.

Although the beta coefficients are naturally smaller than the correlations in Table 1, the regression results confirm most of the bivariate patterns seen there. First, Catholic church membership contributes substantially to European identity, while Protestant affiliation still moves the other way. Interestingly, when everything is in the equation, religious commitment has no consistent effect among either Catholics or Protestants but there is already a hint in 2006 of later findings that by this time religiosity in all traditions was bolstering “European” sentiments, at least when all other variables are considered (cf. Keulman and Koós 2014; Nelsen and Guth 2020). “Other Christians” are consistently less European, and the net impact of their devotion is mixed, but tends again toward the positive side. The Orthodox, on the other hand, do not differ from the omitted reference group of the religiously unaffiliated. “Other Religions” are slightly more attached to the EU and less fearful about identity loss, but identify less with the EU flag. Atheists differ from the unaffiliated only in being slightly less fearful of identity loss, perhaps revealing weaker attachments to their national cultures.

Ideology also exerts considerable direct influence. Neo-liberalism is the most powerful, while postmaterialism tempers fears of identity loss and encourages fealty to the flag, but has no independent impact on attachment to the EU or Europe. With everything in the equation, the Right is consistently less European, while the Left shows small net tendencies toward European identity. Both attachment to country and trust of one’s national government has a consistent positive effect, except for concern about identity loss exhibited by those attached to their country. Knowledge about the EU has a strong, consistent positive effect on European identity, bolstered by higher education and political discussion. Demographics have modest effects. Higher occupational status produces a slightly greater propensity to be European on three of the five measures, men have a small advantage on four of five, older folks are less European on all the measures, as are rural dwellers, although the coefficients are small and not always significant. Finally, emigres are more likely to identify with Europe on all five indicators.

Economic assessments are important to development of a European identity. Here, the “sociotropic” evaluation of the future national economy carries more weight than a similar assessment of personal fortunes. Those who think their national economy will be better off in five years are more European, while the assessment about personal finances adds a modest bit of explanatory power, consistent with studies that show the stronger impact of global assessments over individual “pocketbook” concerns when citizens evaluate the EU (McLaren 2006).

All this suggests two general conclusions. First, European identity is a complex affair, with all of the theoretical perspectives discussed above finding at least some support in the data. Second, although coefficients for the independent variables differ slightly from equation to equation (depending in part on measurement characteristics of each variable), the patterns are quite consistent and the equations predict substantial variance in all five scores: adjusted R squares range from 0.098 for the limited identity loss measure to 0.286 for the composite score, a very respectable showing vis à vis other studies.

9. Adding Contextual Influences: The Impact of Socialization

In

Table 3. European Identity, Individual Traits, and National Context (OLS regression betas).

	Attached to EU (“Civic”)	Attached to Europ (“Cultural”)	Do Not Fear Identity Loss	Identify with EU Flag	European Identity Score
Religious Variables
Majority Catholics	0.075 ***	−0.049 **	0.071 ***	0.046 ***	0.032 **
X commitment	0.061 ***	0.087 ***	−0.008	−0.005	0.039 ***
Majority Protestants	−0.007	−0.063 ***	−0.058 ***	−0.068 ***	−0.071 ***
X commitment	−0.027	0.040 *	0.022	0.006	0.017
Other Christians	−0.060 ***	−0.010	−0.030 *	−0.011	−0.040 ***
X commitment	0.054 ***	−0.031 *	0.016	−0.007	0.011
Orthodox	−0.066 **	0.009	−0.038	0.019	−0.024
X commitment	0.038	−0.096 ***	−0.002	−0.001	−0.022
Other religions	0.025 **	0.007	0.028 ***	−0.012 *	0.010
Atheist/agnostic	0.009	−0.013	0.021 ***	−0.002	0.000
Ideological Factors
Neo-liberalism	0.139 ***	0.114 ***	0.074 ***	0.158 ***	0.170 ***
Postmaterialism	0.004	0.003	0.019 **	0.049 ***	0.036 ***
Right	−0.068 ***	−0.018 *	−0.054 ***	−0.015 *	−0.047 ***
Left	−0.016	0.037 ***	0.019 **	0.007	0.015 *
National Orientations
Attached to Country	0.191 ***	0.234 ***	−0.075 ***	0.101 ***	0.128 ***
Trust National Government	0.085 ***	0.098 ***	0.057 ***	0.075 ***	0.102 ***
Cognitive Mobilization
Knowledge of EU	0.201 ***	0.175 ***	0.020 **	0.170 ***	0.187 ***
Discuss Politics	0.037 ***	0.064 ***	−0.012	0.067 ***	0.054 ***
Education	0.018 *	0.044 ***	0.046 ***	0.050 ***	0.061 ***
Demographics
Higher Occupation	0.026 **	−0.004	0.036 ***	0.005	0.021 ***
Male	0.020 *	−0.028 ***	0.009	0.039 ***	0.019 ***
Born in Other Country	0.059 ***	0.069 ***	0.030 ***	0.028 ***	0.058 ***
Age	−0.009	−0.018 *	0.004	−0.074 ***	−0.040 ***
Rural	−0.042 ***	−0.014	−0.020 ***	−0.040 ***	−0.040 ***
Economic Expectations
Positive National	0.119 ***	0.119 ***	0.066 ***	0.060 ***	0.114 ***
Positive Personal	0.034 ***	0.007	0.053 ***	0.057 ***	0.055 ***
Contextual Socialization
National tradition	0.073 ***	0.060 ***	0.141 ***	0.079 ***	0.117 ***
Number of formations	0.031 **	0.053 ***	0.147 ***	0.098 ***	0.132 ***
Late accession	−0.063 ***	0.035 ***	0.070 ***	−0.032 ***	0.029 ***
Adj. R squared	0.253	0.236	0.133	0.223	0.310
N	12,184	12,509	25,589	25,589	25,589

Source: Eurobarometer 65.2 (2006). *** p < 0.001; ** p < 0.01; * p < 0.05.

we replicate the analysis, but add three indicators of contextual socialization. Much of the literature on support for European integration and on the development of European identity has stressed national differences. These differences have often been attributed to “national tradition”, a catchall term summarizing the varied historical experiences of member nations (Eichenberg and Dalton 1993). Presumably, the same influences that shape support for integration should also shape the development of European identity. Others have emphasized “transnational” socialization, based on common participation in the emerging European political system and its “public space”. We tap these possibilities with (1) the mean national rating for the EU, (2) the number of “formations” a country participates in, and (3) the length of national membership in the EU and its predecessors.

National tradition does matter, as a nation’s generalized support for the EU produces modest net positive effects on all four measures of European identity, as does participation in EU formations. Additionally, when everything is in the equation, a late accession date into the EU has a mixed effect: reducing somewhat the attachment of a country’s citizens to the EU and identification with the EU flag, but enhancing attachment to Europe and reducing the fear of identity loss, with a net positive effect on the composite measure. On the whole, then, contextual socialization variables have a positive, if fairly limited role in explaining European identification. Together, these variables increase the variance explained from 0.245 to 0.253 on attachment to the EU, from 0.230 to 0.236 on attachment to Europe, from 0.098 to 0.133 on fear of identity loss, from 0.209 to 0.223 on the flag, and from 0.286 to 0.310 on the composite measure. But it is clear that our contextual socialization measures do not dominate the explanation.

Without belaboring the point, we will simply note that inclusion of the socialization variables leaves most other predictors in significant roles, although with some changes from the coefficients in Table 2. This is clear in the case of religion, arguably an important part of national tradition. The inclusion of the national tradition variable reduces somewhat the negative coefficients for majority Protestant affiliation, but this is not unexpected: Protestantism has been a central aspect of national culture in Britain and the Nordics since the Reformation and its “Euroskeptic” posture is long-standing and embedded in national tradition. In the same vein, the coefficients for “Other Christians” remain consistently negative, although not always statistically significant. Catholic affiliation, on the other hand, survives the incorporation of national tradition estimates, befitting a religion that has always been transnational, historically suspicious of the nation state, and supportive of European integration (Scherer 2020). Catholic affiliation is clearly supportive of a “European” identity as well.

The impact of membership in other religious groups remains very modest and inconsistent. “Other Religions” show some tendencies toward European orientations, mostly in line with the comparable coefficients in Table 2, but once again atheists do not differ from the reference group on any indicator, raising again the question about whether secular citizens can provide the cutting-edge of “European identifiers”. Nor does the inclusion of contextual socialization measures eliminate the effects of many other individual-level variables identified in earlier studies. Neoliberals, national patriots, the cognitively mobilized, business and professional people, males, urbanites, transnationals, and the economically optimistic are still most likely to give positive responses to different measures of European identity. Clearly that identity issues from a complex interaction of religious, ideological, demographic, and economic factors.

10. Support for the EU and European Identity

We need to address an additional possibility. The integration literature is bedeviled by a crucial question of causal order: does European integration require supportive “Europeans” for its development? Or does the EU’s success produce more “Europeans” in its wake? If the latter possibility is dominant, those who approve of the EU’s contemporary functioning should be more “European” in their responses. Indeed, “European identity” may be nothing more than approval of the EU’s performance. If true, the incorporation of generalized EU support in the analysis should absorb all the influences considered, especially that of religion, with its considerable impact on evaluations of the EU.

To test this possibility, we ran two OLS regressions on the identity score, using first individual variables and then adding socialization measures, while incorporating an especially robust measure of EU support, based on nine items (alpha = 0.87). Not surprisingly, approval of the EU produces a strong positive coefficient in both (

Table 4. European Identity, Individual Traits and National Context (OLS regression betas).

	Model 1	Model 2
Support for European Union	0.559 ***	0.551 ***
Religious Variables
Majority Catholics	0.062 ***	0.021 *
X commitment	−0.010	0.014
Majority Protestants	−0.091 ***	−0.041 ***
X commitment	0.022 *	0.015
Other Christians	−0.071 ***	−0.028 **
X commitment	0.021 ***	0.010
Orthodox	−0.063 ***	−0.073 ***
X commitment	−0.008	−0.007
Other religions	0.001	0.001
Atheist/agnostic	0.006	0.004
Ideological Factors
Neo-liberalism	0.061 ***	0.063 ***
Postmaterialism	0.033 ***	0.031 ***
Right	−0.020 ***	−0.016 ***
Left	0.009	0.007
National Orientations
Attached to Country	0.105 ***	0.109 ***
Trust National Government	−0.018 ***	−0.029 ***
Cognitive Mobilization
Knowledge of EU	0.089 ***	0.088 ***
Discuss Politics	0.059 ***	0.054 ***
Education	0.038 ***	0.041 ***
Demographics
Higher Occupation	0.021 ***	0.018 ***
Male	0.008	0.011 *
Born in Other Country	0.036 ***	0.044 ***
Age	−0.028 ***	−0.029 ***
Rural	−0.019 ***	−0.025 ***
Economic Expectations
Positive National	0.034 ***	0.031 ***
Positive Personal	−0.012 *	0.005
Contextual Socialization
National tradition	--	0.009
Number of formations	--	0.128 ***
Late accession	--	0.031 ***
Adj. R squared	0.475	0.486

Source: Eurobarometer 65.2 (2006); N = 25,198. *** p < 0.001; ** p < 0.01; * p < 0.05.

).

To make a long story very short: those approving the EU’s performance are indeed much more likely to feel “European”. Nevertheless, that approval does not entirely eliminate the other hypothesized direct influences. Catholics are still more “European”, while Protestants and “Other Christians” are less so. The Orthodox are also less European, a tendency reinforced by religiosity. “Other Religions” exhibit the same mild pro-Europe trends we saw in the previous table, but once again atheists are not at all distinctive. The coefficients for neo-liberalism and postmaterialism hang in there, although markedly reduced as independent factors. National orientations become somewhat mixed, however, with trust in the national government now having a net negative effect. Cognitive mobilization factors also have a mixed effect, once approval of the EU is in the equation.

The demographic factors retain the same modest impacts revealed in earlier tables, except that gender drops out. Economic expectations almost disappear, however, no doubt absorbed by the EU approval scores. If we add the contextual factors to the equation, only participation in EU formations remains, providing modest support for the notion that deeper participation in EU “public space” produces more “Europeans”. Note that the variance explained climbs substantially with inclusion of the EU approval score (from 0.310 to 0.486). Without resolving the causal order issue, we can hazard a prediction that the growth of the “European” population will depend increasingly on positive assessments of the EU by its citizens (cf. Mitchell 2016).

11. Regional Differences in European Identity

Finally, we must address another possibility on the role that religion has played in the development of European identity: that Catholicism (and Catholic devotion) may have buttressed support for the EU and the growth of broader identities in the early stages of integration, but with the accession of Eastern European states with a distinctly less “democratic” religious history, the influence of religion, especially that of Catholicism, has been reversed, especially among the devout, who reject the EU’s social and cultural liberalism. To test these suggestions and to consider the differential impact of other causal factors, we ran the individual level regressions on the composite identity score for three accession cohorts: citizens of the EU “Nine”, the next major accession group, and the ten most recent Eastern European member states.

The results in

Table 5. European Identity: OLS betas for EU Nine, Later and Eastern Accessions.

	EU Nine	Later Accessions	Eastern Accessions
Religious Variables
Majority Catholic	0.119 ***	0.056 *	0.034
X commitment	0.043 **	−0.051 *	−0.023
Majority Protestant	−0.152 ***	−0.083 **	−0.028
X commitment	0.015	0.011	−0.017
Other Christian	−0.106 ***	0.005	−0.035
X commitment	0.026 *	−0.041	0.016
Orthodox	−0.018	−0.061	−0.111 **
X commitment	0.034 *	−0.027	0.016
Other Religions	0.003	0.033 **	−0.001
Atheist	0.010	−0.013	−0.032 *
Ideological Factors
Neo-Liberalism	0.155 ***	0.217 ***	0.181 ***
Postmaterialism	0.045 ***	−0.008	0.024 *
Right	−0.084 ***	−0.020	0.011
Left	0.018 *	0.035 **	0.004
National Orientations
Attachment to Country	0.093 ***	0.163 ***	0.155 ***
Trust National Government	0.155 ***	0.096 ***	0.063 ***
Cognitive Mobilization
Knowledge of EU	0.178 ***	0.237 ***	0.137 ***
Discuss Politics	0.072 ***	0.020	0.041 **
Education	0.048 ***	0.049 ***	0.029 *
Demographics
Higher Occupation	0.037 ***	0.004	0.005
Male	0.022 **	0.028 *	0.009
Born in other EU country	0.056 ***	0.039 **	0.008
Age	−0.027 ***	−0.034 **	−0.054 ***
Rural	−0.045 ***	−0.007	−0.024
Economic Expectations
National (5 year)	0.132 ***	0.119 ***	0.052 ***
Personal (5 year)	0.031 ***	0.046 **	0.096 ***
R squared	0.304	0.296	0.195
N	14,387	5425	5777

Source: Eurobarometer 65.2 (2006). *** p < 0.001; ** p < 0.01; * p < 0.05.

reveal some consistency but also fascinating differences among accession groups. Catholic identity is a positive influence on European identity across all three groups, but with some diminution among more recent EU entrants. Catholic commitment, however, “changes sign” from the EU Nine to later groups, going from a positive to a negative influence on European identity, perhaps reflecting the more conservative, ationalistic orientation of the devout in the East, especially in Poland. Protestant identity goes from a very negative impact in the Nine to lesser effects in the next accession group to very little in the Eastern European nations (with few Protestant citizens, to be sure). A similar pattern is shown by “Other Christians” and the Orthodox, while atheists are slightly less pro-European in the East, but do not differ from the unaffiliated in the other groups.

Ideology is important everywhere, but neo-liberalism is even more vital to European identity in the two more recent accession groups than in the Nine. Postmaterialism produces European identity primarily in the Nine. The Right is less European in the Nine, but not elsewhere, while the Left is slightly more so in the first two membership groups. Perhaps surprisingly, the positive impact of national attachments appears everywhere, but the impact of trust in national government declines in importance across entry groups. Additionally, although all three indicators of cognitive mobilization are important in “old Europe”, they appear to have less effect in the “New Europe”. A mild business-labor cleavage appears in the Nine, but is absent elsewhere.

Of the demographic variables, only age has much of an impact in the second and third entry groups, with older citizens being less “European”. Sociotropic assessments of the economic future substantially bolster European identity in the first two categories, but in Eastern Europe, personal assessments appear more important. Note that the analysis accounts for just under a third of the variance in the first two groups of countries, but only a fifth in the newest members, indicating that European identity is less structured there.

12. Summary and Conclusions

Scholarship on European identity shows that it exists on the Continent, but its distribution is decidedly lumpy and is influenced by a complex of factors, most of which have been identified by previous analysts. Central to our concern here, we find that religion must also be considered when evaluating “European” identity. Whatever measure we use to tap that identity, we can say with some confidence that Catholics are not only more likely to support the integration process, as demonstrated by past scholarship, but they are also more prone than other citizens to identify with Europe—especially if they are observant. On the other hand, Protestants are much more attached to their nation—and skeptical of European identity, although the impact of religiosity among Protestants and “Other Christians” (also less apt to be “European”,) may be shifting to a more positive role, at least when everything else is considered.

Despite the expectations of some scholars, atheists and agnostics (the epitome of “secular cosmopolitans”) do not seem to be a trustworthy vanguard of the “New Europeans”, as they usually differ little from the large bulk of religiously unaffiliated citizens, who are themselves consistently less “European” than Catholics. “Europeans” do cluster somewhat at the cosmopolitan end of the social spectrum, but it is less demography than ideology, political engagement, and economic optimism that shape identity. Those who best understand the new entity emerging in Brussels, possess neo-liberal economic convictions, and anticipate better economic conditions in the future are more likely to identify as “Europeans”. And although European elites have consciously and creatively attempted to construct a European identity with staying power, our analysis indicates that construction efforts may not have had dramatic effects. They were not wrong to engage in identity construction, but may have overestimated the likely impact of their actions.

Finally, we see some negative omens for identity creation. The numerical decline and decreasing devotion of European Catholics has reduced one long-term force working for European identity, even as many still observant Catholics, especially in Eastern Europe, become more skeptical about the EU. Many of these folks may be tempted by the neo-Christian appeals of new nationalist or populist parties feeding off the immigration issue and anti-Muslim sentiment. The continuing economic crisis—or at least stasis—of much of Europe has also eroded the national and personal economic optimism that is another major source of EU approval and European identity—and perhaps the only major source in some new member nations. It seems unlikely, then, that the best identity construction efforts can surmount these religious and economic headwinds, so European enthusiasts must hope for religious and economic revivals—or find some new foundation for building European identity (Ciaglia et al. 2020).