List of Tables and Figures
1.1 Research Question and Procedure
1.2 The Treaty of Lisbon
2.1 EU Referendums
2.2 The Media
2.4 Framing Effects
2.4.1 Equivalency Framing Effects
2.4.2 Emphasis Framing Effect
2.5 Framing EU Referendums
2.5.1 Equivalency Framing of the Lisbon Treaty
2.5.2 Emphasis Framing of the Lisbon Treaty
3.2 Article Selection Process
3.3 Measuring Frames
3.4 Methodological Problems
4.1 Presentation of the Results
4.2 Discussion of the Results
Examples of negative framing statements
Complete List of Framing Issues
List of Tables and Figures
Table 1: Results of the Article Selection Process
Figure 1: Figure 1: The multiplicity of possible relationships between the framing concepts
Figure 2: The complete explanation of the influence of framing of the treaty in the news coverage on the referendum result
Figure 3: Percentage of positive and negative framing statements
Figure 4: Percentage of the 1a and 1b framing strategy for positive framing statements and of the 0a and 0b framing strategy for negative framing statements during the first and the second period
Figure 5: The percentage of the framing statements referring to the most important issues in the first and the second period
We voted No to divorce, then Yes. The Nice Treaty was rejected first time around, but was turned around by a newly humbled Bertie Ahern who effectively said: "What you really meant to say was Yes!" (As one No voter told me at the time: "Maybe I should have said: 'No, thank you'.") (Fottrell 2008:3)
1.1 Research Question and Procedure
On June 12, 2008 over 53% of the Irish voters rejected the Lisbon treaty in a nationwide referendum. On October 3, 2009, only 17 months later, 67% of the Irish electorate voted in favor of a ratification of exactly the same treaty. These occurrences are the starting point for this study.
These have not been the first referendums on European integration and scholars already developed very sensible explanations for the voting behavior in the EU. Most prominent is the discussion between the advocates of the second-order theory, which assumes that voters make their decisions in EU referendums in order to punish incumbent governments (Franklin 1995, Schneider and Weitsman 1996) and the issuevoting theorists, who stand that voters decide on the issue of the referendum, whether they are in favor of it or not (Svensson 2002, Downs 1957). But as recent research has shown, both theories can explain some of the referendums, but not all of them. It seems that voters sometimes want to punish their government, and sometimes make their decision on the issue of the referendum (Marsh 2007).
Therefore, a branch of research shed some light on the question why voters vote on different issues on similar referendums. The characteristics of the referendum campaigns and the framing of referendums in the news have been identified as important determinants to explain different voting behavior on similar topics (de Vreese 2007, Hobolt 2006a).
The Eurobarometer (2007) recently showed that for 41% of EU citizen daily newspapers are the main source of information about European topics. Also for those citizens, who decided on the adaption of the Lisbon treaty in the two nationwide referendums in Ireland. But as many media researchers showed, news coverage is biased (Entman 1993, Scheufele 2003). It has to be biased, because it is simply not possible to report the whole reality in one article (Downs 1957).
One aspect of news bias is known under the term framing. Entman (1993: 52) defines to frame as follows “to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.”
A different framing of the same ballot can induce individuals to make different choices. Such effects are known as framing effects (Scheufele 2003, de Vreese 2005a). This paper follows these hints and examines the news coverage before the two Irish referendums on the Lisbon treaty. Applying a content analysis of the two most selling Irish newspapers, The Irish Times and the Irish Independent, it aims to make statements about the framing of the treaty in the news and to give evidence if the question in the title just sounds good or contains some truth.
In the theoretical part I give an overview of the theoretical landscape dealing with EU referendums. Then I specify the function of the media in society and their relationship. After that I introduce the concept of framing and framing effects. The theoretical part concludes with a transfer of this framework to the study of voting in EU referendums to deduce expectation about the framing of the treaty in the news before the two referendums. The empirical part contains the content analysis. It gives further information about the article sampling, the coding procedure and discusses methodological shortcomings of the approach. The fourth chapter presents the results, discusses and analyses them. On the basis of the results the conclusion answers the question, if the treaty was framed differently in the news coverage before the two referendums and if therefore the platitude: “You vote what you read” can be justified to some degree.
1.2 The Treaty of Lisbon
At July 18, 2003 the Convent for the Future of Europe, assigned the „Treaty establishing a Constitution for Europe (TCE), to the Italian Council Presidency. After a negotiating process the treaty was subscribed in Rome on October 29, 2004 (Weiner 2008).
After the two negative referendums on the constitutional treaty in France and the Netherlands, the European Council entered a period of reflection in June 2005. The period took over two years. In this period, the treaty was redesigned mostly during the German Presidency into a reform treaty on the basis of the already existing treaties. This new Treaty was subscribed in Lisbon on December 12, 2007 and was therefore called the “Treaty of Lisbon” (Maurer 2009).
France and the Netherlands renounced to hold a referendum on the treaty. Therefore, Ireland was the only of the 27 EU member states to hold a referendum on the treaty. According to the Irish constitution, the Irish government needs to have referendums on every new treaty concerning an EU integration topic. The first referendum was held on June 12, 2008. In this referendum 53% of the Irish voters voted against the treaty, with a turnout of 53%. This result would have meant a final failure of the treaty in whole Europe, but the government decided to ask to vote a second time on the treaty.
Immediately after the first referendum the EU Representation in Ireland requested a Flash Eurobarometer (2008) to find out the reasons for non-participation and a ‘yes’ or a ‘no’ vote. The Irish Taoiseach1, Brian Cowen, set out the main concerns of the Irish people, identified by public opinion polls, to the Council of the European Union (Council), in order to negotiate special concessions for Ireland as a part of the treaty. This process ended up in a protocol to the treaty that sets out guarantees for the Irish people, concerning the main arguments for a ‘no’ vote. But as the Council (2009) stated, the guarantees clarify but do not change either the content or the application of the treaty. The content of the Lisbon treaty was the same in the first referendum as in the second one. On October 3, 2009 the Irish electorate voted a second time on the treaty. This time 63% of the voter voted ‘yes’, the turnout was around 58% this time.
2.1 EU Referendums
The literature about European integration referendums arranges around two different theories concerning the explanation of voting behavior (see Hobolt 2006b: 155). The first one is the second-order theory. This theory attests that voters in EU referendums merely vote on national issues and domestic party politics. This can be either a positive vote to express a approval to the government or a negative vote in order to punish the domestic government. (Franklin 1995). Schneider and Weitsman (1996) describe referendums as “punishment traps”, in which voters punish governments for bad economic performance (see Hobolt 2006a: 626)
The second theory assumes that voters make their choices on the basis of their general attitudes towards European integration, often described as attitude voting (Svensson 2002). Because there are only two possibilities to vote in European referendums, a ‘yes’ vote or a ‘no’ vote, it is very close to “issue voting”, which evolved from the American two-party system (Downs 1957). Voters vote for the option which is closest to their own ideal point.
But the voter is confronted with a fundamental problem. European legislation is complicated. To find out which choice is closer to his ideal point the voter needs information about the consequences of a ‘yes’ or a ‘no’ vote (Alvarez 1997, Down 1957). To avoid these efforts he could simply rely on what his favored national party recommends him to vote for, so called “partisan voting”. But this does not seem to be the case for the Irish referendums on Lisbon treaty. The government as well as the largest opposition parties recommended a ‘yes’ vote. These parties together held over 80 per cent of the votes in the 2007 Irish parliament elections (van der Eijk and Marsh 2009).
Otherwise voters could vote according to their basic attitudes about Europe and European integration, ignoring the content of the treaty (Franklin et al 1994). According to the fact that the Eurobarometer constantly indicates a widespread support for the EU, even among the ‘no’ voters of the first referendum 59 per cent support the statement that Ireland will benefit in the future from EU membership, this explanation seems inappropriate (Eurobarometer 2007, Leahy 2008).
Therefore, it is unclear on what kind of standards the Irish citizen made their decisions on the Lisbon referendums, have they punished their political class as a whole, that is government and the main opposition parties or have they made their decisions on the treaty itself and their general attitudes towards European integration? As the results of Garry, Marsh and Sinnot (2005) show for the case of the two Irish Nice referendums, both explanations are right to some degree. In the negative first referendum some voters punished the government; others were against the perceived consequences of the treaty. In a more successful second campaign the pro side managed to decouple the unambiguous issues from the treaty and managed it to persuade voters, that the referendum is mainly about the enlargement of Europe and not an evaluation of the Irish government or a question on Ireland´s neutrality (Marsh 2007, Garry et al. 2005).
This shows that the electorate is heterogeneous. Not only in its underlying voting explanation, but also between the first and the second Irish referendum on the Nice treaty people had a different idea what they are voting about. They changed their opinion about the treaty without changing their general attitudes. Marsh (2007: 81) concludes: “To understand why people vote as they do […] it is vital to understand what people think they are voting about.” And he points to Cohen (1963: 13) who stated, that the media cannot tell people what to think, but what to think about.
Hobolt (2006a) shows that the framing of the referendum can play a pivotal role for the voters´ choice. Being one of the main information providers parties “can frame the meaning of the choice that voter face in referendums” and therefore, they “can convince voters to vote in favor of (or against) a ballot proposal by framing the ballot proposal as close to the ideal point of the median voter and the reversion point as more extreme.” (ibid. p.641). In Europe the main source of information is the media. Thus, the next two chapters examine the role of the media in referendum campaigns and introduce the concept of framing and framing effects.
2.2 The Media
As society has outgrown the town meeting, questions have arisen as to how far democratic communication is possible in a mass society. In our society, extension of the political debate far beyond the limits of face-to-face contact is made possible by the existence of the mass media. Newspapers, magazines, radio, and now television are essential for the process of “making sense of the meeting” when the meeting involves more than fifty millions participants. (Berelson et al. 1957: 234)
What Berelson, Lazarsfeld and Mc Phee (1957) assure for their study of different US presidential election campaigns between 1940 and 1952, still stays true for the first decade of the 21st century. Of course the internet joined the list, but like the Eurobarometer (2008) recently showed newspapers are still the most important source of information for European citizens, concerning EU topics.2
Of course, every European treaty is available in the internet and can be ordered as a hardcopy, so they are available for everybody, but treaties are written in a formal, juridical language and therefore difficult to read. Even the Irish EU Commissioner McCreevy admitted these difficulties when he said that he does not expect “any sane and sensible person” to read it cover by cover (Independent 24.5.2008). Newspapers inform the voter about the object of the referendum.
But the media has to filter the information of the treaty or change the form of the information. Otherwise the newspaper has to print the full treaty, which probably would not be very sales promoting and impossible because space in newspaper is limited. This gets more obvious for reports on events during the referendum campaign. The reporter cannot report any detail of the reality. He has to simplify, interpret and select the reality and its information to produce an article. A second selection step takes place in the editorial department. The next chapter discusses these selection criteria as “input frames” (Scheufele 2003, de Vreese 2005).
Every reporting has to be biased because there always has to be a selection of the reality. For the news coverage on the Lisbon treaty this means that there cannot be a full presentation of all possible consequences of a ’yes’ or a ‘no’ vote, or a complete presentation of the positive and negative aspects of the treaty. But this does not necessarily mean that the reporter propagates one interpretation or another: “When we speak of reporting as biased, we are not implying, that the data therein are false, since we have assumed all data are accurate, nor that the reporter is immoral, since bias cannot be avoided.“ (Downs 1957: 212).
This quota implies a rather functional understanding for the relationship between media and society. This means that the media does not have an own agenda to show a biased coverage. It needs neither to be in favor or against the treaty, nor does it need to be seen as an instrument to any kind of elite as some critical analysts suggest (Murdock and Golding 1995, Bonfadelli 2002).
Of course, one cannot separate the elite, the media and the voters. There are many connections and intersections between them. But a theoretical simplification is necessary for the purpose of this paper. System theory offers a fruitful starting-point. I follow Barker (1951) who arranges all parts of society in a “system of discussion” (Barker 1951: 41). This does not imply that they are autonomous subsystems as Luhmann (1995) suggests. On the contrary: They are interdependent. The elite, such as political parties and interest groups give the input, the media serves as a “stage” to present this input and the audience consist of the general citizen. (see Berelson et al. 1957: 234)
Of course, everyone who holds a specific opinion on the treaty or has a certain interest in a specific referendum result wants to be visible on this stage in order to promote his opinion and to persuade voters. But as already mentioned above a selection takes place and, to speak in the stage-metaphor not everybody has the same time on this stage.
It is likely that the more often opinions, arguments, evaluation and interpretation are present in the media, the higher is the probability that voters take them into consideration for their opinion building. But also the other way around, the dominant public opinion can serve as an input and enter the media stage.
Until now I only spoke very vague about biased reporting. The next chapter introduces the concept of framing to give a more precise understanding of biased news coverage and its possible effects.
This chapter has a closer look on the concepts of framing. It gives a simplified overview of the different applications of frame concepts, by arranging them into different stages. Having its origin in the study of human psychology the term “frame” has made an interdisciplinary academic career (Bonfadelli 2002). Thus, it is not surprising that a number of definitions are in use. In a rough outline, the research focusing on the framing of news has three different research agendas and different definitions of frames. For a better understanding how they are linked, I arrange them into three stages: The input stage, the output stage and the outcome stage.
The first concept emphasizes the input to the media. On this stage frames play an important role for the mobilization of social movements. A frame for these researchers is a schemata which “is an interpretation that enable individuals to locate, perceive, identify and label occurrences within their life space and the world at large” (Snow et al. 1986: 464, quote from Scheufele 2003). In this vein a frame is a necessary prerequisite in order to mobilize possible supporters and to get the frame into public discourse and in the media (Benford and Snow 2000, Scheufele 2003).
Another aspect of the input stage is the journalist centered approach. It describes in which way the cognitive habits, constraints and routines of journalists result in a selection of certain themes. Frames allow journalists to handle the flood of information in their everyday work (see Gitlin 1980: 7). Through categorization and typologies journalists classify events and produce “news-themes”: “A news theme allows journalists to cast an incident as an instance of something.” (Fishman 1978: 534, quote from Scheufele 2003 (italics in original)). This branch of research examines frames mainly as a dependent variable (de Vreese 2005).
The second stage is the media coverage itself, so to speak the output stage. Frames on this stage “describe an attribute of the news itself” (Entman 1991: 7). Because articles in newspapers are direct results of the journalists´ working process, the output frames cannot be strictly separated from the input frames. But because this paper only deals with the effects, not with the “production” of news, it is analytically necessary to separate them. So I only consider the output stage, consequently the news coverage. Frames on this stage merely serve as an independent variable for research. For this purpose the definitions given above are unwieldy, although they are not unimportant (de Vreese 2005, Scheufele 2003).
I apply a definition of framing given by Entman (1993: 52) for the study of news coverage: “To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” A frame is the readable or visible (if pictures are included) result of this process. Although Entman´s (1993) research design identifies descriptively different frames in news coverage and does not use them as a variable, the definition is working to identify and use frames as an independent variable, too.
On the outcome stage these communicating frames can influence the individual frame in thought of the consumer, which can be a reader, a listener or a viewer. Of course, this individual frame can become a communicating frame again (Scheufele 1999, Druckman 2001).
Figure 1: The multiplicity of possible relationships between the framing concepts
illustration not visible in this excerpt
Source: Compiled by the author
1 The Taoiseach is the Irish Prime Minister(Constitution of Ireland §13.1)
2 The terms media and newspaper are from now on used as synonyms.
- ISBN (eBook)
- ISBN (Book)
- File size
- 2.3 MB
- Catalog Number
- Institution / College
- University of Constance – Politik- und Verwaltungswissenschaft
- Framing news coverage Framing effects voting behavior campaign election referendum Ireland Lisbon Treaty EU Referendum content analysis public opinion issue voting second order Referendum Theory Framing Theory decision making media theory Ireland EU Referendum Irish Referendum