Scholarly article on topic 'Shifting Europe's Boundaries: Mass Media, Public Opinion and the Enlargement of the EU'

Shifting Europe's Boundaries: Mass Media, Public Opinion and the Enlargement of the EU Academic research paper on "Political Science"

0
0
Share paper
Academic journal
European Union Politics
OECD Field of science
Keywords
{""}

Academic research paper on topic "Shifting Europe's Boundaries: Mass Media, Public Opinion and the Enlargement of the EU"

European Union Politics

DOI: 10.1177/1465116508089087 Volume 9 (2): 243-267 Copyright© 2008 SAGE Publications Los Angeles, London, New Delhi and Singapore

Shifting Europe's Boundaries:

Mass Media, Public Opinion and the Enlargement of the EU

^ Jürgen Maier

Kaiserslautern University of Technology, Germany

+ Berthold Rittberger

University of Mannheim, Germany

ABSTRACT

This article demonstrates that public attitudes towards EU enlargement are strongly affected by exposure to the mass media. It reveals 'priming' effects by showing that media exposure affects the standards by which individuals evaluate the accession of potential candidate countries. To gain a more refined understanding about media effects on enlargement attitudes, we analytically separate three different factors that underlie EU enlargement support for a given candidate country: its economic performance, its state of democracy and its perceived cultural 'match' with the EU. Employing an experimental design, we probe the media-induced effects of these factors on EU enlargement attitudes.

KEY WORDS

■ enlargement

■ experiment

■ mass media

■ public opinion

Introduction

The accession of Bulgaria and Romania has not put an end to the enlargement agenda of the European Union (EU). Croatia, the Former Yugoslav Republic of Macedonia and Turkey are three candidate countries already on the (bumpy) road to accession. In the Western Balkans, Albania, Bosnia and Herzegovina, Montenegro and Serbia are classified by the EU as 'potential candidate countries'. Furthermore, for countries in the EU's 'neighbourhood', such as Moldova or Ukraine, EU accession in the medium to long run is not out of the question.

The recent enlargement rounds have sparked a broader debate in the media as well as in academia about the EU's (ultimate) boundaries (see, among others, Flora, 2000; Gerhards, 2005; Zielonka, 2005). The discussion about the possible accession of Turkey to the EU is a case in point. Yet we know very little about the conditions under which EU member states support or oppose enlargement to a particular applicant state (see Schimmelfennig and Sedelmeier, 2005: 6, 21). Thus, the more countries accede to the EU, the more pressing is the question about 'where to draw the line'. What prompts EU member states to 'shift' the EU's geographical border?

The EU enlargement literature is strongly centred on elite politics and preferences and has only recently begun to explore public attitudes towards enlargement either in the EU member states (see e.g. Jones and van der Bijl, 2004; Karp and Bowler, 2006; McLaren, 2007) or in accession countries (see e.g. Vetik et al., 2006). Yet, in order to address the question about the borders of the EU, exploring the determinants of public attitudes towards EU enlargement is essential. In light of the evaporating 'permissive consensus' (Lindberg and Scheingold, 1970) - which has been identified for some time (see e.g. Reif, 1993; Hooghe and Marks, 2006) - the political relevance of public opinion for the future of European integration is likely to be paramount, a fact that was forcefully demonstrated by the double rejection of the Constitutional Treaty by referendum in two of the 'founding states', France and the Netherlands. Since leading politicians across different EU member states have pledged to make EU accession of current candidate countries conditional upon a positive referendum outcome in the home country, most EU member states, especially those intending to ask the public about the issue (e.g. Denmark, Sweden and the UK), delayed the ratification process. Moreover, in 2006 the plans for a Constitutional Treaty were officially put on hold by the European Council. Hence, public attitudes towards EU integration in general, and EU enlargement in particular, play an increasingly important role in the enlargement decision process, which has hitherto been dominated by political elites in the national capitals.

In order to explain public support for the EU and European integration, the literature emphasizes two types of determinants. According to the first perspective, citizens base their support on an evaluation of the economic consequences of EU membership. Support for integration thus stems from individual cost/benefit calculations. Those who perceive more advantages than disadvantages - either for their own 'pocket book' or for the national economy - are more likely to support European integration than are those who expect integration to offer a negative cost/benefit balance (see e.g. Hooghe and Marks, 2005: 420-2; Kaltenthaler and Anderson, 2001: 144). In fact, numerous studies show that macroeconomic conditions have a significant impact on EU support levels (see e.g. Anderson and Reichert, 1996; Eichenberg and Dalton, 1993; Gabel, 1998a, 1998b, 1998c; Gabel and Palmer, 1995; Sánchez-Cuenca, 2000). According to the second perspective, support for European integration is conditioned by citizens' affective relationship with their nation-state as well as their orientation towards the EU (see e.g. Kaltenthaler and Anderson, 2001). There is ample evidence that the level of support for the EU and the European integration process depends on individual attitudes towards one's nation, the level of support for vertical integration (i.e. to support the transfer of sovereignty from national to EU level) and more general experiences of the EU (see e.g. Eichenberg and Dalton, 1993; Franklin et al., 1994; Inglehart, 1977; Martinotti and Steffanizzi, 1995; Shepherd, 1975). However, the direction of the impact on EU support levels is not always clear cut. For instance, the relationship between national identity and support for European integration can pull in opposite directions: on the one hand, national identity and European integration can be reinforcing; on the other hand, national identity can also undermine support for European integration (see Hooghe and Marks, 2004; Risse, 2003, for an overview).

With few exceptions the literature on public opinion and EU integration has thus far neglected the role of media effects on attitudes towards European integration. This is striking, since ordinary citizens usually do not have 'firsthand' experience of the EU, but depend on mass media coverage for information. This holds especially true for those levels of the political system that are perceived as 'far away' from everyday life - first and foremost national politics as well as supranational politics (see Norris, 2000). Even though the public opinion literature increasingly acknowledges the relevance of 'cues' in activating or priming citizens' underlying interest in and valuing of EU integration (see Hooghe and Marks, 2005; Steenbergen et al., 2007), the linkages between 'cue-givers', such as political parties, political leaders and institutions, and the 'cue-recipients' remain largely under-explored (see Carey and Burton, 2004, for an exception). These linkages are provided by the mass

media. As we will demonstrate below, the mass media have a significant effect on attitudes towards enlargement.

Public opinion and EU enlargement

The literature on public opinion and EU enlargement takes the two 'traditional' explanatory routes mentioned above as points of departure, emphasizing the economic costs and benefits of enlargement as well as the affective relationship with one's own country and the EU. In one of the first applications of the EU-related public opinion literature concerning the question of enlargement attitudes, Karp and Bowler (2006) advance three sets of arguments - pertaining to respondents' identification with the EU, the costs and benefits of EU membership and satisfaction with the EU institutions - in order to explain public attitudes towards EU enlargement. Although Karp and Bowler (2006) find evidence for all three types of argument, they demonstrate that instrumental self-interest is the most powerful predictor for explaining attitudes towards EU enlargement.

Jones and van der Bijl (2004) present a transaction-affinities approach to explain public support for EU enlargement. Employing aggregate national response data from Eurobarometer, they explore 'whether the level of transactions between existing member states and specific candidate countries can explain support in the member states for the accession . . . of specific candidate countries' (Jones and van der Bijl, 2004: 334). The study finds that trading relationships between 'old' member states and accession countries and geographical proximity are the driving factors explaining levels of support. Jones and van der Bijl (2004: 347) also demonstrate that 'policy-relevant concerns - specifically in relation to agriculture' - have an impact on aggregate response rates. These findings are in line with the 'instrumentalist' argument whereby respondents weigh the expected costs and benefits of enlargement. Yet the study also shows that identity-related factors (historic relationships between countries) as well as cultural aspects (Catholicism) feature in explaining aggregate national support for EU enlargement.

In a recent study on public attitudes towards Turkish EU membership, McLaren (2007) advances different theoretical arguments to explain variation in citizens' levels of support for Turkish EU membership in different EU member states. One line of argumentation echoes previous studies by referring to respondents' economic self-interest as a determinant of attitudes towards Turkish membership. McLaren also introduces into the debate a novel identity-related argument that emphasizes group interests. She argues that 'explanations for differences in opinion about Turkish candidacy are

likely to relate to out-group rejection and its counterpart, in-group identity and protectiveness' (McLaren, 2007: 257; see also McLaren, 2002; De Vreese and Boomgaarden, 2005). In-group and out-group sentiments are connected to the resources possessed by a group, such as jobs and social security benefits, as well as cultural factors, such as symbols, myths and traditions (see McLaren, 2007: 257). Consequently, McLaren argues that both threats to resources and threats to culture should condition public attitudes towards Turkish membership.

Even though the reviewed literature has considerably improved our knowledge about the determinants of mass attitudes towards EU enlargement, Hooghe and Marks (2005) identify one general shortcoming in the existing literature on public attitudes towards the EU. They argue that, in order to become politically salient, the factors 'explaining' mass attitudes towards integration or enlargement need to be 'primed' (Hooghe and Marks, 2005: 424). Cue theory draws from cognitive and social psychology and examines 'how political cues - grounded in ideology or in elite communication - mediate the effect of economic calculation and community membership' (Hooghe and Marks, 2005: 420). It has been demonstrated that ideology and domestic political institutions, first and foremost political parties, shape and frame attitudes to European integration. According to Hooghe and Marks (2005: 425), 'public opinion is constrained by political ideology, political parties, and political elites in . . . domestic arenas'. Carey and Burton (2004) have advanced a similar argument. They rightly claim that existing studies exploring the determinants of public attitudes towards the EU commonly assume that expectations about the expected costs and benefits of EU integration are formed by 'cues', i.e. they 'filter through individuals through media, parties, trade unions, personal contacts, and so on, although these links are not directly investigated' (Carey and Burton, 2004: 638).

The missing link: Media effects and public opinion

It is obvious that the vast majority of citizens have no first-hand experience of politics, whether domestic, European or international. Consequently, citizens rely on the mass media to obtain information about political issues. But how do the mass media affect political attitudes? Research on political communication commonly employs three major processes in relation to attitude formation and change: agenda-setting, priming and framing (see Scheufele and Tewksbury, 2007, as well as Semetko, 2004, for overviews). The focus of this article is on framing and priming processes. According to framing theory, the characterization of an issue, an event or an actor by the

media influences the audience's understanding of the subject. At the macro level, framing refers to the modes journalists choose to present complex information in a practical way ('media frames'). By emphasizing certain aspects of an object while neglecting or suppressing others, particular issues are made 'more salient through different modes of presentation' (Scheufele and Tewksbury, 2007: 15). If those presentations match the often rather simple schemes people use for information-processing (so-called 'audience frames'), media frames can influence orientations at the micro level, i.e. the individual perceptions of a particular object as well as attitudes toward an object.

Through framing, the mass media highlight special aspects or attributes of an issue or object, thereby shaping its image. By framing information, the media are able to initiate priming processes at the micro level, thereby influencing the standards by which individuals evaluate objects or issues (Iyengar and Kinder, 1987: 63). The media have the power to shift those standards because individuals predominantly utilize information that has been most recently 'activated' and is therefore at the 'top of their heads' instead of employing the full range of cognitively available information to evaluate objects (Zaller, 1992: 48). As a result, the media tell citizens not only what to think about but also how to think about an issue or object.

The analysis of media effects on public opinion about the EU is an embryonic field of study. Most recently, De Vreese and Boomgaarden (2006: 421) note that it is 'striking that the relationship between media and public opinion on European integration . . . has remained so underdeveloped in the scientific literature'. This is even more surprising, considering that 'the vast majority of citizens across Europe repeatedly identify news media as their most important and preferred source of information about European integration' (De Vreese and Boomgaarden, 2006: 421).

With regard to the study of EU enlargement, there are presently two studies that explicitly address the question of how the media affect EU enlargement attitudes. In their pioneering study, De Vreese and Boomgaarden (2006) demonstrate that public attitudes about EU enlargement are conditioned by media coverage and, more particularly, by its evaluative content and the direction of this content, i.e. whether evaluations of EU enlargement are portrayed positively, negatively or mixed:

where the news media coverage was considerable in amount and positive in tone we found respondents ... to be gain-seeking and endorse the enlargement of the EU. In the situation where news media messages were less visible and mixed in character we did not find the news media to exert an influence on the dynamics of public opinion formation. (De Vreese and Boomgaarden, 2006: 430)

Schuck and De Vreese (2006) employ evaluations of EU enlargement in terms of risks and opportunities as 'valence frames' in order to assess the influence

of these frames on public support for EU enlargement. Using a multi-method design - their study employs both content analysis and experimental data -Schuck and De Vreese (2006: 21) find the 'risk frame' leads to lower levels of support for EU enlargement among respondents than the 'opportunity frame'.

Explaining EU enlargement attitudes

The use of valence frames has improved our understanding of media effects on enlargement attitudes, but we propose to employ a more differentiated set of frames. Instead of exclusively focusing on the effects of 'positive' and 'negative' issue-framing associated with EU enlargement, we argue that we are potentially losing a lot of information about the determinants of enlargement attitudes and the role of media effects through this form of issue amalgamation. By studying the effects of positive and negative evaluation of different sets of determinants to explain EU enlargement attitudes, we expect to gain a more refined understanding of the variable strength of media effects.

Following the public opinion literature discussed in the previous section, we assume that respondents' enlargement attitudes respond to different evaluations of socioeconomic as well as identity-related issues. With regard to socioeconomic issues, we argue that positive evaluations of a candidate country's economy and expected positive economic externalities of accession for the existing member states will have a positive effect on respondents' level of support. Conversely, when these evaluations are framed negatively, i.e. where the economic externalities of membership are considered to be negative and economic performance poor, we expect a negative effect on respondents' enlargement attitudes.

In the context of EU enlargement, identity-related issues matter for the construction of an 'in-group' and an 'out-group' or of a notion of 'us' versus 'them' (see McLaren, 2007). These intersubjective constructions lend themselves to the formation of symbolic boundaries, i.e. 'conceptual distinctions made by social actors to categorize objects, people, practices, and even time and space. They are tools by which individuals and groups struggle over and come to agree upon definitions of reality' (Lamont and Molnar, 2002: 168). Symbolic boundaries promote identification with an in-group by generating feelings of similarity and group membership. Symbolic boundaries are distinct from social or political boundaries: whereas the latter manifest themselves as groupings of individuals and holders of entitlements (such as citizenship rights), symbolic boundaries are established and sustained intersubjectively. Yet both types of boundaries 'should be viewed as equally real'. At the causal level, symbolic boundaries can be thought of as a necessary but

insufficient condition for the existence of social boundaries' (Lamont and Molnar, 2002: 169). By analysing the effect of identity-related issues on EU enlargement attitudes, we follow an important line of enquiry in order to assess the EU's symbolic boundaries.

How do we go about exploring the impact of identity-related issues on enlargement attitudes? Most indicators employed in research tapping the level of EU identity in public opinion have one fundamental shortcoming concerning issue amalgamation. We agree with Bruter (2003) that Eurobarometer offers poor indicators and data to tap European identity. He argues that '[w]hen two individuals claim to "feel European," they might mean totally different things in terms of both the intensity of the feeling they describe and the imagined political community they refer to' (Bruter, 2003: 1154). Taking recourse to the literature on identity-formation, we distinguish two approaches to explore European identity in order to gain a better understanding of 'what people mean when they claim to feel European or not' (Bruter, 2003: 1154).

First, essentialist theories of identity-formation place a strong emphasis on 'cultural background variables' (Cederman, 2001: 10). Essentialism is most commonly associated with the theory of ethno-nationalism. From this perspective, the nation-state continues to exercise a strong emotional pull based on common memories and common understandings of tradition, religion and culture (see Cederman, 2001: 14). Even though essentialism carries a strong connotation with nationalism, it is possible to conceptualize culture-driven processes of identity-formation with regard to non-nation-based political communities. A European identity based on culture 'is best described as individuals' perceptions that fellow Europeans are closer to them than non-Europeans' (Bruter, 2003: 1155). The bonds that hold a culture-based political community together 'include any form of common history; moral, religious, or ethnic traditions; philosophical, political, and moral norms and values' (Bruter, 2003: 1156). From the essentialist perspective, European identity can be conceived in pan-nationalist terms, i.e. an identity shared by fellow Europeans forming a distinct civilization with its own history, culture, tradition and religion (Cederman, 2001: 15-16; Huntington, 1996).

Second, constructivist theories of identity-formation start from the assumption that politics - not culture - is the driving force behind identity-formation. From this perspective, culture-based nationalism is a modern, 'historically contingent . . . political principle', which can be replaced by a 'thin' political culture that lends itself to the formation of what Habermas has labelled 'constitutional patriotism' (Cederman, 2001: 15; Habermas, 1992). From this post-nationalist perspective, cultural differences and national attachments, although 'real' because they are shared and agreed

upon intersubjectively, can be superseded and eventually replaced by a civic identity that is characterized by the degree to which individuals 'feel that they are citizens of a European political system, whose rules, laws, and rights have an influence on their daily life' (Bruter, 2003: 1155). With regard to civic identity, we expect that exposure to positive evaluations of the political system of a candidate country, such as its democratic performance, implementation of the rule of law and human rights standards, will have a positive effect on respondents' expression of support for EU accession (and vice versa). As for cultural identity, we expect that respondents who are exposed to positive evaluations of a candidate country's identification with Europe, as well as to reference to shared traditions and values, are likely to be influenced positively with regard to enlargement support (and vice versa). Based on the triple distinction between socioeconomic interests and cultural and civic components of respondents' identities, we expect to tap more finegrained differences in EU enlargement attitudes than are provided in the existing literature.

In sum, this article explores three interrelated propositions, which will be tested in the empirical section. First, we posit that media reception affects the attitudes of respondents towards the EU accession of Macedonia. Second, we hypothesize that positive information about Macedonia's economy, state of democracy and 'cultural match' with the EU affects respondents' attitudes towards Macedonian EU accession positively, and vice versa. Third, we hypothesize that media exposure affects the standards by which respondents evaluate the issue of Macedonian EU accession (priming). By providing information about a specific aspect of current developments in Macedonia (i.e. Macedonia's economy, culture and state of democracy), this information should be more important than other criteria in explaining individual support for the EU accession of Macedonia.

Method and design

This article builds and expands on existing research in the field of media effects and public opinion in the EU. It does so by analysing how media exposure affects public attitudes to EU enlargement and by uncovering the priming processes through which the standards employed by respondents for evaluating EU membership of neighbouring countries are altered.

The experimental approach taken in this article differs from existing research in this field. First, studies employing media effects usually link survey data on media exposure to mass attitudes or match survey data with data gained from content analysis. The measured impact of media reporting

on political attitudes is usually very small and often even insignificant, which is attributable to the methodology (see Zaller, 1992; Price and Zaller, 1993). Second, this article seeks to establish the causality between media effects and respondents' enlargement attitudes by focusing on priming processes. Although panel survey designs probe the impact of media coverage on support for EU enlargement, they establish an indication but not a proof of causality (see De Vreese and Boomgaarden, 2006). This proof can be established only by non ex post facto research designs, i.e. experimental designs (see e.g. Behnke et al., 2006: 81).

This study sets up a three-factor experimental pre-test/post-test design to make within- and between-condition comparisons in order to answer the question of whether and how media exposure affects the different standards - economic, political and cultural - through which respondents assess EU enlargement to neighbouring countries. The experiment was conducted at the Faculty of Social Sciences at the University of Kaiserslautern in an undergraduate lecture course. The pre-test was carried out in December 2006. The students completed a questionnaire comprising a wide range of questions on attitudes towards European integration and EU enlargement. The post-test was carried out six weeks later in January 2007. In total, 95 students participated in both the pre-test and the post-test.1 The respondents were not informed about the actual purpose of the study, they were not told that the first questionnaire would be followed up by a second survey, and we opted for a time lag of six weeks between the pre-test and the post-test to limit the problem of panel sensitization. One problem that potentially arises with pre-test/post-test designs is that media use in the interim period may have enhanced or mitigated some of the observed effects. Unfortunately, in the context of this study, we have no information on the media use of our respondents.

During the post-test phase, all participants were randomly assigned to one of seven groups (six experimental groups and one control group). Following a randomization check, we found that the groups did not differ significantly in terms of their social and political composition. Each respondent from the seven groups read a copy of a newspaper article. In the case of the six experimental groups, these entirely fictitious articles raised the issue that Macedonia might shortly start negotiations to join the EU. Macedonia (and not another actual or potential candidate country) was selected in order to stimulate reactions to media content for several reasons: we opted for a country for which levels of prior knowledge were rather low. Low levels of prior knowledge as well as the absence of reliable heuristics are preconditions to create credible experimental conditions. A number of countries do not meet these criteria because respondents have ample first-hand experience either as a result of a country's geographical proximity to Germany (e.g. Switzerland),

or because respondents have been exposed to media coverage about a particular country in the recent past (e.g. Turkey), or because they can make educated guesses about a country's economic situation, its democratic standards and its cultural values (e.g. Norway, Iceland). Macedonia meets these preconditions. Based on our measurement of factual knowledge in the post-test, only 6% of the sample were able to name the capital of Macedonia. In addition, only 40% knew one of Macedonia's neighbouring countries. Combining the answers to the two factual questions, it turned out that 59% had no knowledge about Macedonia at all.

Taking into account the three potential determinants of EU enlargement, we designed pairs of newspaper articles that addressed (1) the state of Macedonia's economy and trade relations, (2) the development of Macedonian democracy and rule of law, and (3) the cultural 'match' between Macedonia and current EU member states. For each of these three conditions, one article framed the issue negatively, the other positively.2 The control group was asked to read a non-fictitious newspaper article that made no reference to Macedonia, the EU or the European integration process.3 After reading the articles, the students were asked to complete a second questionnaire, which included most of the questions asked in the first wave. In order to ensure that our respondents paid attention to the three factors that we included in our stimulus material and understood their messages, the manipulated newspaper articles offered ample verbal references to Macedonia's economy, democracy and culture. These terms appeared in the subheadings of the respective articles as well as in the body text. During debriefing, we discussed the articles and asked questions about what our respondents had read and understood.

In order to measure the impact of mass media content on attitudes towards EU enlargement, we compared support for the EU membership of Macedonia expressed in the pre-test and the post-test (our dependent variable4) using a i-test for dependent samples and ANOVA techniques, respectively. To analyse the priming effects induced by the different newspaper articles, we related the level of support for Macedonia's EU membership to the evaluation of its economy, democracy and culture (our independent variables5) in both the pre-test and the post-test employing ordinary least squares (OLS) regression.

Results

Our findings offer support for the literature on media effects on attitudes towards EU enlargement (De Vreese and Boomgaarden, 2006; Schuck and De Vreese, 2006). The data analysis shows that exposure to the mass media

has an impact on support for Macedonia's accession to the EU (see Table 1). Respondents exposed to newspaper articles emphasizing that Macedonia's economy performs well, that democracy and the rule of law are firmly enshrined and that cultural values match those of the current EU member states display a significant increase in their support for Macedonian EU membership: the sceptical attitude vis-à-vis Macedonia's membership that we observed in the pre-test is replaced by a positive attitude in the post-test. In contrast, exposure to negative press reports across the three dimensions -economics, democracy and culture - leads to a significant decline in support: the level of opposition from the pre-test intensified in the post-test. As expected, the opinions of the control group did not change significantly.

The consequences of exposing respondents to different tones of media coverage are striking. Analysing the differences in support levels of the three different groups - respondents exposed to positive articles, respondents exposed to negative articles, and the control group - we obtain an unequivocal message: whereas the respondents' attitudes on Macedonia's accession to the EU are negative (on average) in the pre-test - with the differences between the groups not reaching levels of statistical significance - this picture changes dramatically in the post-test. Our media stimulus polarized individual attitudes on the issue of accession. This polarization, i.e. the scope and direction of attitude change between the pre-test and the post-test, is pronounced: the experimental groups differed significantly in their views on whether or not to admit Macedonia to the EU between the pre-test and the post-test.

Furthermore, it turns out that, independently of the issue dimension addressed in the press reports (economy, democracy, culture), negative information always leads to a decline in support for Macedonian EU accession (see Table 2). This finding strongly supports the studies carried out by De Vreese and Boomgaarden (2006) and Schuck and De Vreese (2006). In the case of our culture variable, the decline in support is statistically significant. In contrast, positive press reports generally resulted in an increase in support

Table 1 Impact of positive and negative press reports about Macedonia on attitudes towards Macedonian EU membership

Pre-test Post-test Difference p (t-test) N

Article: positive -.38 .10 .49 < .01 39

Article: negative -.29 -.67 -.38 < .05 45

Control group -.18 -.36 -.18 > .05 11

p (analysis of variance) > .05 < .01 < .01

Table 2 Impact of press reports about Macedonia's economy, democracy and culture on attitudes towards Macedonian EU membership

Pre-test Post-test Difference P (t-test) N

Economy: positive -.09 .55 .64 < .05 11

Economy: negative .07 -.21 -.29 > .05 14

Democracy: positive -.47 .33 .80 < .01 15

Democracy: negative -.47 -.67 -.20 > .05 15

Culture: positive -.54 -.54 .00 > .05 13

Culture: negative -.44 -1.06 -.63 < .05 16

Control group -.18 -.36 -.18 > .05 11

p (analysis of variance) > .05 < .01 < .01

for Macedonian EU membership. The only exception is the group exposed to information about a cultural match between the EU and Macedonia; here we recorded no change in attitudes. This does not mean, however, that the individuals assigned to this group did not change their opinions about this issue. In fact, they did: the correlation between attitudes in the pre-test and the posttest is only r = .59. Obviously, these changes are not systematically related to the information provided in the press reports.

Shifting from evaluative tone to content, i.e. probing for the impact of factors relating to Macedonia's economy, democracy and culture, we again find that mass media coverage polarizes attitudes. Analysing the variance in attitudes towards the EU membership of Macedonia among our seven different groups, we recorded no significant differences as a result of the pre-test. The difference between the most supportive and the least supportive group is .61. This pattern clearly changes after the post-test. Now the difference between these two groups is 1.61 and the seven groups differ significantly from each other. Content obviously increases the polarizing impact of the mass media. If we compare the variance of the dependent variable explained by the design used in Table 1 (i.e. controlling only for the evaluative tone of media reporting, i.e. 'positive' or 'negative' attributions) with the design in Table 2 (i.e. controlling for evaluative tone and content simultaneously), it turns out that the explanatory power of the two models does not substantially differ in the pre-test (n2 = .00 vs. n2 = .04), yet it does in the post-test (n2 = .11 vs. n2 = .24).

The impact of the mass media on attitudes toward the EU membership of Macedonia remains significant (p < .05 for democracy, and the effects for both economy and culture are almost statistically significant)6 once we try to explain support for EU enlargement in the post-test by including other

variables using OLS regression (i.e. support for enlargement in the pre-test, sex, age, identification with a party represented in the national parliament, interest in politics, interest in European politics, post-materialism, knowledge about Macedonia, attitudes toward Macedonian citizens, contact with Macedonian citizens, previous visits to Macedonia; see Table 3). Whereas the pre-test level of support for Macedonia's accession to the EU turns out to be a highly significant determinant of post-test support, no other variable accounts for the changes in attitudes. After this rigorous test, we can confidently claim that it was exposure to media information that caused the observed shift in the respondents' attitudes to further EU enlargement - and not any accidental differences in the composition of the assigned groups with regard to demography, political attitudes and values, cognitive and political involvement, or experiences of Macedonia and its inhabitants.

Exposure to the newspaper articles not only changed the level of support for Macedonia's accession to the EU, it also affected Macedonia's perceived image (i.e. a consolidated mental conception of Macedonia; see Table 4). As expected, exposure to news about a particular issue (such as Macedonia's economic situation, its democratic performance or its cultural 'match' with the EU) leads, first and foremost, to significant changes in how respondents evaluate that issue (see numbers in bold).7 For example, focusing on the

Table 3 Determinants of post-test support for Macedonian EU membership

R2 .40

Article: economy .37 (.18)

Article: democracy .51*** (.26)

Article: culture .35 (.18)

Sex -.15 (.22)

Age .01 (.02)

Strength of party identification -.05 (.06)

Interest in politics -.05 (.14)

Interest in European politics .04 (.15)

Knowledge about Macedonia -.09 (.17)

Sympathy toward Macedonians .07 (.14)

Contact with Macedonians -.34 (.20)

Previous visits to Macedonia -.13 (.26)

Post-materialism -.07 (.22)

Support for Macedonia's EU accession (pre-test) .48*** (.09)

Constant .00 (.53)

Note: OLS regression, unstandardized regression coefficients, standard errors in parentheses. Significance levels: *** p< .01

Table 4 Impact of press reports on Macedonia's image

Article Economy Democracy Culture Min. N

Economy: positive .82** .27 .27 11

Economy: negative -.64 -.50 -.36** 14

Democracy: positive .20 .80*** .33 15

Democracy: negative .07 -.67** -.13 15

Culture: positive .08 .31 .54 13

Culture: negative -.31 -.69** -.69*** 16

Control group -.09 .50 .27 10

Note: Significance levels - ** p < .05, *** p < .01

Macedonian economy provoked an update of the assumption of how the Macedonian economy performs. In this context, positive articles intensified positive attitudes and negative articles led to more negative evaluations.

However, these attitude changes resulting from exposure to one particular newspaper article did not relate only to the specific issue addressed (economy, democracy or culture); they also significantly affected evaluations of other aspects of Macedonia's image (at least partially): evaluations of Macedonia's economy, democracy and culture are strongly correlated with each other.8 The 'transfer' of evaluations from one dimension to another often occurs if negative information is provided. In other words, whereas exposure to positive information about the Macedonian economy resulted in more positive evaluations of the performance of the Macedonian economy, but neither improved evaluations of the state of democracy in Macedonia nor improved the image of a cultural match between Macedonia and the EU, exposure to negative information about Macedonia's economy tended significantly to intensify negative evaluations of this dimension as are evaluations of Macedonian democracy and the cultural 'match' issue. How can we account for this finding? We argue that positive and negative information about Macedonia is processed in different ways. In the pre-test, we observed that the participants in our study had stereotypes about Macedonia: on average, the perceived performance of its economy, its assumed level of democracy and the perceived cultural 'match' were low (economy: -.56; democracy: -.16; culture: -.33). Hence, negative information provided by the press reports underscored the negative image of Macedonia. Since individuals access and cognitively process information that confirms existing stereotypes more easily than information contradicting these prejudices (see Fiedler and Bless, 2001), we thus not only identify significant attitude changes for the

issue addressed in a particular article, but also witness the reinforcement of other aspects about Macedonia that are mentally associated with this issue. In contrast, to process information that contradicts the stereotype (i.e. positive articles about Macedonia), processing cannot rely on heuristics and is consequently more systematic and elaborated.9 Given the lack of knowledge about Macedonia, the information provided by a particular newspaper article -irrespective of whether it is framed positively or negatively - causes respondents to update their perception of a particular aspect of Macedonia. Since positive information contradicts already existing views on Macedonia and does not contain convincing arguments forcing subjects to adjust their attitudes on other aspects of Macedonia, we observe no 'diffusion effects' for positive information, whereas we do for negative information.

If newspaper reports have an impact on Macedonia's perceived image -which is particularly pronounced for the issue dimension in question - it seems plausible that the information received by respondents changes the standards by which the EU membership of Macedonia is evaluated. In other words, it is conceivable that the newspaper articles prime attitudes toward a south-eastern enlargement of the EU. This proposition can be assessed empirically by comparing the relative importance of the factors explaining support for Macedonia's accession to the EU before and after exposure to the respective articles and by comparing the relative importance of the primed factor between the different groups. Alternatively, priming effects can be validated by demonstrating that the impact of the experimental treatment on support for Macedonia's accession to the EU was fully mediated by the respondents' beliefs about the Macedonian economy, democracy and culture (see Baron and Kennedy, 1986). We opted for the first route.

The structure of support is straightforward in the pre-test: for all except the control group, the only significant determinant was our cultural 'match' factor (see Table 5). Respondents who perceived the EU and Macedonia to be culturally 'close' showed a significantly higher level of support for Macedonia's accession to the EU than respondents who thought the opposite. In other words, the cultural 'match' factor - not instrumental considerations or issues related to civic or democratic performance - was driving public attitudes towards the location of the EU's borders. In contrast, perceptions of a 'cultural gap' between a candidate state and the EU members caused a rejection of further EU enlargement.

This picture partially changed after exposure to the information provided in the newspaper articles. Whereas articles emphasizing economic and cultural aspects display no significant impact on the structure of support,10 we observe priming effects for press reports emphasizing Macedonia's state of democracy (|i| = 2.55, df = 56, p < .05). For respondents who read about

recent developments in Macedonian democracy, attitudes about the level of democracy are now the most important factor explaining support for Macedonian accession to the EU. In contrast, the impact of evaluations regarding the 'culture gap' between Macedonia and the EU is no longer statistically significant. In addition (and in contrast to the pre-test), the impact of beliefs about Macedonian democracy in the post-test is significantly stronger for those respondents who read an article about the state of democracy than for those who received an article about economic issues (|i| = 2.15, df = 51, p < .05) or cultural issues (|i| = 3.78, df = 55, p < .01). In contrast, the structure of support did not change for the control group: in the post-test, as in the pre-test, none of the three factors is a statistically significant predictor of support for Macedonia's accession to the EU.

Discussion

In this contribution, we started by indicating that research on public support for the EU and European integration in general and for EU enlargement in particular predominantly refers to two types of explanations: instrumental economic considerations on the one hand (i.e. the individually perceived costs and benefits of European integration and EU enlargement), and affective considerations (i.e. the strengths of citizens' psychological identification with their nation-state as well as with the EU) on the other. Since the vast majority of citizens have no first-hand experience of domestic or European politics, we contend that they rely on mass media coverage to form their opinions. Unfortunately, few studies analyse the influence of mass media reports on EU-related attitudes; this is especially true for the research on EU enlargement. Furthermore, we criticized the fact that media impact studies usually engage in issue amalgamation by emphasizing either 'positive' or 'negative' media frames associated with EU enlargement, thereby collapsing different potentially interesting issues dimensions. Consequently, we suggest a more fine-grained understanding of the strength of media effects by analytically separating three different issue dimensions that underlie EU enlargement support for a given candidate country: its economic performance, its state of democracy and its perceived cultural 'match' with the EU.

Obviously, our data and analysis have several limitations. First, our sample consisted of university students. We do not know if the media effects that we observed in this study would also hold for ordinary citizens. In other words, we cannot make inferences from our sample to any 'real' population. Second, our analysis is based on an experimental study. The advantage here is, of course, that we can be certain about the causal relationship between

Table 5 Determinants of support for Macedonian EU membership

Article

Economy Democracy Culture Control group

Pre-test Post-test Pre-test Post-test Pre-test Post-test Pre-test Post-test

R2 .22 .25 .49 .54 .38 .42 .63 .43

Economy 0.06 (.24) 0.05 (.20) 0.21 (.22) -0.06 (.25) 0.06 (.35) 0.16 (.21) -0.05 (.44) -0.33 (.73)

Democracy 0.06 (.22) -0.04 (.35) -0.14 (.32) 0.85*** (.22) 0.39 (.29) -0.20 (.17) 0.10 (.75) 0.55 (.56)

Culture 0.58** (.25) 0.67** (.30) 0.78*** (.24) 0.04 (.27) 0.68** (.27) 0.70*** (.23) 1.41 (.62) 0.57 (.39)

N 25 30 29 10

Notes: OLS regression, unstandardized regression coefficients, standard errors in parentheses. Significance levels: ** p < .05, *** p < .01

media reception and attitude change. To demonstrate this relationship was our central objective. The disadvantage is that experimental studies suffer from an external validity problem: the reception situation was artificial and the attention our respondents paid to the article was probably above average. Again, we cannot be sure if the effects measured here will be replicable in a more realistic reception situation. Third, we do not know how persistent the induced media effects are. Fourth, we do not know whether the effects observed in our study are Germany specific or whether we could potentially observe country effects by conducting a similar experiment in other EU member states.

Our results indicate that media coverage has a strong impact on citizens' attitudes to EU enlargement. By conducting an experiment, we were able to demonstrate that information on the economic, political or social situation in a candidate country provided by only a single newspaper article significantly changes the level of public support for EU accession by this country. According to studies testing the cumulative impact of mass media reports (e.g. Noelle-Neumann, 1973; see also Peter, 2004), we can assume furthermore that exposure to a larger number of articles, newscasts, etc. - provided they pull in the same direction - would further increase changes in support levels. We demonstrated that positive information about a candidate country generally causes an increase in support for accession whereas negative information leads to a decline in support for further EU enlargement.

Furthermore, exposure to newspaper articles on recent democratic developments in Macedonia shifted the standards by which a candidate country's accession to the EU is evaluated (priming). As a consequence, public support for EU membership no longer depends on cultural aspects but depends on the perceived level of democracy. Of the three conditions tested, only the democracy condition displays significant priming effects. One explanation for the presence of priming in this condition could be related to sensitivity and prior knowledge about the criteria for EU accession, the so-called 'Copenhagen criteria', among which the 'political' criteria play a central role: the stability of institutions guaranteeing democracy, the rule of law, human rights and respect for and protection of minorities. Given the media presence of these issues in the run-up to the 2004 and 2007 enlargement rounds, it is conceivable that our respondents possessed knowledge about these criteria that was activated and brought to the 'top of their heads' by the democracy stimulus. But how can we account for the presence of priming effects in the democracy condition when priming is absent in the economy and culture conditions? With regard to the culture condition, we have shown that culture strongly drove attitudes towards Macedonia's membership in the pre-test; culture was hence already at the 'top of the heads' (Zaller, 1992) of

our respondents. Given that culture was already an important determinant of respondents' attitudes towards Macedonia's accession, 'priming' this aspect did not significantly amplify its importance in the post-test. The absence of priming effects in the economic condition can be related to the recent enlargement round. Since 2004, a large number of rather economically weak Central and East European countries have joined the EU and, consequently, the possible negative effects of Macedonian EU membership on the EU and German economies might be seen as rather negligible by the respondents in our study. We wish to emphasize that all these explanations are tentative in nature. Evidently, further research is required to assess whether and how priming effects vary by choosing different sample populations and by engaging in cross-national comparisons.

Since the 'revocation' of the 'permissive consensus' between the political elites and the European public, public opinion features ever more prominently in affecting the pace, content and direction of the European integration project. The negative referendum outcomes in France and the Netherlands on the European Constitutional Treaty are not the first time that citizens have said 'No' to the politics of integration pursued by most of the national and European political elites. A key to understanding why ideas about and trajectories of the future development of the EU differ between the elites and the public is the role of the media. We argue that models explaining support for the EU and European integration should pay more attention to the mass media. Press, radio, television and the Internet serve as a link between 'cue-givers' at the macro level and 'cue-recipients' at the micro level. Unfortunately, media coverage of the European Union tends to have an increasingly negative tone (Peter et al., 2003; for content analyses of the German media coverage, see e.g. Brettschneider and Rettich, 2005). Our experimental results show unequivocally that media reporting has a remarkable impact on shaping public attitudes towards further enlargement of the EU and we are rather sceptical that countries wishing to join the Union will be welcomed by European citizens. Bearing in mind the media's evaluative tone and the role of public opinion in future enlargement rounds, we expect that citizens will give politicians and their plans for further integration a hard time.

This research was funded by a grant from the Kaiserslautern University of Technology. We wish to thank Bernhard Miller, Wolfgang C. Müller and three anonymous referees for their helpful comments. We owe particular thanks to Stefan-Fabian Lutz, Michael Müller and Jens Schmidt for their assistance in organizing the experiment and in collecting the data set.

1 Of the respondents, 41% were female. The average age was 24.9 years. Party attachments were distributed as follows: 23% CDU/CSU, 35% SPD, 14% Green Party, 1% Linkspartei.PDS, 1% 'other' party, 26% 'no party attachment'. In total, we report 36 'drop-outs', i.e. students who participated only in the pre-test. Checking for social and political differences between respondents who participated in both the pre- and the post-test and 'drop-outs', there is only one significant difference for the distribution of party identification (p < .05): in the group of post-test-only respondents, supporters of the SPD (40%) and the FDP (14%) were slightly overrepresented whereas supporters of the CDU (20%) and the Green Party (6%) and persons without party attachment (20%) were slightly underrepresented. Students affiliated to Linkspartei.PDS as well as to other parties were absent from this group. This difference should have no consequence for the results of our study because the German political parties are generally in favour of the European integration process.

2 The length of each article was approximately 350 words. The heading of each article was 'European Union deliberates on South-Eastern enlargement'. The subheading varied according to the frame, yet it clearly indicated the evaluation of the possible accession of Macedonia to the EU and also provided the reasoning behind this evaluation (e.g. 'Support for Macedonia's possible accession due to favourable economic structures'). Each article was divided into four paragraphs. The first paragraph stated that the European Council would decide whether or not the EU would start accession negotiations with Macedonia the coming weekend. In addition, the respondents obtained information about the position taken by EU member states on this issue, the Commission's recommendation and the main reason why the EU was expected to start or to withhold accession negotiations (i.e. economic concerns, the level of democracy or cultural aspects). The second paragraph presented 'facts' and 'statistics' that bolstered the positioning of the EU. The third paragraph highlighted the position of the German government (which is always in line with the position of the EU) and offered some statements by leading German politicians. In order to prevent the information presented in the article from being evaluated by partisan or ideological beliefs ('cues'), we cited politicians from different political camps. The final paragraph emphasized that, if the EU decided to negotiate with Macedonia, the negotiation process might start as early as March 2007 and - in the event of successful negotiations - Macedonia might join the EU in mid-2008. In order to minimize the effects of different wording, we held most of the textual information constant. As a consequence, 'positive' articles differed from 'negative' articles only by switching attributes from the positive to the negative (and vice versa) while the rest of the text remained unchanged, and by slightly modifying the politicians' quotations. In addition, we kept the structure, the basic information provided and the actors mentioned in the different articles as comparable as possible. To increase the credibility of the articles, we employed the house style of the Süddeutsche Zeitung, one of the major German national quality newspapers. In addition, the source of the article (newspaper and date) was handwritten in the margin.

3 'Crude carrier instead of pipeline: New terminals, new ships, new providers - liquid natural gas might become the petroleum of the 21st century', Süddeutsche Zeitung, 6 January 2006 (our translation).

4 'To what extent do you agree or disagree that the following non-EU countries should become an EU member: Macedonia?'; five-point scale ranging from +2 ('completely agree') to -2 ('not agree at all').

5 'What do you think: (a) how congruent is the culture between the EU and _, (b) how does the economy perform in_, (c) what is the level of democracy in_. . . the following countries: Macedonia'; five-point scales ranging

from +2 ('completely share our values'/'the economy performs very well' /'high level of democracy') to -2 ('do not share our values at all'/'the economy performs very poorly'/'not a democracy at all').

6 Coding scheme: -1 ('exposure to negative article on the particular issue'); 0 ('exposure to another article, including control group'); +1 ('exposure to positive article on the particular issue').

7 Up to 6% of the sample had missing values either in the pre-test or in the post-test as regards the evaluation of the Macedonian economy, democracy and culture. In order to prevent a reduction in the sample size, missing values in the pre-test were substituted by post-test values and vice versa. This treatment is conservative, i.e. it assumes stability and therefore reduces change rates and decreases standard deviations. As a consequence, the reported impact of press reports on attitudes is underestimated.

8 The correlation coefficients are r = .53 (pre-test) and r = .42 (post-test) between economy and democracy, r = .34 and r = .46 between economy and culture and r = .48 and r = .57 between democracy and culture. All correlations are statistically significant (p < .01, minimum N = 94).

9 For a discussion of the Heuristic-Systematic Model (HSM) see e.g. Chaiken et al. (1989).

10 The regression coefficients for the pre-test and the post-test do not yield any significant differences (economy: |t| = .03, df = 46, p > .05; culture: |t| = .06, df = 54, p > .05).

References

Anderson, Christopher J. and M. Shawn Reichert (1996) 'Economic Benefits and Support for Membership in the European Union: A Cross-National Analysis', Journal of Public Policy 15(3): 231-49.

Baron, Reuben M. and David A. Kennedy (1986) 'The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations', Journal of Personality and Social Psychology 51(6): 1173-82.

Behnke, Joachim, Nina Baur and Nathalie Behnke (2006) Empirische Methoden der Politikwissenschaft. Paderborn: Schöningh.

Brettschneider, Frank and Markus Rettich (2005) 'Europa - (k)ein Thema für die Medien', in Jens Tenscher (ed.) Wahl-Kampf um Europa: Analysen aus Anlass der Wahlen zum Europäischen Parlament 2004, pp. 136-56. Wiesbaden: Verlag für Sozialwissenschaften.

Bruter, Michael (2003) 'Winning Hearts and Minds for Europe: The Impact of News and Symbols on Civic and Cultural European Identity', Comparative Political Studies 36(10): 1148-79.

Carey, Sean and Jonathan Burton (2004) 'Research Note: Influence of the Press in Shaping Public Opinion towards the European Union in Britain', Political Studies 52(3): 623-40.

Cederman, Lars-Erik (2001) 'Political Boundaries and Identity Trade-Offs', in Lars-Erik Cederman (ed.) Constructing Europe's Identity: The External Dimension, pp. 1-32. Boulder, CO: Lynne Rienner.

Chaiken, Shelly, Akiva Liberman and Alice H. Eagly (1989) 'Heuristic and Systematic Information Processing within and beyond the Persuasion Context', in James S. Uleman and John A. Bargh (eds) Unintended Thought, pp. 212-52. New York: Guilford Press.

De Vreese, Claes and Hajo G. Boomgaarden (2005) 'Projecting EU Referendums: Fear of Immigration and Support for European Integration', European Union Politics 6(1): 59-82.

De Vreese, Claes and Hajo G. Boomgaarden (2006) 'Media Effects on Public Opinion about the Enlargement of the European Union', Journal of Common Market Studies 44(2): 419-36.

Eichenberg, Richard C. and Russell J. Dalton (1993) 'Europeans and the European Community: The Dynamics of Public Support for European Integration', International Organization 47(4): 507-34.

Fiedler, Klaus and Herbert Bless (2001) 'Social Cognition', in Miles Hewstone and Wolfgang Stroebe (eds) Introduction to Social Psychology: A European Perspective, 3rd edition, pp. 115-49. New York: Springer.

Flora, Peter (2000) 'Externe Grenzbildung und interne Strukturierung. Europa und seine Nationen. Eine Rokkan'sche Forschungsperspektive', Berliner Journal für Soziologie 10(2): 151-65.

Franklin, Mark, Michael Marsh and Lauren McLaren (1994) 'Uncorking the Bottle: Popular Opposition to European Unification in the Wake of Maastricht', Journal of Common Market Studies 32(4): 455-72.

Gabel, Matthew (1998a) Interests and Integration: Market Liberalization, Public Opinion, and European Union. Ann Arbor, MI: University of Michigan Press.

Gabel, Matthew (1998b) 'Economic Integration and Mass Politics: Market Liberalization and Public Attitudes in the European Union', American Journal of Political Science 42(3): 936-53.

Gabel, Matthew (1998c) 'Public Support for European Integration: An Empirical Test of Five Theories', Journal of Politics 60(2): 333-54.

Gabel, Matthew and Harvey D. Palmer (1995) 'Understanding Variation in Public Support for European Integration', European Journal of Political Research 27(1): 3-19.

Gerhards, Jürgen (2005) Kulturelle Unterschiede in der Europäischen Union: Ein Vergleich zwischen Mitgliedsländern, Beitrittskandidaten und der Türkei. Wiesbaden: Verlag für Sozialwissenschaften.

Habermas, Jürgen (1992) 'Citizenship and National Identity: Some Reflections on the Future of Europe', Praxis International 12(1): 1-19.

Hooghe, Liesbet and Gary Marks (2004) 'Does Identity or Economic Rationality Drive Public Opinion on European Integration?' PS: Political Science and Politics 37(3): 415-20.

Hooghe, Liesbet and Gary Marks (2005) 'Calculation, Community and Cues: Public Opinion on European Integration', European Union Politics 6(4): 419-43.

Hooghe, Liesbet and Gary Marks (2006) 'Europe's Blues: Theoretical Soul-Searching after the Rejection of a European Constitution', PS: Politics and Political Science 39(2): 247-50.

Huntington, Samuel P. (1996) The Clash of Civilizations and the Remaking of World Order. London: Simon & Schuster.

Inglehart, Ronald (1977) The Silent Revolution: Changing Values and Political Styles among Western Publics. Princeton, NJ: Princeton University Press.

Iyengar, Shanto and Donald R. Kinder (1987) News That Matters: Television and American Opinion. Chicago, IL: University of Chicago Press.

Jones, Erik and Niels van der Bijl (2004) 'Public Opinion and Enlargement: A Gravity Approach', European Union Politics 5(3): 331-51.

Kaltenthaler, Karl C. and Christopher J. Anderson (2001) 'Europeans and Their Money: Explaining Public Support for the Common European Currency', European Journal of Political Research 40(2): 139-70.

Karp, Jeffrey A. and Shaun Bowler (2006) 'Broadening and Deepening or Broadening versus Deepening: The Question of Enlargement and Europe's "Hesitant Europeans"', European Journal of Political Research 45(3): 369-90.

Lamont, Michèle and Virâg Molnar (2002) 'The Study of Boundaries in the Social Sciences', Annual Review of Sociology 28(1): 167-95.

Lindberg, Leon and Stuart A. Scheingold (1970) Europe's Would-Be Policy: Patterns of Change in the European Community. Englewood Cliffs, NJ: Prentice Hall.

McLaren, Lauren (2002) 'Public Support for the European Union: Cost/Benefit Analysis or Perceived Cultural Threat?' Journal of Politics 64(2): 551-66.

McLaren, Lauren (2007) 'Explaining Opposition to Turkish Membership of the EU', European Union Politics 8(2): 251-78.

Martinotti, Guido and Sonia Steffanizzi (1995) 'Europeans and the Nation State', in Oskar Niedermayer and Richard Sinnott (eds) Public Opinion and Internationalized Government, pp. 163-89. Oxford: Oxford University Press.

Noelle-Neumann, Elisabeth (1973) 'Return to the Concept of Powerful Mass Media', Studies of Broadcasting 9(1): 67-112.

Norris, Pippa (2000) A Virtuous Circle? Political Communication in PostIndustrial Democracies. Cambridge: Cambridge University Press.

Peter, Jochen (2004) 'Our Long "Return to the Concept of Powerful Mass Media" - A Cross-National Comparative Investigation of the Effects of Consonant Media Coverage', International Journal of Public Opinion Research 16(2): 144-68.

Peter, Jochen, Holli A. Semetko and Claes de Vreese (2003) 'EU Politics on Television News: A Cross-National Comparative Study', European Union Politics 4(3): 305-27.

Price, Vincent and John Zaller (1993) 'Who Gets the News? Alternative Measures of News Reception and Their Implications for Research', Public Opinion Quarterly 57(2): 133-64.

Reif, Karlheinz (1993): 'Ein Ende des "Permissive Consensus"? Zum Wandel europapolitischer Einstellungen in der öffentlichen Meinung der EG-Mitgliedsstaaten', in Rudolf Hrbek (ed.) Der Vertrag von Maastricht in der wissenschaftlichen Kontroverse, pp. 24-40. Baden-Baden: Nomos.

Risse, Thomas (2003) 'Social Constructivism and European Integration', in Antje

Wiener and Thomas Diez (eds) European Integration Theory, pp. 159-76. Oxford: Oxford University Press.

Sánchez-Cuenca, Ignacio (2000) 'The Political Basis of Support for European Integration', European Union Politics 1(1): 147-72.

Scheufele, Dietram A. and David Tewksbury (2007) 'Framing, Agenda Setting, and Priming: The Evolution of Three Media Effect Models', Journal of Communication 57(1): 9-20.

Schimmelfennig, Frank and Ulrich Sedelmeier (2005) 'The Politics of EU Enlargement: Theoretical and Comparative Perspectives', in Frank Schimmelfennig and Ulrich Sedelmeier (eds) The Politics of European Union Enlargement: Theoretical Approaches, pp. 3-29. London: Routledge.

Schuck, Andreas R.T. and Claes De Vreese (2006) 'Between Risk and Opportunity: News Framing and Its Effects on Public Support for EU Enlargement', European Journal of Communication 21(1): 5-23.

Semetko, Holli A. (2004) 'Media, Public Opinion and Political Action', in John D.H. Downing, Dennis McQuail, Philip Schlesinger and Ellen A. Wartella (eds) The SAGE Handbook of Media Studies, pp. 351-74. Thousand Oaks, CA: SAGE.

Shepherd, Robert J. (1975) Public Opinion and European Integration. Lexington: Lexington Books.

Steenbergen, Marco, Erica E. Edwards and Catherine E. de Vries (2007) 'Who's Cueing Whom? Mass-Elite Linkages and the Future of European Integration', European Union Politics 8(1): 13-35.

Vetik, Ravio, Gerli Nimmerfelft and Marti Taru (2006) 'Reactive Identity versus EU Integration', Journal of Common Market Studies 44(5): 1079-102.

Zaller, John R. (1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.

Zielonka, Jan (ed.) (2005) Europe Unbound: Enlarging and Reshaping the Boundaries of the European Union. London: Routledge.

About the authors

Jürgen Maier is Junior Professor of Social Science Research Methods, Faculty of Social Sciences, Kaiserslautern University of Technology, D-67663 Kaiserslautern, Germany. Fax: +49 631 205 4968 E-mail: maier@sowi.uni-kl.de

Berthold Rittberger is Chair of Political Science and Contemporary History, Faculty of Social Sciences, University of Mannheim, D-68131 Mannheim, Germany. Fax: +49 621 181 2059

E-mail: berthold.rittberger@uni-mannheim.de