Public service media in a digital media environment: performance from an audience perspective

For decades, public service broadcasting has played an important role in the provision of news and information in many European countries. Today, however, public service media (PSM) are confronted with numerous challenges, including the need to legitimise their role in an increasingly digital media environment. Against this background, this study examines the audience perspective on the topic with an international comparative approach. It analyses the population’s assessment of, and attitudes towards, the performance of PSM. The aim is to identify what relevance is attributed to PSM by the public in the digital age and how they see PSM’s role in comparison to other more recent (digital) media offerings. An online survey was conducted in three specifically selected countries: Germany, France, and the UK. Overall, the findings show that respondents attribute a clear role to PSM and distinguish it from other media offerings in the increasingly digital media environment. They rate the information quality offered by PSM as higher than that of most other media offerings. Respondents are more likely to value social media platforms for entertainment purposes than PSM. The findings also reveal differences in the evaluation of PSM depending on PSM news use, interest in news, political interest, as well as on demographic variables. On the other hand, differences between the individual countries overall were surprisingly small, pointing to the fact that PSM across the countries sampled are—with deviations—perceived to be performing better than (most) other media, despite being confronted with changes and challenges in their environment.


Introduction
For decades, public service broadcasting has played an important role in the provision of news and information in many European countries.Public service broadcasters have traditionally been required to fulfil public service obligations, such as universality and diversity in access and coverage, as well as to provide highquality national programmes grounded in an independent, impartial, and accountable approach (Schweizer & Puppis, 2018, pp. 114-115).Today, however, public service media (PSM) are confronted with numerous inter-nal and external challenges: (1) PSM have a high reach regarding their news on the traditional radio and television channels, but for online news-in relation to their offline reach as well as compared to leading newspaper websites-it is less high (Newman, Fletcher, Schulz, Andı, & Nielsen, 2020;Schulz, Levy, & Nielsen, 2019, pp. 12-14), resulting in a pressure for these traditional radio and television organisations to adapt to the digital media environment and to develop strategies for digital, mobile, and social media news distribution (Sehl, Cornia, & Nielsen, 2016); (2) PSM particularly struggle to reach young and hard-to-reach audiences in this new environ-ment (Schulz et al., 2019, pp. 15-19), where they have to compete for attention with newer actors such as social media platforms or digital news start-ups as well as with purely non-journalistic offerings; (3) political and legal constraints challenge PSM (Brevini, 2013).Private sector media and some political parties accuse them of unfairly distorting competition through their public funding and call for more narrowly defined roles and remits.However, research has found little evidence of this so far (for an overview see Nielsen, Fletcher, Sehl, & Levy, 2016, pp. 56-77;Sehl, Fletcher, & Picard, 2020;Sjøvaag, Pedersen, Owren, & Thomas, 2018).Some countries go beyond this and are debating the legitimacy of PSM from both a liberal and a populist perspective.Rightwing populists claim that PSM coverage is distorted in favour of the establishment and is biased towards a proimmigration and politically left-wing cultural elite (Sehl, Simon, & Schroeder, 2020).
Against this background, this study examines the audience perspective on the topic with an international comparative approach including three specifically selected countries: France, Germany, and the UK.It analyses the population's assessment of, and attitudes towards, the performance of PSM for news and information in relation to various context variables such as PSM news use, interest in news, political interest, and demographic characteristics.The aim is to identify, what relevance the public attributes to PSM, especially for news and information, in the digital age and how they see PSM's role in comparison to other more recent (digital) media offerings.In doing so, this study builds on studies about PSM reach and attitudes towards PSM in single countries (e.g., Holtmannspötter & Breunig, 2018; an exception for reach in multiple countries is Schulz et al., 2019).These existing single-country studies are often not independent, as they are commissioned by the PSM themselves.This study addresses these issues, as it is independent and comparative in design.The comparative approach allows discussion of whether the assessment is mainly influenced by country context or whether there is, at least for the sample countries, a tendency for how the public perceives PSM as distinct from other media.
To this end, an online survey was conducted in Germany, France, and the UK, representing different media systems (Brüggemann, Engesser, Büchel, Humprecht, & Castro, 2014;Hallin & Mancini, 2004).Apart from funding, the institutional and legal conditions under which PSM operate may either constrain or enable attempts by governments or other political elites to influence reporting (Hanretty, 2009).To understand the expected differences in the findings, it is therefore important to emphasise that while they are all PSM, they are not all the same.
The article is structured as follows: First, the question of how (public service) media should perform is discussed from a normative perspective, before research on media performance from an audience perspective is presented.On this basis, research questions are formulated and the method of the study laid out.In the penul-timate section, findings are presented.The article concludes by summarising and interpreting the findings, discussing the limitations of the approach, and providing perspectives for future research.

Public Service Media and Media Performance
How media should perform is a question that can be answered from different perspectives, for example from an economic, an audience perspective, or from the professionals' point of view etc.Moreover, quality criteria are always derived from certain norms, principles, standards, or regulations, which must be stated (Wyss, 2002, p. 98).
How PSM specifically should perform is laid down in their legal basis and regulations in each country.However, these legally defined public service obligations are often relatively general and vague.Therefore, on this basis, the European Broadcasting Union (EBU), an association which represents PSM in Europe and beyond, as well as the ARD, a joint organisation of Germany's mainly regional public service broadcasters, have developed value systems which are more operationalisable and also address the challenges PSM are facing today in the increasingly digital media environment.Their focus is on PSMs' value to society, and they actively aim to stimulate a societal debate around them (van Eimeren, 2019, p. 452).Consequently, the perspective on media performance here is a normative one.From this perspective, media performance is defined according to the function of news media in democratic societies (e.g., McQuail, 1992).
The EBU describes six core values and public service obligations of PSM: Universality (in access, reach, and content), independence, excellence, diversity, accountability, and innovation (EBU, 2012).These correspond partly to media performance and related criteria previously developed in communication studies for media in democracies (e.g., McQuail, 1992;Rager, 1994;Schatz & Schulz, 1992).Urban and Schweiger (2014) reviewed those different criteria catalogues and arrived at six basic quality dimensions: Diversity, relevance, accuracy, comprehensibility, impartiality, and compliance of ethical standards.In comparison, it is obvious that the EBU-criteria focus specifically on PSM and their obligations which partly differ from the expectations towards private sector media (e.g., regarding universality).However, some criteria explicitly mentioned in other studies are not mentioned in the EBU-list, possibly because dimensions and subcategories are arranged differently.It is also conceivable, however, that PSM in Europe have differed too much in terms of their legal requirements and values to arrive at a common denominator.In a nutshell, PSM tend to develop a higher degree of professionalism and independence from political control in countries with a strong democratic tradition than in countries where this democratic tradition is less pronounced or which have emerged from authoritarian regimes.In the latter, political clientelism and state paternalism are more frequent (Brevini, 2015).This article builds on these theoretical considerations and empirical studies and adds to them by not only evaluating PSM, but by putting them in a context with other media types, legacy media and new digital media offerings, as well as by not focusing on one country only, but being comparative in its approach.

Media Performance from the Audience Perspective
Apart from the perspective of public value and the accompanying media qualities formulated by journalists, media experts, and scholars, there is the perspective of the audience itself.This is less researched than the normative perspective, however, the number of studies has grown in the last ten years.
Studies on how recipients evaluate different media offerings and/or whether they are able to determine the normative quality of reporting focus on different levels: On the macro-and meso-level, they analyse recipients' evaluations of media types such as radio, TV, newspaper, online (e.g., Holtmannspötter & Breunig, 2018, differentiated between private sector broadcasting and PSM; Neuberger, 2014), or specific media brands or programmes (e.g., Gehrau, 2008;Gscheidle & Geese, 2017).The findings of these surveys show that recipients indeed differentiate between different media types or brands in their quality evaluations and, for example, rate German PSM radio and TV offerings higher for information-centred qualities than those of privatesector radio and TV (Holtmannspötter & Breunig, 2018), or specific PSM TV news programmes higher than those of private-sector TV (Gscheidle & Geese, 2017).On the micro-level, they examine single news items, usually in a combination of content analysis and survey or experiment (e.g., Jungnickel, 2011;Urban & Schweiger, 2014).Overall, they indicate that recipients realise quality differences to a certain extent, but not fully.Furthermore, Urban and Schweiger (2014) identify media image as a factor influencing the assessment of a news item.
Apart from the question of the evaluation of different media offerings, studies have also analysed (role) expectations of the audience, often in relation to those of journalists.However, the focus here is on the audience.Tsfati, Meyers, and Peri (2006) compared Israeli public and journalist perceptions of what constitutes good and bad journalism.They found that the public rated verifying facts, neutrality, and not publishing rumours more important than public interest or interpreting the news.In a study comparing US newspaper journalists and US citizens, de Zúñiga and Hinsley (2013) observed that the public rated getting information quickly, covering stories that should be covered, and verifying facts most highly, while being a watchdog for the public, helping people andinterestingly-being objective was rated lowest.Heise, Loosen, Reimer, and Schmidt (2014) compared journalistic role expectations of journalists and audience members of a German PSM news programme (Tagesschau).
Here, the audience rated-contrary to de Zúñiga and Hinsley (2013)-objectivity highest, followed by explaining complex topics and criticising problems.To provide useful information for the audience in an advisor/guide role, entertainment and relaxation and opportunity for the audience to communicate among themselves was rated least important.Like de Zúñiga and Hinsley (2013), Willnat, Weaver, and Wilhoit (2019) focused on US journalists and US citizens.Their findings show that for citizens it was most important to get information to the public quickly, to not publish unverified content, and to investigate government claims.To develop intellectual or cultural interests, to influence the political agenda or entertainment instead was seen as least important.In addition, the study revealed that traditional news media use and social media use for work/news predicted higher support for traditional journalistic roles among citizens.Also, Vos, Eichholz, and Karaliova (2019) explored how the American audience assesses normative journalistic roles and compared their assessments to those of journalists.Most important to citizens was to "report things as they are," "educate the audience" and "provide information people need to make political decisions," while aiming to "influence public opinion," "be an adversary of the government" and "set the political agenda" were ranked lowest.(Vos et al., 2019(Vos et al., , p. 1022) ) Taken together, the studies show that citizens overall rate traditional journalistic role expectations important that are in line with normative media performance such as reporting quickly, on relevant topics, and in an objective manner, while they differentiate other roles such as acting in particular interests.These studies are an important context for the current study insofar as they provide information on how important the performances explored here are to audiences.
Fawzi (2020) analysed, based on a representative survey of the German population, to what extend the media are able to fulfil expected performances from a citizen's perspective; her findings show that the audience overall is satisfied with the media.However, while the respondents were satisfied with how the media fulfilled their watchdog, information, and opinion formation function, larger parts did not say this to the same extent regarding integration, articulation, and orientation functions.Fawzi further explored how different individual political variables and media use relate to the evaluation of media performance.She found that political interest, satisfaction with democracy, political confidence, and satisfaction with the current economic situation had a positive influence on the evaluation of media performance.The type of media (e.g., PSM or tabloids newspapers) respondents used were not significant predictors.Perceived media dependency and presumed media influence were positively related to the assessment of media performance.
Finally, studies are investigating the role of quality perceptions in selecting a media product: de Zúñiga and Hinsley (2013) found a positive association between the citizens' assessment of media performance and their news and infotainment use.In line with that, Tsfati (2010) demonstrated that mainstream news exposure is related to trust in media, while exposure to non-mainstream news websites is associated with media scepticism.
To sum up, we have seen in the literature overview that the audience: (1) can evaluate media performance to a certain degree; (2) their expectations to journalistic roles highly overlap with normative media performance criteria; and (3) they are more likely to consume news they assess as being high performance and in which they trust than those news offerings they rate less highly or they are more sceptical about.
Nevertheless, it is important to underline that the audience perspective and the normative perspective do not need to be in line as it is expressed in the assumption that PSM are seen as merit goods (e.g., Holtz-Bacha, 2015, p. 38).Merit goods, often expensive to produce while providing little financial return, are not provided by supply and demand because what they produce is socially desirable (for an overview of studies on the social and political impact of PSM see Nielsen et al., 2016) while not paid enough for by the consumers (Ali, 2016).

Research Questions
This empirical study seeks to analyse audience perceptions of PSM performance compared to other media types in today's changing and increasingly digital media environment.Here actors compete for attention with legacy media such as commercial radio, TV, and newspapers, as well as newer actors like social media platforms.The aim is to identify, in a comparative design for France, Germany, and the UK, what relevance the public attributes to PSM, especially for news and information, in the digital age and how they see PSM's role compared to other and more recent (digital) media offerings.In this context, it is important to say that categories like social media or video platforms are, when evaluated e.g., regarding trust, usually evaluated overall in spite of the fact that they host a variety of actors and offer very diverse content-from private posts over public relations to propaganda content and professional journalism.However, little is also known about how much the audience actually differentiates between different actors and contents on these platforms.
The research questions for this three-country study are as follows: RQ1: How do citizens rate the overall performance of PSM for information against other types of media?RQ2: What specific performances do citizens associate with PSM compared to other types of media?RQ3: How do PSM news use, interest in news, political interest, and demographics influence the evaluation of perceived PSM information quality?

Method
The study is based on an online survey conducted in three countries, France, Germany, and the UK, with respect to the assessment of and attitudes towards PSM compared to other types of media.In each of the three countries, 1,000 citizens, representative of gender and age (18-69 years in 10 years groups, e.g., 30-39 years), were sampled and surveyed via an ISO-certified online access panel provider (Respondi) in September 2019.As is common with online access panel providers, participants received a small financial incentive to take part in the survey.
The questionnaire (documented for transparency in the Supplementary File) was professionally translated to the relevant languages and scheduled to take 12 minutes to complete.Before answering questions, respondents were given a short explanation of what was meant by PSM in the questionnaire: For France, PSM was defined as all offerings of Radio France and France Télévisions; for Germany as all offerings of the ARD (including all regional ARD organisations, i.e., BR, HR, mdr, NDR, Radio Bremen, rbb, SR, SWR and WDR), of the ZDF and of Deutschlandradio; and for the UK, for the purposes of this questionnaire, PSM were defined as all services of the BBC.
This article is based on the following variables, which are described in the following paragraphs.
Sources of news: Respondents were asked where they get information about current events from, at least once a week.More than one answer was possible.Examples for each media type were provided specifically to each country context.For the UK, Channel 4 (publicly-owned and commercially-funded public service broadcaster) was treated as a special and separate case.The answer format of the multi-response questions was dichotomous.
Assessment of overall information quality of different types of media: For each type of media, this was measured separately by means of a four-point scale (very poor, generally poor, generally good, very good, don't know).
Assessment of specific performance of different types of media: The statements were taken from an ARD/ZDFsurvey series Media and their Audiences (Medien und ihr Publikum, see Holtmannspötter & Breunig, 2018).Holtmannspötter and Breunig (2018) used those for a comparison of PSM and private sector radio and TV in Germany.In this respect, they cover normative values of PSM (similar to van Eimeren, 2019), but also categories relevant to private sector media in order to compare different types of media.The answer format of the multiresponse questions was dichotomous.
Interest in news and interest in politics: The interest in both was measured by means of a four-point scale (no interest, not very strong, strong, very strong).
Demographics: Gender was measured as female, male, diverse.Age was measured in 10 year-age groups (e.g., 30-39 years, for 18-69 years old).The highest educational qualification was adapted to each country context.These demographic variables were selected as measures on how PSM fulfil their mission of universality including different genders, age groups, and citizens from various educational backgrounds.
The three countries were chosen according to strategic sampling, since each country represents a different Western media system (Brüggemann et al., 2014;Hallin & Mancini, 2004).As such, the survey covers media markets with varying conditions, including Germany with well-funded PSM as a representative of the democratic corporatist model; the UK as a representative of the liberal model, although this classification has been contested by some scholars due to its strong PSM and polarised press (e.g., Norris, 2009, pp. 333-334); and France within the polarised pluralist model, in which commercial media are often dominant.While in the UK and Germany direct influences of the political system on PSM content are unusual, PSM in France are in practice characterised by a higher level of state influence (see also Nord, 2015, p. 184).As a consequence of these differences, Kuhn (2006, p. 20) summarises: "France Télévisions does not enjoy the same status or legitimacy in the French media landscape that the BBC has in the UK equivalent." Table 1 gives an overview of PSM in the three countries with key characteristics in order to provide some context for the comparative analysis.
Also when it comes to distributing their legacy programmes as well as additional services on the Internet, Brevini (2011, p. 175) argues that Southern Europe is behind the UK or Germany in this respect "as the expansion of public broadcasters into new media is a more recent phenomenon."Furthermore, in the UK and Germany-like in many European countries-PSM conduct so-called publicvalue tests (Drei Stufen-Test in Germany), which aim at balancing the public interest of new online offerings with their market impact.In France, however, the control of state aid, as generally required by the European Commission for the member states, was not an issue for the online activities of PSM and a public-value test not implemented (Gransow, 2018).
In line with this, the overview of key data on PSM in the three countries shows that they have a high weekly reach for news offline in the UK and Germany, while their reach in France is somewhat lower.Only the BBC in the UK is able to nearly match their weekly offline reach for news online, while in France and Germany more respondents said social media is a source of news for them than get their news at least weekly from PSM online.
Apart from what has just been said above, this also corresponds to the fact that the BBC transformed its organisation to digital relatively early on, while in France, Radio France and France Télévisions are still separate organisations, both generally including online, which only in September 2016 introduced a joint PSM news website, franceinfo (Sehl, Cornia, & Nielsen, 2017).In Germany, it is important to note that the remit of PSM online activities is strictly limited by legislative interventions.The online offers generally have to be related to broadcast programmes.Furthermore, there are time limitations for digital content to be accessed online (Nissen, 2015, p. 100).
While in all three countries PSM is mainly funded by public revenues (licence fees) in order to avoid direct influence by the state or business, the public revenues per capita vary and are almost double for Germany than France, with the UK being in between.Also normalised Notes: † Calculations include for France, Public radio (France Inter etc.) (Radio France), France Télévisions (France Télévisions); for Germany, ARD News (Tagesschau, Tagesthemen), ZDF News (heute, heute-journal etc); for the UK, BBC News.‡ Calculations include for France, France Info (online); for Germany, ARD News online (tagesschau.deetc.), ZDF News online (heute.de);for the UK, BBC News.* The methodology of the Reuters Institute Digital News Report measures ARD and public regional TV news separately.Public regional TV news is listed with 20 percent weekly reach.However, the PSM funding figures are for the ARD as a whole, including the regional organisations.§ The figure for BBC total revenues includes the turnover of BBC Worldwide and other global commercial activities as well as licence fee revenues, in line with the methodology of the European Audiovisual Observatory (see also Sehl et al., 2016, p. 12).Sources: European Audiovisual Observatory (2016), Eurostat (2020), Newman et al. (2020), and World Bank (n.d.).
by the gross domestic product (GDP) at market prices, the German PSM together have a higher budget then the British and, at the bottom of the list, the French PSM.However, Germany also has, as explained above, a comprehensive PSM system with two main national PSM of which one (ARD) consists of nine regional PSM organisations (and a public international broadcaster, but this one is not funded by the licence fee, but mainly by the German federal tax budget).

Findings
RQ1: How do citizens rate the overall performance of PSM for information against other types of media?
Respondents in all three countries (France, Germany, and the UK) were asked how they rate the quality of information from different media types, including PSM.Across all three countries, the findings, in general, show that PSM are at the top of the ratings compared to other types of media (see Table 2).PSM radio and TV share first place with the national quality newspapers and local and regional newspapers.PSM online services rank directly behind.They are followed by commercial radio and TV, and at the bottom of the list are video platforms, social media platforms, and tabloids.Kruskal-Wallis H tests show that there are statistically significant differences in PSM evaluations between the countries: for PSM radio:  2 (2) = 11,313, p = 0.003, with a mean evaluation score of 1,360.98 for Germany, 1,289.32 for the UK, and 1,260.08 for France; for PSM TV:  2 (2) = 15,367, p = 0.000, with a mean evaluation score of 1,466.30for Germany, 1,405.43 for the UK, and 1,337.86 for France, and for PSM online:  2 (2) = 59.922, p = 0.000, with a mean evaluation score of 1,243.98 for Germany, 1,268.12 for the UK, and 1,049.17for France.
Post-hoc-tests (Dunn-Bonferroni-Tests) show significant differences between France and Germany regarding PSM radio (z = 3,268, p = .003)as well as PSM TV evalu-ations (z = 3,918, p = .000).However, in both cases, the effect sizes according to Cohen (1992): r = 0.06 resp.0.07 are small.There is no evidence of a difference between the other pairs for PSM radio and TV.Instead, for PSM online evaluations, there is evidence not only of differences between Germany and France (z = 6,324, p = .000),but also the UK and France (z = 7,146, p = .000).Here, the effect sizes are a bit larger but still small (r = 0.13 resp.0.15).These findings could possibly be due to the previously discussed lower status of PSM in France as well as the fact that until recently, there was no central news website of PSM in France, but rather news sites of both Radio France and France Télévisions with numerous subsites (Sehl et al., 2017).RQ3 below will address how PSM news use impacts on the quality assessment for each PSM platform and country.
Beyond PSM, the findings (see Table 2) show that online offerings are generally rated better in the UK than in Germany and France; this applies to social media platforms as well as video platforms.Tabloids, which play a more important role in the UK media market, are also rated better there than in France and Germany.On the other hand, national quality newspapers and local and regional newspapers are perceived worse in the UK than in France and Germany.Overall, however, the differences are minor and cannot always be explained by the media systems.
RQ2: What specific performances do citizens associate with PSM compared to other types of media?
Information from the local and regional environment is a basic need for many people.Newspapers are the most popular here in France and Germany, as this category also includes local and regional newspapers (see Table 3).However, the respondents stated that PSM are the second-most likely to provide this service (and in the UK even the most likely), far ahead of all other types of media.Notes: Question "Below are some statements regarding services provided by the media.Please tick all the media to which each statement applies in your case.More than one answer may be given" (n France = 1,000, n Germany = 1,000, n UK = 1,000).* = dominant category/categories in the respective country.
Interestingly, there is a quite consistent difference between the audience assessments of PSM media performance in Germany and the UK versus France when it comes to several information-oriented categories.For "provide reliable and credible information," "provide high-quality journalism," "convey the values of our society," and "provide comprehensive background information on many topics" PSM in Germany and France (radio, TV, online) receive the highest levels of agreement, while in France this applies to newspapers.
However, in the category "allow all sides to express their views on social issues" only in Germany are PSM the category with the highest percentage.In France and the UK, most respondents associate this performance with social media platforms-possibly because of their focus on user-generated content.
There is also another remarkable point: Only 15 percent (France) and 19 percent (UK) of respondents perceive PSM as "independent of state, politics, and business".In France, the highest percentage of respondents state this for social media platforms or say "neither of these media" (22%).This can be interpreted cautiously in such a way that for many users, the democratic potential of the platforms to create and distribute their own content is obviously in the foreground, rather than the commercial goals of the platform companies (e.g., Nielsen & Ganter, 2017) or the political goals of some of the actors there.In the UK, the dominant group of respondents "don't know" (28%).In Germany, the percentage of respondents who see PSM as independent of state, politics, and business is slightly higher (28%), but also percentages for this category are relatively low across all types of media.These findings generally show that a significant part of the population is sceptical about the independence of PSM and journalism in general.
Finally, PSM are seen as a little less "enjoyable or amusing".Here they are overtaken at least in Germany and the UK by social media platforms.So overall, PSM are associated with information-oriented characteristics, while other types of media, especially social media platforms, complement them especially in respect of their being enjoyable or amusing.RQ3: How do PSM news use, interest in news, political interest, and demographics influence the evaluation of perceived PSM information quality?This research question explores the effect of various factors on the evaluation of the quality of information in PSM.In addition to the interest in news and political interest, various central demographic factors (age, gender and formal education) have been included, based on the theoretical assumption that PSM should address society at large.
In various multiple linear regressions, the effects were calculated separately for PSM radio, TV, and online-and separately for the three countries.The use of PSM as sources of information about current events (at least weekly), also separated by media genre, were used as control variables.This was done in blocks, with PSM news use as the control variables first, followed by interest in news and political interest, and finally, demographic data were included in the regression model.
To begin with: In all cases, the explained variance after the first block of PSM news use accounted for a large part of the variance explained by the entire model.Nonetheless, the other two blocks-interest in news and political interest as well as demographic data-led to a further explanation of variance.
The linear regression models show a quality score of .148(Germany), .070(UK), and .106(France; adjusted R-squared) for the evaluation of the quality of information in PSM radio in the three countries (see Table 4).In addition to the use of PSM news on legacy platforms, the most significant and meaningful predictor is, with slight differences between countries, interest in news.Gender (male) also has a significant, albeit weak, influence on the rating of the information quality of PSM radio programmes.In contrast, political interest is a significant predictor only in Germany and has a negative sign, meaning that political interest impacts negatively on the quality assessment.A possible interpretation could be that citizens that say that they are more interested in politics are therefore also more critical towards news and information.Age and formal education have no significant effect in any of the three countries.
Similarly, the significant effects can be seen in the evaluation of the information quality of PSM TV programmes in the three countries (see Table 5).Here the model achieves a quality of .142(Germany), .059(UK), and .112(France; adjusted R-square).Again, in addition to the use of the PSM radio and TV for news, the interest in news has a significant influence and is a strong predictor in all three countries.Political interest, on the other hand, has a significant influence only in France and Germany, again with a negative sign.Furthermore, for  gender (male) a significant, albeit weak influence, can be observed in all three countries, whereas formal education is only significant in the UK.Age has no relevant effect in any of the three countries.Finally, it is interesting to note the deviations that arise when assessing the information quality of PSM online (see Table 6).Here the model achieves a quality of .167(Germany), .068(UK), and .082(France; adjusted R-square).In terms of media use, significant effects are shown in all three countries for the news use of PSM online offerings themselves, while the news use of offline PSM platforms only partly has significant effects.For the third time, the interest in news in all three countries is a significant predictor for the evaluation of the information quality of PSM offerings, here online.The influence of political interest is for Germany significant, but negative.
For the first time, however, age is a significant predictor in all three countries.Obviously, online shows age effects that do not exist in this form for the traditional radio and TV distribution channels.The sign is negative.A possible explanation could be that there are differences in how digitally savvy different generations are and that, in this respect, age impacts on the evaluation.Furthermore, gender (male) is a significant, albeit weak, predictor in Germany and England, but not in France.
Overall, the use of the same offerings and interest in news have a clear influence across all PSM genres and countries and in Germany also political interest, however with a negative sign.In contrast, the demographic characteristics of gender, age, and formal education play a subordinate role at most.Overall, the explained variance is small; in other words, other factors better explain the evaluation of the PSM offerings.On the one hand, this is not surprising, and on the other hand, it is even a positive finding regarding the mission of PSM of universality and to be attractive to different groups in society, independent of gender, age, or education.From previous research, it is furthermore known that for the evaluation of quality, criteria such as interest in a topic (Jungnickel, 2011), image of media brand (Urban & Schweiger, 2014) etc. instead play an important role.

Conclusion and Discussion
To summarise, the findings have shown that respondents attributed a clear role to PSM and clearly distinguished it from other media offerings in the increasingly digital media environment.They rated the information quality offered by PSM higher than that of most other media offerings.National quality newspapers as well as local and regional newspapers also performed well in terms of perceived information quality, while social media platforms and video platforms lagged behind.In this respect, the findings correspond to previous research which found that the audience can perceive differences in media performance at least to some degree (e.g., Jungnickel, 2011;Urban & Schweiger, 2014).However, the actual normative media performance needs to be measured by means of content analysis.Furthermore, as mentioned earlier on, a methodological limitation of current studies, including this one, is that in surveys they often measure social media or video platforms as overall categories when in fact they host a variety of actors and offer very diverse content.Nevertheless, obviously, an expansion of digital services has also not meant that legacy media services are not seen as important for informing society any more, at least not when it comes to the assessment of media performance.Interestingly, there were no major differences across the three sampled countries, but more minor deviations that can partly be interpreted as differences in the media systems (e.g., tabloids, which play a more important role in the UK media market [Hallin & Mancini, 2004], were also rated better there than in France and Germany).Furthermore, the posthoc-tests showed significant differences between France and Germany regarding the evaluations of PSM radio, TV, and online information quality as well as between the UK and France regarding PSM online information quality.Although the effect sizes were in all cases small, these findings could possibly be due to the previously discussed lower status of PSM in France and, specifically for online, might directly reflect on recent developments in the French PSM as they only introduced a joint news website of both Radio France and France Télévisions in 2016 (Sehl et al., 2017).
In the assessment of specific performances of the various media, PSM scored particularly well in all categories in which information functions were queried.In this respect, the findings have shown that respondents saw traditional journalistic role expectations found in various studies actually largely fulfilled by PSM, such as reporting quickly, on relevant topics and in an objective manner (e.g., de Zúñiga & Hinsley, 2013;Heise et al., 2014;Tsfati et al., 2006;Vos et al., 2019;Willnat et al., 2019).Social media platforms were more likely to be valued by respondents in terms of entertainment.
There was one interesting difference between the audience assessments of PSM performance in Germany and the UK versus France.In a number of informationoriented categories in Germany and the UK, PSM (radio, TV, online) received the highest percentages of agreement, while in France this applied to newspapers.At the same time, PSM in France had the lowest percentages of agreement in comparison to the other two countries when it came to independence of state, politics, and business.This finding indeed corresponds to studies which describe high levels of influence of politics on PSM content in France (executive interventions, politicised councils; see Nord, 2015).
Nevertheless, overall, differences in assessments of and attitudes towards PSM between the individual countries were small.This was surprising, considering that these countries are characterised by differences, for example in (PSM) news use, PSM funding, in the organisational structures of PSM, the differences in PSM legal frameworks, and their insulation from direct political influence.This suggests that PSM across the sample countries enjoy-with deviations-a relatively strong position when it comes to their perceived media performance as distinct to other media, in spite of their being confronted with changes and challenges in the environment, including the need to deliver content across the various distribution platforms used by different target groups, as well as the rise of social media platforms.While in detail the country context mattered, e.g., in how (fast) they have adapted to digital for various reasons such as their legal frameworks, the findings have demonstrated that across these countries, the audience perceived a clear information-oriented profile of PSM and attested, with deviations especially for France, PSM a comparatively high quality.
However, the findings have also shown that a significant part of the population, albeit with some variation in country context, was sceptical about the independence of PSM and journalism in general.This aspect is impor-tant to observe, especially as right-wing populism is gaining ground in many European countries and right-wing populists often attack the media and PSM specifically, calling it a 'state broadcaster' ('Staatsfunk';Niemeier, 2018), trying to cut its remit, or even abolish it-or at least its public funding (on populist attacks on PSM see also Sehl, Simon et al., 2020).
It would have been desirable to give more contexts to PSM and challenges in individual countries.In future studies, it will therefore also be useful to deepen the quantitative findings of this survey, for example in group discussions.In this way, more can be discovered about the arguments for certain evaluations, and different groups, such as age groups or politically ideological groups, can be better understood.
The findings thus highlight that PSM in Western Europe still have a firm place in the opinion of respondents, despite the large number of available digital sources for news and information.However, these findings are not set in stone, and it would be helpful to observe them in longitudinal studies, as the media environment and media use are likely to continue to change.At the same time, there are also major challenges for PSM if they are to survive in this increasingly digital media environment.They must be able to drive the digital transformation in their own news organisations.Only then will they also become attractive for younger target groups, who attest to their being of good quality but who no longer use them as naturally as the older generations do, instead informing themselves via social media platforms.Here, PSM must continue to break new ground in news distribution and must adapt their formats accordingly in order to reach younger target groups-at least within the framework to which they are legally bound.In this respect the regression analyses of this study have shown for PSM online services-contrary to their offline services-that age indeed was a significant predictor of the assessment.Consequently, PSM need to live up to their mission of universality in order to legitimise themselves for society in the long term.
PSM need to find ways to communicate the (democratic) value of their services for society to the wider society, especially as they are paid by the public and discussions regarding (the amount of) the licence fee, not only by populists, will probably not keep silent.This may also include seeking discourse with those who do not support the idea of a shared and universal service, offering diversity in content and opinion in an increasingly fragmented and polarised society.
Journalism Studies Division.Furthermore, she is grateful to students of a bachelor course on audience research at the Universität der Bundeswehr München for discussing and pre-testing the German version of the questionnaire.United Kingdom.Word Bank.Retrieved from https://data.worldbank.org/indicator/SP.POP.TOTL?end=2014&locations=DE-FR-GB&most_ recent_year_desc=false&start=1960&view=chart Wyss, V. (2002).Redaktionelles Qualitätsmanagement: Ziele, Normen, Ressourcen [Editorial quality management: goals, standards, resources].Konstanz: UVK.About the Author Annika Sehl is Professor of Digital Journalism at the Universität der Bundeswehr München, Germany, and a Research Associate at the Reuters Institute for the Study of Journalism at the University of Oxford, UK.Her research is often comparative and focuses on how digital change has affected news organisations, especially PSM, news production, and news use.Her research has been published in books, book chapters, and journals such as Journalism, Journalism Studies, the European Journal of Communication, and the International Communication Gazette.Annika Sehl is Associate Editor of the 2019 Wiley Blackwell-ICA Encyclopedia of Journalism Studies.She serves as Vice-Chair of the Journalism Studies Division of the International Communication Association.https://orcid.org/0000-0002-8949-569X

Table 1 .
Context variables on PSM as well as social media in France, Germany, and the UK.

Table 2 .
Assessment of information quality of various types of media in France, Germany, and the UK.

Table 3 .
Assessment of specific performances of various types of media in France, Germany, and the UK.

Table 3 .
(Cont.) Assessment of specific performances of various types of media in France, Germany, and the UK.

Table 4 .
Effects on the assessment of information quality of PSM radio in France, Germany, and the UK (multiple linear regression).

Table 5 .
Effects on the assessment of information quality of PSM TV in France, Germany, and the UK (multiple linear regression).Columns showing standardised beta coefficients following by standard error in parentheses.* p < .05,** p < .01,*** p < .001.n differs by country due to missing data for DV ["Don't know"-option] and "Other"-option for IV "Formal education".

Table 6 .
Effects on the assessment of information quality of PSM online in France, Germany, and the UK (multiple linear regression).