Introduction
The gap between the European Union (EU) and its citizens is widely acknowledged, and is a frequent topic of academic and public debate (e.g. Haller, Reference Haller2009; Hooghe and Marks, Reference Hooghe and Marks2009). According to the European Commission, EU citizens ‘know little about the EU and feel they have little say in its decision-making process. Communication is essential to a healthy democracy. It is a two-way street. Democracy can flourish only if citizens know what is going on, and are able to participate fully’ (Commission of the European Communities, 2006: 2). This study argues that the mass media can improve the democratic performance of the EU and function as a ‘two-way street’ of communication, if the media fulfil two tasks: (a) to provide information on EU governance to citizens and (b) to include the views of EU citizens in the news coverage on EU governance so their opinions become visible to policy makers. Especially in the case of the EU, which a large number of people perceive as a distant and elitist institution (Follesdal and Hix, Reference Follesdal and Hix2006), the visibility of citizens in the news coverage can enhance communication between the institution and its constituents.
A citizen is understood as an individual who is entitled to certain rights and duties by virtue of his or her membership in a democratic political community. In representative democracies, the right to vote is an important citizenship right. Elections function as a feedback mechanism and election campaigns facilitate institutionalized communication between citizens and their representatives (Asp and Esaiasson, Reference Asp and Esaiasson1996: 93). Politicians need to know what their electorate wants in order to be able to act according to the interests of their voters and to be responsive to them (cf. Pitkin, Reference Pitkin1967). Therefore communication is crucial, especially during times of elections. The mass media play an important role in maintaining communication flows between politicians and citizens, as they ‘keep citizens informed of what public officials are doing and public officials informed of what citizens want’ (Milbrath, Reference Milbrath1972: 144). While different normative democratic theories and related models of the public sphere have diverging ideas about the extent to which citizens ought to participate in politics and public debates (Ferree et al., Reference Ferree, Gamson, Gerhards and Rucht2002), with the exception of the elitist model of democracy, most theories perceive citizens as important political actors.
The media have an even more important role in the context of EU governance, as supranational governance decreases citizens’ opportunities to directly participate in the political process (Dahl, Reference Dahl1994). Although previous research has indicated that there is a ‘communication deficit’ (Meyer, Reference Meyer1999) between the EU and its citizens, less attention has been paid to whether this deficit might be the result of the low visibility of EU citizens in news coverage of the EU, and particularly during European elections. This study investigates the visibility of EU citizens in the media coverage as it has significant consequences not only for the mutual understanding of citizens and political elites, but also for citizens’ representation at the EU level. The underlying assumption here is that citizens who become visible in the news coverage represent public opinion, which is not to say that it is representative of the population.
The visibility of political actors in the news coverage is a precondition for the functioning of representative democracy at the national and European levels (cf. de Vreese, Reference De Vreese2003). Generally, there is little research on the visibility of citizens in news coverage. At the national level, research analysing the use of opinion polls, vox pops, exemplars and, to a certain extent, episodic framing and human interest frames, has provided information on citizen visibility in the national news (Daschmann, Reference Daschmann2000; Brookes et al., Reference Brookes, Lewis and Wahl-Jorgensen2004; Lewis et al., Reference Lewis, Wahl-Jorgensen and Inthorn2004; Hopmann and Shehata, Reference Hopmann and Shehata2011). At the European level, the amount of EU news reported by the media and the visibility of EU governmental actors has been investigated in depth (e.g. de Vreese, Reference De Vreese2004; Brüggemann and Kleinen-von Konigslöw, Reference Brüggemann and Kleinen-von Konigslöw2009; Boomgaarden et al., Reference Boomgaarden, Vliegenthart, de Vreese and Schuck2013), while little is known about the visibility of EU citizens. Previous analyses of EU news have been predominantly elite focussed, concentrating primarily on quality media and the visibility of governmental actors (see, e.g. de Vreese, Reference De Vreese2004; Brüggemann and Kleinen-von Konigslöw, Reference Brüggemann and Kleinen-von Konigslöw2009; Boomgaarden et al., Reference Boomgaarden, Vliegenthart, de Vreese and Schuck2013; Gattermann, Reference Gattermann2013; Gattermann and Vasilopoulou, Reference Gattermann and Vasilopoulou2015). With only a few exceptions (e.g. Boomgaarden et al., Reference Boomgaarden, de Vreese, Schuck, Azrout, Elenbaas, van Spanje and Vliegenthart2010, Reference Boomgaarden, Vliegenthart, de Vreese and Schuck2013; Kleinen-von Königslöw, Reference Kleinen-von Königslöw2012; Schuck et al., Reference Schuck, Xezonakis, Banducci and de Vreese2013), past studies have been descriptive in scope and have not tried to further explain different levels of visibility. The visibility of citizens in EU news has mainly been neglected by previous research (for an exception, see Wessler, Reference Wessler2007) and no study has thus far aimed to explain different levels of visibility of EU citizens in the news coverage. This study aims to contribute to the literature by shedding light on a crucial actor in European political communication: EU citizens.
Depending on the extent to which EU citizens are visible in the national news media, citizens of the EU member states might get a different impression as to what extent ordinary EU citizens are able to participate in debates on EU governance. As a consequence, citizens of member states with high levels of visibility of EU citizens might perceive the European public sphere as more participatory and have the impression that their opinions are taken into account in the decision-making process at the EU level. Citizens of EU member states with low levels of visibility of EU citizens, on the other hand, are more likely to perceive debates in the European public sphere as elitist, which might contribute to perceiving the EU as an elitist institution. In the context of this important omission, this study asks: Are EU citizens visible in EU news and does their visibility vary across the EU member states?
When starting from the premise that the visibility of EU citizens in the European public sphere is important for the legitimacy of the EU, it is not sufficient to investigate the levels of visibility, but it is also necessary to understand what factors influence their visibility. Identifying what factors influence the visibility of EU citizens allows the detection of the factors that can stimulate higher levels of visibility. This is especially important, as potentially low levels of visibility of EU citizens are related to the EU’s communication deficit. Understanding the factors that lead to an increased visibility of EU citizens is a first step towards opposing this deficit and, in the long run, might improve the EU’s overall legitimacy. Hence, the second research question of this study is: What factors promote the visibility of EU citizens?
The paper is structured as follows. The first section discusses the concept of the EU citizen and gives a brief overview of the European public sphere and the visibility of EU citizens within it. The research questions, hypotheses, and methodology are presented next, followed by the study’s findings, which are based on a secondary analysis of content analysis data of the 2009 European Parliament election news coverage.
The European public sphere and EU citizens
It should be noted that the EU’s transnational multi-level system of governance is inevitably linked to the member state level and ‘national governments are formidable participants in EU policy making’ (Hooghe and Marks, Reference Hooghe and Marks2001: 2). For example, the nationally elected heads of state or government of the EU member states constitute the European Council and take decisions at the EU level. This means, with regard to the actors that become visible in EU news coverage, one can generally distinguish between (a) actors from the national arena that appear in the context of EU governance and (b) European actors that are directly related to the EU level. This study makes the case that this distinction also applies to EU citizens. The argument is outlined after a brief definition of the European public sphere.
The European public sphere has been theorized about and discussed by an extensive body of research (Eriksen, Reference Eriksen2005; Machill et al., Reference Machill, Beiler and Fischer2006; Neidhardt, Reference Neidhardt2006; Koopmans, Reference Koopmans2007; Sifft et al., Reference Sifft, Brüggemann, Kleinen-von Königslöw, Peters and Wimmel2007; Wessler et al., Reference Wessler, Skorek, Kleinen-von Königslöw, Held, Dobreva and Adolphsen2008). Scholars assume that there are two forms in which a European public sphere could potentially exist: (1) as a single transnational European public sphere based on an integrative media system or (2) via the Europeanization of the national public spheres of the EU member states (see Gerhards, Reference Gerhards1993). Europeanization in this context means that the public spheres of the EU member states become more European by including either EU affairs and EU actors (vertical Europeanization) or events and actors from fellow EU member states (horizontal Europeanization) to a greater extent in their coverage (Koopmans and Erbe, Reference Koopmans and Erbe2004; Brüggemann and Kleinen-von Konigslöw, Reference Brüggemann and Kleinen-von Konigslöw2009). Extant research suggests that a European public sphere has emerged, but that the degree of Europeanization of the public spheres of EU members differs by country (see, e.g. Eriksen, Reference Eriksen2005; Wessler et al., Reference Wessler, Skorek, Kleinen-von Königslöw, Held, Dobreva and Adolphsen2008; Brüggemann and Kleinen-von Konigslöw, Reference Brüggemann and Kleinen-von Konigslöw2009).
An analysis of the visibility of EU citizens requires defining and discussing the potential ways in which EU citizens can become visible actors in the European public sphere. The notion of EU citizenship was a topic of discussion in the 1970s, when there was a shift from a purely economic European integration to a supplementary political integration (cf. Wiener, Reference Wiener1998). Even though citizens of the member states were granted citizenship rights related to the EU, such as the right to vote in European elections, it was not until 1993 that EU citizenship was legally introduced with the Maastricht Treaty. EU citizenship was implemented during an era of declining support for (and trust in) the European project (Thomassen and Schmitt, Reference Thomassen and Schmitt1999). The vast importance of citizenship is grounded in the political legitimacy of democratic governance (Beetham, Reference Beetham1991; Coicaud, Reference Coicaud2004). Governance is considered legitimate if citizens approve of the political outcome and the political process is accountable, meaning that citizens can assign responsibilities to political actions (Scharpf, Reference Scharpf1999; Andeweg and Thomassen, Reference Andeweg and Thomassen2005). Furthermore, the government has to be formed and implement policies according to citizens’ preferences in order to be considered responsive (Powell, Reference Powell2004; Markowski, Reference Markowski2011).
In this study, an EU citizen is understood as an ordinary person from an EU member state who is a member of the EU’s political community and is, based on his or her legal status, entitled to certain rights and duties. An active EU citizen exercises these rights and duties and participates in the EU’s political community. Legally, holding the nationality of an EU member state is a precondition of EU citizenship (Treaty of Maastricht, Art. 8.1). The Treaty of Amsterdam stipulated that EU citizenship shall complement rather than replace national citizenship (Art. 2.9), which means that citizens from EU member states are simultaneously national and EU citizens.
The idea that a citizen can be a member of various political communities is not new. Easton describes how different levels of political community can be nested within each other: ‘For a person to say that he is a Parisian, a Frenchman, and a European indicates three different levels of political community to which he simultaneously adheres’ (Reference Easton1965: 181). In this sense, the member states represent the lower-level political communities that are nested within the higher-level political system of the EU. To differentiate between these different levels, Easton suggests considering which political level is responsible for resolving particular political issues. Following this argument, in the European public sphere, citizens from EU member states can be identified as EU citizens and differentiated from national citizens by reference to EU governance. All citizens from EU member states who appear in EU-related news coverage are thus by definition EU citizens. However, EU citizens can be presented in the news coverage in two different ways: either directly related to the EU level or more closely linked to the member states. In case of the latter, the EU citizenry appears to be fragmented into the citizens of the EU member states. In this study, this type of EU citizen is called national EU citizen – for example, French citizens who vote in European Parliament elections. They can be identified as EU (not national) citizens because they participate in the European (not national) political community, and are from France.
The other form of representation of EU citizens in the news coverage accounts for an integrated European citizenry that will be henceforth referred to as the supranational EU citizens. The member state of origin is unknown for these citizens, who are exclusively defined by a reference to EU governance – for example, ‘European voters’ who cast a vote in the European election. The European voters are EU citizens and as such entitled to participate in EU governance. To be entitled to that right, one needs to be a native citizen from one of the member states. However, for supranational EU citizens the member state of origin is unknown; instead, they account for the EU citizens as a collective. In this sense, supranational EU citizens represent a truly European citizenry and comprise the citizens of the EU as a whole. From a normative perspective, supranational EU citizens denote a more advanced and integrated European citizenry, because they represent the citizens of all EU member states at once and are directly related to the EU level, while references to national EU citizens simultaneously link to national sentiments and only account for certain fragments of EU citizens.
If previous studies have taken EU citizens into account, they have usually done so in the context of attitude research (see, e.g. Gabel, Reference Gabel1998; Hooghe and Marks, Reference Hooghe and Marks2005; Boomgaarden et al., Reference Boomgaarden, Schuck, Elenbaas and de Vreese2011). There are currently only four substantial empirical projects that partially address how (EU) citizens are represented in the European public sphere.Footnote 1 These studies have either investigated references to ‘the Europeans’ as indicators of the visibility of a European identity in EU news (Sifft et al., Reference Sifft, Brüggemann, Kleinen-von Königslöw, Peters and Wimmel2007; Kleinen-von Königslöw, Reference Kleinen-von Königslöw2010; Hepp et al., Reference Hepp, Brüggemann, Königslöw, Lingenberg and Möller2012)Footnote 2 or examined the visibility of news frames that refer to the EU and its citizens (Schuck et al., Reference Schuck, Vliegenthart, Boomgaarden, Elenbaas, Azrout, van Spanje and de Vreese2011). Other studies have considered the responses of citizens to EU online news (de Wilde et al., Reference De Wilde, Michailidou and Trenz2013; see also Michailidou and Trenz, Reference Michailidou and Trenz2010) or focussed on citizens’ visibility in the transnational European public sphere (Garcia-Blanco and Cushion, Reference Garcia-Blanco and Cushion2010). Another study by Wessler (Reference Wessler2007) (see footnote 2) compared different actor types that become visible in discursive (e.g. comments) EU news or stories on other EU member states and found that ‘ordinary citizens’ are almost invisible in the news (below 3%), which is seen as a result of the high visibility of governmental actors that push other groups of actors into the background. However, as non-EU news items are included in the sample, it cannot be concluded that citizens who become visible are, in fact, EU citizens when defined by a reference to EU governance. Furthermore, as only discursive news stories are included in the sample, it cannot be assumed that the results can be generalized to all EU news coverage.
Overall, there is a lack of research on EU citizens, and clear definitions of what constitutes an EU citizen (in contrast to a national or non-European citizen) in the European public sphere are so far missing. In order to identify EU citizens in the European public sphere, this paper focusses on media coverage of EU governance. The next section presents the research questions and hypotheses regarding the visibility of EU citizens in the European public sphere.
Research questions and hypotheses
The previous section highlighted the relevance of the European public sphere and the visibility of EU citizens as actors. The initial question addressed by this study is: Are national and supranational EU citizens, as defined in this study, visible actors in EU news? If so, are they equally visible across EU news of the EU member states? While these research questions are rather explorative, they provide the basis for further explanatory investigation. The follow-up research question is: How can potential cross-country differences in the visibility of EU citizens be explained, and what factors foster their visibility? Related hypotheses regarding this explanatory part of the paper are presented in the following and are divided into three parts, concerning (1) media-specific factors, (2) a member state’s relationship with the EU, and (3) aspects at the national level that are independent of EU governance.
Medium
Studies on the European public sphere typically examine the extent to which different degrees of ‘Europeanness’ of the news coverage, in terms of EU topics and actors, can be explained by media characteristics. This paper does not aim to analyse the degree of Europeanization, but to explain the visibility of EU citizens within EU news coverage. Nevertheless, there might be variance between different media types and outlets. Because television news coverage is generally more personalized than the print media (Bentele and Fähnrich, Reference Bentele and Fähnrich2010: 54–56) and relies to a greater extent on episodic framing (Iyengar, Reference Iyengar1991), EU citizens might be more visible on television. Personalization refers to news that is presented ‘in terms of individual people, rather than abstractions’ (O’Neill and Harcup, Reference O’Neill and Harcup2009: 165). As a stylistic device, journalists use personalization to reduce complexity and help their audience understand the information more easily (Bentele and Fähnrich, Reference Bentele and Fähnrich2010), for example, by relating news to people’s everyday life. Episodic framing, inter alia, portrays news in terms of individual incidents rather than providing a broader picture and background information. Because of the more pronounced tendency towards personalization and episodic framing of television news coverage, the first hypothesis is that (Hypothesis 1) national and supranational EU citizens are more visible in television news than in newspaper coverage. Furthermore, previous research has shown that the degree of Europeanization varies between media outlets. The findings indicate that quality media is generally more Europeanized than popular media. Quality media tend to report more EU news (e.g. de Vreese et al., Reference De Vreese, Banducci, Semetko and Boomgaarden2006; Kleinen-von Königslöw, Reference Kleinen-von Königslöw2010) and the visibility of EU-level actors is higher in quality than popular media (Peter and de Vreese, Reference Peter and de Vreese2004; de Vreese et al., Reference De Vreese, Banducci, Semetko and Boomgaarden2006). This might also apply to the visibility of EU citizens. Therefore, it is expected that (Hypothesis 2) national and supranational EU citizens are more visible in quality than in popular media outlets. As mentioned above, supranational EU citizens represent an image of a more integrated European citizenry. If quality media has an effect on the coverage of EU citizens, it is more likely to be found for supranational than for national EU citizens.
EU migrant and European identity
One crucial aspect of EU citizenship is the right to move and reside freely within the territory of the EU member states (Treaty on the Functioning of the European Union, Art. 21.1). While it has been argued that ‘citizens alone enjoy an unconditional right to remain and reside in the territory of a state’ and that ‘the modern state has fundamental interest in territorial closure’ (Brubaker, Reference Brubaker1992: 24), this no longer holds true for EU member states. One aspect that might determine the extent to which citizens are aware of EU citizenship in the first place, but also the visibility of EU citizens in the news media, is the number of fellow EU citizens living in a member state. The higher the number of EU migrants, the more interaction and exchange takes place among EU citizens. As Putnam notes, immigration can have a negative effect on the cohesion of a society, but it also has the potential to construct new and ‘more encompassing identities’ (Reference Putnam2007: 139). Therefore, it is expected that (Hypothesis 3) the number of EU migrants living in an EU member state is positively related to the visibility of supranational EU citizens in the EU news coverage.
Another aspect that is related to the number of fellow EU citizens living in a country is European identity, because ‘immigrants in Europe (…) can challenge nationality as an exclusive reference point’ (Delanty, Reference Delanty1997) of citizenship. Citizenship and identity are closely related concepts (Isin and Wood, Reference Isin and Wood1999). Research findings indicate that citizens of EU member states with strong national identities are less supportive of European integration (Carey, Reference Carey2002; Hooghe and Marks, Reference Hooghe and Marks2004). The flipside of the coin is that, the more developed a European identity among the citizens of a member state, the more supportive its citizens are of the EU. The notion of a European identity is more closely related to the concept of supranational EU citizens, as they represent a truly European citizenry without any reference to nationality. It can also be assumed that in states with a more advanced European identity, the mediated image of a supranational, European citizenry is also more prevalent in the news media coverage. Hence, this study argues that (Hypothesis 4) the more developed a European identity, the more visible supranational EU citizens are in the EU news coverage.
Active national citizenry
In addition to media and country-specific factors that are related to the EU, national-level factors that are independent of EU governance may impact citizens’ general visibility in the public sphere. As such, this research takes associational membershipFootnote 3 as an indicator of active citizenship at the national level into account. As intermediary organizations, voluntary associations facilitate between citizens and governments (van Deth and Kreuter, Reference Van Deth and Kreuter1998) and play an important role in determining a society’s degree of civic involvement. The more participatory a society is, the more visible and better represented citizens are in the political process in terms of politicians being able to perceive their demands. Journalists might also become more aware of citizens’ actions and perceive them as relevant actors in the political process. By being socialized in a society with an active citizenry, journalists might be predisposed to see citizens as relevant political actors who should have a say in public debates. Linking back to the visibility of EU citizens, member states with a more active citizenry that participates in the political process might include citizens in public debates to a greater extent because journalists are generally paying more attention to citizens as political actors. Yet, references to supranational EU citizens measure a truly European and more abstract concept of EU citizenship that is expected to be more closely associated with EU-related factors. An active national citizenry at the member state level should be less relevant. Hence, the hypothesis is that only (Hypothesis 5) national EU citizens are more visible in EU member states with higher degrees of associational membership.
Data and methodology
To analyse the visibility of EU citizens across the European public spheres of the EU member states, this study draws on a large-scale content analysis of the 2009 European Parliament election carried out by the Providing an Infrastructure for Research on Electoral Democracy in the European Union (PIREDEU) project (Schuck et al., Reference Schuck, Xezonakis, Elenbaas and Banducci2010). The data have been analysed previously, for example, by Boomgaarden et al. (Reference Boomgaarden, Vliegenthart, de Vreese and Schuck2013), to examine variation in EU news coverage across countries and media. Yet, no previous study has investigated which factors can explain potential variation in the visibility of EU citizens in the news coverage. For explanatory variables on the country level, the content analysis is supplemented by Eurobarometer and European Values Study data as well as Eurostat statistics.
European Parliament elections are an ideal setting for research on the European public sphere, as they take place simultaneously and are relevant events for all member states. Furthermore, European elections are crucial as they allow citizens to directly participate in EU governance. This arguably also impacts the visibility of EU citizens in the European public sphere during this time period. As it is EU citizens who determine the outcome of an election, their visibility in the news coverage is arguably higher during elections than between them. Therefore this analysis is carried out under favourable conditions that probably increase the visibility of EU citizens in the EU news.
Sample
The 2009 PIREDEU media study includes content analysis data from all 27 member states, which means that it provides an inclusive image of the mediated European public sphere. For each country, two broadsheets and one tabloid newspaper, and two television news programmes from the most widely watched public and private television stations, were selected for coding.Footnote 4 Depending on the election day in each country, the sample period covers the 3 weeks from 14 May to 4 June up to 17 May to 7 June (for more details, see the data documentation report in Schuck et al., Reference Schuck, Xezonakis, Elenbaas and Banducci2010). For the purpose of this study, the data set was limited to EU news stories that discuss EU affairs extensively, or the main focus of which is EU policy, polity, or politics. This step was taken to maximize the likelihood that citizens are mentioned in the context of EU governance, as such a reference is crucial for the definition of EU citizens.Footnote 5 This leaves a total of 12,850 news stories. Although this approach might include some wrongly classified cases, it was chosen as the data provide a unique opportunity to focus on variations across EU member states. For the PIREDEU data set, the actor coding and the visibility of EU news in the media proved to be reliable (cf. Schuck et al., Reference Schuck, Vliegenthart, Boomgaarden, Elenbaas, Azrout, van Spanje and de Vreese2011). Hence, the analysis can be expected to derive reliable results.
Method
To examine the visibility of EU citizens in the EU news, the analysis is conducted at two different levels. The initial descriptive analysis is carried out at the news story level (N=12,850). For the explanatory analysis, the data were aggregated to the media outlet level. Thus, the media outlet is the unit of analysis, and the analysis is based on 143 cases. The outlet level was chosen in order to account for the different levels of visibility of EU citizens, which are influenced by media-specific characteristics. Even though media outlets might share common characteristics across countries, it is likely that outlets from the same country are not independent of each other, which potentially leads to biased ordinary least squares (OLS) estimates that could distort the results (Arceneaux and Nickerson, Reference Arceneaux and Nickerson2009). Therefore, this study uses a multi-level approach that applies an OLS regression with random effects estimation and robust standard errors (cf. Schuck et al., Reference Schuck, Vliegenthart, Boomgaarden, Elenbaas, Azrout, van Spanje and de Vreese2011). The model is able to account for the clustered structure of the data set. The number of clusters is limited to 27 to coincide with the total number of EU member states in 2009, which is considered sufficient to derive reliable estimates (Arceneaux and Nickerson, Reference Arceneaux and Nickerson2009).
Measures: dependent variablesFootnote 6
An EU citizen has been defined as an ordinary person from an EU member state when matters of EU governance are concerned. Hence, this study explicitly excludes intermediary organizations (i.e. political parties and civil society associations) that represent and mediate civic interests in the public sphere, because it assumes that from a normative point of view, the citizen should be directly visible in the European public sphere, especially during election periods. The operationalizations of the dependent variables are presented next.
National EU citizens: In the case of national EU citizens, EU citizens are divided into nation-specific subgroups. In the data set, up to six actorsFootnote 7 (i.e. persons, groups, institutions, or organizations) were coded for each news story if they were referred to at least twice. An actor can be persons, but also groups, institutions, or organizations. The first actor coded is the main actor in the story, which was identified by the number of mentions and general importance in the news item. The remaining actors were coded in order of their appearance. To operationalize national EU citizens, I drew on the PIREDEU actor scheme of the content analysis codebook, in which for each EU member state, ‘non-organized ordinary citizen(s), non-organized population groups or the population of a country as a whole (e.g. voters, people or the public; for television, e.g. country’s ordinary citizens being interviewed)’ were coded. Together they are operationalized as national EU citizens in this study. A categorical variable was coded 1 if a news story included national EU citizens as actors, and 0 otherwise.Footnote 8 In the explanatory analysis, national EU citizens account for the percentage of EU news stories by media outlet that includes national EU citizens as actors.
Supranational EU citizens: This variable refers to an overarching European citizenry, within which the respective citizen or group of citizens cannot be distinguished by nationality. To operationalize supranational EU citizens, I drew again on the PIREDEU actor scheme. At the EU level, ‘non-organized ordinary citizen(s), non-organized population groups or the EU population as a whole (e.g. Europeans, European voters, European citizens, European population or European public)’ were coded, which are operationalized as the supranational EU citizens in this study. A categorical variable was coded 1 if a news story included supranational EU citizens as actors, and 0 otherwise. In the explanatory analysis, supranational EU citizens account for the percentage of EU news stories by media outlet that includes this actor type.
Measures: independent variables
TV: A categorical variable separating television from newspapers was created and coded 1 for TV channels and 0 for newspapers to analyse whether there are differences in the visibility of EU citizens in different types of media.
Quality media: Broadsheets and public broadcasters were grouped together as quality media, while tabloids and private television channels were classified as popular media outlets. A categorical variable was coded 1 for quality media and 0 for popular media.
EU migrants: Using Eurostat data, this variable indicates the share of the foreign population that lived in 2008 in an EU member state and had the citizenship of one of the fellow EU member states.
European identity: To measure the strength of a European identity, this study follows previous research (cf., e.g. Hooghe and Marks, Reference Hooghe and Marks2005) that has largely relied on the Eurobarometer question (Eurobarometer 67.1):Footnote 9 ‘In the near future do you see yourself as [NATIONALITY] only, [NATIONALITY] and European, European and [NATIONALITY] or European only?’ and takes into account the percentage of respondents that see themselves as exclusively (European only) or predominantly European (European and [NATIONALITY]).
Associational membership: Associational membership is measured as the percentage of the population of the EU member state that is a member of at least one voluntary association. The data are drawn from the European Values Study (EVS 2008), based on the question: ‘Please look carefully at the following list of voluntary organizations and activities and say (a) which, if any, do you belong to?’.
Controls
Length of EU membership: This variable measures the number of years a country had been an EU member as of 2009.
Results
The first research question asked whether national and supranational EU citizens are at all visible in mediated discussions on EU governance. The answer is yes; as can be seen in Table 1, both concepts are present in EU news. Yet, references to national EU citizens (20%) clearly outweigh the visibility of supranational EU citizens (5%). To assess whether the visibility of EU citizens as actors is high or low, it is helpful to compare their visibility with other actor groups within EU news coverage. Not surprisingly, and in line with previous research, governmental actorsFootnote 10 from the EU and national level dominate the news coverage. While the level of visibility of supranational EU citizens seemed initially relatively low, the findings show that when compared with other EU-level actors, their representation is still higher than that of political groups of the European Parliament and EU-level civil society actors, including interest and professional groups, as well as individual actors such as activists, experts, journalists, and the media. What is remarkable with regard to national EU citizens is that they are more visible than national civil society actors and members of (and candidates for) the European Parliament, who should be among the key actors of the election coverage.
Visibility of actor groups in European Union (EU) newsFootnote a
MEP=Members of the European Parliament; MP=Members of the Parliament.
a For the visibility of actor groups, I use categorical variables indicating whether or not a certain actor was mentioned in a news story. As more than one actor was coded per news story, the results do not add up to 100%.
Turning now to the second research question, which concerns cross-country differences in visibility, Figure 1 shows considerable differences in the visibility of national and supranational EU citizens across the EU news coverage of the EU member states. Furthermore, the presence of national EU citizens does not follow the same pattern across countries as supranational EU citizens.
The visibility of national and supranational EU citizens across EU member states.
The visibility of national EU citizens in EU news varies from about 8 to 39%. While visibility is lowest in Italy, Germany, and Lithuania, national EU citizens are most present in the news coverage of Finland, Latvia, and Estonia. The visibility of supranational EU citizens is in all countries significantly lower, and is even close to non-existent in some countries (e.g. Lithuania), while they are included in a maximum of 14% of news stories in other states such as Romania. How can these differences be explained?
Explaining the visibility of EU citizens in EU news
A multi-level analysis was carried out to explain the variation in the visibility of EU citizens in the European public spheres of the EU member states. The results of the two models (Table 2) indicate that there are significant differences in the factors that explain the visibility of national EU citizens compared with supranational EU citizens. According to Hypothesis 1, both national and supranational EU citizens are more visible on television than in print media due to the more extensive use of personalization and episodic framing. The empirical analysis for national EU citizens shows that television has a significant positive impact on visibility, but the same cannot be said for supranational EU citizens. The second media-related hypothesis argued that the media outlet might have an effect, and national and supranational EU citizens were expected to be more visible in quality media outlets (Hypothesis 2). Yet, the media outlet does not significantly influence visibility, and there is no evidence that quality media have a leading role in promoting the image of the supranational EU citizenry.
Explaining visibility of national and European Union (EU) citizensFootnote a
Table 2 shows the results of the multi-level regression model where Level 1 is the media outlet and Level 2 the EU member state. Robust standard errors in parentheses, clustered by country in the RE specifications. Calculations were made using the xtreg option in Stata 13.
a To test the robustness of the results, instead of the TV and quality media dummy, separate dummy variables for the respective media outlets were tested. The results in both models are very similar. European identity was replaced by the respective measure for the strength of national identity, which is producing very similar results, but has a significantly negative effect on the visibility of supranational EU citizens, which supports the hypothesis. The length of EU membership was replaced with a dummy variable that was coded as 1 for EU member states who joined the EU until 1995, while EU member states that joined the EU after 1995 were coded as 0.
*P<0.05; **P<0.01; ***P<0.001.
At the country level, it was expected that a higher number of EU migrants would produce more interactions among EU citizens and increase awareness of EU citizenship, which would be reflected in the media coverage. Therefore, it was expected that the number of EU migrants living in an EU member state would be positively related to the visibility of supranational EU citizens in EU news (Hypothesis 3). Contrary to this initial expectation, the results show that the share of EU migrants has a significantly negative effect on the visibility of national EU citizens, while no significant effect was found for the visibility of supranational EU citizens. As mentioned previously, national EU citizens represent a European citizenry that has links to national sentiments. One potential explanation for the negative effect might be that a higher share of EU migrants leads to a reporting style that is less centred on the nation state and nationalities. However, as this does not go hand in hand with an increase in the visibility of supranational EU citizens, it cannot be said with certainty.
Turning to the role of European identity, Hypothesis 4 stated that a more developed European identity is positively related to the visibility of supranational EU citizens. The results suggest that a European identity is significantly and positively related to the visibility of both supranational and national EU citizens. Hence, a more developed European identity seems to generally improve the visibility of EU citizens in the news coverage. The extent to which a European identity has emerged in an EU member state is the only significant factor found that can explain the visibility of supranational EU citizens.
Besides these differences, variance in visibility might be a cause of country-level characteristics that are independent of EU-related factors. Hypothesis 5 predicts that higher membership rates in voluntary organizations have a positive effect on the visibility of national EU citizens, as such rates indicate a more active citizenry that participates in the political process. Due to socialization processes, journalists of EU member states with higher numbers of associational memberships might include citizens to a greater extent in public debates. The results show that higher rates of associational membership increase the visibility of national EU (but not supranational EU) citizens – which supports Hypothesis 5.
The control variable for the length of EU membership also has a significant effect on visibility. The results show that the longer the EU membership, the lower the visibility of national EU citizens in the EU news coverage. One potential explanation for this finding might be the context of the 2009 European Parliament election campaign. For two new EU member states, Bulgaria and Romania, the 2009 elections were the first European elections their citizens participated in. As voters from the new member states were less familiar with European Parliament elections, media in those countries might have paid more attention to voters during the election campaign.
Taking both models into account, it can be concluded that both the media and country-level factors influence the visibility of EU citizens in the EU news. Yet, cross-country differences seem more important than media-specific characteristics, as the amount of the variance explained is larger on the country than the media outlet level in both cases.
Discussion and conclusion
This study set out to determine whether EU citizens are visible actors in the European public sphere. This question has not been sufficiently answered by previous research, but is highly relevant to the accountability and legitimacy of EU governance. This study has shown that the visibility of EU citizens is not marginal when compared with other actor groups at the national and European level. The presence of EU citizens even exceeds the visibility of other actors in the European public sphere, such as members of the European Parliament and civil society actors. This study has not only provided empirical evidence for the visibility of EU citizens in the European public sphere, but has also contributed to its theoretical conceptualization by introducing the distinction between the concept of the national EU citizen and supranational EU citizen. Both concepts proved to be present in the EU-related news coverage, yet references to national EU citizens clearly prevailed. This indicates that a truly European citizenry is still underdeveloped, at least concerning its visibility in the European public sphere.
Yet, the results have also shown that the visibility of both national and supranational EU citizens varies strongly across the EU member states. This finding is important, as citizens of member states with high levels of visibility of EU citizens in the news coverage are more likely to perceive that their opinions are taken into account in the decision-making process at the EU level and, as such, develop further confidence in the EU. Citizens of EU member states with low levels of visibility of EU citizens, on the other hand, are more likely to see the EU as an elitist institution running the risk that their confidence in the EU will decrease. When assuming that the visibility of EU citizens in the news coverage is important to facilitate exchange between the citizens of the EU member states and decision makers at the EU levels (for a similar argument see, e.g. Milbrath, Reference Milbrath1972), then identifying factors that can foster their visibility is important.
The explanatory analysis has shown that national EU citizens are more visible on TV, and in EU member states with higher rates of associational membership and more developed European identities. The strength of a European identity has also a significantly positive effect on the visibility of supranational EU citizens. European identity is the only explanatory factor that is relevant for the visibility of both national and supranational EU citizens. This indicates that where EU citizens see themselves more as a group and as part of the European political community, the visibility of national, but also the more advanced concept of supranational EU citizens increases. Yet, this study has also identified that there are factors that can hinder the visibility of EU citizens in discussions on EU governance. For national EU citizens, the results show that their visibility is lower in old EU member states and in countries with higher levels of migrants.
Higher levels of visibility of EU citizens ought to have positive effects on the responsiveness of EU governance, as politicians are better able to perceive the preferences and concerns of their citizens. Therefore, fostering a European identity and feeling of solidarity (especially in EU member states with higher levels of migration) among citizens can improve the visibility of EU citizens in the mediated European public sphere and it might, in the long run, have a positive impact on the legitimacy of the EU. While the visibility of both national and supranational EU citizens is relevant for an exchange between the EU and its citizens, these findings underline the importance of civic elements for European integration and for the emergence of a truly European and cross-national understanding of EU citizenship.
A limitation of this study is that it only examined EU citizens in the European public sphere during a single European Parliament election, which probably amplifies their visibility. Analytically, it only included EU news articles with a dominant or extensive focus on EU affairs in order to increase the likelihood that national EU citizens would appear in the context of EU governance. As a result, this approach might include some wrongly classified cases, but it was chosen as the advantages of a data set including all 27 EU member states outweigh the disadvantages. Nevertheless, this study provides new and important insights into the presence of one of the central actors in the EU news: EU citizens. Overall, the relative high degree of visibility can be interpreted as a positive sign for the European public sphere. Yet, further investigation is needed that (1) considers the presence of EU citizens in the European public sphere during routine periods of EU governance and (2) examines whether EU citizens are given an active voice with which to articulate their opinion in the news coverage, or if the media simply use them to set the scene.
Knowing whether EU citizens are present in the European public sphere is important with regard to the legitimacy of EU governance. As contemporary politics is primarily mediated (Bennett and Entman, Reference Bennett and Entman2001), the visibility of EU citizens in the European public sphere is an important precondition for adequate communication between the EU and the citizens of its member states. The results of this study also indicate the importance of an active citizenry at the national level for the visibility of EU citizens. EU citizenship therefore requires more than a top-down legal framework. For EU citizenship to become a politically meaningful concept, there must be a general feeling of belonging to the EU and the active participation of the citizens of the EU member states. In the long run, a visible and active EU citizenry in the European public sphere might help to overcome the gap between the EU and its citizens.
Acknowledgements
The author would like to thank Rüdiger Schmitt-Beck, Hartmut Wessler, Patricia Moy, and Christopher Wlezien for their feedback on earlier versions of this paper. The author also thanks Yannis Theocharis, Lucy Kinski, Anne Schäfer, Mona Krewel, and Julia Partheymüller for their helpful comments. The author is grateful for the feedback and suggestions provided by three anonymous reviewers and the Editors of European Political Science Review, as well as to the PIREDEU team at the University of Amsterdam.
Supplementary Material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1755773915000363
