Framing Nation- States and the E.U. in
the Light of the Refugee Crisis
The Responsibility Burden
Graduate School of Communication Master’s Thesis
Master’s Programme Communication Science
Submitted by Tsatsani Marianthi- Asimina 10915834 Supervised by dr. R. (Rachid) Azrout
2 Abstract
Formation of nation images has, due to its importance, been the object of many studies. However, not enough attention has been paid in the unintentional nation image building as a result of a crisis. In this spirit, the main objective of this study is to examine how are nation images, as well as the portrayal of European Union influenced in the light of the current refugee crisis, focusing on the question of responsibility. Moreover, this research constitutes an effort to find and record differences in nation-images formation through framing, across the German, Greek and British press in the coverage of the refugee crisis. To answer our hypotheses, a content analysis of the coverage of the refugee crisis was employed. Using the theoretical tools of nation- images formation, ethnocentrism and framing, we collected and analysed nine widely circulated newspaper outlets from Germany, Greece and the United Kingdom. The results indicated that the depiction of the European Union was more favourable in the German, compared to both the British and Greek press. Nonetheless, all three countries seek a European solution to the refugee crisis, probably for different reasons. When it comes to the portrayal of Greece in the foreign press, it was found that there is a more favourable treatment in the German press. Finally, looking for differences within the Greek press, it was shown that the responsibility framing of Greece was more prominent in the tabloid, compared to the quality press while the opposite was witnessed regarding the victim framing of Greece. Overall, the study indicated that nation image building in times of crisis is to a large extent ruled by national goals.
Keywords: refugee crisis, nation images, responsibility framing, victim framing, Greece,
3 Introduction
Nation branding is a vital part of public diplomacy that focuses on promoting a nation's image (Kunczik, 2000). As defined by Kunczik (2000), a nation image is “the cognitive
representation that a person holds of a given country, what a person believes to be true about a nation and its people”. This representation can intentionally be influenced, for instance when a nation state dedicates resources to build its reputation in the international system, or
unintentionally when certain events contribute to the establishment of a state image. Media coverage of foreign affairs is what defines the dominant image of a nation state (Kunczik, 2000; Samaras, 2007). The formation of nation images has been the object of many studies (Boulding, 1959; Herrman, 2003; Kunczik, 1997, Manheim & Albritton, 1984). Its
importance lies on the ascertainment that it affects greatly how nations perceive each other and therefore their relations (Dell'Orto, Dong, Schneeweis and Moore, 2004). Nevertheless, for the present study, an even more important motive, is the influence of nation images projected in the media on readers’ perceptions (Dell’Orto et al., 2004). According to Perry (1987), the effects of news reporting tend to be stronger when the readers have no personal experience on the topic, when they do not consider the issue to be of great importance and finally when they can contrast the issue with a reality they have knowledge of. All three criteria generally apply in the reporting of foreign news (Perry, 1987). He additionally found that audiences are apt to build the perception/ image of a certain group or nation, based on misleading news (Perry, 1987). This is what constitutes the study of nations are represented in foreign media vital.
A typical situation that triggers unintentional nation- image formation, is that of a crisis. Currently, the world is undergoing one of the harshest crises of the past decades, as turmoil in Middle East and Northern Africa and other conflict areas have driven millions of people to embark on dangerous journeys to seek asylum. Emotionally charged images and
4
news reports have acquainted the issue to the wider public and triggered a discourse in the political and public sphere. Consequently, questions have been generated about Europe’s response to the crisis in terms of humanitarian action, as well as policies dealing with the increasing numbers of refugees coming to Europe. In turn, press engages in blame games, regarding who is burdened with the responsibility of both creating and handling the refugee crisis. This blame game constitutes the focal point of this study. Moreover, special attention will be paid in the image of Greece, which during the past decade, has suffered extensively due to the coverage of the economic crisis at first and the refugee crisis subsequently. As both of these crises constitute besides of internal problems, a source of uncertainty for Europe as a whole, they received extensive coverage in the European media.
Therefore, this research is an attempt to provide an answer to the question of responsibility attribution and imprinting the process of framing of Greece as well as the European Union. Moreover, an attempt will be made to connect the two by examining the extent to which the attitude a country holds towards the EU, affects the way its media attribute responsibility. The importance of this study lies on the fact that it expands the literature on nation- images formation, by investigating how an international crisis can constitute a source of unintentional nation- image building. To our knowledge, the pre-existing literature focused more on the impact of international reporting regarding politics and international relations on nation- state images (Kunczik, 1997; Boulding, 1959). Therefore, the current refugee crisis allows us to explore this territory. The first part of this study constitutes a review of the theoretical background. Particularly, the theoretical part will elaborate on nation images projected by the media, on the different approaches of the European Union as observed and studied in the under examination countries and the process of framing. Consequently, the methodology of the paper will be outlined, as well as the results of the reliability analysis. Finally, the results of the analysis will be presented, followed by the discussion and
5
conclusion.
Theoretical Framework
The mediated images are largely affected by ethnocentrism as a factor of the news making process. Proximity between nation states and communities, either on geographical or on cultural terms, constitutes an important criterion for the selection of events to be reported (Gans, 1979; Harcup and O’Neill, 2001). Gans (1979) defines ethnocentrism as one of the news selection criteria according to which, a particular event that takes place in a foreign nation is more likely to be covered by the media when it touches upon national interests or is an object of foreign affairs policy. From a different scope, Sumner used the term
ethnocentrism to define the tendency to view things as one's own group (in- group) is in the centre of everything, and all others (out- group) are scaled and rated with reference to it (Duckitt, Callaghan and Wagner, 2005). As a result, the assessments of in- groups and out- groups are much biased in favour of the in- group, leading to a biased media coverage and consequently a distorted own as well as other nations image formation (Müller, 2013).
Because ethnocentrism has been associated with the perception of superiority of the in- group in comparison with the out- group, it carries a fairly negative connotation (Duckitt et al., 2005). Nevertheless, independently of ethnocentrism’s negative consequences for intercultural communication, the tendency to evaluate others according to one’s own standards is natural (Samaras, 2007). And as a consequence of its fast and unconscious nature, it is to a great extent inevitable (Lamont & Molnár, 2002; Samaras, 2007).
Ethnocentric behaviours often stem from the diverging national identities (Samaras, 2007). Wimmel (2009) in his study focused on the effects of the European borders on
ideological position, in the case of the under discussion integration of Turkey in the European Union. He distinguished two main approaches; namely the state- dependent or camp-
6
number one factor that defines ideological positioning. On the other hand, in camp-
dependency, social groups take the lead and overcome national borders in shaping attitudes on a public discourse. The present study takes on the first approach, thus the state- dependency. Notwithstanding it would be very interesting to investigate within- country differences between opposing ideological backgrounds, we chose to focus on the differences between countries.
Therefore, apart from the frames that derive from journalistic routines and news making process, as defined by Nossek (2004), there is also the “national framing” that can act as a mega- frame promoting ethnocentrism of the press. According to the national “mega- frame”, when it comes to politics, despite of any changes in the political, economic and communication field during the past century, foreign news reporting is always an outcome of an interaction between national identity and interests (Nossek, 2004; Handley, 2009).
Moreover, the type of reporting is defined by the question: ‘is it ‘our’ event or ‘theirs’? When the answer is ‘theirs’, reporting is governed by journalistic professional norms. However, when national interests are at stake, journalists tend to adopt a certain narrative and frame events in a way that promotes and protects these interests (Cottle, 2009; Handley & Ismail, 2010; O' Regan, 2007; Handley, 2009). Moreover, in times of transnational crises, like the current refugee crisis, the way in which an event is reported varies in different national media (Cottle, 2009). One important factor contributing to this development is the need for the reporting to respond to the national norms and culture (Cottle, 2009). With that in mind and taking into consideration the questions regarding national security that have been raised as a result of the large influx of refugees, we expect that the ‘national narratives’ will prevail when it comes to the coverage of the crisis.
This massive displacement of populations, which started in the summer of 2015 and is still ongoing, transformed the status quo of the European borders and rose a public discourse
7
regarding what the next step towards dealing with the crisis should be. The extent to which national identity influences the coverage of the refugee crisis as well as the attribution of responsibility, depends on the countries themselves and the structure of their public sphere. However, as the refugee crisis constitutes a European ‘puzzle’, it is expected that the media coverage will be inextricably related to the vision and attitude of each country towards the evolution of the European Union. In this spirit, the three countries that were selected for this study, namely Germany, Greece and the United Kingdom, represent a different mind-set towards the European construction.
In the German case, European integration has always been in Germany’s essential interests, something that Bulmer addressed as Europeanization of German politics
(Marcussen, Risse, Engelmann- Martin, Knopf & Roscher, 1999). The end of the World War II left the German state in need of a new identity, cut off from its nationalistic past. This void was covered by the development of a Europe- oriented nation state identity (Marcussen et al., 1999). In this spirit, Germany appears to desire further integration and empowerment of the European Union and its institutions (Wimmel, 2009). On the other hand, the United Kingdom has always been sceptical and reserved towards its participation in the European Union. Over twenty years since the UK’s entry in the European Communities, and Britain is still
considered as ‘the awkward partner’, who is ‘semi- detached’ from the European Union (Marcussen et al., 1999). Regarding identities, in the case of Britain there is still a very prominent division between ‘us’ (England) and ‘them’ (the Continent). This feeling expands in the political discourse as well, where Europe is considered nothing more than a friendly other (Marcussen et al., 1999). Thus, any further empowerment of the European institutions and political cooperation are rather undesirable for the United Kingdom (Wimmel, 2009). Last but not least, in regards with Greece, the national identity serves as a reminder of belonging to the in- group and not the out- group, while the European identity is understood
8
as a status of cooperation, interaction and sharing common benefits with other nations (Sereti, 2004). The ethnocentrism of the Greek media that stems from the supremacy of national identity acts as a restraint to further integration (Samaras, 2007).
Drawing from this theoretical background and bearing in mind that an effective European solution to the crisis might demand further integration and granting part of
sovereignty and control to European institutions, it is expected that the German press is more likely to promote a European solution to the refugee crisis and thus build up a narrative, in which the European Union is presented as an entity. On the other hand, the United Kingdom, as it already abstains from several European institutions, is more likely to view the European Union not as an entity but as a flawed supranational organisation where each of the members promote their national interests against the Union’s founding principles and thus, count on a national solution to the refugee crisis. Finally, in the case of Greece, considering the severity of the refugee crisis as it unfolds in the Greek islands, it is expected that contrasting the former findings, Greek press will also be more likely to support a European solution. Therefore:
H1: The European Union is more likely to be perceived as an entity in the German
and Greek compared to the British press.
H2: Germany and Greece are more likely to support a European solution to the
refugee crisis, whereas the UK is more likely to promote a national solution. Additionally, regarding the portrayal of Greece, it is expected to be more positive in the German than in the British press. This hypothesis is built based on the views each country holds about the European Union, as well as Sumner’s definition of
ethnocentrism as aforementioned. Because of Germany’s European orientation, it is expected that Greece is considered part of the in-group. On the other hand, in United Kingdom’s case, Greece is considered to be part of the ‘other’, thus outgroup. Hence,
9
due to the natural tendency to favour the in-group, the German press is more likely to be, if not positive, less negative towards Greece.
H3: The evaluation of Greece will be overall more positive in German press in
comparison to the British press.
A very important dimension of ethnocentrism is framing theory (Gurevitch, Levy & Roeh, 1991). A frame has been defined as the central organizational idea or narrative
structure, which offers meaning to an unfolding flow of events (Gamson & Modigliani, 1989). By the process of framing, some aspects of an issue are stressed against some others and are, thus, rendered able to offer a certain meaning to the issue (Entman, 1993). In his study, Entman (1993) recognizes several functions when it comes to framing, which are summarized as follows; definition of the problem, diagnose of the cause, moral judgments and
recommendation of treatment. Furthermore, interpretative frameworks serve two purposes; they constitute on the one hand the way in which news is structured and on the other hand the way the audience perceives reality (de Vreese, 2005). Therefore, they are distinguished in frames used by the media and individual frames, thus intellectual constructions that each member of the audience calls forth to process and comprehend incoming information (de Vreese, 2005; Cappella & Jamieson, 19971).
When it comes to the typology of news framing, scholars have not reached a consensus regarding the identification of news frames (de Vreese, 2005). Nevertheless, researchers have distinguished two main types of framing; namely the issue- specific and generic news frames (de Vreese, Boomgaarden & Semetko, 2011). The issue- specific frames refer and are
applicable only to certain topics or situations (de Vreese, 2005). On the contrary, generic framing is not subject to theme- related boundaries and can be spotted throughout different
1
This distinction provokes the vital question; to what extent the frames generated by the media influence the way certain cognitive structures are activated on an individual level, hence affect the audience’s perception of reality. Several past studies have found evidence that indeed different interpretative frameworks carry the capacity to affect individual perceptions and consequently public opinion (Kahneman & Trevsky, 1984; Iyengar, 1991; Aarøe, 2011).
10
topics, time periods as well as cultural contexts (de Vreese, 2005). Several studies have focused on generic framing. Iyengar (1991) who studied the effects of news reporting on the way the audience evaluate politicians and attribute responsibility for the ills of society, distinguished two main framing types: thematic and episodic. Episodic interpretative framework focuses on specific events, while thematic framing examines political and social issues without isolating them from the broader context. A different approach to generic framing was made by Neuman, Just and Crigler (1992) who identified several main types of framing commonly used in news reporting including ‘conflict’, ‘economic consequences’, ‘human impact’ and ‘morality’ frames (Semetko & Valkenburg, 2000). Drawing from the aforementioned studies by Iyengar (1991) and Neuman et al. (1992) as well as the theoretical work of several other scholars, Semetko and Valkenburg (2000) constructed a generic
framing typology that consists of five frames. In particular, they identified the ‘conflict
frame’, ‘human interest frame’, ‘economic consequences frame’, ‘morality frame’ and last but not least ‘responsibility’ frame.
The present study is based on the typology of generic frames as developed by Semetko and Valkenburg (2000). More specifically, the focus of this analysis lies on the use of
responsibility framing in the coverage of Greece with regard to the refugee crisis, as well as other countries involved in and affected by the refugee crisis. The rationale that guided the choice of this frame, is that it responds better to a crisis situation. By definition, a crisis as ‘an event for which people seek causes and make attributions’ (Coombs and Holladay, 2004, p. 97), justifies the tendency of news media to adopt the ‘blame game’ and place responsibility for the crisis (Seon- Kyoung and Gower, 2009). According to the Semetko and Valkenburg typology, the responsibility frame refers to stressing the aspects of an issue in a way that attributes responsibility for either causing or coping with it, to an individual, a group or the government.
11
H4a: Compared to the Greek press, the responsibility framing of Greece will be
more prevalent in the German and British press.
H4b: The responsibility framing of Greece will be higher in the British compared
to the German press.
Furthermore, in the study by Semetko and Valkenburg (2000), the ‘attribution of
responsibility’ while also very popular in the sensational press, it appeared to be the most widely used frame in quality newspapers. Drawing from this conclusion, we will indicatively test for within- country differences in the Greek press. It is expected that the responsibility framing will be more widely used in the quality newspapers.
H5: The responsibility framing of Greece will be slightly more prominent in the
Greek quality newspapers than in the Greek tabloid.
The coverage of the European refugee crisis, besides the attribution of responsibility, is defined by the dynamics of events and is loaded with features of compassion. As the reporting passes from a crisis itself to its causes and responsible actors and the dominant narratives are formulated, processes of counter- framing also make their appearance (Samaras, 2009). Therefore, in the study we focus not only on the responsibility framing of states, but also the victim framing. The function of each of the frames, results to different portrayals of nation states that directly influence the nation image. When used in crisis situations, the victim frame is in constant and inevitable interaction with responsibility attribution (Seon- Kyoung and Gower, 2009). The victim frame as employed in this study, focuses on the humanistic side of the issue, in an effort to dramatise the situation, bringing an emotional angle to the
presentation of an event. For this study, we look into the images of nation- states, thus ‘the cognitive representation that a person holds of a given country, what a person believes to be true about a nation and its people’ (Kunczik, 2001). Therefore, we will in particular look into
12
the victim framing of people, the government and all the elements that constitute a nation- state. Victim framing presents many similarities to the human interest frame, as defined by Semetko and Valkenburg (2000). Even though human interest frame was not entirely applicable in this study, it guided us through our expectations regarding victim framing. Particularly, it has been found that the human interest frame is notably more prominent in the ‘sensational’, thus ‘tabloid’ press (Semetko and Valkenburg, 2000; Van Cauwenberge, Geldens and Joris, 2009). Based on this knowledge, we expect the victim framing to be more prominent in the tabloids, compared to quality newspapers. Moreover, research has shown that internationally oriented reporting tends to adopt more pragmatic frames, as consequences or responsibility, whereas in nationally oriented reporting the human interest is more present (Van Cauwenberge et al., 2009). Europe as a whole is facing one of the most severe crises of the past decades. Nevertheless, Greece remains one of the countries that have been most affected. To that spirit, it is expected that in the Greek press, the refugee crisis will be treated as a rather national crisis and therefore the human interest frame will be more widely used. On the other hand, when it comes to German and British press, it is expected that the framing process will more likely be focused on responsibility attribution. Building on the
aforementioned theories, the following hypotheses will be examined:
H6a: In comparison to the British and German press, the victim framing of Greece
will be more prominent in the Greek press.
H6b: The victim framing of Greece will be more prominent in the German press,
compared to the British one.
Methodology
The rising interest in framing theory has resulted to a vast number of methods for identifying media frames and their effects (David, Atun, Fille & Monterola, 2011; Matthes, 2009). The
13
focus of the present study lied on nation images projected by the media and furthermore, in a comparison between media outlets from Germany, Greece and the United Kingdom.
Therefore, a manual quantitative content analysis was employed.
Sample
For this research, press from Germany, Greece and the United Kingdom was examined. In particular, from each country, we selected two quality newspapers and one tabloid. The choice of quality newspapers was made after consideration of mainly two factors: circulation as well as political orientation. The chosen newspapers are some of the most popular and high on sales in their genre. Furthermore, for each country under examination, quality newspapers from opposing sides of the political spectrum were preferred in order to avoid biased results and to be able to make generalizable conclusions regarding the national coverage of the refugee crisis. Additionally, for the sample, we chose to use Sunday papers, where applicable, or weekend editions. The rationale behind this choice, is that in the case of Greece and
Germany, the weekend editions tend to report important events of the whole week, presented in a more extensive and analytical manner. Last but not least, research has indicated that weekly editions appear to have similar variation in content measures as daily ones (Lacy, Robinson & Riffe, 1995), therefore offer a very similar portrayal of events. Regarding the British press, even though daily newspapers demonstrate higher sales we had to follow the rule of Sunday newspapers sampling. To reach our choice of outlets, we also took into consideration the circulation, choosing newspapers that present the higher sales. Therefore, from the German quality press, ‘Süddeutsche Zeitung’ (centre – left) and ‘Die Welt’ (centre – right, moderately conservative) were selected. From British and Greek press, ‘Sunday
Telegraph’ (right – wing, conservative), ‘The Observer’ (centre – left) and ‘Kathimerini’ (right – wing), ‘I Efimerida ton Syntakton’ (centre – left) were selected respectively. In the selection of tabloids, the main criteria were popularity but also availability, which led to the
14
choice of ‘Bild’ (Germany), ‘The Mail on Sunday’ (UK) and ‘Proto Thema’ (Greece). From these newspapers, only articles that have a direct reference to the current refugee crisis were analysed.
News articles were retrieved in various ways. The German outlets were partly available on LexisNexis. To find the articles we used the search-terms ‘Flüchtling OR Geflüchtete Asylsuchende’. Our search in the newspaper ‘Die Welt’ resulted in 327 articles with the refugee crisis as the main topic, out of which 78 were randomly selected. In ‘Die Süddeutsche Zeitung’ our search resulted in 122 related articles, out of which a sample of 79 was randomly selected. Finally, the ‘BILD am Sonntag’ was found in hard copy version in the library archive of the ‘Staatsbibliothek zu Berlin’. In this case, the issues were scanned manually using the same search criteria. From each issue, approximately two articles were randomly selected, resulting in 75 articles. The British newspapers were all available on the digital database Lexis Nexis. The terms used for the search were ‘refugees OR refugee crisis’. The term ‘migrant’ was also used in the first search. However, because of the timelessness of the migration situation in the United Kingdom, the majority of the results were irrelevant to our object of interest. Therefore, the term ‘migrant’ was dropped. The Greek newspapers were not available in Lexis Nexis. After contacting the newspapers’ offices, we were offered free access to the online archive of ‘Kathimerini’. Because of time restriction we did not manage to find a similar solution for the other newspapers on time and ‘I Efimerida ton Syntakton’ and ‘Proto Thema’ were bought from the online platform ‘www.readpoint.com’. The terms used for the search of the articles were ‘πρόσφυγες ή προσφυγική κρίση’, which are the equivalent of ‘refugees OR refugee crisis’ in Greek.
Time Frame
The European refugee crisis started already in 2015, when more than a million migrants and refugees attempted the treacherous journey to Europe. The time frame we chose for this study
15
begins from the first of July 2015 and ends on the twenty-fourth of April in 2016, which resulted in a sample consisting of 43 issues2 per newspaper (1st July 2015- 24th April 2016). The number of papers differed in the case of Greece because of strikes of journalists’
association, giving us 37 issues of ‘Kathimerini’, 36 of ‘I Efimerida ton Syntakton’ and 39 of ‘Proto Thema’. Defining our time frame was accompanied by some difficult choices. The final decision was guided by our wish to include in our sampling period, events that played a vital part in the configuration of the disposition of the press towards the refugee crisis itself as well as the construction of nations’ images. Indicatively, we wanted to include in our sample, the publishing of the picture of the dead kid in the shore Alan Kurdi, the Paris attacks, the Köln attacks and finally the agreement of the European Union with Turkey about controlling the influx of refugees through Greece. The final sample consisted of 684 articles3 related to the refugee crisis.
Coding Procedure
The first step of this research consisted of the collection, overview and processing of the newspapers from 1st July 2015 until 24th of April 2016. Τhe processing of the outlets was a combined effort of a native German speaker and a Greek one, both of which are fluent in English. The newspapers were thus distributed equally among the coders. With the help of a codebook we processed the retrieved articles which referred to the current refugee crisis. This codebook was constructed by some ‘borrowed’ from past studies variables, as well as some
2
The selection of articles from the retrieved outlets differed according to the database they were retrieved from. For the articles found in Lexis Nexis we used systematic sampling, coding every 3 or 4 articles as they appeared in the search results, depending on the total number of articles. For the rest of the newspapers, we decided to code 2 to 3 articles per issue. The articles were randomly selected from a pool of articles related to our research topic.
3
These 684 articles were distributed among the selected newspapers as follows; 79 articles from ‘Süddeutsche Zeitung am Sonntag’, 75 from ‘Die Welt am Sonntag’, 78 from the ‘Bild’, 79 were from ‘the Sunday Telegraph’, 71 were from ‘the Observer’, 74 were from ‘the Sunday Mail’, 76 were from ‘Kathimerini’, 77 were from ‘I Efimerida ton Syntakton’ and 75 from ‘Proto Thema’.
16
newly invented categories that served our research purposes. Our final coding manual consisted of two parts.
The first part included the categories which referred to the article characteristics, framing of the refugee crisis itself, as well as the reference of the article to Greece. More specifically, in this stage we coded the placement of the article, namely the page number on which it was published, its size, the type of publication, namely if it constitutes a pure news article, a background analysis, an interviews or an opinion piece, and the sources present in
i
the article. In this stage, the image of the European Union was also coded. Το measure EU image, we coded the suggested degree of unity and cooperation between member states. The coding options were the following five: portrayal of EU as a united entity (7%), EU divided in groups of countries (6.6%), EU as an entity where nation- states act independently according to their own interests (8.5%), EU mentioned in the article but without an evaluation (30.6%) and no reference to the EU (47.3%). Subsequently, the framing of refugees was coded in terms of intruder and victim framing. However, this part of the codebook was not used in the analysis of the present study. Last but not least, reference to Greece was coded using 3
categories, namely the centrality of reference, the actor that is mentioned in the article, as well as the evaluation of the actor. For the items measuring the evaluation of actors, 5-point scales were used, with values ranging from -2 as very negative to +2 as very positive, depending on whether or not intense evaluative judgments were to be found in the articles.
The second part of the codebook focused on the presence of interpretative frameworks in the reference of nation states. As aforementioned, the focal point of this study lies in the responsibility attribution as well as the victim framing of nation states and other entities we included in our analysis, which was considered to be the antipode of responsibility attribution. In this stage of coding, we divided responsibility attribution in two items, namely
17
solutions and responsibility for dealing with the crisis. Each of these variables was followed by a question regarding the nature of the responsible actor. Subsequently, the presence of victim framing was coded, assessing to what extent each of the coded countries was presented as a victim of the current circumstances. To measure the occurrence of responsibility and victim frames we used 3 answer options, not present – present – very present. However, in the analysis, the framing variables were recoded into binary variables, by merging the present- very present options.
Reliability Analysis
To make sure the coding procedure was reliable, before examining the hypotheses, an inter-coder reliability analysis was conducted using all the variables of the analysis. Out of the operational sample of 684 published articles, 64 cases were selected for the inter – coder reliability analysis, using percent agreement and the Krippendorff's Alpha, which have been found to be the most accurate estimators of coding reliability (Hayes & Krippendorf, 2007). The factor that defined the selection of articles was mainly the language, which had to be fluently spoken by both coders. Therefore, the reliability analysis was based only on articles from the British press. The results indicated that the majority of our variables were reliable, taking into consideration that the KALPHA was higher than .80. Nevertheless, the results also indicated some problematic variables. Firstly, in the objective scale measuring victim-
intruder framing of refugees the statements about moral duty to help refugees and the use of victim metaphors, gave us a KALPHA of .54 and .65 respectively. The low reliability results can be explained by the subjective nature of the judgment regarding what constitutes appeal to moral duty and emotionally charged use of metaphors, as well as by the difficulty to
distinguish between the answer options ‘a little’ and ‘yes’. Another variable that indicated relatively low reliability results was the one recording the responsible actor for creating or not dealing successfully with the crisis, indicating a KALPHA of .64. An explanation for this
18
result is the number of responses, as this question was double- filtered resulting to a number of responses smaller than the number of cases in the sample. However, mainly due to time restrictions, we decided to include these variables in our analysis, holding some reservations about the generalisability of the final results. The analytical results of the reliability tests are presented in Appendix II.
Data Analysis
In order to examine the first hypothesis, descriptive statistics were used. In particular, a cross- tab analysis was conducted, using the image of the E.U. variable (nominal) and newspapers variable (nominal).For our third hypothesis, because of the dependent variable violating the normality assumption, we employed a non- parametric test to conduct a comparison between different groups. Particularly, two Mann- Whitney U tests were run, using the country of origin of the outlets as a grouping variable and the evaluation of Greek politicians and citizens as dependent variables. The rest of the hypotheses were analysed using binary logistic
regressions.
Moreover, to test the second hypothesis, we ran a logistic regression after splitting the dataset according to the country of origin of the outlets. In this model, attribution of
responsibility for handling the refugee crisis was used as a dependent variable. Using the variable of countries mentioned in the article, we selected references to Germany, Greece, United Kingdom and European Union and after recoding them into dummies, they were used as independent variables. For the analysis of hypotheses 4a and 4b we recoded the
responsibility framing variable into a binary, selecting only the cases in which Greece was coded. Using the aforementioned variable as dependent variable and the binary variables of newspaper outlets as independent we ran the binary logistic regression. For hypothesis 5, another logistic regression was conducted, where the responsibility framing of Greece was
19
used a dependent variable. For this hypothesis, the variable newspaper outlet, was recoded into 3 dummy variables, after selecting only the Greek newspaper. The recoded dummies were inserted as independent variables in the model. Similar process was followed for the hypotheses regarding victim framing.
Results
H1: The European Union is more likely to be perceived as an entity in the German and
Greek compared to the British press.
In order to test our first hypothesis, a cross- tabulation analysis was employed. For the analysis we only took into consideration the articles that carried a reference and an evaluation of the European Union, which summed 755 across the newspapers outlets. The results indicated that there is a significant relationship between that country of origin of the newspaper and the portrayal of the European Union (p<.001), but this relationship is moderate (Cramer’s V= .30).
Table 1.
Portrayal of the European Union
German newspapers English newspapers Greek newspapers E.U. characterised by unity and equality 43.3% 32.7% 25.8% E.U. divided in groups of countries 56.7% 20% 25.8%
Every member state
for itself 0% 47.3% 48.5%
Total number of
articles (N) 150 275 330
Particularly, the analysis indicated that in the German press, the majority of articles portrayed the European Union as divided in groups of countries (56.7%). Nevertheless, about 43% of
20
the articles perceived European Union as an entity defined by unity and equality among the member states. These percentages, as presented in Table 1, differed considerably in the case of Greece. The majority of articles portray the E.U. as a union in which member states act according to their national interests (48.5%), while 25.8% of the articles depicts the European Union as divided in groups of countries. In turn, the rest 25.8% portrayed the E.U. as a unity. When it comes to the British press, the majority of articles presented E.U. in an ‘every member for itself’ way (47.3%). About 20% of the articles portrayed the European Union divided in groups of countries. Finally, the European Union was characterised by unity and equality in about 33% of the examined articles. Therefore, the first hypothesis was partly supported. In the German press, unity of the E.U. was indeed more prominent than in the case of the United Kingdom. Nevertheless, the hypothesis was not confirmed in portrayals of the Union in the Greek press.
H2: German and Greek press are more likely to support a European solution to the
refugee crisis, whereas the British press is more likely to promote a national solution. This hypothesis4 was tested using a logistic regression. The tests of the full models against constant only models were statistically significant, indicating that the predictors reliably predicted the likelihood of occurrence of the responsibility attribution framing (χ2= 71.473, p < .001 with df = 4; χ2= 56.384, p < .001 with df = 4; χ2= 78.826, p < .001 with df = 4). In the case of German press, the model explained 25.2% (Nagelkerke R2) of the variance in the responsibility framing of the E.U. and national government and correctly classified 78.2% of cases. The corresponding model about the British press explained 14% of the variance in the dependent variable and classified correctly 70% of the cases. Finally, when it comes to Greece, 20% of the variance was explained, while
4
This model was also ran using ideological background of the newspaper outlets as an independent variable. The analysis indicated that ideological background of the newspaper does not significantly predict the occurrence of responsibility framing (p>.05).
21
67% of cases were classified correctly. In particular, as presented in Table 2, results indicated that in the German press, the chance of appearance of the responsibility framing is higher when there is reference of the European Union compared to when the national government is mentioned. When it comes to Greece, the responsibility of dealing with the crisis also increases when there is reference to the European Union, confirming our expectations. Finally, in the British press, the responsibility frame is also more likely to occur in combination with a reference to the EU, compared to the British government. Therefore, our hypothesis appears to be partly supported. Indeed, in the Greek and German press attribution of responsibility for solving the refugee crisis is more likely to appear in references of the European Union, but the same results were witnessed for the British press.
Table 2.
Predictors of Responsibility Framing Occurrence
Model B SE P-value Odds ratio German EU 2.902 .384 .00 18.217 Nat. Government 1.294 .311 .00 3.646 British EU 1.834 .271 .00 6.206 Nat. Government 1.006 .256 .00 2.733 Greek EU 2.423 .348 .00 11.280 Nat. Government .078 .219 .72 1.081
The attribution of responsibility was further analysed with a cross- tabulation analysis, which indicated a significant relationship between the variables (χ2 <.001). The results of the crosstab, are analytically presented in Chart 1. This analysis indicated that the pattern of responsibility attribution appears to differ across the three countries.
22 Chart 1.
Particularly, attribution of responsibility to the European Union seems to be the highest in the German press (37.5%), followed by the Greek (34.7%) and lastly by the British press
(32.4%). When it comes to the national government, contrary to our expectations, the United Kingdom presents the lowest percentage (25%). However, the responsibility attribution to other European countries is more prominent in the British press (28.5%) with a percentage that is almost triple of the corresponding percentage of Greek press 5 and slightly more than double of the percentage in the German outlet. We can therefore conclude that in the UK press, attribution of responsibility to national government is prominent, but not when it comes to the British government. This finding is in line with the theory of the in- group, out- group
5
In the Greek press, the percentage of responsibility attribution to non- European countries was approximately 34%, which is quite high compared to German and British press. Particularly, the majority of the non- European references in terms of responsibility attribution concern Turkey and the agreement with the E.U. regarding control of the refugee influx. This percentage can be explained by the historical rivalry between Greece and Turkey which lurked in most of the articles about the refugee crisis in the Greek outlets.
37,50% 32,40% 34,70% 38,50% 25,00% 31,40% 12,50% 28,50% 9,60% 11,50% 14,10% 33,90% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% G E R M A N P R E S S B R I T I S H P R E S S G R E E K P R E S S
RESPONSIBILITY FOR SOLVING THE
CRISIS
Non- European Countries Other European countries
Responsibility for solving the refugee crisis on the national government Responsibility for solving the crisis on the European Union
23
theory as outlined in reference to European identities in the theoretical section. Nevertheless, it does not confirm our hypothesis.
H3: The evaluation of Greece will be overall more positive in German press in comparison to
the British press.
Before proceeding with the analysis, a test of normality was conducted, in order to ensure that the normal distribution criterion was met in all groups of the independent variable. The results of the Kolmogorov- Smirnov test, indicated that the normality hypothesis was rejected
(p<.05), thus, we resorted to a non- parametric test. To test whether the evaluation of Greece differed significantly in the German press compared to the British one, we employed two Mann- Whitney U tests, using the evaluation of political actors and Greek citizens as
dependent variables respectively. In both tests the country of origin of the outlets was used as independent variable. The results indicated a significant difference between the German and British press in the evaluation of political actors (one- sided p= .028). Indeed, the evaluation of political actors was slightly more neutral in the German compared to the British press. Difference in the evaluation of Greek citizens was marginally not significant, with the
evaluation being more positive in the British press. Therefore, the third hypothesis was partly supported. The evaluation of Greece was more positive in the German press, but only when it came to the political actors.
Table 3.
Comparison of the evaluation of Greece in German and British press
Dependent Variables Outlets Mean Standard Deviation Mann- Whitney U p- value Evaluation of German 00 .50 205.5 .028
24 politicians English -.35 .69 Evaluation of citizens German .42 .79 18.0 .055 English .80 .42
H4a: Compared to the Greek press, the responsibility framing of Greece will be more
prevalent in the German and British press.
H4b: The responsibility framing of Greece will be higher in the British compared to the
German press.
These hypotheses were tested using logistic regression analysis. The test of the full model against a constant only model was statistically significant, indicating that the independent variable reliably predicted the likelihood of occurrence of the responsibility attribution framing (χ2 = 164.787, p < .001 with df =2). The model explained 11% of the variance while 92% of the cases were correctly classified. The first model was run, using the Greek press as a reference group indicated that the likelihood of attributing responsibility to Greece was lower in the German (B= -2.039, Exp(B)=.130, p<.001) and British (B= -1.293, Exp(B)=.275, p<.001) compared to the Greek press. Therefore, hypothesis 4a is rejected. Contrary to our expectations, the responsibility framing of Greece was most prominent in the Greek press. In a further comparison between the British and German press, which is presented in Table 4, it was indicated that the appearance of Greece in terms of responsibility was more likely to occur in the British press (Exp(B) = 2.110, p<.001), when compared to the German press, where the likelihood of the responsibility framing of Greece occurring is the lowest. Hence, hypothesis 4b was supported as indeed the responsibility framing of Greece was higher in the British press, compared to the German one.
25 Table 4.
Predictors of Responsibility Framing of Greece Occurrence
B SE P-value Odds ratio
British Press .747 .232 .001 2.110
Constant -3.668 .188 .000 .026
1
In this model, German Press was used as a reference category.
H5: The responsibility framing of Greece will be slightly more prominent in the Greek
quality newspapers than in the Greek tabloid6.
To test this hypothesis, we used a logistic regression. Testing the full model against a constant model only, indicated significance confirming that the predictor reliably predicted the likelihood of occurrence of the responsibility attribution to Greece (χ2 = 15.916, p <.001 with df =2). The overall success of prediction reached 65.2%. As
outlined in Table 5, newspaper I Efimerida ton Syntakton was significantly less likely to suggest attribution of responsibility to Greece, compared to Proto Thema, indicating significance <.05. Kathimerini also appeared to be less likely to attribute responsibility to Greece. Nevertheless, this likelihood was marginally not significant. Furthermore, the results indicated that the responsibility framing of Greece is the most likely to occur in the Greek tabloid, compared to both right and left quality newspapers, which indicated odds ratios values of .512 and .235 respectively. As a result, our hypothesis that responsibility framing of Greece would be more prominent in the quality press is not supported.
Table 5.
6The same regression analysis was conducted using the British outlets. The results indicated significance of the
model as a whole (χ2 = 11.341, p = .03 with df =1) and the overall success of prediction was 92%. However, the
prediction of responsibility framing was significant only in the case of the Sunday Telegraph (p = .006, Exp(B) = .505). Therefore, applied in the British press, this hypothesis is partly supported with the responsibility framing of Greece being more prominent in the centre-right quality newspaper.
26 Occurrence of Responsibility Framing of Greece in Quality Newspapers and Tabloids
B SE
P-value
Odds ratio
Kathimerini -.670 .385 .081 .512
I Efimerida ton Syntakton -1.450 .375 .000 .235
Constant 1.045 .267 .000 2.842
1
In this model, the newspaper outlet Proto Thema was used as a reference category.
H6a: In comparison to the British and German press, the victim framing of Greece will be
more prominent in the Greek press.
H6b: The victim framing of Greece will be more prominent in the German press, compared to
the British one.
To examine these hypotheses, we used logistic regression. The first step before running the analysis was recoding the victim framing variable, selecting only the cases in which Greece was coded and reducing the answer options into two, by merging present and very present options. The new variable, victim framing of Greece was used as a dependent in this model. The dummy variables of the outlets’ countries of origin were used as independent. A test of the model against a constant only model, showed significance (χ2 = 13.656, p = .001 with df =2). About 7% of variance was explained by the model, while 62% of the cases were correctly classified. Running the model using Greek press as a reference category indicated that victim framing is less likely to occur in the German press (p= .013, Exp(B)=.249), but more likely to appear in the British press (Exp(B)= 1.781). Nevertheless, the difference when it came to the British press, was marginally significant (p=.051). Therefore, hypothesis 5a was partly confirmed as victim framing appears more often in the Greek press, only in comparison to the German press. Moreover, in a comparison using German press as a reference category
27
as presented in Table 6, it was found that victim framing is more likely to appear in the British press (p =.001, Exp(B) = 7.154). Therefore, hypothesis 5b was not supported.
Table 6.
Predictors of Victim Framing of Greece Occurrence
B SE P-value Odds ratio
British Press 1.968 .602 .001 7.154
Greek Press 1.391 .560 .013 4.018
Constant -1.792 .540 .001 .167
1
In this model the German press was used as a reference category.
Discussion & Conclusion
This study was an attempt to assess how are nation images, as well as the portrayal of
European Union influenced in the light of the current refugee crisis, focusing on the question of responsibility. We aimed our attention on the national press of Germany, Greece and the United Kingdom, knowing that these are three countries with an entirely different approach to the European Union and thus, would attribute responsibility differently amongst the countries/ actors involved in the crisis. For this purpose, we focused on theories regarding the
ethnocentricity of the press leading to in- group bias in reporting (Müller, 2013; Duckitt et al., 2005), as well as the process of framing in news reporting. When it comes to framing, we chose to direct our analysis towards the generic frame which is most commonly used in times of crisis, namely the responsibility attribution (Coombs and Holladay, 2004) and victim framing (Samaras, 2009).
The first finding is related to the depiction of the European Union in the sampled articles. Particularly, it was found, that in the German outlets, the European Union was presented as a unity more often compared to the British press, as hypothesised, but also
28
compared to the Greek press. Contrary to our expectations, the dominant assessment of the E.U. in the Greek press was that of a ‘fractured’ union where member- states serve their national interests, which are placed above the overall interest of the E.U. However, this finding was not entirely surprising. Its roots can be traced in the coverage of the financial crisis in the European media, as well as the way Greece’s partner countries handled the crisis. Moreover, at the dawn of the financial crisis, the first reaction of the European Union was characterised by inertia, which came with a massive price, as valuable time was lost and the crisis turned into a financial disaster (Kaitatzi-Whitlock, 2014). Accordingly, the coverage of Greece in the international press was dominated by an anti -Hellenic campaign (Kaitatzi-Whitlock, 2014). This experience can be held responsible for the lack of trust towards the European Union on Greece’s part and thus the portrayal of E.U. in terms of inequality, division and lack of solidarity.
The hypothesis regarding a European solution to the refugee crisis, only found partial support. This hypothesis was based on UK’s attitudes towards the European Union in general, its abstention from several European institutions and its lack of desire for further integration (Marcussen et al. 1999; Wimmel, 2009). Analysis indicated that promoting a European solution to the refugee crisis, was not only trait of the German and Greek press, but also the British one. However, taking a closer look to the further analysis, it is revealed that in the British press there also is much higher percentage of responsibility attribution to other
European countries, compared to the German and Greek press. To that extent, the results were in line with the theory. The British press was more likely to attribute responsibility for solving the crisis on a European level. Nevertheless, taking into consideration the high percentage of responsibility attribution to other European countries, we can conclude that in the UK press responsibility attribution to national governments is very visible, but not when it comes to the British government. Regarding the generalisation of this finding, it should be taken into
29
consideration, that the two categories were not mutually exclusive. The responsibility attribution was coded differently for each country mentioned in the article. Therefore, in several articles, responsibility was attributed to both the European Union and the national government.
Moving further to the evaluation of Greece, results indicated that the evaluation of Greece differed significantly between the German and British press. However, contrary to our expectations, the evaluation was more positive in the German press only when it came to political actors. The evaluation of citizens was more positive in the British outlets. Taking a closer look at the results, we can see that the evaluations of both government and citizens in the German press are closer to neutral. A possible explanation for this outcome could be that in the British press, the percentage of opinion articles (24%) was noticeably higher in
comparison with both the German (11%) and the Greek press (14%). According to McQuail (1992), the principle of objectivity requires a coherent distinction between the presentation of events and their commenting. And although objectivity is a normative standard, it is not entirely implemented. To the extent it is implemented, the reporting of hard news is directed by the dynamic of events and at times the sources, whereas opinion articles are led by the journalistic point of view (Samaras, Iordanidou & Dogani, 2016). Therefore, the different allocation of articles across varying types of publication, and in particular higher number of opinion pieces in the British outlets could have functioned as a magnifying glass when it comes to evaluative judgements, either positive or negative. Furthermore, when it comes to the evaluation of political actors, more negativity in the British press can be justified by UK’s stances towards the E.U. and by extension to other member states, as well as our findings regarding responsibility attribution on the Greek government, which are described in the following paragraphs.
30
According to our analysis, the attribution of responsibility to Greece was more
prominent in the British compared to the German press, confirming our expectations. Overall, our results indicate a more ‘favourable’ treatment of Greece in the German press, which is consistent with the theoretical background. However, when it comes to responsibility attribution to Greece being lower in the national press, our hypotheses were contradicted. In an attempt to reason our finding, a cross- tab analysis was ran using the Greek newspaper outlets and the responsibility framing of Greece. The results indicated that the majority of responsibility attributions were present in the right newspapers. Looking back into Hallin and Mancini’s classification of media systems, Greece is a typical example of the Mediterranean (Polarized Pluralist) model. One of the main characteristics of this model is political
parallelism, thus the existence of links between media outlets and certain political- ideological camps (Brüggerman, Engesser, Büchel, Humprecht & Castro, 2014). The associations with certain ideologies influence news reporting, bringing political bias in the equation. Therefore, we can assume that in this hypothesis, political parallelism overarched what was described as an in-group bias in the theoretical section. Repeating the analysis for within- country
differences in the occurrence of responsibility framing, it was revealed that attribution of responsibility to Greece is more likely to appear in the Greek tabloid, compared to both quality newspapers contradicting our expectations. However, when the same analysis was conducted in the British press, the results were in line with the Semetko and Valkenburg’s finding (2000), thus the responsibility frame was more widely used in the quality press. Taking these discoveries into consideration, as well as the right ideological background of Proto Thema and the leftish roots of the Greek government, political parallelism in the Greek media could be the cause also for this hypothesis’ disconfirmation.
Finally, our hypotheses regarding the victim framing of Greece in the German, British, as well as Greek press were not entirely supported. The victim framing was found more
31
present in the British press compared to both the Greek and German press. This finding could be explained with the same tools we used for the evaluation hypothesis, thus the existence of a bigger number of opinion pieces in the British press. As in opinion articles the journalistic point of view prevails (Samaras et al., 2016), we could argue that adopting a more humanistic angle on the issue was more or less subsequent.
One dubious point of this study is our choice to sample only weekend editions of newspapers. This sampling can be considered a limitation when it comes to generalisation of the results. However, as mentioned in the methodology section, this choice was guided by the fact that weekend editions tend to be longer, focusing on the important events of the whole week (Song & Chang, 2012) and providing more background analysis. To investigate the main objective of this study, we used several variables that required coding judgement and therefore needed more information in order to be coded with greater precision. This was offered by the weekend editions. Therefore, the sampling method constitutes at the same time a weakness and strength of this study. Another weakness that should be pointed out is the lack of certain variables from the codebook, which was noted during the process of coding, but time restrictions did not allow us any adjustments. One of these variables was whether or not there is a process of image building in the article. During the coding, we encountered some cases, where even though there was a negative evaluation of the country, there was no process of nation- image building. In these cases, the negativity was driven by the dynamic of the crisis developments rather than an attempt to construct a negative nation- image.
To conclude, the differences we found by comparing the coverage of the refugee crisis in the German, Greek and British press, were not the ones we expected. However, it can be argued that this study confirmed that media coverage in times of crisis is to a large extent influenced by national interests, especially when the crisis takes place in a foreign country. Past research has shown that nation- images projected by the media do not necessarily
32
coincide with the images projected by political actors, as news reporting is not a precise reflection of reality but a construction (Samaras, 2009). What this study showed is that the mediated images in crisis situations are defined by national goals. More interesting though, is that it expanded Wimmel’s findings (2009) by showing that attitudes towards the EU and its finality greatly affect reporting, not only in regards with policy debates, but also in regards with crises. These findings can constitute a useful tool that will help us predict and explain the coverage in any future European crisis. Last but not least, mediated images carry the capacity to radically change the existing image the audience holds of a certain country, cause partial alteration by adding new elements, or have no impact whatsoever when the incoming information has no influence on a strong pre-existing image. This question remains to be answered in future research.
References
Aarøe, L. (2011). Investigating Frame Strength: The Case of Episodic and Thematic Frames. Political Communication, 28 (2), 207-226.
Boulding, K. E. (1956). The Image: Knowledge in Life and Society. Ann Arbor, MI: University of Michigan Press.
Boulding, K. E. (1959). National Images and International Systems. The Journal of Conflict Resolution, 3(2), 120-131.
Brewer, P.R., Graf, J., and Willnat, L. (2003). Priming or framing media influence on attitudes toward foreign countries. Gazette, 65(6), 493-508.
Brüggerman, M., Engesser, S., Büchel, F., Humprecht, E., & Castro, L. (2014). Hallin and Mancini Revisited: Four Empirical Types of Western Media Systems. Journal of Communication, 64, 1037-1065.
33
Cappella, J.N., & Jamieson, K.H. (1997). Spiral of Cynicism. Oxford: Oxford University Press.
Coombs, W.T., & Holladay, S.J. (2004). Reasoned action in crisis communication: An attribution theory-based approach to crisis management. Responding to crisis: A rhetorical approach to crisis communication, 95-115.
Cottle, S. (2009). Global Crisis Reporting: Journalism in the Global Age. Berkshire: Open University Press.
David, C.C., Atun, J.M., Fille, E., & Monterola, C. (2011). Finding frames: Comparing two methods of frame analysis. Communication Methods and Measures, 5(4), 329-351.
De Vreese, C.H. (2005). News framing theory and typology. Information Design Journal and Document Design, 13 (1), 51-62.
De Vreese, C.H., Boomgaarden, C.G., & Semetko, H.A. (2011). (In)direct Framing Effects: The Effects of News Media Framing on Public Support for Turkish Membership in the European Union. Communication Research, 38(2), 179-205.
Dell’Orto, G., Dong, D., Moore, J., & Scheeweis, A. (2004). The Impact of Framing on Perception of Foreign Countries. Ecquid Novi, 24(3), 294-312.
Duckitt, J., Gallaghan, J., & Wagner, C. (2005). Group Identification and Outgroup Attitudes in Four South African Ethnic Groups: A Multidimensional Approach. Personality and Social Psychology Bulletin, 31(5), 633-646.
Entman, R. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51-58.
Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American journal of sociology, 1-37.
34
Gurevitch, M., Levy, M., & Roeh, I. (1991). The Global Newsroom: Convergences and Diversities in the Globalisation of Television News. Communication and citizenship, 195-216.
Hacker, K.L. (1995). Candidate Images in Presidential Elections. Westport, Connecticut: Preager.
Harcup, T., & O’Neill, D. (2001). What Is News? Galtung and Ruge revisited. Journalism Studies, 2(2), 261-280.
Handley, R.L. (2009). The Conflicting Israeli-Terrorist Image: Managing the Israeli- Palestinian Narrative in the New York Times and Washington Post. Journalism Practice, 3(3), 251-267.
Handley, R. L. & Ismail, A. (2010). Territory Under Siege: ‘Their’ News, ‘Our’ News and ‘Ours Both’ News of the 2008 Gaza Crisis. Media, War & Conflict, 3(3), 279- 297. Hayes, A. F., & Krippendorff, K. (2007). Answering the call for a standard reliability measure
for coding data. Communication Methods and Measures, 1, 77- 89.
Herrman, R.K. (2003). Image Theory and Strategic Interaction in International Relations. In D.O. Sears, H. Leonie & R. Jervis (Ed.), Oxford Handbook of Political Psychology (pp. 285-314). Oxford and New York: Oxford University Press.
Iyengar, S. (1991). Is anyone responsible? How Television Frames Political Issues, Chicago: University of Chicago Press.
Kahneman, D., & Trevsky, A. (1984). Choice, values and frames. American Psychologist, 39, 341-350.
Kaitatzi-Whitlock, S. (2014). Greece, The Eurozone Crisis and the Media: The Solution is the Problem, Javnost - The Public, 21 (4), 25-45.
Kunczik, M. (1997). Images of Nations and International Public Relations. Mahwah, NJ: Lawrence Erblaum Associates.
35
Kunczik, M. (2000). Globalisation: News media, images of nations and the flow of international capital with special reference to the role of rating agencies. Paper presented at the IAMCR Conference, Singapore, July 17-July 20, 2000.
Lacy, S., Robinson, K., & Riffe, D. (1995). Sample Size in Content Analysis of Weekly Newspapers. Journalism & Mass Communication Quarterly, 72(2), 336-345.
Lamont, M., & Molnár, V. (2002). The Study of Boundaries in the Social Sciences. Annual review of sociology, 167-195.
Manheim, J.B, & Albritton, R.B. (1984). Changing National Images: International Public Relations and Media Agenda Setting. American Political Science Review, 78, 641-657. Marcussen, M., Risse, T., Engelmann- Martin, D., Knopf, H.J., & Roscher, K. (1999).
Constructing Europe? The evolution of French, British and German nation state identities. Journal of European Public Policy, 6 (4), 614-633.
Matthes, J. (2009). What’s in a Frame? A Content Analysis of Media Framing Studies in the World’s Leading Communication Journals, 1990-2005. Journalism & Mass
Communication Quarterly, 86(2), 349-367.
Müller, P. (2013). National identity building through patterns of an international third- person perception in news coverage. The International Communication Gazette, 75(8),732-749. Neuman, W.R., Just, M., & Crigler, A. (1992). Common knowledge: News and the
construction of social meaning. Chicago: University of Chicago Press.
Nossek, H. (2004). Our News and Their News: The Role of National Identity in the Coverage of Foreign News. Journalism, 5(3), 343-368.
O' Regan, M. (2007). Explaining Media Frames of Contested Foreign Conflicts: Irish
National ‘Opinion Leader’ Newspapers’ Frames of the Israeli- Palestinian Conflict (July 2000 to July 2004). Journal of the MeCCSA Postgraduate Network, 1(2).
36 Journalism Quarterly, 64(2), 416-433.
Samaras, A.N., Iordanidou, S., & Dogani, M. (2016). The Dialectic of International Relations and Political Communication: The Impact of the Financial Crisis Upon the Nation Image of Germany; A Comparative Analysis in Greek and UK Press. In O.G. Luengo (Ed.), Political Communication in Times of Crisis (pp. 121- 144). Berlin: Logos Verlag Berlin.
Semetko, H.A., & Valkenburg, P.M. (2000). Framing European Politics: A Content Analysis of Press and Television News. Journal of Communication, 50(2), 93-109.
Seon- Kyoung, A., & Gower, K.K. (2009). How do the news media frame crises? A content analysis of crisis news coverage. Public Relations Review, 35(2), 107-112.
Song, Y., & Chang, T.K. (2012). SELECTING DAILY NEWSPAPERS FOR CONTENT ANALYSIS IN CHINA. Journalism Studies, 13(3), 356-369.
Van Cauwenberge, A., Gelders, D., & Joris, W. (2009). Covering the European Union. Javnost - The Public, 16(4), 41-54.
Wimmel, A. (2009). Beyond the Bosphorus? Comparing public discourses on Turkey’s EU application in the German, French and British quality press. Journal of Language and Politics, 8(2), 223-243. Σαμαράς, Αθ. Ν. (2007). Εθνοκεντρισμός και Ευρωκεντρισμός στις Αναπαραστάσεις της Ευρωπαϊκής Ένωσης: Ζητήματα Πολιτικής Επικοινωνίας. Στο Παπαθανασόπουλου Σ., (Επιμ.), Ανάλεκτα, (σελ. 234- 269). Αθήνα: Ινστιτούτο Οπτικοακουστικών Μέσων. Σαμαράς, Αθ. Ν. (2009). Μέσα Μαζικής Επικοινωνίας και η Εικόνα των Κρατών: Η Περίπτωση των ΗΠΑ στη Μετά την 11/9 Εποχή. Στο Γιαλλουρίδη, Χ. και Λαγκίδη, Αφ. (Επιμ.), Ο Οργανισμός Ηνωμένων Εθνών από την Αμφικτυονία έως τον 21ο Αιώνα, Αθήνα: Ευρωπαϊκό Πολιτιστικό Κέντρο Δελφών – Εκδόσεις Λιβάνη.
37
Σερέτη, Ν. (2004). Μια Ποσοτική Ανάλυση της Εθνικής και Ευρωπαϊκής ταυτότητας στην Ελλάδα. Στο Κασιμάτη, Κ. (Επιμ.), Εθνική και Ευρωπαϊκή Ταυτότητα, Συγκλίσεις και Αποκλίσεις (σελ. 101-126). Αθήνα: Gutenberg.
