Does News Diversity Work Among Audiences? A Citizen Experiment

ABSTRACT

Scholarly and policy debates alike frequently highlight the normative value of news diversity as a necessary prerequisite for journalism to optimally fulfil its watchdog function in democratic societies. Various studies have sought to conceptualise news diversity and/or to operationalise it from the vantage point of journalists and newsrooms as news content producers. However, very little remains known about how news diversity is received and perceived by audiences. In this paper, we outline an audience experiment carried out among Flemish (Belgian) adult citizens (N = 208) to examine whether enhanced viewpoint diversity in news content affects its perceived diversity and credibility. Using news content focusing on one controversial and one more politically moderate topic to assess differences in the perception of news content diversity, we find that high viewpoint diversity is observed by audiences whereas low viewpoint diversity is not. Moreover, news articles with only one viewpoint that matched their own were perceived as more credible. We conclude that researchers and policymakers can only advance the ongoing important debate on news diversity if they choose to pay more attention to how audiences (don’t) engage with it.

KEYWORDS:

• News diversity
• viewpoint diversity
• media bias
• audience experiment
• news content
• journalism

Democracy and Diversity, Thick as Thieves?

Democratic societies are trademarked by a necessity for a well-informed citizenry able to participate in political decision-making processes, such as voting in free and fair elections (Milner 2002). To facilitate this, a free and independent press is required to independently provide citizens with factual information (Schultz, Brennan, and Castles 2009; Thompson 1995). The need for a pluralistic media landscape where various opinions and viewpoints are represented continues to be underlined (Hendrickx and Van Remoortere 2023; Mattis et al. 2021; Napoli 2011), particularly in the light of recent, detrimental global trends including the erosion of democratic values and free news media (see for instance the most recent World Press Freedom Index and the Democracy Index which both confirm these tendencies).

Media regulators across policy levels have enacted laws and guidelines to foster media freedom and pluralism. For instance, in 2022, the proposed and discussed European Media Freedom Act focuses on safeguarding specifically media pluralism and independence. Media scholars, meanwhile, have made attempts at conceptualising and operationalising news diversity for decades. Nevertheless, the concept remains ill-defined due to a continued lack of a shared understanding (for recent structured literature reviews, see Hendrickx et al. 2020; Joris et al. 2020). Other studies have looked at the diversity of news content in terms of the outlets, topics, actors and/or viewpoints in reporting from various countries using content (e.g., Badr 2021; Beckers et al. 2017; Hendrickx and Ranaivoson 2019; Magin et al. 2023; Sjøvaag 2014; Vogler, Udris, and Eisenegger 2020). They reached disparate conclusions due to different research scopes and questions, again confirming the lack of a shared consensus. Thus far, significantly less attention has been devoted to how news content diversity is perceived and received by citizens. This so-called “diversity as received” or “exposure diversity” (Helberger, Karppinen, and D’Acunto 2018; McQuail 1992; van der Wurff 2011) has recently gained academic momentum in the light of digitisation and, among other things, algorithmic news recommender systems (Hendrickx et al. 2021; Smets, Hendrickx, and Ballon 2022).

Along with the normative relevance of news diversity in a healthy democratic society, the recently acknowledged audience turn in journalism studies centring the agentic power of individuals to engage with news content in various ways (Costera Meijer 2020; Swart et al. 2022) serves as the second foundation for our study. The long-held view that citizens will automatically have a more varied media diet if news media offer diverse content is no longer taken for granted (van der Wurff 2011). This paper will examine whether audiences notice news content diversity and whether they perceive it to be more credible. On the one hand, there are reasons to expect that audiences will perceive diverse news items to be credible. Diverse viewpoints are one of the main principles of “impartial” journalism (Schultz 1998), which is expected to be perceived to be of higher quality by audiences. On the other hand, audiences may find one-sided stories to be more comprehensible and less complex (Benson 2009), making them more credible, especially if they align with their own views on the topic (e.g., Garrett and Stroud 2014; Kim 2015). Investigating perceptions of news credibility is vital because research has established a robust correlation between these perceptions and news consumption (Fletcher and Park 2017; Nelson and Kim 2021).

This paper focuses on an audience experiment in the Belgian context, investigating if and how viewpoint diversity affects audiences’ perceived credibility of news content and diversity perceptions. To our knowledge, it is one of the first recent studies that seeks to answer the pivotal question of whether or not actively fostering news content diversity is received and preferred by audiences or not.

News Diversity as a Normative Concept

News diversity, previously defined as “the various ways of producing, disseminating and consuming news for and on different platforms within one news brand, one media company and/or one media market, directly and indirectly influenced by overarching media market and company characteristics” (Hendrickx et al., 2020, 1766), is considered as a crucial element in facilitating the processes of democratic societies (Badr 2021; Hendrickx and Van Remoortere 2024; Mattis et al. 2021). It is predominant both within concepts of liberal and deliberative democracies, wherein news diversity is routinely situated in (also highly normative) journalistic standards such as neutrality, rationality, civility, and discursivity (Loecherbach et al. 2020; Magin et al. 2023). However, in its basic understanding, it is a highly normative concept that has largely failed to prove its usefulness in considering how citizens actually engage with news (Raeijmaekers and Maeseele 2015). Thus, despite the recent heightened scholarly attention to news diversity (Hendrickx et al. 2020; Joris et al. 2020), the concept remains problematic.

While its relationship with democratic societies and values is universally shared within relevant scholarship, different models of democracy across the world hamper a joint understanding of news diversity (Magin et al. 2023). Its intrinsically normative features and foundations, rooted in idealised functions and positions for news media in society, have created different journalistic and scholarly traditions in vogue simultaneously. This has led to the finding that existing news diversity research is effectively “largely incomparable due to different conceptualizations” (Joris et al. 2020, 1905). In turn, the diversity of viewpoints comprises just one dimension of news content diversity. Various studies have assessed the actors present in news reporting (Beckers and Van Aelst 2019; Masini et al. 2018), sharing the view that who receives a platform is as important as what is being said.

Philip Napoli’s (1999) conspicuous classification of diversity consisted of ownership, source, content and exposure diversity. This spawned similar operationalizations of news diversity existing across different sections, such as Sjøvaag’s (2016) organisational, structural, production, output and reception diversity, as well as ownership, brand, production, content and consumption diversity as schematised by Hendrickx et al. (2020). These taxonomies share the embeddedness of news content diversity within a wider, more holistic framework of various components which together envelop the general news diversity moniker. They stem from the assumption that external factors such as the composition of media systems and markets and the available production means and freedoms all have the power to alter news diversity and its facets, in different sequences and directions. This also includes the range of modes audiences have at their disposal to engage with news content and instrumentalise it to form their own opinions on important subject matters. Put differently, and building on existing literature, we argue that news content diversity is just as prone to being adapted by changes in audience behaviour as vice versa. Applying an online experiment among news audiences, we explore this rather unknown relationship.

News Diversity and its Relationship with Audiences

In exploring the bond between the diversity in news content and audience behaviour, we emphasise the increasingly important role ascribed to citizens as collective audiences in news-related research. Several scholars have in recent years argued that the rippling effects of media convergence and digitisation include the blurring of boundaries between content producers and consumers (Bruns 2008), a rising dependence on “black box algorithms” to recommend said content (Helberger, Karppinen, and D’Acunto 2018) and the growth of individuals’ agency in choosing how to engage with digital media (Hendrickx 2022), coupled with a tendency to track and monitor their behaviour through audience analytics omnipresent in the contemporary newsroom (Lamot and Paulussen 2019). Even though recent research has indicated that news audiences and scholars “share a common understanding of news media quality” (Bachmann, Eisenegger, and Ingenhoff 2022, 32), very little remains known, however, on to what extent news audiences are aware of diversity in news content.

We build on the experimental study of van der Wurff (2011) on how Dutch and Irish citizens receive diversity when confronting them with news content focusing on European integration. The author found that interest and knowledge on the topic served as more useful predictors of receiving diverse ideas than exposure itself. In this study, we further examine how viewpoint diversity affects not only the reception but also the perception of news content and its credibility among citizens. At the time of writing, only a narrow body of scholarship exists that studies the relationship between news diversity and perceived quality, with the multi-method study of Choi, Shin, and Kang (2021) finding that “respondents’ perception of journalistic values is the strongest predictor of their news quality evaluation” (96). In a similar vein, others have suggested that citing scientific sources and using statistical information (Henke, Leissner, and Möhring 2020) as well as quoting both official and non-official actors in reporting can enhance the perceived credibility of news reporting (Cozma 2006). Recent studies have also started analyzing diversity as received in the context of news recommenders. Using an experiment on diverse news recommenders presenting either politically diverse or one-sided content, Heitz etf al. (2022) found that exposure to diverse content resulted in more tolerant attitudes among participants, as well as a bigger preference for news with different views. Despite indications that it might be the case, thus far, no (experimental) study has specifically assessed the effects of viewpoint diversity on the perceived credibility of news content, despite its ascribed relevance and importance by media regulators and scholars alike.

The latter points are crucial as scholarship has argued that citizens’ trust in traditional news media is a major predictor of their news consumption and that, as such, trust represents another important component in maintaining the democratic status quo (Kalogeropoulos et al. 2019). Credibility has been used somewhat interchangeably with trust in relevant literature, with a popular classification distinguishing source, medium, and message credibility (Metzger, Hartsell, and Flanagin 2020). In this study, we are mostly interested in observing the credibility of news content from the perspective of audiences as news consumers (see also Naab et al. 2020). Therefore, we focus on studying message credibility, previously defined as “an individual’s judgment of the veracity of the content of communication” (Appelman and Sundar 2016, 63), and explicitly link this to the diversity of viewpoints and how the presence or absence thereof affects audiences’ credibility of news content. The findings from our literature overview lead us to formulate the following research questions for our user experiment study:

RQ1: How aware are audiences of the degree of viewpoint diversity in news content?

RQ2: Does more viewpoint diversity in news content lead to higher credibility perceptions than content with less viewpoint diversity?

Methodological Framework

To test the existence of a causal relationship between viewpoint diversity and perceived diversity and credibility, we conducted an online survey experiment with 208 Flemish (Belgian) citizens between 18 and 65 years old. The participants were recruited through online advertisements and were randomly assigned to fictitious news items on one of two politically relevant, but not main news topics at the time. These topics were deliberately selected to have varying levels of controversy: one focused on socio-cultural aspects related to what was termed the “Congo Commission,” and the other centred around socio-economic factors, specifically focused on Airbnb. The first article covers a controversial discussion held in the Belgian parliamentary committee about colonial reparation payments of the Belgian government towards its former African colony. The article discusses differences in opinion between commission members from different political parties. The second article covers the decrease in the supply of Airbnb rentals in Belgium because of a new fiscal policy. In 2022, the company had to inform the Belgian tax authorities for the first time about what the landlords have earned, and the article covers the discussion around the desirability of these measures. No source was provided with the news items, to avoid an influence of the news platform on our results. The full basic versions of both news items that were kept identical across conditions are available in Appendix A.

The length and format of the articles were kept as similar as possible. The actors that were discussed in the articles were also identical to exclude the influence of actor diversity. In the Congo Commission articles, both quoted actors were members of the commission, in the Airbnb articles, both actors were landlords. Participants were randomly assigned to the following four conditions:

1. Control condition: an article without explicit viewpoints;

2. Against-condition: a non-diverse article, an article with only a point of view against;

3. In favour-condition: a non-diverse article, an article with only a point of view in favour;

4. Diverse condition: a diverse article, an article with both a viewpoint in favour and against

Data was collected in April 2022 through the survey programme Qualtrics. In total, 224 respondents completed the survey. However, as exposure to the experimental news item was important, all respondents who spent less than fifteen seconds on the page with the experimental stimulus were removed, resulting in a final sample of 208 respondents. Before the respondents were exposed to the experimental stimuli, their knowledge of the two news topics from the experiment was tested using three knowledge questions per subject (based on the measurement by van der Wurff 2011). The questions were kept simple, yet sufficiently specific, ensuring that respondents who closely followed the news on the subject could answer them correctly. There were four answer options, only one of which was correct. Next, we inquired about people’s pre-existing viewpoints on the issue by asking how much they agreed or disagreed with six opinion statements on the issues on a 5-point scale going from completely disagree (1) to completely agree (5). The specific question wording can be found in Appendix B.

Our main dependent variable was news item credibility. This was measured using the news credibility scale of Gaziano and McGrath (1986), e.g., The article was: “fair”, “biased” (mirrored), “reliable”, measured on a 5-point scale going from completely disagree to completely agree; Cronbach’s alpha = .89. Secondly, we were interested in respondent’s awareness of viewpoint diversity in an article. This was measured with the item “The article portrays a diversity of viewpoints”, going from completely disagree (1) to completely agree (5).

After exposure to the stimulus news items and the main dependent variables, we measured respondents’ age (in years), gender, level of education, and political interest (How interested are you in politics in general; not interested at all (1)—very interested (5)). On average, the respondents were 36 years old (M = 35.6, SD = 14.0; min = 19, max = 65), 59.6% were female, and the majority had followed higher education (primary school = 0.5%; secondary education = 16.3%; higher professional education = 38.9%, higher academic education = 44.2%). Political interest was moderate with an average of 2.7 (SD = 1.1) on a 5-point scale.

We tested whether the topics were indeed considered more politically contested (Congo Commission) versus more moderate (Airbnb) by the respondents. On average, people agreed a bit more with the statements in favour of the goal of the Congo Commission (M = 3.67, SD = 0.87) compared to Airbnb’s new fiscal rules (M = 3.52, SD = 0.64). The variance is lower for the Airbnb issue, so respondents’ views on Airbnb are therefore more closely concentrated around the average than at the Congo Commission. Or, in other words, responses for the Congo Commission were more diverse. This is also displayed by the scores on the three knowledge questions we asked for each issue (M _{knowledge Congo}= 2.32, SD = 0.64; M _{knowledge Airbnb} = 1.51, SD = 0.64). It thus seems that public opinion on the Congo Commission is more divided, which serves as some sort of indication of how politicised the topic is.

Results

Diversity Perceptions

First, we tested whether people “saw” diversity in news content in the first place. Did people notice when different viewpoints were present in a news item? And were they aware when only one side of the issue was covered? To test this, we queried the degree to which people thought the article portrayed a diversity of opinions (Figure 1). For both topics, we find a similar pattern: respondents recognised an article containing diverse viewpoints as being diverse, as this condition has a significantly higher diversity score compared to the other conditions; F(3, 410) = 63.17, p < 0.001. When an article was not diverse and explicitly contained only one viewpoint, there was no difference with the control condition containing no viewpoints. Hence, people were able to recognise viewpoint diversity in an article. However, when an article was one-sided, people did not consider it less diverse than an article containing no viewpoints.

Figure 1. Diversity perceptions across various experimental conditions (Higher scores indicate a greater perception of diversity).

News Item Credibility

Congo Commission

Our main interest is how viewpoint diversity influenced audiences’ credibility perceptions. Table 1 shows the results for the credibility scores of the politically sensitive article on the Congo Commission based on the experimental conditions (with the control article as a reference category) in model I. We see that the article giving viewpoints against the Congo commission has the lowest credibility score. People who saw a news item with opinions against the Congo Commission rated the article as less credible than the article without opinions. We do not find such a difference for the condition in favour or the diverse condition. Regarding the control variables, we find that the more people are in favour of the goals of the Congo Commission, the more credible they judge the article to be. Looking at model II including the interactions of the experimental stimuli and people’s opinions about the issue, we find a significant interaction effect of the condition against and people’s own opinions on the topic.

Table 1. Regression models with news item credibility as dependent variable (Congo Commission).

	Model I		Model II
	B (SE)	Beta	B (SE)	Beta
Gender (men = referent)	−0.126 (0.081)	−0.109	−0.127 (0.078)	−0.11
Age	−0.001 (0.003)	−0.027	−0.001 (0.003)	−0.026
Education	0.008 (0.054)	0.011	0.002 (0.052)	0.003
Political interest	−0.02 (0.048)	−0.037	−0.013 (0.046)	−0.024
Interest in political news	−0.014 (0.040)	−0.032	−0.010 (0.039)	−0.021
Issue knowledge	−0.018 (0.060)	−0.021	−0.007 (0.058)	−0.007
Opinion on the issue (=in favour)	0.181*** (0.049)	0.276	0.265** (0.083)	0.404
Experimental stimuli (control = referent)
Condition against	−0.324** (0.110)	−0.246	−0.324** (0.520)	1.112
Condition in favour	−0.137 (0.107)	−0.108	−0.137 (0.404)	−0.292
Condition diverse	0.097 (0.109)	0.074	0.097 (0.446)	0.525
Interaction effects (control = referent)
Condition against * issue opinion			−0.475*** (0.135)	−1,385
Condition in favour * issue opinion			0.070 (0.106)	0.206
Condition diverse * issue opinion			−0.159 (0.117)	−0.459

R² Model I = 0.157; R² Model II = 0.235.

Figure 2 visualises this interaction effect. Here we see that higher support of the Congo Commission resulted in a higher perceived credibility of the article for the control, in favour, and diverse condition. In the against condition, we find the opposite effect. The stronger respondents’ own opinion was against the Congo Commission, the more credible they found the article with the arguments against the Congo Commission, i.e., the ones in line with their own viewpoint. The diverse condition was particularly interesting: even though it was perceived as being more credible than the article going against people’s own opinion, it still was perceived as slightly less credible than the article containing only a viewpoint in the direction of their own viewpoint. So merely including a viewpoint against one's own beliefs, even if people’s own viewpoint is also represented, reduces an article's credibility. Put more positively, a more diverse article slightly increased an article's credibility compared to articles going against one’s viewpoint.

Figure 2. Interaction effect between experimental condition (viewpoint diversity) and pre-existing opinions on the issue on news item credibility.

Airbnb

Our second, less “controversial” issue covered new fiscal rules for landlords on Airbnb in Belgium. Looking at Model I with the main effects in Table 2 below, we see that both stimulus articles containing only one viewpoint were judged as being less credible than the diverse and neutral articles. However, when the interaction terms were added to the model (Model II), the main effects of the conditions disappeared. Although the interaction terms seemed to have a relatively strong effect, they failed to reach significance. We did, however, see a similar trend as with the article on the Congo Commission; people’s credibility judgments were higher when the opinion in the article corresponded to their own opinion on the issue. The fact that we found less pronounced results for this issue may be due to the fact that people had weaker opinions about the more technical Airbnb fiscal topic compared to the issue of the Congo Commission and the proposed reparation payments for Belgium’s colonial past. The latter is more morally charged and likely to elicit stronger opinions from people.

Table 2. Regression models with news item credibility as dependent variable (Airbnb).

	Model I		Model II
	B (SE)	Beta	B (SE)	Beta
Gender (men = referent)	−0.110 (0.083)	−0.094	−0.107(0.082)	−0.091
Age	−0.001(0.003)	−0.029	−0.001(0.003)	−0.029
Education	0.063(0.056)	0.081	0.047(0.056)	0.061
Political interest	0.025(0.051)	0.045	0.023(0.051)	0.041
Interest in political news	−0.012(0.042)	−0.026	−0.015(0.043)	−0.032
Issue knowledge	−0.109(0.062)	−0.121	−0.101(0.062)	−0.112
Opinion on the issue (=in favour)	−0.051(0.066)	−0.056	0.078(0.140)	0.086
Experimental stimuli (control = referent)
Condition against	−0.381*** (0.115)	−0.283	0.655(0.674)	0.487
Condition in favour	−0.333*** (0.112)	−0.258	0.545(0.623)	0.422
Condition diverse	−0.020 (0.114)	−0.015	−0.476(0.702)	−0.356
Interaction effects (control = referent)
Condition against * issue opinion			−0.292(0.187)	−0.789
Condition in favour * issue opinion			−0.252(0.176)	−0.685
Condition diverse * issue opinion			0.125(0.194)	0.342

R² Model I = 0.107; R² Model II = 0.141.

Discussion

In this paper, we have operationalised an online experiment among 208 Flemish (Belgian) citizens to harmonise three different strands of relevant literature on contemporary news journalism. First, the importance ascribed to diversity in its many shapes and forms as the underlying principle of news media as the independent fourth estate, but which as a concept remains opaque and predominantly in the eye of the beholder, scholar or regulator (Hendrickx et al. 2020; Joris et al. 2020; Magin et al. 2023). Second, the rising academic interest for citizens’ ways of engaging with news in the digital era and the awareness of online audiences’ agency when selecting which media content to consume (Costera Meijer 2020; Swart et al. 2022). Third, the trustworthiness and credibility of news journalism and the many factors that have the power to influence these significantly (Cozma 2006; Henke, Leissner, and Möhring 2020; van der Wurff 2011), with viewpoint diversity thus far predominantly absent in the studied factors. This study aimed to fill gaps in prior research by investigating whether audiences can identify a diversity of viewpoints and whether they perceive such diversity as credible in an experimental context.

Combining insights from these disparate bodies of literature, we constructed a conceptual and methodological framework that sought to assess whether audiences are aware of viewpoint diversity in news content in the first place and whether a high degree of viewpoint diversity in news content leads to higher credibility perceptions as opposed to content with less viewpoint diversity. We exposed a sample of adult citizens to fictitious news articles on two different topics (the Congo Commission as discussed in the Belgian parliament and tax authorities cracking down on Belgian Airbnb landlords) and randomly assigned the survey respondents to four conditions ranging in viewpoint diversity, also taking into account the direction of the opinions provided in the articles.

Among our principal findings, our study establishes a significant relationship between high viewpoint diversity in news content and readers’ awareness of this diversity. When only one viewpoint was present, people did not perceive differences in diversity with articles containing no explicit viewpoints. The news topic itself did not affect this awareness in either way. Put succinctly, and to answer RQ1, we find that audiences seem to be aware of the presence of viewpoint diversity, but much less so of its absence. We cannot definitely explain this finding. Relating back to the work of van der Wurff (2011), we highlight his exploratory study’s finding that across audiovisual and print media “the reception of diversity did not vary” (339), thus not concluding that our experiment with written articles could be considered the main culprit. One possible explanation, relevant for Flanders specifically, could be that scholarship has indicated how news coverage often is not very diverse in terms of actors or viewpoints (Walgrave and Kuypers 2021) even though Flemish audiences have an above-average trust in journalism as well as high levels of journalistic professionalism and norm adherence such as providing reliable information and acting as a watchdog (Truyens and Picone 2021, 269). News generally is dominated by elite perspectives (Carlson 2009; Hopmann and Shehata 2011) and often does not present a diversity of views within a single news item (e.g., Beckers and Moy 2023; Hopmann, Van Aelst, and Legnante 2012). It consequently might be the case that one-sided news does not stand out to audiences, while diverse news does.

While previous research (e.g., Cozma 2006; Henke et al. 2020) suggests that increasing viewpoint diversity in news content would automatically enhance its perceived credibility, our findings thus indicate a more nuanced reality (RQ2). Particularly for the more highly polarised topic, our study reveals that pre-existing attitudes on the issue influence which opinions readers deem credible, aligning with prior research on selective exposure and confirmation bias (Garrett and Stroud 2014; Kim 2015). Hence, audience behaviour does not always align with normative journalistic ideals. Nonetheless, our research also highlights that even in such scenarios, news content with greater viewpoint diversity is still perceived as more credible than content containing viewpoints contradicting one's own perspective.

These findings have significant implications for journalists. Although the inclusion of a viewpoint that goes against one's own opinion in a diverse news item results in lower credibility perceptions, people still perceive such articles as more credible compared to news items that only give voice to one-sided actors, especially if those actors contradict their own point of view. Therefore, it is essential for the overall credibility of news coverage to incorporate diverse opinions, particularly when dealing with more polarised topics in public opinion. Only in this way can diverse news reporting continue to fulfil its important remit in democratic society, which is informing citizens and enabling them to take part in said society. Even though people might consume news for different reasons (e.g., emotional appeal or entertainment value; Knobloch-Westerwick 2014), research has consistently shown that credibility is one of the primary factors influencing individuals’ decisions about which news sources to consume (Fletcher and Park 2017; Nelson and Kim 2021). As argued in our conceptual framework, the normative notion of news diversity tends to be routinely linked to similarly normative understandings such as neutrality and civility (Loecherbach et al. 2020; Magin et al. 2023). This only cements our point that the relationship between the need for quality, diverse journalistic work remains crucial to uphold news and viewpoint diversity.

Concluding Remarks

In the first part of the twenty-first century, journalism and wider communication research have seen a dramatic rise in research devoted to the concept of diversity, with a vast and still rapidly expanding body of literature assessing predominantly production and content diversity through field studies and content analysis (Beckers et al. 2017; Hendrickx and Picone 2020; Sjøvaag 2014; Vogler, Udris, and Eisenegger 2020). This study has expanded this area of scholarly work by incorporating viewpoint diversity as perceived by news consumers.

As with all studies, we acknowledge that our study has its limitations. First, our experiment was conducted in the small media market of Flanders (Belgium). However, we believe that the mechanisms behind credibility perceptions are universal, and previous studies have established the effect of counterattitudinal information on audiences’ unfavourable perceptions and judgments of news in various contexts (Garrett and Stroud 2014). Second, our “diverse” viewpoint conditions consisted of two clearly opposing viewpoints in one news story. However, striving for balance within individual news items might not always be advantageous. This is because it could lead to a situation referred to as “false equivalency” of opinions (Jamieson and Waldman 2003), where opposing viewpoints might not accurately represent public opinion in all cases. Moreover, viewpoint diversity can extend to more intricate and nuanced forms, for instance by having diverse frames in one news item (Huang 2009). Third, we focused solely on viewpoint diversity in our experiment while keeping actor diversity constant. Nevertheless, prior research has shown that actor and viewpoint diversity are highly correlated and that both are seen as important indicators of news content diversity (Masini et al. 2018). Therefore, researchers could investigate the relationship between these two factors and audiences’ judgments and perceptions of the news.

Despite its limitations, we strongly believe our study and its findings can impact and advance the scholarly debate on the normative need for news diversity as well as the crucial role audiences play in selecting and engaging with news content in different ways. We integrate our findings to conclude that researchers and policymakers can only advance the ongoing important debate on news diversity if they choose to pay more attention to how audiences (do not) engage with it. Only by further advancing the audience turn in journalism (Costera Meijer 2020; Swart et al. 2022) can we as scholars become more aware of how important viewpoint and other types of news diversity are of particular importance to individual citizens and collective news audiences alike. This again highlights the importance of the holistic view on news diversity with various facets and, in a similar vein, necessitates a multifocal perspective on the audience’s perceptions and relationship with diversity (Hendrickx et al. 2020; Napoli 1999; Sjøvaag 2016). We thus urge fellow scholars to continue to (a) experiment (pun intended) with novel frameworks and studies to enlighten scientific knowledge on what pushes and pulls (perceived) diversity and credibility, and (b) to critically and consistently consider studied effects of changes in production and/or content diversity on the rippling effects they mean for the diversity of news consumption and the perception of news quality and credibility. In order to achieve this, a better conceptual understanding of these concepts and their linkage remains desirable.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Appendices

Appendix A. Experimental Stimuli (Control Conditions)

Discussion About Reparations Puts Congo Commission under Time Pressure

The Special Parliamentary Committee Congo—Colonial Past (Congo Commission) has asked for an extension of its deadline. Due to the continuing disagreement between members about the reparations, the committee wants to be given time to conduct additional research.

The Federal Parliament established the Congo Commission in June 2020 in response to the Black Lives Matter protests. With three months to go, it's time to consider the final leg of her mission. With the establishment of the commission, the Federal Parliament wanted not only a description of the colonial past but also a clear step-by-step plan for the future.

Possible practical recommendations for the future could range from anti-racism plans for education to reparation payments. And although each option has pros and cons, it is mainly the latter that gets the committee stuck. In concrete terms, reparations mean that the Belgian government would pay financial compensation to the descendants of colonised persons in their own country and the former colonies. The committee's expert report already showed that they too did not agree on the importance of reparations.

Now that difference of opinion continues among the other members of the committee. For example, according to the official statement, there are discussions about the role of reparations throughout history. There is also uncertainty about who should bear the financial responsibility for the suffering caused. In any case, it is still awaiting the response of the House to find out whether the committee will be given extra time to investigate this

Supply of Airbnb Belgium Decreases by Twenty Percent

The supply of Belgian holiday homes on the Airbnb platform has decreased by twenty percent. That was announced by the platform itself.

Since the start of this year, the offer on the holiday rental platform Airbnb has fallen by twenty percent. This is evident from figures released by the platform itself today. In the survey that Airbnb puts out to every departing landlord, nine out of ten indicated that “the changed tax legislation” was the reason for their departure.

Since the beginning of this year, there are some new tax obligations for Airbnb. For example, in March 2022, the company had to inform the tax authorities for the first time about what the landlords have earned over the past year thanks to the platform. Landlords who offer extra services, such as a breakfast or cleaning service, will also pay six percent VAT on their income from 1 July.

Whether Airbnb plans to take further action is still unknown. Since the law has only just been enacted, it seems unlikely that lobbying would get them to change it. However, recognition as an official platform for the sharing economy would allow the platform to benefit from a more favourable tax regime. But the company has so far always refused that recognition.

Appendix B. Question Wording Knowledge and Opinion Questions Regarding the Experimental Topics

Knowledge Questions

The survey will focus on two topics, namely the Congo Commission and Airbnb. The questions below only serve to get a general idea of how informed Flemish people are about the subjects. It is no problem if you give a wrong answer.

What is the Congo Commission?

1. A special committee in the Belgian federal parliament.

2. A special committee in the Flemish Parliament.

3. An association made up of people who love Congolese culture.

4. An association consisting of Belgians with Congolese roots.

What is the Aim of the Congo Commission?

1. Preserving and spreading Congolese culture.

2. Investigate Belgium's colonial past and propose measures for the future.

3. Set agenda items for talks between King Philippe and Congolese statesmen during his visit to Congo.

4. Standing up for the rights of people with Congolese roots in Belgium.

What are Reparations?

1. Payment obligations that may be imposed on the attacking party to a conflict for all damage caused to the attacked party.

2. The costs associated with the restoration of art destroyed during the colonisation of a country.

3. The total amount of money needed to pay for the material damage caused during the colonisation.

4. The amount that the Belgian royal family annually donates to charities in Congo.

What is Airbnb?

1. An online marketplace for the rental and booking of accommodations from private individuals, hotels, and investors.

2. An online network where members can offer a free place to sleep in their homes to other members.

3. An online community for bed & breakfast operators.

4. An international website where participants can add reviews and photos of cities and places of interest.

What New Tax Legislation Must Airbnb Follow as of this Year?

1. Make information about the total revenue of the platform known to the Belgian government.

2. Make information about the income that users earn through the platform known to the Belgian government.

3. Make information about the total income of the platform known to the Flemish government.

4. Make information about the income that users earn via the platform known to the Flemish government.

What New Tax Legislation Must Users on Airbnb Follow since this Year?

1. All users must pay six percent VAT on the revenue they receive.

2. Users must pay six percent VAT on the amount they pay for their membership of the website.

3. Users must pay six percent VAT when they book a stay through the platform.

4. Users must pay six percent VAT on their earnings if they offer additional services, such as a breakfast or cleaning service.

Pre-existing Opinions

Please indicate how much you agree or disagree with the following statements (1 = completely disagree, 5 = completely agree).

Congo Commission

“Belgium must take measures to repair the suffering caused in Congo.”; “Belgium must pay financial compensation for the colonisation of Congo.”; “The Belgian government is not responsible for the colonisation of Congo.”; “Reparations to the relatives of colonised people are useless.”; “The Congo Commission is a bad initiative.”; “Countries that have colonised other populations must admit their mistakes.”

Airbnb

“Taxes are a good way to redistribute wealth in society.”; “Landlords should not earn more on Airbnb than on the regular rental market.”; “The government should interfere as little as possible in the operation of the free market.”; “Hosts on Airbnb should not be subject to stricter government regulations.”; “International companies must follow the same tax rules as other companies in the country.”; “The government is violating the privacy of Airbnb hosts by requesting their income.”