Never Mind the Source? The Drivers of User Engagement With Politicians’ Online News-Sharing Posts in 15 European Countries

Abstract

Politicians on social media curate the flow of information their followers are exposed to, and they share news as a way to do so. Previous research has shown that they are strategic in their selection of news stories—they select specific news items and outlets to appeal to their online audiences. The reach of politicians’ news-sharing posts and their possible effects, are, however, influenced by the extent to which individual social media users engage with them. Combining comparative survey data (N = 22,145) with Facebook posts (N = 21,113) by 2,021 MPs in 15 European countries, this article zooms in on user engagement with politicians’ online news-sharing posts. Results show that, in general, news content and message cues can drive user engagement, and that news outlet characteristics do not. Nevertheless, the ideological closeness between a politician and the news outlet, which was found to predict politicians’ online news-sharing behavior, does positively influence reactions in media systems with high levels of political parallelism. These findings shed a new light on partisan selective exposure following politicians’ news recommendations in the digital information environment.

Keywords

news sharing social media user engagement politicians news media

Introduction

Although social media allow politicians to communicate with their electorates without the interference of news media professionals, politicians also highlight news articles online. Doing so, they act as secondary gatekeepers (Singer 2014), suggesting news items to their followers, and thus curating the news and information those followers are exposed to (Thorson and Wells 2016). Previous research has shown that politicians do not share a representative sample of news coverage in general, as the news they suggest to their online followers is a strategic and select subset of articles and outlets that appeal to their electorates and online audiences (Buyens, Van Aelst, and Vaccari 2024; Heidenreich et al. 2022).

The selectivity of politicians’ online news sharing is especially relevant in the context of the digital information environment, in light of the multiplication of news outlets and the fragmentation of news consumption (see Van Aelst et al. 2017). The internet allowed for the establishment of a variety of new, alternative news outlets and for a diversification of the ways in which people access or encounter news. Politicians, especially those connected to radical and right-wing parties, have been shown to play a role in the distribution of alternative news (see Buyens, Van Aelst, and Paulussen 2024), and they are more likely to share news when it is published by an outlet whose audience aligns with their ideology (Buyens, Van Aelst, and Vaccari 2024; Heidenreich et al. 2022). Considering the risk of partisan selective exposure to news on social media, it is important to know the impact of politicians’ online news-sharing posts. In this study, the impact of those posts will be gauged by studying social media user engagement.

Because the digital distribution of information is based on “virality” (Klinger and Svensson 2015), the ways in which social media users engage with online content are crucial for its reach. The premise of this article is that, while politicians’ online audiences are rather limited, their posts may reach and impact large groups of social media users indirectly, because of user engagement. Therefore, this research is contextualized in recent scholarly literature that translates of the seminal two-step flow of information (Katz and Lazarsfeld 1955) to the digital information environment, illustrating that “followers are opinion leaders” (Karlsen 2015) and, thus, that individual social media users can be influential in relaying information to their networks.

When exposed to news they did not actively seek out, social media users can determine its informational utility, based on a variety of different characteristics, like those of the news content, the news curator, and the news provider (Boehmer and Tandoc 2015; Kümpel 2019). Building on previous research on the factors that predict user engagement with politicians’ social media posts (see Bene et al. 2022), user engagement with online news more generally (Kümpel 2019), and politicians’ news-sharing behavior (Buyens, Van Aelst, and Paulussen 2024; Buyens, Van Aelst, and Vaccari 2024; Heidenreich et al. 2022), this article zooms in on the drivers of user engagement with politicians’ news-sharing posts specifically. To do so, Facebook posts made by MPs in 15 European countries are analyzed in parallel with a comparative survey on political preferences and news consumption in those same countries. The main research question that guides this study is as follows.

RQ: Which characteristics explain user engagement with politicians’ online news-sharing posts?

Politicians as Online News Curators

Digital and social media have significantly impacted the interrelationships between politicians and news media. One of the most prominent changes is that social media allow politicians to publicly use and interact with journalistic output for the advancement of their own political agenda (Ekman and Widholm 2015). Politicians who share news items on their social media pages can affect the distribution of the news because it is highlighted to relatively large groups of social media users. When sharing news online, politicians act as secondary gatekeepers (Singer 2014), influencing the news exposure of primary and secondary groups of online connections (Karlsen 2015; Vaccari and Valeriani 2015). Thorson and Wells (2016) identified politicians as actors who have strategic incentives to curate the information their followers are exposed to in the digital information environment. The authors conceptualize curation as “the production, selection, filtering, annotation, or framing of content” (Thorson and Wells 2016: 310). When politicians share news on social media, they curate the flow of online news by selecting, annotating, and framing news items, effectively influencing if and how social media users are exposed to it.

Politicians’ online news-sharing behavior in the European context is, at least to some extent, explained by ideological factors, as research has shown that politicians are more likely to share news published by outlets that are favored by their electorates (Heidenreich et al. 2022), and they share more news from ideologically aligned news sources (Buyens, Van Aelst, and Vaccari 2024). This is true across parties and across countries. Moreover, this behavior is linked to the concept of political parallelism, or the extent to which news media in a given system are connected to political parties. These connections are traditionally “by organization, loyalty to party goals and the partisanship of its readers” (Seymour-Ure 1974: 173), but they are in contemporary media systems considered to be shaped by “the degree to which audiences consume media according to their political preferences” (Brüggemann et al. 2014: 1040). Comparative research has indeed shown that the ideological connection between politicians and a news outlet more strongly predicts their news-sharing behavior in countries with higher political parallelism (Buyens, Van Aelst, and Vaccari 2024). This means that politicians in media systems with higher levels of political parallelism may further reinforce trends of news audience isolation and polarization.

Moreover, radical and right-wing politicians show similar tendencies of outspoken, ideologically selective news sharing (Buyens, Van Aelst, and Vaccari 2024). This is ever more relevant because digital media have allowed for the establishment and proliferation of novel and alternative news outlets, which proclaim or are perceived to be corrective, opposing the perceived dominant mainstream media in a given system (Holt et al. 2019). Alternative news media have been shown to cater to ideologically homogenous audiences (Bailey et al. 2007; Buyens and Van Aelst 2022; Müller and Schulz 2021). Studying the predictors of user engagement with politicians’ online news-sharing posts, this article will consider party radicalism and political parallelism as moderating variables to study whether citizens, as politicians, take them into consideration when engaging with online content.

Online News Exposure

The increased importance of online intermediaries and curators, like politicians, and the proliferation of alternative and ideologically outspoken news outlets illustrate the growing fragmentation of contemporary online information environments. Social media are the main points of access for news for many people (Newman et al. 2023). On the one hand, those platforms foster incidental exposure to news, showing—possibly counter-attitudinal—content to users who did not actively seek it out (see Tewksbury et al. 2001). On the other hand, social media have been said to boost selective exposure to news, leading users to consume supportive information over unsupportive information as a means of reducing dissonance (Iyengar and Hahn 2009). While Klinger and Svensson (2015) argue that the network media logic could also lead to more selective exposure, other studies have shown that social endorsements also play a crucial role in incidental exposure to news on social media (Anspach 2017; Messing and Westwood 2014). It is likely that the exposure to selective as well as counter-attitudinal news on social media following social endorsements is largely dependent on the ideological heterogeneity of one’s social media network. For example, one could hypothesize that social endorsements in ideologically homogenous social media networks reinforce partisan selective exposure.

The multiplication of the supply of news outlets and the fragmentation of news consumption in the digital information environment has raised concerns about “echo chambers” (see Jamieson and Cappella 2008; Sunstein 2007), in which people are mostly exposed to like-minded views, and “filter bubbles” (Pariser 2011), in which people are supplied with information they will likely agree with, which could amplify ideological segregation (see Flaxman et al. 2016). Empirical research has shown that news consumption on social media is associated not only with higher ideological segregation but also with exposure to counter-attitudinal information (Flaxman et al. 2016). Nevertheless, Weeks et al. (2017) found that incidental exposure to counter-attitudinal information could lead partisans to intentionally and selectively seek out pro-attitudinal information, and that those partisans are more likely to share political information after pro-attitudinal exposure.

When news is shared on social media, it will not just reach the direct connections of the one who shared it. Rather, it will spread beyond those primary networks and lead to incidental exposure of users who were not actively seeking it out. If news is shared by a politician, not only their direct followers will be exposed to it. Following the logic of virality (Klinger and Svensson 2015), other citizens may also see it (see Karlsen 2015; Vaccari and Valeriani 2015), boosting the number of users who are exposed to this content. The spread of politicians’ news-sharing posts, as well as their impact on citizens’ exposure to news, however, is affected by the extent to which social media users engage with them.

User Engagement With News-Sharing Posts

The reach and impact of politicians’ social media communication depends not only on the amount of followers they have but also on the user engagement their posts can trigger, which can be seen as “popularity cues” (see Haim et al. 2018). Thus, even though politicians do not seem to attract very large and diverse groups of followers directly (Nielsen and Vaccari 2013), their posts might indirectly find their way to other online networks. Therefore, it is of importance to know which elements affect user engagement with politicians’ news-sharing social media posts (1) to understand the effectiveness of politicians’ strategic online communication for generating more traffic to their social media platforms, and (2) to better be able to gauge to what extent politicians’ strategic information curation affects what news content social media users are exposed to.

User engagement decisions are informed by different elements of the social media post. Boehmer and Tandoc (2015) pointed out source, message and user characteristics as the three main groups of factors that predict retweeting intentions. Kümpel (2019) proposed a more elaborate framework that is focused on online news-sharing posts. She identified four information scent cues that inform engagement decisions, which are the perceptions of characteristics of the news provider, the news content, the news curator, and the news recommendation (Kümpel 2019: 170).

In this article, I will focus on news-sharing posts by politicians and take into account (1) the news provider, or the news outlet that published the item that is shared, and how it relates to the news curator, (2) the message that is added by the politician in their news-sharing post, and (3) the news content, to study which of these elements affect user engagement. More specifically, I will study whether the ideological closeness between the news curator and the news provider affects user engagement and test the moderating influence of party radicalism and political parallelism.

Expressions of user engagement depend on the platform one looks at. For example, Facebook uses likes, comments, shares, and emotional reactions, and X uses likes, comments, and reposts. Moreover, previous studies have operationalized user engagement as individual expressions (likes, comments, reposts. . .) or an aggregate measure of those. This paper does not differentiate between expressions of user engagement in the hypotheses, but the analyses will explain the number of likes, shares, comments, and emotional reactions individually, to facilitate the interpretation of the effects of those expressions on the distribution of information on social media.

News Outlet

News brand images and perceptions affect individuals’ selection of and exposure to news content in high-choice media environments (Arendt et al. 2019; Iyengar and Hahn 2009; Medders and Metzger 2018; Winter and Krämer 2012, 2014). Therefore, this article focuses on two characteristics of the news provider that could influence social media user engagement with politicians’ online news-sharing posts.

Firstly, I will focus on ideological closeness to the politician who shared it. Research has shown that politicians share more news that is published by outlets of which the readership aligns with the politician’s own ideology (Buyens, Van Aelst, and Vaccari 2024; Heidenreich et al. 2022). One possible reason for this finding is that politicians appeal to the familiarity that their online audiences have with that outlet and the trust they have in it. When a political actor shares news published by an outlet of which the audience aligns with their ideology, it is expected that the general attitude of the politician’s followers towards that news outlet positively influences user engagement intentions, leading to higher user engagement with that social media post. In other words, the ideological closeness between the politician and the readership of the news outlet will positively affect user engagement.

H1: User engagement with politicians’ news-sharing posts will be higher when the ideology of the readership of the news outlet aligns with that of the politician.

Building on earlier scholarship that shows that the ideological closeness between a politician and a news outlet is a stronger predictor of news sharing (1) among radical party politicians, and (2) in countries which are characterized by higher political parallelism (Buyens, Van Aelst, and Vaccari 2024), this article will explore if those variables also moderate the relationship between ideological closeness and user engagement. Politicians reinforce partisan selective exposure in contexts that are more vulnerable to it to begin with, but do social media users close the vicious circle and propel those ideologically aligned news-sharing posts by engaging more with them?

RQ1: Does party radicalism moderate the relationship between ideological closeness and user engagement with politicians’ news-sharing posts?

RQ2: Does political parallelism moderate the relationship between ideological closeness and user engagement with politicians’ news-sharing posts?

Secondly, this article considers news outlet popularity. When incidentally exposed to a news article on social media, social media users take into account the perceived societal relevance of the outlet that published it when deciding whether or not to read it (Kaiser et al. 2021), as well as its perceived credibility (Winter and Krämer 2014). In other words, social media users base their decisions to engage with online news not only on their own familiarity with the outlet, but also on its impact in society, more generally. Thus, it is expected that the popularity of a news outlet, or the extent to which it is used in a given media system, will be positively associated with user engagement when a politician shares a news item published by that outlet.

H2: User engagement with politicians’ news-sharing posts will be higher when the news outlet that is shared is more popular.

Message

On top of source cues and curator characteristics, the content of the news-sharing posts, made up of limited news item titles and lead information, and an optional message added by the politician, will affect user engagement. Several characteristics of politicians’ social media posts have been shown to affect different expressions of user engagement. Facebook posts on salient issues were shown to attract more “Anger” reactions (Eberl et al. 2020), although Bene et al. (2022) found mixed results for the issues of immigration and environment when studying reactions, shares, and comments. Most notably, the sentiment expressed in Facebook posts by political actors matters for user engagement (Peeters et al. 2023). Negative tone leads to more comments and shares (Bene 2017; Heiss et al. 2019) and more reactions (Bene et al. 2022). Although most studies do not find the same results for positive sentiment, Peeters et al. (2023) conclude that positive and negative sentiment similarly boost user engagement. Moreover, Eberl et al. (2020) found that it does significantly decrease “Anger” reactions, and it increases “Love” reactions. Jost (2020) found that posts that contain higher emotionality attract more “Anger” reactions and likes, while they attract fewer “Love” reactions. Heiss et al. (2019) similarly found that posts that contain negative or positive emotions attract more likes and comments, while the latter also attract more shares. Although findings on the sentiment of politicians’ online posts and user engagement are mixed, the abovementioned research illustrates that sentiment and emotional language are appealing in the sense that they trigger some sort of reaction among social media users. In this article, I will add to this literature by testing if the sentiment of politicians’ news-sharing posts affects user engagement, and thus if politicians’ use of positive and negative language affects the popularity and the reach of the news they share. It is expected that the sentiment of a politician’s news-sharing post, both positive and negative, will boost user engagement with that post.

H3a: User engagement with politicians’ news-sharing posts will be higher when message in the post contains more negative sentiment.

H3b: User engagement with politicians’ news-sharing posts will be higher when the message in the post contains more positive sentiment.

News Item

Lastly, this study accounts for the content of the news. Politicians’ news-sharing posts do not only contain their own message, but they also include a link to a news item that, usually, is displayed with an image and a (shortened version of the) title and the lead of the news article. User engagement with news on social media has been studied through the lens of newsworthiness (Trilling et al. 2017; Weber 2014; Ziegele et al. 2014). This study will control for the sentiment of the news shared in the post to test if social media users’ engagement decisions are informed by the content characteristics of the news.

H4a: User engagement with politicians’ news-sharing posts will be higher when the news preview contains more negative sentiment.

H4b: User engagement with politicians’ news-sharing posts will be higher when the news preview contains more positive sentiment.

Method

Case Selection and Data Collection

To study user engagement with politicians’ news-sharing social media posts across media systems, I select fifteen countries in which varying levels of political parallelism have been identified (Hallin and Mancini 2004; Humprecht et al. 2022). The sample includes seven democratic corporatist media systems (Austria, Switzerland, Germany, Denmark, the Netherlands, Norway, and Sweden), four hybrid systems (Humprecht et al. 2022), two of which were positioned by Hallin and Mancini between the democratic corporatist and polarized pluralist models (Belgium and France), one as a polarized pluralist system (Italy), and one as a liberal model (the United Kingdom), and four polarized pluralist systems (Hungary, Poland, Romania, and Spain). This broad range of European countries, which have been identified as having different media systems with varying levels of political parallelism, allows for a comparative approach to user engagement with politicians’ news-sharing posts online in diverse political and journalistic contexts. Comparing the predictors of user engagement with politicians’ news-sharing posts across countries allows for a test of the consistency of the relationship and provides an opportunity to differentiate between contexts, which is done by looking at political parallelism on the level of the media system.

Facebook posts made by all MPs in the abovementioned fifteen European countries between October 01, 2021 and October 01, 2022¹ were collected through Meta’s social monitoring platform CrowdTangle. All members of the lower house of representatives (or the house of representatives in unicameral systems) as of October 01, 2021 (N = 5,281) were included. Of that group of MPs, 3,612 had a personal, accessible public Facebook page at the time of the data collection.

To include insights on political preferences, news consumption, and political parallelism, the content analysis data is supplied with a survey² fielded by the Network of European Political Communication Scholars (NEPOCS) in December 2019. Data were collected with the support of Dynata, and population quotas were used for age, gender, and metropolitan region. For each of the countries, the sample is fairly representative of the population (see Castro et al. 2022; https://nepocs.eu/research/). The number of respondents per country ranges from 1,393 to 1,605; exact numbers are shown in Table 1 in the Supplemental Material file. Questions in the survey gauge political knowledge, opinions, attitudes, and behavior. I focus on measures of respondents’ online news consumption (whether they accessed a news outlet—up to 16 per country—in the last 30 days), ideological self-placements (left-right on an 11-point scale), and voting preferences (propensity to vote for a party on an 11-point scale). The exact question wording of the items used for this study can be found in Table 4 in the Supplemental Material file. The inclusion of this survey data, instead of other secondary data sources, has the added benefit that all relevant variables are measured at the same time and with the same respondents, which in turn allows for the calculation of measures on the party, outlet and country level.

Variables

User Engagement

User engagement metrics were collected through CrowdTangle and serve as the main dependent variable in this study. In the analyses, the focus is on likes, comments, shares and reactions (sum of the number of likes, “love,” “wow,” “haha,” “sad,” “angry,” “care”).

Independent Variables

News Outlet Left-Right Position

The ideological position of the audience of a news medium was calculated by averaging the self-reported left-right positioning of all respondents who reported having visited the news outlet online in the last 30 days. News outlet’s left-right position ranges from 0 (extremely left-leaning) to 10 (extremely right-leaning).

News Outlet Popularity

The popularity of a news outlet was calculated as the share of survey respondents who reported having visited that news outlet online in the last 30 days. This variable ranges from 0 to 1.

Party Left-Right Position

Politicians’ party left-right position was calculated by averaging the self-reported left-right positioning of respondents who rated that party highest when reporting their probability to vote for the main parties in the political system of their country. This voter-derived estimate of parties’ left-right positions is strongly correlated with the 2019 Chapel Hill Expert Survey data (Jolly et al. 2022) (r[98] = 0.93, p < 1.1e-16). Table 2 in the Supplemental Material file shows the party left-right scores for each party in the dataset.

Party Radicalism

As an addition to the party left-right position measure, party radicalism gauges the radical nature of a politician’s party. The measure was constructed by first centering the left-right positions around 0, thus transforming the range from 0–10 to −5 to +5, and taking the squared value of that centered left-right position.

Ideological Closeness

The ideological closeness between a politician and a news outlet is measured as the closeness between the politician’s party and the news outlet on a left-right scale. The left-right positioning of the political parties, as well as the news outlets’ audiences, was measured as the average left-right positioning of their electorates and audiences, respectively. This variable was calculated by subtracting 10 from the absolute difference between the politician’s and the news outlet’s left-right position and taking the absolute value of the result of that subtraction. For example, if a politician from a radical right-wing party (L.-R. = 7.19) shares a news item by a right-wing alternative news website (L.-R. = 7.93), the ideological closeness between them would be 9.26 (||7.19–7.93| − 10|). If a radical left-wing politician (L.-R. = 2.80) would share the same news item, the ideological closeness would be 4.87 (||2.80–7.93| − 10|).

l r c l o s e n e s s = | | l r_{p o l i t i c i a n} - l r_{o u t l e t} | - 10 |

Message Sentiment

The sentiment of the message in the post was determined in two steps. In a first step, all non-English posts were automatically translated to English using the Google Translate API. In a second step, the English (translations of the) messages were automatically coded for sentiment using the 2015 Lexicoder Sentiment Dictionary (Young and Soroka 2012). Although recent scholarly debate has highlighted the shortcomings of dictionaries for automated sentiment analysis (Van Atteveldt et al. 2021), it has also illustrated its value (Dun and Soroka 2021). The Lexicoder Sentiment Dictionary was designed for the analysis of political and media texts, and has been applied in similar cases (e.g., Chinn et al. 2023; Soroka et al. 2018). The message positive sentiment and message negative sentiment scores were calculated by dividing the number of positive and negative words, respectively, by the total number of words. This approach makes use of two separate variables, one for positive sentiment and one for negative sentiment, and thus purposefully accounts for the nuance in posts that contain some positive and some negative language. Theoretically, these variables range from 0 (no positive words for message positive sentiment; no negative words for message negative sentiment) to 1 (only positive words for message positive sentiment; only negative words for message negative sentiment).

News Sentiment

The sentiment of the news shared within the post was determined in a similar way to the sentiment of the message. All available information about the news item, like the title and the lead, as shown in a Facebook feed, was stitched together. Then, all non-English data was translated, and all text was automatically coded for sentiment using the Lexicoder Sentiment Dictionary. Two separate variables, news positive sentiment and news negative sentiment, were calculated by dividing the number of positive and negative words, respectively, by the total number of words. The theoretical range of these variables is the same as those of the message sentiment variables.

Political Parallelism

In this study, political parallelism is measured as the extent to which vote choice is explained by news consumption in a given country. In other words, it gauges “the degree to which audiences consume media according to their political preferences” (Brüggemann et al. 2014: 1040), in line with the contemporary expression of the concept in Western media systems. Estimated country-level political parallelism is calculated following the operationalization proposed by Van Kempen (2007: 307), which is widely used in research (e.g., Brüggemann et al. 2014; Humprecht et al. 2022). Analyzing the survey data, we regressed party preference scores, which were measured as propensity to vote for each party in the political system, on online exposure to news outlets (up to sixteen per country) in the last 30 days, for each party in every country. We subsequently calculated country-level political parallelism as the mean of adjusted explained variance scores weighted for each party’s size, measured as the number of seats in parliament. Theoretically, this variable ranges from 0 to 100. However, practically, a political parallelism score of 100 is near impossible because news audiences overlap (Van Kempen 2007). Political parallelism scores in our data range from 2.71 (Norway) to 18.16 (Poland). An overview of country-level political parallelism scores is shown in Table 3 in the Supplemental Material file.

Table 1 shows the main descriptive statistics for the independent and dependent variables included in the analyses below.

Table 1.

Descriptive Statistics for the Variables Included in the Analysis.

Variable	Mean	SD	Median	Min.	Max.	Skew	Kurt.
Number of likes	200.0	763.8	26.0	0.0	29,379.0	12.4	261.6
Number of comments	65.7	234.59	5.0	0.0	7405	9.32	144.23
Number of shares	26.8	133.0	2.0	0.0	5,703.0	16.9	430.3
Number of reactions	264.4	913.8	36.0	0.0	38853	11.9	277.5
Ideological closeness	8.6	0.8	8.8	6.1	10.0	−0.3	−0.7
Message pos. sentiment	0.04	0.04	0.03	0.0	1.0	3.1	29.1
Message neg. sentiment	0.04	0.04	0.03	0.0	1.0	3.0	30.8
News pos. sentiment	0.03	0.03	0.02	0.0	0.3	1.5	3.6
News neg. sentiment	0.04	0.04	0.03	0.0	0.4	1.3	2.1
Number of followers	39,380.4	139,717.5	6,367.0	71.0	2,443,541.0	9.8	127.2
Outlet popularity	0.4	0.2	0.3	0.0	0.8	0.6	−0.8
Party left-right	5.3	1.7	5.5	1.7	7.8	−0.1	−1.4
Party radicalism	3.0	2.3	2.9	0.0	10.7	0.5	−0.4
Political parallelism	7.7	5.2	7.1	2.7	18.2	0.9	−0.3

Statistical Analysis

To test the hypotheses, user engagement expressions with politicians’ online news-sharing posts are analyzed using multilevel regression models with random intercepts for politicians, parties, and countries.³ This approach was taken because social media posts are grouped by politicians, politicians are grouped by parties, and parties are grouped by countries. Moreover, the regression models specify random slopes for ideological closeness on the politician level. Because the different expressions of user engagement, which are analyzed as dependent variables, are count variables, all models shown in this article are negative binomial regression models, in line with recent research on social media user engagement (e.g., Bene et al. 2022; Peeters et al. 2023). All independent variables are standardized to facilitate the interpretation of the coefficients. The main dependent variables remain unscaled. All models include the hypothesized predictors (message sentiment, news sentiment, ideological closeness, and outlet popularity) and the variables that will be used to test the moderating effect of party radicalism and political parallelism. The models include two control variables. They control for the number of followers at the time of posting, and for politicians’ parties’ left-right position. The latter is included to isolate the effect of party radicalism.

Results

Table 2 shows the main regression models for testing the hypothesis set out above. The dependent variables are the number of likes, comments, shares and reactions (likes, “love,” “wow,” “haha,” “sad,” “angry,” “care”). The models show that, contrary to H1, the ideological closeness between a politician’s party and the audience of a news outlet does not positively predict user engagement when that politician shares news published by that outlet, controlling for news outlet popularity. The coefficient of the direct effect of ideological closeness is nonsignificant across all user engagement metrics. Table 2 also shows two interaction effects to disentangle the relationship between ideological closeness and social media user engagement. Firstly, the coefficients show that there is no significant interaction effect between ideological closeness and radicalism. In other words, the ideological closeness between the politician and the news outlet does not predict user engagement with a politicians’ news-sharing post for radical and moderate politicians. Secondly, although the direct effect is nonsignificant, ideological closeness between politicians and news outlets has a slight positive effect on likes and reactions in media systems that are characterized by higher levels of political parallelism. Despite the absence of a direct effect of ideological closeness on user engagement with politicians’ online news-sharing posts, these interaction terms illustrate that it does play a role in some contexts. More specifically, there are indications that in politically parallel media systems the ideological closeness between the politician and the news outlet that is shared matters for user engagement. Crucially, this seems to be the case for user engagement metrics that are important for the spread of the post on the social media platform (reactions and likes). Table 2 also shows that, contrary to H2, politicians’ news-sharing posts that contain a reference to a more popular news outlet do not trigger more user engagement in terms of comments and shares. News-sharing posts that contain a reference to a more popular news outlet seem to trigger significantly more likes and reactions, but these coefficients are rather small, as the predicted number of likes and reactions increases by 2 and 3 percent for every 1 SD increase in outlet popularity, respectively.

Table 2.

Multilevel Negative Binomial Regressions Modeling Likes, Shares, Comments, Reactions. Random Intercepts for Politicians, Parties, and Countries; Random Slopes for Politicians (N = 17,625).

Predictors	Likes				Shares				Comments				Reactions
Predictors	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR	B (SE)	IRR
(Intercept)	3.76 (0.14)***	42.97	3.76 (0.14)***	43.05	1.53 (0.17)***	4.64	1.54 (0.17)***	4.67	2.04 (0.17)***	7.69	2.06 (0.17)***	7.85	3.99 (0.14)***	53.80	4.00 (0.14)***	54.85
Post-level
Message pos. sentiment	0.08 (0.01)***	1.08	0.08 (0.01)***	1.08	−0.03 (0.01)**	0.97	−0.03 (0.01)**	0.97	−0.03 (0.01)**	0.97	−0.03 (0.01)**	0.97	0.07 (0.01)***	1.07	0.07 0.01)***	1.07
Message neg. sentiment	0.00 (0.01)	1.00	0.00 (0.01)	1.00	0.04 (0.01)***	1.04	0.04 (0.01)***	1.04	0.05 (0.01)***	1.05	0.05 (0.01)***	1.05	0.01 (0.01)	1.01	0.01 (0.01)	1.01
News pos. sentiment	−0.01 (0.01)	0.99	−0.01 (0.01)	0.99	−0.01 (0.01)	0.99	−0.01 (0.01)	0.99	−0.06 (0.01)***	0.94	−0.06 (0.01)***	0.94	−0.02 (0.01)**	0.98	−0.02 (0.01)**	0.98
News neg. sentiment	−0.02 (0.01)*	0.98	−0.02 (0.01)*	0.98	0.04 (0.01)***	1.05	0.04 (0.01)***	1.05	0.00 (0.01)	1.00	0.00 (0.01)	1.00	0.01 (0.01)	1.01	0.01 (0.01)	1.01
Politician-level
Party left-right	0.06 (0.06)	1.06	0.05 (0.05)	1.05	0.08 (0.09)	1.08	0.08 (0.09)	1.09	0.10 (0.07)	1.10	0.09 (0.07)	1.10	0.05 (0.06)	1.05	0.04 (0.06)	1.04
Number of followers	0.86 (0.05)***	2.37	0.87 (0.05)***	2.38	0.80 (0.05)***	2.23	0.80 (0.05)***	2.23	0.98 (0.06)***	2.66	0.98 (0.06)***	2.65	0.88 (0.05)***	2.41	0.88 (0.05)***	2.41
Party radicalism	0.20 (0.06)***	1.22	0.19 (0.06)***	1.20	0.23 (0.09)**	1.26	0.23 (0.09)**	1.26	0.17 (0.07)*	1.18	0.17 (0.07)*	1.18	0.18 (0.06)**	1.20	0.18 (0.06)**	1.20
Outlet-level
Ideological closeness	0.01 (0.02)	1.01	0.01 (0.02)	1.01	0.03 (0.03)	1.03	0.03 (0.03)	1.03	−0.02 (0.03)	0.98	−0.02 (0.03)	0.98	0.00 (0.02)	1.00	−0.01 (0.02)	0.99
Outlet popularity	0.02 (0.01)*	1.02	0.02 (0.01)*	1.02	−0.02 (0.01)*	0.98	−0.02 (0.01)*	0.98	−0.00 (0.01)	1.00	0.00 (0.01)	1.00	0.03 (0.01)***	1.03	0.03 (0.01)***	1.03
Country-level
Political parallelism	0.14 (0.14)	1.15	0.12 (0.15)	1.13	0.46 (0.18)*	1.58	0.46 (0.18)*	1.58	0.08 (0.19)	1.08	0.07 (0.18)	1.07	0.21 (0.15)	1.23	0.20 (0.15)	1.22
Interactions
Ideological closeness* Pol. parallelism			0.06 (0.02)**	1.07			−0.01 (0.03)	0.99			0.05 (0.03)	1.05			0.05 (0.02)*	1.05
Ideological closeness* Party radicalism			0.01 (0.02)	1.01			0.01 (0.03)	1.01			0.04 (0.03)	1.04			0.03 (0.02)	1.03
AIC	175,579.4		175,574.8		102,949.9		102,950.7		128,851.0		128,850.8		183,541.0		183,538.1
Var. politicians (N = 2,021)	1.36		1.36		1.59		1.58		2.13		2.13		1.40		1.40
Var. parties (N = 102)	0.12		0.11		0.45		0.46		0.15		0.15		0.16		0.16
Var. countries (N = 15)	0.22		0.23		0.31		0.31		0.38		0.37		0.22		0.22

***

p < .001. **p < .01. *p < .05.

In terms of sentiment, the findings are inconclusive. Politicians’ news-sharing posts that contain more negative language yield more comments and shares (respectively 5% and 4% increase, for every 1 SD increase in negative sentiment), while negativity has no significant impact on likes and reactions. This means that findings are in line with H3a, but only for comments and shares. Positive language in news-sharing posts, by contrast, leads to more likes and reactions (8% and 7% increase per 1 SD increase in positive sentiment), but fewer comments and shares (3% decrease per 1 SD increase in positive sentiment), meaning that the results are in line with H3b for likes and reactions, but contradict the expectation for comments and shares. The content of the news preview also matters. Posts that reference more negative news receive more shares (in line with H4a), although they get fewer likes (contrary to H4a). H4b cannot be confirmed, as politicians’ posts that reference more positive news yield fewer comments and fewer reactions, while the coefficients for likes and shares are insignificant. These results highlight that there is no straightforward effect of sentiment on user engagement, and they illustrate that negative language sparks conversation and leads to the post being picked up and spread further, while positive language invites social media users to share their attitude towards the post with emojis, but not to substantively interact with it or spread it within their own online networks.

Table 5 in the Supplemental Material file shows multilevel negative binomial regression results predicting the sum of all user engagement expressions (likes, comments, shares, and reactions). These models confirm that, contrary to H1 and H2, there is no straightforward positive effect of ideological closeness or outlet popularity. In line with H3a and H3b, the more positive and negative language politicians’ news-sharing posts contain, the higher the number of total user engagement. Negative sentiment in the news does not significantly predict user engagement, and, contrary to H4b, positive sentiment in the news leads to less aggregate user engagement. The interaction terms of ideological closeness with radicalism and political parallelism are insignificant. Despite their simplicity, these models lack the nuance discussed above.

Discussion and Conclusion

Social media allow politicians to directly engage with their online followers, and to do so in many ways. One of the ways politicians can appeal to social media users is by sharing news items. When they do so, politicians effectively impact the distribution of that news item, and the exposure of their followers to it. However, the impact of politicians’ news-sharing posts is dependent on user engagement which highlights the relevance of a post among its audiences and plays a role in the “virality” logic of social media (Klinger and Svensson 2015). This article combines insights on user engagement with politicians’ online posts, news on social media and political parallelism to test the determinants of user engagement with politicians’ news-sharing posts on social media.

Results show that, despite previous findings that politicians share more news of which the audience aligns with their own ideology (Buyens, Van Aelst, and Vaccari 2024; Heidenreich et al. 2022), user engagement with social media posts in which politicians reference ideologically aligned news is not higher than user engagement with posts in which they share ideologically dissonant news sources. This is a sign that, even though politicians’ online news-sharing behavior seems to provide the circumstances for partisan selective exposure to news, social media users do not close the vicious circle by engaging more with that type of content. If user engagement with ideologically aligned news-sharing posts were systematically higher, such tendencies would lead to more ideologically isolated and, possibly, polarized news audiences. So, on the one hand, I conclude that the further distribution of ideologically congruent news shared by politicians is not so much boosted by their online audiences, which may have beneficial implications for our understanding of partisan selective exposure in the digital information environment. On the other hand, this implies that politicians are not driven to share ideologically congruent news sources by the incentive of more user engagement.

Nevertheless, the findings of this study also highlight that in some specific contexts, like those in more politically parallel media systems, the ideological closeness between the politician and the news outlet does matter for user engagement with news-sharing posts. Because politicians in countries with higher political parallelism are also more likely to share ideologically aligned news (Buyens, Van Aelst, and Vaccari 2024), there is a reason for concern that the ideologically congruent news-sharing behavior of politicians is further stimulated by social media users. For example, while ideological alignment may not boost user engagement with politicians’ news-sharing posts in Belgium and Denmark, it may still do so in Hungary and Poland. This trend may encourage ideological segregation in digital information networks and lead to more opportunities for partisan selective exposure.

While the news outlet that is shared does not have a strong general effect on user engagement, the content of the message and of the news matters more. Negative sentiment in politicians’ news-sharing posts triggers comments and shares, while positive sentiment in their message leads to more likes and reactions. Like negative sentiment in the post, negative news also seems to be more “shareable” (see Trilling et al. 2017). Although the effects of sentiment on user engagement are diverse, they show that positive and negative sentiment have distinct effects on the ways in which social media users process the content they are exposed to. This is in line with previous research that has shown the impact of negativity on shares and comments (Bene 2017; Bene et al. 2022; Heiss et al. 2019), indicating that it is more successful in triggering discussion and making social media users further distribute the post within their own online networks.

There are several limitations to this research. Firstly, there is a considerable time gap between the survey and the Facebook data collection which might have implications for our findings as the COVID-19 pandemic was shown to have affected news consumption and perceptions of political institutions (Kritzinger et al. 2021; Van Aelst et al. 2021). Secondly, apart from message and news sentiment, this study does not consider other variables that may impact user engagement, like the topic of the news or criticism toward the outlet that published it. Moreover, this study takes into account only one variable on the country level. Therefore, it cannot confirm or rule out that any other characteristics of the political system or media environment play a role in the relationships under study. Lastly, we can only study and discuss the effects of news and post characteristics on user engagement metrics, but we cannot be sure what these metrics mean for actual news exposure and consumption, or for social media users’ responses to these posts or the news that is shared within. Future research should take into account the receivers’ perceptions associated with user engagement and explain why social media users engage with politicians’ news-sharing posts.

Notwithstanding the limitations of this study, its findings illustrate that news content and message cues matter most for user engagement with news shared by politicians on social media. Crucially, the ideological closeness between the politician and the news outlet does not predict user engagement. Nevertheless, there are reasons to believe that in contexts that are already more vulnerable to partisan selective exposure, like those in countries with higher political parallelism, social media users reward politicians with higher user engagement when they share news that is published by outlets that are congruent with their own ideology. Not only may this behavior reinforce politicians in their ideologically selective news-sharing behavior, it also substantially impacts the visibility and the impact of this content for other social media users. This finding further illustrates the role that politicians play in the polarization and ideological segregation of online information environments.

Supplemental Material

sj-docx-1-hij-10.1177_19401612251334941 – Supplemental material for Never Mind the Source? The Drivers of User Engagement With Politicians’ Online News-Sharing Posts in 15 European Countries

Supplemental material, sj-docx-1-hij-10.1177_19401612251334941 for Never Mind the Source? The Drivers of User Engagement With Politicians’ Online News-Sharing Posts in 15 European Countries by Willem Buyens in The International Journal of Press/Politics

Footnotes

Acknowledgements

I thank the members of the Network of European Political Communication Scholars (NEPOCS) for granting me access to their survey data.

Data Availability Statement

The data that support the findings of this study are available from the author, Willem Buyens, upon reasonable request.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Research Foundation—Flanders (FWO) under Grant G096621N.

Ethical Approval

The survey reported in this manuscript was approved by the ethics committee of the Amsterdam School of Communication Research (ASCoR secretariat, Ethics Committee, University of Amsterdam, Postbus 15,793, 1001 NG Amsterdam; +31(0)205253680; ascor-secr-fmg@uva.nl).

ORCID iD

Willem Buyens

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biography

Willem Buyens is a postdoctoral researcher at the department of Communication Studies of the University of Antwerp. His work focuses on the relationships between politics and news media in the evolving digital information environment. More specifically, he studies the ways in which political actors interact with and share news content online.

References

Anspach

N. M.

2017. “The New Personal Influence: How Our Facebook Friends Influence the News We Read.” Political Communication 34(4):590–606. doi:10.1080/10584609.2017.1316329.

Arendt

Northup

Camaj

2019. Selective Exposure and News Media Brands: Implicit and Explicit Attitudes as Predictors of News Choice. Media Psychology, 22(3): 526–543. https://doi.org/10.1080/15213269.2017.1338963

Bailey

O. G.

Cammaert

Carpentier

2007. Understanding Alternative Media. Maidenhead, UK: McGraw-Hill Education (UK).

Bene

2017. “Go Viral on the Facebook! Interactions Between Candidates and Followers on Facebook During the Hungarian General Election Campaign of 2014.” Information, Communication & Society 20(4):513–29. doi:10.1080/1369118X.2016.1198411.

Bene

Ceron

Fenoll

Haßler

Kruschinski

Larsson

A. O.

Magin

Schlosser

Wurst

A.-K.

2022. “Keep Them Engaged! Investigating the Effects of Self-centered Social Media Communication Style on User Engagement in 12 European Countries.” Political Communication 39(4):429–53. doi:10.1080/10584609.2022.2042435.

Boehmer

Tandoc

2015. “Why We Retweet: Factors Influencing Intentions to Share Sport News on Twitter.” International Journal of Sport Communication 8(2):212–32. doi:10.1123/IJSC.2015-0011.

Brüggemann

Engesser

Büchel

Humprecht

Castro

2014. “Hallin and Mancini Revisited: Four Empirical Types of Western Media Systems.” Journal of Communication 64(6):1037–65. doi:10.1111/jcom.12127.

Buyens

Van Aelst

2022. “Alternative Media, Alternative Voices? A Quantitative Analysis of Actor Diversity in Alternative and Mainstream News Outlets.” Digital Journalism 10(2):337–59. doi:10.1080/21670811.2021.1929366.

Buyens

Van Aelst

Paulussen

2024. “Curating the News. Analyzing Politicians’ News Sharing Behavior on Social Media in Three Countries.” Information, Communication & Society 1–17. doi:10.1080/1369118X.2024.2317901.

10.

Buyens

Van Aelst

Vaccari

2024. “The Same Views, the Same News? A 15-Country Study on News Sharing on Social Media by European Politicians.” Political Communication 42(1): 27–56. doi:10.1080/10584609.2024.2352795.

11.

Castro

Strömbäck

Esser

Van Aelst

de Vreese

Aalberg

Cardenal

A. S.

Corbu

Hopmann

D. N.

Koc-Michalska

Matthes

Schemer

Sheafer

Splendore

Stanyer

Stępińska

Štětka

Theocharis

2022. “Navigating High-Choice European Political Information Environments: A Comparative Analysis of News User Profiles and Political Knowledge.” The International Journal of Press/Politics 27(4):827–59. doi:10.1177/19401612211012572.

12.

Chinn

Hiaeshutter-Rice

Chen

2023. “How Science Influencers Polarize Supportive and Skeptical Communities Around Politicized Science: A Cross-Platform and Over-Time Comparison.” Political Communication 41(4):627–48. doi:10.1080/10584609.2023.2201174.

13.

Dun

Soroka

2021. “Dictionaries, Supervised Learning, and Media Coverage of Public Policy.” Political Communication 38(1–2):140–58. doi:10.1080/10584609.2020.1763529.

14.

Eberl

J.-M.

Tolochko

Jost

Heidenreich

Boomgaarden

H. G.

2020. “What’s in a Post? How Sentiment and Issue Salience Affect Users’ Emotional Reactions on Facebook.” Journal of Information Technology & Politics 17(1):48–65. doi:10.1080/19331681.2019.1710318.

15.

Ekman

Widholm

2015. “Politicians as Media Producers: Current Trajectories in the Relation Between Journalists and Politicians in the Age of Social Media. Journalism Practice 9(1):78–91. doi:10.1080/17512786.2014.928467.

16.

Flaxman

Goel

Rao

J. M.

2016. “Filter Bubbles, Echo Chambers, and Online News Consumption.” Public Opinion Quarterly 80(S1):298–320. doi:10.1093/poq/nfw006.

17.

Haim

Kümpel

A. S.

Brosius

H.-B.

2018. “Popularity Cues in Online Media: A Review of Conceptualizations, Operationalizations, and General Effects. Studies in Communication | Media 7(2):186–207. doi:10.5771/2192-4007-2018-2-58.

18.

Hallin

D. C.

Mancini

2004. Comparing Media Systems. Cambridge: Cambridge University Press.

19.

Heidenreich

Eberl

J.-M.

Tolochko

Lind

Boomgaarden

H. G.

2022. “My Voters Should See This! What News Items Are Shared by Politicians on Facebook?” The International Journal of Press/Politics 29(1):5–25. doi:10.1177/19401612221104740.

20.

Heiss

Schmuck

Matthes

2019. “What Drives Interaction in Political Actors’ Facebook Posts? Profile and Content Predictors of User Engagement and Political Actors’ Reactions. Information, Communication & Society 22(10):1497–513. doi:10.1080/1369118X.2018.1445273.

21.

Holt

Ustad Figenschou

Frischlich

2019. “Key Dimensions of Alternative News Media.” Digital Journalism 7(7):860–9. doi:10.1080/21670811.2019.1625715.

22.

Humprecht

Castro Herrero

Blassnig

Brüggemann

Engesser

2022. «Media Systems in the Digital Age: An Empirical Comparison of 30 Countries.” Journal of Communication 72(2):145–64. doi:10.1093/joc/jqab054.

23.

Iyengar

Hahn

K. S.

2009. “Red Media, Blue Media: Evidence of Ideological Selectivity in Media Use.” Journal of Communication 59(1):19–39. doi:10.1111/j.1460-2466.2008.01402.x.

24.

Jamieson

K. H.

Cappella

J. N.

2008. Echo Chamber: Rush Limbaugh and the Conservative Media Establishment. 1st Edition. Oxford: Oxford University Press.

25.

Jolly

Bakker

Hooghe

Marks

Polk

Rovny

Steenbergen

Vachudova

M. A.

2022. “Chapel Hill Expert Survey trend file, 1999–2019.” Electoral Studies 75:102420. doi:10.1016/j.electstud.2021.102420.

26.

Jost

2020. “Populism Fuels Love and Anger: The Impact of Message Features on Users’ Reactions on Facebook.” International Journal of Communication 14:2081–102.

27.

Kaiser

Keller

T. R.

Kleinen-von Königslöw

2021. “Incidental News Exposure on Facebook as a Social Experience: The Influence of Recommender and Media Cues on News Selection.” Communication Research 48(1):77–99. doi:10.1177/0093650218803529.

28.

Karlsen

2015. “Followers are Opinion Leaders: The Role of People in the Flow of Political Communication on and Beyond Social Networking Sites.” European Journal of Communication 30(3):301–18. doi:10.1177/0267323115577305.

29.

Katz

Lazarsfeld

P. F.

1955. Personal Influence: The Part Played by People in the Flow of Mass Communications. London: Routledge.

30.

Klinger

Svensson

2015. “The Emergence of Network Media Logic in Political Communication: A Theoretical Approach. New Media & Society 17(8):1241–57. doi:10.1177/1461444814522952.

31.

Kritzinger

Foucault

Lachat

Partheymüller

Plescia

Brouard

2021. “‘Rally Round the Flag’: The COVID-19 Crisis and Trust in the National Government.” West European Politics 44(5–6):1205–31. doi:10.1080/01402382.2021.1925017.

32.

Kümpel

A. S.

2019. “The Issue Takes It All? Incidental News Exposure and News Engagement on Facebook.” Digital Journalism 7(2):165–86. doi:10.1080/21670811.2018.1465831.

33.

Medders

R. B.

Metzger

M. J.

2018. The Role of News Brands and Leads in Exposure to Political Information on the Internet. Digital Journalism 6(5): 599–618. https://doi.org/10.1080/21670811.2017.1320770

34.

Messing

Westwood

S. J.

2014. “Selective Exposure in the Age of Social Media: Endorsements Trump Partisan Source Affiliation When Selecting News Online.” Communication Research 41(8):1042–63. doi:10.1177/0093650212466406.

35.

Müller

Schulz

2021. “Alternative Media for a Populist Audience? Exploring Political and Media Use Predictors of Exposure to Breitbart, Sputnik, and Co. Information, Communication & Society 24(2):277–93. doi:10.1080/1369118x.2019.1646778.

36.

Newman

Fletcher

Eddy

Robertson

C. T.

Nielsen

R. K.

2023. Reuters Institute Digital News Report 2023. Reuters Institute for the Study of Journalism.

37.

Nielsen

R. K.

Vaccari

2013. “Do People ‘Like’ Politicians on Facebook? Not Really. Large-Scale Direct Candidate-to-Voter Online Communication as an Outlier Phenomenon.” International Journal of Communication 7: 2333–56.

38.

Pariser

2011. The Filter Bubble: What the Internet is Hiding from you. London: Penguin Books.

39.

Peeters

Opgenhaffen

Kreutz

Van Aelst

2023. “Understanding the Online Relationship Between Politicians and Citizens. A Study on the User Engagement of Politicians’ Facebook Posts in Election and Routine Periods. Journal of Information Technology & Politics 20(1):44–59. doi:10.1080/19331681.2022.2029791.

40.

Seymour-Ure

(1974). The political impact of mass media (Vol. 4). London: Constable.

41.

Singer

J. B.

2014. “User-Generated Visibility: Secondary Gatekeeping in a Shared Media Space.” New Media & Society 16(1):55–73. doi:10.1177/1461444813477833

42.

Soroka

Daku

Hiaeshutter-Rice

Guggenheim

Pasek

2018. “Negativity and Positivity Biases in Economic News Coverage: Traditional Versus Social Media.” Communication Research 45(7): 1078–98. doi:10.1177/0093650217725870.

43.

Sunstein

C. R.

(2007). Republic.com 2.0. Princeton, NJ: Princeton University Press.

44.

Tewksbury

Weaver

A. J.

Maddex

B. D.

2001. “Accidentally Informed: Incidental News Exposure on the World Wide Web.” Journalism & Mass Communication Quarterly 78(3):533–54. doi:10.1177/107769900107800309.

45.

Thorson

Wells

2016. “Curated Flows: A Framework for Mapping Media Exposure in the Digital Age.” Communication Theory 26(3):309–28. doi:10.1111/COMT.12087.

46.

Trilling

Tolochko

Burscher

2017. “From Newsworthiness to Shareworthiness.” Journalism & Mass Communication Quarterly 94(1): 38–60.

47.

Vaccari

Valeriani

2015. “Follow the Leader! Direct and Indirect Flows of Political Communication During the 2013 Italian General Election Campaign. New Media & Society 17(7):1025–42. doi:10.1177/1461444813511038.

48.

Van Aelst

Strömbäck

Aalberg

Esser

de Vreese

Matthes

Hopmann

Salgado

Hubé

Stępińska

Papathanassopoulos

Berganza

Legnante

Reinemann

Sheafer

Stanyer

2017. “Political Communication in a High-Choice Media Environment: A Challenge for Democracy? Annals of the International Communication Association 41(1):3–27. doi:10.1080/23808985.2017.1288551.

49.

Van Aelst

Tóth

Castro

Štětka

de Vreese

Aalberg

Cardenal

A. S.

Corbu

Esser

Hopmann

D. N.

Koc-Michalska

Matthes

Schemer

Sheafer

Splendore

Stanyer

Stępińska

Strömbäck

Theocharis

2021. “Does a Crisis Change News Habits? A Comparative Study of the Effects of COVID-19 on News Media Use in 17 European Countries.” Digital Journalism 9(9):1316–46. doi:10.1080/21670811.2021.1943481.

50.

Van Atteveldt

Van der Velden

M. A. C. G.

Boukes

2021. “The Validity of Sentiment Analysis: Comparing Manual Annotation, Crowd-Coding, Dictionary Approaches, and Machine Learning Algorithms.” Communication Methods and Measures 15(2):121–40. doi:10.1080/19312458.2020.1869198.

51.

Van Kempen

. 2007. “Media-Party Parallelism and Its Effects: A Cross-National Comparative Study.” Political Communication 24(3):303–20. doi:10.1080/10584600701471674.

52.

Weber

2014. “Discussions in the Comments Section: Factors Influencing Participation and Interactivity in Online Newspapers’ Reader Comments. New Media & Society 16(6):941–57. doi:10.1177/1461444813495165.

53.

Weeks

B. E.

Lane

D. S.

Kim

D. H.

Lee

S. S.

Kwak

2017. “Incidental Exposure, Selective Exposure, and Political Information Sharing: Integrating Online Exposure Patterns and Expression on Social Media.” Journal of Computer-Mediated Communication 22(6):363–79. doi:10.1111/jcc4.12199.

54.

Winter

Krämer

N. C.

2012. Selecting Science Information in Web 2.0: How Source Cues, Message Sidedness, and Need for Cognition Influence Users’ Exposure to Blog Posts. Journal of Computer-Mediated Communication 18(1): 80–96. https://doi.org/10.1111/j.1083-6101.2012.01596.x

55.

Winter

Krämer

N. C.

2014. “A Question of Credibility—Effects of Source Cues and Recommendations on Information Selection on News Sites and Blogs. Communications 39(4):435–56. doi:10.1515/commun-2014-0020.

56.

Young

Soroka

2012. “Affective News: The Automated Coding of Sentiment in Political Texts.” Political Communication 29(2):205–31. doi:10.1080/10584609.2012.671234.

57.

Ziegele

Breiner

Quiring

2014. “What Creates Interactivity in Online News Discussions? An Exploratory Analysis of Discussion Factors in User Comments on News Items.” Journal of Communication 64(6):1111–38. doi:10.1111/jcom.12123.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB