Abstract
Societally salient issues, like migration, stimulate user engagement with political parties on social media. This user engagement, in turn, is associated with political behavior, such as voting. Nonetheless, few studies so far have investigated the interaction between these factors. We examine how anti-immigration attitudes are associated with user engagement with political parties on social media. In this study, user engagement is understood as following political parties on social media. Through online data that were collected in October 2019 among adults (N= 1,000) in Belgium, we investigate how attitudes and user engagement are associated with voting behavior. Results suggest that attitudes towards migration are associated with user engagement with both left and right-wing parties on social media. Moreover, these attitudes and user engagement – and the interaction between the two – are related to voting behavior: being against (being in favor of) migration and following right-wing (left-wing) parties on social media is associated with a higher likelihood of voting for a right-wing (left-wing) party.
1 Introduction
Over the past years, migration has become an inescapable theme in elections in several European countries. A major factor was the so-called refugee crisis between 2014 and 2016, during which the European Union received over 1.3 million asylum applications (UNHCR, 2017). These large numbers have stimulated public attention to migration. The years following the crisis saw widespread attention, mostly through negative framing, being paid by (news) media and by political actors to migration and integration of migrants in European societies (Georgiou and Zaborowski, 2017). This arguably has contributed to a surge in support for right-wing populist parties across Europe (Bieber, 2019; Damstra, Jacobs, Boukes, and Vliegenthart, 2021). Previous studies have shown that media play a vital role in disseminating information about migration and integration to the public, while also being significantly related to individual attitudes and public opinion towards migration (De Coninck, 2020; Heidenreich, Eberl, Lind, and Boomgaarden, 2020). Heath and Richards (2019) found that particularly in countries where large numbers of refugees arrived during the crisis, citizens favor more restrictive migration policies and hold rather negative attitudes towards migrants/migration. From 2014 to 2018, Belgium received more than 80,000 asylum applications (Statistiek Vlaanderen, 2020). While the responses to these refugees were positive and welcoming at first, as the refugee-crisis continued media outlets started to focus on adverse economic consequences, danger to citizens’ well-being, and cultural differences. In line with Heath and Richards (2019), the growing numbers of refugees was linked to more negative attitudes. The tension regarding the refugee crisis led to the collapse of the Belgian federal government in 2018. Consequently, migration was a key campaigning topic in the Flemish election of 2019.
Besides the impact of traditional news media on opinions about migration, which has been extensively studied over the past decades (Eberl et al., 2018), social media play an increasingly important role in the diffusion of political ideas about migration (Heidenreich et al., 2020). The presence of political actors on social networking sites (SNS) is rapidly increasing, and some authors argue that social media have become so important that politicians now use them as the preferred venue for propagating new policies or ideas (Bossetta, 2018; Esser and Pfetsch, 2020) or as a vital part of a larger communication strategy that includes traditional media (Ross and Bürger, 2014). The political role of social media, including how social networks are linked to user engagement and political participation among voters, has quickly become one of the more established themes in political communication research (Davis, 2010; Gil De Zúñiga, Puig-I-Abril, and Rojas, 2009; Kim and Baek, 2018; Pennington and Winfrey, 2021). The growth and popularity of SNS suggest that, regardless of their actual impact on increasing political participation, their popularity is unlikely to wane in the short or even medium term (Larsson and Moe, 2012).
Content analyses of social media messages about migration by political actors has shown that the depiction of migration varies across political parties. Van Leuven, Deprez, Joye, and Ongenaert (2019) found that – as in television and print media – negative frames are most commonly used by politicians on social media. More specifically, right-wing politicians frame migrants as intruders, while other parties emphasize the victim-frame. More moderate political players address migration less frequently than players on the extreme left and right (Heidenreich et al., 2020). To date there is limited authoritative account of the exact role of social media platforms in this (political) migration debate. With this study, we aim to contribute to the current literature of user engagement on social media and political participation by focusing on the issue of migration in two ways. First, although SNS have become sources of information and avenues for organizing activities and making decisions, little is known about who acts upon or reacts to social media messages from politicians (Lee and Nerghes, 2018) and what prompts them to do so. Here, we want to investigate how citizens’ attitudes towards migration are related to user engagement towards specific parties or politicians on SNS (i. e., clicktivism). Second, since exposure to political messages on social media may be associated with voting intention, and successful communication on social media may lead to a higher vote share in elections (Gibson and McAllister, 2014; Kruikemeier, van Noort, Vliegenthart, and de Vreese, 2014), we also investigate if migration attitudes and social media activities are associated with political behavior in the ‘offline’ world, more specifically with voting behavior. This allows our study to provide some insights into the agenda-setting power of political actors. Given the growing role that social media play in many aspects of societal life and have on attitudes and behaviors, it is particularly relevant to provide further insights into the ways in which social media behavior may be associated with political behavior like voting, particularly when dealing with a societally salient issue like migration. Although we cannot make any claims regarding the causality of this relationship – as it may be reciprocal –, insights provided here can aid scholars and policy makers to understand who ‘supports’ their social media presence, and whether this is linked to real-life support.
Migration politics on (social) media
During and in the wake of the European refugee crisis, migration has become one of the most salient topics on news media (Heidenreich et al., 2020). As previous studies have shown, the volume and intensity of migration-related news influence public’s perceptions of, and attitudes towards, migration (Boomgaarden and Vliegenthart, 2009; van Klingeren, Boomgaarden, and de Vreese, 2017). Furthermore, the way in which (political actors on) media frame migration is related to attitudes towards migration. In particular, frames that provide a somewhat negative view of societal themes are significantly related to attitudes (Soroka, Young, and Balmas, 2015). Most studies on this topic have investigated the effects of traditional media content (television, radio, newspapers), while less is known about how the intensity and framing of migration on social media is related to migration attitudes.
With the growing impact and integration of social media in people’s daily lives (Metz, Kruikemeier, and Lecheler, 2020), these platforms have become increasingly important in the diffusion of, and communication about, political ideas (McGregor, 2017). Politicians increasingly take to SNS to propagate new policies or ideas (Bossetta, 2018; Esser and Pfetsch, 2020), often in combination with appearances on television and radio (Ross and Bürger, 2014). The presence of politicians in social media users’ digital space is beneficial to them for several reasons. On social media, politicians cultivate authenticity in campaigns or political messages by blurring the lines between their private and public selves, something which appeals to voters (Louden and McCauliff, 2004; Starke, Marcinkowski, and Wintterlin, 2020). Furthermore, the interactive and participatory nature of SNS means that individuals can also comment, like or share politicians’ messages (Heiss, Schmuck, and Matthes, 2019), which further strengthens individuals’ perceived involvement. Personalized messages from politicians arouse more engagement from the public and intensify perceptions of social presence, of parasocial interaction, and vote intentions for candidates (McGregor, 2017). However, the public’s political social media engagement varies by political orientation and may also depend on which topics are discussed – and how – by politicians on SNS.
In that regard, Heidenreich et al. (2019) found that the visibility of migration-related party discourses on Facebook significantly increased in several European countries during the refugee crisis. Looking at the tonality of this migration discourse, they found surprisingly few differences across ideologies, although extreme right and left-wing parties spoke more about migration than more moderate parties did (Heidenreich et al., 2019). However, other studies indicated that right-wing actors are more critical of immigration than moderate parties (Heiss and Matthes, 2020; Schemer, 2012; Van Gorp, Figoreux, and Vyncke, 2018; Van Leuven, Deprez, Joye, and Ongenaert, 2019). In line with this, Van Leuven et al. (2019) analyzed 1,528 tweets produced by Flemish politicians, investigating the occurrence of four common migration frames (i. e., intruder, victim, wealth gap, and win-win, see Van Gorp et al., 2018). The intruder frame occurred most and across all tweets (73.1 %), followed by the victim frame (22.0 %), the wealth gap frame (3.7 %), and the win-win frame (1.2 %). The intruder frame was by far the most prominent in tweets by politicians of the far right-wing party Vlaams Belang and the Flemish nationalist party N-VA, while the other four (moderate to left-wing) politicians mostly utilized the victim frame (Van Leuven et al., 2019). Vlaams Belang (38.8 %) had the most framed statements, followed closely by the N-VA (35.9 %). This is also reflected in Heidenreich et al.’s study (2020), which shows that migration was more prominent in the Facebook posts of more ideologically extreme parties – particularly on the right – in Germany, Sweden, Austria, the UK, Spain, and Poland.
User engagement
Given the increased salience of migration on social media as a result of the refugee crisis, notwithstanding the different intensity and frames by which some parties or politicians to post or communicate about it (Van Leuven et al., 2019), some parties seem more likely to engage users on SNS than others. Heidenreich et al. (2019, p. 36) found in their study on Facebook that “migration-related status posts receive distinctly more interactions. While non-migration-related status posts received approximately the same amount of average interactions during two years after the introduction of the reactions, interactions for migration-related status posts almost continuously increased during the same time span”. These interactions build up users’ personal interaction with politicians or parties (Heiss et al., 2019), which can be illustrated by reacting to, or sharing, messages, or liking or following a politician or party on SNS (Eberl, Tolochko, Jost, Heidenreich, and Boomgaarden, 2020). In this context, we argue that clicktivism is becoming an increasingly relevant concept in online political participation literature. Clicktivism is an online, noncommittal type of political participation, which does not draw on specialized knowledge of digital environments, engages an established political object, such as politicians or parties, and includes an action perpetrated by an individual (Halupka, 2014; Keller and Kleinen-von Königslöw, 2018). It is a reflexive political act contained within a moment of spontaneity (Halupka, 2014; Štětka and Mazák, 2014). Heiss et al. (2019) addressed the question of which type of messages or posts – regardless of the specific theme they address – elicit clicktivism. First, they found that user engagement varies by the type of profile that a message is posted on. Posts from general party pages were less successful in stimulating clicktivism than posts from private politician pages, for example. Second, political messages with a negative tone or frame have a larger impact on users in terms of engagement and political mobilization than positive messages (Eberl et al., 2020; Heiss et al., 2019; Xenos, Macafee, and Pole, 2017). Given the negative framing of migration on SNS by right-wing parties and the salience of this issue among online audiences (Heidenreich et al., 2019; Van Leuven et al., 2019), it seems reasonable to assume that a link exists between people’s opinions about migration and their likelihood of digitally interacting with political parties with corresponding viewpoints (Heidenreich et al., 2019). In line with this reasoning,
we expect that users who hold negative attitudes towards migration are more likely to follow accounts of right-wing parties than of other parties (Hypothesis 1a),
and that users who hold positive attitudes towards migration are more likely to follow accounts of left-wing parties than of other parties (Hypothesis 1b).
Voting behavior
Previous studies have shown that reading and/or interacting with online political communication by audiences is related to greater political participation (Gibson and McAllister, 2014; Xenos and Moy, 2007). Although getting audiences to interact with their (online) political communication (e. g., by following them or sharing posts) is not the end goal of a political party or politicians, it “must be seen as part of an ongoing political process” whereby “one ‘like’ often leads to another—and in rare cases even to hundreds of thousands—which can have serious political consequences” (Keller and Kleinen-von Königslöw, 2018, p. 2). In recent elections, such as presidential elections in the United States, the Brexit campaign, and elections in many Western European countries, social media campaigning has become an increasingly important part of political parties’ election campaigns (Gerbaudo, Marogna, and Alzetta, 2019). In many of these election campaigns – particularly following or during the European refugee crisis – migration was a key theme (Walgrave, Lefeverem and Tresch, 2020). We will focus on two well-established theoriesfound in earlier studies (Boomgaarden and Vliegenthart, 2007; Burscher, van Spanje, and de Vreese, 2015; Walgrave and De Swert, 2004) to explain how exposure to (media) messages about migration is associated with a greater probability to vote for a right-wing (or anti-immigrant) party.
In agenda-setting theory (McCombs and Shaw, 1972), it is theorized that concerns about a certain issue or topic are transferred to the public through (news) media messages. In line with this and applied to the current research, studies have shown that exposure to migration-related messages of political actors on SNS is associated with greater concern about migration among voters (Heidenreich et al., 2019). Following issue ownership theory (Budge and Farlie, 1983; Petrocik, 1996), which states that parties and politicians attempt to mobilize voters by emphasizing their competence on specific key issues on which they position themselves, we also expect that concerns about migration are associated with the likelihood for voting for a certain party (Burscher et al., 2015).
When taken together, these theories explain how issue-related news – or, in our case, social media messages – are associated with right-wing party voting. As previously shown, and in accordance with agenda-setting theory, exposure to messages about migration on social media increases the perceived salience of this topic among voters (Heidenreich et al., 2019), and this increased salience is likely to be related to voting for a party which “is associated with the issue and/or has a reputation of handling the issue” (Burscher et al., 2015, p. 60). So far, most studies investigating the relationship between issue-related media messages and voter turnout have focused on messages by traditional media. However, we expect that many of the same theoretical assumptions are true for social media messages, given the increased normalization of social media as a tool during electoral campaigns (Gibson and McAllister, 2014). This leads us to:
expect that following right-wing parties on SNS is associated with a greater likelihood of voting for a right-wing party (Hypothesis 2a),
while following left-wing parties on SNS is associated with a greater likelihood of voting for a left-wing party (Hypothesis 2b).
The original issue-ownership model, as developed by Petrocik (1996), was tested by focusing on issues on which all voters and parties share the same goal (e. g., reducing unemployment, fighting crime) (Burscher et al., 2015). However, Walgrave, Lefevere, and Tresch (2012) argue that for so-called positional issues – issues on which voters and parties may hold different preferences – the effect of issue-ownership on party preferences is conditioned on positional agreement between voters and issue owners. Migration as an issue has certain positional elements to it on which voters and parties can disagree. We therefore expect that the link between SNS engagement to right-wing parties with voting for a right-wing party particularly occurs among voters who already held negative attitudes towards migration in the first place – and vice versa for left-wing parties. Heiss and Matthes (2020) have found support for this in Austria. In their study on the effect of right-wing populists’ communication on social media, they showed that anti-immigrant attitudes drove selective exposure on Facebook: Users who held anti-immigrant attitudes were more likely to seek out right-wing populists’ messages, which in turn reinforced these negative attitudes. To explain this, they argued that right-wing populist messages (attempt to) activate feelings of group threat. They argue that “RWP [right-wing populist] actors strategically channel and strengthen specific group-based identities by stimulating antagonism toward political elites and by strategically nurturing perceived threats attributed to immigrants” (Heiss and Matthes, 2020, p. 305). In short, we expect the H2a effect to be stronger among voters who hold negative attitudes towards migration and the H2b effect to be stronger among voters who hold positive attitudes towards migration. This leads to the following hypotheses:
Individuals with negative attitudes towards migration and SNS engagement with right-wing parties show a greater likelihood of voting for a right-wing party (Hypothesis 3a).
Individuals with positive attitudes towards migration and SNS engagement with left-wing parties show a smaller likelihood of voting for a left-wing party (Hypothesis 3b).
Despite a variety of studies on issue ownership claiming that issue perceptions affect party preference or voting intention (Walgrave et al., 2015), a growing criticism of this theory focuses on the possibility of reverse causation, that is, the possibility that party choice affects issue perceptions (Craig and Cossette, 2020; Vliegenthart and Lefevere, 2018). Recent longitudinal evidence by Vliegenthart and Lefevere (2018) among Dutch voters provides mixed evidence for the assumption of reverse causation. However, issue ownership perceptions mattered for party preferences, but the reverse relationship – from party preference to issue ownership perceptions – is much stronger. Thus, as we move on to the empirical part of this article, we avoid making causal claims but rather emphasize the presence or absence of a relationship or association between indicators.
2 Data and methodology
Data
We distributed an online questionnaire among adults in Flanders, the northern Dutch-speaking region of Belgium, in October 2019. The survey was fielded for two weeks until a sample size of 1,000 respondents was reached. The polling agency we cooperated with (iVOX) drew a quota sample out of its large-scale panel of 150,000 Belgians, with a cooperation rate of about 29 % – comparable to other online surveys in Belgium about outgroup attitudes (see De Coninck et al., 2019, 2021). The total sample mirrored the Flemish adult population by gender, age, and educational attainment. Respondents were contacted via e-mail with the request to cooperate in an unspecified study. The survey itself was distributed via the polling agency’s own survey tool, and the survey language was Dutch. Each question in the survey was presented on a different page, and respondents did not have the option to return to previous questions and change their answer. We received ethical approval for this study from the KU Leuven Social and Societal Ethics Committee (case number 2017 07 854), and informed consent was obtained from all participants.
Measures
Anti-immigration attitudes. We measured anti-immigration attitudes through a proxy variable: the evaluation of migration-integration policies. This proxy variable was chosen since a few weeks prior to this study a new Flemish government had been formed. In the government’s public communication, plans concerning migration-integration policies were prominently discussed. As migration had been a key theme in this election campaign which involved heated debates regarding policy preferences – and general preferences – regarding migration-integration among the public and political elites, we believed that assessing policy preferences as a proxy for general migration attitudes presented a timely, salient, and societally relevant measure. The practice of assessing policy preferences as proxies for other concepts has a long history (Nosek, Hawkins, and Frazier, 2012). For example, the Modern Racism Scale (McConahay, 1986; Nosek et al., 2012) indirectly measured racial attitudes by asking individuals about policy positions with racial implications.
Two items asked whether respondents believed that the migrant-integration policy plans of the Flemish government would contribute to the integration of migrants into Flemish society, and if they believed that the Belgian migration policies during the refugee crisis were a good way to deal with this crisis. It is important to emphasize that migration policies at the time were made by a center-right government with a Secretary of State for Asylum and Migration from N-VA, the Flemish nationalist party, which was the largest Belgian party at the time of the refugee crisis (Puschmann, Sundin, De Coninck, and d’Haenens, 2019). The migration-integration policies of this government could be summarized as follows: decrease waiting times for asylum applications, strict return policies for those whose applications were denied, and a hard-line approach to illegal immigration and border policies (Commissie Bossuyt, 2019).
Answer options ranged from 1 (strongly disagree) to 5 (strongly agree), with two additional options for respondents who had no opinion (6) or did not know what these policies entailed (7). Respondents who indicated one of these last two options were excluded from analyses (N = 271) since we were unable to derive opinions on migration policies from them. We calculated the mean score of these two indicators, with high scores indicating support for the (right-wing) government’s migration-integration policies and, thus, negative views towards migration.
Following political parties on social media. We asked whether respondents followed political parties on social media. Presenting all major Flemish political parties, respondents had to indicate for each party whether they followed the main party page. We grouped the data from left-wing (PVDA, Groen, sp.a), moderate (CD&V, Open VLD), and right-wing (N-VA, Vlaams Belang) parties into a set of three dichotomous variables to indicate whether the survey respondents follow political parties from a certain political orientation (0 = do not follow party, 1 = follow party). Once respondents followed one of the parties in a category, they were classified into the corresponding category. For example, if respondents only followed PVDA but not Groen or sp.a., they were classified as following a left-wing party. Respondents who were not active on social media accounts (n = 163) were excluded.
Voting behavior
In May 2019, Belgium held federal elections. We asked respondents which (Flemish) party they voted for in these elections[1] (ranked here from far-left to far-right): PVDA, Groen, sp.a (left-wing); CD&V, Open VLD (moderate); N-VA, Vlaams Belang (right-wing). Respondents had the option to indicate if they were not yet eligible to vote, did not vote or cast a blank vote, or if they did not want to disclose their choice. These respondents (n = 186) were categorized as missing. Results indicate that a large share of the sample voted for N-VA (23.3 %) and Vlaams Belang (16.9 %), two right-wing parties. CD&V (10 %) and Open VLD (9.2 %), two centrist parties, and left-wing party Groen (9.4 %) received similar shares of votes. PVDA (5.6 %) and sp.a (7 %), also two left-wing parties, received the smallest shares of votes. This distribution is in line with the outcome of the 2019 Belgian federal election in Flanders (Federale Overheidsdienst Binnenlandse Zaken, 2019). In subsequent analyses, we grouped the left-wing, centrist, and right-wing parties into three categories.
Socio-demographic indicators
Data on gender (1 = male, 2 = female), educational attainment (1 = uneducated, 2 = primary education, 3 = lower secondary education, 4 = higher secondary education, 5 = higher non-university education, 6 = university education), political ideology (1 = far-left, 11 = far-right), perceived income (1 = very difficult to make ends meet, 6 = very easy to make ends meet), and year of birth were also collected. For a descriptive overview of the sample, see Table 1.
Descriptive overview of the sample.
Min |
Max |
% |
N |
|
Gender |
||||
Male |
0 |
1 |
49.7 |
299 |
Female |
0 |
1 |
50.3 |
303 |
Voting behavior in 2019 |
||||
Left-wing (PVDA, Groen, sp.a) |
0 |
1 |
24.5 |
147 |
Moderate (CD&V, Open Vld) |
0 |
1 |
27.2 |
164 |
Right-wing (N-VA, Vlaams Belang) |
0 |
1 |
48.4 |
291 |
User engagement with parties |
||||
Follow no parties |
0 |
1 |
74.5 |
449 |
Follow left-wing party |
0 |
1 |
8.0 |
48 |
Follow moderate party |
0 |
1 |
5.0 |
30 |
Follow right-wing party |
0 |
1 |
12.5 |
75 |
Min |
Max |
Mean (SD) |
N |
|
Age |
18 |
86 |
44.12 (13.18) |
602 |
Educational attainment |
1 |
6 |
4.48 (1.11) |
602 |
Perceived income |
1 |
6 |
3.51 (1.07) |
602 |
Political ideology |
1 |
11 |
6.59 (2.27) |
602 |
Anti-immigration attitudes |
1 |
5 |
3.20 (1.09) |
602 |
3 Results
We were interested in understanding how anti-immigration attitudes related to following political parties on social media, how following parties was associated with voting behavior, and how the interaction of both elements was related to voting behavior among Flemish adults. Firstly, we determined the size of the subsample. As indicated, respondents who did not know what the migration-integration policies entailed (n = 271), did not vote in the recent elections (n = 186), or were not active on social media (n = 161) were removed from the analyses. Since some of these categories overlapped, we ended up with a subsample of 602 respondents. This subsample is similar to the total sample in terms of gender and educational attainment but is somewhat younger (mean age: 44 years). Respondents who followed no parties were combined with those who followed moderate parties. The reason for doing this was twofold. First, we were primarily interested in the effects of right versus left-wing parties, making a single ‘moderate’ category more appealing to compare with. Second, a one-way ANOVA showed that anti-immigration attitudes did not differ significantly between individuals who did not follow any parties on social media and those who followed moderate parties (p > 0.9), while significant differences were found with both right and left-wing party followers (see Table 2). Combining these various categories resulted in a new user engagement indicator, whereby 1 = following left-wing party, 2 = following moderate or no party, and 3 = following right-wing party. We ran Pearson correlation analyses between user engagement, migration attitudes, voting behavior, and political ideology (see Table 3), and results indicated weak to moderate correlations between most indicators.
Tukey post-hoc tests of one-way ANOVA of anti-immigration attitude scores by following political parties on social media.
Mean difference |
SE |
p-value |
||
No party (n = 449) |
Left-wing party Moderate party Right-wing party |
1.01 0.10 –0.55 |
0.12 0.21 0.11 |
.00 .96 .00 |
Left-wing party (n = 48) |
No party Moderate party Right-wing party |
–1.01 –0.91 –1.57 |
0.12 0.24 0.15 |
.00 .00 .00 |
Moderate party (n = 30) |
No party Left-wing party Right-wing party |
–0.10 0.91 –0.66 |
0.21 0.24 0.23 |
.96 .00 .02 |
Right-wing party (n = 75) |
No party Left-wing party Moderate party |
0.55 1.57 0.66 |
0.11 0.15 0.23 |
.00 .00 .02 |
Correlations between relevant variables.
1. |
2. |
3. |
4. |
5. |
6. |
|
1. Anti-immigration attitudes |
1 |
|||||
2. Follow left-wing party |
–.33** |
1 |
||||
3. Follow moderate party |
–.02 |
.26** |
1 |
|||
4. Follow right-wing party |
.20** |
.09** |
.30** |
1 |
||
5. Voting behavior |
.57** |
–.38** |
–.01 |
.28** |
1 |
|
6. Political ideology |
.55** |
–.35** |
.01 |
.31** |
.71** |
1 |
Note: * p < .05; ** p < .01.
Multinomial logistic regression with following parties on social media as outcome variable (n = 602).
Left-wing party |
Right-wing party |
|
Constant |
2.03* (.88) |
–6.70*** (.95) |
Age |
1.01 (.01) |
0.99 (.01) |
Gender (ref: female) |
||
Male |
1.92** (.23) |
0.79 (.29) |
Perceived income |
0.98 (.10) |
0.72** (.13) |
Educational attainment |
0.87 (.11) |
0.90 (.13) |
Political ideology |
0.73*** (.07) |
1.88*** (.09) |
Anti-immigration attitudes |
0.62** (.18) |
1.76** (.22) |
Interaction attitudes and political ideology |
0.91 (.07) |
1.13* (.06) |
Nagelkerke R² |
0.35 |
|
Note. + p < .10; * p < .05; ** p < .01; *** p < .001. Reference category of outcome variables: follow moderate party. Odds ratios are presented, standard errors between brackets.
In a first step, a multinomial logistic regression was performed to ascertain the association of age, gender, educational attainment, perceived income, political ideology, and anti-immigration attitudes with the likelihood that participants follow a party’s social media page (Table 4). We found that negative migration attitudes were related to a greater likelihood of following a right-wing party on social media (confirming Hypothesis 1a) and that positive migration attitudes were associated with a greater likelihood of following a left-wing party (confirming Hypothesis 1b) compared to a moderate party. Political ideology also appeared to be related to following parties on social media: It was significantly associated with following SNS pages of left-wing and right-wing parties (compared to following a moderate party). The interaction of political ideology with attitudes indicated that the effect of political ideology on following a right-wing party on SNS compared to following a moderate party was stronger among respondents with negative views on migration than among respondents with positive views on migration.
Socio-demographic characteristics played a limited role in explaining the likelihood of following a political party on social media. We found that women were less likely to follow left-wing parties compared to moderate parties than men. Furthermore, a more negative evaluation of the respondent’s income was associated with a decreased likelihood of following a right-wing party compared to a moderate party.
Multinomial logistic regression with Flemish voting behavior in the 2019 Belgian federal elections as outcome variable (n = 602).
Left-wing party vote |
Right-wing party vote |
|
Constant |
6.65*** (1.39) |
–3.37** (1.09) |
Age |
1.00 (.01) |
1.00 (.01) |
Gender (ref: female) |
||
Male |
0.88 (.31) |
0.88 (.26) |
Perceived income |
0.85 (.15) |
0.89 (.13) |
Educational attainment |
0.74+ (.16) |
0.74* (.13) |
Political ideology |
0.46*** (.11) |
1.69*** (.09) |
Anti-immigration attitudes |
0.78 (.21) |
1.84*** (.17) |
User engagement |
||
Follow left-wing party |
5.92** (.58) |
0.57 (.88) |
Follow right-wing party |
0.12** (.55) |
4.27** (.95) |
Interaction attitudes – user engagement |
||
Positive attitudes – Left-wing following |
1.64** (.43) |
0.70** (.64) |
Negative attitudes – Right-wing following |
0.60** (.49) |
1.43** (.50) |
Nagelkerke R² |
0.71 |
|
Note. + p < .10; * p < .05; ** p < .01; *** p < .001. Reference category of outcome variables: follow moderate party. Odds ratios are presented, standard errors between brackets.
In Table 5, we present the results of multinomial logistic regression to ascertain the association of migration attitudes, user engagement on SNS, and socio-demographic characteristics with the likelihood of voting for a party. Positive evaluations of the Belgian and Flemish migration policies – and thus, negative views on migration – were associated with an increased likelihood of voting for a right-wing party compared to voting for a moderate party, while no association with left-wing voting was found. As for the role of following political parties on SNS, results indicated that following a right-wing party was associated with a greater likelihood of voting for a right-wing party, confirming Hypothesis 2a, and decreased the likelihood of voting for a left-wing party compared to voting for a moderate party. Furthermore, following a left-wing party was associated with a greater likelihood of voting for a left-wing party, confirming Hypothesis 2b.
The interaction between user engagement and attitudes towards migration showed that respondents who held negative views of migration and followed a right-wing party on SNS were more likely to vote for a right-wing party than a moderate party, while respondents who held positive views on migration and followed a left-wing party were more likely to vote for a left-wing party than a moderate party, confirming Hypotheses 3a and 3b. The inclusion of socio-demographic characteristics did not yield many results aside from the finding that decreased educational attainment is related to a lower likelihood of voting for both a left-wing and right-wing party compared to a moderate party.
Robustness check
We ran additional analyses to show that findings remain valid when different dependent variables were used. In a first robustness check, we swapped the indicator for following a political party with following a specific politician of a certain party. This indicator was measured in the same way as following a political party was and coded using the same categories. We tested the relationship between attitudes towards migration and user engagement (now operationalized as following a politician) (see Table A1; mirroring the analysis in Table 4) and the relationship between these attitudes and user engagement with voting behavior (see Table A2; mirroring the analysis in Table 5). All substantial results were robust to these changes.
4 Discussion
Following the refugee crisis from 2014 to 2016, migration has become a key theme in public discourse and election campaigns throughout Europe. The salience of this issue was evident through its large-scale coverage on traditional media as well as social media (Georgiou and Zaborowski, 2017). Opinions on salient issues like migration are associated with political participation (e. g., voting behavior) (Keller and Kleinen-von Königslöw, 2018), which is why it is important to understand individuals’ user engagement with specific political parties on social networking sites in relation to their opinions about migration and how this may be related to voting behavior. Our findings suggest that: 1) users who held negative attitudes towards migration were more engaged with pages of right-wing parties than with those of other ideologies, while users who held positive attitudes towards migration were more engaged with pages of left-wing parties than with those of other ideologies. With these results, our study shows that user engagement with political parties on SNS (or clicktivism) is associated with their migration preferences. Through social media activity, individuals engage with politicians and political ideas. Although clicktivism in itself is noncommittal, given that interacting with a party on SNS is not necessarily associated with voting behavior, our results do show that it is strongly associated with voting behavior as well. In line with Halupka (2014), we argue that clicktivism should not simply be regarded as a noncommittal and ‘easy’ act of political participation, but that it can be perceived more as a political act that also relates to voting behavior of citizens and potentially to other types of political engagement.
Political actors may influence opinions about, and the importance of, specific topics (like migration) through social media. The strong association between attitudes towards migration and the following of political parties with similar attitudes suggests that social media are a place for agenda-setting by political actors. Social media pages can be used to focus on an issue, influence opinions and stimulate interest, especially when that message has a negative frame. Although this study has shown that social media users engage with the SNS pages of political parties with similar views on migration as their own, we do not know at which point and why they decide to follow a given party page. Most research focuses on the perspectives of politicians and their pages (e. g., Keller & Kleinen-von Königslöw, 2018; Metz et al., 2020) without looking into the decision-making process of citizens.
Furthermore, we also found that following SNS pages of right-wing parties is associated with a greater likelihood of voting for a right-wing party and that following SNS pages of left-wing parties is associated with a greater likelihood of voting for a left-wing party. There is also evidence of an interactive effect between user engagement and migration attitudes, as individuals who hold negative attitudes towards migration and who follow a right-wing party on SNS have a higher likelihood of voting for a right-wing party, and that individuals with positive attitudes towards migration and who follow a left-wing party on SNS have a higher likelihood of voting for a left-wing party. This shows that salient topics like migration not only relate to user engagement on social networking sites but also to actual political behavior in the ‘offline’ environment. Statements and information about actions and political ideas may engage users to follow a politician or political party and might spark interest in the given theme. These results show the interaction between the agenda-setting theory (Heidenreich et al., 2019) and the issue ownership theory (Walgrave et al., 2012), as exposure to migration messages is associated with the salience of this issue among individuals (Burscher et al., 2015). This explains the parallel between attitudes towards migration and engagement with parties in line with their pro- or anti-immigration ideology. We also found that the relation between issue-ownership and voting behavior is dependent on positional agreement between voters and issue owners: The association between SNS-engagement to right-wing parties and voting for a right-wing party is most pronounced among voters who already held negative attitudes towards migration in the first place – and the mirrored association is found for left-wing parties.
Although our study tackles several interesting issues and enters new territory, there are some limitations. First, in line with a growing criticism of issue ownership theory (Vliegenthart and Lefevere, 2018), it is possible that social media users engage with a party or politician after they have voted, and hence their voting behavior has also influenced their social media activity. After all, both this and many previous studies that draw on insights from issue ownership theory (Walgrave, Tresch, and Lefevere, 2015) use cross-sectional data which do not allow researchers to draw causal inferences. Although experimental or longitudinal studies are more suitable to providing conclusive evidence regarding the causal relation between these indicators, this study adds to a large body of cross-sectional literature that provides support for the association between elements of issue ownership and party preference. Second, the sample was representative of the Flemish adult population, but the results cannot be generalized to other populations. We therefore call for replication of this study in other countries and among other populations to confirm the mechanisms we found. Nevertheless, our study suggests that social media are influencing political decisions and opinions, a trend that will most likely continue to grow as the current generations grow up with social media as a normal part of life. The focus on specific topics on social media (issue-ownership) will play a bigger role in future elections, as will issues of agenda-setting of political actors. Third, we used the following of a political party (or a politician in the robustness checks) on SNS as a proxy for user engagement. Obviously, user engagement could also be illustrated in other ways: simply liking or following specific posts, following politically active groups that are not officially affiliated with political parties, reposting/sharing political posts, etc. However, for analytical purposes we had to develop a specific way in which user engagement could be operationalized, but we acknowledge that other interpretations may also be appropriate. Fourth, the measure of migration policy evaluation may be somewhat biased due to the specific wording of the items used to construct it (‘The migrant-integration policy plans of the Flemish government contribute to the integration of migrants into Flemish society’, and ‘Do you believe that the Belgian migration policies during the refugee crisis were a good way to deal with this crisis?’), as they may also signal approval of the current government.
As we move towards a society in which the presence of political actors on social media plays an increasingly important role (in some cases already eclipsing their presence on traditional media) (Ross and Bürger, 2014), especially among younger age cohorts, it will become even more important to study the interplay between opinions on societally salient topics, the engagement with voters, and the subsequent voting behavior. This study is a first step in understanding this issue and hopes to stimulate further research into the impact of social media on the political process. Following up on this study, it would be interesting to understand exactly how users engage with the SNS pages of political parties. When does one decide to follow certain pages? And which messages motivate voters to change or decide on their party vote? Would it be possible to predict voting shares based on follow numbers? And how will this impact on future political interactions between citizens and politicians? Clearly, social media will play an important role in future politics, and research into the subject will have to take the different aspects and platforms into consideration to understand this role. We must consider that following political parties that align with one’s own political preferences may eventually lead to an information bias. This is often called the “filter bubble”, an information bias that may contribute to a stronger polarization of society (Pariser, 2011). Although the existence of the “filter bubble” is highly contested, this study points to interesting ways in which an information bias comes into existence and has real-life consequences. Although it makes sense to follow political pages aligned with one’s political preferences, it might steer society back to a new age of pillarization. The UNHCR predicts that migration will continue to increase, and hence the topic will stay a key issue during the upcoming election periods in Europe. Considering the impact social media have according to our study, we recommend politicians to consider ways to tackle this subject on their social media pages and keep the importance of the social networking sites in mind. Looking at it from a user perspective and in support of critical thinking, media literacy is an essential skill in the digital age for putting the tone and content of SNS messages in the right context and checking the facts.
Acknowledgment
This research was funded by the Belgian Federal Science Policy Office, Grant Number: BR/165/A4/IM2MEDIATE.
References
Bieber, F. (2019). How Europe’s nationalists became internationalists. Retrieved May 25, 2021 from https://foreignpolicy.com/2019/11/30/how-europes-nationalists-became-internationalists/.Search in Google Scholar
Boomgaarden, H. G., & Vliegenthart, R. (2007). Explaining the rise of anti-immigrant parties: The role of news media content. Electoral Studies, 26(2), 404–417.10.1016/j.electstud.2006.10.018Search in Google Scholar
Boomgaarden, H. G., & Vliegenthart, R. (2009). How news content influences anti‐immigration attitudes: Germany, 1993–2005. European Journal of Political Research, 48(4), 516–542.10.1111/j.1475-6765.2009.01831.xSearch in Google Scholar
Bossetta, M. (2018). The digital architectures of social media: Comparing political campaigning on Facebook, Twitter, Instagram, and Snapchat in the 2016 US election. Journalism & Mass Communication Quarterly, 95(2), 471–496.10.1177/1077699018763307Search in Google Scholar
Budge, I., & Farlie, D. J. (1983). Explaining and predicting elections: Issue effects and party strategies in twenty-three democracies. Unwin Hyman.Search in Google Scholar
Burscher, B., van Spanje, J., & de Vreese, C. H. (2015). Owning the issues of crime and immigration: The relation between immigration and crime news and anti-immigrant voting in 11 countries. Electoral Studies, 38, 59–69.10.1016/j.electstud.2015.03.001Search in Google Scholar
Commissie Bossuyt (2019). Interim verslag van de Commissie voor de evaluatie van het beleid inzake de vrijwillige terugkeer en de gedwongen verwijdering van vreemdelingen [Interim report of the Commission for the evaluation of policies on the voluntary return and forced removal of foreigners]. Belgian Federal Government.Search in Google Scholar
Craig, S. C., & Cossette, P. S. (2020). Who owns what, and why? The origins of issue ownership beliefs. Politics & Policy, 48(1), 107–134.10.1111/polp.12338Search in Google Scholar
Damstra, A., Jacobs, L., Boukes, M., & Vliegenthart, R. (2021). The impact of immigration news on anti-immigrant party support: Unpacking agenda-setting and issue ownership effects over time. Journal of Elections, Public Opinion, and Parties, 31(1), 97–118.10.1080/17457289.2019.1607863Search in Google Scholar
Davis, A. (2010). New media and fat democracy: The paradox of online participation. New Media & Society, 12(5), 745–776.10.1177/1461444809341435Search in Google Scholar
De Coninck, D. (2020). Migrant categorizations and European public opinion: Diverging attitudes towards immigrants and refugees. Journal of Ethnic and Migration Studies, 46(9), 1667–1686.10.1080/1369183X.2019.1694406Search in Google Scholar
De Coninck, D., d’Haenens, L., & Joris, W. (2019). Investigating intergroup attitudes in Europe: Cross-national data on news media, attitudes towards newcomers, and socio-psychological indicators. Data in Brief, 26, 104535.10.1016/j.dib.2019.104535Search in Google Scholar
De Coninck, D., Duque, M., Schwartz, S. J., & d’Haenens, L. (2021). Public attitudes towards immigration, news and social media exposure, and political attitudes from a cross-cultural perspective: Data from seven European countries, the United States, and Colombia. Data in Brief, 39, 107548.10.1016/j.dib.2021.107548Search in Google Scholar
Eberl, J.-M., Meltzer, C. E., Heidenreich, T., Herrero, B., Theorin, N., …, & Strömbäck, J. (2018). The European media discourse on immigration and its effects: A literature review. Annals of the International Communication Association, 42(3), 207–223.10.1080/23808985.2018.1497452Search in Google Scholar
Eberl, J.-M., Tolochko, P., Jost, P., Heidenreich, T., & 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), 48–65.10.1080/19331681.2019.1710318Search in Google Scholar
Esser, F., & Pfetsch, B. (2020). Political communication. In D. Caramani (Ed.), Comparative politics (pp. 336–358). Oxford University Press.10.1093/hepl/9780198820604.003.0019Search in Google Scholar
Federale Overheidsdienst Binnenlandse Zaken (2019). Verkiezingen [Elections]. Retrieved May 25, 2021 from https://verkiezingen2019.belgium.be/nl/verkiezingen?el=CK.Search in Google Scholar
Georgiou, M., & Zaborowski, R. (2017) Media coverage of the “refugee crisis”: A cross-European perspective. Council of Europe.Search in Google Scholar
Gerbaudo, P., Marogna, F., & Alzetta, C. (2019). When “positive posting” attracts voters: User engagement and emotions in the 2017 UK election campaign on Facebook. Social Media + Society, 5(4).10.1177/2056305119881695Search in Google Scholar
Gibson, R. K., & McAllister, I. (2014). Normalising or equalising party competition? Assessing the impact of the web on election campaigning. Political Studies, 63(6), 529–547.10.1111/1467-9248.12107Search in Google Scholar
Gil De Zúñiga, H., Puig-I-Abril, E., & Rojas, H. (2009). Weblogs, traditional sources online and political participation: An assessment of how the internet is changing the political environment. New Media & Society, 11(4), 553–574.10.1177/1461444809102960Search in Google Scholar
Halupka, M. (2014). Clicktivism: A systematic heuristic. Policy & Internet, 6(2), 115–132.10.1002/1944-2866.POI355Search in Google Scholar
Heath, A., & Richards, L. (2019). How do Europeans differ in their attitudes to immigration? Findings from the European Social Survey 2002/03 – 2016/17. Report for OECD Social, Employment and Migration Working Papers. Report no. 222 OECD Publishing.Search in Google Scholar
Heidenreich, T., Eberl, J.-M., Galyga, S., Lind, F., Boomgaarden, H. G., Jiménez, B. H., Montero, E. L. G., & Berganza, R. (2019). Political migration discourses on social media across countries and over time. Report for REMINDER Working Papers. University of Vienna.Search in Google Scholar
Heidenreich, T., Eberl, J.-M., Lind, F., & Boomgaarden, H. G. (2020). Political migration discourses on social media: A comparative perspective on visibility and sentiment across political Facebook accounts in Europe. Journal of Ethnic and Migration Studies, 46(7), 1261–1280.10.1080/1369183X.2019.1665990Search in Google Scholar
Heiss, R., & Matthes, J. (2020). Stuck in a nativist spiral: Content, selection, and effects of right-wing populists’ communication on Facebook. Political Communication, 37(3), 303–328.10.1080/10584609.2019.1661890Search in Google Scholar
Heiss, R., Schmuck, D., & Matthes, J. (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), 1497–1513.10.1080/1369118X.2018.1445273Search in Google Scholar
Keller, T. R., & Kleinen-von Königslöw, K. (2018). Followers, spread the message! Predicting the success of Swiss politicians on Facebook and Twitter. Social Media + Society, 4(1).10.1177/2056305118765733Search in Google Scholar
Kim, S., & Baek, T. H. (2018). Examining the antecedents and consequences of mobile app engagement. Telematics and Informatics, 35(1), 148–158.10.1016/j.tele.2017.10.008Search in Google Scholar
Kruikemeier, S., van Noort, G., Vliegenthart, R., & de Vreese, C. H. (2014). Unraveling the effects of active and passive forms of political internet use: Does it affect citizens’ political involvement? New Media & Society, 16(6), 903–920.10.1177/1461444813495163Search in Google Scholar
Larsson, A. O., & Moe, H. (2012). Studying political microblogging: Twitter users in the 2010 Swedish election campaign. New Media & Society, 14(5), 729–747.10.1177/1461444811422894Search in Google Scholar
Lee, J. S., & Nerghes, A. (2018). Refugee or migrant crisis? Labels, perceived agency, and sentiment polarity in online discussions. Social Media + Society, 4(3).10.1177/2056305118785638Search in Google Scholar
Louden, A., & McCauliff, K. (2004). The “authentic candidate”: Extending candidate image assessment. In K. Hacker (Ed.), Presidential candidate images (pp. 85–103). Rowman & Littlefield Publishers.Search in Google Scholar
McCombs, M., & Shaw, D. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36, 176–187.10.1086/267990Search in Google Scholar
McConahay, J. B. (1986). Modern racism, ambivalence, and the Modern Racism Scale. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 91–125). Academic Press.Search in Google Scholar
McGregor, S. C. (2017). Personalization, social media, and voting: Effects of candidate self-personalization on vote intention. New Media & Society, 20(3), 1139–1160.10.1177/1461444816686103Search in Google Scholar
Metz, M., Kruikemeier, S., & Lecheler, S. (2020). Personalization of politics on Facebook: Examining the content and effects of professional, emotional and private self-personalization. Information, Communication & Society, 23(10), 1481–1498.10.1080/1369118X.2019.1581244Search in Google Scholar
Nosek, B. A., Hawkins, C. B., & Frazier, R. S. (2012). Implicit social cognition. In S. T. Fiske & C. N. Macrae (Eds.), The SAGE handbook of social cognition (pp. 31–51). SAGE Publications.10.4135/9781446247631.n3Search in Google Scholar
Pariser, E. (2011). The filter bubble. What the internet is hiding from you. Penguin Books.10.3139/9783446431164Search in Google Scholar
Pennington, N., & Winfrey, K. L. (2021). Engaging in political talk on Facebook: Investigating the role of interpersonal goals and cognitive engagement. Communication Studies, 72(1), 100–114.10.1080/10510974.2020.1819844Search in Google Scholar
Petrocik, J. R. (1996). Issue ownership in presidential elections, with a 1980 case study. American Journal of Political Science, 40(3), 825–850.10.2307/2111797Search in Google Scholar
Puschmann, P., Sundin, E., De Coninck, D., & d’Haenens, L. (2019). Migration and integration policy in Europe: Comparing Belgium and Sweden. In L. d’Haenens, W. Joris, F. Heinderyckx (Eds.), Images of immigrants and refugees in Western Europe. Media representations, public opinion, and refugees’ experiences (pp. 21–36). Leuven University Press.Search in Google Scholar
Ross, K., & Bürger, T. (2014). Face to face(book). Political Science, 66(1), 46–62.10.1177/0032318714534106Search in Google Scholar
Schemer, C. (2012). The influence of news media on stereotypic attitudes toward immigrants in a political campaign. Journal of Communication, 62(5), 739–757.10.1111/j.1460-2466.2012.01672.xSearch in Google Scholar
Soroka, S., Young, L., & Balmas, M. (2015). Bad news or mad news? Sentiment scoring of negativity, fear, and anger in news content. The Annals of the American Academy of Political and Social Science, 659(1), 108–121.10.1177/0002716215569217Search in Google Scholar
Starke, C., Marcinkowski, F. & Wintterlin, F. (2020). Social networking sites, personalization, and trust in government: Empirical evidence for a mediation model. Social Media + Society, 1–11.10.1177/2056305120913885Search in Google Scholar
Statistiek Vlaanderen (2020). Internationale bescherming [International protection]. Retrieved June 2, 2021 from https://www.statistiekvlaanderen.be/nl/internationale-bescherming#:~:text=Belgi%C3%AB%20was%20goed%20voor%203,lag%20dat%20aantal%20op%2011.Search in Google Scholar
Štětka, V., & Mazák, J. (2014). Whither slacktivism? Political engagement and social media use in the 2013 Czech parliamentary elections. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 8(3).10.5817/CP2014-3-7Search in Google Scholar
UNHCR (2017). Global trends. Forced displacement in 2017. Retrieved June 1, 2021 from https://www.unhcr.org/5b27be547.pdfSearch in Google Scholar
Van Gorp, B., Figoureux, M., & Vyncke, B. (2018). Twitterende politici over migratie: Welke effecten heeft de gehanteerde framing? [Twittering politicians on migration: What effects does the framing used have?] Institute for Media Studies, KU Leuven.Search in Google Scholar
van Klingeren, M., Boomgaarden, H. G., & de Vreese, C. H. (2017). Will conflict tear us apart? The effects of conflict and valenced media messages on polarizing attitudes toward EU immigration and border control. Public Opinion Quarterly, 82(2), 546–563.10.1093/poq/nfw051Search in Google Scholar
Van Leuven, S., Deprez, A., Joye, S., & Ongenaert, D. (2019). How do Flemish politicians talk about migration? A study on the political framing of migration in Belgium 2016–2018. Centre for Journalism Studies, UGent.Search in Google Scholar
Vliegenthart, R., & Lefevere, J. (2018). Disentangling the direction of causality between competence issue ownership and party preference. International Journal of Public Opinion Research, 30(4), 663–674.10.1093/ijpor/edx017Search in Google Scholar
Walgrave, S., & De Swert, K. (2004). The making of the (issues of the) Vlaams Blok. Political Communication, 24(4), 479–500.10.1080/10584600490522743Search in Google Scholar
Walgrave, S., Lefevere, T., & Tresch, A. (2012). The associative dimension of issue ownership. Public Opinion Quarterly, 76(4), 771–782.10.1093/poq/nfs023Search in Google Scholar
Walgrave, S., Lefevere, T., & Tresch, A. (2020). Position, competence, and commitment: Three dimensions of issue voting. International Journal of Public Opinion Research, 32(1), 165–175.10.1093/ijpor/edz006Search in Google Scholar
Walgrave, S., Tresch, A., & Lefevere, J. (2015). The conceptualisation and measurement of issue ownership. West European Politics, 38, 778–796.10.1080/01402382.2015.1039381Search in Google Scholar
Xenos, M. A., Macafee, T., & Pole, A. (2017). Understanding variations in user response to social media campaigns: A study of Facebook posts in the 2010 US elections. New Media & Society, 19(6), 826–842.10.1177/1461444815616617Search in Google Scholar
Xenos, M. A., & Moy, P. (2007). Direct and differential effects of the internet on political and civic engagement. Journal of Communication, 57, 704–718.10.1111/j.1460-2466.2007.00364.xSearch in Google Scholar
Appendix
Multinomial logistic regression with following politicians on social media as outcome variable and odds ratios of socio-demographic characteristics and interactions of evaluations of migration policy and political ideology.
Left-wing politician |
Right-wing politician |
|
Constant |
9.22** (1.42) |
–4.71*** (.75) |
Age |
0.97 (.01) |
0.99 (.01) |
Gender (ref: female) |
||
Male |
1.03* (.20) |
0.28 (.29) |
Perceived income |
1.27 (.15) |
1.00 (.10) |
Educational attainment |
0.37* (.11) |
0.57 (.08) |
Political ideology |
0.67*** (.18) |
2.77*** (.09) |
Anti-immigration attitudes |
0.91** (.22) |
3.01** (.17) |
Interaction migration attitudes and political ideology |
0.86 (.08) |
1.01* (.09) |
Nagelkerke R² |
0.45 |
|
Note. + p < .10; * p < .05; ** p < .01; *** p < .001. Reference category of outcome variables: follow moderate politician.
Multinomial logistic regression with Flemish voting behavior in the 2019 Belgian federal elections as outcome variable and odds ratios of socio-demographic characteristics, user engagement, migration-integration policy evaluations, and interaction effects.
Left-wing party vote |
Right-wing party vote |
|
Constant |
6.76*** (1.24) |
–3.18** (1.05) |
Age |
1.00 (.01) |
0.99 (.01) |
Gender (ref: female) |
||
Male |
1.00 (.31) |
0.91 (.27) |
Perceived income |
0.87 (.14) |
0.86 (.12) |
Educational attainment |
0.77+ (.16) |
0.73** (.14) |
Political ideology |
0.46*** (.11) |
1.74*** (.09) |
Anti-immigration attitudes |
0.70+ (.18) |
1.79** (.15) |
User engagement |
||
Follow left-wing politician |
1.17** (.55) |
0.02 (.85) |
Follow right-wing politician |
0.85** (.57) |
6.43* (.90) |
Interaction attitudes – user engagement |
||
Positive attitudes – Left-wing following |
1.52** (.41) |
0.27* (.62) |
Negative attitudes – Right-wing following |
0.69** (.44) |
2.02** (.51) |
Nagelkerke R² |
0.68 |
|
Note. + p < .10; * p < .05; ** p < .01; *** p < .001. Reference category of voting behavior = moderate party vote.
© 2022 Walter de Gruyter GmbH, Berlin/Boston
