Reach or Trust Optimisation? A Citizen Trust Analysis in the Flemish Public Broadcaster VRT

In democracies, one of Public Service Media’s (PSM) main roles is to inform the public. In a digital news ecosystem, where commercial, citizen, and alternative news sources have multiplied, questions about the ability and need for PSM to fulfil this role are increasingly being raised. While the role of PSM can and should be scrutinized, a too-narrow a focus on an informed citizenry may obfuscate aspects, other than audience reach and objectivity, that are key to this information role, such as trust. Against this background, this article studies whether and to what extent citizens still trust the news and information services of their public broadcaster, asking if that trust is still high, whether there is a difference between groups in the population, and if trust is in line with reach. Based on a representative survey of news users in Flanders, the Dutch-speaking community of Belgium, the article studies the reach and trust scores of the brands of VRT, Flanders’ PSM, and compares them to those of its main competitors, with a specific focus on differences in terms of age, education levels, and political orientation. The results suggest that VRT struggles more than the main commercial players to reach young people and the lower-educated, but still leads when it comes to trust. The data show the continued importance of widening our assessment of PSM beyond market-focused indicators of reach.


Introduction
Public Service Media (PSM) can be defined as the provision of media services, across devices and platforms, that contributes to the democratic, cultural and social wellbeing of society (see contributions in Lowe & Martin, 2013;Lowe, Van den Bulck, & Donders, 2018). Often, the delivery of such services is entrusted to public broadcasters. For all of them, strengthening informed citizenship is an, if not their most, important task. PSM has a "responsibility for the health of the political process and the quality of public discourse generated with it" (Blumler, as cited in Hesmondhalgh, 2019, pp. 152). Having said that, four trends have hindered public broadcasters in their work to inform the citizenry.
First, the online environment results in lower reach of public broadcasters' news services and, particularly with younger audiences, as well as encouraging the rapid consumption of news (Cola & Prario, 2012). Second, the rise of disinformation comes with lower trust in traditional media outlets (Fletcher & Park, 2017). The line between what is true and what is false seems to have become more blurred and public broadcasters are finding it difficult to position themselves in a 'post-truth society' (Gibson, 2018); even if that latter concept has been rightfully criticised by some (e.g., Fuller, 2018). Third, and re-lated to the former, a rise of populism in a multitude of EU Member States puts additional pressure on legacy news media and allegedly 'leftist' and progressive public broadcasters (Wettstein, Esser, Schulz, Wirz, & Wirth, 2018). In some countries such as Poland and Hungary, the re-balancing of powers and the strive for pluralism between conservative and progressive, left-and rightwing ideologies, etc., has resulted in a complete political capture of the PSM system (Reporters Without Borders, 2019). In other countries, it at least creates the possibility of self-censorship with journalists who fear #lügenpresse, #MSM, #fakenews, etc. Fourth, most European public broadcasters have faced budget cuts over the last few years, making it more difficult to maintain investment in journalism (European Broadcasting Union, 2018).
Against this background, we studied whether and to what extent citizens still trust the news and information services of their public broadcaster, asking if that trust is still high, whether there is a difference between groups in the population, and if trust is in line with reach.
The focus of our analysis is Flanders, the Dutchspeaking region of Belgium with 6.5 million inhabitants. The Flemish media market is rather concentrated with only four (cross)-media companies supplying a significant portion of the audience with news: DPGMedia, Mediahuis, Roularta, and VRT. DPG Media has been particularly wide-ranging, covering newspapers, magazines, online brands, television and radio. Mediahuis and Roularta are mainly active in print and online. VRT offers services on radio, television, and online. DPGMedia leads in print and online news, followed by Mediahuis, while VRT has the highest market share in radio and television (Vlaamse Regulator voor de Media, 2020). There are smaller online news sites, such as Doorbraak.be, SCEPTR and Apache, but these are niche publications. The first two are more partisan right-wing outlets although they do impact public opinion, whereas the latter has greater impact on political debate.
Public service broadcasting takes a key position in both the French and Dutch speaking communities, which are autonomous in their decision-making on PSM. While both can be seen as democratic corporatist systems (Hallin & Mancini, 2004), there are notable differences though with more pressure in Flanders from commercial competitors and right-wing politicians to limit VRT in its activities, specifically online. That has also resulted in declining government funding since 2007, the rejection of a pre-school children's channel after an ex ante evaluation, and a Government Agreement in 2019 which mentioned the Flemish Government's intentions to limit the amount of text in online news. Remarkable according to some (Donders, Van den Bulck, & Raats, 2019) given that the Flemish media market is so concentrated and the public broadcaster thus adds to not only internal but also external pluralism.
While applauded for its high-quality service delivery, criticism from the main right-wing parties for its alleged political bias as well as for distorting the online market is on the rise (Donders, Van den Bulck, & Raats, 2018). It is not clear whether this trend in opinion is also likely to be observed among audiences. Based on a representative survey among Flemish citizens, we find high levels of trust in the Flemish public broadcaster VRT, although having that said, results are slightly different for young people and right-wing voters. The article consists of five parts. First, we formulate a problem statement and key research questions in the introduction. Second, we theorise what task public broadcasters have in the area of news and information, relating this to the need for citizens' trust in the impartiality of public broadcasters. Third, we explain the methodology underlying the representative survey, then we present our findings, and finally, we outline our conclusions and reflect on the importance of trust in PSM.

PSM and Informed Citizenship
Information provision has been a core task of public broadcasters ever since their creation in the 1930s (Price & Raboy, 2003). Public broadcasters should inform citizenships, confront people with different viewpoints, and in so doing strengthen democracy (Van den Bulck, 2016). Political citizenship is thus not only about being informed, but also about having access to different interpretative frameworks and deliberative fora where information can be discussed and evaluated. Political citizenship requires pluralism, which is not the same as an abundance of content. Structural pluralism, so Beata Klimkiwiecz (2010, p. 907) says, "refers to a condition where diverse, independent media entities exist within a system and are arranged together in a particular way." Allen, Connolly, and Hargreaves Heap (2017, p. 47) see media pluralism as "an essential pillar in the right to information and freedom of expression." It requires the representation of all relevant opinions and the potential for citizens to engage in a debate rather than acting as mere spectators. Citizens should be part of a process of interaction and genuine dialogue within some sort of public sphere (Habermas, 1991) and this should not be eroded to the point where news is only consumed (Scannell, 1995, pp. 23-24;see also McQuail, 1998, p. 140).
As maintained by Picard and Pickard (2018, p. 16) "a healthy democracy requires opportunities for citizens to deliberate in public spaces that are largely independent of state and market forces." The European understanding of media pluralism as a necessary condition for political citizenship goes beyond rejecting government control over media; it extends to avoiding commercial interests from becoming so overly dominant that they inhibit the free, pluralistic exchange of media services (Czepek, Hellwig, & Novak, 2009). That opinion diverges significantly from the US model where the freedom of individual media owners is placed above the equal right of all citizens to receive information as well as to express their opinions (Humphreys, 1996). While the 'free marketplace of ideas' has not delivered-look for example at the low voter turnouts-also several EU Member States have adopted media policies that are inspired by libertarian ideas.
There is disagreement on whether public broadcasters have contributed to informed citizenship. Research indeed shows higher levels of current affairs knowledge among citizens who access strong public broadcasting systems (Soroka et al., 2013). People watching public broadcasters learn more about domestic and international affairs than those watching commercial news. Factors such as independence, adequate public funding, and audience share are relevant factors in determining levels of hard news in particular (Soroka et al., 2013).
Some scholars have been more critical though, pointing to public broadcasters' insufficient investment in investigative journalism (Cordell, 2009) and their lack of criticism of ruling parties which largely results from funding issues and/or institutional weakness (Stetka & Örnebring, 2013). Furthermore, they have been criticised for reporting in overly dramatic manners without paying adequate attention to the historical context of serious issues such as the financial crisis in 2008/2009 (Berry, 2013), as well as for plainly being mouthpieces of government (Mungiu-Pippidi, 2006). Public broadcasters, some research shows, have gone along with the polarization of public discourse, without critically questioning statements made by politicians (e.g., related to the Brexit referendum, see Cushion & Lewis, 2017), or have failed to represent all opinions in society, specifically those of ethnic minorities (Panis, Paulussen, & Dhoest, 2019). Some also argue that there has been a heavy emphasis on one-directional knowledge transfer (Bardoel & Lowe, 2007) rather than a two-way understanding of political processes, current affairs, and events in society. Overall, one could conclude that public broadcasters in Western and Northern European countries have demonstrably contributed to informed citizenship while being imperfect in delivering that objective in several ways.

Informed Citizenship and Trust
The discussion on the role of PSM has intensified in a context of internationalisation, further commodification of media users, and disinformation. While some argue that the role of public broadcasters becomes more important (Ramsey, 2018), some market failure thinking on PSM is on the rise again. The questioning of the BBC license fee in the UK by conservative politicians is a case in point. They point at the digital environment which has rendered the BBC a dinosaur as well as at its allegedly biased reporting.
Underlying these various contesting views on the performance of PSM is the ideal of an informed citizenry, one that consists of rational, information-seeking, politically engaged citizens-which is still the default view amongst media professionals (Graves, 2017(Graves, , p. 1242. Indeed, as is clear from the above, this view is dominant in newsrooms and with policymakers and academics as well, even if it has been contested before (Graeff, 2019). Will this view hold against the background of a 'global democratic recession' during which we witness, among others, the growing popularity of right-wing nationalist parties, a severe EU political crisis, and the unnerving political theatre of the US president (Graves, 2017(Graves, , p. 1239? The sense of crisis revolves around the erosion of accountability in the media system. Van Aelst et al. (2017, p. 12) consider that the rise of partisan media forms a key challenge for the political information environment as it leads to "opportunity structures for selective exposure based on political attitudes and beliefs" rather than on factual information. In its most radical articulation, this means that facts and truth do not matter, and the current distributed media ecosystem seems to further amplify those who pursue 'post-truth politics' (Suiter, 2016) which challenges the democratic public sphere and promotes societal conflict.
If facts do not matter, then what are the benefits of an informed citizenry? Such a context forces us to revisit our understanding of information in relation to citizenship. An interesting case concerns fact-checking as a solution to false information (Nieminen & Rapeli, 2018). While still inconclusive, various studies seem to suggest that fact-checking might increase factual knowledge, but that this does not necessarily affect citizens' ideological beliefs and political choices (see Tucker et al., 2018). In the US, for example, when the mainstream media went to great lengths to challenge conspiracy reports that linked Hillary Clinton to a child trafficking ring run from a pizza shop, the many people who distrust 'liberal' media saw this rather as a confirmation that there was something to investigate (Boyd, 2017).
At this point, trust enters the equation. As Strömbäck et al. (2020) suggest, informative news media will hardly lead to the democratic ideal of an informed citizenry if citizens do not consume or do not trust the news. We might thereby tend to think that the consumption of a certain news source implies the source is trusted. The authors are however keen to stress the complexity of the relationship between news trust and news use: Overall previous research suggests that media trust is associated with greater use of news media while media distrust is associated with greater use of nonmainstream news sources, but that the relationship between media trust and media use is quite modest. (Strömbäck et al., 2020, p. 8) On top of that, news use is not always an instrumental practice driven by a rational, informative and selective orientation towards the media, but can also be more ritual, driven by distraction, affection and habits (Rubin, 2002, pp. 534-535). This can even lead to people using news they do not trust, e.g., when their need for cog-nition is higher than their distrust in the news (Tsfati & Cappella, 2005) or their orientation towards convenience leads them to choose less trusted over more trusted news (Jarvis, Stroud, & Gilliland, 2009).
What is clear though, is that news use and news trust do not necessarily match. In other words, reach is not necessarily an indicator of trust. Exactly this differentiation is often lacking in the way news organisations, PSM included, assess their goals. Traditionally, VRT in Flanders, in line with other PSMs in Western democracies, has been required through their charter agreements with the subsidising governments, to cater to a wide audience (e.g., Vlaamse Regering & VRT, 2011, 2015. The idea that a PSM organisation should be there for all citizens has been a leeway for reach to become a central measurement to assess a PSM's success, much in line with commercial logics (Donders, 2012). In the pressured news ecosystem described above, this focus on reach, further emphasized by editorial analytics entering the newsroom, has led to critical scrutiny.
Journalism scholar Jay Rosen (2018) put it well when he wondered what a news organization would look like if it were not optimized for reach, but for trust? Admittedly, optimising for trust is not without problems; alternative media, populist politicians and fearmongers on any side of the political spectrum have all mastered ways to gain people's trust, regardless of the truth. Rosen himself has addressed this issue by recalibrating his question of how to generate trust by publishing news that still adheres to high standards of verification. It is exactly at this intersection between information, reach and trust that one of journalism's key challenges materialises. As partisan media do not need to balance trust with factuality and commercial media might find it challenging because of stringent market conditions, PSM organisations-at least those who are financially and politically independentare the media actors par excellence to take up the challenge to combine reach with trust.
Coming back to the notion of an informed citizenry, this approach requires us to rethink its centrality in our conception of democratic, factual, and trustworthy news offerings. In her critical assessment of the conception of trust in news media, Fisher (2016, p. 461) points out the issue regarding the normative link between news and informed citizenry: Relying on the assumption that the news consumer will interpret trust based on traditional conceptions of reliability and accuracy bound up in the ideal of the informed citizen, does not adequately accommodate how and why people are accessing news media.
The kind of citizenship she is hinting at is one with "an orientation towards a public world, including politics and broader public issues, beyond matters of purely private concern" (Couldry, Livingstone, & Markham, 2007;p. xv). Such a view on citizenry adheres much more to Schudson's (1995, p. 169) notion of a monitorial citizen: We cannot always be fully informed about what happens in society, but we should be sufficiently informed to recognize possible threats to our personal and collective wellbeing.
If the objective is to have a widely informed public, then reach is an important metric. If the goal is to make sure that when needed, people can turn to relevant information, a degree of trust becomes essential. The question is then: Do citizens still trust PSM organisations? This article will, therefore, look at the performance of PSM organisations in terms of reach and trust, and in doing so, complement a view that 'maximizes on reach' and relates to an 'informed citizenry' with a view that 'maximises on trust' and relates to a 'publicly connected citizenry.'

Operationalising Trust in the News
While research into trust in the media has a long research tradition, a single definition or reliable operationalisation of news trust is lacking in the literature (Fisher, 2016, p. 455). An important part of that conceptual vagueness is due to the notion of 'media' variably referring across studies to media in general, to different media types, to media as institutions, to individual media outlets, to journalists, as well as to the content or topic of media coverage (Strömbäck et al., 2020, p. 9). In an effort to increase conceptual clarity, Strömbäck et al. (2020, p. 10) suggest a central focus on trust in the information coming from news media rather than trust in the media as institutions or in individual news producers.
To a large extent, we adhere to this perspective. In the study, respondents were asked to denote whether they agree or disagree, via a 1-5-point Likert scale, with the following statement: 'I think you can trust most news most of the time.' An equivalent question was asked for trust in the news one consumes, news in social media, and news in search engines, always leaving the concept open to the interpretation of the respondent (see also Fletcher & Park, 2017, p. 1290. However, the argumentation proposed by Strömbäck and his colleagues once more puts central the idea of an informed citizenry that needs verified, reliable, and factual information to play its democratic role. Then indeed, the most important thing is that citizens trust the information in itself. But when considering a publicly connected citizenry, trust in the news providers becomes equally important. Rather than continuously engaging in informationgathering, people are mainly busy living their own lives, and in doing so might often just scan the headlines, while trying to remain alert and ready to respond to news that affects their lives (Graves, 2017(Graves, , p. 1243. In navigating the high choice news environment, they develop epistemological strategies to assess which news to trust. One such strategy is pragmatic trust, where news users confide in specific news sources based on personal experience but also institutional reputation (Schwarzenegger, 2020, p. 371). In this study, we therefore also look at trust in news brands. Respondents were invited to answer the question: 'How trustworthy would you say news from the following brand is?'-on the condition that they had heard of the brand and without differentiating for the device or channel they used to access the brand. A scale from 0 to 10 was used where 0 is 'not at all trustworthy' and 10 is 'completely trustworthy.'

Methodology
In this article, we use the raw data of the 2020 Digital News Report (Newman, Fletcher, Schulz, Andi, & Nielsen, 2020). The survey was conducted by YouGov with respondents from an online panel in January 2020-just before the COVID-19 pandemic hit most countries-across 40 markets spread over six continents. The data set used for this article concerns 980 news users from Flanders (the Dutch-speaking region of Belgium), is representative of the online population and is weighted according to targets on variables such as age, gender, and education.
Our focus of the analysis was mainly on questions relating to reach and trust, limiting the analysis to the brands of VRT and its main commercial counterparts. We break down the results according to three sociodemographic dimensions: age, education level, and political orientation. For age, we choose to differentiate within the younger age groups when presenting the data, as younger people are more likely to have a preference for non-mainstream news sources than older people (Fletcher & Park, 2017, p. 1292. We work with three brackets: 18 to 24 years, 25 to 34 years, and 35 years and older. For education levels, we asked 10 categories, which we recoded into lower education (all levels up to upper secondary education), middle (from upper secondary education to short-cycle tertiary education), and higher education (bachelors, masters and doctorates). For political orientation, we asked respondents to position themselves on a left-right political scale ranging from 'very left-wing' and 'fairly left-wing' to 'slightly left of centre,' 'centre' and 'slightly right of centre,' to 'fairly right-wing' and 'very right-wing,' recoding them accordingly into left, centre, and right. For each of these sociodemographic variables, statistical significance was tested via a One-way ANOVA test for the dependent variable (brand trust scores) and calculated using a Tukey post-hoc test (see Supplementary File Annex 2).
For reach, the question was divided into one for offline and one for online brands. For offline media, the question was formulated so as to encompass all forms of use, including delayed viewing of radio and television; for online media, the question encompassed all channels or devices used, including via apps and social media. The respondents were asked to mark the brands they used three days a week or more. In the analysis, we interpret this as use on a regular basis. For VRT, this means that the following channels were included: the main television channel één, the information and cultural channel Canvas, the information radio channel Radio 1, as well as the popular radio channel Radio 2, the alternative niche radio channel Stubru, and the niche entertainment radio MNM. In the presentation of the data, we grouped the television channels under VRT TV, the generalist radio channels under VRT Radio (broad), and the niche radio channels under VRT Radio (niche).
We benchmark the results against the most popular news brands in Flanders on the one hand and key 'quality' news brands on the other hand. The most popular brands include the television station VTM, the radio station Q Music and the newspaper Het Laatste Nieuws, all owned by DPG Media, and the newspaper Het Nieuwsblad, owned by Mediahuis. The quality news brands include the newspapers De Standaard, owned by Mediahuis, and De Morgen, owned by DPG Media. This gives a similar picture online. For VRT, the general news brand VRT NWS is probed-also in the brand trust question. For the popular media, the online counterparts of Het Laatste Nieuws, Het Nieuwsblad, De Standaard and De Morgen were included. It is important to mention that VTM and Q Music are run based on advertisements only and hence offer their content for free, whereas all the other newspapers here operate some form of hybrid/metered paywall, offering a combination of free and a paid news offerings.

Findings
Before looking specifically at the levels of (self-reported) trust in the public broadcaster amongst Flemish news users, we analyse how the public broadcasters' offline and online channels compare to those of commercial players in terms of use.

Reach of PSM News Is Big but Differs Significantly across Age and Education
Our data show, firstly, that VRT is still widely used for news in Flanders. All VRT channels and brands combined are regularly used by 60% of Flemish news users. Secondly, while VRT is dominant offline, it has not been able to hold that same position online. Its television channels are regularly used by 33% of Flemish news users, slightly more than main commercial channel VTM. Online, however, only 23% of Flemish news users regularly turn to its news site. Here, VRTNWS is outperformed by the online offering of the commercial newspaper Het Laatste Nieuws (39%) and closely followed by Het Nieuwsblad (22%). Thirdly, behind the general numbers lie important differences in terms of the target audiences and competition with commercial partners. Let us zoom in on how VRT's reach for news differs across age, education level, and political orientation.

Age
VRT regularly reaches 52% of news users between 18 and 24 years old (see Table 1 in Supplementary File Annex 1). VRT reaches the young with news mainly through its ded-icated radio stations (29%), Stubru and MNM. Television is clearly a more difficult story. Reach is almost twice as high amongst regular users over 35 years old (38%) as amongst those under 35 years (20%). Is VRT reaching more young people via its website or social media (see Table 2 in Supplementary File Annex 1)? No. Online reach (21%) is only slightly higher than television amongst the 18 to 24 years old. What is maybe most remarkable here is that VRTNWS is performing particularly poorly amongst the 'millennial' age group of 24 to 34 year olds (14%) and has the highest reach in the over 35 age group (25%). So, VRT reaches younger audiences mainly through its dedicated radio stations and has not yet managed to convert its decreasing TV audience into an online audience.
One might, of course, suspect young people to be a tough audience. So how does VRT compare to other popular news brands in this regard? Once more, VRT's niche radios perform well as they reach more young people than any other news brand. Admittedly, we are talking about hourly news updates here, not in-depth coverage. When we turn to television, één (VRT) and VTM (DPGMedia) provide the only two television news bulletins in Flanders, which are comparable in terms of length, frequency, and format. Both provide additional current affairs shows, but VRT offers significantly more of these through its second channel Canvas. Even with both channels combined, VTM (26%) is performing better amongst the youngest age group than VRT (19%). Popular newspaper Het Laatste Nieuws reaches roughly the same amount of 18 to 24 years old (18%) as the public broadcasters' television stations. Online, Het Laatste Nieuws leads the pack (36%) by a large margin, but VRT NWS (21%) is doing a much better job in reaching young people online than the other newspapers' websites (see Table 2 in Supplementary File Annex 1). VTM and Het Laatste Nieuws are also part of the same media group, DPG Media, and since the latter took full ownership of VTM, cross-promotion between the newspaper and the television channel has increased. In short, VRT's main commercial counterpart, DPG Media, is doing a better job at reaching younger audiences both offline and online.

Education
When it comes to education levels (see Table 3 in Supplementary File Annex 1), VRT brands reach more higher-educated (70%) than middle-(49%) or lowereducated (56%) citizens. Whereas VTM (37%) reaches a greater number of lower educated citizens through television than VRT (28%), VRT reaches more of them than the popular newspaper Het Laatste Nieuws (25%).
We see a similar picture online (see Table 4 in Supplementary File Annex 1): VRT NWS (32%) and Het Laatste Nieuws (37%) both reach a significant portion of the higher-educated, but whereas Het Laatste Nieuws reaches a similar percentage of the lower-(41%) and middle-educated (41%), VRT NWS reaches much less of them (18%), being surpassed also by Het Nieuwsblad (20%) amongst these groups. Again, compared to De Standaard (5%) and De Morgen (4%), VRT performs better here. So, VRT reaches fewer lowerand middle-educated people than popular brands but reaches more of them than other quality brands. It thus takes a middle-ground position.

Political Orientation
When looking at political orientation (see Table 5 in Supplementary File Annex 1), we see a rather nuanced view. Overall, VRT brands do not show a great deal of variation between left-, centre-, and right-wing citizens, reaching 65%, 64%, and 58% of each group, respectively. Through its television stations, VRT reaches more citizens on the left (40%) than the right (30%), but this is not the case through its broad radio networks (22% vs. 27%). VTM is even more 'polarised' in terms of the political orientation of its audience (19% vs. 35%), leaning more towards the right. Online (see Table 6 in Supplementary File Annex 1), this discrepancy is almost completely absent in the audience of VRT NWS, especially compared to its competitors.

PSM is the Most Trusted News Brand across Education and Political Orientation
Trust in the news, in general, is relatively high in Flanders, even if a slight downward trend has recently been detected. In 2020, Flanders was ranked 6th amongst the 40 countries surveyed in Newman et al. (2020). When it comes to social media and search engines, on the other hand, Flanders is much more aligned with other countries.
In our sample, when asked whether they agreed that most news can be trusted most of the time, 56% said yes, although when asked the same question about news via search engines and via social media, only 28% and 16% agreed, respectively. Clearly, news obtained through the intermediary of powerful technology companies is less trusted. Only 39% of the 18 to 24 age group agreed that they generally trusted the news compared to 56% of those over 35: a notable difference. Across educational levels, we only see slightly higher trust levels amongst the higher-educated, and regarding political orientation, there are slightly higher trust levels on the left. For trust in social media, one might suspect trust to be higher amongst younger generations as they are more likely to turn to social media for news. Hoverer, only 18% of those under 35 years old agree that news via social media can be trusted most of the time, compared to 15% of the group above 35 years old. Trust and use thus do not always match.
Against that background, VRT NWS succeeds in being the most trusted of the brands in our analysis, with an average score of 7.3 out of 10, outperforming not only the news of main commercial broadcaster VTM (7), newspaper/site Het Laatste Nieuws (6.5), and radio Q Music (6.2), but also main quality newspapers like De Standaard (7) and De Morgen (6.6). The public broadcaster might not always be the most used news brand, but it is still the most trusted. That observation is confirmed also by the high market share of VRT NWS during the Coronavirus crisis (PDE, 2020). On the other hand, the scores of VRT and its main competitors are not that far apart.

Age
When looking at trust in relation to age (see Figure 1), we can conclude that, in general, Flemish news users under 35 years old tend to score news brands lower than those above 35, albeit with some notable exceptions. Also worth noting is that for every brand, the 25 to 35 age group is the least trusting, which could be due to their specific life phase. Still, amongst this age group, as amongst those over 35, the public broadcaster remains the most trusted brand, even holding up against quality newspapers. This is not the case amongst the 18 to 24 year olds. Not only do they trust VRT significantly less than those older than 35, but they also trust VTM more than VRT. VTM combines its higher reach in this age group with a (slightly) higher trust score. And while the use of De Standaard and De Morgen is much lower amongst this age group, the quality news brands resonate in terms of trust. This shows that while young people have much lower trust in the news in general, they do trust the main news brands, including the public broadcaster. It might indicate that young people, in particular, have developed a more cautious way of navigating the news, being critical about news in general, but using a 'traditional' compass which guides them towards the more trusted brands.

Education
We see more remarkable differences when considering education levels (see Figure 2). Higher-educated citizens trust VRT's news radio Radio 1 significantly more than lower-educated citizens, who in turn trust commercial brands like Het Laatste Nieuws and VTM Nieuws more The question was 'How trustworthy would you say news from the following brands is? Please use the scale below, where 0 is "not at all trustworthy" and 10 is "completely trustworthy."' See Table 1 in Supplementary File Annex 2 for exact numbers and statistical significance tests. The question was 'How trustworthy would you say news from the following brands is? Please use the scale below, where 0 is "not at all trustworthy" and 10 is "completely trustworthy."'See Table 2 in Supplementary File Annex 2 for exact numbers and statistical significance tests.
than higher-educated users. This seems to correspond broadly with reach: VRT and DPG Media reach more higher-and lower-educated citizens respectively and are also trusted more amongst those groups. But while trust in brands such as Het Laatste Nieuws and VTM drops significantly amongst the higher-educated, this does not hold for trust in VRT NWS amongst the lower-educated. VRT NWS, and also Radio 2, is much less differentiating in terms of education levels, especially when compared to VTM and Het Laatste Nieuws. Except for its information radio Radio 1, trust in VRT is quite similar across education levels.

Political Orientation
A similar pattern can be seen between people on the right showing slightly lower trust scores across all media, except for popular news brands Het Nieuwsblad en Het Laatste Nieuws, where trust is lower amongst leftwing citizens (Figure 3). In particular, trust in Het Laatste Nieuws is significantly lower amongst left-wing citizens. Trust in the public broadcaster, while still high, is in turn significantly lower amongst people on the right than on the left and in the centre. Here, it is VTM that seems to be the least 'polarising' brand. This right-left divide in trust is in line with the discourse and policies of right-wing political parties in Flanders, which have been critical about the scope and editorial decisions made by VRT. Still, even amongst citizens leaning to the right of the political spectrum, VRT remains the most trusted source amongst the brands covered. Again here, what might be a more accurate description is that trust in VRT NWS amongst rightwing voters is high and particularly high amongst leftwing voters.

Conclusions
Our research shows that VRT still reaches a lot of citizens and that its reach is higher among older and more highly educated people. While being a market leader in radio and television, VRT has less reach with its online news offerings. The popular news brands of DPG Media and Mediahuis lead, with VRT being in the middle, and quality newspapers following with their news sites. VRT reaches fewer young people although, unlike commercial media, the discrepancy between age groups manifests itself mostly in relation to television rather than its online news offers. For radio, channels such as Stubru and MNM seem to be effective means of aiming information at young people. Essentially, data shows that VRT is no exception, and follows the major trends in news consumption that Western and North European PSM are generally confronted with (Schulz, Levy, & Nielsen, 2019). The status of online news differs somewhat with some public broadcasters such as the British BBC and Irish RTE being very competitive with their commercial online news offers. Is the low reach of VRT with young people, compared to tabloid-like brands such as Het Laatste Nieuws and its website hln.be, a problem? Not per se. It might be logical that VRT reaches fewer young people compared to brands with more celebrity news, human interest, etc., as younger audiences, paradoxically, do not necessarily prioritise the consumption of 'serious' news over 'light' news, even if they deem the former more important (Costera Meijer, 2007). As most commercial news websites are increasingly putting their articles behind pay-walls, reach with young people might fall as willingness-to-pay is low within this group (Flamingo, 2019). At the same time, radio brands such as Stubru and MNM are experimenting with online-only content, also in the news domain, and might see their reach increase further. As such, VRT, through its various brands, is still feeding information with alternative perspectives into the high-choice news environment for people to consume at their own pace.
When we look at trust instead of reach, we see a different picture. VRT remains the most trusted news brand, even though most of its commercial counterparts also show high trust levels. The high level of trust in the news The question was 'How trustworthy would you say news from the following brands is? Please use the scale below, where 0 is "not at all trustworthy" and 10 is "completely trustworthy."' See Table 3 in Supplementary File Annex 2 for exact numbers and statistical significance tests. and most news brands in Flanders remains remarkable and might be the result of strong news brands (including those of VRT), the relatively slow take-up of digital news and a lack of big media scandals. Here too, breaking down the numbers shows a more nuanced situation. The more highly educated people have higher trust levels in news overall whereas older people and people with a left-wing orientations trust VRT more than the young and those with right-wing orientations. The latter have lower trust in the news generally, and their trust in VRT is still high. Young people (18 to 24 years old) trust the commercial broadcaster VTM more than VRT. That is somewhat concerning and should be looked into further. Still, if we consider VRT as an organisation that should be optimised for trust rather than reach, then it would definitely meet its goals. This is important when we conceive citizens as publicly connected rather than strictly informed citizens. In that case, the presence of a trusted PSM in the ecosystem remains key, even if not always used. It also opens up the debate on competition. The scope of the PSM should not be to compete with others to attract the largest audience, nor should PSM be evaluated on that. Rather, it should be assessed on trust and how it works to maintain trust.
Even when putting trust first though, challenges can arise. If we look at young people, we see that in terms of trust, VTM scores slightly better. A commercial player, too, can succeed in providing news that is trusted by a younger audience. That offers interesting avenues for hypotheses. The main question here is: would VTM ever have reached this level of trust if it had not been forced to compete with VRT? Or the other way round, would VRT have the reach it has if it had not been forced to take into consideration the more 'user-friendly' approach of VTM?
The study at hand is leading us to question our view on the place of PSM in the news ecosystem. We conceive their impact to be valued, not so much on the level of a single organisation finding an online audience, but of an organisation injecting a specific kind of information and journalistic practice in the system for the whole system to benefit from. Previous studies have linked the presence of PSM to an overall higher degree of trust in both institutions and individuals and have linked the individual use of PSM with higher levels of social trust and trust in the media (Kalogeropoulos, Suiter, Udris, & Eisenegger, 2019, p. 3678). Assessing PSM from the perspective of trust and defining performance accordingly carves out a specific place for PSM in the public domain. This is a place that should be less governed by reaching people with information but by injecting trustworthy information into the ecosystem.
This perspective steps away from an all too marketdriven logic of PSM as broad crowd-pleasers, but on the other hand does not focus on a too informed public either, which would reduce it to a mere hard news information provider. Rather, it puts forward the need for PSM to be a trusted source. In Flanders, despite what right-wing politicians might claim, VRT still is that trusted source amongst its main stakeholders, namely its citizens.
Still, further reflection is needed on how to study and assess this role. Such an assessment will always depend on the indicators used. Increased competition of global tech giants and editorial analytics permeating public and private newsrooms do not offer the most fertile ground for discussing audience metrics that go beyond reach. Indicators valuing trust next to reach require a less information-centric view on citizenship to start with. In order to develop such an indicator, further research will be necessary that also tackles some of the limitations of this study. Firstly, qualitative research can help us understand why young people prefer one news source over another. It will also shed light on their trust in news media and their appreciation of core PSM values such as independence, impartiality, and quality. Secondly, while the focus of this article was on news consumption and trust, a more comprehensive account of news repertoires and their social stratification will add to a more layered and contextualized understanding of news consumption.