News diversity: a Longitudinal and Comparative Perspective

You are holding a manuscript that has been written within the Steunpunt Media (2012-2016), a policy instrument and knowledge centre that was set up to monitor the pluriformity of the Flemish news media landscape, initiated by the former Flemish Minister of Media Ingrid Lieten. The starting point of our research is the notion that content diversity, by itself, ought to be protected within a democratic society (Lieten, 2011; Napoli, 1999; Raeijmaeckers & Maeseele, 2015). Furthermore we intended to explore in what way reflecting about news diversity is important and useful for Flemish journalists.

Following the method of the European Monitoring Study (EU Media Pluralism Monitor), we approached news diversity as a multidimensional concept from three perspectives: that of the content, the journalist and the audience. Traditionally, most studies only focus on content diversity (e.g. Glasser, 1984; Hoffmann-Riem, 1987; Owen, 1977, 1978; McQuail & Van Cuilenburg, 1983). The content perspective in this study was operationalised along four dimensions: diversity of actors, diversity of opinions, diversity of topics, and geographical diversity (based on Hoffmann-Riem, 1987). Furthermore, we also add the perspective of the journalist as well as that of the audience.

Two normative frameworks are referred to so as to evaluate the level of diversity as ‘good enough’: reflective diversity and open diversity. Strictly speaking, diversity can be interpreted as a reflection of societal reality, which Napoli (1999) defined as demographic diversity: the extent to which the various demographic groups are given a voice in direct proportion to their prevalence in society. More broadly interpreted, reflective diversity entails the extent to which a dimension in the (news) content is distributed in proportion with the prevalence in society (after e.g. Van Cuilenburg, 1998, 1999, 2016). Theoretically speaking, representative diversity equals open diversity: the extent to which divergent categories within a dimension are in absolute terms equally (i.e., statistically uniformly) represented in news content (after McQuail, 1992; Van Cuilenburg 1998, 1999, 2005, 2016).

This manuscript consists of four empirical chapters. Each chapter brings another perspective to news diversity. While chapters 2 and 3 shed light on the content perspective, Chapter 4 looks at the journalistic perspective and Chapter 5 considers the audience perspective. We will give a succinct overview of the most important conclusions derived from each of these studies.

In Chapter 2 we measured the four dimensions (diversity of actors, opinions, topics and geographical diversity) of content diversityacross several media. The sample included two weeks of 2012, consisting of nine Flemish news brands (Het Journaal (VRT), Het Nieuws (VTM), De Morgen, De Standaard, Het Belang van Limburg, Het Nieuwsblad, Het Laatste Nieuws, hln.be and hbvl.be) of three different media types (television news, newspapers and online news sites). Diversity of actors was operationalised as the prevalence of women and ethnic minorities in news stories. In this study, opinion diversity referred to diversity of sources along four categories (gender, ethnicity, status position and geographical distance). Diversity of topics is based on a preset list of news categories (national politics, law, crime, economy & business, wellbeing & migration, environment, culture & entertainment, science, technology & education, disasters, mobility, war & peace, sports, and international & European politics). Geographical diversity was operationalised as the distribution of domestic, mixed and foreign news based on the countries mentioned in the news stories.

None of the studied news brands showed enough gender diversity according to the reflection norm. The two television news brands, Het Journaal (VRT) and Het Nieuws (VTM), are sufficiently diverse according to the 80%-rule of the reflection norm (i.e. the equivalent to 80% of the maximum reflective diversity according to the valid norm for sufficient gender diversity within the management context, see Ali, Kulik, & Metz, 2011). In television news we came across significantly more female news sources than in newspapers and online news sites.

Again according to the reflection norm, ethnic diversity of actors and sources was not sufficient in any of the news brands under study. At hln.be, the 80%-rule of the reflection standard for ethnic diversity of actors was achieved. However, ethnic minorities were portrayed in traditional roles and contexts such as that of a politician or a criminal in news stories about national politics and law. This is consistent with previous findings (e.g. Devroe, 2007; Hussain, 2000; Ter Wal, d'Haenens, & Koeman, 2005). The VRT newscast showed more ethnic minorities both as actor and as source than the commercial news broadcaster VTM. This finding refutes a previous study (Van den Bulck & Broos, 2011).

Every news brand under study was insufficiently open diverse when it came to the subdimension status position within the diversity of sources. Flemish journalists based their news stories primarily on elite sources. This corresponds to the journalistic practice to frame facts based on the professional authority of expert sources (Albæk, 2011). However, in relative terms, we came across more “vox populi” on television than in newspapers and on online news sites.

The subdimension geographical distance within diversity of sources was in general sufficiently open diverse, an unexpected exception being the newspaper Het Belang van Limburg that used few sources in a foreign context due to its regional focus. The diversity of topics and geographic diversity were generally sufficient for all types of news media.

In conclusion, Flemish news media were found to be sufficiently diverse at the level of what was told and where news stories took place. However, news stories were insufficiently diverse at the level of who was implicated in the news stories (diversity of actors).

A second striking finding was that the news media types differed only slightly from each other. At the level of the choice of news stories we found mainly similarities. Thus, it seems that independently of the news media type the same news stories were selected due to the same evaluation of newsworthiness, which reaffirms the importance of media logic associated with news production (after Ghersetti, 2014).

Finally, Het Journaal (VRT) was the most open diverse news brand in the sample with the highest average value for all dimensions of the Simpson’s Dz measure (an unambiguous diversity measure for the distribution of proportions in different categories).

In Chapter 3 levels of gender and ethnicity diversity were examined longitudinally. We looked at the extent to which women and ethnic minorities are given voice in the news content. We operationalised gender and ethnic diversity as the prevalence of women and ethnic minorities in various topics and status positions. Our sample included four news brands (Het Journaal (VRT), Het Nieuws (VTM), De Standaard en Het Laatste Nieuws) in three years (2003, 2008 and 2013).

In line with the results of Chapter 2, none of the news brands under scrutiny in any of the three years were sufficiently gender diverse according to the reflection norm. However, according to the 80%-rule of the reflection standard, Het Nieuws (VTM) in 2008 and 2013 and Het Journaal (VRT) in 2008 (only for the diversity of actors) resulted to be sufficiently gender diverse. Moreover, gender diversity increased over the years in Het Journaal (VRT), Het Nieuws (VTM) and Het Laatste Nieuws. In all, we found no clear impact of the explicit diversity policy instruments of the public broadcaster as in any of the years under study, Het Nieuws (VTM) was the most gender diverse news medium.

Like in 2012, for each of the three years almost all of the four news brands showed too little reflective ethnic diversity, an exception being the ethnic diversity of actors in 2003 in Het Nieuws (VTM). In that year actually more actors with an ethnic minority background happened to be mentioned in the news in proportion to their relative occurrence in Belgian society. It is doubtful, however, whether this indicates a more diverse picture. The three most common roles for ethnic minorities were politicians (17%), suspects (14%) and terrorists (14%). The three most frequent news topics were law (38%), wellbeing & migration (24%) and national politics (17%). This was in line with the representation of ethnic minorities at hln.be in 2012 as well as in previous studies on ethnic representation (e.g. Devroe, 2007; Hussain, 2000; Ter Wal et al., 2005).

We wondered whether the public broadcaster’s newscast was more ethnic diverse compared to the other news brands. In 2013 most ethnic minorities on the screen were given a voice in the VRT’s newscasts. Particularly in the case of “vox populi”, journalists of the VRT seemed to make more efforts to quote ethnic minorities in the news stories in comparison with their colleagues of commercial news brands. As a result, the value for Simpson’s Dz value exceeded the reflection norm.

In conclusion, the impact of the explicit diversity policy instruments of VRT seems to be rather limited in the case of gender and ethnic diversity (see also De Swert & Hooghe, 2010; Van den Bulck & Broos, 2011). However, the prevalence of ethnic minorities increased during our period of research as the public broadcaster ended up being the most ethnically diverse news brand in 2013. Particularly the proportion of “vox populi” with an ethnic minority background was high, which affirms awareness of the diversity issue among Flemish journalists.

In Chapter 4, we examined the extent to which Flemish print journalists incorporate gender and ethnic diversity in their reporting. This was a measurement for journalistic diversity. Based on 16 in-depth interviews, we distinguished five different arguments about diversity in the newsroom: (1) Diverse representation as incompatible with the news making process, (2) Equality as the norm, (3) Diverse representation as an active search process, (4) News practice knows no diversity, and (5) Reality knows no diversity. The first three arguments concern the conceptual relevance of diversity, while the latter two focus on practical obstacles to implement diversity in news reporting. The main conclusion of this study is that most Flemish print journalists, like their American colleagues (Robinson & Culver, 2016), are not actively nor consciously preoccupied with giving a voice to those for whom a voice has been denied such as women and ethnic minorities. Journalists selected their news sources from their fixed familiar contact lists of experts, andthese lists seem to continue to be dominated by “white” men.

Chapter 5 investigated diversity of the audience. This study uniquely linked the rating figures including demographic characteristics (gender, age, education level and social group) of Het Journaal (VRT) and Het Nieuws (VTM) with the content diversity of topics present in the newscasts. The sample was composed of the same three years (2003, 2008 and 2013) as in the study in Chapter 3.

A striking conclusion was that, irrespective of demographic background, the Flemish audience watched the news broadcasts from beginning to end. This implies that Flemish television news viewers are mainly passive viewers (after Van der Wurff, 2004): they consume the news menu as it is offered. In the case of the traditional news medium television, there was only a slight lack of alignment between what the public wanted (demand) and what the news providers offered (supply) (Boczkwoski & Mitchelstein, 2013).

However, the audience dropped out of some news topics in 2003 and 2013. This was not the case in 2008. One of the most prominent results in this respect was that the audience of both newscasts did not follow all news about the 2003 federal elections. The perceived excess of national political news on both channels translated into lower average rating figures. However, this was not the case for all demographic groups. Surprisingly, for example, is that Flemish youth and young adults watched the political news as it was offered, both in Het Journaal (20-24 year olds) and Het Nieuws (15-19 and 25-34 year olds). This finding is consistent with the interest for political news measured among British youth (Newman, Levy & Nielsen, 2015)

From this study on diversity of the audience, we can derive two conditions to achieve diversity in television news: a big – diverse – enough audience who consumes the newscasts and a sufficient amount of news content diversity. Both the public and private Flemish broadcasters seem to be quite successful in this, although it remains a challenge at both European and Flemish levels to reach younger news consumers via television (e.g. DigiMeter, 2016, Newman, Flechter, Levy, & Nielsen, 2016).

One of the major recommendations of this study is to improve the awareness on the diversity issue among journalists. We bring forward four arguments why professional journalists have an advantage if they are aware of diverse news reporting. Firstly, there is the commercial argument. The more diverse news content is, the more likely a larger audience will recognise itself and consume this content (e.g. Costera Meijer, 2012; Reich & Godler, 2015). Secondly, diversity is connected with the professional standard of truthful reporting. This is enshrined in Article 1 of the Code of the Flemish Press Council. Truthfulness is defined as reflecting reality within a representative democracy (Raeijmaekers & Maeseele, 2015), which is related to the reflection standard of diversity. Thirdly, diversity is linked to the professional standard of reportin in a balanced way. This means that all sides of the story should be heard, both from a majority and a minority perspective (e.g. Hackett, 1984; Skovsgaard, Albæk, Bro, & the Vreese, 2013). In other words, everyone, regardless of his or her status, should be able to do his or her story in the news, which is in turn related to the openness standard of diversity. Finally, there is the argument of social responsibility. This requires commitment from journalists by looking actively for solutions to social problems. In the case of diversity, this implies that journalists should seek consciously and actively to give a voice to consistently underrepresented groups such as women and ethnic minorities.