Julian Schärdel & Pascal König

Transcription

1 Eurosceptic Newcomer Parties in the National Media An Empirical Analysis of Newspaper Coverage on New Eurosceptic Parties in three 2013 National Elections Julian Schärdel & Pascal König Department for Political Science, University of Freiburg D Freiburg Paper prepared for the ECPR General Conference 2014, Glasgow. Panel 073: Distrust and Populism: The Rise of Euroscepticism in Contemporary Crisis, chaired by Hans-Jörg Trenz. First draft, please do not cite without permission by the authors. Abstract This study sheds light on how new Eurosceptic parties are received and made sense of by the media. It does so by analyzing the quality press coverage on three specific cases: The M5S (Italy), the AfD (Germany), and the TS (Austria). These cases represent non-extremist Eurosceptic challengers to the political establishment which capitalized upon the euro crisis and competed in national elections in 2013 for the first time. We look at their representation in the coverage, which issues they were mainly associated with, and how they were pictured overall. The analysis yields differences primarily in presentational style: a strong personalization and a focus on events are very pronounced for the M5S and the TS, while issues played a slightly greater role in comparison for the AfD. The three parties coverage, however, converges in a generally peripheral role of issues, in an overall negative bias, and in depicting them as dangerous and irresponsible populist protest parties. 1

2 1. Introduction The economic and financial crisis has meant a hard time for the European spirit. Economic uncertainty and contentious measures on a national as well as European level for dealing with the euro crisis were a severe test for public support of national governments and the European Union (Braun and Tausendpfund 2014; Kriesi 2014). During this crisis, several political challengers have spawned that combined a Eurosceptical stance with an anti-establishment orientation and mild or strong populist traits. This is even true for countries that have strongly supported the European integration process so far, like Germany. This transformation in European national party systems seems to support Kriesi s New Cleavage Hypothesis, according to which Euroscepticism on the one, and integration into the European Union on the other side, are part of a new structural conflict (Kriesi 2007, 85). As new parties, not only do they have to make themselves, their issues, and positions heard. They also have to struggle to find their place in the party system as well as in the eyes of the public, since there is no objective public perception of such new phenomena. Rather, citizens will form their interpretations and opinions of such new parties in the context of the existing constellation. This is where the media come into play, as they fulfill an essential role in transmitting and evaluating political information and setting the political agenda (Schoen 2005, 505; Jenkins 1999, 430). In this study, we look at three new non-extremist Eurosceptic challenger parties that have entered the national arena and contended national elections for the first time in 2013: the AfD ( Alternative für Deutschland, Alternative for Germany) in Germany, the M5S ( Movimento Cinque Stelle, Five Star Movement) in Italy, and the TS (Team Stronach) in Austria. What makes these parties interesting, besides their status as newcomers on the national political scene, is that none of them is an extremist party. Hence, in comparison with this kind of parties they should, ceteris paribus, have been considerably more likely to get acknowledged as a legitimate political competitor by the political mainstream and to get specific issues like Euroscepticism on the public agenda. The aim of the study is to investigate how national media depicted these parties just before those national elections, i.e. during the election campaigns. In doing so, we seek to answer the following questions: To what degree were they represented in the media and what overall portrayal of the parties could recipients find in the media based on the characterizations conveyed by them? The analysis is also concerned with the degree to which their media coverage involved specific issues that may have built on or extended existing political cleavages. Of particular interest in this context is, then, what role EU related matters played in the coverage on Eurosceptic challenger parties during a national election campaign. Are there maybe even discernable aspects of first order European elections, to variegate the 2

3 often-cited phrase of second order national elections (Reif and Schmitt 1980)? To answer these questions, we analyze the coverage of quality newspapers from the three countries during the crucial phase of the electoral campaign (Wilke and Leidecker 2010, 340), the last month before the election date. The following section presents the theoretical and empirical background for the analysis. Section three briefly describes the design, data and the method before turning to the analysis in the fourth section. The paper concludes with a summary after a focused discussion of the results. 2. Eurosceptic challenger parties, campaign communication and the media Before outlining the framework for the analysis, this section reviews the relevant theoretical contributions and empirical insights, which together form a theoretical background for the inquiry. What follows is a brief account of a) the concept of new challenger parties in general and with regard to Euroscepticism and populism in particular, b) the role of the media in campaign communication and more specifically with regard to those challenger parties. New parties, populism, and the issue of European integration While it has been noted that challenging parties do not necessarily have to enter the party system and compete based on specific issues at all (Lucardie 2000; Tavits 2008; Sikk 2012), it is generally presumed that new parties emerge as the consequence of established parties not having incorporated (new) relevant issues (Hug 2001). Their success is often fueled by voters disappointment with incumbents (Tavits 2007) and according to Abedi (2002, ) such parties profit from a diminished range of political offers following a convergence of the established parties (and even more so in polarized party systems). Discontent with the range of political options can also be said to be essential in the spawning of anti-political-establishment parties, which emerge as and provide a viable vote option for disenchanted and disappointed citizens (Dalton and Weldon 2005, 940). Euroscepticism has often been a characteristic of such protest parties. Eurosceptic parties frequently emerge as a consequence of disaffection with the current situation, although this disaffection does not necessarily have to be directly connected to European issues. Particularly during deteriorating socioeconomic situations, the European level has often been blamed for economic difficulties, while the problem-solving capacity of the nation state has been highly praised compared to the allegedly bloated bureaucracy in Brussels even by (mainstream) parties in government (Risse and Kleine 2007; Wenzelburger 2011). Nevertheless, Eurosceptic parties are usually opposition parties that try to differentiate themselves from the established 3

4 parties through a hostile attitude towards European integration either entirely or in certain regards while the Eurosceptic attitudes often decline when these parties assume the responsibilities of government (Hooghe, Marks, and Wilson 2002, 985; Aust 2003, 261; Oberkirch and Schild 2010, 43). Eurosceptic positions can, however, not be firmly tied to a broader ideological stance and have in fact shifted over time: While Euroscepticism has become weaker in the left ideological spectrum, it seems to have shifted somewhat to more of an issue of rightist parties. Among new Eurosceptic parties, Paul Taggart differentiates protest-based parties, which have added the dimension of European politics into their general protest attitude, and so called single-issue Eurosceptical parties. As this denomination already implies, these movements explicitly focus on an antagonistic stance toward the European Union as a means of mobilizing voters (Taggart 1998: 368). In both cases Euroscepticism is often accompanied by populist tendencies. Populist parties define themselves as anti-establishment or anti-elitist and criticize the governing class for not being understanding of popular concerns while they themselves claim to have a strong connection to the people as a whole (Hartleb 2011, 7). With their usually charismatic leaders, their provocative and polarizing style as well as messages evoking a strong we against the others -identity, they are more prone to please media criteria for newsworthy content (Kriesi, Bernhard, and Hänggli 2009, 360). The role of the media and election campaigns Receiving media coverage is not only important for the populist strategy to publicly challenge and discredit their opponents. It is essential for any new party, as the mass media provide the most important source of information and a central link between the political elite and the public (Kriesi 2004; Kriesi et al. 2009: 351). As pointed out by several scientists, the media constitute the main forum for the public political discourse (Pfetsch 2004; Schoen 2005). Especially for new parties, media reception is crucial as nobody is going to vote for a party or based on an issue that he or she is not aware of. Through agenda-setting (McCombs and Shaw 1972), priming (Iyengar and Kinder 1987), and framing (Chong and Druckman 2007; Scheufele and Tewksbury 2007, 11), the media are said to decide which actors are present in the public discourse, and which issues are covered, which issues are important for public evaluations, and in what light these issues are presented and perceived respectively. 1 The 1 A demonstrative example of how the presentation may differ substantially between several media could be observed in Germany after the parliamentary election in 2013 and with regard to one of the parties investigated below. The election result of the new Eurosceptic party Alternative für Deutschland (AfD), which had just failed to surmount the five percent threshold (required for entry into the German Bundestag), was analyzed by the news channel N-TV under the heading: Is this the end of the AfD? Eurosceptics failed to get into the 4

5 importance of the mass media should be considerable, especially during the heated phase of election campaigns when public and media attention is particularly high and new and unknown parties have an opportunity to gain broad popular attention (Jenkins 1999, 430). 2 They have to make themselves known to the voters and make an effort in conveying their messages without having the privileged access that some of their adversaries have (Kriesi 2004, ). One important dimension of getting out one s messages during the election campaign lies in propagating and occupying certain issues, in other words asserting issue ownership (Kriesi, Bernhard, and Hänggli 2009, ). According to Petrocik (1996), issue ownership involves the perception that a party is more capable to deal with specific problems. Due to their role as a main information source, the media are of obvious importance for learning about such issue associations. However, they are not simply transmitters of messages but have acquired an increasingly independent role (Page 1996; Mazzoleni and Schulz 1999; Trenz 2007, 89). The idea of an autonomous operating mode of the media has been termed as media logic, according to which media select and tread content in a way as to foster a presentation that adheres to schedules, entertainment values and the style and grammar of media production (Altheide and Snow 1979; Mazzoleni 1987). This media logic leads to an orientation towards news values, common perceptions of what is newsworthy and interesting to the public (Altheide 1991; Galtung and Ruge 1965). In their attempts to gain media attention and utilize the media for their objectives, political actors thus have to accommodate criteria and routines of selection and production (Strömbäck 2008, ). Empirical studies have attempted to carve out how this media logic manifests in the coverage of election campaigns. Mazzoleni (1987) first conducted a study on the media coverage of the election campaign before Italian national elections in 1983 in order to test the relative influence of these operating mechanisms. According to his findings, the media gave an extensive coverage of issues, but elements of a media logic were also very much visible in aspects like a spectacularization of the election campaign and a stronger focus on the election politics itself 2 Bundestag (N-TV 2013). In contrast, a regional newspaper from Hamburger described the result as follows: Hamburg-based Professor Lucke and the AfD hidden winner of the elections Even if they narrowly failed to enter into the Bundestag, the Eurosceptic Party won a hundred thousand of votes from the established parties. A success of their leading candidate Bernd Lucke (Hamburger Abendblatt 2013). While previous work contended that campaigns and campaign communication have minimal effects on voters at best, more recent work has been successively bolstering the view that it does indeed have a considerable impact (Kriesi, Bernhard, and Hänggli 2009, 345). And, despite the recent developments in the information and communication technologies, the traditional mass media are still essential for political actors chances of winning support, and modern campaigning is still very much mass media-centered (Jerit 2008; Kriesi, Bernhard, and Hänggli 2009; Norris 2006; Römmele 2003). This may be due to the fact that the traditional reporting is usually perceived to be rather objective and that it is thus more likely to have an influence on voters opinion (Pfetsch 2004, 32 34; Schoen 2005, 510). 5

6 closer toward the end of the campaign (Mazzoleni 1987, 91 99). Brants and van Praag (2006) adopted a historical perspective and investigated the importance of media logic in three Dutch election campaigns distributed over roughly half a century. They find that since the 1950s the media logic has become much more discernible and conclude that there was on the whole less substantive and more horse race and poll driven reporting in the national election in 2003 (Brants and van Praag 2006, 38). Van Aelst et al. (2008, ) similarly find in their analysis of media attention to candidates in the Belgian election campaign of 2003 that both a media logic and a political logic that represents the importance of political relevance seem to be at work. In addition to that, election polls became increasingly important in the news coverage. In sum, there seems to have been an evolution in the last decades which has transformed election campaigns into a sort of sports competition and struggle for votes as suggested by the notion of horce race journalism (Schoen 2005, 519). This trend has been accompanied by an increasing personalization of election campaigns and their coverage. With the declining strength of partisan ties, strategic political communication and a stronger concern for the image and the media presence of leading candidates have become increasingly important (Holtz- Bacha 2007; Negrine 2007)(Ro ssler 2011, 305). This stronger focus on candidates can be seen as another aspect of the media logic: A personalized presentation of political issues is much easier to sell than an analysis of party manifestos and it thus fits better with the media s aim to attract a broader audience (Wattenberg 1994, 92 98; Holtz-Bacha 2003, 20). As implied above, anti-establishment populist parties can capitalize on these dimensions of the media logic because they are usually conflict-oriented and tend to have strong leading personalities (Mudde 2004, 558). As Jenkins (1999, ) notes in his analysis of the Canadian Reform Party, media reporting can surge with the rise of political support for a new party and in turn operate as a push factor driving party support. And while increased media attention for political challengers seems to go along with a strong negativity bias especially in tabloid media (Pfetsch 2004, 32 34) populist parties may even benefit from such coverage (Mazzoleni 2008, 56; Diani 1996, 1064): It allows them to portray the media as an instrument of suppression by the establishment and to present themselves as an underdog. Although this summary of the literature suggests a predominant role of the media logic, this is not to say that substantial political aspects have become irrelevant. Issues still play a basic role in serving as evaluation frameworks for the candidates and their parties (Ro ssler 1997, 9). Following Walgrave and De Swert (2007), who conduct an analysis on Belgian politics, the issues reported in the media in combination with certain parties correspond to the issue salience 6

7 of these parties based on their respective manifestos. Thus, although the media select and emphasize certain aspects that conform to the idea of a media logic, they also seem more or less to faithfully convey information on political parties and candidates that are important from a voter perspective. In addition, however, Walgrave and De Swert (2007, 63 64) also find that besides the party communication through manifestos and parliamentary action parties media coverage seem to have an independent and reinforcing role in shaping the public s perception of issue ownership. They also find that this shows most strongly for new parties, as they were most clearly associated with specific issues. 3. Research Design Analytical Framework The research review above serves as a basis from which to extract relevant aspects for the intended analysis of media coverage on the three selected parties. The following aspects are chosen in order to give a differentiated picture and with the aim to provide answers to the following questions: 1. How often were the parties mentioned, 2. which issues were associated with them, 3. what weight did the coverage on such political issues have in comparison to more newsworthy elements of the politics dimensions (like personalization, special events or election polls), and 4. how were they depicted? The first aspect of interest, which falls under the logic of agenda-setting, is the extent to which the three Eurosceptic parties were represented in the national press and whether this representation is proportionate to its political relevance in relation to other parties. Is there a higher attention for new parties as predicted by several scientists or does the quality press disadvantage Eurosceptic challenger parties as part of a pro-european consensus? A second aspect related to agenda-setting and priming is the coverage of issues in combination with the chosen parties. This is important for two reasons: first, the corresponding coverage can have consequences for associations of issue-ownership with the parties and second, it indicates if parties have been able to set certain issues on the media agenda. It can also be informative with regard to whether the challenger parties in the media coverage build on existing and established political conflict dimensions or introduce other issues or cleavages like Euroscepticism. 7

8 Thirdly, the coverage of specific issues is also relevant under aspects of media logic: The relative weight of issue reporting can be reduced by a strong horse-race reporting that focuses more on polls or on the election campaign itself and the conflict that goes along with it. Features that usually go along with this kind of reporting are eventization and personalization, all of which can also be expected in the coverage of the new parties. Finally, a closer look at the media coverage can reveal what kinds of characterizations have (consistently) been used with regard to the Eurosceptic challenger parties. This should yield an overall portrayal of the parties that indicates whether a negative or positive bias is present in the coverage and if new actors were depicted as credible, trustworthy, responsive, or competent political agents (of the citizens) all of which are aspects that can be expected to be of prime relevance for the citizens evaluation of the parties. This portrayal should form part of the media s effort to make sense of the challenger parties, an effort that should also comprise an account of these parties role in their political and party systems in relation to other political actors: e.g. did typical populist strategies such as the identification as underdog or antiestablishment attacks manifest in their media coverage. Case selection and data For a comparable larger setting, the study focuses on a rather short period, the year 2013 in which several national parliamentary elections took place. Three of these elections were held in countries with long-standing democratic systems that are traditionally rather Europhile, and in which Eurosceptic challengers to the established mainstream parties tried to enter the national parliament for the first time. These three parties, the Italian M5S, the German AfD, and the Austrian Team Stronach, were founded or contested the national elections as a reaction to the euro crisis. All three parties combine Eurosceptic positions with a critical stance toward the established political parties and with at least mildly populist traits. Despite these similarities, they are of different political-ideological orientation: While Grillo s M5S adheres more to leftist stance (Bordignon and Ceccarini 2013; Hartleb 2013), the AfD and TS come close to a conservative orientation (see Table 1). Another difference lies in the role of the parties leaders and/or candidates. Whereas the M5S and TS had very strong and publicly visible leaders, there was no figure within the AfD that struck out to a comparable degree. More striking, however, are the wide differences in their electoral successes. While the M5S ran at 15 percent in polls two months before the election and eventually reached 25.6 percent (as the second largest party), TS fell from formerly about 8

9 15 percent in the opinion polls to 5.7 percent in the election. The German AfD gained 4.7 percent and closely failed to pass the electoral threshold of five percent. Table 1: Brief description of the parties Movimento 5 Stelle The Five Star Movement (short M5S) lead by former Italian comedian Beppe Grillo was already founded as a movement in 2010 with the five core programmatic issues being water, environment, transport, connectivity and growth (representing the five stars). Programmatically the party has a leftist stance but the voter base has broadened to the right. Because of its new protest party status, the M5S was also able to absorb many voters from the Lega Nord and the former PdL (Bordignon and Ceccarini 2013: ). The political outsider Grillo and his movement position themselves in opposition to the political establishment (including the mass media), especially the parties. The M5S claims the political system to be inherently corrupt and has shown a major concern for scandals and corruption in the political and the economic sphere. The strongly anti-elitist party depicts its own role as being close to the citizens by emphasizing anti-professionalism in politics and the direct participation of citizens. Grillo has organized large-scale meetings in public places and the party makes extensive use of new information and communication technology to mobilize and include citizens. At the same time, there are strong elements of top-down organization within the party, with Grillo as a very visible leader at the top of the hierarchical and professionalized organization. As Bordignon and Ceccarini (2013, 431, 441) argue, the party has capitalized on the debt crisis in Europe and the social discontent together in face of tough austerity measures. In a climate of political distrust, the party gained ground as a new political force in the local elections of May In the national election in February 2013, after trending in the last polls (ending two weeks before the election date) at 15 percent, the party eventually gained 25.5 percent, surpassing even the party led Silvio Berlusconi, Popolo della Libertà. Alternative für Deutschland The free-market conservative party Alternative for Germany (short AfD) was founded in 2013 by several economics professors and members of the political or economic elite who were unsatisfied with the German European policies in the debt crisis. The AfD s main programmatic issue is the abolition of the single currency and the dissolution of the euro zone. As the party emerged only a few months before the national elections, the AfD s election manifesto contained only four pages, half of which dealt with European Union issues. Moreover, they call for an extension of direct democracy in Germany and Europe and for stricter immigration policy. Some observers therefore classify the AfD as a right-wing populist party (Häusler 2013, 92). With a share of 4.7% of the votes in the national elections and 4.1% in the regional elections in Hessen in 2013, the AfD respectively failed to pass the five percent threshold, but finally celebrated a victory in the European elections in 2014 with a share of the votes of 7.1%. The party s voter base is relatively diverse, the AfD attracted voters from all established parties, most of them from the liberal party FDP and the leftwing party Die LINKE, i.e. from very different positions in the ideological range. A survey by the research institute infratest dimap points to a large basis of protest voters contributing to the aforementioned result: 57% of the interviewed AfD voters stated that their motivation was dissatisfaction with the established parties rather than the AfD s political aims (Infratest Dimap 2013). Team Stronach The Team Stronach (short TS, or TEAM, or FRANK) was founded by the Austro-Canadian businessman Frank Stronach in Their main programmatic issues were the transformation of the common currency into several national Euros, the reduction of bureaucracy, a reformation of the army and the national tax system as well as a budget consolidation. As a protest-party (Hofer 2014, 18), the TS did not only attract voters of its own free-market conservative camp, but also disappointed citizens of other political colors, notably men. After the party s creation, some parliamentarians from other parties joined the TS, so that it was already represented in the national parliament before the elections in After some remarkable successes in regional elections in 2013, the TS failed to live up to previous expectations in the national elections by the party as much as by the polls with a share of the votes of only 5.7%. In October 2013, Stronach announced his retirement from politics and resigned from parliament after having been present in two sessions. 9

10 Besides differences in party characteristics, the political and economic environment of the three parties varied as well. All three selected countries are established parliamentary democracies with strong consensus-democratic features (Lijphart 2012). However, whereas relatively strong populist Eurosceptic forces such as the Lega Nord and the FPÖ had already existed in Italy and Austria respectively for several years, there had not been any Eurosceptic party of political relevance in Germany prior to the AfD. Especially in Italy, the prevalence of anti-party sentiments seems to make mobilization by populist parties easier (Diani 1996). Moreover, with regard to the economic situation, Italy stands out in that it was hit much harder by the financial crisis than Germany and Austria, which provided a more fertile ground for public protest in Italy. Taken together, i.e. in terms of the economic conditions and political distrust, Italy provided the best opportunity structures for the success of a populist Eurosceptic challenger party, while they were the most unfavorable in Germany due to its strong pro- European social and political consensus. Hence, even if the country specific circumstances and the ideological make-up of these parties differ, there are also some remarkable similarities: Firstly, there is the larger crisis context, which occasioned these parties to compete at the national elections. Secondly, they protest, have populist traits, and explicitly challenge eliteconsensus, e.g. through a Eurosceptic position. The data selection is in line with other empirical studies in the area of media content research (Eilders and Voltmer 2003; Trenz 2004; Trenz 2007; Wessler 2008; Gerhards, Offerhaus, and Roose 2009). For every country, the two most influential daily quality newspapers were chosen for the analysis: La Repubblica and Il Corriere della Sera for Italy; Die Süddeutsche Zeitung and Die Frankfurter Allgemeine Zeitung for Germany and Der Standard and Die Presse for Austria. This selection does of course not capture the entire public discourse of the three Eurosceptic challenger parties. However, these newspapers of record are an important reference for other media so that their influence goes far beyond their own readership (Trenz 2004, 314; Trenz 2006, 120; Peters and Weßler 2006, 126). More importantly, these quality newspapers are most likely to deal with the three parties in a comparatively differentiated way in order to make sense of them. This can hardly be expected of visual media, which inherently tend to simplify more. Thus, the quality press is where we can expect a coverage that deals with the political issues to be more likely which is important for a fruitful comparison in terms of issues between the three cases, especially with regard to Euroscepticism. The textual basis consists of all newspaper issues from the last month before the respective national election date. The relevant articles have been selected by whether they contained the 10

11 party name and/or the name of their top candidates. 3 In this sense, we have opted for a full sample of articles satisfying the described criteria. A qualitative content analysis was then performed on the selected texts. The categories for the analysis were generated inductively from the identified material and structured according to the analytical framework delineated above: The main categories, which were predefined, are issues/topics, kind of source of reference (about the party), mentions of the top candidate/party leader and the party, and characterizations in combination with valence (positive, negative or neither). The coding units are statements, which can be of variable length, from several words to several phrases. The category scheme for the sources of references to the party turned out with a fourfold distinction between direct citations of the party (including politicians, members, and followers), indirect citations (at least no sentence cited), external sources (either directly or indirectly quoted), and references by the article s author, which comprises all statements about the party that can be ascribed to the newspaper itself. This schema allows for a more fine-grained analysis of the party reference in the coverage and of where specific characterizations and a certain evaluative stance derive from. The categories for these characterizations were condensed in several steps. Each of them is subsumed under one of the three evaluative categories. The attributes were coded in such a way that they were fixed with regard to the evaluative direction: E.g. the category comedian as a neutral characterization (for Grillo) was not coded in statements that depict him as a clown or associate comedian directly with negative traits. In such cases the statements were given a code that corresponded to the respective traits. If a statement was not overtly negative or positive, it was considered as neutral. The quantitative scores for personalization are based on the share of articles in which the party s top candidate (party leader for the M5S) is mentioned compared with the share of articles in which the party is mentioned. Finally, in order to determine to what extent the challenger parties were represented in the media in comparison with other relevant political parties, we researched the numbers of articles in which such other parties or their top candidates were mentioned (in the same way as the articles for the studied Eurosceptic parties were selected). For this purpose, we chose the two largest parties and main contenders as well as the party whose election vote share came closest to the respective new Eurosceptic party in each country. The resulting article shares for each party were then complemented with the 3 In case of the M5S, which did not have a top candidate, Beppe Grillo was chosen since he was the obvious figurehead of the party. 11

12 corresponding relative vote shares based on polling data: Gallup in Austria, IPSOS in Italy, and Forschungsgruppe Wahlen in Germany. 4 By means of the polling data it should be possible to ascertain whether the coverage on the Eurosceptic challenger parties was disproportionate to their relative political importance in terms of public endorsement. 4. Empirical Analysis Media representation of the challenger parties and kinds of references Turning first to the question how often the parties were mentioned in the press coverage, we cannot detect an overrepresentation of the three parties, as the notion of the media logic and newsworthiness might suggest. This can be concluded from the scores in Table 2. It shows the vote shares (percentage relative to the sum of shares for the respective four parties) for the two main competitors, a third party which came closest to the selected challenger party in each country, and the challenger party itself. The table also contains the shares for the number of articles in which the party or its top candidate were mentioned. If anything, the challenger parties have been underrepresented in all three countries, although the discrepancies are not huge. The most striking aberration is found in the Germany with regard to FDP, which received an immensely disproportionate coverage. Its high article share could be due to an incumbency bonus, but it is more likely that the increased media attention derives from the historical situation in which it was doubtful that the FDP would make it into the Bundestag which indeed it did not. The FDP, however, is not the only smaller party that is overrepresented in the data. The party led by former prime minister possibly giving him a certain attention bonus Mario Monti, Scelta Civica, and the Austrian NEOS both have higher relative article shares than corresponding vote shares. Finally, one can also see from the table that the coverage on the two largest parties is more balanced than the relation between their vote shares according to the campaign polls. In sum, while the largest parties seem to have been underrepresented in the coverage for a more equalized reporting on the two main competitors, the smaller parties FDP, NEOS, and Scelta Civica got more media attention than their relative vote shares would lead to suggest. The Eurosceptic challengers AfD, M5S, and Team Stronach, in contrast, consistently received comparatively less attention. 4 For Italy, polling is prohibited after two weeks before the election. The available polling data is also not given for the individual parties but for the election alliances, which are typical for Italian politics. The vote shares for these alliances are thus used as proxies for the selected Italian parties. Since the M5S did not compete in a party alliance, its vote share is estimated higher by this proxy in relation to the other individual parties. This also means that in relative terms, for the corresponding article shares based on the individual party names, the M5S should by trend be slightly overrepresented. 12

13 Table 2: Relative voter support and representation in the coverage Party Vote share a Article share Proportion Germany: AfD CDU 53.2% 34.6% 0.6 SPD 34.5% 39.1% 1.1 FDP 7.7% 22.8% 3.0 AfD 4.6% 3.6% 0.8 Austria: Team Stronach SPÖ 47.5% 38.7% 0.8 ÖVP 34.7% 41.3% 1.2 NEOS 4.9% 8.1% 1.6 TS 12.8% 12.0% 0.9 Italy: M5S b PD 40.3% 36.7% 0.9 PdL 28.3% 30.7% 1.1 Scelta Civica 16.9% 21.6% 1.3 M5S 14.5% 11.1% 0.8 Annotations: a = shares according to the averaged polling data (see the previous section for the data sources) for the four respective parties only; the percentage scores are thus relative to the sum of shares for only these respective parties; b = The vote shares for the PD, PdL, and Secelta Civica are based on the vote shares of the vote alliances these parties led respectively (used as a proxy). Looking at these parties themselves more in detail reveals that the intensity of the coverage varies substantially between them. The number of articles per day for the German case (2.8) is half that of the Italian one (5.9). With 7.9 articles per day the Austrian newspaper coverage on the Team Stronach is even higher, whereas the daily number of party (or candidate/leader) references are about the same as the average for the coverage on the M5S. The distribution of this coverage on the three challenger parties over the last month before the election is shown in Figure 1. The graphs show a comparable pattern for the M5S and the AfD: Both show an upward trend with the most intense coverage in the last seven to ten days, albeit at markedly different levels. Unlike the other two cases, the development over time in Austria shows no increase toward the end of the observation period. Instead, there are several peaks until about the end of the first half. These peaks from days 30, 25, 23, and 16 before the election can easily be explained as they were each preceded by a TV debate in which Frank Stronach participated. 5 The first debates with the participation of the TS top candidate Frank Stronach were marked by intensive disputes and several unorthodox political ideas uttered by Stronach and thus provided material for heated public discussions. As to be expected, the Austrian evidence thus clearly speaks for an event-focused coverage. 5 The Austrian public television ORF organized a series of TV debates with the top candidates of the biggest parties. 13

14 Figure 1: Coverage over time: number of articles and references by kind of source Italy: M5S Author External Direct quote Indirect Articles Germany: AfD Author External Direct quote Indirect Articles Austria: Team Stronach Author External Direct quote Indirect Articles Annotations: x-axis = day before the election; y-axis left = number of references; y-axis right = number of articles. The occurrence of eventization in the German coverage is hard to ascertain as there were practically no spectacular campaign events in which the AfD was involved. There was, however, one incident that evoked greater media attention: the peak on day 27 before the election can be attributed to an attack on an AfD campaign event by leftist activists, which had occurred the day before. It also seems that the emergence of and the debate on statements about coalition options, proposals, and rejections has led to a temporary rise of coverage over days ten to eight in the respective graph in Figure 1. In contrast to the AfD, Beppe Grillo s Five Star 14

15 Movement with its leader s Tsunami Tour offered the media more newsworthy occasions for event-focused coverage. The increased media attention on days 27, 21, 5, and 2 can be attributed to preceding major tour events at which Grillo rallied masses of people in greater cities (in Ancona, Turin, Milan, and Rome respectively). Thus, on the meta-level, i.e. regarding the selection and form of media content, there are some similarities between the three cases, which however yield different patterns in the intensity of coverage over time as well as its average level. Moreover, there are also striking differences regarding the ways the parties and their leaders have been referred to. As Table 3 illustrates, while party references by (directly or indirectly cited) external sources have about the same share in all three cases (between 18 and 23 percent), the shares of references by the article s author and especially direct and indirect citations of the party (and its politicians, members or followers) vary considerably more. In Italy, 35 percent of the statements regarding the party are direct quotes of party members, whereas indirect citations of the parties positions were almost absent, amounting to only six percent. For the German AfD this ratio is more or less reversed, with 39 percent of the references to the AfD in the German newspapers being indirect citations 6 and direct citations accounting for only nine percent of the references (20 citations). The Austrian coverage shows a more balanced picture regarding these two kinds of references, while the share of references by the authors of the articles is comparatively high with 46 percent. However, even though there is considerable variation in the shares of direct and indirect citations of the parties, taken together they amount to roughly two fifth in all three cases. Table 3: Total party references by kind of source Author External Direct quote Indirect Italy: M5S 37% 23% 35% 6% Germany: AfD 31% 21% 9% 39% Austria: Team Stronach 46% 18% 21% 16% Issues and Topics Judging from these numbers, the proportion of direct and indirect citations suggests that, basically, the challenger parties had a chance to place their issues/topics in the media and portray themselves in a positive light. Turning to the former aspect first, one has to note that the coverage on policy issues or topics associated with the investigated parties is far from a pivotal aspect in all three cases. At the same time, there are discernible differences between the 6 It should be noted though that this is due to a frequent description of the AfD as Eurosceptic, which was coded as an indirect citation since it expressed the party s self-proclaimed political position. 15

16 coverage of the three parties with regard to case-specific issues as well as the volume of mentioned issues in general. Specific political issues play a negligible role in the Italian newspapers coverage of the M5S. As can be glanced from Table 4, which contains the top five issues and topics (which together make up about 80 percent for the Italian case), the last element that made it on the list was coded only eight times in 178 articles. Even the specific policy issue that was reported the most in relation to the party, economic and social policy, was referred to only 24 times. The other more noteworthy issues are corruption and scandals, the media, and the political system and democracy. The high ranks of corruption and the media at least seem to mirror core issues of the party: Corruption and scandals were in fact core concerns of the M5S and were strongly politicized by the Grillo and his party (see Sberna and Vannucci 2013, 580). Likewise, they show a very negative attitude towards the media, which they see as part of the corrupt political system (Bordignon and Ceccarini 2013, 433). Notably, EU matters were referred to in the context of references to the M5S only five times; the party s position toward the EU, European integration and the Euro can hence hardly be discerned from the newspaper coverage. The clear majority of coverage was about the election campaign as such including electoral prognoses, which often drew upon the astonishing rise of the challenger hence showing distinct symptoms of a horse race journalism. Again, for the Eurosceptic AfD the pattern is more or less the inverse. Not only did the newspapers mention specific issues more frequently in relation to the number of articles. The German newspaper coverage also had quite a lot to say about the party s stance toward the EU and European Integration as this issue has been reported the most in relation to the AfD. It should be noted, however, that this issue association frequently only takes the brief form of simply explicitly calling the AfD a Eurosceptic party and involves no further discussion of its position. Yet even when considering this, the EU-issue is still much more pronounced than in the Italian coverage. What is nevertheless very similar to the Italian coverage is that the election campaign as such is a major object of the newspaper coverage. Together with EU matters, it even dwarfs the three remaining of the top five issues/topics: Economy and social policy, immigration, and family policy. The Austrian challenger party, the Team Stronach, lies somewhat between the other two cases with regard to the mentioned issues. As in the Italian case, the election campaign is the most frequently reported topic. However, policy issues carry comparatively greater weight than in the Italian coverage with economic and social policy issues and law and order being the two most important ones in terms of reporting, followed by EU matters and education. The high 16

17 ranking of law and order is due to a rather special development during the election campaign. It stems from a controversy that Frank Stronach almost incidentally started by declaring that he was in favor of introducing the death penalty for hit men. In sum, while the parties stance toward the EU, Euro, and European integration was mentioned at least with some regularity only in two of the three cases, economic and social policy is the only political issue that is covered notably in all three cases. The predominant topic of reporting in all three countries, however, is the election campaign and the reporting on election polls. These results are in line with the idea of a media logic that leads to an eventization of the coverage and a presentation of the electoral process as a horse race. Table 4: Top five reported issues or topics by party Italy: M5S Germany: AfD Austria: Team Stronach Issue/Topic Issue/Topic Issue/Topic Elections 142 European Union Elections 55.3% matters 46.3% 45.4% Economic and Social Economic and Social 47 Elections Policy 9.3% 34.9% policy 12.6% Corruption, scandals 20 Economy and Social 9 35 Law and Order 7.8% Policy 5.1% 9.4% Media 11 7 European Union 28 Immigration 4.3% 4.0% matters 7.5% Democracy Family Education 3.1% 2.9% 4.0% Sum Top Five % 93.1% 79.0% Total Number of articles Personalization, characterizations, and valence Based on the findings regarding the policy issues one might tentatively conclude that the coverage of the AfD was driven less by the media logic than that of the other two challenger parties. The differing extent to which this coverage put the parties top candidates at the center is also in line with this conjecture. The Italian newspapers focused heavily on the aspect of action and the dimension of politics and geared their coverage mainly to the party leader Beppe Grillo. A cursory look at the quantitative evidence in Table 5 underscores this strong personalization: Only 7 percent of the articles in the sample do not refer to him while this is true for his party in 36 percent of the articles. These quantitative measures only spell out what becomes obvious from the reading of the selected material: The coverage in the Italian newspapers is strongly centered upon the party leader and figurehead of the M5S. The newspaper articles are generally more interested in the party and less centered upon Lucke, which is probably why there are only few direct citations (see Table 3 above). 17

18 Table 5: Articles with references to the party and their top candidates Country/Publication Top candidate/party leader Party Italy: M5S Frequency Articles Article share Frequency Articles Article share Corriere della Sera % % La Repubblica % % Total % % Germany: AfD FAZ % % SZ % % Total % % Austria: Team Stronach Standard % % Presse % % Total % % Annotations: Frequency denotes the number of times the top candidate/the party has been mentioned; articles is the number of articles in which the top candidate/the party has been mentioned; article share is the percentage of these articles with such references in relation to all articles. The reason for the low personalization can be seen in the fact that the economics professor presented himself more as a discreet politician and chose a serious and demure tone, which was not the case with regard to Beppe Grillo and Frank Stronach. Accordingly, the leader and top candidate of the Team Stronach, like Beppe Grillo, was much more the focus of newspaper coverage than was his party: The latter was mentioned in only 40 percent of the articles while Stronach himself was referred to in twice as many articles. This strongly suggests that actions and statements by Frank Stronach were a more newsworthy occasion for media reporting. The aforementioned differences become less pronounced as several similarities become visible when turning to the overall portrayal of the parties (and their leaders): First of all, when we look at the traits that were ascribed to the three Eurosceptic challenger parties, a strong negative bias can be observed. This negativity bias in all three cases can be read from Figure 2, which contains the total valuation of the characterizations for the parties by source of the statement: Firstly, the total number of negative depictions clearly outweighs positive ones in all cases. The graphs also show, secondly, that overall, direct or indirect quotes are naturally more positive, while the tilting of the coverage towards the negative side stems mainly from cited external sources and even the newspapers themselves, i.e. the article s authors. Thirdly, it is notable that the largest share of neutral characteristics are found in the reporting on the M5S with about 60 percent of corresponding attributes. Although there are also quite a number of direct quotes from M5S politicians, members and followers containing party characterizations, the majority of these are neither overtly negative nor positive, too. Only for the cited external 18