Ordinary People over Journalists? Young People’s Use of Different Curating Actors for News

Abstract

Most young people today consume news through social media, where diverse sources (e.g., influencers and journalists) curate news for them. However, we know little about which curating actors young people rely on for news, and which factors predict their use. Addressing these gaps, we draw on the curated flows framework and a quota-based survey of N = 569 young people aged 16 to 25 in Austria, conducted in December 2023 and January 2024. We found that ordinary people and influencers were young people’s most popular information sources on social media, favored by those who consume opinionated and soft news and encounter news incidentally. Traditional journalism and politicians were less relevant and primarily sought out intentionally, and for hard news.

1. Introduction

In the contemporary media environment, individuals access news not only via traditional news sources like newspapers, TV or online websites, but also increasingly via social media platforms (Newman et al., 2024; Wunderlich et al., 2022). This trend is especially pronounced among youth, who use social media as the primary source for consuming news (Hasebrink et al., 2021). Here, different actors select, modify and share—in short: curate—news (Thorson & Wells, 2016). Through the direct contact with audiences enabled by social media, actors such as politicians, influencers or celebrities1 have become increasingly relevant sources of information online (Monaci & Persico, 2022). In this role, they have challenged journalists’ monopoly as gatekeepers of information. In recent years, influencers and celebrities seem to have even established themselves as the most popular source of news on platforms such as YouTube or TikTok (Newman et al., 2021, 2023, 2024).

Prominent examples of influencers, distinguished from ordinary people through their “high follower counts and growing authority” (Harff & Schmuck, 2023, p. 147), creating news are British TikToker Dylan Page, (‘News Daddy’), who has over 10 million followers and covers a range of serious and lighter current affairs issues for young audiences (Newman et al., 2024), or Fidias Panayiotou, a former entertainment-influencer, who won 19.4% of Cypriot votes in the EU elections 2024 and now uses his popularity to create educational news videos covering his political work (Fidias, 2024; Wong, 2024). Meanwhile politically active celebrities also add to the news on social media. For instance, Taylor Swift regularly speaks up about current political events such as the 2024 presidential elections (Swift, 2024) or the Supreme Court’s overturn of Roe v. Wade (Swift, 2022), while Leonardo DiCaprio has been talking about climate change and the environment for many years (Robeers & Van Den Bulck, 2021).

While these examples offer anecdotal evidence for the importance of social media personalities as producers of news, we lack a systematic comparison of the popularity of different curating actors for news use among young people. Based on the curated flows framework (Thorson & Wells, 2016), we argue that five core curating actors on social media can be distinguished: (1) journalists and their organizations, (2) politicians or political activists, (3) ordinary people, (4) influencers, and (5) celebrities (Newman et al., 2024). Yet, we know little about young people’s propensity to use these actors for news, although these insights are crucial to understand key democratic processes such as political participation, knowledge acquisition, or opinion polarization, which may be differentially influenced by these actors.

Second, while existing research has investigated the predispositions and motivations that predict the use of single actors like influencers (Harff & Schmuck, 2025b) or politicians (e.g., Heiss & Matthes, 2017) on social media, we lack a systematic comparison of these drivers for different news-curating actors. Considering the diversity of sources that young people can choose from (Klopfenstein et al., 2024), it is critical to analyze which factors characterize the use of these different sources for news. Young people may, for instance, actively seek out more traditional news sources because of their specific interests, but also often engage in more passive news consumption patterns (Costera Meijer & Groot Kormelink, 2015), which may link with exposure to news from actors engaging in social curation (e.g., friends or colleagues; Bergström & Jervelycke Belfrage, 2018). Insights not only into the relationships between consumption of certain news-curating actors and news exposure modes but also young people’s personal characteristics (e.g., gender) and motives for news consumption (e.g., wanting to receive humorous content) enable us to gain a more thorough understanding of youth’s experiences with news on social media. Moreover, they provide important information for stakeholders like journalistic organizations trying to reach young people, and allow for educators or literacy campaigns to tailor their programs to specific user types among this age group.

To address these crucial research gaps, we conducted a quota-based survey of a national youth sample (aged 16 to 25 years) in Austria, which provides systematic insights into the frequency of use of different curating actors and into key predispositions, motivations, exposure modes, and topic interests that predict using them for news. Against this background, the present study examines the overarching question of which curating actors young people rely on for news on social media, and which factors are associated with the use of these actors for this purpose.

1.1. Curated News Flows and Young People’s News Consumption

To map out the multiple ways in which news reaches young people and to better conceptualize differences between various news sources on social media, we draw from the curated flows framework established by Thorson and Wells (2016). This model identifies key information flows in the modern media environment. Central to their argument is the term ‘curation,’ which describes the “production, selection, filtering, annotation, or framing of content” (p. 310). The curated flows framework outlines five different flows through which information can reach people, reflecting that the visibility and presentation of information are nowadays shaped not only by journalists, but also by personal preferences, people’s social contacts, third parties’ strategic interests, and algorithms (Thorson & Wells, 2016). Extending this framework, we outline five central actors who curate news, which we understand as a specific subtype of the broader category of information2. We distinguish between journalists, influencers, celebrities, politicians and ordinary people (Newman et al., 2024), who all engage in either journalistic, social, and/or strategic curation of news.

First, journalistic curation involves professional news producers, like journalists and their organizations, who share information with the public (Thorson & Wells, 2016). While traditional news sources such as printed newspapers are regarded as somewhat antiquated (Zerba, 2011), they still likely constitute relevant components of young people’s news diets, even on social media (Newman et al., 2024; Swart & Broersma, 2024).

While journalists were traditionally also assigned a gatekeeper role—meaning they determined not only the content people would see, but also whose voices would be heard—social media has diminished this function and enabled non-journalistic actors to directly address their audiences without journalists acting as intermediaries (Gunn & Rosenberg, 2018). In this context, politicians or political activists do not necessarily produce but at least promote news as a form of strategic curation, disseminating it to underline their own political and ideological stances. Prior research has indicated that the majority of politicians engage in news sharing, suggesting that it is a highly common practice among this source type (Buyens et al., 2025).

Meanwhile, social curation describes curation performed by the human social network, e.g., friends, family or colleagues. Social curation is most closely related to the two-step flow of communication, whereby opinion leaders filter media messages and distribute them among their social contacts (Lazarsfeld et al., 1968). This process is also encouraged on social media, where peers and other ordinary people “pass on everyday news on different topics” (Bergström & Jervelycke Belfrage, 2018, p. 595). Advancing the original concept, Stehr et al. (2015) introduced parasocial opinion leadership to describe the influence that media personalities can have on followers through one-sided, unreciprocated, so-called parasocial relationships (Horton & Wohl, 1956). Influencers and celebrities could be interpreted as opinion leaders following this logic, since young people often grow intimate ties with them (see also Harff & Schmuck, 2025b). Influencers in particular have become key news sources for young people, offering content perceived as more modern, relevant, and comprehensible than traditional news (Pew Research Center, 2024; Zimmermann et al., 2022). Similarly, celebrities have been posting news-related content on their personal profiles (e.g., Johnson et al., 2023) and functioned as news sources for adolescents (Newman et al., 2024). Yet, classifying a dominant flow for influencers and celebrities is challenging, as influencers and celebrities have diverse motivations for posting news: On the one hand, they promote community building (Lou, 2022), which leads followers to view them as friends or peers, allowing influencers and celebrities to become part of followers’ social flows (Harff & Schmuck, 2023). Some celebrities, and especially many influencers, however, also pursue strategic curation by aligning content with economic interests, including sponsored deals (Borchers & Enke, 2022), endorsing politicians (Goodwin et al., 2020), or even running for office themselves (Stamouli, 2024). Additionally, collaborations between influencers and journalistic actors are increasingly emerging. For example, Germany’s public service broadcaster launched ‘funk’ to connect with younger audiences through social media personalities (Lichtenstein et al., 2021), blurring the line between typical influencer content and journalism (Fischer et al., 2022; McEnnis, 2023).

As a fourth form of curation, personal curation describes the process through which users shape their own information repertoire by choosing which accounts to follow and which information to consume (Thorson & Wells, 2016), closely corresponding with the active audience paradigm in media research (Ruggiero, 2000). Finally, algorithmic curation shapes how people receive their information through the automated promotion and prioritization of specific types of content (Thorson & Wells, 2016).

The curated flows framework and actor categories used in this study do not refer to the exact same analytical levels and can therefore not be equated. The curated flows model describes more general logics through which information becomes visible, circulates, and is encountered, whereas the curating actors examined here refer to groups of actors who may operate within and across several of these flows. Influencers and celebrities may be embedded in social curation through parasocial ties while also engaging in strategic curation when aligning content with self-branding, sponsorships, or political positioning (Thorson & Wells, 2016). In this sense, the boundaries between journalistic, social, and strategic curation are not fixed, but fluid and potentially reciprocal.

Personal and algorithmic curation are closely intertwined with journalistic, social, and strategic curation. The former both shape and are shaped by the latter forms of curation. For instance, a person may initially follow a health influencer because of an interest in an active lifestyle (personal curation) or encounter this influencer’s content through recommendations in the platform feed (algorithmic curation). Over time, this person may develop an increasing parasocial connection with the influencer and engage more frequently with the content, while the influencer may simultaneously post content that combines socially embedded communication with strategic goals, such as promoting a nutrition brand. Such encounters can then feed back into future curation processes: prior exposure and engagement may affect whom users actively choose to follow (personal curation), while platform algorithms may respond to such engagement by making similar content more visible in the future (algorithmic curation). Personal and algorithmic are thus embedded in recursive feedback loops with journalistic, social, and strategic curation.

Taken together, through the different curated flows (see Supplementary Material, Figure S1 for an overview), various curating actors can crucially determine how (frequently) young people encounter news on social media. Yet, existing research has mostly examined the relevance of these actors separately (Harff & Schmuck, 2025b). The exception in this context is the latest editions of the Digital News Report, which repeatedly found that ‘personalities’—i.e., influencers and celebrities—rather than journalistic outlets were the most used source of news on social media like Instagram and TikTok (Newman et al., 2021, 2023, 2024). In their role as news providers, these personalities proved to be particularly popular among young people (Newman et al., 2021), which is further supported by qualitative research (Wunderlich et al., 2022). Yet, these reports did not distinguish between influencers and celebrities (Newman et al., 2024), even though they have a different origin of fame (Khamis et al., 2017) and divergent content creation habits and motives (Gonzalez et al., 2024). Furthermore, these two actor groups can be differentiated along two additional criteria: while celebrities usually engage in low levels of personalized communication in their predominantly third-party-curated public networks, influencers’ personalized communication in their primarily self-built public networks is more pronounced (Harff et al., 2025). We thus argue that consumption of their news-related content should also be investigated separately. Still, in comparison to the other source types and given these prior findings (e.g., Newman et al., 2024), we expect both influencers and celebrities to emerge as the two most used sources of news among young people. This possible source preference may be explained by personalities’ ability to combine news with entertaining elements (Zimmermann et al., 2022) or connecting news-worthy elements with their own life stories and experiences (e.g., Suuronen et al., 2022). With this type of content, they take a lighthearted approach to news that has been shown to strongly resonate with young audiences (e.g., Costera Meijer, 2007; Schwaiger et al., 2022). We thus postulate:

H1: 

Influencers and celebrities are the most used source of news for young people on social media.

1.2. Individual Predispositions and Usage of Curating Actors

In this article, we also seek out to explore which factors are associated with using specific curating actors for news. Following the selective exposure paradigm (e.g., Knobloch-Westerwick & Kleinman, 2012) and the notion of social curation (Thorson & Wells, 2016), people’s own media choices at least partially determine the content and source types that they are exposed to on social media. In this context, personal predispositions can offer some insights into which youth prefer which curating actors for news. Researchers typically consider individual predispositions such as education, attitudinal factors or age as possible determinants of general information consumption (York & Scholl, 2015). Regardless of the specific source type, people with lower levels of news interest generally tend to avoid news, while those with higher levels actively increase their news exposure (Geers, 2020). High levels of political news consumption from traditional media among young people have been connected with variables such as political interest, political knowledge (Geers, 2020), or political talk inside the family (York & Scholl, 2015). Meanwhile, high political interest and being male (Marquart et al., 2020) and having a strong political ideology and high levels of (online) political participation (Fisher et al., 2019) are predictors among youth for following politicians on social media. Young people who are male, show high levels of political participation, and have low levels of subjective political knowledge are more likely to consume political information from influencers (Harff & Schmuck, 2025b). Meanwhile, no research has yet investigated predictors of using celebrities for news. Moreover, no study has so far compared predispositions that determine the use of different curating actors. Especially among youth, such predictors for using varying kinds of news sources are scarcely researched, even though influencers and celebrities today likely play a key role in young people’s news diets (Newman et al., 2024). Thus, to identify different predispositions for using curating actors, we ask the following research question:

RQ1: 

Which individual predispositions predict the use of different curating actors for news on social media?

Additionally, we are interested in investigating which usage motives are associated with reliance on certain curating actors for news. Aside from traditional uses, following the uses and gratifications approach (Katz et al., 1973), there are specific qualities that young people seek in (political) information: youth especially appreciate when news is not only informative, but also enjoyable (Costera Meijer, 2007). Though traditional news media are potentially lacking in this respect, they seem to be valued by young people who search for reliable and objective news content (Wunderlich et al., 2022). For adolescents, content on social media is additionally judged more favorably when it is personalized, mobilizing, and opinionated (Marchi, 2012; Schwaiger et al., 2022). To systematically relate young people’s consumption motives to the content of novel curating actors, we thus ask:

RQ2: 

Which usage motives predict the use of different curating actors for news on social media?

1.3. Modes of Receiving News from Curating Actors

Consumption of news from different curating actors may also be tied to specific exposure modes. We distinguish between intentional exposure, meaning the active selection of content and source types, and incidental exposure. The latter concept describes exposure to content that recipients originally did not search for (Lee & Kim, 2017), is either algorithmically promoted (algorithm curation) or shared inside one’s own network (social curation), and then encountered by people by accident. Upon seeing content incidentally, social media users have two choices: simply scanning content and scrolling away (referred to as first-level incidental exposure) or engaging with and effortfully processing the recommended content (second-level incidental exposure; Matthes et al., 2020). Either choice reflects personal curation, as users decide whether to disregard the content or pay more attention to it. The distinction between these two levels of incidental exposure is also relevant due to potential differences in outcomes: learning effects and higher levels of online political participation have been observed for second-level incidental exposure, while first-level incidental exposure typically does not have these consequences (Nanz & Matthes, 2020, 2022). Thus, while the initial accidental contact with content may not have any implications for the individual, consequently engaging with a post does. Yet, so far, it remains unclear what kind of content is incidentally encountered by people in the first place, and which content is then further processed by adolescents. For instance, incidentally encountered content from novel curating actors like influencers or celebrities may be processed in more depth due to the parasocial relationships they tend to build with their followers (Lou, 2022)—but there is no evidence yet for such a link. However, it is also important to consider self-selection in the first step, as some source types may be specifically sought out for news, and this content would then not be discovered by chance. As such, the use of different curating actors among young people could be systematically related to intentional exposure or either of the two forms of incidental exposure. To analyze how people receive news from curating actors, we ask the following research question:

RQ3: 

Which exposure modes predict the use of different curating actors for news on social media?

1.4. Exposure of Young People to Different News Topics

For young people, a preference for soft news has been identified (Boukes & Boomgaarden, 2015). Hard news traditionally features breaking events and issues such as politics or the economy, and stylistically often contains unemotional and rational language (Boukes & Boomgaarden, 2015; Patterson, 2000). Soft news as a counterpart can be defined as more sensationalistic and entertaining, and traditionally treats issues of societal relevance in less depth and focuses instead on practical, service-oriented information. However, the boundary between soft news and hard news has, in some cases, been described as blurry (e.g., Boukes & Boomgaarden, 2015): soft news is sometimes connected to broader societal important issues, and hard news can be presented in a ‘softer’ way to keep audiences engaged. On the one hand, consuming soft news has been linked to political cynicism (Boukes & Boomgaarden, 2015) and lower political knowledge and efficacy (Andersen et al., 2016). Yet, when ‘softer’ topics are connected with broader societal issues, they can be engaging for young adults (Patterson, 2000; Reinemann et al., 2016). Importantly, such softer formats may serve as an entry point to political information for audiences that are less attracted to traditional hard news (Andersen, 2019). This mix of lifestyle content with political elements—sometimes referred to as lifestyle politics—appears to be especially prominent in the communication of novel curating actors like influencers (Suuronen et al., 2022). Here, we seek to compare in more detail which topics are consumed through which news sources. We thus ask the following research question:

RQ4: 

Which news topics predict the use of different curating actors for news on social media?

2. Method

To address our research questions, we conducted a cross-sectional online survey in Austria among young adults aged 16 to 25, representing ‘Generation Z’, which consists of individuals who grew up with and are avid users of social media platforms (Kullolli & Trebicka, 2023). The questionnaire, which was distributed in German, was part of a larger data collection in which additional variables were assessed that were unrelated to the present study. The dataset, questionnaire and detailed analysis can be found in the Supplementary Material on OSF3. This project received ethical approval from the institutional review board of the Department of Communication at the University of Vienna [20231126_057].

2.1. Sample

Participants were recruited using a private survey company (Gapfish; Berlin, Germany) in December 2023 and January 2024. In total, the survey company recruited 1789 potential respondents. Of these, 740 completed the survey, while the remaining cases were excluded due to lack of consent, failure to meet quota criteria or prior participation in the survey. Among the completed surveys, 170 failed both attention checks and one respondent did not report their age, resulting in a final analytical sample of N = 569.

The participants’ age ranged from 16 to 25 years (M = 21.21, SD = 2.58). Among the respondents, 49.82 percent identified as male, 49.82 percent as female, while 0.35 percent identified as non-binary/other—thus closely resembling the gender quota for Austrian youth (Statistik Austria, 2024a).

Most participants (65.38 percent) had no migration background, 23.20 percent were born in Austria to at least one foreign-born parent, and 11.42 percent were foreign-born along with their parents. We grouped the latter two categories for the analyses, resulting in 34.62 percent of the participants having some kind of migration background. Again, our sample approximates proportions in this age group in Austria with regard to migration background (Statistik Austria, 2024b). The sample was educationally diverse: 1.23 percent completed primary school only; 12.30 percent lower secondary school (“AHS Unterstufe”/“Mittelschule”); 23.02 percent finished professional school (“Berufsschule”); 41.83 percent finished higher secondary school (“AHS Oberstufe”/“BHS”/“BMS”); and 20.03 percent completed either their Bachelor’s, Master’s or PhD degree. A total of 1.58 percent completed another form of education. Our sample thus showed a slant towards academic participants compared to the proportions in this age group in Austria (Statistik Austria, 2023).

2.2. Procedure

After asking the participants to provide informed consent, definitions of and examples for news, as well as influencers and celebrities, were given. News was defined in accordance with young people’s preferences (Boukes & Boomgaarden, 2015) as including both soft news and hard news: “News can be anything that is currently happening in the world and in Austria, e.g., news about politics, the economy, but also sporting events, music, culture, lifestyle or science”. Influencers and celebrities were defined as stated in the introduction section of this paper4 (Gonzalez et al., 2024; Khamis et al., 2017). Examples for each curating actor were provided in the questionnaire. Following this section, we posed diverse questions regarding young people’s news use.

2.3. Measures

A full item list can be found in Supplementary Material, Item S1. The whole questionnaire can be found in the online repository OSF (Washington, DC, USA)5.

2.3.1. Dependent Variables

Regarding our dependent variables, we asked about the use of different curating actors. To assess the importance of these different news sources, our items were inspired by those in the Digital News Report 2023 (Newman et al., 2023). However, we further differentiated between celebrities and influencers, and added other important news sources for young people. We opted for a 6-point Likert scale to measure the news consumption of these actors. For the question: “How often do you consume news from the following sources,” we defined 1 as ‘never’ and 6 as ‘several times a day’. For journalistic news sources, we created an index of the means of using (1) personal social media accounts of journalists, (2) daily newspapers on social media, (3) tabloids on social media, and (4) the public service broadcaster ‘ORF’ in digital environments6.

2.3.2. Independent Variables

Consistent with our research questions, we captured a range of independent variables. These variables were measured for general news consumption and in later analysis connected to different single actors.

Individual predispositions. First, individual predispositions included sociodemographic variables and news-related/political predispositions. To capture sociodemographic variables, we included several dummy variables: male and other gender to investigate potential gender differences (with female gender functioning as the reference group), migration background (any kind of migration background vs. none), and higher education and professional school (with other types of education functioning as the reference group). Additionally, age was included as a predictor. Regarding news-related and political predispositions, we assessed political ideology by asking the respondents to place themselves on an 11-point Likert scale (European Social Survey, 2022) with 1 being ‘left’ and 11 being ‘right’ (M = 5.80, SD = 2.49). News interest was measured on a 5-point Likert scale (Newman et al., 2023), ranging from 1 = ‘not at all’ to 5 = ‘very much’ (M = 3.57, SD = 1.15). We also measured traditional news use (which included the means of using printed or online newspapers, printed or online tabloids, TV and radio). For the question: “How often do you consume news from the following sources?”, we defined 1 as ‘never’ and 6 as ‘several times a day’ (M = 2.93, SD = 1.16).

Usage motives. To assess different kinds of news consumption motives, we specifically focused on motives that have been derived from the literature on young people’s expectations towards news (e.g., Costera Meijer, 2007). Receiving personally important news ranked first (M = 3.76, SD = 1.21), followed by news being exciting (M = 3.47, SD = 1.26), finding out about opinions of others (M = 3.40, SD = 1.28), having something to talk about (M = 3.30, SD = 1.33), receiving humorous news (M = 3.12, SD = 1.31) and the aim to get more involved in a political or social issue (M = 2.92, SD = 1.35; Schwaiger et al., 2022; Zimmermann et al., 2022). These motives were assessed by stating: “News can be consumed for several reasons. I am consuming news because…” and answered using a 6-point Likert scale ranging from ‘do not agree at all’ to ‘fully agree’.

Exposure modes. We posed questions drawing from the PINE model (Matthes et al., 2020) to assess whether adolescents actively search for news and, if not, whether they continue to consume news content upon incidentally coming across it. We assessed active searching behavior/intentional exposure by having the participants react to the statement “I actively search for news on social media” (M = 2.66, SD = 1.36). First-level incidental exposure was measured by assessing answers to the item “I do not actively search for news on social media, and when I see any, I don’t look at it further” (M = 2.31, SD = 1.31). Second-level incidental exposure was captured by measuring agreement with the statement “I do not actively search for news on social media, but when I see any, I look at it further” (M = 3.54, SD = 1.27). The responses were given on a 5-point Likert scale ranging from ‘do not agree at all’ to ‘fully agree’.

News topics. To analyze preferred news topics of adolescents, we included frequently discussed issues on social media (e.g., consumer behavior or climate change; Suuronen et al., 2022). We extended this list with topics that adolescents consume via traditional media such as economics (Wunderlich et al., 2022) and other important news topics for young people, e.g., health issues (Swart & Broersma, 2024). We asked for consumption of these topics using a 6-point Likert scale ranging from ‘never’ to ‘several times a day’. After performing a principal component analysis (see Supplementary Material, Item S2), we opted to split news topics into two groups: hard news (M = 2.97, SD = 1.14) and soft news (M = 3.69, SD = 1.19). Hard news consisted of politics, economy/finance, climate/environment, animal rights and social issues. Soft news consisted of sports, brands/products, health/food and entertainment, culture, and music/lifestyle. Technology was omitted due to doubled factorial loadings and since it was difficult to assign it strictly to one category from a theoretical perspective. A detailed overview of the used topics can be found in Supplementary Material, Figure S2.

Control Variables. We additionally included usage of different social media as control variables: here, we included the five most common social media networks according to the Digital News Report (Newman et al., 2023). We opted for a 6-point Likert scale asking about the frequency of using WhatsApp (M = 5.41, SD = 1.12), YouTube (M = 4.44, SD = 1.54), Instagram (M = 5.14, SD = 1.48), Snapchat (M = 4.64, SD = 1.88) and TikTok (M = 4.49, SD = 2.00) ranging from 1 ‘never’ to 6 ‘several times a day’.

2.4. Data Analysis

For H1, we performed a repeated measures ANOVA7, comparing the use of influencers, celebrities, politicians, ordinary people and journalistic sources for news. For RQ1 to RQ4, we conducted a single hierarchical regression per actor type to account for changes in R². We added blocks step by step in the order of the research questions. In the case of each DV, the first block of predictors thus included sociodemographic variables: namely gender, age, migration background and education. The second block comprised political ideology, news interest and traditional news use (i.e., the individual predispositions that are news-related or related to politics). Third, we added the five most important social media networks for young people: WhatsApp, YouTube, Instagram, Snapchat and TikTok; however, we only included them as control variables. Fourth, usage motives were included; fifth, exposure modes (intentional exposure and both types of incidental exposure); and finally, exposure to hard news and soft news. Importantly, this approach does not allow for formal statistical comparisons between the different curating actors. Each source was analyzed in a separate model, and we therefore describe relationships between specific predictors and individual sources without implying that these relationships are meaningfully different across sources. In the hierarchical regressions, predictors within blocks should be understood as analytically distinct but theoretically related. The block-wise modeling strategy serves to assess incremental explanatory power and to organize the analysis. We therefore interpret changes in explained variance across model steps as descriptive indicators of how much each domain contributes beyond the preceding blocks. The results of the ANOVA can be found in Supplementary Material, Table S1, and the hierarchical regression models in Supplementary Material Tables S3–S7.

3. Results

We present the findings in the order of the hypothesis and research questions. Because each research question is examined across five curating actors, this structure highlights how different analytical dimensions, such as predispositions, motives, exposure modes, and topic preferences, relate to actor use. To facilitate comparison across actor types and to provide a clearer overview of the broader pattern, we additionally summarize the main findings for each curating actor at the end of the results section.

H1: 

News sources on social media

Regarding H1, we found that on social media, young people most often used ordinary people for news (M = 3.85, SD = 1.72), followed by influencers (M = 3.55, SD = 1.75) and celebrities (M = 3.22, SD = 1.77). Journalistic sources for news on social media only ranked fourth (M = 2.80, SD = 1.22) and politicians/political activists were the least consumed news source (M = 2.22, SD = 1.51). The repeated measures ANOVA, F (4, 2256) = 149.99, p < 0.001, and pairwise comparisons showed significant differences across all the news sources. Correlations between usage of the different curating actors were all positive and significant, indicating that the respondents who used one type of source frequently also tended to use the others more often (see Supplementary Material, Table S2). Since ordinary people were the most used news source on social media in this sample, we were unable to support H1. For an additional analysis, in which we compared the use of the different curating actors to every single news source that was part of the index for journalistic sources, please consult Supplementary Material, Item S3; for an overview of individual sources, see Supplementary Material, Figure S3.

RQ1: 

Individual predispositions

Regarding RQ1, traditional news use was a positive predictor for using journalistic sources (b = 0.41, SE = 0.04, p < 0.001), politicians (b = 0.33, SE = 0.05, p < 0.001), ordinary people (b = 0.17, SE = 0.07, p = 0.015), influencers (b = 0.20, SE = 0.07, p = 0.004) and celebrities (b = 0.22, SE = 0.07, p = 0.001) for news on social media. Controlling for participants with other genders, the results indicated that females were more likely to consume news content of ordinary people (b = −0.45, SE = 0.14, p = 0.002), influencers (b = −0.54, SE = 0.15, p < 0.001), and celebrities (b = −0.36, SE = 0.14, p = 0.012) than males. Having a migration background negatively predicted the use of influencers (b = −0.30, SE = 0.14, p = 0.036) for news. Higher education (higher secondary school and bachelor’s, master’s or PhD degree) positively predicted use of journalistic sources (b = 0.30, SE = 0.11, p = 0.008) and ordinary people (b = 0.43, SE = 0.20, p = 0.030) for news compared to people with lower levels of education. Professional school positively predicted the use of ordinary people (b = 0.51, SE = 0.22, p = 0.022) compared to other types of education for news. Identifying as politically left predicted the use of journalistic sources for news (b = −0.04, SE = 0.02, p = 0.009).

RQ2: 

Usage motives

Seeking out information to get more involved in a social or political issue positively predicted news use of politicians (b = 0.11, SE = 0.04, p = 0.007). “Having something to talk about” as a reason for consuming news negatively predicted the use of ordinary people for news (b = −0.14, SE = 0.06, p = 0.020). The desire for opinionated content predicted influencer use (b = 0.14, SE = 0.06, p = 0.019) and consumption of news content of ordinary people (b = 0.17, SE = 0.06, p = 0.006). Wanting personally important news negatively predicted news content use of politicians (b = −0.11, SE = 0.05, p = 0.020), but positively predicted the use of news content curated by ordinary people (b = 0.15, SE = 0.06, p = 0.016).

RQ3: 

Exposure modes

Active searching for news/intentional exposure only predicted use of sources of the journalistic flow (b = 0.09, SE = 0.03, p = 0.009). First-level incidental exposure did not significantly predict any news sources. Second-level incidental exposure was positively related to the use of influencers for news (b = 0.17, SE = 0.06, p = 0.004).

RQ4: 

Hard and soft news consumption

Soft news consumption negatively predicted the use of politicians for news content (b = −0.13, SE = 0.05, p = 0.017) and positively predicted use of ordinary people (b = 0.32, SE = 0.07, p < 0.001), influencers (b = 0.56, SE = 0.07, p < 0.001) and celebrities (b = 0.53, SE = 0.07, p < 0.001) for news. Hard news consumption positively predicted the use of journalistic sources (b = 0.29, SE = 0.05, p < 0.001) and politicians (b = 0.55, SE = 0.06, p < 0.001) for news.

3.1. Summary of Results by Curating Actor

Journalistic sources. Use of journalistic sources was positively associated with traditional news use, higher levels of education, intentional news exposure, and hard news consumption. It was also more common among the left-leaning respondents.

Politicians. Use of politicians as curating actors was positively related to traditional news use, motivation to become more involved in a social or political issue, and hard news consumption. By contrast, receiving personally important news and soft news was negatively associated with this actor type.

Ordinary people. Ordinary people emerged as the most frequently used source of news on social media. Their use was positively associated with traditional news use, receiving personally important news, opinion-oriented news use, and soft news consumption, but negatively with the motive of having something to talk about. The female respondents, as well as the respondents with higher or professional school education, were more likely to rely on ordinary people for news.

Influencers. Influencer use was positively associated with traditional news use, opinion-oriented news use, second-level incidental exposure, and soft news consumption. The female respondents were more likely than the male respondents to use influencers for news, whereas respondents with a migration background were less likely to do so.

Celebrities. Celebrity use was positively associated with traditional news use and soft news consumption. The female respondents were more likely than the male respondents to rely on celebrities for news.

Across all the models, traditional news use was the most consistent positive predictor, while topic preferences clearly differentiated actor types: hard news consumption predicted use of journalistic sources and politicians, whereas soft news consumption predicted use of ordinary people, influencers, and celebrities.

3.2. Contribution of Predictor Blocks to Explained Variance and Their Internal Correlations

Looking across the hierarchical models, the contribution of predictor blocks varied across curating actors. Overall, the largest increases in explained variance tended to occur when adding news-related and political predispositions and, in several models, topic preferences. News-related and political predispositions explained much of the variance in consuming news from journalistic sources, politicians, and celebrities, whereas topic preferences were particularly important for politicians, influencers, and celebrities. For ordinary people, explanatory power was distributed more evenly across several blocks, including predispositions, social media use, and usage motives. For a block-wise overview of incremental explained variance across the hierarchical models, see Supplementary Material, Table S8. To provide greater clarity on the empirical associations among predictors within the same analytical blocks, we added within-block correlation tables in Supplementary Material, Tables S9–S14. These correlations should be interpreted with caution, as some variables differ in their level of measurement and several sociodemographic variables are dummy-coded.

4. Discussion

In this study, we assessed the use of different news sources among young people, drawing from the curated flows concept (Thorson & Wells, 2016). To this end, we conducted a survey among N = 569 young people in Austria aged between 16 and 25 years. In our analyses, we regressed the use of different news sources on social media on variables such as usage motives, individual predispositions, and differential modes of exposure. We were thereby able—for the first time in extant research—to offer a comprehensive picture of young people’s use of different news sources on social media and the variables that may predict consumption from specific curating actors.

First, somewhat unexpectedly, ordinary people emerged as the most used source of news among young people on social media, although this aligns in part with extant research identifying that social curation shapes much of young people’s news exposure on social media (e.g., Bene, 2017). During late adolescence and young adulthood, close social contacts are particularly important reference points for individuals (Bergström & Jervelycke Belfrage, 2018), possibly explaining why they also emerge as important curators of news. Additionally, the prominence of ordinary people may point to the broader importance of interpersonal and semi-private communication contexts in young people’s social media-based information environments. This is particularly relevant because, in some contexts, messaging apps such as WhatsApp may function as part of young people’s broader social media repertoires, even though they differ from more public-facing platforms in their communicative structure. In such environments, news may become visible not only through original producers, but also through friends, family members, peers, or other ordinary users who share, comment on, or react to current information (Matassi et al., 2019; Thorson & Wells, 2016). However, because our study measured WhatsApp use only as general platform use and not as a news-specific channel, we cannot determine whether WhatsApp contributed to the prominence of ordinary people in our findings. Future research should therefore investigate more directly how young people classify messaging apps and how they use them for news.

However, our findings also raise the question of what young people understand as ‘news’ in social media environments. In the survey, we deliberately used a broad definition of news aligned with young people’s preferences, including both hard and soft news topics such as politics and the economy, but also sports, music, culture, lifestyle, and science (Boukes & Boomgaarden, 2015). This broader understanding of news is not only aligned with young people’s news preferences but also supported by recent audience-centered research showing that news on social media is not always experienced according to traditional journalistic boundaries. Swart and Broersma (2024) identify four ways in which young people make sense of news on Instagram: traditionalism, where news is understood in relation to established journalistic outlets; compartmentalization, where users distinguish between different levels or types of news; homogenization, where news is understood broadly as novel or recently encountered information; and reconceptualization, where classic news values such as relevance or objectivity are reinterpreted in relation to users’ everyday social media experiences. These findings suggest that a broader conceptualization of news is necessary to capture how young people encounter and interpret news-like information in social media environments. At the same time, self-reported survey data cannot fully capture how the respondents interpreted this concept in practice. Especially in social media contexts, ‘news’ may be understood not only as current information of general public relevance, but also as socially shared and personally relevant updates circulating in users’ networks. This may partly explain the prominence of ordinary people as curating actors in our findings. A useful follow-up step for future research would therefore be to investigate more directly how young people define and distinguish news from other kinds of information in their everyday media use.

Influencers and celebrities ranked only second and third in terms of popularity as news sources among youth. This may be because algorithms tend to prioritize stories and posts from friends over those from influencers, making content from peers more frequently visible to young people (DeVito, 2017). By differentiating between influencers and celebrities, this study made an even clearer distinction between two important types of online personalities—discovering that influencers may be slightly more relevant for youth than traditional celebrities. This finding may be due to the notion that influencers, whose success depends on continuously producing posts and being visible on social media (Khamis et al., 2017), are simply more active content creators than traditional celebrities. The relevance of influencers as news sources also confirms qualitative work that suggests that young people increasingly consume news from these actors, not least because it is viewed as more suited for their information preferences (Wunderlich et al., 2022).

In contrast, traditional news sources such as social media channels of newspapers or the public service media emerged as less popular news sources in the context of social media. This finding is in line with other work suggesting that young people rely increasingly less on these sources when browsing these platforms (Schwaiger et al., 2022; Tamboer et al., 2022; Wunderlich et al., 2022). Yet, politicians rank even lower, suggesting that their communication has comparably little relevance for youth.

Addressing RQ1, we identified several important predispositions for using specific curating actors as a source of news. In line with research on usage intention of online news apps, we showed that traditional news use was a predictor of all the curating actors on social media for news (Guo, 2024). News interest did not account for additional variance in using curating actors on social media for news, likely due to its strong connection with general news consumption (Weeks et al., 2022). Instead, motive- or exposure-related factors may better explain the remaining variation.

Based on our results, we can conclude that female participants may use influencers and celebrities for news on social media to a greater extent than male respondents and participants of other genders. When investigating general usage patterns, previous studies have also shown that female users follow and engage more with influencers than male users (Leite et al., 2023). However, a different study suggests that being male is positively related to relying on influencers for political information (Harff & Schmuck, 2025b). In contrast to that study, we did, however, investigate consumption of news in general and not specifically the use of political information from these sources. Moreover, we did not ask about specific influencers and celebrities and their importance as information sources, but about these actors as a group—potentially explaining the divergent findings. For celebrities, this pattern of women being more receptive to their news is novel, and we thus provide first insights into gender differences that emerge when exploring consumption of news shared by traditional celebrities. We also found that young people with a migration background may be less likely to use influencers as news sources. Based on the principle of self-categorization rooted in social identity theory (e.g., Tajfel & Turner, 1986), it could be expected that people with a migration background prefer to follow influencers who also have a migration background. Some popular influencers in Austria have a migration background (e.g., dariadaria, Austriankiwi), the most followed, however, do not (Hansen, 2024; Likeometer, 2024), which could provide a partial explanation for this phenomenon. Last, our finding that left-leaning people tend to use more traditional journalistic sources corroborates the tendency of right-leaning people to be critical of traditional journalistic media (Haller & Holt, 2019).

Meanwhile, our results regarding RQ2 can especially confirm one central motive for using influencers, namely seeking opinionated content. This result indicates that influencers may offer such content for youth, thereby providing orientation in relation to specific topics and helping young people build up attitudes on certain topics, as also suggested by prior research (Schwaiger et al., 2022; Wunderlich et al., 2022). With both these previous studies using qualitative methods, our survey provides quantitative support for the notion that influencers are popular news providers for young people by acting as ‘commentators’ (Harff & Schmuck, 2025a). On the other hand, news consumption through ordinary people is positively predicted by seeking personally relevant news. Networks of friends are often established through shared interests (Li et al., 2020), and news curated by peers or other ordinary people may thus be considered a close match with young people’s personal beliefs and values. For using politicians for news on social media, receiving more information to engage with a social or political topic emerged as a central usage motive. Yet, the desire for personally important information in news was a negative predictor of this dependent variable, which could indicate that politicians—due to their focus on formal political topics—are less able to provide relatable content that connects with lifestyle, identity or other aspects that are vital components of young people’s everyday lives (Suuronen et al., 2022).

Regarding RQ3, our results suggest that young people actively search for journalistic sources on social media, while the use of influencers is connected with second-level incidental exposure. Thus, influencer content may often be incidentally found and is then processed in-depth (Matthes et al., 2020). Since influencer content often matches young people’s preferences for news (Harff & Schmuck, 2025a)—i.e., providing information that is opinionated, modern, humorous and personally relevant (Marchi, 2012; Wunderlich, 2023)—young people may be motivated to further engage with it. This distinction between active search and incidental exposure also reflects the role of curation: active search is linked to personal curation, whereas incidental exposure may be shaped by algorithmic curation, for example, when content is prioritized in users’ feeds based on previous engagement (Thorson & Wells, 2016). Incidental news exposure can furthermore lead to opportunities for generally less-interested people to engage with news (Weeks et al., 2022) and can, in consequence, lead to more (online) political participation among young people (Nanz & Matthes, 2020, 2022). However, since content from influencers is also occasionally linked to misinformation (e.g., Mena et al., 2020), further studies should look into the potential negative effects of second-level incidental exposure to their messages. Future research may also investigate how baseline exposure to non-news content from different sources shapes incidental news consumption and subsequent engagement.

In line with existing research (Gonzalez et al., 2024; Reinemann et al., 2016), we also found that use of soft news was linked to consumption of news from influencers, celebrities and ordinary people, while exposure to hard news was connected with higher reliance on traditional journalistic news sources and politicians for news. Exposure to soft news, however, does not necessarily mean that topics are fully nonpolitical: they may still reflect societally relevant issues, e.g., sports news may cover racist attacks on football players. This notion is also reflected by the concept of lifestyle-based politics, which concerns a broader range of issues that result from the politicization of principally nonpolitical issues related to topics such as sports, health or entertainment (Suuronen et al., 2022). Interestingly, when we conducted regression models with only one news category included, hard news consumption was also a positive predictor of influencer and celebrity use for news, while soft news consumption predicted journalistic sources positively, suggesting that both hard and soft news may also be prominent in the content of these sources, but just to a lesser extent.

4.1. Limitations

Of course, our study does not come without its limitations. Given that our research is based on cross-sectional survey data, we cannot make inferences about causality. Thus, some of the relationships we proposed here may be reversed or reciprocal in nature, such as the link between traditional news use and use of different news sources on social media. We also note that the predictors we used here are formulated in a general way and thus do not pertain to a specific source type. Therefore, associations we detected should be approached with some caution: for example, we found a relationship between second-level incidental exposure to news and consumption of news from influencers. However, we cannot say for certain that youth are incidentally exposed to information from them. Yet, the link we found here is highly plausible from a theoretical perspective, since young people are known to mainly use influencers for reasons other than news (Croes & Bartels, 2021), and may thus be primarily accidentally exposed to their news-related content. Conversely, it is intuitive that journalists would be followed specifically because of the news content that they offer.

Another limitation of this study concerns the measurement of the usage of ordinary people for news, for which respondents’ interpretations may vary. When news content produced by journalistic outlets is shared or reposted by friends, respondents may differ in whether they attribute the exposure to the original producer or to the curating intermediary. This ambiguity is not unique to our study, as existing surveys often rely on broad source measures and do not distinguish between original producers and curating actors (e.g., the Digital News Report asking for ‘ordinary people’ who are being paid attention to for news; Newman et al., 2025) or even equate news consumption on Facebook with news consumption from peers (Bene, 2017). Future research should therefore develop more fine-grained measurement approaches that explicitly differentiate between production and curation in social media news use.

A natural limitation of trying to assess consumption of news from different sources in surveys is reliance on self-reported data, which may be somewhat inaccurate. Young people may have difficulty indicating how often they see content from these different sources. In addition, it may be hard for them to judge how often they are exposed to a source type in general. In relation to a specific person (e.g., Donald Trump), frequency of exposure may be easier to report. Yet, by employing such an approach, it becomes a challenge to track the importance of entire actor groups (e.g., politicians) as news curators for young people.

In any case, shortcomings related to self-report data could be addressed in research that uses large-scale data donations or web/mobile tracking data, whereby exposure could be more accurately measured. Still, we believe that our survey study presents some important novel insights into why certain sources may be used by young people and how they are differentially encountered by them. To expand on these insights, we encourage scholars to replicate this study in other contexts in order to explore whether findings are generalizable beyond the specific country case.

4.2. Implications

Despite these limitations, this study helps develop a clearer picture of youth news consumption, and shows how young people’s individual predispositions, information preferences and different exposure modes relate to the use of different curating actors on social media. These findings not only provide a deeper understanding of news curation on social media but also have practical implications for the news actors that we focused on here. It is, for instance, noteworthy that preferences for personally relevant news are negatively associated with using politicians for news. Youth generally criticize that they feel that their interests and needs are not acknowledged by politicians (Dahl et al., 2018). The findings of this study suggest that this lack of representation may potentially extend beyond electoral politics to politicians’ communication on social media. Politicians may thus want to tailor their messages to young people and more frequently address topics that they value to elevate their position in the news diets of youth. Failure to do so might result in increased political apathy among youth.

This relatable type of content can currently be offered by ordinary people. Considering that ordinary people and influencers are also generally popular as news sources, this article further offers theoretical implications, highlighting that—despite the presence of journalistic actors on social media—interpersonal influence/social curation may have the highest relevance in this environment (Lazarsfeld et al., 1968). This might indicate a shift in the agenda-setting power of traditional media outlets, as audiences increasingly rely on content curated by peers or influencers.

Our findings also highlight that many of the actors relevant in young people’s social-media-based news repertoires—such as ordinary people, influencers, celebrities, or algorithmically shaped feeds—may not primarily engage in original journalistic reporting. However, the curated flows framework does include the “production, selection, filtering, annotation, and framing of content” (Thorson & Wells, 2016, p. 310), and compared with journalistic actors, these non-journalistic actors may be especially influential in selecting, recirculating, personalizing, commenting on, and framing information produced elsewhere. This also connects to classic and more recent notions of opinion leadership, as these actors may shape not only which information reaches young audiences, but also how this information is interpreted, evaluated, and socially legitimized (Lazarsfeld et al., 1968; Stehr et al., 2015). This may have mixed implications for the quality of public information. On the one hand, non-journalistic curators can make news more accessible, relatable, and engaging for young audiences. On the other hand, a stronger reliance on such actors may contribute to more selective, opinionated, simplified, or decontextualized forms of information exposure. For democratic public communication, the question is therefore not only from which institutions news originates, but also whose selections, annotations, and interpretive frames shape how news reaches and matters to young people in everyday social media use.

Last, journalistic actors were still used in a rather traditional way, in that they were actively sought out for hard news, and especially by left and highly educated users. Given that both forms of incidental exposure were also unrelated to the use of journalistic sources and politicians for news on social media, we can conclude that traditional journalistic actors may struggle to become part of young people’s social media experiences unless their content is specifically looked for. These actors could try to imitate influencers’ techniques when posting content, as influencers not only seem to make their way into young people’s feeds through incidental exposure, but also consequently manage to capture their attention. Strategies that journalistic actors could employ include a focus on more opinionated content and preparing hard news in ways that speak to the everyday lives and experiences of young people. At the same time, the pressure to adapt to these platforms and communication logics might challenge the established journalistic standards such as neutrality or nuance and may consequently affect media trust. However, ending on a more positive note, these developments and the popularity of these competitors online could push journalistic actors to develop innovative formats that translate hard news into accessible, relatable and engaging content, potentially revitalizing journalism’s role among younger audiences and elevating their visibility on social media.