A Crisis Far and Near
The Scandinavian Media Coverage of the Refugee Crisis in 2015
Ida Eri Sørbye 11081856 Master’s Thesis
Graduate School of Communication Master’s programme Communication Science
Supervisor Andreas R. T. Schuck May 27th 2016
Abstract
This study investigates the use of generic and issue specific frames in the coverage of the refugee crisis in Sweden, Norway and Denmark over time and in different news outlets. Using content analysis the study examines one broadsheet and one tabloid newspaper from each of the three countries from September to November in 2015. The study finds that the generic frames human interest, conflict, attribution of responsibility, and morality, are the most commonly used frames in the coverage. Four issue specific frames are detected: the threat frame, refugee frame, the deportation frame, and the asylum policies frame. There are distinct differences in the use of frames between the countries, and the use of frames changes throughout the time period. News outlets in Denmark overall focused less on the
humanitarian aspect of the crisis and more on restriction of asylum policies. The news coverage in Norway and Sweden changes from September to November. The reporting changes from a humanitarian focus of the crisis towards focusing on restrictions and asylum seekers as a threat to society. The findings are discussed in light of the importance of
comparative research on the use of generic and issue specific frames in the news in
Scandinavia and the possible societal consequences of negative immigration coverage in the media.
A Crisis Far and Near
The Scandinavian Media Coverage of the Refugee Crisis in 2015
In the beginning of 2015 the European media started giving attention to the rising refugee crisis. For a while it was a distant crisis on the Greek islands or on the roads of eastern Europe. The refugees´ ultimate goal, however, was northern Europe. The Scandinavian countries started to feel the weight of the numbers in the fall. The distant
refugee crisis moved on to become a crisis in the national immigration system, and at the peak 10 000 refugees arrived in Sweden during one week in November (Första ganger över, 2015). Crises become known through the media (Pantti et al., 2012, p. 33), and the refugee crisis has indeed been high on the media agenda all over Europe in 2015. It hence provides a good topic for comparison of media coverage across countries.
Journalists make choices on how to report on the refugee crisis, on what to emphasize and what to leave out; they use different types of frames. Frames can be measured as
characteristics of a news text (De Vreese, 2005, p. 54), and hence be used as effective tools for comparison across countries. Frames used in the media have been found to have effect on public opinion (see e.g. Lecheler et al., 2015) and political attention (see e.g. Vliegenthart & Roggeband, 2007; Walgrave & Van Aelst, 2006). The framing of the refugee crisis in Scandinavia can therefore have implications for how the local people view the immigrants and how the political debate unfolds on the issue.
However, comparative research on news frames in Scandinavia is scarce.
Thorbjørnsrud (2013), and Strömbäck and Aalberg (2008) have studied the election coverage on Scandinavian public broadcasters and election coverage in Norwegian and Swedish newspapers respectively. In an overview of the media research on immigration in the Nordic countries Horsti (2008, p. 279) notes that there in general is “a conspicuous lack of
comparative media research on Scandinavia generally and on immigration in Scandinavian news specifically, the refugee crises provides a relevant and important backdrop for a
comparative study on the frames used in the Scandinavian media. Following this the research question of the study is:
RQ: What characterizes the newspaper coverage of the refugee crisis in Norway, Sweden and
Denmark in 2015?
This study has two aims. Firstly, to test the applicability of generic frames and issue specific frames on the topic of immigration in the Scandinavian media. Secondly, the aim is to look for similarities and differences in how the frames are applied across the countries in Scandinavia, throughout the time period of the study and between the newspaper outlets, such as broadsheets and tabloids. The study will be conducted using content analysis to examine the highest circulating broadsheet and tabloid newspaper in Norway, Sweden and Denmark. The months of September, October and November in 2015 will be examined, as these were the months with the heaviest coverage of the refugee crisis.
Theoretical framework Framing Immigration in Europe
Frames make the world´s events accessible to the public in the limited format of a news article; frames are “principles of organization” (Goffman, 1974, p. 10). Frames present issues in alternative ways by focusing on different aspects of the issue (De Vreese, 2005, p. 53). Vliegenthart and Roggeband (2007, p. 296) states that even though no consensus has been reached on the conceptualization of framing, most of the definitions have commonalities with Entman´s concept of framing: ‘‘To frame is to select some aspects of a perceived reality and
make them more salient in a communicating context, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment
choices of how to frame every issue in every news story. Some frames have been found to be present across time, across topics and countries (De Vreese, 2005, p. 54), these are generic frames. Because of the broad scope of the frames it has been possible to use them to compare coverage across cultures (De Vreese et al., 2001, pp. 108-109).
Issue specific frames differ from generic frames on their high level of “issue sensitivity” (De Vreese, 2005, p. 55), which allows for detailed analysis of the issue in question (De Vreese et al., 2001, 108). Although this issue sensitivity makes it difficult to generalize across topics, it does allow for a more in-depth analysis of a specific topic. By focusing on issue specific frames it might be possible to “capture specific aspects of selection, organization, and elaboration” (Ibid.) on an issue, which will not be captured by the generic frames. A combination of generic and issue specific frames will hence allow for both generalizations and in-depth knowledge of the framing of the topic. Matthes and Kohring (2008, pp. 262-263) criticize deductive frames for being inflexible when it comes to discovering emerging frames in the text. By combining the issue specific and the generic frames, this study seeks to alleviate some of this limitation. The following section will give a further account of the generic frames before moving on to the issue specific frames on immigration.
It is a need for a “reliable set of content analytic indicators” (Semetko & Valkenburg, 2000, p. 94), or standardized instructions and variables (Aalberg et al., 2011, p. 12) to be able to compare generic frames across countries, outlets, and the change of frame use over time. Through an extensive literature review, Semetko & Valkenburg (2000) identified five frames in the literature, found to be commonly present in the news. The conflict frame, attribution of responsibility frame, human-interest frame, morality frame, and economic consequences frame, accounted for the majority of frames that had been found in previous studies (for the full overview see Semetko &Valkenburg, 2000, pp. 95-96). The conflict frame focuses on
disagreements between groups, individuals and countries. The attribution of responsibility frame places the responsibility of an event at the level of the government, or an individual or group. The human-interest frame brings the individual angle in the story to the forefront. The morality frame raises questions about an issue by focusing on religion or morality. Lastly, the economic consequences frame focuses on the economic costs, gain or losses in the story.
In Europe, the use of generic frames has been compared across countries and media outlets on different topics such as the framing of the launch of the euro between four EU countries (De Vreese et al., 2001), the European parliamentary elections in all the 27 member states (Schuck et al., 2013b) and election coverage in Sweden and Norway (Strömbäck & Aalberg, 2008). There is however, a lack of cross-country comparative research on the use of generic frames on the topic of immigration.
Previous studies have found outlet to be one of the factors affecting the framing of the news story. Whether an outlet is broadsheet or tabloid has been shown to produce different styles and quality of reporting (De Vreese et al., 2006, p. 483). When measuring the use of frames, tabloids have been found to produce more stories with the human-interest frame, whereas broadsheets have a stronger focus on the attribution of responsibility frame (Semetko & Valkenburg, 2000, pp. 103-104). Schuck et al. (2013b, p. 19) found that the conflict frame was more prominent in broadsheet than in tabloid newspapers. In the recent study on the European refugee crisis, Berry et al. (2015, p. 236) found that refugee sources were used more extensively to make a narrative in the news story in the Swedish tabloid Aftonbladet, than in the broadsheet newspapers. Following this, the expectations for the use of generic frames in the different outlets are:
H1: Tabloids will use the human-interest frame more extensively in their reporting than the
H2: Broadsheets will use the conflict frame more extensively in their reporting than the
tabloid newspapers.
H3: Broadsheets will use the responsibility frame more extensively in their reporting than the
tabloid newspapers.
In Europe some studies have looked at issue specific framing of immigration in the news. There is however, a lack of comparative studies looking at how immigration is framed across countries, and how the use of issue specific framing differs between outlets and time. Horsti (2007, p. 147) notes that the asylum issue has been high up on the political agenda in Europe for a long time, but specific studies on the coverage of the topic is not as common as might be assumed. Her inductive study focusing on news agency stories found that the frames of illegality and control are used when covering asylum seekers. Hence other possible angles such as focusing on a human rights frame are never used. The topic of asylum seekers only got into the news if the story involved state actors or illegality (Ibid., p. 157).
Roggeband and Vliegenthart (2007) developed five issue specific frames by inductively coding policy documents on migration in Netherlands. They found the
multicultural frame, emancipation frame, restriction frame, victimization and Islam-as-threat frame to be represented in the documents. When analyzing the use of these frames in the media, they found that Islam-as-threat frame was the dominant frame throughout the years in the study, 1995-2004, except between 1998-2000. The multicultural frame was found to be the second most common frame. In another study (Vliegenthart & Roggeband, 2007) they looked at the relationship between the same frames and the use in the parliament documents and newspaper articles. Long-term influences of frame use were found to go both ways. Effects were also found from external events, having a larger effect on the media frames than the parliament frames. Lecheler et al. (2015) looked at the role of emotions, and how they can explain the effect of the emancipation, multicultural, assimilation and victimization frame on
people´s opinion on immigration in the Netherlands. The study found that all the frames led to emotional reactions and some functioned as mediators for the effect on frames on opinions on immigration.
Van Gorp (2005) studied the use of frames in the Belgian media coverage of the immigration and asylum issue. He looked at newspapers framing asylum seekers as either intruders or victims, and to what extent the frames where used differently depending on the outlet and the ideological background of the newspaper. Asylum seekers can be portrayed as the typical victim of an unbearable situation, having the need for protection in a new country. On the other hand, they can also be portrayed as threatening strangers, as “suspects or even as the associates of human traffickers” (Ibid., p. 489), not in need for a safe new home or of the empathy of the reader. The two frames are likely to suggest different solutions to the issue; granting people asylum, or discouraging the people to come into the country. The study found that the presence of one of the frames was strongly correlated with a low presence of the other. It also found support for the intruder frame being used more in the tabloid newspapers than in broadsheets (Ibid., p. 499). Based on this, we form the hypothesis:
H4: Tabloids use the intruder frame more than the broadsheet newspapers.
Communication is a dynamic process, and frames develop in the interplay between journalists and other stakeholders such as politicians, interest groups and public opinion (De Vreese, 2005, p. 52; Van Gorp, 2005, p. 490). In their study of the early stages of the refugee crisis in Europe, Berry et al. (2015) found that there were shifts in the coverage of the refugee crisis from the first period of study, 2014 to early 2015, to the second period in April 2015. The periods differed in the explanations for the conflict and which solutions that were visible in the reporting (Ibid., p. 11). As the refugee crisis developed further in the year of 2015, it moved from being a distant crisis in the Mediterranean to a crisis at the Scandinavian doorstep. The events and developments of the refugee crisis are likely to have changed the
framing of the news, as framing is “vulnerable to specific events” (Roggeband &
Vliegenthart, 2007 p. 532). Brosius and Eps (1995, p. 408) found that “exceptional events” could lower the threshold for reporting on cases similar to the event, and have lasting effects on the coverage of an issue. Frames can also themselves cause a “media hype”: Van Gorp (2005, p. 502) found that during the Christmas week the victim frame was applied more often and the newspapers sought out stories that would fit the frame. However, as this is a relatively unexplored issue in Scandinavia, the study will abstain from hypothesizing about the direction of the change over time, and rather pose a research question:
RQ2: Will the use of frames in the media change during the time period of the research? Comparing Scandinavian media
The three Scandinavian countries Sweden, Norway and Denmark are very similar in many regards, such as language, culture and history. The countries are seen as belonging to the same model, the democratic corporatist model, in Hallin and Mancini´s (2004) seminal work on media systems, and also in the same Northern type in the further development of the system by Brüggemann et al. (2014). Strömbäck et al. (2008, p. 268) warns taking for granted that the Nordic countries are similar in all regards. However, they find that the Nordic
countries, which include Finland and Iceland, share several common traits among each other and with the Democratic corporatist model. This includes a highly developed newspaper marked, formerly a high degree of political parallelism, and specifically for the Scandinavian countries; high degree of journalistic professionalism and state intervention in the
development of radio and television. This does not mean however, that it is not important to look for both similarities and differences between countries that from the outside seem to be the same. Strömbäck et al. (Ibid., p. 272) calls for more comparative research to gain a better understanding of the Nordic system. Horsti (2008b, p. 286) states that comparative research between the Nordic countries is scarce. This research scarcity is surprising, she notes, as the
people in the region can understand each other’s languages and research therefore easily can be conducted across borders and languages.
Lund (2007, p. 121) states that it is important to pay attention to divergence, as not to fall into unwanted generalizations about the Scandinavian media. According to De Vreese et al. (2001, p. 119), a study comparing countries can “render the invisible visible, especially with respect to the extent to which factors internal and external to journalism influence the emergence of frames in the news”. Strömbäck et al. (2008, p. 12) also notes; “differences between the Nordic countries might be significant in a Nordic perspective while
simultaneously being insignificant from a global or an international perspective”.
Comparative research can also help prevent “naïve universalism” and the assumption that findings from research in one Scandinavian or other Western European country automatically apply to other countries (Blumler & Gurevitch, 1995, p. 75). Furthermore, they argue, it is only by comparing countries the relative features of a country can be understood and placed in a broader context (Ibid., p. 76).
With a relatively uniform population and culture, the media landscape in Scandinavia tends to conform to the same editorial ideas within each country (Lund, 2007, p. 124). This aligns with Berry et al. (2015, p. 10), they find that the press is more different between countries than within a country when reporting on the refugee crisis. The media in Sweden is found to have some of the most homogenous coverage in the study, which also includes the countries Italy, UK, Germany and Spain. The newspapers tended to “use the same language, report on the same themes and feature the same explanations and responses” (Ibid.). Van Gorp´s study (2005, p. 499) finds that popular and broadsheet newspapers within each region in Belgium reports on issues in the same manner, and use each other as a source in the news stories. Therefore, we hypothesize that:
H5: Newspapers within a country will report more similarly on issues than newspapers
between countries.
Although it is not clear who the most powerful partner in the journalist and politician couple is, a reciprocal relationship does exist between media and politicians (see e.g. Van Aelst & Vliegenthart, 2014). As noted earlier a long-term influence has been found in the use of frames between the parliament documents and the media on the topic of immigration (Vliegenthart & Roggeband, 2007). Knowing that politicians receive more attention the more powerful they are (Hopman et al., 2011, p. 276) it is clear that it is important to take the country specific political contexts into consideration when comparing countries, where the policy discussions about immigration take place and from which the media report.
Green-Pedersen and Krogstrup (2008, p. 615) compared immigration as a political issue in Denmark and Sweden over a period of 20 years. They found that much more attention was given to the issue in Denmark, even though Sweden has a share of immigrants that is substantially higher. They explain this by pointing to the united mainstream conservative parties in Denmark, which were able to make immigration a hot topic across the political spectrum. This subsequently led to restrictions on immigration and asylum policies. Denmark has traditionally had a lower support for the social democratic party than especially Sweden, but also Norway, and a high level of issue voting (Esmark & Ørsten, 2008, p. 32).
In Sweden the main conservative party has generally been weaker and has not tried to politicize the issue of migration. The country had its first non-socialist majority government formed after the 2006 elections, which was seen as a result of the conservative party moving towards the center of the political spectrum (Strömbäck & Nord, 2008, p. 107). The topic of immigration has traditionally been consensual and the issue has not received a lot of attention (Green-Pedersen & Krogstrup, 2008 p. 626).
Denmark ranks higher than the other Scandinavian countries when measuring racist attitudes in opinion surveys from Eurobarometer, with Norway in the middle and Sweden having the least racist attitudes (Horsti, 2008b, p. 277; Horsti, 2007, p. 146). In the spring of 2015, 57% of Danes were negative to immigration from outside the EU, compared to 60% of Swedes seeing immigration as something positive (Eurobarometer 83). Horsti (2008a) compared the Swedish and the Finish coverage of African boat refugees arriving in Spain in 2006, and observed that Sweden framed the issue as a humanitarian crisis, whereas Finland focused on the issue as a border control crisis. Berry et al. (2015, p. 11) also find Sweden to be the most positive towards refugees and the country has few examples of focusing on the refugees as a threat in the study. Based on the above, we will hypothesize that:
H6: Danish media will use the intruder frame more than the Swedish media. Method
To study the coverage of the refugee crises in Scandinavia, a deductive quantitative content analysis was carried out on newspapers in Sweden, Norway and Denmark. According to Reinard (2006, p. 7) the method of content analysis allows for analyzing written text production systematically. When collecting big amounts of data on complex phenomena, quantitative methods makes it possible for the researcher to examine this data efficiently (Ibid.). The deductive approach hence makes it possible to find differences between media outlets and across countries and time. The refugee crisis was covered extensively in all of the three countries in the fall of 2015, and gives an opportunity to study how the generic and issue specific frames are applied to the topic of immigration, and to study how the use of frames differ between the countries.
Sampling
One broadsheet and one tabloid were included from each country, Sweden, Denmark and Norway, as the study aims to look at differences of coverage between countries, but also
between outlets within countries. The newspapers in the study are the highest circulating national broadsheet and tabloid newspaper in each of the three countries1. In Denmark this includes the broadsheet Politiken and the tabloid Ekstra Bladet, in Sweden Dagens Nyheter and the tabloid Aftonbladet, and in Norway Aftenposten and the tabloid Dagbladet. Two newspapers from each country cannot be said to represent the whole media landscape in Scandinavia. However, as Lund (2007, p. 124) states, the media within each country tend towards conformity. The newspapers in the study, being the most widely read newspapers, will therefore be assumed to represent the general lines of the national media landscape. Time frame
The study aimed to look at changes in the coverage over time. The focus of the study is the three months September, October and November 2015. These months were chosen after gathering all the news articles covering the refugee crisis in the last six months of 2015. The three months with the highest number of news articles was thereafter chosen for the study. In the last six months of 2015 UNHCR registered a record numbers of refugees reaching the Greek islands, peaking in October (“Refugees/Migrants Emergency,” 2016). This period also saw the peak of the refugee influx in the Scandinavian countries, reaching up to 10 000 per week in Sweden in November (“Första ganger över,” 2015).
Unit of data collection
The unit of data collection in the study is the individual news item. The articles from Norway and Sweden were collected through the Scandinavian media database Retriever. The articles from the Danish broadsheet newspaper Politiken were collected through the database
Factiva. Finally, the articles from the Danish tabloid Extra Bladet were collected through the
Danish database Infomedia. Different databases were used due to the accessibility of the newspaper in question. The search for articles was conducted by limiting the search to the last six months of 2015 and using the search string “refugee OR refugee crisis OR refugees OR
migrants OR the refugee crisis” in the local language2.
All articles were selected manually. Only news articles were chosen, not editorial pieces or opinion pieces by external voices, as the aim of the study is to look at frames in news articles. The articles had to contain more than 100 words to be chosen. In addition to containing at least one of the search words, the topic of the story had to relate to the refugee crisis. The Swedish debate about Roma people begging on the street, who they call EU-migrants, were excluded from the sample. The chosen articles were collected in one PDF-file per newspaper, arranged by date, starting with August ending with December. All of the articles per month were counted and September, October and November chosen for further analysis, as described above. To make sure of the same proportion of articles within the same time periods in each country, the study used stratified random sampling, randomly selecting 20 articles per month for each newspaper (Reinard, 2006, p. 29). A random numbers
generator online was used to generate the numbers. 20 articles were chosen because this was the maximum number of articles in Dagbladet in September, which was the lowest number of articles for a newspaper during the three months. Hence, the total number of articles in the study was 360.
Frames
To analyze the stories and find out whether the generic codes appear in the news articles about the refugee crisis, the questions from the study of Semetko and Valkenburg (2000, p.100) were replicated. Babbie (Ibid., p. 147) suggests using established measures to ensure reliability when conducting studies. 19 of the 20 questions from the study were used, the exception being the question “Does the story contain visual information that might generate feelings of outrage, empathy-caring, sympathy or compassion?” from the human-interest frame as the study wanted to focus on the textual elements of the news story. The questions represented the five generic frames with three to five questions per category.
For the attribution of responsibility frame one of the questions were: “Does the story suggest that an individual or group is responsible for the issue or problem?” The questions for the human-interest frame focused on the people in the story, for example: “Does the story employ adjectives or personal vignettes that generate feelings of outrage, empathy-caring, sympathy, or compassion?” The conflict frame has questions such as: “Does the story refer to two sides or more of the issue?” The morality frame focused around questions such as: “Does the story make reference to morality, God or other religious tenets?” Lastly the economic consequences frame: ”Is there a mention of the costs/degree of expense involved?”
For the issue specific frames on immigration, the study of Van Gorp (2005) was used to develop the questions for the coding of the news articles. By investigating texts that used frames about immigrants for rhetorical reasons, such as political parties, and
non-governmental organizations working for refugees, the study built a matrix representing the framing and reasoning devices used for the intruder frame and the victim frame (Ibid., p. 490). The matrix builds on Entman´s (1993, p. 53) definition of the elements in a frame, namely the “problem definition, causal interpretation, moral evaluation, and/or treatment recommendation”. In Van Gorp´s study these elements are labelled the problem definition, problem source, responsibility, policy solution, and moral basis.
Each of the five elements described above have reasoning devices which in this present study are translated into questions to measure the intruder frame and the victim frame in the Scandinavian coverage. This resulted in five questions constituting the intruder frame and five for the victim frame. The problem definition in the victim frame for example, is turned into the question: “Does the article discuss how the country will receive refugees, which are victims of persecution and forced to leave their country? Where as for the intruder frame the problem definition is turned into: “Does the article discuss how to stop people who are a threat to `our` culture and social services from coming into the country?” See Appendix
A for the codebook and coding instructions. Procedure
Based on the operationalization of the concepts, a codebook with the questions for the generic and the issue specific frames was developed and pre-tested on a random sample of 20 articles before the actual coding was conducted. Each text was coded for all the questions in the codebook, using binary variables registering the presence (1) or the absence (0) of the element. One main coder coded all the news articles in the sample. The coder is fluent in the three Scandinavian languages. In addition to the main coder, one other coder was trained and supervised with the codebook. The second coder is fluent in one of the languages, and has good knowledge of the two other Scandinavian languages. The second coder coded a random selection of articles, constituting of 10% of the full sample, to ensure an acceptable level of inter-coder reliability.
The inter-coder reliability showed to be satisfactory, with all variables having a Krippendorff’s alpha value above 0.63. Krippendorff (2004, p. 241) suggested a lower limit of .667 to document sufficient reliability. However, “in the absence of any formal rule about acceptability levels”, Van Spanje and De Vreese (2014, p. 332) find a Krippendorff’s alpha value above 0.6 to be sufficient. Other studies have also stressed that Krippendorff’s alpha shows to provide especially conservative reliability estimates for variables with a skewed distribution (see Schuck et al., 2016, p. 191; Schuck et al., 2013a, p. 304). As some of the reliability estimates in the current study are below the suggested lower limit3, the findings should be interpreted with the appropriate caution.
Results Identifying news frames in media content
In the following section the paper will present the results of the analysis of the use of generic and issue specific frames, and how they differ across countries, time and outlets. First,
a principal component analysis with varimax rotation was conducted on the 30 items to see if they reflected the predefined generic and issue specific frames as they had been
conceptualized based on theoretical grounds (see discussion in the previous section) (for the same approach, see e.g., Semetko & Valkenburg, 2000). The analysis yielded nine factors with Eigenvalues greater than one, explaining 64.4% of the total variance. With a KMO level of 0.75 and a Bartlett´s test significant at p<.01 the items are considered appropriately related for factor analysis. All of the individual items had factor loading higher than .53, and where included in the scales constituting the frames, which were then used for further analysis.
As Appendix B shows, the items formed nine factors, i.e. frames, in the rotated component matrix. The two items focusing on economic gains or losses were collapsed into one item, as in the original study of Semetko and Valkenburg (2000). Two items were removed from the analysis as they made up two individual components in the matrix and showed to be unrelated to the other items: “Does the story refer to winner and losers” and “Does the article suggest that democratic countries bear the responsibility for the refugees”. Conceptually, this decision can be justified as it can be argued that a single indicator is not enough to constitute a frame. Theoretically, a frame represents a coherent construction of an issue for an audience and thus more than just a single or isolated perspective. The items from the original generic frames, except for two items in the attribution of responsibility frame formed the expected frames of the attribution of responsibility frame, human-interest frame, morality frame, conflict frame, and economic consequences frame.
The items developed from the issue specific frames in Van Gorp´s (2005) study on immigration split into four distinct patterns, i.e. frames, in the rotated component matrix: the threat frame, the deportation frame, the refugee frame and the asylum policy frame. The item focusing on an individual or group being responsible for the problem from the attribution of responsibility frame make up the new threat frame together with two of the items from the
previous intruder frame, namely the item focusing on how to stop people who are a threat to ‘our’ culture from coming into the country and the item focusing on the importance of protecting the interests of the native people. Likewise the item focusing on the government being responsible for the issue from the attribution of responsibility frame formed a new refugee frame together with the items from the previous victim frame focusing on how to receive refugees and refugees having a legitimate reason to come to the country.
The frames are weakly to moderately correlated (r=.001 to r=.294). The human-interest frame and the deportation frame show the highest negative correlation (r=.294, p<.01). The asylum policy frame and the deportation frame have the highest positive correlation (r=.275, p<.01).
Framing the refugee crisis
The scores of the items that constitute each of the nine frames where summed up and divided by the number of items per frame to get the mean presence score of each frame (see e.g., Semetko & Valkenburg, 2000; Schuck & de Vreese, 2006). This produced a scale indicating the presence of each frame ranging from 0 (not present) to 1 (fully present). See Figure 1 for an overview. Overall the generic human-interest frame is the most commonly used frame in the study, (M=0.33, SD=0.44). The three generic frames conflict (M=0.32,
SD=0.36), attribution of responsibility (M=0.29, SD=0.32) and morality (M= 0.23, SD=0.26)
follow as the most commonly used frames. Thereafter follows the issue specific refugee frame (M=0.18, SD=0.26), the deportation frame (M=0.15, SD=0.28) and the threat frame (M=0.11, SD=0.22). The economic consequences frame (M=0.11, SD=0.25) and the asylum policy frame (M=0.05, SD=0.17) are the least often used frames in the study.
<Figure 1 about here> News outlets
Looking broadly at the type of news outlet, there are no big differences between tabloids and broadsheets in their use of frames, an exception being the conflict frame. An
independent-samples t-test was run to compare the use of frames between tabloids and broadsheets. Broadsheets used the conflict frame significantly more (M=0.37, SD=0.37) than tabloids (M=0.27, SD=0.33) (t(353.4)=2.58, p<0.05). This lends support to the H2 that tabloids will use the conflict frame more than the tabloids. It does not give support to the H1 that the tabloids will use the human-interest frame more extensively than the broadsheet, nor the H3 that broadsheets will use the responsibility frame more extensively. H4 is also not supported, assuming that tabloids will use the intruder frame more than broadsheets, as none of the four frames constituted from the original issue specific items are used significantly different between the outlets.
Country differences
The media in Scandinavia use the generic frames quite similarly across countries when reporting on the refugee crisis. In the following the country name will be used to refer to the national media. A one-way ANOVA was conducted to test for significant differences between the countries regarding the use of frames. Four of the five generic frames showed no
significant difference in their average presence between the countries. Only the attribution of responsibility frame differed significantly between the countries at F(2, 357)=11.16, p<.01. The Scheffe´s post hoc test revealed that Norway (M=0.38, SD=0.35) used the attribution of responsibility frame significantly more than Denmark (M=0.19, SD=0.27). It is however worth noting that the Norway is also close to reporting this frame significantly more (p=.058) than Sweden (M=0.29, SD=0.32), and Sweden is close to reporting significantly more than Denmark (p=.068).
Looking at the issue-specific frames, interestingly there are more differences between the countries; three out of the four frames are used significantly different.
The victim frame is used significantly different between the countries (F(2, 357)=4.34,
Denmark (M=0.15, SD=0.22). It is also worth noting that Norway is close to using the victim frame significantly more (p=.079) than Sweden (M=0.16, SD=0.26). The asylum policy frame is also significantly different between the countries (F(2, 357)=4.05, p<.05), as Norway (M=0.08, SD=0.22) uses the frame significantly more than Sweden (M=0.02, SD=0.12). Denmark (M=0.21, SD=0.32) used the deportation frame significantly more (F(2, 357)=4.29,
p<.05) than Sweden (M=0.10, SD=0.26). This gives partial support for H6 that Denmark will
use the intruder frame more than Sweden in its reporting. However, as the original intruder frame is split into new frames in the analysis, this study will remain cautious with giving full support for the hypothesis, as the deportation frame is only constituted of two of the items from the original intruder frame. The few significant differences between the types of outlets in addition to the findings of country differences lends support to the H5 that newspapers within countries will report more similarly on issues than newspapers between countries. Changes throughout the time period
Next in our analysis we turn to the different use of frames throughout the time period under investigation. This is where we find the greatest difference in both the use of the generic frame as well as the issue specific frames. Six of the nine frames are used
significantly different, and two come close to significance. Turning first to the generic frames the one-way ANOVA shows that the human-interest frame has a significantly higher use (F(2, 357)=3.12, p<.05) in September (M=0.40, SD=0.45) compared to later on in November (M=0.26, SD=0.42). The economic consequences frame was used significantly more (F(2, 357)=10.59, p<.01) in October (M=0.19, SD=0.30) than both in September (M=0.07,
SD=0.22) and in November (M=0.06, SD=0.19). The attribution of responsibility was also
used significantly more (F(2, 357)=4.69, p<.01) in September (M=0.34, SD=0.37) than in November (M=0.22, SD=0.25). It is also worth mentioning the trends one can detect
(F(2, 357)=2.69, p=.069) comes close to being significantly more (p=.072) used in October (M=0.37, SD=0.37) than in September (M=0.26, SD=0.33). The morality frame (F(2, 357)=2.99, p=.052) is also close to being significantly more (p=.057) used in September (M=0.27, SD=0.25) than in November (M=0.19, SD=0.26).
Looking at the issue-specific frames we see that three of the four frames are used significantly differently. The threat frame (F(2, 357)=5.65, p<.01) was used significantly more in November (M=0.16, SD=0.27) than earlier on in September (M=0.07, SD=0.18). The deportation frame is also used significantly more (F(2, 357)=17.08, p<.01) in November (M=0.26, SD=0.33) than earlier on in both September (M=0.06, SD=0.21) and October (M=0.13, SD=0.26). The refugee frame (F(2, 357)=5.26, p<.01) on the other hand is used significantly more in September (M=0.24, SD=0.28) than in November (M=0.14, SD=0.23). The findings show that the frames vary throughout the time period of the research, and we can answer affirmatively to RQ2, the frames change throughout the time period. The differences between the months trend towards a humanitarian focus in September, which has changed to a more negative media climate towards the refugees in November.
Changes over time in each of the countries
We now turn to the change in frame presence during the time period at the country level. The picture then becomes more multifaceted, and it becomes clear that the different countries vary in how much the frame use changes over time. As Figure 2 and Figure 3 shows the trends are visible over time.
<Figure 2 and 3 about here>
Denmark only shows significant differences in the use of the deportation frame (F(2, 117)=4.44, p<.05) throughout the time period, which is used more in November (M=0.33,
SD=0.35) than in September (M=0.13, SD=0.29). Sweden, on the other hand, shows
human-interest frame (F(2, 117)=5.33, p<.01) is used significantly more earlier on in September (M=0.51, SD=0.45) than both in October (M=0.22, SD=0.41) and in November (M=0.27,
SD=0.43). The same difference is found in the use of the morality frame (F(2, 117)=4.58, p<.05), which is used significantly more in September (M=0.33, SD=0.23) than in both
October (M=0.18, SD=0.24) and November (M=0.19, SD=0.25). Finally, the refugee frame (F(2, 117)=5.55, p<.01) is also used significantly more in September (M=0.26, SD=0.28) than in November (M=0.08, SD=0.14). The economic consequences frame (F(2, 117)=5.21, p<.01) shows a different pattern. It is used significantly more in October (M=0.18, SD=0.29) than in September (M=0.03, SD=0.10), and close to significantly more (p=.082) than in November (M=0.08, SD=0.21). The deportation frame (F(2, 117)=3.91, p<.05) is used significantly more in November (M=0.19, SD=0.31) than earlier on in September (M=0.04, SD=0.17), showing the same trend as in Norway and Denmark.
Norway shows significant differences in four frames. The attribution of responsibility frame (F(2, 117)=4.80, p=.01) is used significantly more in September (M=0.49, SD=0.38) than in November (M=0.26, SD=0.24). The economic consequences frame (F(2, 117)=4.81,
p=.01) is used significantly more in October (M=0.24, SD=0.32) than in November (M=0.07, SD=0.19), and close to significantly more (p=.082) than in September (M=0.11, SD=0.27),
following the same pattern as in Sweden. The threat frame (F(2, 117)=6.26, p<.01) on the other hand was used significantly more in November (M=0.21, SD=0.30) than in September (M=0.04, SD=0.13). The deportation frame (F(2, 117)=10.86, p<.01) was also used
significantly more in November (M=0.28, SD=0.32) than in both September (M=0.03,
SD=0.11) and October (M=0.13, SD=0.25). We see a move in Sweden from a humanitarian
outlook in September, to a trend towards more focus on discouragement and deportation in November, which is evident in all of the three countries. We now turn to a discussion of the findings.
Discussion
The purpose of this study was to look at the characteristics of the coverage of the 2015 refugee crisis in the Scandinavian media. The research had two aims, namely to test the applicability of generic frames and issue specific frames on the topic of immigration in the Scandinavian media. Secondly, it was to look at how the frames were applied across the three countries in Scandinavia in general, throughout the time period of the study spanning several months, and between the different types of newspaper outlets.
Framing the refugee crisis in a globalized world
The study has shown that the presence of the generic frames in news coverage is quite consistent with previous literature. The strong presence of the generic frames throughout the countries and the time period further establishes the strength of generic frames as tools to compare coverage across countries and across topics, in line with De Vreese (2005) and De Vreese et al. (2001). When comparing the findings to the study of Semetko and Valkenburg (2000), the use of the frames differs, however. In their study from the Netherlands on the topics of European integration and crime, the responsibility frame was the most common frame, the conflict frame coming second. Morality was barely used in the stories.
In Scandinavia, and on the topic of the refugee crisis a different picture emerges. The human-interest frame is the most common frame overall, and morality is the fourth most common frame of the nine frames in the study. This result might be explained by the difference in topic. Semetko and Valkenburg (Ibid., p. 104) found that frame use differed depending on the topic, media using the attribution of responsibility frame more on European stories, and conflict more when reporting about crime. The high use of the human interest and morality frame in the refugee crisis coverage might be explained by the media trying to grab the readers attention with emotional expressions, encouraging people to engage in the news
they read (Pantti et al., 2012, p. 61). Pantti et al. argues that emotional discourses matter, exactly because they move and motivate people to take action. The Swedish tabloid
Aftonbladet illustrates this well in the articles in the sample by openly displaying the donation
information of various aid organizations alongside gripping stories about refugees and volunteers.
Further, the research showed that the items constructed from Van Gorp´s (2005) immigration study in Belgium did not form the expected intruder frame and victim frame. The items representing the frames were present in the news articles, but divided into four distinct patterns: the refugee frame, the deportation frame, the threat frame and the asylum policy frame. This result may be explained by the situational differences of the studies. Although Van Gorp´s study from Belgium cover refugees and the asylum issue, the situation in the country is not marked by unusual events. The year of 2015 however, was extraordinary in the sense that record numbers of asylum seekers arrived in Europe and Scandinavia (Asylum statistics, 2016), with huge implications for many functions in society. The results suggest that when big and long-term events, such as the refugee crisis, takes place the issue specific frames of immigration are too narrow to catch the complexity of the issue. Rather than being treated under the topic of immigration, it might be better understood “under a broader
conceptualization of crisis” in a global context (Pantti et al., 2012, p. 18).
Underlining the global scope of the crisis might also explain some of the formations of the frames in the study. The threat frame was composed of one of the items from the
attribution of responsibility frame suggesting an individual or group was responsible for the problem in the story, and the items in the intruder frame referring to how to stop people who are a threat to the society and to protect the local people. This can be linked to the focus on terror in the media, especially after the attack in Paris in November, where a Syrian passport was found beside one of the suicide bombers. The refugee crisis is seen as interlinked with
other issues when the media reports on the topic, not “confined within the borders of particular nation states”, but as a part of interconnected global issues (Cottle, 2011, p. 78). Rendering the invisible differences visible
The research showed that there are few differences between the broadsheet and the tabloid newspapers. This contrasts earlier findings in research on generic frames (Semetko &Valkenburg, 2000) and issue specific frames (Van Gorp, 2005). The conflict frame is the only frame used significantly different, which is in line with previous research from Schuck et al. (2013b). Therefore, the overall findings support the notion that each of the Scandinavian countries conforms to the same editorial lines within the country (Lund, 2007, p. 124). The similarities within each country seem to be greater than the differences between the types of outlets in Scandinavia.
The research showed that there are several differences between the countries in the use of frames, but mainly in the issue specific frames. This underlines the importance of
comparing similar countries to avoid the assumption that a characteristic is true for one country based on the findings in another (Blumler & Gurevitch, 1995, p. 75), and to make visible the differences in Scandinavia that might be insignificant in a larger context
(Strömbäck et al., 2008, p. 12). The results partly support the hypothesis that Denmark uses the intruder frame more than Sweden, as the Danish media used the new deportation frame more often, which is constituted of two of the from the original intruder frame. The Danish focus on this frame might be explained by several country specific factors
The social democratic government stepped down after the Danish national elections in June 2015, although still being the biggest party in the country. The liberal conservatives came in third, and formed a minority government with the backing from the second largest party, the right wing anti-immigrant Danish people´s party (Eddy, 2015). One of the main issues of the national election was immigration and further restrictions on the asylum policies.
Denmark has a high level of issue voting (Esmark & Ørsten, 2008, p. 32) and Danes have shown to be much less positive to immigrations from the outside of the EU than Swedes (Eurobarometer 83). Therefore, the Danish government could rely on strong public support for policies discouraging asylum seekers to come to Denmark. The newspaper coverage in the fall of 2015 hence circled around the political discussions on restrictions, how to discourage the flow of refugees from choosing Denmark as their final destination, and thus can thus explain the high use of this frame in the study.
The Norwegian media uses the asylum policy frame more than the Swedish media. This frame can be seen as a middle position. It focuses on lax asylum policies being the reason for people coming into the country, but also emphasizing the importance of a fair asylum system. Hence the frame is not merely focused on discouragement and deportation, like in Denmark. An explanatory factor for this finding can be the current political climate in Norway, and where the most powerful politicians receive the most attention in the media (Hopman et al., 2011). Norway has a minority government constituted of the conservative party and the right wing party. However, the minority government, contrary to the Danish government, has to seek towards the center for support for the policy changes. The two supporting parties are the Liberals and the Christian people´s party. As proposed by Green-Pedersen and Krogstrup (2008, p. 613), this party group aligns with the economic views of the right, but not other issues. This is evident in Norway where especially the Christian people´s party has a strong humanitarian profile. The high use of the asylum policy frame in the news could therefore be seen as a result of an immigration debate the governing and supporting parties must carefully craft between the right wing party´s wish for
restrictions, and the notion of the importance of keeping up a fair and humane asylum system, as demanded from the supporting parties.
The times they are a changin’
The research showed that month matters for frame use. Frames are used differently between the months in the study, and this trend is especially evident from September to November. The results are thus in agreement with the research from De Vreese (2005), Van Gorp (2005), and Roggeband and Vliegenthart (2007) that frames respond to outside events and stakeholders, such as politicians, the media itself and public opinion. Although this could not be modeled empirically in the current study, there are strong signs for this assumption.
Further, the research showed that the countries vary in how much the coverage change in the time period. As the general country differences already pointed towards, the Danish media focused more on restrictions and deportation than the other countries´ media. The patterns throughout the months point towards Denmark having a high focus on deportation and threat overall, and a low humanitarian profile. The Norwegian and Swedish media on the other hand make a big move from the humanitarian angle in September, to aligning with the Danish media in November with a higher focus on refugees as a threat and discouragement to come to the country than the previous months. The focus of the Danish media in the early period of the study might be explained by the fact that immigration already had been a political issue for many years (Green-Pedersen and Krogstrup, 2008).
Generally, the changes in the use of frames might be explained by the developments and the specific events that dominated the media coverage the particular month. The use of the frames focusing on the humanitarian side of the crisis was higher in September than later on in the period of the study. September was also when the growing awareness of the
magnitude of the refugee crisis led to an increased news value of the issue. This is shown by the fact that all the newspapers in the sample had teams of journalists reporting directly from the beaches in Greece and on the refugee trails in Eastern Europe at this point. This can have led to what Van Gorp (2005, p. 502) calls “frame driven” news, where the scope of the crisis
made the journalists actively seek out stories that would fit the humanitarian aspect of the crisis. One specific event also caused heavy media coverage, namely the death of Alan Kurdi, resulting in the now iconic picture of the little boy on the beach. This event could have caused the media to focus more on the refugees and the issues and dangers they faced, in alignment with Brosius and Eps (1995) findings that extraordinary events make the media more prone to cover similar stories. These general developments and specific events are therefore likely to have caused a more human centered coverage of the crisis in September, in particular in Norway and Sweden.
In November the coverage has clearly turned. The media now focus on the threat the refugees pose to society, and stories on discouraging asylum seekers to come to the country, than in the previous months. This can also be seen in light of both general developments, and specific events affecting the coverage. Generally all of the countries were feeling the weight of the numbers, and the migration systems were breaking down in both Norway and Sweden because of the amount of refugees coming into the countries. All of the Scandinavian
countries where debating stricter asylum policies during the fall and the debates were reported extensively in the media, likely causing the increase in the use of the deportation frame in all the countries.
A violent event happening unexpected and causing major damage will according to Brosius and Eps (1995, p. 408) have the potential to cause long term impact on the news coverage. Such an event occurred in Paris the 13th of November. The terror attack was linked directly to the refugee crisis when a Syrian passport was discovered beside one of the suicide bombers, and later found to have been registered in Greece in the midst of the flow of
refugees crossing the Mediterranean. The link made between the flow of unregistered refugees coming into Europe and the major and sudden event, might explain the increase in the threat discourse in November.
The challenges ahead
Overall, the coverage seems to have turned more negative towards the refugees. These findings become especially relevant, as previous research has shown that media coverage have an effect on people’s opinions (see e.g. Lecheler et al., 2015) and political attention (see e.g. Vliegenthart & Roggeband, 2007; Walgrave & Van Aelst, 2006). A coverage
continuously focusing on the negative aspects of immigration might therefore have
implications for society. Pantti et al. (2012, p. 26) warns that crisis can lead people to give up what they normally would protect. The right for asylum and a fair tried case would for
example be something which society under normal circumstances would agree was important to protect. As the numbers of refugees go up, and the intensity of the immigration debate increase, the asylum seeker seems to not necessarily be something a majority would protect in the future.
There are several shortcomings in the study. Firstly, Krippendorff´s alpha is reported in the study, where some of the items are below Krippendorff´s (2004, p. 241) suggested lower limit (α=.667) of inter-coder reliability. Therefore, some caution should be taken when interpreting the results of the study. Another possible limitation of the study is that the items of the frames did not match the predefined frames in the final analysis. This can imply that Matthes and Kohring´s (2008) critique is partly justified in that deductive frames are too inflexible to discover emerging frames. The frames might have more and other indicators, which were not captured in this study. The grand scope of the refugee crisis might also have led to the formation of new frames in the intersection of immigration and crisis coverage.
Future studies should therefore take two approaches. Firstly, an inductive approach should be taken to the framing of extraordinary events such as the refugee crisis, in search of other frames and framing items. Secondly, more comparative research on the topic of
countries. However, the results in the study show the usefulness of investigating the presence of the generic and the issue specific frames on the topic of immigration in Scandinavia.
Although different frames were detected than what was initially expected, the analysis yielded important findings, relating to the differences in time and the political contexts of the
countries. Future studies should also continue to broaden the comparative research agenda on media coverage in Scandinavia, as comparative research between the countries is still scarce (Horsti, 2008b). This should include other topics and issues, to gain a more comprehensive picture of the media coverage in the region, as this study has shown that important differences do exist between the countries.
The limitations of the study notwithstanding, the study provides an important contribution to the comparative research on Scandinavian media generally and the use of generic and issue specific frames on the topic of immigration specifically. In light of the effects of media coverage on public opinion, the findings in the study becomes especially relevant, as they point towards a public debate which can have implications for society in the coming years if the negative coverage continues, as integration and not immigration of the people in question is likely to become the biggest challenge of the day.
References
Aalberg, T., Strömbäck, J., & De Vreese, C. H. (2011). The framing of politics as strategy and game: A review of concepts, operationalizations and key findings. Journalism, 13(2) 1-17
Asylum statistics. (2016, May 11). Eurostat. Retrieved from
Babbie, E. (2005). The basics of social research (3rd). Belmont, CA: Thomson Wadsworth. Berry, M., Garcia-Blanco, I., & Moore, K. (2015). Press Coverage of the Refugee and
Migrant Crisis in the EU: A Content Analysis of Five European Countries. Report
prepared for the United Nations High Commission for refugees.
Blumler, J., & Gurevitch, M. (1995). The crisis of public communication. Psychology Press. Brosius, H. B., & Eps, P. (1995). Prototyping through key events News selection in the case
of violence against aliens and asylum seekers in Germany.European Journal of
Communication, 10(3), 391-412.
Brüggemann, 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(6), 1037-1065.
Cottle, S. (2011). Taking global crises in the news seriously: Notes from the dark side of globalization. Global Media and Communication, 7(2), 77-95.
De Vreese, C. H. (2005). News framing: Theory and typology. Information design journal+
document design, 13(1), 51-62.
De Vreese, C. H., Banducci, S. A., Semetko, H. A., & Boomgaarden, H. G. (2006). The news coverage of the 2004 European Parliamentary election campaign in 25
countries. European Union Politics, 7(4), 477-504.
De Vreese, C. H., Jochen Peter, Holli A. Semetko, C. (2001). Framing politics at the launch of the Euro: A cross-national comparative study of frames in the news. Political
communication, 18(2), 107-122.
Eddy, Melissa (2015, June 18). Anti Immigrant Party Gains in Denmark. The New York
Times. Retrieved from http://www.nytimes.com
Entman, Robert (1993) Framing: Towards a clarification of a Fractured Paradigm. Journal of
Esmark, A., & Ørsten, M. (2008). Media and politics in Denmark. In Strömbäck, J., Ørsten, M., & Aalberg, T. (Eds.), Communicating Politics: political communication in the
Nordic countries, Nordicom.
European Commission (2015) Eurobarometer 83. Brussels: EC DG Press and Communication.
Första ganger över 10 000 asylsökande de senaste sju dagarna (2015, November 9). Retrieved from http://www.migrationsverket.se/Om-Migrationsverket/Nyhetsarkiv/Nyhetsarkiv-2015/2015-11-09-Forsta-ganger-over-10-000-asylsokande-de-senaste-sju-dagarna.html Goffman, E. (1974). Frame analysis: An essay on the organization of experience. Harvard
University Press.
Green-Pedersen, C. & Krogstrup, J. (2008). Immigration as a political issue in Denmark and Sweden. European journal of political research, 47(5), 610-634.
Hallin, D. C., & Mancini, P. (2004). Comparing media systems: Three models of media and
politics. Cambridge University Press.
Hopmann, D. N., de Vreese, C. H., & Albæk, E. (2011). Incumbency bonus in election news coverage explained: The logics of political power and the media market. Journal of
Communication, 61(2), 264-282.
Horsti, K. (2007). Asylum seekers in the news: Frames of illegality and control.Observatorio
(OBS) Journal, 1, 145-161.
Horsti, K. (2008a). Europeanisation of public debate: Swedish and Finnish news on African migration to Spain. Javnost-the public, 15(4), 41-53.
Horsti, K. (2008b). Overview of Nordic media research on immigration and ethnic relations: From text analysis to the study of production, use and reception. Nordicom
Krippendorff, K. (2004). Content analysis: An introduction to its methodology (2nd ed.). Thousand Oaks, CA: Sage.
Lecheler, S., Bos, L., & Vliegenthart, R. (2015). The Mediating Role of Emotions News Framing Effects on Opinions About Immigration. Journalism & Mass Communication
Quarterly, 92(4), 812-838.
Lund, Anker B. (2007). Media Markets in Scandinavia. Nordicom Review, Jubilee Issue, 121-134.
Matthes, J., & Kohring, M. (2008). The content analysis of media frames: Toward improving reliability and validity. Journal of Communication, 58(2), 258-279.
Pantti, M., Wahl-Jorgensen, K., & Cottle, S. (2012). Disasters and the Media. Peter Lang Publications.
Refugees/Migrants Emergency Response – Mediterranean (2016). Retrieved from http://data.unhcr.org/mediterranean/country.php?id=83
Roggeband, C., & Vliegenthart, R. (2007). Divergent framing: The public debate on migration in the Dutch parliament and media, 1995–2004. West European
Politics, 30(3), 524-548.
Schuck, A. R., Boomgaarden, H. G., & de Vreese, C. H. (2013a). Cynics all around? The impact of election news on political cynicism in comparative perspective. Journal of
Communication, 63(2), 287-311.
Schuck, A. R., & De Vreese, C. H. (2006). Between risk and opportunity news framing and its effects on public support for EU enlargement. European Journal of
Communication, 21(1), 5-32.
Schuck, A. R., Vliegenthart, R., Boomgaarden, H. G., Elenbaas, M., Azrout, R., van Spanje, J., & De Vreese, C. H. (2013b). Explaining campaign news coverage: How medium,
time, and context explain variation in the media framing of the 2009 European parliamentary elections. Journal of Political Marketing, 12(1), 8-28.
Schuck, A. R., Vliegenthart, R., & De Vreese, C. H. (2016). Who's Afraid of Conflict? The Mobilizing Effect of Conflict Framing in Campaign News. British Journal of Political
Science, 46(01), 177-194.
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.
Strömbäck, J., & Aalberg, T. (2008). Election news coverage in democratic corporatist countries: A comparative study of Sweden and Norway. Scandinavian Political
Studies, 31(1), 91-106.
Strömbäck, J., Ørsten, M., & Aalberg, T. (2008). Communicating politics: political
communication in the Nordic countries. Nordicom.
Strömbäck, J., & Nord, L. W. (2008). Media and politics in Sweden. In Strömbäck, J., Ørsten, M., & Aalberg, T (Eds.), Communicating politics: political communication in the
Nordic countries, 103-121.
Thorbjørnsrud, K. (2013). The Autonomy of Scandinavian Public Service Broadcasters During Election Campaign Periods. Nordicom Review, 34(1), 63-76.
Vliegenthart, R., & Roggeband, C. (2007). Framing immigration and integration Relationships between press and parliament in The Netherlands. International
Communication Gazette, 69(3), 295-319.
Van Aelst, P., & Vliegenthart, R. (2014). Studying the tango: An analysis of parliamentary questions and press coverage in the Netherlands. Journalism Studies, 15(4), 392-410. Van Gorp, B. (2005). Where is the frame? Victims and intruders in the Belgian press
Van Spanje, J., & de Vreese, C. (2014). Europhile media and Eurosceptic voting: Effects of news media coverage on Eurosceptic voting in the 2009 European parliamentary elections. Political Communication, 31(2), 325-354.
Walgrave, S., & Van Aelst, P. (2006). The contingency of the mass media's political agenda setting power: Toward a preliminary theory. Journal of Communication, 56(1), 88-109.
Endnotes
1. Newspaper circulation numbers retrieved from:
Denmark: http://mediawatch.dk/Statistik/Oplagstal_aviser/article7496701.ece Norway: http://medienorge.uib.no/?cat=statistikk&page=avis&queryID=190 Sweden: http://ts.se/mediefakta-upplagor/dagspress/
2. Search strings in the local languages:
Denmark: (flygtning OR flygtningekrise OR flygtninge OR migranter OR flygtningekrisen OR flygtingekatastrofen)
Norway: (flyktning OR flyktningkrise OR flyktninger OR migranter OR flyktningkrisen) Sweden: (flykting OR flyktingkrise OR flyktinger OR migranter OR flyktingkrisen OR flyktingkatastrofen)
3. Frame items below Krippendorff´s suggested level of reliability:
Does the story suggest that some level of government has the ability to alleviate the problem? (α=.630)
Does the story suggest that an individual (or group of people in society) is responsible for the issue-problem? (α=.650)
Does the story make reference to morality, God, and other religious tenets? (α=.641) 16. Does the story offer specific social prescriptions about how to behave? α=.654
Is there a reference to economic consequences of pursuing or not pursuing a course of action? (α=.657)
Does the article mention that the policy makers, and the asylum seekers themselves are responsible for dealing with the problem? (α=.641)
Does the article mention the importance of protecting the interests of the native people? (α=.657)
Does the article discuss how the country will receive refugees, which are victims of persecution and forced to leave their country? (α=.657)
Does the article mention violence, persecution, poverty and the gap between rich and poor countries as reasons for the people coming to Europe? (α=.650)
Does the article speak of a flexible, cautiously and effectively applied asylum policy? (α=.657)
Does the article mention that we have a moral duty to help people in need? (α=.657)
Appendix A
Coding instructions: 1. Newspaper
Each newspaper is given a number below as exemplified below: 1=Dagens Nyheter 2=Aftonbladet 3=Aftenposten 4=Dagbladet 5=Politiken 6=Extra Bladet 2. Code
Each article is given a code number, following from the example below: 100– Dagens nyheter (101, 102, 103 etc)
200– Aftonbladet (201, 202, 203 etc) 300– Aftenposten 400– Dagbladet 500– Politiken 600– Extra Bladet 3. Title
Write down the 4 first words of the title 4. Date
Write down the date of publishing in this manner: 09.10 Attribution of responsibility:
1. Does the story suggest that some level of government has the ability to alleviate the problem?
The question should be answered with 0=No or 1=Yes. The question should be answered Yes if the article specifically states that some level of national, foreign or international
government can alleviate the problem. The answer should be Yes if the article states what should be done to alleviate the problem and also if it does not specify what should be done. It should be Yes both if the article mentions the possibility of alleviate and/or stopping the problem or issue. The issue can be alleviating the problem of beds for the refugees, stopping the influx of refugees, but not limited to this. In case of doubt, the answer of the question should be No.
Example: “Cameron promised the police in Calais British aid in the form of police dogs and fences to stop the flood of people.”
2. Does the story suggest that some level of the government is responsible for the issue/problem?
The question should be answered with 0=No or 1=Yes. The question should be answered Yes if it states that some level of a national, foreign or international level of government is
